article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
in the currently - favoured hierarchical model of structure formation , clusters of galaxies have formed from the collapse of gravitational potential wells ( e.g. @xcite ) and are powerful cosmological probes .
for instance , since they are sensitive to the expansion history of the universe , their abundance and spatial distribution ( e.g. @xcite ) and their mass function ( e.g. @xcite ) depend on cosmological parameters , such as the dark energy equation of state parameter @xmath6 ( e.g. @xcite ) , or the power spectrum normalisation @xmath5 ( e.g. @xcite ) .
several observational methods now permit the use of clusters of galaxies as cosmological probes , such as x - ray observations , weak gravitational lensing and the sunyaev - zeldovich effect .
due to improvements in telescopes and techniques , x - ray studies are able to constrain cluster physics and the mass scaling relation with ever greater precision . for instance , the self - similarity for clusters of galaxies ( @xcite ) has been observationally verified .
nevertheless , adiabatic simulations still predicts a mass - temperature relation with double the observed normalisation ( e.g. @xcite ) and the self - similarity assumption could break down at low temperatures ( @xmath7 ) .
thus , a steepening of the mass - temperature ( m - t ) relation is expected if galaxy groups underwent a preheating by supernovae , or a surge of entropy , in their early days ( @xcite ) .
recent evidence for this steepening was found by e.g. @xcite , or arnaud , pointecouteau & pratt 2005 ( app05 hereafter ) , but could not be seen by e.g. @xcite .
moreover , the m - t normalisation estimation is currently limited by systematics in measuring cluster masses from their x - ray profiles .
this limitation can be lifted by using probes which are independent of the physical state of the cluster . beyond galaxy cluster physics ,
the m - t relation is needed by x - ray experiments to estimate the power spectrum normalisation @xmath5 . measuring
this parameter has triggered much effort in several observational areas .
for instance , cmb experiments ( e.g. @xcite ) tend to a low value for @xmath5 ( @xmath8 ) , weak lensing experiments tend to higher values ( @xmath9 ) , and x - ray observations provide intermediate values .
gravitational lensing does not depend on the underlying physics of clusters of galaxies or dark matter , but only on their potential wells , and on the universe s geometry .
strong gravitational lensing has been used for galaxy clusters physics ( e.g. @xcite ) and measurement of @xmath5 ( e.g. @xcite ) .
weak gravitational lensing is more difficult to measure ( for reviews , see e.g. @xcite ) ) , and has taken longer to be detected ( @xcite ) .
since then , particular attention has been given to cosmic shear , i.e. statistical cosmological weak lensing ( e.g. @xcite ) , in attempts to measure @xmath6 and @xmath5 .
it has also begun to be used as a tracer of the cosmic web ( e.g. @xcite ) , and a way to detect and catalogue mass overdensities ( e.g. @xcite , gavazzi & soucail 2007 , gs07 hereafter , and @xcite ) . beside the constraints it can bring to cosmology , it can be used as a complement to x - ray analyses of clusters of galaxies . thanks to the physics - independent estimation of cluster masses
, it appears as a unique method to calibrate the mass - temperature relation for clusters of galaxies ( e.g. @xcite ) .
it has been shown that the uncertainty in the normalisation of the mass - temperature relation is the largest source of error in @xmath5 measurements inferred from x - ray cluster analyses ( @xcite ) .
an accurate mass - temperature relation , obtained from combined weak lensing and x - ray analyses , will thus provide new insights not only on the @xmath5 discrepancy , but also on galaxy cluster physics . in this paper , we present the first joint analysis of weak gravitational lensing and x - ray wide - area surveys of a randomly - selected patch of sky .
the weak lensing survey is derived from the cfhtls , and the x - ray survey from the xmm - lss .
the weak lensing analysis uses shapelets ( @xcite ) , a new generation shear measurement technique , which has been shown to achieve a few percent accuracy in shear measurement from ground based telescopes ( @xcite ) .
we analyse one square - degree of the cfhtls deep survey ( the d1 field ) and four contiguous square degrees of the cfhtls wide survey , which enclose the d1 field .
we create convergence maps for this region of the sky and give a catalogue of detections .
we show how counting weak - lensing - selected clusters can provide an estimate of the power spectrum normalisation @xmath5 .
we then show how the combination of weak lensing and x - ray analyses of clusters provides an estimate of the mass - temperature relation normalisation @xmath10 , independent of clusters physical state .
finally , we investigate the impact of a joint weak lensing and x - ray survey strategy on the accuracy of the @xmath5 and @xmath10 measurement .
we consider deep and wide weak lensing surveys , with the cfhtls characteristics , combined with a blind x - ray survey of the same region of the sky .
the organisation of the paper is as follows .
section [ data ] presents the surveys used in this paper , namely the cfhtls and the xmm - lss .
the methods that we use are described in section [ method ] .
we show how we estimate the weak lensing effect using shapelets , and how we generate convergence maps and detect clusters .
we also briefly describe how the x - ray properties of clusters are obtained .
section [ results ] presents the convergence map we inferred from our weak lensing analysis , and gives a catalogue of the galaxy clusters that we detect .
we then give our estimates of the normalisation of the power spectrum and of the m - t relation .
we then show in section [ discussion ] that combined blind surveys are necessary to get the best insights about those normalisations .
the impact of survey strategy on those parameters estimations is discussed in section [ impact ] .
we conclude in section [ conclusion ] .
the canada - france - hawaii telescope legacy survey " ( cfhtls ) , a joint france - canada project , consists of three different surveys , namely the very wide survey , the wide synoptic survey ( referred to as wide survey " hereafter ) , and the deep survey .
once complete , the wide survey will cover 170 deg@xmath0 ( divided into four distinct patches ranging from 49 deg@xmath0 to 72 deg@xmath0 ) in five filters ( u*,g,r,i,z ) , down to a magnitude @xmath11 .
its main goal is the study of large scale structures by weak gravitational lensing and galaxy counts .
the deep survey covers 4 different fields , each with an area of 1 deg@xmath0 , in the same five filters , down to @xmath12 .
it is primarily intended for type ia supernovae studies but it is also useful for measuring large - scale structures .
the cfhtls images were obtained from observations with the megacam camera , made of a 36 ccd mosaic , of 2048 @xmath13 4196 pixels each , with a 1 deg@xmath0 field of view ( @xcite ) . in this paper , we present the weak gravitational lensing based mapping of 4 deg@xmath0 of the wide survey ( w1 patch ) , which include the 1 deg@xmath0 field of the deep survey ( d1 field ) , using both w1 and d1 images . the geometry of the fields that we use is shown in figure [ fig_data ] .
the data processing ( astrometry , photometric calibration , stacking of images ) has been done by the cfht community and terapix .
we use w1 images optimised for weak lensing : each field is the combination of 7 stacked images , each of 620 seconds exposure time .
we use the t0003 release of the d1 field , consisting of 275 stacked images , with a total 37.4 hour exposure time .
the average seeing is 0.7 arcsec .
we masked parts of the images with saturated stars and/or too high a noise , by hand , so as not to bias our weak lensing results .
this operation removes 10% of the original area covered by the data .
we do not mask the ghosts created by spurious reflections on the telescope optics around saturated stars , but we eventually remove the galaxies that they cover from our catalogues , since they are too noisy .
the average galaxy density is 28 arcmin@xmath14 in the d1 image , and 13 arcmin@xmath14 in the w1 images .
designed to cover an area of several tens of square degrees up to a redshift @xmath15 , the xmm - lss survey aims at detecting a significant fraction of clusters of galaxies , in order to constitute a sample suited to cosmological studies ( @xcite ) .
its nominal exposure times are 10 ks , and have been raised up to 20 ks for the _ xmm _ medium deep survey ( xmds : @xcite ) , a 2 deg@xmath0 region included in the xmm - lss , which covers the cfhtls d1 field . in this paper
, we use 4 deg@xmath0 of the xmm - lss which cover our 4 deg@xmath0 cfhtls data .
the _ xmm _
pointings are shown on figure [ fig_data ] .
the raw x - ray observations reduction is presented in @xcite .
introduced in @xcite , @xcite and @xcite , shapelets are a complete , orthogonal , set of basis functions with which one can analytically decompose galaxy shapes .
they can be seen as fundamental shapes : a particular galaxy can be represented by a particular sum of shapelets basis functions @xmath16 , weighted by coefficients @xmath17 .
their rich formalism provides an intuitive and analytic form for geometrical transformations ( such as smear , shear , rotation ) and for ( de)convolution .
hence , they allow one to analytically describe the smearing of the point spread function ( psf ) and the shear of galaxies , properly correcting for the psf .
the shear estimation they provide has been shown to reach the needed accuracy for the cfhtls specifications by the step project ( @xcite ) .
our full pipeline will be described in an upcoming paper ( berg et al . in prep ) .
we briefly summarise it here .
each sufficiently bright and non - saturated star is first decomposed into shapelets .
a polynomial interpolation of each shapelet coefficient then provides a model of its spatial variations across the image .
we are thus able to reconstruct the shape of the psf at the position of each galaxy , the condition necessary for deconvolving it from the galaxies shapes .
several stringent tests then validate our psf model .
in particular , we require that the ellipticity distribution , and the two point correlation functions of the ellipticity , of the residuals between observed stars and their shapelet models are consistent with zero .
we also require that the cross - correlation between stars and galaxies ellipticity , when corrected from the psf , is consistent with zero .
the shape of galaxies is finally measured by decomposing them into shapelets , while deconvolving from the psf , as shown in @xcite .
a shear estimator is created from the shapelet decompositions of galaxies , as prescribed by @xcite : @xmath18 where the shear susceptibility @xmath19 is fitted on the magnitude - size plane for galaxies .
the coefficients are complex numbers .
the shear @xmath20 of equation ( [ gamma ] ) is the complex notation for shears , @xmath21 .
to increase the signal to noise ratio of our measurements , we give to each galaxy @xmath22 a minimum variance weight @xmath23 , where @xmath24 is the error on shape measurement for galaxy @xmath22 , @xmath25 the error on the measurement of its shear susceptibility , and @xmath26 is the intrinsic ellipticity dispersion , set to @xmath27 . slightly changing @xmath26
would be equivalent to giving more or less weight to our measurement errors , and would mostly affect the error bars in the shear measurement .
the eventual peak detection would not be affected by such slight changes .
this weighting scheme is equivalent to smoothly selecting the most useful galaxies for shear measurement .
for instance , the faintest are down - weighted .
it therefore provides us with effective densities of @xmath28 and @xmath29 useful galaxies per square arcminute , in the d1 and w1 images , respectively .
then , a direct inversion in fourier space of the pixelised shear map allows us to infer a convergence ( i.e. projected mass ) map ( @xcite ) of the images
. structures in this mass map are extracted from the noise using a gaussian filter .
figures [ massmapw1 ] and [ massmap ] show the convergence maps that we inferred from our data .
these are described in section [ catalogue ] .
while constructing a convergence map , we also create a signal - to - noise map , the signal - to - noise ratio being defined as @xmath30 where @xmath31 is the convergence at the ( @xmath32 ) sky coordinates , and @xmath33 its r.m.s error .
the r.m.s error @xmath33 is computed using monte - carlo simulations in which the input galaxies are positioned like the observed ones but with randomised shape orientations .
structures are then searched for in the filtered convergence map , and their astrometry provided , by the sextractor software ( @xcite ) .
they are extracted according to their signal - to - noise peak , read from the signal - to - noise map .
hereafter , we define a ` significant structure ' as a detection with a signal - to - noise ratio greater than 2.5 . their mass is related to their integrated convergence through the lensing geometry , and can be estimated when their redshift and the redshift distribution of background galaxies are known . to account for the latter , we use the normalised distribution @xmath34\ ] ] where the parameters @xmath35 are given for the wide images by @xcite @xmath36 .
to account for @xmath37 in the d1 image , we fit @xcite s photometric redshift distribution in the cfhtls d1 field , and obtain @xmath38 . @xcite
have shown that errors in the @xmath37 fit are subdominant compared to poisson noise and sample variance for the measurement of cosmological parameters .
we thus neglect them hereafter .
we measure a cluster s virial mass by averaging its convergence in an aperture large enough that we can assume that the entire cluster is captured .
the aperture corresponds to the region enclosed in the 2@xmath39 level of the cluster s convergence map .
this technique is similar to using a @xmath40-statistic ( @xcite ) , with infinitely large annulus around the cluster , provided that the convergence in the entire field averages to 0 .
we verified this latter point , thus validating our choice .
note that because of the small number density of background sources , we can not reliably fit a shear profile around clusters ( see e.g. @xcite for an example of mass estimation using two profile fits around the galaxy cluster abell 209 ) .
we then convert the virial mass into @xmath41 , the mass enclosed in the sphere of mean overdensity 200 times higher than the critical density , using the recipe from @xcite .
hereafter , we will note @xmath41 more briefly @xmath42 .
weak lensing is affected by the entire mass distribution along the line of sight . as a result
, the weak lensing mass measurement of one cluster is biased by projection effects .
it has been shown , using different mass estimators , that large - scale structures in the line of sight , and near the target cluster , introduce errors ranging from a few percent ( @xcite ) to a few tens of percent ( @xcite ) . in this paper
, we assume that they produce a 20% error , added in quadrature to the shear measurement error .
the x - ray cluster detection pipeline has been described in @xcite .
it takes account of the poisson nature of the x - ray images , to extract and analyse clusters of galaxies in a two - step procedure .
clusters are first detected by a multi - resolution wavelet filter ( @xcite ) .
then , each source is analysed using a maximum likelihood profile fitting procedure , and its x - ray properties assessed .
three classes of extended sources have been defined ( @xcite ) : ( 1 ) the c1 class contains the highest surface brightness sources , and is uncontaminated ; ( 2 ) the c2 class allows for 50 % contamination , and contains less bright extended sources ; ( 3 ) finally , the c3 class contains optically confirmed sources with apparent x - ray emission , that were not selected as c1 or c2 . in this paper
, we only consider c1 class detections , representative of the most massive objects seen in the xmm - lss .
the redshift of detected clusters has been measured using spectroscopic observations from a number of telescope and instrument combinations detailed in table 2 of @xcite .
their temperature estimation is described in @xcite .
@xcite have extracted and analysed 29 c1 clusters from 5 deg@xmath0 of the xmm - lss data ( shown on fig .
[ fig_data ] ) , that contain our 4 deg@xmath0 optical data . among other things ,
they have measured their luminosity and temperature . here , we take into account their 16 clusters which are enclosed in the fields of our cfhtls data , making use only of their temperature and redshift . they are listed in table [ cataloguetable ] .
note that the cluster xlssc053 is in the g12 xmm - lss pointing ( shown by the cross in the d1 field , on fig .
[ fig_data ] ) , which was not used when @xcite analysed the xmm - lss observations .
this pointing has been re - observed , and the x - ray characteristics of xlssc053 are listed in table [ cataloguetable ] .
in this section , we give the properties of clusters of galaxies detected with our weak lensing pipeline .
counting the weak lensing detections allows us to constrain the matter power spectrum normalisation @xmath5 .
we then use the weak lensing mass of the detected groups to calibrate the mass - temperature relation for clusters of galaxies .
@xmath43 & significance & @xmath42(wl ) & @xmath44 & @xmath42(x)@xmath45 & notes + cluster i d & number & i d & ( deg ) & ( deg ) & & d1/w1 & ( @xmath46 ) & ( kev ) & ( @xmath46 ) + wl01 & 013 & cl03 & 36.8497 & -4.5481 & 0.31 & 3.61 / - & @xmath47 & @xmath48 & 2.1 + wl02 & - & - & 36.6589 & -4.7516 & - & 3.09 / - & - & - & - + wl03 & - & cl04 & 36.3628 & -4.1886 & 0.32 @xmath49 & 2.91 / - & @xmath50 & - & - + wl04 & 053 & cl02 & 36.1229 & -4.8341 & 0.50 @xmath51 & 2.90 / - & @xmath52 & @xmath53 & 5.0 & xmm - lss pointing + & & & & & & & & & & not observed in + & & & & & & & & & & @xcite + wl05 & 041 & cl14 & 36.3723 & -4.2604 & 0.14 & 2.62 / - & @xmath54 & @xmath55 & 3.5 + wl06 & 044 & - & 36.1389 & -4.2384 & 0.26 & 2.48 / - & @xmath56 & @xmath57 & 3.7 & just below detection threshold + & & & & & & & & & & in gs07 s catalogue + wl07 @xmath58 & 022 & cl07 & 36.9167 & -4.8606 & 0.29 & 2.42 / - & - & @xmath59 & 5.3 & near a mask + - & 025 & cl05 & 36.3375 & -4.6925 & 0.26 & - / - & - & @xmath60 & 6.5 & under a mask + - & - & cl10 & 36.8167 & -4.1269 & - & - / - & - & - + - & 029 & - & 36.0172 & -4.2260 & 1.05 & - / - & - & @xmath61 & 13.9 & too high redshift + - & 011 & - & 36.5410 & -4.9680 & 0.05 & - / - & - & @xmath62 & 1.0 & + - & 005 & - & 36.7866 & -4.2995 & 1.05 & - / - & - & @xmath63 & 16.5 & too high redshift + - & 006@xmath64 & - & 35.4382 & -3.7717 & 0.43 & x / - & - & @xmath65 & 30.4 & near an edge + - & 040@xmath64 & - & 35.5232 & -4.5463 & 0.32 & x / - & - & @xmath66 & 6.8 + - & 049@xmath64 & - & 35.9892 & -4.5880 & 0.49 & x / - & - & @xmath67 & 5.0 + - & 018@xmath64 & - & 36.0079 & -5.0903 & 0.32 & x / - & - & @xmath68 & 8.0 + - & 021@xmath64 & - & 36.2338 & -5.1340 & 0.08 & x / - & - & @xmath69 & 1.8 + - & 001@xmath64 & - & 36.2378 & -3.8156 & 0.61 & x / - & - & @xmath70 & 14.3 + - & 008@xmath64 & - & 36.3367 & -3.8014 & 0.30 & x / - & - & @xmath71 & 2.1 + - & 002@xmath64 & - & 36.3841 & -3.9198 & 0.77 & x / - & - & @xmath72 & 9.6 + + + + + + [ cataloguetable ] figure [ massmapw1 ] shows the 4 deg@xmath0 of the w1 field that we considered .
no significant overdensity ( i.e. with @xmath73 ) has been detected . as we will quantitatively show in section [ selfct ] , this is consistent with the expected cluster counts for this survey .
the black square in the image shows the position of the d1 field .
since there are around 20 galaxies per arcmin@xmath0 , we expect more significant detections in this deep field ( see section [ selfct ] ) .
the upper panel of fig .
[ massmap ] shows the convergence map that we obtained from our weak lensing analysis of the d1 field . due to the varying level of noise in our map , which varies independently of @xmath74 , two peaks with the same @xmath74 value
do not necessarily have the same significance .
that results in the rejection of seemingly significant structures , such as the peak around ( @xmath75)=(@xmath76 ) .
significant structures are marked out by the white contours , which start at 2.2@xmath39 , with an increment of 0.5@xmath39 .
the lower panels show individual candidate clusters in more details . in these ,
contours start at 2.2@xmath39 , with an increment of 0.2@xmath39 .
even though we consider as significant a structure with @xmath77 , we plot the 2.2@xmath39 contours as a way to show the extension of our detections . dashed circles mark x - ray clusters , and dotted circles @xcite s weak lensing detections . @xcite
measured the shear in the d1 field using the ksb method ( @xcite ) .
table [ cataloguetable ] summarises the measured characteristics of the clusters that we detect , together with all x - ray and @xcite s detections in the region .
the clusters that we detect through our shapelets weak lensing analysis are labelled with wl_id _ , where _ i d _ runs from 00 to 07 , and are sorted by decreasing significance .
their labels are listed in the first column . their official _ xmm _
names are given in column ( 2 ) , and @xcite s ids in the third column .
x - ray clusters marked by a @xmath64 are outside the d1 field .
columns ( 4 ) and ( 5 ) give their position .
column ( 6 ) lists their spectroscopic redshifts , except for clusters wl03 , for which a tomographic redshift is given , and wl04 for which a photometric redshift is given ( aussel et al . in prep ) .
the significances of the weak lensing detections are listed in column ( 7 ) , in d1 and w1 .
a ` - ' means that the cluster is not detected ; a ` x ' means that the cluster is outside the d1 field .
columns ( 8) and ( 9 ) give their weak lensing mass @xmath42(wl ) and x - ray temperature , respectively . column
( 10 ) gives the mass estimate from the x - ray data , @xmath78 .
as in @xcite , these were evaluated under the assumption of an isothermal @xmath79-model gas distribution in hydrostatic equilibrium with the cluster s potential well
. in this earlier work , the associated statistical errors were generally dominated by the temperature measurement uncertainty ( @xmath80 of order 10 - 25% ) . here , the error on the emission profile can also become quite significant because we estimate the masses within @xmath81 instead of @xmath82 where the x - ray emission starts to vanish .
moreover , it was shown by @xcite and @xcite that the isothermal @xmath79-model assumption leads to an underestimation of the total mass , by up to 40% for low mass systems .
for these reasons , we decided not to provide error bars for our x - ray masses
. finally , column ( 11 ) gives some details about weak lensing detections , explaining for instance why we chose not to take them into account , or why we do not detect a cluster seen by another method . among the rejection criteria are the proximity to an edge or to a masked region , the mass inversion procedure being sensitive to missing data and to edge effects . in that sense , and for clarity , the detections closest to edges have been removed from fig .
[ massmapw1 ] and fig .
[ massmap ] .
clusters wl01 , wl04 and wl05 have unequivocal counterparts both in our x - ray catalogue and in @xcite s ksb one .
no significant c1 x - ray source has been selected around wl02 , and it remains invisible to @xcite .
moreover , a visual inspection of the optical images does not show any galaxy overdensity around it .
no significant c1 x - ray source has been detected at the position of wl03 , even though it is also seen by @xcite .
cluster wl06 lies just below our detection threshold ( @xmath83 ) . nevertheless , since @xcite detect it ( though just below the detection threshold they use for their analyses )
and since it is also detected by our x - ray analysis , we decided to list it , and to assess its weak lensing characteristics .
the significance of cluster wl07 is even lower ( @xmath84 ) .
since it is found close to the xlssc022 cluster ( which coincides with @xcite s cl07 ) , we show its contours in fig .
[ massmap ] and list it in table [ cataloguetable ] .
however , it lies near an edge and a mask , so that its weak lensing characteristics are likely to be biased .
we thus do not measure its mass , and will not take it into account in what follows . cluster xlssc025 ( @xcite s cl05 ) is under a mask , and can not be detected by our pipeline . in summary , out of our 7 shapelet weak lensing detections , 4 ( wl01 , wl04 , wl05 and wl07 ) have a counterpart both in our x - ray catalogue and in the ksb weak lensing catalogue by @xcite , even though we remove wl07 from our subsequent analyses .
one detection ( wl02 ) appears only in our catalogue .
one ( wl06 ) has an x - ray counterpart , and appears in @xcite s map , but just below the detection threshold they use for their analysis .
finally , one ( wl03 ) has a counterpart in @xcite s catalogue , but is not selected as a c1 x - ray cluster .
this proves a good agreement between the three cluster detection methods used in those observations .
x - ray clusters xlssc005 and xlssc029 are at too high a redshift to be detected with our surveys .
cluster xlssc011 is too close and not massive enough to be detected by weak lensing , as will be shown in section [ selfct ] .
figure [ fig_nugs ] compares the significance of our detections with that given by @xcite .
we make use of wl06 , even though @xcite did not use it , but gave its significance . given our weak lensing measurement characteristics ( @xmath85 , @xmath27 ) and theirs ( @xmath86 , @xmath87 ) , eq .
( [ sf ] ) below allows us to compute the expected proportionality factor between our detections significance @xmath88 and theirs , @xmath89 .
we expect @xmath90 .
this relation is shown by the dashed line on fig .
[ fig_nugs ] .
the significance of clusters in both catalogues scale as expected .
one should note that this relation depends on the measurement characteristics for both methods in those particular experiments , and should not be used as a final comparison between ksb and shapelets .
more comparison on real data will be needed in order to explore this issue .
while x - ray masses listed in column ( 10 ) of table [ cataloguetable ] must be taken with caution , they can be compared to the weak lensing masses listed in column ( 8) . although one can notice an order of magnitude agreement between @xmath42(wl ) and @xmath42(x ) , masses estimated from x - ray data seem slightly underestimated .
this is consistent with the previously mentioned bias arising from the isothermal beta - model parametrisation .
while they do not provide strong statistics , our detections can be used to estimate @xmath5 , as shown below .
four detections have an x - ray counterpart and can thus be used to constrain the mass - temperature relation , provided that we add clusters from another catalogue .
this is described below . .
, width=302 ] a weak gravitational lensing selection function can be computed analytically ( see e.g. @xcite ) from the signal - to - noise ratio of a halo in a given cosmology and weak lensing survey parameters .
we derive such a selection function , using an optimal match filter , in berg , amara & rfrgier ( in prep ) . in an observation characterised by a number density of background galaxies @xmath91 , an nfw halo of convergence
@xmath74 has signal - to - noise ratio : @xmath92 where @xmath93 is the r.m.s shear error per galaxy , and where we neglect projection effects and sample variance , which have been shown to have subdominant effects ( @xcite ) .
our selection function is shown in figure [ fig_mlim ] in the mass - redshift plane , for our deep ( @xmath94 arcmin@xmath14 , @xmath95 , thick black ) and our wide ( @xmath96 arcmin@xmath14 , @xmath97 , red ) surveys , in a cosmological model based on the three - year _ wilkinson microwave anisotropy probe _ results ( wmap3 ; @xcite ) , ( @xmath98)=(0.73,0.127,0.0223,0.76,-1 ) .
we use the redshift distributions for background galaxies given by equation ( [ eq_nz ] ) .
figure [ fig_mlim ] shows , from bottom to top , the minimum detectable mass for a halo at a @xmath99 , a @xmath100 and a @xmath101 detection threshold .
the deep and wide selection functions have different slopes , as a consequence of their different redshift distribution for background galaxies .
symbols represent the position , in the redshift - mass plane , of clusters listed in table [ cataloguetable ] .
we use the weak lensing mass @xmath42(wl ) for wl01 , wl03 , wl04 , wl05 and wl06 ( thick square symbols ) and the x - ray mass @xmath42(x ) for other clusters ( diamonds ) .
although we detect cluster xlssc022 ( wl07 ) , we do not assess its gravitational mass , and thus show its x - ray mass in fig .
[ fig_mlim ] .
cluster xlssc025 should be detectable ( and is detected by @xcite ) , but it is under a mask in our analysis .
triangle symbols ( labeled @xmath64 in table [ cataloguetable ] ) correspond to c1 x - ray clusters in w1 that are outside the d1 region .
it is clear from figure [ fig_mlim ] that they can not be detected by our weak lensing analysis of w1 .
only xlssc006 should be seen , at the 2@xmath39 level .
however , its detection is plagued by its proximity to the edge of the image .
also shown on fig .
[ fig_mlim ] are the x - ray selection functions , for 50% and 80% detection probabilities ( dashed lines ) as estimated by @xcite .
figure [ fig_mlim ] shows an excellent agreement between the clusters characteristics and their predicted detectability by weak gravitational lensing .
, @xmath91=20 arcmin@xmath102 , redshift distribution as eq .
( [ eq_nz ] ) ) and w1 ( red ; @xmath103 , @xmath91=9 arcmin@xmath102 , redshift distribution as eq .
( [ eq_nz ] ) ) , assuming a wmap3 cosmology in each case . from bottom to top , lines correspond to 2@xmath39 , 3@xmath39 and 4@xmath39 significance .
dashed lines show the x - ray selection function , corresponding to 50% , and 80% detection probability ( @xcite fig .
18 , lower and upper curves , respectively ) .
thick square symbols are our detections in the d1 data , labeled by their i d ; they are not detectable in the w1 data .
diamonds are clusters detected either by @xcite or by x - ray analysis , in d1 , that we do not detect for reasons listed in the text .
red triangles are c1 x - ray clusters lying outside the d1 region . except for the thick square symbols ( for which we use the weak lensing mass @xmath42(wl ) ) , we use the x - ray mass @xmath42(x ) . except xlssc006 (
labeled for visibility as x006 ) , they can not be detected by a weak lensing experiment in the w1 data .
xlssc006 is not detected because of its proximity to an edge.,width=302 ] ( @xmath104 ) .
the dot - dashed line is the expected number counts for the wide survey , with @xmath104.,width=302 ] from equation ( [ sf ] ) , the expected number of haloes detected above a certain significance can be computed using a press - schechter approach ( @xcite ) . for this purpose ,
we use the @xcite mass function to estimate the number of haloes that we can detect , as a function of significance threshold . curves on figure [ fig_nvsnu ] show such counts for different @xmath5 and survey depths .
@xcite already used this statistic to discriminate between halo profile models .
it is used here to measure @xmath5 .
most of our detections are validated by corresponding objects either in the catalogue of @xcite or our x - ray c1 cluster catalogue . nevertheless , despite its relative high significance
, wl02 does not have such independent support .
we indeed consider it as a false detection , and do not take it into account for cluster counts .
we then estimate the number of false detections from monte carlo simulations . for this purpose ,
we input galaxies at the position of the actual ones , but randomise their shear , and look for detections with significance higher than 2.5@xmath39 .
the convergence maps that we infer from them show only false detections .
we find that , in this particular experiment , we expect only one false detection above the 2.5@xmath39 level .
this is thus consistent with removing wl02 . for this counting purpose ,
we remove wl06 and wl07 from our catalogue , since they do not reach the 2.5@xmath39 level . the symbols on fig .
[ fig_nvsnu ] represent our cumulative counts , corrected from false detections .
their error bars include the effects of shot noise and sample variance , computed from hu & kravtsov s ( 2003 ) analytic formula .
we then fit the expected number counts to our data as a function of @xmath5 , keeping all other parameters constant . in order to avoid covariance between our data points in the cumulative counts depicted by fig .
[ fig_nvsnu ] , we performed the fit on the expected differential number counts @xmath105 .
we find @xmath106 ( at the 68.3% confidence limit ) , for @xmath2 . despite large error bars
, we can set interesting constraints thanks to the strong dependence of these counts on @xmath5 , as shown by the difference between the solid and dashed curves in figure [ fig_nvsnu ] .
the dashed line shows the expected cumulative number counts for the deep survey in a wmap3-like universe ( @xmath107 ) .
the solid line is our best fit ( @xmath104 ) .
the dot - dashed line shows the expected number density on the wide survey , with @xmath104 .
, for our group sample ( diamonds ) . to improve the statistics , we added clusters from @xcite ( triangles ) .
we use x - ray temperature and weak lensing mass .
the solid line is our best fit ( eq . [ eq_mt_obs ] ) .
the dashed and dotted line are @xcite m - t relation , when they consider all clusters or only those with @xmath108 , respectively.,width=302 ] under the virial equilibrium assumption , the mass and temperature of a cluster are related by the following scaling relation ( @xcite ) @xmath109^{-1/2 } \left[1 - 2 \frac{\omega_\lambda(z)}{\delta_c(z ) } \right]^{-3/2}\ ] ] where @xmath110 is the virial mass , @xmath111 is the virial temperature , @xmath10 is a normalisation factor , and @xmath112 .
@xmath113 is the overdensity inside the virial radius , in units of the critical density .
we compute it using the fitting formula by @xcite for @xmath114 , which is very similar to an earlier approximation by @xcite given by @xcite for a universe with arbitrary @xmath115 .
a more general relation often used to fit observations makes use of a related normalisation factor @xmath3 and is given , at redshift @xmath116 , by @xmath117 where @xmath42 is the mass inside the sphere of mean overdensity 200 times higher than the critical density and @xmath118 in the hydrostatic equilibrium assumption ( e.g. @xcite ) .
hereafter , to account for redshift evolution , we normalise all our temperatures to @xmath116 by dividing them by @xmath119 . measuring @xmath5 from x - ray counts is affected by the degeneracy @xmath120 ( @xcite ) .
@xcite have shown that the uncertainty in @xmath3 is the main concern in measurements of @xmath5 from x - ray cluster observations alone .
such data is limited by the requirement that the cluster masses be inferred from the x - ray profiles .
@xcite have also shown that unrelaxed clusters , being hotter than relaxed clusters , provide a supplementary bias to the @xmath5 estimate .
it is thus important to have a mass estimate independent of the hydrostatic equilibrium assumption .
weak gravitational lensing gives such an estimate .
combined with x - ray temperature , it can be efficiently used to constraint the m - t relation , independently of the cluster physical state . @xcite
have already used it to measure the m - t relation normalisation . as described above
, we have the weak lensing mass and x - ray temperature of only four groups .
hence , to increase our statistics , we add @xcite s clusters to our catalogue , providing us with 11 additional clusters .
@xcite estimated cluster masses by fitting an nfw model to their tangential shears .
we should note here that since @xcite s mass estimation and ours are based on different techniques , our subsequent analysis of the mass - temperature relation could be slightly biased due to possible calibration differences .
figure [ fig_mt ] shows the relation between the temperature and the weak lensing mass @xmath42 for the combined catalogues .
diamonds are our groups , labelled wl_id _ , triangles are @xcite s clusters , labelled a_id_. @xcite proceeded to the weak lensing analyses of massive haloes , the temperature of which were obtained by @xcite and @xcite ; particularly , they estimated their weak lensing mass and measured the scale relations for those clusters .
they fitted their sample by varying both @xmath121 and @xmath3 , and found a large slope , far from the hydrostatic equilibrium assumption , @xmath122 .
doing the same analysis on the larger range in mass that the addition of both catalogues probes , from galaxy groups to galaxy clusters , we find : @xmath123 which is in good agreement with @xcite ( whether they use all clusters or only the most massive ones ) , @xcite ( when they fix @xmath124 ) , @xcite or @xcite .
the solid line on fig .
[ fig_mt ] is our best fit .
the dashed line is the best fit from @xcite , when they consider all clusters in their catalogue .
the dotted line is their best fit when they consider their most massive clusters ( @xmath125 ) .
our result is consistent with self - similarity evolution for galaxy clusters down to low temperatures .
it is also consistent with previous measurements which observe a steepening of the m - t relation at the low mass end , due to the expected self - similarity breaking for such masses ( e.g. nevalainen et al .
2000 ; finoguenov et al .
2001 ; @xcite ) .
moreover , one must be aware that the galaxy groups we consider were detected just above our weak lensing selection function ( fig .
[ fig_mlim ] ) . due to the expected scatter in the m - t relation
, those groups can represent only the most massive ones with temperature ranging about 1 kev .
our group sample could thus bias our fit towards a flat slope for the m - t relation .
the analysis of more low temperature groups will be needed to further explore this issue .
- @xmath126 plane .
the shaded region shows the constraints given by our @xmath5 and @xmath10 measurements .
thick lines are our best fits .
the slanted black region correspond to the @xmath127 constraints on the @xmath120 relation from @xcite .
the vertical , light blue , shaded region shows the @xmath128 error on @xmath10 from @xcite .
cmb derived constraints of @xmath5 ( @xcite ) are shown by the horizontal dark blue shaded region .
cosmic shear @xmath5 estimations from @xcite and @xcite are marked by the red and green horizontal shaded regions .
@xcite s constraints are typical of cosmic shear results
. they are higher than x - ray estimations , marked by the intersection between @xcite s and @xcite s allowed domains.,width=302 ]
the power spectrum normalisation @xmath5 has been measured with different probes , such as x - ray clusters of galaxies , cmb , and cosmic shear ( i.e. statistics of weak gravitational lensing ) .
some discrepancies have emerged between the preferred value from those measurements .
recent cmb observations favour a low @xmath5 and cosmic shear used to emphasise a high value ( see e.g. @xcite for a review ) .
x - ray clusters provide intermediate measurements ( see e.g. @xcite for a review ) . the dominant discrepancy between cosmic shear and x - ray clusters has recently been reduced by @xcite , who measured @xmath129 for @xmath130 when using cosmic shear alone , followed by @xcite , who used @xcite s improved galaxies photometric redshifts , and measured @xmath131 for @xmath2 , and @xcite who found a consistent value . to clarify these discrepancies
, one needs to measure both the power spectrum and the m - t relation normalisations @xmath5 and @xmath3 , as we discuss here .
the regions allowed for by different measurements on the @xmath10 - @xmath126 plane are shown by figure [ fig_constraints ] .
the shaded region shows the constraints given by our @xmath5 and @xmath10 measurements .
our best fits are shown by the thick lines .
the slanted black band on the figure is the 68.3% bound on the @xmath120 relation from @xcite using x - ray clusters .
its intersection with the vertical light blue band ( @xcite s @xmath10 estimation ) gives the current value for @xmath5 favoured by x - ray cluster observations , @xmath132 for an @xmath133 universe ( e.g. @xcite ) , which corresponds to @xmath134 .
this value is higher than that measured by @xcite from cmb analyses of wmap3 ( dark blue ) , but lower than most cosmic shear analysis , like that of @xcite made with cfhtls wide data ( red ) .
this highlights the discrepancy between x - ray and weak lensing estimates of @xmath5 mentioned above .
however , @xcite give a lower estimate for @xmath5 , consistent with x - ray measurements ( green ) .
this could be a sign that other cosmic shear analyses did not take some systematics into account , and have thus overestimated @xmath5 . according to @xcite , previously published analyses made use of insufficiently - accurate galaxy photometric redshifts . using @xcite s redshifts yielded a lower value of @xmath5 both for cosmic shear ( @xcite ) and for our cluster count analysis .
we found a 5% decrease in our @xmath5 estimation when going from previous redshift distributions to @xcite s ones .
this is less than the change reported by @xcite , and our best fit still tends to favour a higher value for @xmath5 , but is limited by low statistics .
@xcite have analysed the bias from unrelaxed clusters in @xmath5 measurement using lensing clusters and the m - t relation .
they found that unrelaxed clusters are 30% hotter than relaxed clusters : using unrelaxed clusters can provide 20% overestimates of @xmath5 .
this is enough to explain the large range of measured @xmath5 , from @xmath135 0.6 to @xmath135 1 .
they estimated @xmath136 .
estimates from x - ray alone can also be affected by systematics , such as the mass estimate from x - ray profiles of clusters .
for example , a slight decrease of @xmath10 would cause an increase of the x - ray estimate for @xmath5 . a better insight into this will come from an accurate measurement of @xmath10 , preferably with mass estimation methods independent of cluster physics .
large combined weak lensing and x - ray surveys will be needed to disentangle the situation .
they will provide both independent constraints on @xmath5 , and insights on @xmath10 .
in the following , we investigate the impact of future combined blind weak lensing and x - ray surveys on the measurement precision of the power spectrum and the mass - temperature relation normalisations .
we take the wmap3 ( @xcite ) cosmology as our fiducial model .
we consider two different ground based survey strategies for our weak gravitational lensing analysis : deep and wide surveys similar to the cfhtls deep and wide surveys .
we use their observed weighted number density of useful background galaxies to be @xmath94 arcmin@xmath14 and @xmath29 arcmin@xmath14 , respectively , distributed according to eq .
( [ eq_nz ] ) .
we also assume the intrinsic ellipticity and shape measurement error to be @xmath95 in both cases . following the cfhtls scheduling ,
we take for exposure times 40 hours per square degree for the deep survey and 1 hour per square degree for the wide survey .
, from clusters counts in a weak gravitational lensing survey , as a function of survey size and integration time .
all other parameters are kept constants .
we assume that 1 deg@xmath0 of wide requires 1 hour of observation time , and 1 deg@xmath0 of deep requires 40 hours of observation time .
that is , the lower x - axis shows area as well as the wide survey exposure time ; the upper x - axis shows the deep survey exposure time .
the thick solid line corresponds to a deep survey , and the thick dashed line to a wide survey .
the flat lines show the current error measurement from cosmic shear statistics ( dash - dot , hoekstra et al .
2006 ; benjamin et al .
2007 ) , and from x - ray clusters ( dash - dot - dot , pierpaoli et al . 2003 ; app05).,width=302 ] we first investigate the impact of future surveys on the @xmath5 measurement .
using the press - schechter approach described in section [ sect_counts ] , we estimate the number of weak lensing detections with significance higher than 2.5 , taking into account shot noise and sample variance .
we assume that all clusters have a spherically symmetrical nfw profile .
we thus neglect the effect of haloes asphericity shown by clowe et al .
( 2004 ) : triaxial haloes oriented along the line of sight appear more massive than triaxial haloes of the same mass , but perpendicular to the line of sight , and thus have a higher signal - to - noise ratio .
clowe et al . (
2004 ) have shown that this approximation does not yield any difference in the mass measurement dispersion .
figure [ fig_prosp1 ] shows the 68.3% relative error on @xmath5 that can be reached by counting weak lensing detected clusters as a function of their significance , for a deep ( thick solid line ) and a wide ( thick dashed line ) surveys , as a function of survey s size and observing time .
because of the higher number density of clusters it allows one to detect , a deep survey provides errors 3 times lower than a wide survey of the same size .
however , for a given exposure time , a wide survey provides errors 2.1 times lower than a deep one .
that means that the gain due to the coverage ( and detectable clusters number ) increase is faster than the one due to depth increase .
a larger coverage is also advantageous in that it makes sample variance fall down rapidly .
moreover , a wide survey detects the most massive haloes , the physics of which is better understood .
consequently , in a survey strategy driven by exposure time , one should prefer a wide survey .
the flat dashed - dot line shows the current constraints provided by cosmic shear analyses ( @xcite ) .
the flat dashed - dot - dot line shows the current constraints from the combination of x - ray @xmath3 measurement ( @xcite ) and x - ray cluster counts ( @xcite ) .
detecting and counting clusters on a 10 deg@xmath0 deep survey will be competitive with current cosmic shear measurements , whereas 20 deg@xmath0 of coverage is needed to compete with current x - ray clusters measurements .
those figures transform as 100 deg@xmath0 and 200 deg@xmath0 for a wide survey .
that is , to compete with current cosmic shear surveys , one needs 400 hours of deep survey exposure , or 100 hours of wide survey exposure .
double these times are required to compete with x - ray surveys . a wide survey , less demanding in exposure time than a deep one ,
should then be used .
counting clusters on the entire planned cfhtls wide survey 170 deg@xmath0 will provide a 6% fractional error on the @xmath5 measurement .
reaching the 1% fractional error will require a 7000 deg@xmath0 wide survey , or a 700 deg@xmath0 deep survey .
future surveys ( e.g. pan - starrs , lsst , dune ) will be able to achieve such errors . from combined weak gravitational lensing and x - ray surveys , as a function of survey size and integration time .
all other parameters are kept constants . the lower x -
axis shows area as well as the wide survey exposure time ; the upper x - axis shows the deep survey exposure time .
solid lines show the errors for a deep survey , assuming the fractional error on weak lensing mass measurement is 20% ( thick black ) and 30% ( red ) .
dashed lines , with the same colour indexing , show the errors for a wide survey .
the current error measurement from x - ray clusters is shown by the flat dashed - dot - dot line ( pierpaoli et al .
2003 ; app05 ) .
we made the same assumption about the relation between survey area and observation time as in fig .
[ fig_prosp1].,width=302 ] we now turn to the precision that can be reached on @xmath10 measurements by future joint surveys .
what matters here is not cluster counts as a function of significance , but as a function of mass .
as we see from figure [ fig_mlim ] , low mass clusters of galaxies can not be seen through weak gravitational lensing since they do not create high enough signal - to - noise ratios .
a deep survey captures lower mass clusters than a wide survey , but deep and wide surveys give access to the same number of massive clusters .
therefore , a deep survey is naturally focused on the physics of galaxy groups ( e.g. it can probe similarity breaking at the low mass end of the mass - temperature relation ) .
a wide survey gives the same statistics on massive haloes , generally used to measure the mass - temperature relation normalisation : for this purpose , one should then choose a wide survey . to compare the merits of both deep and wide survey on the @xmath10 estimation s precision , we simulate m - t relations for both types of survey .
we take a realistic scatter into account , @xmath137 for the logarithmic m - t relation ( @xcite ) .
we assume that masses are measured through weak gravitational lensing .
we measure @xmath10 and the error on its estimate , by assuming ( 1 ) that our cluster sample is complete , ( 2 ) that we only make use of those clusters detected in our blind survey , and ( 3 ) that we know the x - ray temperature of each of them .
we also investigate the influence of the mass estimation fractional error @xmath138 .
figure [ fig_prosp2 ] shows the 68.3% error on @xmath10 that can be reached from a combined blind x - ray and deep ( solid line ) or wide ( dashed line ) weak gravitational lensing surveys , as a function of survey s size and integration time . here again , as for the error on the power spectrum normalisation , a deep survey gives errors 2.3 times lower than a wide one with the same sky area coverage . on the other hand
, a wide survey gives errors 2.7 times lower than a deep one with the same exposure time .
the dependence on area underlines the @xmath10 estimation s reliance on the number of useable haloes for the m - t relation fitting .
figure [ fig_prosp2 ] also shows the sensitivity of the @xmath10 estimation to the mass measurement errors .
the black lines assume @xmath139 , and the red ones @xmath140 , which are the current fractional errors on mass measurement from weak lensing .
going from @xmath140 to @xmath141 allows one to reduce the error on @xmath10 by a factor of 1.3 ( resp .
1.2 ) for a deep ( resp .
wide ) survey of a given sky area .
the flat solid line represents the current error on @xmath10 from x - ray clusters ( @xcite ) . assuming a 20% error measurement on weak lensing masses
, one needs a 50 ( resp .
300 ) deg@xmath0 weak lensing deep ( resp .
wide ) survey to reach the current error . reaching the 1% fractional error ( for our fiducial model with @xmath142 ) will require a 2500 deg@xmath0 wide survey , or a 500 deg@xmath0 deep survey .
weak lensing surveys like lsst or dune combined with x - ray surveys like erosita will be able to reach such a limit . in a survey strategy driven by exposure time , a wide survey of 2500 deg@xmath0 ( 2500 hours ) will be able to reach the 1% accuracy both on @xmath5 and @xmath10 , at a much cheaper expense than a deep survey .
nevertheless , a deep survey will still be useful to probe high redshift regions ( @xmath143 ) , and to study low mass clusters of galaxies ( @xmath144 ) .
we have presented the first shapelet analysis of weak gravitational lensing surveys .
we have constructed convergence maps of the cfhtls deep d1 field , and of 4 deg@xmath0 of the cfthls wide w1 field , which include the d1 field .
we have detected six clusters of galaxies , through the lensing signal they generate .
our d1 map is in good agreement with that of @xcite , precedently created using the ksb shear measurement method .
we combined our weak lensing data with the x - ray analysis of xmm - lss c1 clusters lying in the same region of the sky ( @xcite ) .
these three clusters catalogue are consistent .
all our shapelet detections have either an x - ray counterpart or a ksb detection .
counting our detections and accounting for the weak lensing selection function allowed us to constrain the power spectrum normalisation @xmath145 .
the combination of lensing masses and x - ray temperatures provided us with a new measurement of the mass - temperature relation normalisation @xmath10 ( or equivalently @xmath3 ) for clusters of galaxies , @xmath4 .
our results , though limited by low statistics and sample variance , are consistent with other current estimates .
we also measured the slope of the mass - temperature relation , and found it consistent with self - similarity for low mass clusters , @xmath146 .
we have shown that one must measure both @xmath5 and @xmath10 from combined weak lensing and x - ray surveys to investigate the discrepancy between independent measurements of @xmath5 from different probes .
weak lensing surveys are becoming more and more effective , and are currently being optimised for best extracting cosmological information .
optimal surveys will allow us to provide more accurate estimates of @xmath5 and @xmath10 , and to disentangle the current @xmath5 issue ( @xcite ) .
we have compared the merits of weak lensing deep and wide blind surveys , based on the cfhtls , at estimating @xmath5 .
we also looked at their merits at estimating @xmath10 while combined with an x - ray survey on their region of the sky .
we found that for experiments driven by exposure time constraints , a wide survey will give @xmath135 3 times lower errors on the estimates of both @xmath5 and @xmath10 . to secure the measurement of @xmath5 and @xmath3 with the current statistical accuracy , a 200 deg@xmath0 and a
300 deg@xmath0 wide surveys will be needed respectively .
we finally found that a 7000 deg@xmath0 wide survey will be able to reach the 1% accuracy both on the power spectrum and mass - temperature relation normalisations .
the authors wish to thank herv aussel , krys libbrecht , jean - baptiste melin , yannick mellier , sandrine pires , trevor ponman , jean - luc starck , genevive soucail and romain teyssier for useful discussions , and cathy horellou for comments on the first version of the paper .
99 amara a. , refregier a. , 2007 , mnras , 381 , 1018 arnaud m. , aghanim n. , neumann d.m . , 2002 , a&a , 389 , 1 arnaud m. , pointecouteau e. , pratt g.w . , 2005 , a&a , 441 , 893 ( app05 ) bacon d. , refregier a. , ellis r. , 2000 , mnras , 318 , 625 bacon d. , massey r. , refregier a. , ellis r. , 2003 , mnras , 344,673 bardeau s. , soucail g. , kneib j.p . , czoske o. , ebeling h. , hudelot p. , smail i. , smith g.p . , , 2007 , a&a , 470 , 449 bartelmann m. , schneider p. , 2001
, phys . rep .
, 340 , 291 basilakos s. , 2003 , apj , 590 , 636 basilakos s. , voglis n. , 2007 , mnras , 374 , 269 benjamin j. et al . , 2007 , mnras , 381 , 702 bertin e. , arnouts s. , 1996 , a&as , 117 , 393 bialek j.j .
, evrard a.e .
, mohr j.j . , 2001 ,
apj , 555 , 597 boulade o. et al . , 2003 ,
spie , 4841 , 72 castillo - morales a. , schindler s. , 2003 , a&a , 403 , 433 chiappetti l. et al . , 2005 , a&a , 439,413 clowe d. , de lucia g. , king l. , mnras , 350 , 1038 de putter r. , white m. , 2005 , new astronomy , 10 , 676 eke c.r . , navarro j.f .
, frenck c.s . , 1998 , apj , 503 , 569 ettori s , de grandi s. , molendi s. , 2002 , a&a , 391 , 841 fahlman g. , kaiser n. , squires g. , woods d. , 1994 , apj , 437 , 56 finoguenov a. , reiprich t.h . , bhringer h. , 2001 , a&a , 368 , 749 fu l. et al , 2008 , a&a , 479 , 9 gavazzi r. , soucail g. , 2007 , a&a , 462 , 459 ( gs07 ) hamana t. , takada m. , yoshida n , 2004 , mnras , 350 , 893 heymans c. et al . , 2005 , mnras , 361 , 160 hjorth j. , oukbir j. , van kampen e. , 1998 , mnras , 298 , 1l hoekstra h. , 2001 , a&a , 370 , 743 hoekstra h. , 2003 , mnras , 339 , 1155 hoekstra h. et al . , 2006 , apj , 647 , 116 hoekstra h. , 2007 , mnras , 379 , 317 horellou c. , berge j. , 2005 , mnras , 360 , 1393 hu w. , kravtsov a.v . , 2003 , apj , 584 , 702 huterer d. , white m. , 2002 , apj , 578 , 95l ilbert o. et al , 2006 , a&a , 457 , 841 jarvis m. , jain b. , bernstein g. , dolney d. , 2006 , apj , 644 , 71 jenkins a. , frenck c.s . ,
white s.d.m , colberg j.m . ,
cole s. , evrard a.e .
, couchman h.m.p . ,
yoshida n. , 2001 , mnras , 321 , 372 kaiser n. , squires g. , 1993 , apj , 404 , 441 kaiser n. , squires g. , broadhurst t. , 1995 , apj , 449 , 460 kitayama t. , suto y. , 1996 , apj , 469 , 480 kneib j .-
p . , ellis r.s . , smail i. , couch w.j . , sharples r.m
. , 1996 , apj , 471 , 643 kneib j .-
, mellier y. , pello r. , miralda - escude j. , le borgne j .- f . , boehringer h. , picat j .-
, 1995 , a&a , 303 , 27 lacey c. , cole s. , 1993 , mnras , 262 , 627 lokas e.l . , 2001 , acta phys .
b , 32 , 3643 lokas e.w . ,
bode p. , hoffman y. , 2004 , mnras , 349 , 595 luppino g.a .
, gioia i.m .
, hammer f. , le fvre o. , annis j.a . , 1999 ,
a&a , 136 , 117 marian l. , bernstein g. , 2006 , phreviewd , 73 , 123525 maor i. , lahav o. , 2005 , jcap , 07 , 003 manera m. , mota d.f . , 2006 , mnras , 371 , 1373 massey r. et al . , 2007c , mnras , 376 , 13 massey r. , refregier a. , 2005 , mnras , 363 , 197 massey r. , refregier a. , bacon d. , ellis r. , brown , m.l .
, 2005 , mnras , 359 , 1277 massey r. et al . , 2007b , nature , 445 , 286 massey r. et al .
, 2007a , apjs , 172 , 239 massey r. , rowe b. , rfrgier a. , bacon d. , berg j. , 2007d , mnras , 380 , 229 mellier y. , 1999 , araa , 37 , 127 mellier y. , fort b. , kneib j .-
, 1993 , apj , 407 , 33 metzler c.a . , white m. , loken c. , 2001 , apj , 547 , 560 miyazaki s. et al . , 2002 , apj , 580 , 97l miyazaki s. , hamana t. , ellis r.s . , kashikawa n. , massey r.j .
, taylor j. , refregier a. , 2007 , apj , 669 , 714 muanwong o. , thomas p.a . , kay s.t . ,
pearce f.r . , 2002 , mnras , 336 , 527 munshi d. , valageas p. , van waerbeke l. , heavens a. , 2006 , astro - ph/0612667 nakamura t.t . , 1996 , master s thesis , univ .
tokyo nevalainen j. , markevitch m. , forman w. , 2000 , apj , 532 , 694 nunes n.j .
, da silva a.c . , aghanim n. , 2006 , a&a , 450 , 899 ota n. , mitsuda k. , 2004 , a&a , 428 , 757 pacaud f. et al . , 2006 , mnras , 372 , 578 pacaud f. et al . , 2007 , mnras , 382 , 1289 padmanabhan t. , 1993 , structure formation in the universe , cambridge univ .
press , cambridge paulin - henriksson s. , antonuccio - delogu v. , haines c.p . , radovich m. , mercurio a. , becciani u. , 2007 , a&a , 467 , 427 pedersen k. , dahle h. , 2007 , apj , 667 , 26 peebles , p.j.e . , 1980 , the large - scale structure of the universe , princeton univ . press , princeton , nj pierpaoli e. , borgani s. , scott d. , white m. , 2003 , mnras , 342 , 163 pierre m. et al .
, 2004 , jcap , 9 , 11 pierre m. et al .
, 2006 , mnras , 372 , 591 press , w.h . , schechter p. , 1974 ,
apj , 187 , 425 rasia e. et al . , 2006 , mnras , 369 , 2013 reblinsky k. , bartelmann m. , 1999 , a&a , 345 , 1 refregier a. , 2003a , araa , 41 , 645 refregier a , 2003b , mnras , 338 , 35 refregier a. , bacon d. , 2003 , mnras , 338 , 48 rhodes j. , refregier a. , groth e.j . , 2001 , apj , 552 , l85 schrabback t. et al . , 2007 , a&a , 468 , 823 seljak u. , 2002 , mnras , 337 , 769 semboloni e. et al . , 2006 , a&a , 452 , 51 smail i. , ellis r.s . , dressler a. , couch w.j . , oemler a.j .
, sharples r.m . , butcher h. , 1997 , apj , 479 , 70 smith g.p . , edge a.c . ,
, nichol r.c . , smail i. , kneib j .-
, 2003 , apj , 570 , l79
smith g.p . ,
, smail i. , mazzotta p. , ebeling h. , czoske o. , 2005 , mnras , 359 , 417 spergel d.n .
et al . , 2007 , apjs , 170 , 377 starck j.l .
, murtagh f. , bijaoui a. , 1998 , image processing and data analysis : the multiscale approach .
cambridge univ . press ,
cambridge van waerbeke l. et al . , 2000 ,
a&a , 358 , 30 van waerbeke l. , white m. , hoekstra h. , heymans c. , 2006 , astropart .
phys . , 26 , 91 viana p.t.p , liddle a.r . , 1996 , mnras , 281 , 323 vikhlinin a. , kravtsov a. , forman w. , jones c. , markevitch m. , murray s.s . , van speybroeck l. , 2006 , apj , 640 , 691 wang l. , steinhardt p.j .
, 1998 , apj , 508 , 483 weinberg n.n .
, kamionkowski m. , 2003 , mnras , 341 , 251 willis j.p .
et al . , 2005 , mnras , 363 , 675 wittman d.m . , dellantonio i.p .
, hughes j.p . , margoniner v.e . , tyson j.a . , cohen j.g . , norman d. , 2006 , apj , 643 , 128 wittman d.m . , tyson j.a . , kirkman d. , dellantonio i. , bernstein g. , 2000 , nature , 405 , 143 zhang y.y . , finoguenov a. , bhringer h. , kneib j.p . , smith g.p . , czoske o. , soucail g. , 2007 , a&a , 467 , 437 | we present a joint weak lensing and x - ray analysis of 4 deg@xmath0 from the cfhtls and xmm - lss surveys .
our weak lensing analysis is the first analysis of a real survey using shapelets , a new generation weak lensing analysis method .
we create projected mass maps of the images , and extract 6 weak - lensing - detected clusters of galaxies .
we show that their counts can be used to constrain the power spectrum normalisation @xmath1 for @xmath2 .
we show that despite the large scatter generally observed in the m - t relation derived from lensing masses , tight constraints on both its slope and normalisation @xmath3 can be obtained with a moderate number of sources provided that the covered mass range is large enough . adding clusters from @xcite to our sample , we measure @xmath4 . although they are dominated by shot noise and sample variance , our measurements are consistent with currently favoured values , and set the stage for future surveys .
we thus investigate the dependence of those estimates on survey size , depth , and integration time , for joint weak lensing and x - ray surveys .
we show that deep surveys should be dedicated to the study of the physics of clusters and groups of galaxies . for a given exposure time
, wide surveys provide a larger number of detected clusters and are therefore preferred for the measurement of cosmological parameters such as @xmath5 and @xmath3 .
we show that a wide survey of a few hundred square degrees is needed to improve upon current measurements of these parameters .
more ambitious surveys covering 7000 deg@xmath0 will provide the 1% accuracy in the estimation of the power spectrum and the m - t relation normalisations .
[ firstpage ] gravitational lensing - surveys - dark matter - large - scale structure of universe - cosmological parameters - x - rays : galaxies : clusters |
quite recently , we have constructed a many - body theory for frenkel excitons @xcite similar to the one we have developed for wannier excitons . in these theories , the composite nature of the particles is treated exactly through a set of `` pauli scatterings '' for carrier exchanges between two excitons in the absence of carrier interaction .
exchanges between two or more excitons , which control the many - body physics of these composite particles , are visualized and readily calculated by using a new diagrammatic representation , called `` shiva diagrams '' in relation with their multiarm structure .
alternative approaches which allow one to take into account the pauli exclusion principle in the case of frenkel excitons are based on expressing the fermionic hamiltonian as an infinite series expansion in creation and destruction operators for excitations on atomic sites . in the formalism proposed by agranovich and coworkers @xcite
, these operators obey bosonic commutation rules , while in the method developed by mukamel and coworkers @xcite , the commutation rules for these operators are not bosonic .
the precise forms of the resulting hamiltonian expansions produced by these two approaches also differ from each other .
in contrast to these approaches , the composite boson many - body theory we have constructed @xcite avoids the difficulty of deriving hamiltonian infinite series expansions with an increasing number of operators @xcite , the processes involving any number of excitons being in our formalism treated on the same footing .
this enables us , for the first time in our best knowledge , to self - consistently study interactions between an arbitrary number of frenkel excitons while including the tricky consequence of the pauli exclusion principle in an exact way .
the composite nature of the wannier excitons turns out to be crucial in their many - body physics .
this becomes quite reasonable once we note that the extension of the electron - hole relative motion in an exciton - which is expected to control carrier exchanges - is as large as the scale for coulomb interaction between excitons , this scale being the exciton bohr radius @xmath4 calculated with the semiconductor dielectric constant and the conduction and valence band effective masses .
in contrast , since frenkel excitons are made of electrons and holes on the same atomic site , we could think that , as the extension of their electron - hole relative motion is essentially zero , fermion exchanges between frenkel excitons should reduce to zero ; so that these excitons should behave as elementary bosons . in our previous work on the frenkel
exciton many - body theory @xcite , we have shown that , indeed , the closure relation for @xmath0 frenkel excitons has the same @xmath5 prefactor as the one of @xmath0 elementary bosons , while for wannier excitons , this prefactor is @xmath6 , which is the signature of the fact that wannier excitons are made of two independent fermions .
however , for many other basic properties , frenkel excitons are composite objects definitely .
in particular , they do have pauli scatterings for fermion exchange in the absence of fermion interactions .
these scatterings lead to a normalization factor for @xmath0 identical excitons different from its @xmath7 value for @xmath0 elementary bosons , by a factor @xmath8 which decreases exponentially with @xmath0 .
this normalization factor thus exibits the same `` moth eaten '' effect induced by pauli exclusion as the one found for wannier excitons .
fermion exchanges between @xmath0 frenkel excitons are found to be controlled by the dimensionless parameter@xmath9where @xmath3 is the number of atomic sites in the sample .
it is of interest to note that this parameter is actually similar to the dimensionless parameter @xmath10 which controls fermion exchanges between @xmath0 wannier excitons .
indeed , @xmath11 , like @xmath3 , is the maximum number of excitons a sample of volume @xmath12 can accommodate .
let us stress that , while the expression of @xmath13 found for wannier excitons could make us think that the parameter which controls fermion exchanges is linked to the extension @xmath14 of the electron - hole relative motion wave function , the expression of @xmath13 for frenkel excitons rules out this physical understanding - otherwise @xmath13 for these excitons would reduce to zero due to the tight - binding approximation underlying the frenkel exciton concept .
in the present work on frenkel excitons , we are interested in the density dependence of the @xmath0 ground state exciton energy in the born approximation .
we reach this quantity through a procedure similar to the one we have used for wannier excitons , namely we calculate the hamiltonian mean value in a @xmath0 ground state exciton state with momentum @xmath15@xmath16where @xmath17 creates a frenkel exciton with momentum @xmath18 .
let us recall that , while the interacting part of composite quantum particles can not be written as a potential so that a `` zero order state in the exciton interaction '' can not be properly defined , the state @xmath19 plays this role since @xmath20 cancels if we drop all coulomb scatterings between excitons .
as @xmath21 contains coulomb interaction at first order by construction , @xmath22 thus corresponds to the ground state energy of @xmath0 frenkel excitons at first order in coulomb interaction , i.e. , in the born approximation .
we are going to show that , while the @xmath0 exciton normalization factor reads as@xmath23where @xmath8 , calculated in our general paper on the many - body theory of frenkel excitons @xcite , has an expansion in @xmath13 which comes from fermion exchanges and makes it exponentially decrease with @xmath0 , the hamiltonian mean value is found to be _ exactly equal _
to@xmath24@xmath25 comes from the coulomb scattering of just _ two _ excitons due to both direct coulomb processes and indirect coulomb processes , i.e. , the so - called `` electron - hole exchange '' which is the only process responsible for the excitation transfer in the case of frenkel excitons .
( 1.4 ) thus leads to@xmath26@xmath27\]]this shows that the hamiltonian mean value depends on the dimensionless parameter @xmath13 associated to exciton density through a first order term only , corrections being in @xmath28 and @xmath29 , but not in @xmath30 .
this has to be contrasted with the hamiltonian mean value for @xmath0 wannier excitons that we have found to expand as @xcite@xmath31 in 3d systems for which the exciton ground state energy is @xmath32where @xmath33 is the exciton rydberg . we have explained the presence of terms in density higher than @xmath13 by saying that , while @xmath34 contains one @xmath35 coulomb process only by construction , interactions between more than two excitons exist through fermion exchanges induced by the composite nature of the particles .
if we now return to frenkel excitons , the reason for the disappearance of high order terms is far from trivial .
while the tight binding approximation underlying the frenkel exciton concept makes most coulomb exchange processes between excitons reducing to zero , this is not enough to explain the result since the denominator of @xmath34 contains a density expansion through @xmath8 .
the fact that @xmath36 does not contain terms in @xmath37 with @xmath38 actually means that similar @xmath37 terms exist in the coulomb part of @xmath34 in order to cancel the @xmath37 terms coming from @xmath8 .
this cancellation is however not mathematically exact since it holds at the order @xmath2 , which indeed reduces to zero in the large sample limit .
it is clear that the proof of such a tricky cancellation must rely on a very precise procedure in which the composite nature of the excitons is treated exactly .
this is why the present work heavily uses the composite boson many - body theory for frenkel excitons we have just constructed paper1,paper2 .
the main results of this theory which is very new , are recalled in section ii for the reader to possibly follow the rest of the argument . in section iii , we start with a simpler two - body problem , namely the calculation of the hamiltonian mean value for @xmath39 frenkel excitons , while in section iv , we turn to arbitrary @xmath0 . this mean value is compared to a similar quantity obtained with excitons replaced by elementary bosons , @xmath34 then only having a first order term in density . in section
v , we conclude .
many - body effects with composite excitons can be derived through a set of four commutators @xcite .
the ones linked to fermion exchanges are obtained from@xmath40 = nb_{0}^{\dagger n-1}\sum_{n}\left\ { \lambda \left ( _ { m\ i}^{n\ 0}\right ) + ( m\longleftrightarrow n)\right\ } b_{n}^{\dagger } \tag{2.1}\]]@xmath41 = nb_{0}^{\dagger n-1}\left ( \delta _ { m0}-d_{m0}\right)\]]@xmath42where definitions of the deviation - from - boson operator @xmath43 and the pauli scattering @xmath44 between two excitons follow from these equations taken for @xmath45 . in the same way , many - body effects linked to fermion interactions are obtained from @xcite@xmath46 = nb_{0}^{\dagger n-1}\sum \xi \left ( _ { m\ i}^{n\ 0}\right ) b_{m}^{\dagger } b_{n}^{\dagger } \tag{2.3}\]]@xmath47 = nb_{0}^{\dagger n-1}\left ( e_{0}b_{0}^{\dagger } + v_{0}^{\dagger } \right)\]]@xmath48where definitions of the exciton creation potential @xmath49 and the interaction scattering between two excitons @xmath50 follow from these equations taken for @xmath45 . in the case of wannier excitons , the index @xmath51 stands for ( @xmath52 , @xmath53 ) where @xmath52 is the exciton center of mass momentum and @xmath53 a relative motion index . due to the tight binding approximation which in some sense freezes electron on top of hole and makes frenkel excitons constructed on electron and hole on the same atomic site ,
these excitons do not have relative motion index ; so that they only are characterized by a center of mass momentum @xmath18 .
their creation operator @xmath54 reads in terms of free pairs @xmath55 on atomic site @xmath56 as@xmath57where @xmath3 is the number of these atomic sites and @xmath58 their location . in the same way ,
free pairs read in terms of excitons as@xmath59as easy to check from the system periodicity which makes the following equations valid@xmath60@xmath61 the fact that frenkel excitons are only characterized by their center - of - mass momentum , leads to results linked to their many - body exchanges far simpler than the ones for wannier excitons .
in particular , the pauli scattering of two frenkel excitons is structureless : it only contains the expected momentum conservation between the `` in '' excitons ( @xmath62 , @xmath63 ) and the `` out '' excitons ( @xmath64 , @xmath65)@xmath66more generally , the shiva diagram for carrier exchange between @xmath0 frenkel excitons shown in fig .
1(b ) leads to@xmath67 by using the pauli scattering given in eq .
( 2.9 ) , we find that the commutators ( 2.1 , 2 ) for fermion exchanges reduce to@xmath68 = \frac{2n}{n_{s}}b_{0}^{\dagger n-1}b_{\mathbf{q - q}^{\prime } } ^{\dagger } \tag{2.11}\]]@xmath69 = nb_{0}^{\dagger n-1}\left ( \delta _ { \mathbf{q}0}-d_{\mathbf{q}0}\right ) -\frac{n(n-1)}{n_{s}}b_{0}^{\dagger n-2}b_{\mathbf{-q}}^{\dagger } \tag{2.12}\]]from them , it becomes easy to show that the factor @xmath8 , defined in eq .
( 1.3 ) , which makes @xmath0 frenkel excitons differing from @xmath0 elementary bosons , obeys the recursion relation@xmath70so that this factor is just given by@xmath71this compact expression of @xmath8 as well as the rather simple commutators for fermion exchanges between @xmath0 excitons given in eq .
( 2.11 , 12 ) are of great help to get exact results on frenkel exciton many - body effects as shown below .
if we now turn to the interaction scattering induced by the coulomb interaction which exists between electrons in the various atomic levels , we have been led to split it as@xmath72 actually , the last part @xmath73 of the interaction scattering @xmath74 does not play any role in the many - body physics of frenkel excitons because , when inserted in equations like eq .
( 2.4 ) , it gives@xmath75indeed , this neutrality scattering , which comes from the energy @xmath76 we must pay to separate electron from hole , is equal to@xmath77while , due to eqs .
( 2.5 , 8) and the fact that @xmath78 , we do have@xmath79 the second part @xmath80 of the interaction scattering @xmath74 comes from the _ indirect coulomb _ processes between atomic levels , these processes insuring the excitation transfer over the whole system in the case of frenkel excitons . such a scattering , quite specific to frenkel excitons , can actually be seen as a `` transfer assisted exchange '' , the coupling between the `` in '' excitons ( @xmath81 , @xmath63 ) and the `` out '' excitons ( @xmath64 , @xmath65 ) occurring through the pauli scattering of these excitons@xmath82its energy - like prefactor is related to the @xmath18 dependent part of the `` out '' exciton energy , namely@xmath83where @xmath84 is the coulomb potential between atomic sites at @xmath85 , one electron going from the atomic state @xmath86 to @xmath87 while the other goes from @xmath88 to @xmath89 , the relevant states for frenkel excitons being the ground state @xmath90 and the first excited state @xmath91 .
this coulomb coupling @xmath92 reads in terms of atomic wave functions as@xmath93 note that this transfer scattering @xmath80 which does not depend on the momentum transfer between the `` in '' and `` out '' excitons , but only on the momenta of `` out '' excitons , has similarity with the `` photon assisted exchange scattering '' we have recently identified between two polaritons @xcite . the first part @xmath94 of the interaction scattering @xmath74 in eq .
( 2.15 ) comes from _ direct coulomb _ processes between electrons , between holes and between electrons and holes . as for the coulomb scatterings of two wannier excitons ,
this scattering depends on the momentum transfer between the `` in '' excitons ( @xmath81 , @xmath63 ) and the `` out '' excitons ( @xmath64 , @xmath65 ) through@xmath95the prefactor @xmath96 is similar to @xmath97 but for all possible direct coulomb processes@xmath98 \tag{2.23}\ ] ] if we now construct exchange coulomb scattering between two excitons along the standard procedure , namely@xmath99we find from eq . ( 2.9 ) and the fact that all scatterings conserve momentum , that@xmath100it is then easy to see that , due to eq .
( 2.8 ) , the sum over @xmath18 reduces to zero for @xmath94 as well as for @xmath80 since the @xmath101 term is missing from the sums they contain .
consequently,@xmath102 in the same way , if we construct the exchange scatterings for @xmath80 and @xmath94 between more than two excitons by adding a coulomb line on the shiva diagram shown on fig .
1b , we find from eq .
( 2.10 ) that the coulomb exchange scatterings shown in fig .
2a in the case of four excitons reads as@xmath103@xmath104so that , this four - body exchange , as well as all high order exchanges , contains the same sum over @xmath18 which cancels .
this shows that all coulomb exchange scatterings between frenkel excitons of the type of fig .
2a reduce to zero .
a way to understand the cancellation of these scatterings could be to say that frenkel excitons are made of electron on top of hole ; so that the `` relative motion wave function '' of these pairs reduces to zero .
however this understanding is far too nave since pauli scatterings for carrier exchange between two excitons do exist , as seen from the fact that @xmath8 is not equal to 1 but exponentially small .
actually , there are two types of exchange coulomb scatterings : we can either have the one of fig . 2a in which all the `` in '' and `` out '' excitons are constructed on different pairs , but we can also have the one of fig .
2b in which one exciton stays made with the same pair .
this other coulomb exchange scattering which , due to momentum conservation , reads as@xmath105@xmath106does not contain the sum over @xmath18 appearing in eqs .
( 2.25 ) or ( 2.27 ) ; so that it does not reduce to zero . before going further , it can be of interest to note that coulomb exchange processes like the one of fig .
2c a priori exist for wannier excitons .
however , due to the symmetry between electron and hole in the case of wannier excitons , the @xmath35 direct coulomb scattering of these excitons is such that @xmath107so that coulomb exchange scattering like the ones of fig .
2c reduces to zero for @xmath108 .
in contrast , since the wave functions for electrons in the atomic ground state and first excited state which enter the definition of @xmath109 through @xmath84 defined in eq .
( 2.21 ) are different , such an electron - hole symmetry does not exist for frenkel excitons .
this makes scattering like @xmath110 different from zero for @xmath111 . at this stage
, we can roughly say that exchange coulomb scatterings like the ones of fig .
2a are the relevant ones for wannier excitons , as they are responsible for the @xmath13 expansion of @xmath34 while this type of scatterings reduces to zero for frenkel excitons . in contrast ,
diagrams like the ones of fig .
2b are going to control coulomb exchange processes for frenkel excitons while they reduce to zero for wannier excitons when one of the `` in '' excitons stay in the same state .
all this tends to show that fermion exchanges between frenkel excitons are unimportant in some configurations but important in a few others .
so that a careful calculation of these exchange processes through a secure many - body theory as the one we have developed , is highly desirable to possibly trust the obtained results .
as for many other problems we have already tackled , it is quite valuable to first perform calculations for @xmath39 because these calculations are far simpler while most of the many - body physics already shows up . in order to calculate @xmath34 defined in eq .
( 1.2 ) for @xmath39 , we first note that , due to eqs .
( 1.3 ) , ( 2.3)@xmath112@xmath113while eq . ( 2.4 ) taken for @xmath39 , leads to@xmath114since @xmath115 as easy to check by taking eq . ( 2.4 ) for @xmath45 acting on @xmath116.the hamiltonian matrix element between two ground state excitons thus appears as@xmath117to go further , we use the scalar product of two frenkel exciton states which , according to eqs . ( 2.1 , 2 , 9 ) reads as@xmath118when inserted into eq .
( 3.3 ) , the exchange coulomb scattering generated by the @xmath119 term of this scalar product reduces to zero due to eqs .
( 2.24 ) and ( 2.26 ) .
we are thus left with@xmath120 in view of this result , we are led to expect the hamiltonian expectation value for @xmath0 excitons to have a @xmath121 prefactor in front of the interaction term @xmath122 of @xmath123 since the two excitons involved in this scattering are now taken among @xmath0 ; plus possibly higher order terms in @xmath124 ... , due to fermion exchanges between more than two excitons .
in addition , the term @xmath2 in the denominator of eq .
( 3.5 ) could possibly be replaced by @xmath125 due again to fermion exchanges appearing in @xmath8 .
we are going to show that the amount of @xmath0-body exchanges in the coulomb term exactly cancels the exchanges in @xmath8 ; so that the interaction term of @xmath34 ends by reading as the one for @xmath123 with just a @xmath121 prefactor and nothing else
. let us now prove this unexpected cancellation which actually holds , within terms in @xmath2 .
according to eq . ( 1.3 ) , the denominator of @xmath34 in eq .
( 1.2 ) is just @xmath126 . to calculate the numerator , we use eq .
( 2.4 ) for @xmath127 $ ] .
this readily gives@xmath128 @xmath129 since scattering conserves momentum , @xmath130 differs from zero for @xmath131 only .
so that to get @xmath132 , we must calculate@xmath133as for @xmath134 , this scalar product is just @xmath126 , we in fact need to calculate @xmath135 for @xmath136 only .
since @xmath137 as seen from eq .
( 2.12 ) taken for @xmath45 acting on vacuum , this eq . ( 2.12 ) readily gives@xmath138to go further , we use @xmath139 = 1-d_{\mathbf{qq}}$ ] .
this leads to split the rhs of the above equation as@xmath140@xmath141 @xmath142 \tag{4.4}\]]the first term is just @xmath143 . the second term , calculated from eq . ( 2.11 ) ,
gives @xmath144 ( n-2)!f_{n-2}$ ] . in the last term , we use eq .
( 2.12 ) for @xmath145 .
this allows us to rewrite this term as @xmath146 g_{n-2}(\mathbf{q})$ ] . by collecting these three terms ,
we get a recursion relation between the @xmath147 s .
it reads@xmath148 @xmath149using eq .
( 2.13 ) for @xmath8 , it is then easy to check that this recursion relation is fulfilled for @xmath150so that we end with a surprisingly simple expression for @xmath135 , namely@xmath151when inserted into the sum of eq .
( 4.1 ) , we find that this sum reduces to@xmath152since the sum of coulomb scatterings over @xmath18 gives zero according to eqs .
( 2.25 , 26 ) , we get the following compact form for the hamiltonian mean value of @xmath0 frenkel excitons@xmath153it shows that the interaction term is indeed the one for two frenkel excitons with just a @xmath121 prefactor which corresponds to the number of ways to choose these 2 excitons among @xmath0 .
although the above algebraic derivation is quite easy to follow , it may leave the reader unsatisfied as it does not really show why the amount of exchange processes in the coulomb term is exactly the same as the one appearing in the normalization factor within terms in @xmath2 . to grasp this reason , shiva diagrams once more are quite valuable .
let us consider the coulomb term appearing in @xmath154 , namely@xmath155 in it , two excitons @xmath156 have a direct coulomb interaction to become ( @xmath81 , @xmath63 ) .
they can then possibly mix with the other @xmath157 excitons @xmath156 through fermion exchanges to end as @xmath0 excitons @xmath156 .
this sum thus corresponds to to the diagram of fig .
the simplest of these exchanges corresponds to leave @xmath158 excitons @xmath156 unaffected .
so that these @xmath157 excitons appear as @xmath159 .
the two other excitons ( @xmath81 , @xmath63 ) become ( @xmath156 , @xmath156 ) with or without exchanging their fermions .
this first possibility which corresponds to the diagram of fig .
3b , thus leads to @xmath160 \left [ n(n-1)\left ( \xi \left ( _ { \mathbf{0}\ \mathbf{0}}^{\mathbf{0}\ \mathbf{0}}\right ) -\xi ^{in}\left ( _ { \mathbf{0}\ \mathbf{0}}^{\mathbf{0}\ \mathbf{0}}\right ) \right ) \right ] \tag{4.11}\]]the @xmath161 factor in the second bracket comes from the number of ways to choose the two excitons @xmath156 on the left among @xmath0 . a second possibility is to mix one exciton @xmath156 with the excitons ( @xmath81 , @xmath63 ) while leaving the other @xmath162 excitons @xmath156 apart .
this second possibility corresponds to the diagram of fig .
we have shown in eq .
( 2.27 ) that coulomb exchange scatterings like the one of fig .
2a reduce to zero . by using the value of the exchange shiva diagram of fig .
1b given in eq .
( 2.10 ) , we then find from the diagram of fig . 3c that this second possibility produces a term@xmath163 \left [ n(n-1)(n-2)\frac{1}{n_{s}}\xi \left ( _ { \mathbf{0}\ \mathbf{0}}^{\mathbf{0}\ \mathbf{0}}\right ) ( n-2)\right ] 2 \tag{4.12}\]]in the shiva expansion of @xmath164 .
the factor @xmath124 comes from the number of ways to choose the three excitons @xmath156 of the left among @xmath165 while the factor @xmath157 comes from the number of ways to choose the exciton @xmath156 on the right among @xmath157 .
this term has an overall minus sign because one fermion exchange is involved .
an additional factor 2 comes from the number of diagrams which give the same nonzero contributions and which have the structure like the diagram of fig .
a third possibility is to mix two excitons @xmath156 with the excitons ( @xmath81 , @xmath63 ) .
the nonzero coulomb exchange diagrams are shown in fig .
they leads to@xmath166 \left [ n(n-1)(n-2)(n-3)\frac{1}{n_{s}^{2}}\xi \left ( _ { \mathbf{0}\ \mathbf{0}}^{\mathbf{0}\ \mathbf{0}}\right ) ( n-2)(n-3)\right ] 3 \tag{4.13}\]]and so on ... this shiva diagram expansion of the coulomb term @xmath164 defined in eq . ( 4.10 ) allows us to write it as@xmath167@xmath168@xmath169 the last step is to relate @xmath170 to @xmath8 . a simple way to do
it is to consider the scalar product@xmath171shown in fig .
5a . let us perform a shiva diagram expansion of this scalar product according to the standard rules @xcite , namely we isolate @xmath172 , @xmath162 , ... excitons @xmath156 and write all the possible diagrams in which the remaining excitons @xmath156 are `` not alone '' .
this leads to the diagram of fig .
5c for the second term . and so on ... : by using the results for shiva diagrams in the case of frenkel excitons given in eqs .
( 2.9 , 10 ) , we get for @xmath173 , i.e. , when @xmath174 reduces to @xmath126@xmath175@xmath176 @xmath177 when inserted into eqs .
( 4.14 ) and ( 4.1 ) , this result allows one to readily recover the expression of the hamiltonian mean value for @xmath0 frenkel excitons given in eq .
( 4.9 ) .
it is clear that such a derivation needs to be knowledgeable with shiva diagrams .
it however has the great advantage to pick out in a transparent way from where the final result comes . in the present case
, this is from coulomb exchange processes like the one of fig .
2 . in order to identify the microscopic coulomb processes which control the ground state energy of @xmath0 frenkel excitons , we first use the definition of the interaction scattering given in eqs .
( 2.15 , 16 , 19 ) .
this leads to@xmath178when compared to eq .
( 1.4 ) , the microscopic value of the coulomb contribution to the energy is thus found to read @xmath179 . to get a precise understanding of the various contributions to the hamiltonian mean value ,
let us now write the above quantities in terms of the various elementary coulomb scatterings between atomic sites .
for that , we use eqs .
( 2.16 , 17 , 18 , 19 , 20 ) and the expression of the @xmath18 exciton energy , namely @xmath180 where the free pair energy @xmath181 with @xmath182 is just the energy of one electron in the excited level and one hole in the ground state level , minus the neutrality energy @xmath183 we would pay for not having electron and hole on the same atomic site ( see ref .
this leads to@xmath184@xmath185 \right ] \tag{4.19}\]]since in the last sum , we can replace @xmath186 by @xmath187 , due to eq .
( 2.21 ) .
we see that the interaction term , which cancels for @xmath45 , as expected , contains all possible coulomb processes between the atomic levels 0 and 1 of the different atomic sites , through the sum over @xmath188 . due to the orhogonality of the @xmath189 atomic states , it is of importance to note that the indirect coulomb processes @xmath190 in which electrons change from atomic state 1 to atomic state 0 are expected to be much smaller than the direct coulomb processes @xmath191 ; so that the interaction term of the hamiltonian expectation value is essentially controlled by direct coulomb processes @xmath84 .
let us however stress that , while these indirect scatterings are not that important in the interaction term of the hamiltonian mean value , they play a major role in the frenkel exciton physics because , although extremely small , they are the only processes allowing an excitation transfer between sites as necessary to produce the exciton .
since the coulomb matrix elements @xmath192 entering the hamiltonian mean value are defined in terms of wave functions for electron in the ground and first excited atomic states , they strongly depend on the chemical properties of the material at hand .
this has to be contrasted with wannier excitons which are constructed on free electrons and free holes , the material of interest just appearing through effective masses induced by the lattice periodicity . for these wannier excitons ,
the intraband coulomb matrix element thus reduces to @xmath193 in 3d . as a result ,
the hamiltonian mean value @xmath34 for these excitons only depends on one input parameter which is the exciton bohr radius ; this is why its @xmath13 expansion can be derived explicitly for all materials ( see eq .
( 1.6 ) ) . in contrast
, the prefactor of the @xmath13 term for frenkel excitons highly depends on the material at hand . if we now come back to eq .
( 4.18 ) , we can , in the large sample limit , replace @xmath194 by @xmath3 .
so that , for @xmath0 large compared to 1 , the hamiltonian expectation value expands as given in eq .
( 1.5 ) : it only contains a linear term in density . again , this has to be contrasted with @xmath34 for wannier excitons in which the coulomb term reads as a density expansion in @xmath13 . for completeness ,
let us note that the hamiltonian mean value for @xmath0 identical elementary bosons also has just one linear contribution in density .
this is expectable since the hamiltonian mean value contains the interaction at first order only by construction : when the particles are taken as elementary bosons the @xmath35 interaction couples two bosons among @xmath0 , no fermion exchange exists to possibly mix these two particles with the rest of the system to produce higher order couplings .
this is supported by the precise calculation of the hamiltonian mean value@xmath195which for bosonized excitons such that @xmath196 = \delta _ { \mathbf{q}^{\prime } \mathbf{q}}$ ] , has a denominator which reduces to @xmath7 while , for an effective bosonic hamiltonian written as@xmath197its numerator follows , since @xmath198 , from@xmath199@xmath200this leads to an hamiltonian mean value given by@xmath201by noting that the scattering @xmath202 must decrease as @xmath203 with the sample volume , as necessary for the extensivity of the system , we see that the correction to the energy of one bosonized exciton just increases linearly with the exciton density @xmath204 .
in this paper , we calculate the ground state energy of @xmath0 frenkel excitons in the born approximation through the hamiltonian mean value in the state made of @xmath0 ground state excitons .
we show that the energy change induced by exciton interactions contains one linear term only in the exciton density , while higher order terms appear in the case of wannier excitons
. a naive reason could be that , since frenkel excitons are constructed on highly localized atomic wave functions , their exchange coulomb scatterings must reduce to zero , making frenkel excitons appearing as elementary bosons .
this however does not hold since the normalization factor of @xmath0 frenkel excitons is exponentially small compared to the one for @xmath0 elementary bosons .
the deep nonobvious reason for having one linear term only in the hamiltonian mean value , is the fact that the same exponentially small prefactor also exists in the coulomb term of the hamiltonian matrix element , within corrections in @xmath2 where @xmath3 is the number of atomic sites .
so that , in the end , all the exchange processes appearing in @xmath36 finally disappear , as if frenkel excitons were true elementary particles , within terms in @xmath2 . here again , shiva diagrams enlighten the physical understanding of the fermion exchanges which lead to this rather unexpected cancellation . | by using the composite many - body theory for frenkel excitons we have recently developed , we here derive the ground state energy of @xmath0 frenkel excitons in the born approximation through the hamiltonian mean value in a state made of @xmath0 identical @xmath1 excitons . while this quantity reads as a density expansion in the case of wannier excitons , due to many - body effects induced by fermion exchanges between @xmath0 composite particles ,
we show that the hamiltonian mean value for @xmath0 frenkel excitons only contains a first order term in density , just as for elementary bosons .
such a simple result comes from a subtle balance , difficult to guess a priori , between fermion exchanges for two or more frenkel excitons appearing in coulomb term and the ones appearing in the @xmath0 exciton normalization factor - the cancellation being exact within terms in @xmath2 where @xmath3 is the number of atomic sites in the sample .
this result could make us naively believe that , due to the tight binding approximation on which frenkel excitons are based , these excitons are just bare elementary bosons while their composite nature definitely appears at various stages in the precise calculation of the hamiltonian mean value . |
light - front models are useful to describe hadronic bound states , like mesons or baryons due to its particular boost properties @xcite . however , the light - front description in a truncated fock - space breaks the rotational symmetry because the associated transformation is a dynamical boost @xcite .
therefore , an analysis with covariant analytical models , can be useful to pin down the main missing features in a truncated light - front fock - space description of the composite system . in this respect , the rotational symmetry breaking of the plus component of the electromagnetic current , in the drell - yan frame ,
was recently studied within an analytical model for the spin-1 vertex of a composite two - fermion bound state @xcite .
it was shown that , if pair term contributions are ignored in the evaluation of the matrix elements of the electromagnetic current , the covariance of the form factors is lost @xcite . the complete restoration of covariance in the form factor calculation is found only when pair terms or zero modes contributions to the matrix elements of the current are considered @xcite .
the extraction of the electromagnetic form - factors of a spin-1 composite particle from the microscopic matrix elements of the plus component of the current ( @xmath1 ) in the drell - yan frame ( momentum transfer @xmath2 ) , based only on the valence component of the wave function , presents ambiguities due to the lacking of the rotational invariance of the current model @xcite . in the breit - frame with momentum transfers along the transverse direction ( the drell - yan condition is satisfied ) the current @xmath3 has four independent matrix elements , although only three form factors exist .
therefore , the matrix elements satisfies an identity , known as the angular condition @xcite , which is violated .
several extraction schemes for evaluating the form - factors were proposed , and in particular we consider the suggestion made in ref .
it was found in a numerical calculation of the @xmath4-meson electromagnetic form factors considering only the valence contribution @xcite , that the prescription proposed by @xcite to evaluate the form - factors , produced results in agreement with the covariant calculations . in ref .
@xcite , it was used an analytical form of the @xmath4-quark - antiquark vertex .
later , in ref .
@xcite , it was shown that the above prescription eliminates the pair diagram contributions to the form factors , using a simplified form of the model , when the matrix elements of the current were evaluated for spin-1 light - cone polarization states .
this nice result was thought to be due to the use of the particular light - cone polarization states . here
, we will show in a straightforward and analytic manner that the cancelation of the pair contribution in the evaluation of the form factors using the prescription from ref .
@xcite also happens for the instant form polarization states in the cartesian representation , generalizing the previous conclusion @xcite .
our aim , is to expose in a simple and detailed form , how the pair terms appear in the matrix elements of the current evaluated between instant form polarization states , and their cancelation in the form factors using the correct prescription .
therefore , we conclude that this property is more general than realized before .
but , in the case of spin-0 composite particles ( like the pion ) with the correspondent form of the analytical model , the plus component of the electromagnetic current in the breit - frame , with @xmath5 , does not have contributions from pair terms @xcite .
the electromagnetic current has the following general form for spin-1 particles @xcite : @xmath6 ( p^\mu + p^{\prime \mu } ) - f_3(q^2 ) ( q_\alpha g_\beta^\mu- q_\beta g_\alpha^\mu ) \ , \label{eq : curr1}\ ] ] where @xmath7 is the mass of the vector particle , @xmath8 is the momentum transfer , @xmath9 and @xmath10 is on - shell initial and final momentum respectively . from the covariant form factors @xmath11 ,
@xmath12 and @xmath13 , one can obtain the charge ( @xmath14 ) , magnetic ( @xmath15 ) and quadrupole ( @xmath16 ) form factors ( see e.g. @xcite ) .
the matrix elements of the electromagnetic current @xmath17 in the impulse approximation are written as @xcite : @xmath18 \lambda(k , p_f)\lambda(k , p_i ) } { ( k^2 - m^2+\imath \epsilon ) ( ( k - p_i)^2 - m^2+\imath\epsilon ) ( ( k - p_f)^2 - m^2+\imath \epsilon ) } \ ] ] where @xmath19 and @xmath20 are the polarization four - vectors of the final and initial states , respectively and @xmath21 is the quark mass .
the electromagnetic form - factors are calculated in the breit frame with the drell - yan - west condition , which gives the momentum transfer @xmath22 , the particle initial momentum @xmath23 and the the final one @xmath24 .
we use @xmath25 and @xmath26 .
the polarization four - vectors in instant - form basis are given by @xmath27 for the initial state and by , @xmath28 for the final state .
the regularization function @xmath29 is enough to render finite the photo - absorption amplitude ; the regularization parameter is @xmath30 .
the vertex function @xmath31 for the vector particles utilized ( see ref .
@xcite for details ) is given below : @xmath32 the vector particle is on - mass shell ; @xmath7 is the vector bound state mass , and its four momentum @xmath33 .
after the integration in the light - front energy ( @xmath34 ) , the light - front wave function is writing like : @xmath35 \ .
\label{eq : npwf}\ ] ] the extraction of the electromagnetic form factor with the plus component of the electromagnetic current and the angular condition are discussed next section .
the matrix elements of the electromagnetic current expressed in terms of the current in the instant form spin basis , after the use of the melosh rotation @xcite is : @xmath36 here the matrix elements of the electromagnetic current are calculated with the plus component of the current ,
@xmath37 and the light - front polarizatiion states denoted as @xmath38 . in the breit - frame ( @xmath5 ) , the angular condition is translated by the equation @xcite : @xmath39 in the case of the instant form spin basis , the angular condition is @xmath40 @xcite .
the prescription adopted by grach and kondratyuk @xcite eliminate the matrix elements @xmath41 in order to compute the electromagnetic form factors for spin-1 particles .
the terms of the trace calculated with only the @xmath42 structure of the vertex , of the eq.([rhov ] ) , is writing below : @xmath43_{ji}= tr\left [ \sla{\epsilon}_j^{\alpha } ( \sla{k}-\sla{p}^{\prime } + m ) \gamma^+ ( \sla{k } - \sla{p } + m ) \sla{\epsilon}_i^{\alpha } ( \sla{k } + m ) \right ] \ . \label{trgaga}\ ] ] the trace is calculated with the light - front coordinates , ( @xmath44 ) and the @xmath45 dependence is separate , ( in the eq.([trgaga ] ) ) , then , we get the results below : @xmath43_{ji}^{+z } = \frac{k^-}{2 } tr\left [ \sla{\epsilon}_j^{\alpha } ( \sla{k}-\sla{p}^{\prime } + m ) \gamma^+ ( \sla{k } - \sla{p } + m ) \sla{\epsilon}_i^{\alpha } \gamma^+ \right ] \ .\ ] ] for the polarization four - vectors @xmath46 , the traces above are given by : @xmath47^{+z}_{xx } & = & - k^- \frac{\eta}{8 } r_{gg } \ , \
_ { yy}^{+z } = \ 4 k^- ( p^+ -k^+)^2 \ , \nonumber \\ tr[gg]_{zz}^{+z } & = & \frac18 ~k^- ~r_{gg } \ , \
tr[gg]^{+z}_{zx } = - k^- \frac { \sqrt{\eta } } { 8 } r_{gg } \label{traces } \ , \end{aligned}\ ] ] where @xmath48 \label{badtrace } \ .
$ ] the trace equation , eq.([traces ] ) , yield the following relations : @xmath49^{+z}_{xx } & = & -\eta tr [ gg ] _ { zz}^{+z } ~ , \nonumber \\ tr [ gg ] ^{+z}_{zx } & = & -\sqrt{\eta}~tr [ gg ] ^{+z}_{zz } \ . \label{trrelations}\end{aligned}\ ] ] the zero - modes contributions for the direct term coupling given by : @xmath50 = \lim_{\delta^+ \rightarrow 0_+}\;\sum_{r=4,5\ ; ; \;s=4,6 } \int [ d^4 k]^{+z } \frac { tr[gg ] ^{+z}_{ji } } { [ 1 ] [ 2 ] [ 3 ] [ r ] ^2 [ s ] ^2 } \ , \label{currimgg}\ ] ] where @xmath51^z=[d^4 k]~\theta(p^{\prime+}-k^+ ) \theta(k^+-p^{+})$ ] , and the denominators are given by : @xmath52 = [ k^2-m^2+\imath\epsilon]$ ] , @xmath53 = [ ( k - p)^2-m^2 + \imath \epsilon]$ ] , @xmath54 = [ ( k - p^{\prime } ) ^2 - m^2 + \imath \epsilon]$ ] , @xmath55 = [ ( k - p)^2-m_r^2 + \imath \epsilon]$ ] , @xmath56 = [ ( k - p^{\prime } ) ^2 - m_r^2 + \imath \epsilon]$ ] , @xmath57 = [ k^2- m_r^2 + \imath \epsilon]$ ] and @xmath58 . the integration of the eq.([currimgg ] ) is performed with the _ pole dislocation method _ , developed in the reference @xcite , after that , the contribution of the _ zero modes _ for the matrix element @xmath59 of the electromagnetic current is zero and the matrix element @xmath60 of the electromagnetic current not have a pair term contribuition . using the prescription of the ref.@xcite
, the electromagnetic form factors for spin-1 particles are writing with matrix elements of the electromagnetic current in instant form spin basis as : @xmath61 here the _
superscripts gk _ in the equation above is the initials for the authors of the ref .
@xcite , grach , i. and kondratyuk , l. a. considering the relations in eq.([trrelations ] ) , the prescription given by grach @xcite , cancel out the zero - modes contribution to the electromagnetic form factors for this part of the vertex function ( see the figures 1 and 2 ) . then the angular condition in the cartesian spin basis with the zero - modes contribuition is @xcite : @xmath62 because the zero modes contribuition to the matrix elements @xmath59 of the electromagnetic current is zero ; the angular condition is only due to the valence components of the electromagnetic current , @xmath63 , and to the zero modes contribuition to the electromagnetic current @xmath64 .
the figures 1 and 2 show the results for the electromagnetic matrix elements of the current calculated with the instant form and the light - front approach .
after the zero - modes inclusion , the covariance are restorate .
following the ref .
@xcite we separate the structure of the vertex eq.([rhov ] ) in the valence contribuition and the zero - modes for the spin-1 particles . the prescription sugested by the ref .
@xcite utilized in the calculation of the electromagnetic form factor for spin-1 particles cancel out the zero - modes contribuition to the matrix elements for the electromagnetic current ; then the electromagnetic form factors for spin-1 calculated with the _ grach _ and _ kondratyuk _
@xcite are free of the non - valence contribuition and only the valence part of the electromagnetic current contributed to the electromagnetic form factors .
it is possible to generalize the conclusion found here for the full electromagnetic matrix elements of the spin-1 particles with the vertex function , eq.([rhov ] ) ; i.e , the zero - modes not contributed to the electromagnetic form factors if the prescription of the ref.@xcite it is adopted .
this work was supported in part by the brazilian agencies fapesp ( fundao de amparo pesquisa do estado de so paulo ) and cnpq ( conselho nacional de desenvolvimento cientfico e tecnolgico ) .
j. p. b. c. de melo thanks the organizer of the light - cone 2011 for the invitation .
i.l.grach , l.a .
kondratyuk , and m.strikman : _ is the structure in the deuteron magnetic form - factor at @xmath65 approximately = @xmath66 a new evidence for nuclear core?_. phys .
* 62 * , 387 ( 1989 )
. j. .p .
b. c. de melo , t. frederico , e. pace , and g. salm : frame dependence of the pair contribution to the pion electromagnetic form - factor in a light front approach .
jour . of phys . * 33 * , 301 ( 2003 ) .
frankfurt and m. strikman : _ relativistic description of the deuteron : elastic and inelastic form - factors in the region kinematically forbidden for scattering on a free nucleon .
b*148 * , 107 ( 1979 ) . | the contribution of the light - front valence wave function to the electromagnetic current of spin-1 composite particles is not enough to warranty the proper transformation of the current under rotations .
the naive derivation of the plus component of the current in the drell - yan - west frame within an analytical and covariant model of the vertex leads to the violation of the rotational symmetry . computing the form - factors in a quasi drell - yan - west frame @xmath0 ,
we were able to separate out in an analytical form the contributions from z - diagrams or zero modes using the instant - form cartesian polarization basis . |
almost 200 molecules have been detected in the interstellar medium or circumstellar shells so far . in dark clouds , unsaturated carbon - chain molecules such as ccs are abundant , whereas they decrease in star forming cores @xcite .
these carbon - chain molecules are formed from c or c@xmath12 in the gas phase before carbon atoms are converted into co. on the other hand , saturated complex organic molecules such as ch@xmath3oh and ch@xmath3cn increase in star forming regions , and these regions are called hot cores for high - mass star forming regions and hot corinos for low - mass star forming regions . in hot cores and hot corinos ,
the combination between grain - surface reactions and gas - phase reactions produces various saturated complex organic molecules .
recently , in contrast to the above scenarios , several star forming cores associated with carbon - chain molecules have been discovered .
@xcite showed that various carbon - chain molecules are abundant toward the low - mass star forming region l1527 .
several chemical model calculations showed that ch@xmath13 plays an essential role in efficient formation of carbon - chain molecules in the warm regions ( @xmath14 k ) ( e.g. * ? ? ?
* ; * ? ? ?
@xcite indicated that carbon - chain molecules are formed by a combination of gas - phase reactions and grain - surface reactions following the sublimation of ch@xmath13 .
these low - mass star forming regions were named warm carbon chain chemistry ( wccc ) sources .
cyanopolyynes ( hc@xmath15n ) are one of the representative carbon - chain - molecule series .
@xcite demonstrated that regeneration of carbon - chain molecules results from gas - phase chemistry and showed the primary formation pathway of hc@xmath3n as follows : @xmath16 @xcite presented their time - dependent gas - phase chemical model and showed that cyanopolyynes ( hc@xmath15n ; @xmath17 ) can be formed under hot core conditions .
they considered that c@xmath11h@xmath11 is released from the grain mantle inside hot cores and the neutral - neutral reaction between c@xmath11h@xmath11 and cn ( reaction ( [ r6 ] ) ) proceeds relatively easily for the hot core temperature ( @xmath18 k ) .
their chemical model calculation was supported by detections of hc@xmath19n toward 35 hot cores associated with 6.7 ghz methanol masers @xcite .
therefore , the predicted main formation pathway of hc@xmath3n is reaction ( [ r6 ] ) in both the low - mass star forming region l1527 and the high - mass star forming regions .
deriving @xmath2c isotopic fractionation of carbon - chain molecules by observations is one of the promising methods to constrain their main formation mechanism .
for example , the dominant formation mechanism of hc@xmath3n has been investigated based on its @xmath2c isotopic fractionation @xcite toward the cyanopolyyne peak in taurus molecular cloud-1 ( tmc-1 cp ; @xmath20 pc ) .
the derived abundance ratios of three @xmath2c isotopologues of hc@xmath3n are @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] = @xmath21 ( @xmath4 ) ( @xmath6 ) , and it is suggested that its main formation pathway is the neutral - neutral reaction between c@xmath11h@xmath11 and cn @xcite
. thus investigation of formation mechanism of hc@xmath3n using its @xmath2c isotopic fractionation has proved successful . in the present paper
, we observed the three @xmath2c isotopologues of hc@xmath3n , the shortest cyanopolyyne , toward l1527 ( @xmath20 pc ) and g28.28 - 0.36 ( @xmath22 kpc , @xcite ) with the 45-m radio telescope of the nobeyama radio observatory ( nro ) in order to compare @xmath2c isotopic fractionation of hc@xmath3n and its main formation mechanisms in the two star forming regions .
g28.28 - 0.36 is one of the hot cores that @xcite detected hc@xmath19n , and associates with an ultra compact hii region .
@xcite detected ch@xmath3cn , a hot core tracer , and h@xmath2co@xmath12 toward the hot core .
we carried out observations of the three @xmath2c isotopologues of hc@xmath3n ( h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn ) simultaneously with the nro 45-m radio telescope in 2015 december , 2016 feburary , march and april ( 2015 - 2016 season ) .
we also observed the normal species of the @xmath0 rotational lines simultaneously , when we observed the @xmath0 lines of the three @xmath2c isotopologues .
we used the tz and the t70 receivers for the observations of the @xmath0 transition lines , and the tz receiver for the observations of the @xmath23 transition lines .
we used the tz receiver for the observations of the @xmath0 lines , after we could not use the t70 receiver due to equipment troubles .
the rest frequencies of the observed lines are given in table [ tab : tab2 ] .
both receivers allow us to obtain dual - polarization data simultaneously .
the beam sizes ( hpbw ) and main beam efficiencies ( @xmath24 ) were 18 and 54% for the tz receiver and 20 and 56% for the t70 receiver at 86 ghz , respectively .
the system temperatures were between 120 and 250 k , depending on the weather conditions and the elevations .
we employed the position - switching mode . the observed position and off - source position for l1527 were ( @xmath25 , @xmath26 ) = ( 04@xmath2739@xmath285389 , 2603110 ) and ( 04@xmath2742@xmath28359 , 2553233 ) , respectively .
the telescope pointing was checked using the h40 receiver every 1.5 hours by observing the sio maser line ( @xmath29 ) from nml tau , and the pointing error was less than 3 .
the observed position for g28.28 - 0.36 was ( @xmath25 , @xmath26 ) = ( 18@xmath2744@xmath28133 , -0418033 ) , and the off - source position was set to be @xmath30 away in the declination .
we checked the pointing accuracy by observing the sio maser lines ( @xmath29 ) from oh39.7 + 1.5 every @xmath31 hours , depending on the wind conditions .
the pointing error was within 3 .
we used the sam45 fx - type digital correlator in frequency settings whose bandwidths and resolutions are 125 mhz and 30.52 khz , and 250 mhz and 61.04 khz for l1527 and g28.28 - 0.36 , respectively .
we applied 2 channel binding , and velocity resolutions of final spectra are 0.25 and 0.5 km s@xmath32 for l1527 and g28.28 - 0.36 , respectively .
the spectra of the three @xmath2c isotopologues of hc@xmath3n were taken with signal - to - noise ratios of @xmath33 in l1527 , except the @xmath23 line of h@xmath2cccn , as shown in figure [ fig : f1 ] .
we also show the spectra of the @xmath0 rotational line of the normal species observed with its @xmath2c isotopologues simultaneously .
the values of _
v_@xmath34 are in good agreement with the _ v_@xmath34 value reported for this source ( 5.9 km s@xmath32 ) .
we derived their integrated intensities with the same velocity range , and the results are summarized in table [ tab : tab1 ] .
the ratios of the integrated intensities among the three @xmath2c isotopologues of the @xmath23 lines are consistent with those of the @xmath0 lines .
however , we can not detect h@xmath2cccn using the @xmath23 line with signal - to - noise ratio above 3 , and we do not use the @xmath23 lines in the following analyses .
we evaluated the error of the integrated intensities using the following formula : @xmath35 in equation ( [ error ] ) , @xmath36_t_@xmath37 is the rms noises in the emission - free regions , @xmath38 is the numbers of channels , and @xmath39 is the velocity resolution per channel .
the rms noises are summarized in table [ tab : tab2 ] .
we used 4 ch and 0.25 km s@xmath32 for @xmath38 and @xmath39 , respectively , in l1527 .
the three @xmath2c isotopologues of hc@xmath3n were detected with signal - to - noise ratios of @xmath40 in g28.28 - 0.36 , as shown in figure [ fig : f2 ] .
the values of _ v_@xmath41 agree with one another and the systematic velocity reported for the source ( 48.9 km s@xmath32 @xcite ) .
we derived the integrated intensities of the three @xmath2c isotopologues of hc@xmath3n in the same way as l1527 , and their errors were calculated using equation ( [ error ] ) with 7 ch and 0.5 km s@xmath32 for @xmath38 and @xmath39 , respectively .
we summarize the results in table [ tab : tab1 ] .
cccc l1527 & & & + @xmath0 & 0.021 ( 3 ) & 0.028 ( 3 ) & 0.035 ( 3 ) + @xmath23 & 0.016 ( 4 ) & 0.023 ( 3 ) & 0.032 ( 3 ) + & & & + g28.28 - 0.36 & & & + @xmath0 & 0.055 ( 9 ) & 0.064 ( 7 ) & 0.090 ( 7 ) + @xmath23 & 0.049 ( 5 ) & 0.047 ( 4 ) & 0.059 ( 4 ) + lccccc + + hc@xmath3n & 81.8814677 ( 9 ) & 2154 ( 16 ) & 0.622 ( 6 ) & 5.8 ( 3 ) & 5.5 + h@xmath2cccn & 79.35046 ( 2 ) & 45 ( 6 ) & 0.49 ( 7 ) & 6.0 ( 3 ) & 6.6 + hc@xmath2ccn & 81.53411 ( 2 ) & 52 ( 5 ) & 0.46 ( 6 ) & 6.0 ( 3 ) & 5.9 + hcc@xmath2cn & 81.54198 ( 2 ) & 66 ( 6 ) & 0.47 ( 5 ) & 6.0 ( 3 ) & 6.2 + + h@xmath2cccn & 88.16683 ( 2 ) & @xmath42 & - & - & 7.4 + hc@xmath2ccn & 90.59306 ( 2 ) & 43 ( 5 ) & 0.51 ( 6 ) & 5.9 ( 3 ) & 5.9 + hcc@xmath2cn & 90.60178 ( 2 ) & 39 ( 4 ) & 0.75 ( 9 ) & 5.9 ( 3 ) & 5.9 + + + + hc@xmath3n & 81.8814677 ( 9 ) & 1310 ( 20 ) & 2.19 ( 7 ) & 49.2 ( 5 ) & 6.7 + h@xmath2cccn & 79.35046 ( 2 ) & 34 ( 4 ) & 1.6 ( 2 ) & 49.4 ( 5 ) & 6.6 + hc@xmath2ccn & 81.53411 ( 2 ) & 39 ( 4 ) & 1.6 ( 2 ) & 49.5 ( 5 ) & 5.3 + hcc@xmath2cn & 81.54198 ( 2 ) & 38 ( 3 ) & 2.3 ( 2 ) & 49.3 ( 5 ) & 5.3 + + h@xmath2cccn & 88.16683 ( 2 ) & 28 ( 2 ) & 1.64 ( 17 ) & 49.5 ( 5 ) & 3.9 + hc@xmath2ccn & 90.59306 ( 2 ) & 25 ( 2 ) & 1.86 ( 17 ) & 49.3 ( 5 ) & 2.9 + hcc@xmath2cn & 90.60178 ( 2 ) & 31 ( 2 ) & 1.79 ( 15 ) & 49.0 ( 5 ) & 2.9 + we fitted the spectra with a gaussian profile and obtained the spectral line parameters , as summarized in table [ tab : tab2 ] .
we calculated the column densities of the normal species and the three @xmath2c isotopologues of hc@xmath3n using the local thermodynamic equilibrium ( lte ) analysis as shown in the following formulae @xcite : @xmath43,\ ] ] where @xmath44 and @xmath45 in equation ( [ tau ] ) , _
t_@xmath37 denotes the antenna temperature , _ f _ the beam filling factor , @xmath46 the main beam efficiency , and @xmath47 the optical depth .
we used 1 and 0.54 for _ f _ and @xmath46 ( section [ sec : obs ] ) , respectively .
_ t_@xmath48 is the excitation temperature , and _
t_@xmath49 is the cosmic microwave background temperature ( @xmath50 k ) .
_ j_(_t _ ) in equation ( [ tem ] ) is the planck function . in equation ( [ col ] ) , _ n _ is the column density , @xmath51 is the line width ( fwhm ) , @xmath52 is the partition function , @xmath53 is the permanent electric dipole moment of hc@xmath3n ( @xmath54 esu cm , @xcite ) , and @xmath55 is the energy of the lower rotational energy level .
@xcite derived the excitation temperatures and the column densities of hc@xmath3n using the lte analysis .
the determined excitation temperatures and column densities are @xmath56 k and ( @xmath57)@xmath58 @xmath59 when the @xmath60 and @xmath23 data were included , and @xmath61 k and ( @xmath62)@xmath58 @xmath59 using only the @xmath23 and @xmath63 lines .
we then derived @xmath47 assuming the excitation temperatures of 9.7 k and 16.9 k , respectively , using equation ( [ tau ] ) .
the calculated column densities are summarized in table [ tab : tab3 ] .
the column densities are determined to be ( @xmath64)@xmath58 , ( @xmath65)@xmath66 , ( @xmath67)@xmath66 , and ( @xmath68)@xmath66 @xmath59 using the excitation temperature of 9.7 k , and ( @xmath69)@xmath70 , ( @xmath71)@xmath66 , ( @xmath72)@xmath66 , and ( @xmath73)@xmath66 @xmath59 using the excitation temperature of 16.9k for hc@xmath3n , h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn , respectively .
the two assumed excitation temperatures seem to be the lower and upper limits , and then the derived column densities are also the upper and lower limits .
a change in the assumed excitation temperature by a factor of 2 does not affect the derived column densities of the three @xmath2c isotopologues within 3-sigma errors .
the derived column density of the normal species using the excitation temperature of 9.7 k agrees with that derived by @xcite ( ( @xmath57)@xmath58 @xmath59 ) , while the column density using the excitation temperature of 16.9 k is lower than that calculated by @xcite ( ( @xmath62)@xmath58 @xmath59 ) by 1.5 times .
the abundance ratios of the three @xmath2c isotopologues are derived to be 0.9 ( @xmath4 ) : 1.00 : 1.29 ( @xmath5 ) ( @xmath6 ) for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] . using the column densities of the normal species of ( @xmath57)@xmath58 @xmath59 and ( @xmath62)@xmath58 @xmath59 for the excitation temperatures of 9.7 k and 16.9 k , respectively @xcite
, we also calculated the @xmath74c/@xmath2c ratios of hc@xmath3n , as summarized in table [ tab : tab3 ] .
when the excitation temperature is 9.7 k , the @xmath74c/@xmath2c ratios are determined to be @xmath75 , @xmath76 , and @xmath77 ( @xmath6 ) for h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn , respectively . the @xmath74c/@xmath2c ratios are derived to be @xmath78 , @xmath79 , and @xmath80 ( @xmath6 ) for h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn , assuming that the excitation temperature is 16.9 k. the @xmath74c/@xmath2c ratios do not change depending on the two assumed excitation temperatures within 1-sigma errors .
these ratios would be the lower and upper limits .
we also calculated the column densities of the normal species and the three @xmath2c isotopologues of hc@xmath3n in g28.28 - 0.36 , using the @xmath0 and @xmath23 lines , respectively .
we conducted calculations using equations ( [ tau ] ) - ( [ col ] ) with the excitation temperatures of 50 , 100 , and 200 k , respectively , because the typical temperature in hot cores is @xmath81 k. we summarize the results with the excitation temperature of 100 k in table [ tab : tab4 ] .
the column densities of the normal species and the three @xmath2c isotopologues are derived to be ( @xmath82)@xmath83 , ( @xmath84)@xmath66 , ( @xmath85)@xmath66 , and ( @xmath86)@xmath66 @xmath59 for hc@xmath3n , h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn , respectively , using the @xmath0 lines . using the @xmath23 lines , the derived column densities of
the three @xmath2c isotopologues are ( @xmath87)@xmath88 , ( @xmath89)@xmath88 , and ( @xmath90)@xmath88 @xmath59 for h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn , respectively .
we discuss the differences in the column densities between the @xmath0 and @xmath1 lines in section [ sec : ratio ] .
the values of the column densities do not change depending on the assumed excitation temperatures within 1-sigma errors .
the abundance ratios are found to be 1.0 ( @xmath7 ) : 1.00 : 1.47 ( @xmath8 ) ( @xmath6 ) , and 1.05 ( @xmath91 ) : 1.00 : 1.22 ( @xmath92 ) ( @xmath6 ) for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] using the @xmath0 and @xmath1 lines , respectively .
thus the abundance ratios are consistent between the @xmath0 and @xmath1 lines within 1-sigma errors .
we also derived the @xmath74c/@xmath2c ratios using the @xmath0 lines , and these results are summarized in table [ tab : tab4 ] .
the calculated ratios are @xmath93 , @xmath94 , and @xmath95 for h@xmath2cccn , hc@xmath2ccn , and hcc@xmath2cn , respectively .
the @xmath74c/@xmath2c ratios do not change depending on the assumed excitation temperatures .
lcccc h@xmath2cccn & @xmath96 & @xmath97 & @xmath98 & @xmath99 + hc@xmath2ccn & @xmath100 & @xmath101 & @xmath102 & @xmath103 + hcc@xmath2cn & @xmath104 & @xmath105 & @xmath106 & @xmath107 + lcccccc hc@xmath3n & ( @xmath108 ) @xmath109 & - & - + h@xmath2cccn & @xmath110 & @xmath111 & @xmath112 + hc@xmath2ccn & @xmath113 & @xmath114 & @xmath115 + hcc@xmath2cn & @xmath116 & @xmath117 & @xmath118 +
the abundance ratios of the three @xmath2c isotopologues are derived to be 0.9 ( @xmath4 ) : 1.00 : 1.29 ( @xmath5 ) ( @xmath6 ) and 1.0 ( @xmath7 ) : 1.00 : 1.47 ( @xmath8 ) ( @xmath6 ) for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] in l1527 and g28.28 - 0.36 , respectively .
one possible mechanism producing @xmath2c isotopic fractionation is isotope exchange reactions .
as @xcite discussed , the isotope exchange reactions can be negligible in the case of hc@xmath3n .
hence , the differences in the abundances among the three @xmath2c isotopologues should occur during their formation processes . in both l1527 and g28.28 - 0.36 ,
the abundance ratios of the three @xmath2c isotopologues of hc@xmath3n show the following two characteristics ; 1 .
the abundances of h@xmath2cccn and hc@xmath2ccn are comparable with each other , and 2 .
the abundance of hcc@xmath2cn is higher than the previous two species .
these characteristics imply that the main formation pathway of hc@xmath3n contains two equivalent carbon atoms , and the other carbon atom originates from different parent species .
we investigate possible reactions leading to hc@xmath3n , using the umist database for astrochemistry 2012 .
the four formation pathways are possible as follows .
pathway 1 : the neutral - neutral reaction between c@xmath11h@xmath11 and cn ( reaction ( [ r6 ] ) ) , pathway 2 : the neutral - neutral reaction between c@xmath11h and hnc , pathway 3 : the ion - molecule reactions between c@xmath3h@xmath119 ( @xmath120 ) and nitrogen atoms followed by electron recombination reactions , and pathway 4 : the ion - molecule reactions between c@xmath11h@xmath121 and hcn followed by electron recombination reactions .
if pathway 1 is the main formation pathway of hc@xmath3n , the abundance ratios of the three @xmath2c isotopologues of hc@xmath3n should be @xmath122 , where @xmath123 is an arbitrary value , for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] , because two carbon atoms in c@xmath11h@xmath11 are equivalent and the triple bond between c and n in cn molecule is preserved during proceeding of the reaction @xcite .
our observational results show 0.9 ( @xmath4 ) : 1.00 : 1.29 ( @xmath5 ) ( @xmath6 ) and 1.0 ( @xmath7 ) : 1.0 : 1.47 ( @xmath8 ) ( @xmath6 ) for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] in l1527 and g28.28 - 0.36 , respectively .
therefore , our observational results are well consistent with the expected ratios by pathway 1 . in l1527
, @xcite showed the formation pathways leading to c@xmath11h@xmath11 from ch@xmath13 as follows ; @xmath124 @xmath125 followed by @xmath126 and @xmath127 @xmath128 c@xmath11h@xmath11 then can be efficiently formed from ch@xmath13 , and the neutral - neutral reaction between c@xmath11h@xmath11 and cn seems to occur .
the chemical model calculation supports our conclusion derived from our observations .
@xcite showed that the neutral - neutral reaction between c@xmath11h@xmath11 and cn ( reaction ( [ r6 ] ) ) proceeds under the hot core condition , and c@xmath11h@xmath11 is released from the grain mantle inside hot cores .
our conclusion about g28.28 - 0.36 is consistent with their chemical model calculation .
the validity of their model calculation seems to be supported by detections of hc@xmath19n @xcite and hc@xmath129n ( taniguchi et al .
, _ in prep .
we next consider a possibility that pathway 2 is the dominant formation reaction leading to hc@xmath3n . in this reaction , three carbon atoms are not equivalent , and the expected abundance ratios are @xmath130 for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] , where @xmath123 , @xmath131 , and @xmath132 are arbitrary values . in fact , showed that the [ c@xmath2ch]/[@xmath2cch ] abundance ratio is @xmath133 ( @xmath134 ) .
@xcite also demonstrated their quantum chemical calculations , and their results show that a carbon atom in hnc and a carbon atom with an unpaired electron in cch are connected . from the above two results
, the abundance ratios of hc@xmath3n should be @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] = @xmath135 , where @xmath123 is an arbitrary value , if pathway 2 is the main formation pathway of hc@xmath3n .
pathway 2 can not explain our observational results , and we conclude that the reaction between c@xmath11h and hnc is not the primary formation pathway of hc@xmath3n
. pathways 3 and 4 include the ion - molecule reactions .
there is a possibility that there is no significant differences in abundances among all of the @xmath2c isotopologus , when the main formation mechanism is the ion - molecule reactions @xcite .
we also consider that scrambling may occur during the processes of these ion - molecule reactions .
the clear differences in abundances among the three @xmath2c isotopologues of hc@xmath3n should not be recognized , if scrambling occurs .
hence , these ion - molecule reactions are not the main formation mechanisms of hc@xmath3n in l1527 and g28.28 - 0.36 . in summary ,
the abundance ratios derived by our observations agree with the ratios of only pathway 1 .
we thus propose that the neutral - neutral reaction between c@xmath11h@xmath11 and cn ( reaction ( [ r6 ] ) ) dominates other formation pathways in both l1527 and g28.28 - 0.36 .
this proposal also agrees with the predicted primary formation pathway of hc@xmath3n by chemical model calculations @xcite .
the @xmath74c/@xmath2c ratios of hc@xmath3n in l1527 are summarized in table [ tab : tab3 ] .
when we assume that the excitation temperature is 9.7 k , the @xmath74c/@xmath2c ratios are consistent with the elemental ratio of @xmath136 in the local interstellar medium ( ism ) @xcite within 1-sigma errors . the derived @xmath74c/@xmath2c ratios of h@xmath2cccn and hc@xmath2ccn are slightly higher than the elemental ratio in the local ism , using the excitation temperature of 16.9 k. however , the ratios of these two @xmath2c isotopologues are consistent with the elemental ratio in the local ism within 2-sigma errors .
the @xmath74c/@xmath2c ratio of hcc@xmath2cn agrees with the elemental ratio of the local ism within 1-sigma error .
therefore , the @xmath2c species of hc@xmath3n are not significantly diluted in l1527 , as well as tmc-1 cp @xcite . in table
[ tab : tab4 ] , the @xmath74c/@xmath2c ratios of the three @xmath2c isotopologues in g28.28 - 0.36 are summarized .
these @xmath74c/@xmath2c values are lower than the elemental ratio of @xmath136 in the local ism , because g28.28 - 0.36 is located nearer the galactic center than the earth and l1527 .
the @xmath74c/@xmath2c ratio shows a gradient with galactic distance ( @xmath137 ) @xcite .
we estimated the @xmath137 of g28.28 - 0.36 to be 5.4 kpc using the trigonometry .
the @xmath74c/@xmath2c ratios at @xmath138 kpc are derived to be @xmath139 , using the results obtained by the observations of cn , co , and h@xmath11co ( @xmath74c/@xmath2c = 6.21(1.00)@xmath137 + 18.71(7.37 ) ) @xcite
. the @xmath74c/@xmath2c ratios of hc@xmath3n obtained by our observations are well consistent with the estimated values at @xmath138 kpc .
these suggest that @xmath2c isotopologues of hc@xmath3n are not heavily diluted , which is in good agreement with in the local ism .
the column densities derived by the @xmath23 lines are lower than those derived by the @xmath0 lines by @xmath140 times .
the explanation for the differences in the column densities between the @xmath23 lines and @xmath0 lines is that the spatial distribution of hc@xmath3n shows the ring - like structure with outer strong emission peaks , which is supported by our high spatial resolution map with the very large array ( taniguchi et al .
, _ in prep .
cccccccc l1527 & 0.9 ( @xmath7 ) & 1.00 & 1.29 ( @xmath141 ) & wccc & @xmath142 & @xmath143 + g28.28 - 0.36 & 1.0 ( @xmath4 ) & 1.00 & 1.47 ( @xmath144 ) & hot core & @xmath145 & @xmath146 + tmc-1 cp & 1.0 & 1.0 & 1.4 ( @xmath4 ) & dark cloud & @xmath147 & @xmath148 + serpens south 1a & 0.91 ( @xmath149 ) & 1.00 & 1.32 ( @xmath149 ) & irdc & @xmath150 & @xmath143 + we summarize @xmath2c isotopic fractionation of hc@xmath3n in the four sources , using hc@xmath2ccn as a reference , in table [ tab : tab5 ] .
we also summarize the source types and their typical temperatures and densities in table [ tab : tab5 ] .
we categorize these four sources into the star - forming cores ( l1527 and g28.28 - 0.36 ) and the starless cores ( tmc-1 cp and serpens south 1a ) .
we can also categorize these four sources into the low - mass star forming regions ( tmc-1 cp and l1527 ) and the high - mass star forming regions ( serpens south 1a and g28.28 - 0.36 ) .
we then compare various environment in star forming regions .
although there are wide ranges of temperatures ( @xmath151 ) and densities ( @xmath152 ) , the @xmath2c isotopic fractionation pattern in all of the four sources show the same tendency .
the abundances of h@xmath2cccn and hc@xmath2ccn are comparable with each other , and hcc@xmath2cn is more abundant than the others . in addition ,
the [ hcc@xmath2cn]/[hc@xmath2ccn ] ratios are in good agreement ( @xmath153 ) among the four sources within 1-sigma errors .
these results may imply that the neutral - neutral reaction between c@xmath11h@xmath11 and cn ( reaction ( [ r6 ] ) ) is the universal main formation mechanism of hc@xmath3n .
we carried out observations of the @xmath0 and @xmath1 rotational transitions of the three @xmath2c isotopologues of hc@xmath3n toward the low - mass star forming region l1527 and the high - mass star forming region g28.28 - 0.36 with the nro 45-m telescope .
the abundance ratios are found to be 0.9 ( @xmath4 ) : 1.00 : 1.29 ( @xmath5 ) ( @xmath6 ) and 1.0 ( @xmath7 ) : 1.00 : 1.47 ( @xmath8 ) ( @xmath6 ) for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] in l1527 and g28.28 - 0.36 , respectively .
our observational results suggest that the neutral - neutral reaction between c@xmath11h@xmath11 and cn seem to overwhelm the other formation pathways in both l1527 and g28.28 - 0.36 .
in addition , the @xmath2c isotopic fractionation pattern seen in the two star - forming regions is the same one in starless cores .
the primary formation pathway of hc@xmath3n may be common from low - mass prestellar cores to high - mass star - forming cores .
we would like to express our great thanks to the staff of the nobeyama radio observatory .
the nobeyama radio observatory is a branch of the national astronomical observatory of japan , national institutes of natural sciences .
in particular , we deeply appreciate dr .
tetsuhiro minamidani , ms .
chieko miyazawa , and dr .
hiroyuki nishitani for immediate responding to the equipment problems .
we express our thanks to mr .
mitsuhiro matsuo ( ph .
d student of kagoshima university ) for helping observations at nobeyama .
this work was supported in part by the center for the promotion of integrated sciences ( cpis ) of sokendai .
aikawa , y. , wakelman , v. , garrod , r. t. , & herbst , e. 2008 , , 674 , 993 chapman , j. f. , millar , t. j. , wardle , m. , burton , m. g. , & walsh , a. j. 2009 , , 394 , 221 deleon , r. l. , & muenter , j. s. 1985 , jchph , 82 , 1702 fukuzawa , k. , & osamura , y. 1997 , , 489 , 113 garrod , r. t. , & herbst , e. 2006 , , 457 , 927 green .
, c. -e . , green .
, j. a. , burton , m. g. , et al .
2014 , , 443 , 2252 hassel , g. e. , herbst , e. , & garrod , r. t. 2008 , , 681 , 1385 langer , w. d. , & penzias , a. a. 1993 , , 408 , 539 li , j. , shen , z. , wang , j. , et al .
2016 , , 824 , 136 mcelroy , d. , walsh , c. , markwick , a. j. , et al .
2013 , , 550 , a36 milam , s. n. , savage , c. , brewster , m. a. , ziurys , l. m. , & wyckoff , s. 2005 , , 634 , 1126 m@xmath154ller , h. s. p. , schl@xmath155der , f. , stutzki , j. , & winnewisser , g. 2005 , jmost , 742 , 215 purcell , c. r. , balasubramanyam , r. , burton , m. g. , et al 2006 , , 367 , 553 sakai , n. , sakai , t. , hirota , t. , & yamamoto , s. 2008 , , 672 , 371 sakai , n. , sakai , t. , hirota , t. , & yamamoto , s. 2009 , , 702 , 1025 sakai , n. , saruwatari , o. , sakai , t. , takano , s. , & yamamoto , s. 2010 , , 512 , a31 savage , c. , apponi , a. j. , ziurys , l. m. , & wyckoff , s. 2002 , , 578 , 211 suzuki , h. , yamamoto , s. , ohishi , m. , et al .
1992 , , 392 , 551 takakuwa , s. , kawaguchi , k. , mikami , h. , & saito , m. 2001 , , 53 , 251 takano , s. , masuda , a. , hirahara , y. , et al .
1998 , , 329 , 1156 taniguchi , k. , ozeki , h. , saito , m. , et al .
2016 , , 817 , 147 | we observed the @xmath0 and @xmath1 rotational lines of three @xmath2c isotopologues of hc@xmath3n in l1527 and g28.28 - 0.36 with the 45-m radio telescope of the nobeyama radio observatory in order to constrain the main formation mechanisms of hc@xmath3n in each source .
the abundance ratios of the three @xmath2c isotopologues of hc@xmath3n are found to be 0.9 ( @xmath4 ) : 1.00 : 1.29 ( @xmath5 ) ( @xmath6 ) and 1.0 ( @xmath7 ) : 1.00 : 1.47 ( @xmath8 ) ( @xmath6 ) for @xmath9h@xmath2cccn@xmath10 : [ $ ] hc@xmath2ccn@xmath10 : [ $ ] hcc@xmath2cn ] in l1527 and g28.28 - 0.36 , respectively .
we recognize a similar @xmath2c isotopic fractionation pattern that the abundances of h@xmath2cccn and hc@xmath2ccn are comparable , and hcc@xmath2cn is more abundant than the others . based on the results , we discuss the main formation pathway of hc@xmath3n .
the @xmath2c isotopic fractionation pattern derived from our observations can be explained by the neutral - neutral reaction between c@xmath11h@xmath11 and cn in both the low - mass ( l1527 ) and high - mass ( g28.28 - 0.36 ) star forming regions . |
the 21st century holds great promise for reaching a new era for unlocking the mysteries of the inner quark - gluon structure of the atomic nucleus and the nucleons inside it as governed by the theory of strong interactions ( qcd ) . the experimental program at the upgraded 12 gev jefferson laboratory @xcite , currently nearing completion , will play an important role in this quest , and revolutionize the current understanding of the dynamics of the fundamental quarks and gluons that create the wide and varied structure of hadrons and nuclei .
hall c with its heavily - shielded detector setup in a highly - focusing magnetic spectrometer with large momentum reach will be the optimal hall for certain classes of deep exclusive and semi - inclusive measurements , and in particular those requiring high quality rosenbluth ( or longitudinal - transverse - l / t ) cross section separations .
the hall c base experimental equipment consists of two magnetic spectrometers : the high momentum spectrometer ( hms ) @xcite and the super high momentum spectrometer ( shms ) @xcite . depending on the specific requirements of the experiments , they can detect either negatively or positively charged particles by choosing the magnetic field and the trigger configuration .
the hms is designed to detect secondary products of reactions in the momentum range from 0.5 to 7.3 gev/@xmath4 , while the shms momentum range extends up to about 12 gev/@xmath4 .
both spectrometers are equipped with a pair of drift chambers , pairs of timing scintillator hodoscopes for trigger formation , and various detectors for particle identification purposes , including the aerogel erenkov detector .
a schematic of the shms detector package is shown in fig .
[ shms - hut ] .
particle identification in the shms relies on time - of - flight measurements at lower momentum , an electromagnetic pre - shower and full shower calorimeter and erenkov detectors .
a erenkov detector as filled with noble gases , and the calorimeter will be used for @xmath8 separation , while a separate erenkov detector filled with heavier gases such as c@xmath9f@xmath10o will provide @xmath11 identification above 3.4 gev/@xmath4 . to complete the necessary pid capability of the shms necessary to successfully carry out the strangeness physics part of the hall c program , the shms aerogel erenkov detector descriped in this paper was built to distinguish kaons from protons with momenta above 2.6 gev/@xmath4 .
( color online ) schematic drawings of the shms aerogel detector ( left , a ) and its placement in the shms detector hut ( right , b ) .
the detector stack of the shms consists , right to left , of a gas erenkov , a pair of drift chambers , a pair of @xmath12 segmented scintillator , a gas erenkov ( gray cylinder ) , two possible slots for the aerogel detector , each of 30 cm depth , a second @xmath12 segmented scintillator / quartz bar hodoscope , and an electromagnetic calorimeter . , title="fig:",width=153 ] ( color online ) schematic drawings of the shms aerogel detector ( left , a ) and its placement in the shms detector hut ( right , b ) .
the detector stack of the shms consists , right to left , of a gas erenkov , a pair of drift chambers , a pair of @xmath12 segmented scintillator , a gas erenkov ( gray cylinder ) , two possible slots for the aerogel detector , each of 30 cm depth , a second @xmath12 segmented scintillator / quartz bar hodoscope , and an electromagnetic calorimeter .
, title="fig:",width=153 ] meson production experiments such as the `` scaling study of the l - t separated pion electroproduction cross section at 11 gev '' ( e12 - 07 - 105 ) @xcite and `` studies of the l - t separated kaon electroproduction cross section from 5 - 11 gev '' ( e12 - 09 - 011 ) @xcite , will study the potential for 3d hadron imaging studies and may also allow for form factor extractions in a kinematic regime where the signatures of qcd are quantitatively revealed @xcite .
these experiments will provide high - quality cross section data , requiring detection of scattered mesons with momenta approaching 10 gev/@xmath4 , for which the shms in hall c is exceptionally well suited .
however , additional instrumentation for @xmath13/@xmath14 particle identification is needed for the kaon measurement , which forms an essential part of this program .
the `` @xmath15 in sidis '' experiment ( e12 - 09 - 017 ) @xcite will map the transverse momentum dependence of semi - inclusive electroproduction of charged mesons from both proton and deuteron targets . in this experiment it will be beneficial to also separate kaons from protons .
the `` measurement of @xmath16=@xmath17/@xmath18 in sidis '' ( e12 - 06 - 104 ) @xcite will study the inclusive - exclusive connection in meson electroproduction , where duality has recently been shown to be valid and will be essential in understanding flavor decomposition in semi - inclusive deep inelastic scattering at 12 gev . with @xmath13/@xmath14 separation
this measurement can access the additional flavors of strange systems .
the `` pion transparency '' experiment ( e12 - 06 - 107 ) @xcite will map the transparency of the nuclear medium to pions and measure the approach to the qcd - calculable , or point - like , regime using the @xmath19 dependence of the ratio of cross sections from nucleon and nuclear targets .
kaon transparency measurements may contribute additional information to our understanding of this approach .
the general requirement for these experiments is a high detection efficiency for kaons in the shms and the capability to separate protons from kaons .
the experiments plan to run with an electron beam with energy of up to 11 gev and beam currents between a few and 80 @xmath20a hitting liquid hydrogen or deuterium target cells with lengths about 10 cm , or selected solid targets , yielding particle rates in the shms detector stack of up to 2 mhz . to take full advantage of the existing shms standard detector configuration without compromising its performance , the additional aerogel erenkov detector system was designed with the following specifications : * to have a sensitive area of at least 90@xmath2160 @xmath22 with a thickness of radiator up to 9 cm .
* to fit into a small slot of about 30 cm between the heavy gas erenkov detector and the electromagnetic calorimeter ; * to minimize the material in the particle path to keep the amount of multiple scattering and @xmath23-electrons small ; * to allow easy exchange of the aerogel material to match the required kaon momentum range ; * to provide acceptable time resolution and a high - rate capability . of the approved experiments , three ( e12 - 06 - 104 , e12 - 09 - 017 , and e12 - 07 - 105 ) only require kaon particle identification ( pid ) in the 2.7 - 5 gev/@xmath4 momentum range .
higher momenta are relevant for the kaon factorization experiment ( e12 - 09 - 011 ) , up to 7.2 gev / c , and the pion transparency measurement , where kaon identification up to this momentum would allow extraction of kaon transparencies up to @xmath24 7 ( gev/@xmath4)@xmath25 . to take into account this range of experimental requirements , covering from about 2 to above 7 gev/@xmath4 momentum , the refractive indices characteristic of aerogel material are well suited for this purpose . to maintain good @xmath26 separation over this full kinematic range
, we chose a range of aerogel materials , with nominal refractive indices of @xmath3=1.030 ( sp-30 ) , @xmath3=1.020 ( sp-20 ) , @xmath3=1.015 ( sp-15 ) , and @xmath3=1.011 ( sp-11 ) .
the threshold momenta ( in gev/@xmath4 ) for muons , pions , kaons and protons in these types of aerogel are listed in table [ tab : aerogel - thresholds ] . up to about 4 gev/@xmath4
, @xmath26 separation is achieved with aerogel with a refractive index of @xmath3=1.030 . beyond
, the lower three refractive indices provide excellent @xmath26 separation in the momentum range between 4 and @xmath77 gev/@xmath4 .
the aerogel material is characterized by a constant refractive index , and by wavelength - dependent absorption and scattering lengths , @xmath28 and @xmath29 .
the last 2 parameters were modeled in the simulation based on studies of aerogel optical properties carried out at desy @xcite . for light of wavelengths @xmath30300 nm the transmittance through the aerogel is described to good accuracy by the hunt formula @xcite which assumes constant absorption ( constant @xmath28 ) and rayleigh scattering ( @xmath31 ) . below 300 nm
@xmath28 dramatically decreases , and @xmath29 changes less rapidly .
absorption and scattering lengths of the aerogel used in the simulations include a constant absorption length for @xmath30 300 nm linearly decreasing to 1 cm at @xmath32 200 nm in accordance with ref .
@xcite . the constant value for @xmath30 300 nm was tuned using data from cosmic tests with the shms aerogel erenov detector ( see section [ cosmic_tests_detector ] ) and optimized for the optimum sensitivity of our pmts , around 450 nm .
a value of @xmath28=90 cm gives the best agreement between simulation and data for sp-30 and sp-20 aerogels .
this value is also consistent with our direct measurement of the absorption length at 450 nm as shown in fig .
[ fig : aero_spall_abscoeff ] . for the lower refractive index aerogels , which were produced by a different vendor
, a larger value seems to be more suitable , we find @xmath33220 cm .
the scale of @xmath29 is of less importance for the simulation accuracy of the detector s performance . for the purpose of the present simulations
it was derived from the average value for the hunt clarity factor @xmath34 obtained in ref .
@xcite .
reflectivity of the millipore paper for @xmath35 315 nm is taken from @xcite .
below 315 nm it is linearly extrapolated to 50% at @xmath36 = 190 nm .
the millipore reflectance is taken pure ( 100% ) lambertian .
the reflectance of gore material is taken from ref.@xcite .
the geometry of the pmt spherical window together with adjacent photocathode is coded using dimensions of the xp4500 and xp4572 pmt from the vendor @xcite .
the mean quantum efficiency of the bialcali photocathode is derived from typical spectrum of radiant sensitivity , also from the vendor . | hadronic reactions producing strange quarks such as the exclusive @xmath0 and @xmath1 reactions , or the semi - inclusive @xmath2 reaction , play an important role in studies of hadron structure and the dynamics that bind the most basic elements of nuclear physics .
the small - angle capability of the new super high momentum spectrometer ( shms ) in hall c , coupled with its high momentum reach - up to the anticipated 11-gev beam energy in hall c - and coincidence capability with the well - understood high momentum spectrometer ( hms ) , will allow for probes of such hadron structure involving strangeness down to the smallest distance scales to date . to cleanly select the kaons ,
a threshold aerogel cerenkov detector has been constructed for the shms .
the detector consists of an aerogel tray followed by a diffusion box .
four trays for aerogel of nominal refractive indices of @xmath3=1.030 , 1.020 , 1.015 and 1.011 were constructed .
the tray combination will allow for identification of kaons from 1 gev/@xmath4 up to 7.2 gev/@xmath4 , reaching @xmath5 proton and @xmath6 pion rejection , with kaon detection efficiency better than 95% .
the diffusion box of the detector is equipped with 14 five - inch diameter photomultiplier tubes .
its interior walls are covered with gore diffusive reflector , which is superior to the commonly used millipore paper and improved the detector performance by 35% .
the inner surface of the two aerogel trays with higher refractive index is covered with millipore paper , however , those two trays with lower aerogel refractive index are again covered with gore diffusive reflector for higher performance .
the measured mean number of photoelectrons in saturation is @xmath712 for @xmath3=1.030 , @xmath78 for @xmath3=1.020 , @xmath710 for @xmath3=1.015 , and @xmath75.5 for @xmath3=1.011 .
the design details , the results of component characterization , and initial performance tests and optimization of the detector are presented . |
after the great success of elliptic curves in public - key cryptography , researchers have naturally been drawn to their higher - dimensional generalizations : jacobians of higher - genus curves .
curves of genus @xmath5 ( elliptic curves ) , @xmath2 , and @xmath0 are widely believed to offer the best balance of security and efficiency .
this article is concerned with the security of curves of genus @xmath0 .
there are two classes of curves of genus @xmath0 : hyperelliptic and non - hyperelliptic .
each class has a distinct geometry : the canonical morphism of a hyperelliptic curve is a double cover of a curve of genus @xmath6 , while the canonical morphism of a non - hyperelliptic curve of genus @xmath0 is a birational map to a nonsingular plane quartic curve
. a hyperelliptic curve can not be isomorphic ( or birational ) to a non - hyperelliptic curve . from a cryptological point of view
, the discrete logarithm problem ( dlp ) in jacobians of hyperelliptic curves of genus @xmath0 over @xmath7 may be solved in @xmath8 group operations , using the index calculus algorithm of gaudry , thom , thriault , and diem @xcite .
jacobians of non - hyperelliptic curves of genus @xmath0 over @xmath7 are amenable to diem s index calculus algorithm @xcite , which requires only @xmath9 group operations to solve the dlp ( for comparison , pollard / baby - step - giant - step methods require @xmath10 group operations to solve the dlp in jacobians of genus @xmath0 curves over @xmath7 ) .
the security of non - hyperelliptic genus @xmath0 curves is therefore widely held to be lower than that of their hyperelliptic cousins .
our aim is to construct explicit homomorphisms to provide a means of efficiently translating instances of the dlp from jacobians of hyperelliptic curves of genus @xmath0 to jacobians of non - hyperelliptic curves , where faster index calculus is available . in the context of dlp - based cryptography
, we may assume that our jacobians are absolutely simple .
in this situation , every nontrivial homomorphism of jacobians of curves of genus @xmath0 is an _ isogeny _ : that is , a surjective homomorphism with finite kernel . to be specific ,
suppose we are given a hyperelliptic curve @xmath11 of genus @xmath0 over a finite field @xmath7 , together with an instance @xmath12q$ ] of the dlp in @xmath13 ; our task is to recover @xmath14 given @xmath15 and @xmath16 .
after applying the standard pohlig hellman reduction @xcite , we may assume that @xmath15 and @xmath16 have prime order .
we want to solve this dlp instance by solving an equivalent dlp instance in a non - hyperelliptic jacobian .
suppose we have an isogeny @xmath17 , where @xmath18 is a non - hyperelliptic curve of genus @xmath0 .
further , suppose that @xmath19 is explicit ( that is , we have equations for @xmath18 and an efficient map on divisor classes representing @xmath19 ) and defined over @xmath7 , so it maps @xmath13 into @xmath20 .
provided @xmath21 , we can recover @xmath14 by solving the dlp instance @xmath22\phi(q)$ ] in @xmath20 with diem s algorithm .
the approach outlined above is conceptually straightforward ; the difficulty lies in computing explicit isogenies of jacobians of genus @xmath0 curves .
automorphisms , integer multiplications , and frobenius maps aside , we know of no explicit and general formulae for isogenies from jacobians of hyperelliptic curves of genus @xmath0 apart from those presented below . in [ section : kernel ] through [ section : equations - for - the - isogeny ]
, we derive explicit formulae for isogenies whose kernels are generated by differences of weierstrass points , following the construction of donagi and livn @xcite .
the key step is making recillas trigonal construction @xcite completely explicit .
this gives us a curve @xmath23 of genus @xmath0 and an explicit isogeny @xmath24 .
while @xmath23 may be hyperelliptic , nave moduli space dimension arguments suggest ( and experience confirms ) that @xmath23 will be non - hyperelliptic with an overwhelming probability , and thus explicitly isomorphic to a nonsingular plane quartic curve @xmath18 .
we can therefore compute an explicit isogeny @xmath25 ; if @xmath19 is defined over @xmath7 , then we can use it to reduce dlp instances .
we note that the trigonal construction ( and hence our formulae ) does not apply in characteristics @xmath2 and @xmath0 .
we show in [ section : probabilities ] that , subject to some reasonable assumptions , given a uniformly randomly chosen hyperelliptic curve @xmath11 of genus @xmath0 over a sufficiently large finite field @xmath7 of characteristic at least @xmath26 , our algorithms succeed in constructing an explicit isogeny defined over @xmath7 from @xmath27 to a non - hyperelliptic jacobian with probability @xmath28 . in particular , instances of the dlp can be solved in @xmath9 group operations for around @xmath4 of all jacobians of hyperelliptic curves of genus @xmath0 over finite fields of characteristic at least @xmath26 .
we discuss more general isogenies in [ section : other - isogenies ] . given explicit formulae for these isogenies , we expect that most , if not all , instances of the dlp in jacobians of hyperelliptic curves of genus @xmath0 over any finite field could be reduced to instances of the dlp in non - hyperelliptic jacobians .
our results have a number of interesting implications for curve - based cryptography , at least for curves of genus @xmath0 .
first , the difficulty of the dlp in a subgroup @xmath29 of @xmath27 depends not only on the size of the subgroup @xmath29 , but upon the existence of other rational subgroups of @xmath27 that can be used to form quotients .
second , the security of a given hyperelliptic genus @xmath0 curve depends significantly upon the factorization of its hyperelliptic polynomial .
neither of these results has any parallel in genus @xmath5 or @xmath2 .
the constructions of [ section : kernel ] through [ section : equations - for - the - isogeny ] and [ section : other - isogenies ] require some nontrivial algebraic geometry .
we have included enough mathematical detail here to enable the reader to compute examples , to justify our claim that the construction is efficient , and to support our heuristics .
this article presents an extended version of work that appeared in the proceedings of the eurocrypt 2008 conference @xcite .
the chief results are the same ; we have made some ( minor ) changes to our notation , expanded the derivation in [ section : equations - for - the - isogeny ] , given further details and proofs throughout , and added an appendix with algorithms to compute sets of tractable subgroups .
we will work over @xmath7 throughout this article , and [ section : equations - for - the - isogeny ] are valid over fields of characteristic not @xmath2 or @xmath0 .
] where @xmath30 is a power of a prime @xmath3 .
we let @xmath31 denote the galois group @xmath32 , which is ( topologically ) generated by the @xmath33 power frobenius map .
suppose we are given a hyperelliptic curve @xmath11 of genus @xmath0 over @xmath7 .
we will use both an affine model @xmath34 where @xmath35 is a squarefree polynomial of degree @xmath36 or @xmath37 , and a weighted projective plane model @xmath38 for @xmath11 ( here @xmath39 , @xmath40 , and @xmath41 have weights @xmath5 , @xmath5 , and @xmath42 , respectively ) .
the coordinates of these models are related by @xmath43 and @xmath44 .
the polynomial @xmath45 is squarefree of total degree @xmath37 , with @xmath46 and @xmath47 .
we emphasize that @xmath35 need not be monic . by a _
randomly chosen hyperelliptic curve _
, we mean the hyperelliptic curve defined by @xmath48 , where @xmath45 is a uniformly randomly chosen squarefree homogenous bivariate polynomial of degree @xmath37 over @xmath7
. the canonical _ hyperelliptic involution _
@xmath49 of @xmath11 is defined by @xmath50 in the affine model , @xmath51 in the projective model , and induces the negation map @xmath52 $ ] on @xmath27 .
the quotient @xmath53 sends @xmath54 to @xmath55 in the projective model , and @xmath56 to @xmath57 in the affine model ( where it maps onto the affine patch of @xmath58 where @xmath59 ) . to compute in @xmath27
, we fix an isomorphism from @xmath27 to the group of degree-@xmath6 divisor classes on @xmath11 , denoted @xmath60 . recall that divisors are formal sums of points in @xmath61 , and if @xmath62 is a divisor , then @xmath63 is the _ degree _ of @xmath64 .
we say @xmath64 is _ principal _ if @xmath65 for some function @xmath66 on @xmath11 , where @xmath67 denotes the number of zeroes ( or the negative of the number of poles ) of @xmath66 at @xmath15 . since @xmath11 is complete , every principal divisor has degree @xmath6 .
the group @xmath60 is defined to be the group of divisors of degree @xmath6 modulo principal divisors ; the equivalence class of a divisor @xmath64 is denoted by @xmath68 $ ] .
we let @xmath69 $ ] denote the @xmath70-torsion subgroup of @xmath27 : that is , the kernel of the multiplication - by-@xmath70 map .
if @xmath70 is prime to @xmath30 , then @xmath69({{\overline{\mathbb{f}}}}_{q})$ ] is isomorphic to @xmath71 .
the eight points of @xmath72 where @xmath73 are called the _ weierstrass points _ of @xmath11 .
each weierstrass point @xmath74 corresponds to a linear factor @xmath75 of @xmath45 , which is defined up to scalar multiples .
if @xmath76 and @xmath77 are weierstrass points , then @xmath78 , so @xmath79 = 0 $ ] ; hence @xmath80 $ ] represents an element of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] .
in particular , @xmath80 = [ ( w_2 ) - ( w_1)]$ ] , so the divisor class @xmath80 $ ] corresponds to the pair @xmath82 of weierstrass points , and hence to the quadratic factor @xmath83 of @xmath45 ( up to scalar multiples ) .
[ proposition : partitions - and - tractable - subgroups ] to every @xmath31-stable partition of the eight weierstrass points of @xmath11 into four disjoint pairs , we may associate an @xmath7-rational subgroup of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] isomorphic to @xmath1 .
let @xmath84 be a partition of the set of weierstrass points of @xmath11 into four disjoint pairs .
each pair @xmath85 corresponds to the @xmath2-torsion divisor class @xmath86 $ ] in @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] .
we associate the subgroup @xmath87 : 1 \le i \le 4}\right\rangle}}$ ] to the partition .
observe that @xmath88 = \big[\mathrm{div}\big(w/\prod_{i=1}^4 l_{w_i''}\big)\big ] = 0 ; \ ] ] this is the only relation on the classes @xmath86 $ ] , so @xmath89 .
the action of @xmath31 on @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] corresponds to its action on the weierstrass points , so if the partition is @xmath31-stable , then the subgroup @xmath90 is @xmath31-stable . by `` an @xmath7-rational subgroup of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] isomorphic to @xmath1 '' , we mean a @xmath31-stable subgroup that is isomorphic to @xmath1 over @xmath91 .
we emphasize that the subgroup need _ not _ be contained in @xmath13 . requiring the pairs of weierstrass points in proposition [ proposition : partitions - and - tractable - subgroups ] to be disjoint ensures that the associated subgroup is isotropic with respect to the @xmath2-weil pairing .
we will see in [ section : other - isogenies ] that this is necessary for the quotient by the subgroup to be an isogeny of principally polarized abelian varieties , and hence for the quotient to be an isogeny of jacobians .
we call the subgroups corresponding to partitions of the weierstrass points of @xmath11 as in proposition [ proposition : partitions - and - tractable - subgroups ] _
tractable subgroups_. we let @xmath92 denote the set of all @xmath7-rational tractable subgroups of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] . not every subgroup of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] that is the kernel of an isogeny of jacobians is a tractable subgroup .
for example , if @xmath93 are the weierstrass points of @xmath11 , then the subgroup @xmath94 : ( i , j , k ) \in \ { ( 2,3,4 ) , ( 2,5,6 ) , ( 3,5,7 ) \ } } \right\rangle}}\ ] ] is a maximal @xmath2-weil isotropic subgroup of @xmath95 , and hence is the kernel of an isogeny of jacobians ( see [ section : other - isogenies ] ) .
however , this subgroup contains no nontrivial differences of weierstrass points , and therefore can not be a tractable subgroup . computing @xmath92 is straightforward if we identify each tractable subgroup with its corresponding partition of weierstrass points .
recall that each pair of weierstrass points @xmath96 corresponds to a quadratic factor of @xmath45 ( up to scalar multiples ) . since the pairs are disjoint , the corresponding quadratic factors are pairwise coprime , so we may take them to form a factorization of @xmath45 .
we therefore have a correspondence of tractable subgroups , partitions of weierstrass points into pairs , and sets of quadratic polynomials ( up to scalar multiples ) : @xmath97 the action of @xmath31 on @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] corresponds to its action on the set of weierstrass points , so the action of @xmath31 on a tractable subgroup @xmath90 corresponds to the action of @xmath31 on the corresponding set @xmath98 ( assuming the @xmath99 have been scaled appropriately ) . in particular , @xmath90 is @xmath7-rational precisely when @xmath98 is fixed by @xmath31 .
the factors @xmath99 are themselves defined over @xmath7 precisely when the corresponding points of @xmath90 are @xmath7-rational .
we can use this information to compute @xmath92 .
the set of pairs of weierstrass points contains a @xmath31-orbit @xmath100 if and only if ( possibly after exchanging some of the @xmath101 with the @xmath102 ) either both @xmath103 and @xmath104 are @xmath31-orbits or @xmath105 is a @xmath31-orbit .
every @xmath31-orbit of weierstrass points corresponds to an @xmath7-irreducible factor of @xmath35 , so the size of @xmath92 depends only on the factorization of @xmath35 . a table relating the size of @xmath92 to the factorization of @xmath45 appears in lemma [ lemma : number - of - tractable - subgroups ] below ; this will be useful for our analysis in [ section : probabilities ] .
for completeness , we have included a nave algorithm for enumerating @xmath92 in appendix [ appendix : computing - sh ] .
[ lemma : number - of - tractable - subgroups ] let @xmath106 be a hyperelliptic curve of genus @xmath0 over @xmath7 .
the cardinality of the set @xmath92 depends only on the degrees of the @xmath7-irreducible factors of @xmath45 , and is described by the following table : [ cols="^,^",options="header " , ] this is a routine combinatorial exercise after noting that every @xmath31-orbit of pairs of weierstrass points corresponds to either an even - degree factor of @xmath35 , or a pair of factors of @xmath35 of the same degree .
we will now briefly outline the theoretical aspects of constructing isogenies with tractable kernels . we will make the construction completely explicit in [ section : computing - trigonal - maps ] and [ section : equations - for - the - isogeny ] .
suppose @xmath107 : 1 \le i \le 4}\right\rangle}}$ ] is a tractable subgroup .
we say that a morphism @xmath108 is a _
trigonal map for @xmath90 _ if @xmath109 has degree @xmath0 and @xmath110 for @xmath111 .
given a trigonal map @xmath109 for some tractable subgroup @xmath90 , recillas trigonal construction @xcite specifies a curve @xmath23 of genus @xmath0 and a map @xmath112 of degree @xmath42 .
is an tale double cover ; the trigonal construction we apply here is in fact the flat limit of recillas construction ( see @xcite for details ) . ]
the isomorphism class of @xmath23 depends only on @xmath90 , and is independent of the choice of @xmath109 ( see recillas @xcite , donagi ( * ? ? ?
2.11 ) , and remark [ remark : trigonal - map - descent ] below ) .
theorem [ theorem : isogeny - from - trigonal - construction ] , due to donagi and livn , states that if @xmath109 is a trigonal map for @xmath90 , then @xmath90 is the kernel of an isogeny from @xmath27 to @xmath113 .
[ theorem : isogeny - from - trigonal - construction ] let @xmath90 be a tractable subgroup in @xmath92 , and let @xmath108 be a trigonal map for @xmath90 . if @xmath23 is the curve formed from @xmath109 by recillas trigonal construction , then there is an isogeny @xmath114 ( defined over @xmath91 ) with kernel @xmath90 .
we will give only a brief description of the geometry of @xmath23 here , concentrating instead on its explicit construction ; we refer the reader to recillas @xcite , vakil @xcite , donagi @xcite , and birkenhake and lange @xcite for proofs and further detail . the isogeny of theorem [ theorem : isogeny - from - trigonal - construction ] is analogous to the well - known richelot isogeny in genus @xmath2 ( see bost and mestre @xcite , and donagi and livn @xcite for details ) , and to the explicit isogeny described by lehavi and ritzenthaler in @xcite for jacobians of non - hyperelliptic genus @xmath0 curves . in abstract terms ,
if @xmath115 is the subset of the codomain of @xmath109 above which @xmath116 is unramified , then @xmath23 is by definition the closure of the curve over @xmath115 representing the pushforward to @xmath115 of the sheaf of sections of @xmath117 ( in the tale topology ) .
this means in particular that the @xmath91-points of @xmath23 over an @xmath91-point @xmath15 of @xmath115 represent partitions of the six @xmath91-points of @xmath118 into two sets of three exchanged by the hyperelliptic involution .
the fibre product of @xmath11 and @xmath23 over @xmath58 with respect to @xmath116 and @xmath66 is the union of two isomorphic curves , @xmath119 and @xmath120 , which are exchanged by the involution on @xmath121 induced by the hyperelliptic involution .
the natural projections induce coverings @xmath122 and @xmath123 of degrees @xmath2 and @xmath0 , respectively , so @xmath119 is a @xmath124-correspondence between @xmath11 and @xmath23 .
the maps @xmath125 and @xmath126 induce homomorphisms @xmath127 ( the _ pullback _ ) and @xmath128 ( the _ pushforward _ ) . in terms of divisor classes , the pullback is defined by @xmath129\big ) = \big [ \sum_{p \in h}n_p\!\!\!\!{\sum_{q
\in \pi_h^{-1}(p)}}\!\!\!\!(q ) \big],\ ] ] with appropriate multiplicities where @xmath125 ramifies ; the pushforward is defined by @xmath130\big ) = \big [ \sum_{q\in r}m_q(\pi_x(q ) ) \big ] .\ ] ] composing @xmath131 with @xmath132 , we obtain an isogeny @xmath114 with kernel @xmath90 .
if we replace @xmath119 with @xmath120 in the above , we obtain an isogeny isomorphic to @xmath133 .
thus , up to isomorphism , the construction of the isogeny depends only on the subgroup @xmath90 . the curves and jacobians described above form the commutative diagrams shown in figure [ table : curve - diagram ] .
@xmath134 & r&&&{\ensuremath{{j}_{r}}}\\ h&&x&{\ensuremath{{j}_{h}}}&&{\ensuremath{{j}_{x}}}\\ { \ensuremath{\mathbb{p}^{1}_{{}}}}\\ & { \ensuremath{\mathbb{p}^{1 } _ { { } } } } \ncline[nodesep=3pt]{->}{1,2}{2,1}^{\pi_h}_2 \ncline[nodesep=3pt]{->}{1,2}{2,3}^{\pi_x}_3 \ncline[nodesep=3pt]{->}{2,1}{3,1}<{\pi}>2 \ncline[nodesep=3pt]{->}{3,1}{4,2}^{3}_g \ncline[nodesep=3pt]{->}{2,3}{4,2}^{4}_f \ncline[nodesep=3pt]{->}{2,4}{2,6}_{\phi } \ncline[nodesep=3pt]{->}{2,4}{1,5}<{\pi_h^ * } \ncline[nodesep=3pt]{->}{1,5}{2,6}>{(\pi_x ) _ * } \end{psmatrix}\ ] ] the hyperelliptic jacobians form a codimension-@xmath5 subspace @xmath135 of the moduli space of @xmath0-dimensional principally polarized abelian varieties which , by the theorem of oort and ueno @xcite , is also the moduli space @xmath136 of jacobians of genus @xmath0 curves .
the weil hypotheses imply that @xmath137 for sufficiently large @xmath30 ( cf .
* theorem 1 ) ) .
in particular , for cryptographically relevant sizes of @xmath30 , the probability that a uniformly randomly chosen curve @xmath23 of genus @xmath0 over @xmath7 should be hyperelliptic is negligible .
we will suppose that the same is true for the curve @xmath23 constructed in theorem [ theorem : isogeny - from - trigonal - construction ] for a uniformly randomly chosen @xmath11 and @xmath90 in @xmath92 .
this is consistent with our experimental observations , so we postulate hypothesis [ hypothesis : hyperellipticity ] . [ hypothesis : hyperellipticity ] the probability that the curve @xmath23 constructed by the trigonal construction for a randomly chosen @xmath138 and @xmath90 in @xmath92 is hyperelliptic is negligible for sufficiently large @xmath30 .
suppose we are given a tractable subgroup @xmath90 of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] , corresponding to a partition @xmath139 of the weierstrass points of @xmath11 into pairs .
the first step in the explicit trigonal construction is to compute a trigonal map @xmath109 for @xmath90 .
we will compute polynomials @xmath140 and @xmath141 such that the rational map @xmath142 defines a trigonal map for @xmath90 .
the derivation is an exercise in classical geometry ; we include it here to demonstrate its efficiency and to justify hypothesis [ hypothesis : trigonal - map - rationality ] , which will be important in determining the expectation of success of our reduction in [ section : probabilities ] .
the reader prepared to admit the existence of efficiently computable trigonal maps in the form of may skip the remainder of this section on first reading . by definition
, @xmath108 is a degree-@xmath0 map with @xmath143 for @xmath111
. we will express @xmath109 as a composition @xmath144 , where @xmath145 is the rational normal embedding defined by @xmath146 and @xmath147 is the projection defined as follows .
for each @xmath111 , we let @xmath148 denote the line in @xmath149 passing through @xmath150 and @xmath151
. there exists at least one line @xmath152 intersecting all four of the @xmath148 ( in fact there are two , though they may coincide ; we will compute them below ) .
we take @xmath153 to be the projection away from @xmath152 ; then @xmath154 for @xmath111 , so @xmath155 is a trigonal map for @xmath90 .
given linear equations for @xmath152 in the coordinates @xmath156 , we can use gaussian elimination to compute elements @xmath157 , @xmath158 , @xmath159 , and @xmath160 of @xmath7 such that @xmath161 the projection @xmath147 away from @xmath152 is then defined by @xmath162 so our trigonal map @xmath155 is defined by @xmath163 therefore , if we set @xmath164 and @xmath165 , then @xmath109 will be defined by the rational map @xmath166 . to compute equations for @xmath152
, we will use the classical theory of _
grassmannian varieties_. the elementary lemmas [ lemma : grassmannian - correspondence ] and [ lemma : hyperplane ] will be stated without proof ; we refer the reader to griffiths and harris @xcite and harris ( * ? ? ?
* lecture 6 ) for details .
the set of lines in @xmath149 has the structure of an algebraic variety @xmath167 , called the grassmannian .
there is a convenient model for @xmath167 as a quadric hypersurface in @xmath168 : if @xmath169 are coordinates on @xmath168 , then we may take @xmath170 [ lemma : grassmannian - correspondence ] there is a bijection between points of @xmath171 and lines in @xmath149 , defined as follows . 1
. the point of @xmath171 corresponding to the line through @xmath172 and @xmath173 in @xmath149 has coordinates @xmath174 2 .
the line in @xmath149 corresponding to a point @xmath175 of @xmath171 is defined by @xmath176 ( two of the equations will be redundant linear combinations of the others ) .
[ lemma : hyperplane ] let @xmath152 be the line in @xmath149 corresponding to a point @xmath175 of @xmath171
. the points in @xmath171 corresponding to lines in @xmath149 that intersect nontrivially with @xmath152 are precisely the points lying in the hyperplane defined by @xmath177 ( where the subscripts are taken modulo @xmath178 ) .
suppose @xmath90 is represented by a set @xmath179 of quadratic factors of @xmath45 ( as in [ section : kernel ] ) , with each factor @xmath99 corresponding to a pair @xmath180 of weierstrass points .
applying lemma [ lemma : grassmannian - correspondence ] , we see that the line @xmath148 through @xmath150 and @xmath151 corresponds to the point @xmath181 on @xmath167 .
if @xmath175 in @xmath171 corresponds to a candidate for @xmath152 , then by lemma [ lemma : hyperplane ] we have @xmath182 , where @xmath183 the kernel of @xmath184 is two - dimensional , corresponding to a line @xmath185 in @xmath168 . the kernel is independent of the ordering of the @xmath99 , and does not change if we replace the @xmath99 by scalar multiples ; hence , @xmath185 depends only on the subgroup @xmath90 .
let @xmath186 be a basis for @xmath187 , writing @xmath188 and @xmath189 .
if @xmath90 is @xmath7-rational , then so is @xmath187 , so we may take the @xmath190 and @xmath191 to be in @xmath7 ( see cartier @xcite ) .
we want to find a point @xmath192 where @xmath185 intersects with @xmath167 .
the points @xmath193 on the line @xmath152 in @xmath149 corresponding to @xmath194 satisfy @xmath195 , where @xmath196 by part ( 2 ) of lemma [ lemma : grassmannian - correspondence ] , the rank of @xmath197 is @xmath2 . using the expression @xmath198 ( where the subscripts are taken modulo @xmath178 ) , we see that @xmath197 has rank @xmath2 precisely when @xmath199 : we can therefore solve @xmath199 to determine a value for @xmath200
. finally , we use gaussian elimination to compute @xmath157 , @xmath158 , @xmath159 , and @xmath160 in @xmath201 such that @xmath202 and @xmath203 generate the rowspace of @xmath204 .
we then take @xmath205 , and compute @xmath153 , @xmath206 , and the trigonal map @xmath155 as above .
since @xmath152 is defined over @xmath201 , so is the projection @xmath153 and the trigonal map @xmath109 .
but @xmath200 satisfies a quadratic equation with coefficients in @xmath7 , so @xmath201 is at most a quadratic extension of @xmath7 .
computing the discriminant of @xmath207 , we obtain a criterion for existence of trigonal maps over @xmath7 for a given tractable subgroup .
[ proposition : trigonal - map - rationality - criterion ] suppose @xmath90 is a tractable subgroup , and let @xmath208 be any @xmath7-rational basis of the nullspace of the matrix @xmath184 defined in .
there exists an @xmath7-rational trigonal map for @xmath90 if and only if @xmath209 is a square in @xmath7 , where the subscripts are taken modulo @xmath178 .
from the derivation above , we see that there exists an @xmath7-rational trigonal map for @xmath90 if and only if we can find a @xmath200 in @xmath7 such that @xmath210 . by equation
, we can find such a @xmath200 if and only if the quadratic polynomial @xmath211 has two roots in @xmath7 .
this occurs precisely when the discriminant of this polynomial the expression in above is a square in @xmath7 .
proposition [ proposition : trigonal - map - rationality - criterion ] shows that the rationality of a trigonal map for a tractable subgroup @xmath90 depends only upon whether an element of @xmath7 depending only on @xmath90 is a square .
it seems reasonable to assume that these field elements are uniformly distributed for uniformly random choices of @xmath11 and @xmath90 , and indeed this is consistent with our experimental observations . since a uniformly randomly chosen element of @xmath7 is a square with probability @xmath212 , we propose hypothesis [ hypothesis : trigonal - map - rationality ] . [
hypothesis : trigonal - map - rationality ] the probability that there exists an @xmath7-rational trigonal map for a subgroup @xmath90 uniformly randomly chosen from @xmath92 , where @xmath11 is a randomly chosen hyperelliptic curve over @xmath7 , is @xmath213 .
suppose we have a hyperelliptic curve @xmath11 of genus @xmath0 , a tractable subgroup @xmath90 in @xmath92 , and a trigonal map @xmath109 for @xmath90 .
we will now perform an explicit trigonal construction on @xmath109 to compute a curve @xmath23 and an isogeny @xmath114 with kernel @xmath90 .
we assume that @xmath109 has been derived as in [ section : computing - trigonal - maps ] , and in particular that @xmath214 is defined by a rational map in the form @xmath215 observe that @xmath109 maps the point at infinity to the point at infinity ( that is , @xmath216 ) .
for notational convenience , we define @xmath217 unless otherwise noted , we will view @xmath218 as an element of @xmath219[x]$ ] .
we have @xmath220 we also define @xmath221 , @xmath222 , and @xmath223 to be the elements of @xmath219 $ ] such that @xmath224 let @xmath115 be the subset of @xmath225 above which @xmath116 is unramified . with the notation above , @xmath226)\setminus{\ensuremath{v\!\left ( { ( f_1 ^ 2 - 4f_2f_0)(4g_2 ^ 3g_0 - g_2 ^ 2g_1 ^ 2 - 18g_2g_1g_0 + 4g_1 ^ 3 + 27g_0 ^ 2 ) } \right ) } } .\ ] ]
we will derive equations for an affine model @xmath227 of @xmath228 that is , the open subset of @xmath23 over @xmath115 .
we will not prove here that the normalization of @xmath227 is isomorphic to the curve @xmath23 specified by recillas , but we will exhibit a bijection on geometric points . if @xmath23 is not hyperelliptic , then taking the canonical map of @xmath227 into @xmath229 will give us a nonsingular plane quartic curve @xmath18 isomorphic to @xmath23 . by definition ,
every point @xmath15 in @xmath230 corresponds to a pair of unordered triples of points in @xmath72 , exchanged by the hyperelliptic involution , with each triple supported on the fibre of @xmath116 over @xmath231 . to be more explicit ,
suppose @xmath16 is a generic point of @xmath115 .
since @xmath116 is unramified above @xmath16 , we may choose three preimages @xmath232 , @xmath233 , and @xmath234 of @xmath16 such that @xmath235 viewing unordered triples of points as effective divisors of degree @xmath0 ( that is , as formal sums of three points ) , we have @xmath236 note that @xmath237 and @xmath238 never appear in the same divisor for any @xmath239 .
there is a one - to - one correspondence between effective divisors of degree @xmath0 on @xmath11 satisfying this condition , and ideals @xmath240 where @xmath241 is a monic cubic polynomial and @xmath242 is a quadratic polynomial satisfying @xmath243 ( this is the well - known _ mumford representation _ ( * ? ? ? * iiia ) ) .
for example , @xmath244 corresponds to the ideal @xmath245 where @xmath246 and @xmath242 satisfies @xmath247 for @xmath239 ( with appropriate multiplicities ) ; we may compute @xmath242 using the lagrange interpolation formula .
a divisor is defined over @xmath248 if and only if @xmath241 and @xmath242 are defined over @xmath248 .
the ideal @xmath245 corresponds to @xmath249 if and only if @xmath250 corresponds to @xmath251 ; so each point of @xmath23 over @xmath115 corresponds to a pair @xmath252 of ideals .
we will construct a curve parametrizing these pairs of ideals , and take this as a model for @xmath227 .
suppose @xmath253 is a pair of ideals corresponding to one of the preimages of @xmath16 on @xmath227 .
the product of the two ideals is equal to the principal ideal @xmath254 ; but products of ideals correspond to sums of divisors , so @xmath254 must cut out the divisor @xmath255 on @xmath11 .
this divisor is just @xmath256 , which we know is cut out by @xmath257 ; so we conclude that @xmath258 for every pair of ideals @xmath253 corresponding to a point in @xmath259 . in particular , the generic point of @xmath227 corresponds to a pair of ideals of the form @xmath260 , where @xmath261 , @xmath262 , and @xmath263 are algebraic functions of @xmath264 such that @xmath265 viewing @xmath261 , @xmath262 , and @xmath263 as coordinates on @xmath266 ( over @xmath7 ) , we expand both sides of modulo @xmath218 and equate coefficients to obtain a variety @xmath267 in @xmath268 parametrizing ideals : @xmath269 where @xmath270 the ideals in each pair @xmath271 are exchanged by the involution @xmath272 defined by @xmath273 the curve @xmath227 is therefore the quotient of @xmath267 by @xmath274 . to make this quotient explicit ,
let @xmath275 be the map defined by @xmath276 observe that @xmath277 we have @xmath278 , so @xmath279 where @xmath280 , @xmath281 , and @xmath282 are the polynomials defined by @xmath283 observe that @xmath227 is defined over the field of definition of @xmath109 .
it remains to derive a correspondence @xmath119 between @xmath11 and @xmath227 inducing the isogeny @xmath19 .
we know that @xmath119 is a component of the fibre product @xmath284 ( with respect to @xmath285 and @xmath66 ) .
we may realise the open affine subset @xmath286 as the subvariety @xmath287 of @xmath288 ; decomposing the ideal @xmath289 will therefore give us a model for @xmath119 .
[ lemma : square - root - of - s ] let @xmath290 be the polynomial in @xmath219 $ ] defined by @xmath291{l } f_0 ^ 3 - f_0 ^ 2f_1g_2 - 2f_0 ^ 2f_2g_1 + f_0 ^ 2f_2g_2 ^ 2 + f_0f_1 ^ 2g_1 + 3f_0f_1f_2g_0 - f_0f_1f_2g_1g_2 \\ { } - 2f_0f_2 ^ 2g_0g_2 + f_0f_2 ^ 2g_1 ^ 2 - f_1 ^ 3g_0
+ f_1 ^ 2f_2g_0g_2 - f_1f_2 ^ 2g_0g_1 + f_2 ^ 3g_0 ^ 2 , \end{array}\ ] ] and let @xmath292 be its leading coefficient .
then @xmath290 has a square root in @xmath293 $ ] .
the polynomial @xmath290 is a square in @xmath293 $ ] if and only if each of its roots in @xmath91 occur with multiplicity @xmath2 . in the notation of
, we have @xmath294 so @xmath295 if and only if @xmath296 for some @xmath239 that is , if and only if at least one of the @xmath237 is a weierstrass point of @xmath11 .
but the trigonal map @xmath109 was constructed precisely so that the weierstrass points of @xmath11 appear in pairs in the fibres of @xmath109 : hence exactly two of the @xmath237 must be weierstrass points , and so @xmath297 and @xmath295 with multiplicity @xmath2 . [
proposition : b22-t - equation ] let @xmath290 be the polynomial of lemma [ lemma : square - root - of - s ] , and let @xmath298 , @xmath299 , @xmath300 , and @xmath301 be the polynomials in @xmath302 $ ] defined by @xmath303 on the curve @xmath227 , we have @xmath304 consider again the fibre of @xmath112 over the generic point @xmath305 of @xmath115 ( as in ) .
if @xmath306 is a pair of divisors corresponding to one of the points in the fibre , then by the lagrange interpolation formula the value of @xmath307 at the corresponding point of @xmath267 is @xmath308 where the sum is taken over the cyclic permutations @xmath309 of @xmath310 .
after interpolating for each pair of divisors in the fibre , an elementary but involved symbolic calculation shows that @xmath307 satisfies @xmath311 where @xmath312 with @xmath313 for each cyclic permutation @xmath309 of @xmath310 , and where @xmath314 .
now @xmath315 , @xmath316 , @xmath317 , and @xmath318 are symmetric functions with respect to permutations of the points in the fibre @xmath319 .
they are therefore polynomials in the homogeneous elementary symmetric functions @xmath320 which are polynomials in @xmath264 .
indeed , the @xmath321 are given by the coefficients of @xmath218 : @xmath322 expressing @xmath315 , @xmath316 , @xmath317 , and @xmath318 in terms of @xmath221 , @xmath222 , @xmath223 , @xmath323 , @xmath324 , and @xmath325 , and substituting the resulting expressions into , we obtain .
equation gives us a ( singular ) affine plane model for @xmath23 .
we can also use to compute a square root for @xmath307 on @xmath227 : we have @xmath326 returning to , we observe that @xmath307 is a unit on @xmath227 , since its zeroes and poles occur only at points @xmath16 where @xmath116 is ramified over @xmath327 , and these points were excluded from @xmath115 . since @xmath328 is the square root of @xmath307 , it must also be a unit on @xmath227 . given a point @xmath329 of @xmath227 , the corresponding pair of divisors of degree @xmath0 on @xmath11 is cut out by the pair of ideals @xmath330 this is precisely the decomposition of @xmath289 that we need to compute the correspondence from @xmath331 to @xmath227 : we have @xmath332 , where @xmath333 and @xmath334 on the level of divisor classes , the isogeny @xmath114 is made explicit by the map @xmath335 where @xmath122 and @xmath336 are the natural projections defined by @xmath337 and @xmath338 , respectively . in terms of ideals cutting out effective divisors , @xmath19 is realized by the map @xmath339 .\ ] ] taking @xmath120 in place of @xmath119 in the above gives an isogeny equal to @xmath133 .
it remains to determine the field of definition of @xmath19 .
[ proposition : x - rationality - criterion ] if @xmath90 is a subgroup in @xmath92 with an @xmath7-rational trigonal map @xmath109 defined over @xmath7 , and @xmath340 is the polynomial defined in lemma [ lemma : square - root - of - s ] .
then the explicit trigonal construction on @xmath109 described above yields an isogeny defined over @xmath7 if and only if the leading coefficient of @xmath340 is a square in @xmath7 .
we noted earlier that @xmath227 is defined over the field of definition of @xmath109 .
the correspondence @xmath119 , and hence the induced isogeny @xmath19 , are both defined over the field of definition of @xmath328 , which is the field of definition of @xmath341 , @xmath342 , and @xmath343 . but @xmath301 , @xmath301 , and @xmath298 are all defined over @xmath7 ( cf .
proposition [ proposition : b22-t - equation ] ) , while @xmath299 is defined over @xmath344 where @xmath292 is the leading coefficient of @xmath290 by lemma [ lemma : square - root - of - s ] .
if @xmath19 is not defined over @xmath7 , then the jacobian @xmath113 is in fact a quadratic twist of the quotient @xmath345 ( see [ section : other - isogenies ] ) .
in fact , when @xmath19 is not @xmath7-rational , frobenius exchanges @xmath328 and @xmath346 , hence @xmath119 and @xmath120 , and therefore @xmath19 and @xmath133 .
this is a concrete realization of the galois cohomology referred to in the proof of proposition [ proposition : kernels ] below : the obstruction to the existence of an isomorphism from @xmath345 to @xmath113 over @xmath7 is in fact the interaction of @xmath31 with @xmath347 $ ] on @xmath113 .
if we assume that the leading coefficients of the polynomials @xmath340 are uniformly distributed for randomly chosen @xmath11 , @xmath90 , and @xmath109 , then the probability that @xmath290 is a square in @xmath219 $ ] is @xmath213 .
indeed , it is easily seen that @xmath340 is a square for @xmath11 if and only if it is not a square for the quadratic twist of @xmath11 .
suppose @xmath106 is a hyperelliptic curve .
let @xmath348 be a non - square in @xmath7 , and let @xmath349 be the quadratic twist of @xmath11 .
suppose @xmath90 in @xmath92 is a tractable subgroup , represented by a set @xmath98 of quadratic factors of @xmath45 .
the set @xmath350 is a factorization of @xmath351 , so it represents a tractable subgroup @xmath352 in @xmath353 .
we noted in [ section : computing - trigonal - maps ] that scalar multiples of quadratic polynomials do not affect the construction of trigonal maps ; so if @xmath90 has a trigonal map @xmath109 defined over @xmath7 , then @xmath109 is also a trigonal map for @xmath352 .
let @xmath290 be the polynomial computed from @xmath109 and @xmath90 in lemma [ lemma : square - root - of - s ] , and let @xmath354 be the corresponding polynomial computed for @xmath109 and @xmath352 .
looking at the form of , we see that @xmath355 .
therefore , the leading coefficient of @xmath354 is a square if and only if the leading coefficient of @xmath290 is _ not _ a square . in particular , if @xmath90 has a trigonal map defined over @xmath7 , then so does @xmath352 , and we can construct an isogeny of jacobians with kernel @xmath90 if and only if we can not construct an isogeny of jacobians with kernel @xmath352 .
this suggests that the probability that we can compute an isogeny defined over @xmath7 given a randomly chosen @xmath11 and @xmath90 in @xmath92 with a trigonal map defined over @xmath7 is @xmath213 since we have a @xmath356 chance of being on the `` right '' quadratic twist of @xmath11 .
this hypothesis is consistent with our experimental observations . [
hypothesis : rationality ] for a randomly chosen hyperelliptic curve @xmath11 and a uniformly randomly chosen subgroup @xmath90 in @xmath92 with a trigonal map @xmath109 defined over @xmath7 , the probability that we can compute an @xmath7-rational isogeny @xmath19 with kernel @xmath90 is @xmath213 .
now we will put the ideas above into practice .
suppose we are given a hyperelliptic curve @xmath11 of genus @xmath0 over @xmath7 , and a dlp instance in @xmath13 to solve .
our goal is to compute a nonsingular plane quartic curve @xmath18 and an explcit isogeny @xmath357 defined over @xmath7 , so that we can solve our dlp instance in @xmath20 .
we begin by computing the set @xmath92 of @xmath7-rational tractable subgroups of the @xmath2-torsion subgroup @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] ( see appendix [ appendix : computing - sh ] below ) . for each @xmath90 in @xmath92
, we apply proposition [ proposition : trigonal - map - rationality - criterion ] to determine whether there exists an @xmath7-rational trigonal map @xmath109 for @xmath90 .
if so , we use the formulae of [ section : computing - trigonal - maps ] to compute @xmath109 ; if not , we move on to the next @xmath90 .
having computed @xmath109 , we apply proposition [ proposition : x - rationality - criterion ] to determine whether we can compute an isogeny over @xmath7 .
if so , we use the formulae of [ section : equations - for - the - isogeny ] to compute equations for @xmath23 and the isogeny @xmath114 ; if not , we move on to the next @xmath90 .
the formulae of [ section : equations - for - the - isogeny ] give an affine model of @xmath23 in @xmath358 . in order to apply diem s algorithm to the dlp in @xmath113
, we need a nonsingular plane quartic model of @xmath23 : that is , a nonsingular curve @xmath359 isomorphic to @xmath23 , cut out by a quartic form .
such a model exists if and only if @xmath23 is not hyperelliptic . to find @xmath18 ,
we compute a basis @xmath360 of the riemann roch space of a canonical divisor of @xmath23 .
this is a routine geometrical calculation ; hess @xcite describes an efficient approach . in practice ,
the algorithms implemented in magma @xcite compute @xmath361 very quickly .
the three functions in @xmath361 define a map @xmath362 , mapping @xmath15 to @xmath363 .
up to automorphisms of @xmath229 , the map @xmath364 is independent of the choice of basis @xmath361 , and depends only on @xmath23 . if the image of @xmath364 is a conic ( that is , if the @xmath365 satisfy a quadratic relation ) ,
then @xmath23 is hyperelliptic ; in this situation we move on to the next @xmath90 , since we will gain no advantage from index calculus on @xmath23 .
otherwise , the image of @xmath364 is a nonsingular plane quartic @xmath18 , and @xmath364 restricts to an isomorphism @xmath366 .
if the procedure outlined above succeeds for some @xmath90 in @xmath92 , then we have computed an explicit @xmath7-rational isogeny @xmath367 .
we can then map our dlp from @xmath368 into @xmath369 , and solve it using diem s algorithm .
we emphasize that the entire procedure is very fast : the curve @xmath23 and the isogeny can be constructed using just a few low - degree polynomial operations and some low - dimensional linear algebra ( and hence the procedure is polynomial - time in @xmath370 , the size of the base field ) . for a rough idea of the computational effort involved , given a random @xmath11 over a @xmath371-bit prime field with a tractable subgroup @xmath90 in @xmath92 , a nave implementation of our algorithms in magma computes the trigonal map @xmath109 , the curve @xmath23 , the nonsingular plane quartic @xmath18 , and the isogeny @xmath17 in a few seconds on a 1.2ghz laptop . since the difficulty of the construction depends only upon the difficulty of arithmetic in @xmath7 ( and _ not _ upon the size of the dlp subgroup of @xmath13 ) , we may conclude that instances of the dlp in @xmath371-bit jacobians chosen for cryptography may also be reduced to instances of the dlp in non - hyperelliptic jacobians in very little time . we will give an example over a small field .
let @xmath11 be the hyperelliptic curve over @xmath372 defined by @xmath373 using the ideas in [ section : kernel ] , or the algorithms in appendix [ appendix : computing - sh ] , we find that @xmath27 has one @xmath372-rational tractable subgroup : @xmath374 where @xmath375 is an element of @xmath376 satisfying @xmath377 , and @xmath378 . applying the methods of [ section : computing - trigonal - maps ] ,
we compute a trigonal map @xmath379 for @xmath90 , taking @xmath380 clearly @xmath109 is defined over @xmath372 .
the formulae of [ section : equations - for - the - isogeny ] give us a curve @xmath381 of genus @xmath0 , defined by @xmath382 the map on divisors inducing an isogeny from @xmath27 to @xmath113 with kernel @xmath90 is induced by the correspondence @xmath119 defined as in with @xmath383 computing the canonical morphism of @xmath23 , we find that @xmath23 is non - hyperelliptic , and isomorphic to the nonsingular plane quartic curve @xmath384 composing the isomorphism with the isogeny @xmath24 , we obtain an explicit isogeny @xmath17 .
we can verify that @xmath27 and @xmath385 are isogenous by checking that the zeta functions of @xmath11 and @xmath18 are identical : indeed , direct calculation with magma shows that @xmath386 let @xmath387 $ ] and @xmath388 $ ] be divisor classes on @xmath11 ; we have @xmath389d$ ] . applying @xmath19
, we find that @xmath390 \text{\quad and}\\ \phi(d ' ) & \left [ ( 7 : 23 : 1 ) + ( 6 : 13 : 1 ) - ( 13 : 15 : 1 ) - ( 7 : 18 : 1 ) \right ] ; \end{array}\ ] ] direct calculation verifies that @xmath391\phi(d)$ ] , as expected .
our aim in this section is to estimate the proportion of genus @xmath0 hyperelliptic jacobians over @xmath7 for which the methods of this article produce an @xmath7-rational isogeny and thus for which the dlp may be solved using diem s algorithm as @xmath30 tends to infinity .
we will assume that if we are given a selection of @xmath248-rational tractable subgroups of a given jacobian , then the probabilities that each will yield a rational isogeny are mutually independent .
this hypothesis appears to be consistent with our experimental observations .
[ hypothesis : independence ] for a randomly chosen hyperelliptic curve @xmath11 , the probabilities that we can compute an @xmath7-rational isogeny with kernel @xmath90 for each @xmath90 in @xmath92 are mutually independent .
[ theorem : success - probability ] assume hypotheses [ hypothesis : hyperellipticity ] , [ hypothesis : trigonal - map - rationality ] , [ hypothesis : rationality ] , and [ hypothesis : independence ] . as @xmath30 tends to infinity , the expectation that the algorithms in this article will give a reduction of the dlp in a subgroup of @xmath13 for a randomly chosen hyperelliptic curve @xmath11 of genus @xmath0 over @xmath7 to a subgroup of @xmath20 for some nonsingular plane quartic curve @xmath18 is @xmath392 where @xmath393 denotes the set of integer partitions of @xmath37 and @xmath394 denotes the multiplicity of an integer @xmath14 in a partition @xmath395 , and @xmath396 , where @xmath11 is _ any _ hyperelliptic curve over @xmath7 such that the multiset of degrees of the @xmath7-irreducible factors of its hyperelliptic polynomial coincides with @xmath395 .
suppose @xmath11 is a randomly chosen hyperelliptic curve of genus @xmath0 over @xmath7 .
hypotheses [ hypothesis : hyperellipticity ] , [ hypothesis : trigonal - map - rationality ] , and [ hypothesis : rationality ] together imply that for each @xmath90 in @xmath92 , the probability that we can compute an isogeny with kernel @xmath90 defined over @xmath7 is @xmath397 .
hypothesis [ hypothesis : independence ] implies that we have an equal chance of constructing an isogeny from each @xmath90 in @xmath92 , so the probability that we can compute an isogeny over @xmath7 from @xmath27 is @xmath398 . the expectation that we can compute an isogeny over @xmath7 given
a curve over @xmath7 is therefore @xmath399 where @xmath11 is the curve defined by @xmath48 , and @xmath45 ranges over the set of all homogeneous squarefree polynomials of degree @xmath37 over @xmath7 .
lemma [ lemma : number - of - tractable - subgroups ] implies that @xmath400 depends only on the degrees of the @xmath7-irreducible factors of @xmath45 , so the map @xmath401 is well - defined .
for each @xmath395 in @xmath393 , let @xmath402 denote the number of homogeneous squarefree polynomials over @xmath7 whose multiset of degrees of @xmath7-irreducible factors coincides with @xmath395 .
we can now rewrite as @xmath403 there are @xmath404 monic irreducible polynomials of degree @xmath14 over @xmath7 ( here @xmath405 is the mbius function ) .
clearly @xmath406 , so @xmath407 and @xmath408 .
therefore , as @xmath30 tends to infinity , we have @xmath409 the result follows upon explicitly computing this sum , using the values for @xmath410 listed in lemma [ lemma : number - of - tractable - subgroups ] .
theorem [ theorem : success - probability ] gives the expectation of our ability to construct an explicit isogeny for a randomly selected hyperelliptic curve .
however , looking at the table in lemma [ lemma : number - of - tractable - subgroups ] , we see that we can be sure that a particular curve has no isogenies with tractable kernels defined over @xmath7 if we use only curves whose hyperelliptic polynomials have an irreducible factor of degree @xmath26 or @xmath36 ( or a single irreducible factor of degree @xmath0 )
. it may be difficult to efficiently construct a curve in this form if we are using a cm construction , for example , to ensure that the jacobian has a large prime - order subgroup . in any case , it is interesting to note that the security of genus @xmath0 hyperelliptic jacobians depends significantly upon the factorization of their hyperelliptic polynomials .
this observation has no analogue for elliptic curves or jacobians of curves of genus @xmath2 .
of course , if @xmath411 is an elliptic curve and @xmath35 is completely reducible , then @xmath412 is divisible by @xmath42 , and in particular @xmath412 can not be prime ; but this does not reduce the security of @xmath413 to the extent that a completely reducible hyperelliptic polynomial does for a curve of genus @xmath0 . [
remark : trigonal - map - descent ] we noted in [ section : trigonal - construction ] that the @xmath91-isomorphism class of the curve @xmath23 in the trigonal construction is independent of the choice of trigonal map .
if there is no trigonal map defined over @xmath7 for a given subgroup @xmath90 in @xmath92 , then the methods of [ section : computing - trigonal - maps ] construct a pair of galois - conjugate trigonal maps @xmath324 and @xmath325 ( corresponding to the roots of ) instead . applying the trigonal construction to @xmath324 and @xmath325 , we obtain curves @xmath414 and @xmath415 over @xmath416 .
if the isomorphism between @xmath414 and @xmath415 were made explicit , then we could descend it to compute a curve @xmath23 over @xmath7 in the @xmath91-isomorphism class of @xmath414 and @xmath415 , and hence a nonsingular plane quartic @xmath18 over @xmath7 and an isogeny @xmath357 .
we note that the isogeny may not be defined over @xmath7 , but this approach could still allow us to replace the @xmath417 in and with @xmath213 , raising the expectation of success in theorem [ theorem : success - probability ] to @xmath418 .
let @xmath419 .
note that @xmath153 is a 60-bit prime ; if @xmath11 is a hyperelliptic curve of genus @xmath0 over @xmath420 such that @xmath421 has a large prime - order subgroup and if gaudry
thom thriault
diem index calculus is the fastest algorithm for solving dlp instances in @xmath421 , then @xmath27 has roughly the same security level as an elliptic curve over a 160-bit field .
we generated one million random hyperelliptic curves of genus @xmath0 over @xmath420 using magma .
for each curve @xmath11 , we computed the set @xmath92 of tractable subgroups ; then , for each @xmath90 in @xmath92 we determined whether there was an @xmath420-rational trigonal map for @xmath90 , and if so whether there was an @xmath420-rational isogeny with kernel @xmath90 .
of these curves , @xmath422 ( that is , @xmath423 ) had at least one rational tractable subgroup . between them ,
the @xmath424 curves had @xmath425 rational tractable subgroups , of which @xmath426 had a rational trigonal map ( that is , @xmath427 , which is close to the @xmath356 predicted by hypothesis [ hypothesis : trigonal - map - rationality ] ) . of these subgroups
, @xmath428 led to a rational isogeny ( that is , @xmath429 , which is close to the @xmath356 predicted by hypothesis [ hypothesis : rationality ] ) .
we found that @xmath430 of the curves had at least one @xmath420-rational isogeny , none of which had a hyperelliptic codomain ( this is compatible with hypothesis [ hypothesis : hyperellipticity ] ) .
in particular , we could move a discrete logarithm problem for @xmath431 of these curves ( recall that theorem [ theorem : success - probability ] predicts a success rate of about @xmath4 ) .
so far , we have concentrated on using isogenies with kernels generated by differences of weierstrass points to move instances of the dlp from hyperelliptic to non - hyperelliptic jacobians .
more generally , we could use isogenies with other kernels .
there are two important issues to consider here : the first is a theoretical restriction on the types of subgroups that can be kernels of isogenies of jacobians , and the second is a practical restriction on the isogenies that we can currently compute . let @xmath11 be a hyperelliptic curve of genus @xmath0 .
we want to characterize the subgroups @xmath90 of @xmath27 that are kernels of isogenies of jacobians , combining standard results from the theory of abelian varieties with some special results on curves of genus @xmath0 . for our purposes , it is enough to know that the @xmath70-_weil pairing _ is a nondegenerate , bilinear pairing on the @xmath70-torsion of an abelian variety , which can be efficiently evaluated in the case where the abelian variety is the jacobian of a hyperelliptic curve ; for further detail , we refer the reader to ( * ? ? ?
a.7.8 ) .
let @xmath432 be an abelian variety over @xmath7 , and let @xmath70 be a positive integer coprime with @xmath30 .
we say a subgroup @xmath90 of @xmath433 $ ] is _
maximal @xmath70-isotropic _ if 1 .
the @xmath70-weil pairing on @xmath433 $ ] restricts trivially to @xmath90 , and 2 .
@xmath90 is not properly contained in any other subgroup of @xmath433 $ ] satisfying ( 1 ) .
if @xmath70 is a prime not dividing @xmath30 , then every maximal @xmath70-isotropic subgroup of @xmath434 $ ] is isomorphic to @xmath435 .
the situation is more complicated when @xmath70 is not prime : for example , @xmath81 $ ] is a maximal @xmath42-isotropic subgroup of @xmath436 $ ] , but it is isomorphic to @xmath437 and not @xmath438 . [ proposition : kernels ] let @xmath11 be a hyperelliptic curve of genus @xmath0 over @xmath7 such that @xmath27 is absolutely simple .
let @xmath90 be a finite , nontrivial , @xmath7-rational subgroup of @xmath95 .
there exists a curve @xmath23 of genus @xmath0 over @xmath7 , and an isogeny @xmath114 with kernel @xmath90 , if and only if @xmath90 is a maximal @xmath70-isotropic subgroup of @xmath69 $ ] for some positive integer @xmath70 .
the isogeny @xmath19 is defined over @xmath416 .
the quotient @xmath439 always exists as an isogeny of abelian varieties , and is defined over @xmath7 ( see serre @xcite ) . for the quotient to be an isogeny of jacobians , there must be an integer @xmath70 such that @xmath90 is a maximal @xmath70-isotropic subgroup ( see proposition 16.8 of milne @xcite ) : this ensures that the canonical polarization on @xmath27 induces a principal polarization on the quotient @xmath345 .
the theorem of oort and ueno @xcite therefore guarantees that there will be an isomorphism of principally polarized abelian varieties over @xmath91 from @xmath345 to the jacobian @xmath113 of some irreducible curve @xmath23 ( irreducibility of @xmath23 follows from the fact that @xmath27 , and hence @xmath345 , is absolutely simple ) . composing this isomorphism with the quotient map gives an isogeny of jacobians from @xmath27 to @xmath113 with kernel @xmath90 .
standard arguments from galois cohomology ( see serre @xcite , for example ) show that the isomorphism is defined over either @xmath7 or @xmath416 , and it follows that the isogeny @xmath24 must be defined over @xmath7 or @xmath416 .
proposition [ proposition : kernels ] does _ not _ hold in higher genus : for every @xmath440 , there are @xmath109-dimensional abelian varieties that are not isomorphic to jacobians .
indeed , this is the generic situation : for @xmath441 the moduli space of @xmath109-dimensional abelian varieties is @xmath442-dimensional , with the jacobians occupying a subspace of dimension @xmath443 which is strictly less than @xmath442 for @xmath440 .
we should not therefore expect an arbitrary quotient of a jacobian to be isomorphic to a jacobian in genus @xmath440 .
proposition [ proposition : kernels ] does hold in genus @xmath5 and @xmath2 , and in these cases the isogenies are always defined over @xmath7 .
we can expect the curve @xmath23 of proposition [ proposition : kernels ] to be non - hyperelliptic . to compute an @xmath7-rational isogeny from @xmath27 to a non - hyperelliptic jacobian ,
therefore , the minimum requirement is an @xmath7-rational @xmath70-isotropic subgroup of @xmath95 isomorphic to @xmath435 for some prime @xmath70 .
we emphasize that this subgroup need _ not _ be contained in @xmath13 . indeed
, there may be isogenies from @xmath27 to non - hyperelliptic jacobians over @xmath7 even when @xmath13 has prime order ( which would be the desirable situation in cryptological applications ) .
the major obstruction to using more general isogenies to move dlp instances is the lack of general constructions for explicit isogenies in genus @xmath0 .
apart from integer multiplications , automorphisms , frobenius isogenies , and the construction for isogenies with tractable kernels exhibited above , we know of no constructions for explicit isogenies of general jacobians of genus @xmath0 hyperelliptic curves . in particular
, while we know that the curve @xmath23 of proposition [ proposition : kernels ] exists , we generally have no means of computing a defining equation for it , let alone equations for a correspondence between @xmath11 and @xmath23 that would allow us to move dlp instances from @xmath27 to @xmath113 .
this situation stands in marked contrast to the case of isogenies of elliptic curves , which have been made completely explicit by vlu @xcite .
deriving general formulae for explicit isogenies in genus @xmath0 ( and @xmath2 ) remains a significant problem in computational number theory .
the greater part of this work was completed in the department of mathematics at royal holloway , university of london , where the author was supported by epsrc grant ep / c014839/1 .
the author gratefully acknowledges roger oyono and christophe ritzenthaler for discussions which inspired this research , and steven galbraith and the anonymous referees for their helpful suggestions .
xx birkenhake , c. , lange , h. : complex abelian varieties ( 2e ) , grundlehren der mathematischen wissenschaften 302 .
springer ( 2004 ) bosma , w. , cannon , j. , playoust , c. : the magma computational algebra system .
i. the user language .
j. symbolic comp . 24(3 - 4 ) , 235265 ( 2006 ) bost , j .- b .
, mestre , j .- f .
: moyenne arithmtico - gomtrique et priodes des courbes de genre 1 et 2 .
france 38 , 3664 ( 1988 ) cartier , p. : isogenies and duality of abelian varieties , annals of mathematics 71 no . 2 , 315351 ( 1960 ) diem , c. : an index calculus algorithm for plane curves of small degree .
in : hess , f. , pauli , s. , pohst , m. ( eds . ) ants - vii .
lncs , vol . 4076 , pp .
springer ( 2006 ) donagi , r. : the fibres of the prym map . in : curves , jacobians , and abelian varieties ( amherst , ma , 1990 ) .
contemp .
136 , 55125 ( 1992 ) donagi , r. , livn , r. : the arithmetic - geometric mean and isogenies for curves of higher genus .
ann . scuola norm .
pisa cl .
( 4 ) 28 no . 2 , 323339 ( 1999 ) gaudry , p. , thom , e. , thriault , n. , diem , c. : a double large prime variation for small genus hyperelliptic index calculus . math . comp .
76 , 475492 ( 2007 ) griffiths , p. , harris , j. : principles of algebraic geometry .
wiley and sons ( 1978 ) harris , j. : algebraic geometry : a first course .
springer ( 1992 ) hess , f. : computing riemann - roch spaces in algebraic function fields and related topics .
j. symbolic computation 33 v.4 425445 ( 2002 ) hindry , m. , silverman , j. : diophantine geometry : an introduction .
graduate texts in mathematics 201 .
springer ( 2000 ) lang , s. , weil , a. , number of points of varieties in finite fields .
am . j. math .
, lxxvi no .
4 819827 ( 1954 ) lehavi , d. , ritzenthaler , c. : an explicit formula for the arithmetic geometric mean in genus 3 . experimental math . 16 421440 ( 2007 ) the magma computational algebra system , ` http://magma.maths.usyd.edu.au/ ` milne , j. s. : abelian varieties . in : arithmetic geometry ( storrs ,
conn . , 1984 ) ,
springer ( 1986 ) mumford , d. : tata lectures on theta ii .
birkhuser ( 1984 ) oort , f. , ueno , k. : principally polarized abelian varieties of dimension two or three are jacobian varieties .
j. fac .
tokyo sect .
ia math .
20 , 377381 ( 1973 ) pohlig , g. , hellman , m. : an improved algorithm for computing logarithms over @xmath444 and its cryptographic significance .
ieee trans .
theory 24 , 106110 ( 1978 ) recillas , s. : jacobians of curves with @xmath445 s are the prym s of trigonal curves .
soc . mat .
mexicana ( 2 ) 19 , no . 1 913 ( 1974 ) serre , j .-
: algebraic curves and class fields . graduate texts in mathematics 117 .
springer ( 1988 ) serre , j .-
: galois cohomology .
springer ( 2002 ) smith , b. : isogenies and the discrete logarithm problem in jacobians of genus @xmath0 hyperelliptic curves . in : smart , n. ( ed . ) eurocrypt 2008 .
lncs vol . 4965 , pp .
springer ( 2008 ) vakil , r. : twelve points on the projective line , branched covers , and rational elliptic fibrations .
320 no . 1 , 3354 ( 2001 ) . vlu , j. : isognies entre courbes elliptiques . c. r. acad .
paris , sries a 273 , 305347 ( 1971 )
given a hyperelliptic curve @xmath11 of genus @xmath0 over @xmath7 , we want to compute the set @xmath92 of @xmath7-rational tractable subgroups of @xmath27 . algorithm [ algorithm : subgroup - enumeration ] splits the hyperelliptic polynomial of @xmath11 into galois orbits of factors , before calling the recursive subroutine algorithm [ algorithm : subgroup - subroutine ] to enumerate @xmath92 .
this algorithm is included only for completeness , and is not particularly efficient ( we suggest some optimisations in remark [ remark : tractable - implementation ] below . )
[ algorithm : subgroup - enumeration ] given a hyperelliptic curve @xmath11 of genus @xmath0 over @xmath7 , enumerates the set @xmath92 of @xmath7-rational tractable subgroups of @xmath81({{\overline{\mathbb{f}}}}_{q})$ ] .
each subgroup in @xmath92 is represented by a set of four coprime quadratic factors of @xmath45 .
input : : the hyperelliptic polynomial @xmath446 of @xmath11 .
output : : the set @xmath92 .
step 1 : : let @xmath447 be the set of irreducible factors of @xmath45 over its splitting field , + scaled so that @xmath448 , and set @xmath449 .
step 2 : : choose a polynomial @xmath152 from @xmath447 . set @xmath450 , set @xmath451 , + and set @xmath452 .
step 3 : : set @xmath453 , where @xmath454 denotes the @xmath33 power frobenius map .
+ if @xmath455 , then append @xmath152 to @xmath456 , set @xmath451 , and go to * step 3*. + if @xmath457 , then set @xmath458 ; if @xmath459 , then go to * step 2*. step 4 : : return the result of algorithm [ algorithm : subgroup - subroutine ] applied to @xmath460 .
input : : a set @xmath460 of disjoint sequences of distinct linear polynomials .
each sequence @xmath461 in @xmath460 must satisfy @xmath462 and @xmath463 for @xmath464 , where @xmath454 denotes the @xmath33-power frobenius map .
output : : the set @xmath465 of @xmath31-stable sets of coprime quadratic polynomials such that @xmath466 .
step 1 : : if @xmath460 is empty , then return @xmath467 .
step 2 : : choose a sequence @xmath456 from @xmath460 , and set @xmath468 .
+ if @xmath469 is even , then let @xmath393 be the result of algorithm [ algorithm : subgroup - subroutine ] applied to @xmath470 , + and set @xmath471 . + if @xmath469 is odd , then set @xmath472 .
step 3 : : for each @xmath15 in @xmath473 such that @xmath474 , + step 3i ; ; set @xmath475 .
step 3ii ; ; let @xmath476 be the result of algorithm [ algorithm : subgroup - subroutine ] applied to @xmath477 .
step 3iii ; ; set @xmath478 .
step 4 : : return @xmath465 .
[ remark : tractable - implementation ] as we noted above , algorithms @xmath42 and @xmath26 are not particularly efficient : for conceptual simplicity we worked over the splitting field of the hyperelliptic polynomial , and this can be extremely slow in practice . a number of simple optimizations will significantly improve the performance of this algorithm : the key is to avoid field extensions where possible , and to minimize their degree in any case . before factoring @xmath45 over its splitting field we should factor it over @xmath7 , and then work on a case - by - case basis depending on the degrees of the @xmath7-irreducible factors .
for example , if @xmath45 has an odd number of odd - degree factors , then @xmath92 is empty by lemma [ lemma : number - of - tractable - subgroups ] , and we can simply return the empty set .
if @xmath45 is @xmath7-irreducible , then it is not necessary to factor @xmath45 over its splitting field ( which is @xmath479 ) : there is one tractable subgroup , and it corresponds to the four quadratic factors of @xmath45 that we obtain by factoring @xmath45 over @xmath480 . making similar modifications for the cases where @xmath45 has factors of degree @xmath178 , we can avoid working over any extensions of degree greater than @xmath42 .
if desired , we can further avoid some field extensions in the case where @xmath45 has only low - degree factors .
these modifications resulted in a factor - of-@xmath481 speedup for our experiments with @xmath482-bit prime fields ; the unmodified algorithms [ algorithm : subgroup - enumeration ] and [ algorithm : subgroup - subroutine ] should _ not _ be used in practice . | we describe the use of explicit isogenies to translate instances of the discrete logarithm problem ( dlp ) from jacobians of hyperelliptic genus @xmath0 curves to jacobians of non - hyperelliptic genus @xmath0 curves , where they are vulnerable to faster index calculus attacks . we provide explicit formulae for isogenies with kernel isomorphic to @xmath1 ( over an algebraic closure of the base field ) for any hyperelliptic genus @xmath0 curve over a field of characteristic not @xmath2 or @xmath0 .
these isogenies are rational for a positive fraction of all hyperelliptic genus @xmath0 curves defined over a finite field of characteristic @xmath3 .
subject to reasonable assumptions , our constructions give an explicit and efficient reduction of instances of the dlp from hyperelliptic to non - hyperelliptic jacobians for around @xmath4 of all hyperelliptic genus @xmath0 curves over a given finite field .
we conclude with a discussion on extending these ideas to isogenies with more general kernels . a condensed version of this work appeared in the proceedings of the eurocrypt 2008 conference . [ definition]algorithm |
while on the gravity side , the lovelock gravity coupled nonminimally to a scalar dilaton field appears in the low energy limit of string theory @xcite , the born - infeld action coupled to a dilaton field , appears in the low energy limit of open superstring theory on the electrodynamic side @xcite . although one can consistently truncate such models , the presence of the dilaton field can not be ignored if one consider coupling of the gravity to other gauge fields , and therefore one remains with einstein - born - infeld gravity in the presence of a dilaton field .
some efforts have been done to construct exact solutions of einstein - dilaton gravity in the presence of linear electromagnetic field .
for example , exact static dilaton black hole solutions of einstein - maxwell - dilaton ( emd ) gravity have been considered in refs .
exact rotating black holes solutions with curved horizons have been obtained only for some limited values of the coupling constant @xcite .
recently , in the absence of dilaton potential , a class of asymptotically flat rotating solution in @xmath9-dimensional emd gravity have been constructed by embedding the @xmath9-dimensional myers - perry @xcite solutions in @xmath10 dimensions , and performing a boost with a subsequent kaluza klein reduction @xcite .
this procedure leads to emd black holes in @xmath9 dimensions , for particular values of the dilaton coupling constant . for general dilaton coupling , the properties of rotating charged dilaton black holes only with infinitesimally small charge @xcite or small angular momentum
have been investigated in four @xcite and five @xcite dimensions . when the horizons are flat , exact rotating solutions of emd gravity with liouville - type potential have been constructed in four @xcite and @xmath11 dimensions @xcite . in this paper , we want to consider the black brane solutions of einstein - dilaton gravity in the presence of nonlinear electromagnetic field . the idea of the nonlinear electrodynamics was first introduced in 1934 by born and infeld in order to obtain a finite value for the self - energy of point - like charges @xcite .
although it become less popular with the introduction of qed , in recent years , the born - infeld action has been occurring repeatedly with the development of superstring theory , where the dynamics of d - branes is governed by the born - infeld action @xcite .
it has been shown that , charged black holes solutions in einstein - born - infeld gravity are less singular in comparing with the reissner - nordstrm solution . in other word , there is no reissner - nordstrm - type divergence term @xmath12 in the metric near the singularity while it exist only a schwarzschild - type term @xmath13 @xcite . in the absence of dilaton field , exact solutions of einstein - born - infeld black holes with / without cosmological constant
have been constructed @xcite , while the rotating solutions of ebi gravity with flat horizons have been obtained in @xcite . in the presence of dilaton , exact solutions of the einstein born - infeld dilaton ( ebid ) gravity
are known only in three dimensions @xcite .
numerical studies of the four - dimensional spherically symmetric solutions of ebid gravity have been done in @xcite .
some factorized black hole solutions in four - dimensional ebid theory have been discovered in @xcite .
till now , exact rotating solution of ebid gravity for an arbitrary value of coupling constant have not been constructed . in this paper
, we first introduce the action of @xmath11-dimensional ebid gravity in the presence of a potential and obtain its field equations .
then , we construct the @xmath11-dimensional rotating black brane solution of this theory with cylindrical or toroidal horizons in the presence of liouville - type potential and investigate the effect of dilaton and born - infeld fields on the stability of the solutions . the outline of our paper is as follows . in sec .
[ field ] we introduce the action of the @xmath11-dimensional ebid gravity and obtain the field equations . in sec .
[ charged ] , we construct the @xmath11-dimensional rotating dilaton black branes with a complete set of rotation parameters in the ebid theory and investigate their properties . in sec .
[ therm ] , we obtain the conserved and thermodynamical quantities of the @xmath11-dimensional black brane solutions and show that these quantities satisfy the first law of thermodynamics .
we also perform a stability analysis and show that the dilaton creates an unstable phase for the solutions .
we finish our paper with some closing remarks .
we consider the @xmath11-dimensional action in which gravity is coupled to dilaton and born infeld fields with an action @xmath14 where @xmath15 is the ricci scalar curvature , @xmath16 is the dilaton field , @xmath17 is a potential for @xmath16 and @xmath18 is @xmath19 here , @xmath20 is a constant determining the strength of coupling of the scalar and born - infeld fields , @xmath21 , where @xmath22 is the electromagnetic tensor field,@xmath23 is the vector potential , and @xmath24 is the born - infeld parameter with dimension of mass . in the limit @xmath25 , @xmath18 reduces to the standard maxwell field coupled to a dilaton field @xmath26 on the other hand , @xmath27 as @xmath28 .
it is convenient to set @xmath29 where @xmath30 the last term in eq .
( [ act ] ) is the gibbons - hawking boundary term which is chosen such that the variational principle is well - defined .
the manifold @xmath31 has metric @xmath32 and covariant derivative @xmath33 .
@xmath34 is the trace of the extrinsic curvature @xmath35 of any boundary(ies ) @xmath36 of the manifold @xmath37 , with induced metric(s ) @xmath38 . in this paper
, we consider the action ( [ act ] ) with a liouville type potential , @xmath39 where @xmath40 is a constant which may be referred to as the cosmological constant , since in the absence of the dilaton field ( @xmath41 ) the action ( [ act ] ) reduces to the action of einstein - born - infeld gravity with cosmological constant .
the equations of motion can be obtained by varying the action ( [ act ] ) with respect to the gravitational field @xmath32 , the dilaton field @xmath42 and the gauge field @xmath43 which yields the following field equations @xmath44 g_{\mu \nu } , \label{fe1}\end{aligned}\ ] ] @xmath45 , \label{fe2}\ ] ] @xmath46 in particular , in the case of linear electrodynamics with @xmath47 , the system of equations ( [ fe1])-([fe3 ] ) reduce to the well - known equations of emd gravity @xcite .
[ field]our aim here is to construct the @xmath11-dimensional rotating solutions of the field equations ( [ fe1])-([fe3 ] ) with @xmath48 rotation parameters and investigate their properties . the rotation group in @xmath11-dimensions is @xmath49 and therefore the number of independent rotation parameters for a localized object is equal to the number of casimir operators , which is @xmath50\equiv k$ ] , where @xmath51 $ ] is the integer part of @xmath52 .
the metric of @xmath11-dimensional rotating solution with cylindrical or toroidal horizons and @xmath48 rotation parameters can be written as @xcite @xmath53 where @xmath54 s are @xmath48 rotation parameters .
the functions @xmath55 and @xmath56 should be determined and @xmath57 has the dimension of length which is related to the cosmological constant @xmath40 for the case of liouville - type potential with constant @xmath16 .
the angular coordinates are in the range @xmath58 and @xmath59 is the euclidean metric on the @xmath60-dimensional submanifold with volume @xmath61 . the modified maxwell equation ( [ fe3 ] ) can be integrated immediately to give @xmath62 where @xmath63 , is an integration constant related to the electric charge of the brane . in order to solve the system of equations ( [ fe1 ] ) and ( [ fe2 ] ) for three unknown functions @xmath55 , @xmath56 and
@xmath64 , we make the ansatz @xmath65 using ( [ rphi ] ) , the born - infeld fields ( [ ftr ] ) and the metric ( [ met3 ] ) , one can easily show that the field equations ( [ fe1 ] ) and ( [ fe2 ] ) have solution of the form @xmath66 @xmath67 where @xmath68 here @xmath69 and @xmath70 are arbitrary constants and @xmath71 .
the integral can be done in terms of hypergeometric function and can be written in a compact form as @xmath72 , \left [ \frac{\alpha ^{2}+n-2}{2n-2}\right ] , -\eta \right ) \right ) .\end{aligned}\ ] ] one may note that as @xmath73 these solutions reduce to the @xmath11-dimensional charged rotating dilaton black brane solutions given in @xcite . in the absence of a nontrivial dilaton ( @xmath74 ) , the above solutions reduce to the asymptotically ads charged rotating black brane solutions of einstein - born - infeld gravity presented in @xcite , and in the limit @xmath75 and @xmath76 , these solutions reduce to those of @xcite . using the fact that @xmath77 has a convergent series expansion for @xmath78 , we can find the behavior of the metric for large @xmath79 .
this is given by @xmath80 the last term in the right hand side of the above expression is the leading born - infeld correction to the ads black brane with dilaton field . in order to study the general structure of these solutions ,
we first look for the curvature singularities in the presence of dilaton gravity .
it is easy to show that the kretschmann scalar @xmath81 diverges at @xmath82 , it is finite for @xmath83 and goes to zero as @xmath84 .
thus , there is an essential singularity located at @xmath82 .
the spacetime is neither asymptotically flat nor ( a)ds . as in the case of rotating black hole solutions of the einstein gravity , the above metric given by ( [ met3 ] ) and ( [ f ] ) has both killing and event horizons .
the killing horizon is a null surface whose null generators are tangent to a killing field .
it is easy to see that the killing vector @xmath85 is the null generator of the event horizon , where @xmath86 is the @xmath87th component of angular velocity of the outer horizon which may be obtained by analytic continuation of the metric .
one can obtain the temperature and angular momentum of the event horizon by analytic continuation of the metric .
setting @xmath88 and @xmath89 yields the euclidean section of ( [ met3 ] ) , whose regularity at @xmath90 requires that we should identify @xmath91 and @xmath92 , where @xmath93 and @xmath86 s are the inverse hawking temperature and the angular velocities of the outer event horizon .
one obtains @xmath94 , \left [ { \frac{{3n+\alpha } ^{2}{-4}}{{2n-2}}}\right ] , -\eta _ { + } \right ) , \label{temp}\end{aligned}\ ] ] @xmath95 as one can see from eq .
( [ f ] ) , the solution is ill - defined for @xmath96 .
the cases with @xmath97 and @xmath98 should be considered separately . in the first case where
@xmath97 , as @xmath79 goes to infinity the dominant term is the second term , and therefore the spacetime has a cosmological horizon for positive values of the mass parameter , despite the sign of the cosmological constant @xmath40 . in the second case
where @xmath98 , as @xmath79 goes to infinity the dominant term is the first term , and therefore there exist a cosmological horizon for @xmath99 , while there is no cosmological horizons if @xmath100 .
it is worth to mention that when there exists a cosmological horizon , the solution does not present black hole .
this is due to the fact that the metric function @xmath55 has only one real positive root . in the case of ( @xmath101 and @xmath100 ) the spacetimes associated with the solution ( [ f ] ) exhibit a variety of possible casual structures depending on the values of the metric parameters @xmath20 , @xmath102 , @xmath70 , @xmath63 , and @xmath103 .
one can obtain the casual structure by finding the roots of @xmath104 .
unfortunately , because of the nature of the exponents in ( [ f ] ) , it is not possible to find explicitly the location of horizons for an arbitrary value of @xmath20 .
but , we can obtain some information by considering the temperature of the horizons . using the fact that the temperature of the extreme black brane is zero , it is easy to show that the condition for having an extreme black hole is that the mass parameter is equal to @xmath105 , where @xmath105 is given as @xmath106 /(1-\gamma ) } } { ( \alpha ^{2}-n)(n-1)}\left ( \frac{q_{% \mathrm{ext}}^{2}}{\beta ^{2}\eta _ { \mathrm{ext}}}\right ) ^{[n(1-\gamma ) -\gamma ] /[2(n-1)(1-\gamma ) ] } \times \nonumber \\ & & \left\{\lambda -2\beta ^{2}\left [ 1-\;_{2}f_{1}\left ( \left [ -{\frac{1}{2},% \frac{{\alpha } ^{2}{-n}}{{2n-2}}}\right ] , \left [ { \frac{{\alpha } ^{2}+n{-2}}{% { 2n-2}}}\right ] , -\eta _ { \mathrm{ext}}\right ) \right ] \right\ } , \nonumber \\
\eta _ { \mathrm{ext } }
& = & \frac{\lambda ( \lambda -4\beta ^{2})}{4\beta ^{4}}. \label{mext}\end{aligned}\ ] ] indeed the metric of eqs . ( [ met3 ] ) and ( [ f ] ) has two inner and outer horizons located at @xmath107 and @xmath108 , provided the mass parameter @xmath70 is greater than @xmath105 , an extreme black hole in the case of @xmath109 , and a naked singularity if @xmath110 .
in this section we calculate the action and the thermodynamic and conserved quantities of the solutions . for asymptotically ads solutions ,
the way that one can calculate these quantities and obtain finite values for them is through the use of the counterterm method inspired by ads / cft correspondence @xcite . although , in the presence of a nontrivial dilaton field , the spacetime may not behave as either ds ( @xmath99 ) or ads ( @xmath100 ) , one may use the counterterm method for computing the conserved quantities . since our solutions have flat boundary [ @xmath111 , there exists only one boundary counterterm @xmath112 ^{1/2}. \label{ict}\ ] ]
one may note that the counterterm has the same form as in the case of asymptotically ads solutions with zero curvature boundary when @xmath20 goes to zero .
having the total finite action @xmath113 , one can use the quasilocal definition @xcite to construct a divergence free stress - energy tensor . for the case of manifolds with zero curvature boundary
the finite stress - energy tensor is @xmath114 ^{1/2}h^{ab}\right\ } .
\label{stres}\ ] ] the first two terms in eq .
( [ stres ] ) is the variation of the action ( [ act ] ) with respect to @xmath38 , and the last term is the variation of the boundary counterterm with respect to @xmath38 .
thus a conserved charge @xmath115 can be associated with a spacelike surface @xmath116 ( with normal @xmath117 ) , provided the boundary geometry has an isometry generated by a killing vector @xmath118 @xcite . for boundaries with timelike ( @xmath119 ) and rotational ( @xmath120 ) killing vector fields , one obtains the mass and angular momentum as the conserved quantities . denoting the volume of the boundary at constant @xmath121 and @xmath79 by @xmath122 , the mass and angular momentum per unit volume @xmath123 of the black branes ( @xmath98 ) can be calculated by use of eq .
( [ charge ] ) .
we find @xmath124 @xmath125 for @xmath126 ( @xmath127 ) , the angular momentum per unit volume vanishes , and therefore @xmath54 s are the rotational parameters of the spacetime .
notice that the total mass and angular momentum per unit volume are similar to the mass and angular momentum of @xmath11-dimensional emd solutions presented in @xcite .
next , we calculate the electric charge of the solutions . to determine the electric field we should consider the projections of the electromagnetic field tensor on special hypersurfaces .
the normal to such hypersurfaces is @xmath128 where @xmath129 and @xmath130 are the lapse function and shift vector .
then the electric field is @xmath131 , and the electric charge per unit volume @xmath123 can be found by calculating the flux of the electric field at infinity , yielding @xmath132 the electric potential @xmath133 , measured at infinity with respect to the horizon , is defined by @xcite @xmath134 where @xmath135 is the null generator of the horizon given by eq .
( [ kil ] ) .
one can easily show that the vector potential @xmath43 corresponding to the electromagnetic tensor ( [ ftr ] ) can be written as @xmath136 , \left [ { \frac{{3n+\alpha } ^{2}{-4}}{{2(n-1)}}}\right ] , -\eta \right ) \nonumber \label{vectorpot } \\ & & \times \left ( \xi \delta _ { \mu } ^{t}-a_{i}\delta _ { \mu } ^{i}\right ) \hspace{0.5cm}(\mathrm{no\;sum\;on}\;i),\end{aligned}\ ] ] where @xmath137 .
therefore , the electric potential may be obtained as @xmath138 , \left [ { \frac{{3n+\alpha } ^{2}{-4}}{{2(n-1)}}}\right ] -\eta _ { + } \right ) . \label{pot}\ ] ] black hole entropy typically satisfies the so called area law of the entropy @xcite .
this near universal law applies to almost all kinds of black holes and black branes in einstein gravity @xcite . in our case ,
due to the presence of dilaton field , we calculate the entropy through the use of gibbs - duhem relation @xmath139 where @xmath140 is the finite total action evaluated on the classical solution . using eqs .
( [ act ] ) and ( [ ict ] ) , the finite total action per unit volume @xmath123 can be calculated as @xmath141 where @xmath142 , \left [ { % \frac{{3n+\alpha } ^{2}{-4}}{{2n-2}}}\right ] , -\eta _ { + } \right ) .\end{aligned}\ ] ] now using gibbs - duhem relation and the thermodynamic and conserved quantities , calculated before , one obtains @xmath143 for the entropy per unit volume @xmath123 .
this shows that the entropy obeys the area law for our case where the horizon curvature is zero .
now , we check the first law of thermodynamics for our solutions in einstein - born - infeld - dilaton gravity . in order to do this ,
we obtain the mass @xmath144 as a function of extensive quantities @xmath145 , @xmath146 and @xmath147 . using the expression for the mass , the angular momenta , the charge , and the entropy given in eqs .
( [ mass ] ) , ( [ angmom ] ) , ( [ charge ] ) , ( [ entropy ] ) and the fact that @xmath148 , one can obtain a smarr - type formula as @xmath149 j}{% l(n-\alpha ^{2})\sqrt{z(z-1 ) } } , \label{smar}\ ] ] where @xmath150 and @xmath151 is the positive real root of the following equation @xmath152/(n-1)}\pi j}{(1+\alpha ^{2})b^{\alpha ^{2}}\sqrt{(z-1)}z^{(\alpha ^{2}-1)/(2n-2)}}s^{(\alpha ^{2}-n)/(n-1)}+ \nonumber \\ & & \frac{4\beta ^{2}}{n-1}\left [ 1-\;_{2}f_{1}\left ( \left [ \frac{-1}{2},% \frac{\alpha ^{2}-n}{2n-2}\right ] , \left [ \frac{\alpha ^{2}+n-2}{2n-2}\right ] , \frac{-\pi ^{2}q^{2}}{\beta ^{2}s^{2}}\right ) \right ] = 0 .\end{aligned}\ ] ] one may then regard the parameters @xmath145 , @xmath146 , and @xmath147 as a complete set of extensive parameters for the mass @xmath153 and define the intensive parameters conjugate to @xmath145 , @xmath146 and @xmath147 .
these quantities are the temperature , the angular velocities , and the electric potential .
it is a matter of forward calculations to obtain : @xmath154 @xmath155 @xmath156 where @xmath157 , \left [ { \frac{\alpha ^{2}+{3n-4}}{{2n-2}}}\right ] , -% \frac{{\pi } ^{2}q^{2}}{\beta ^{2}s^{2}}\right)$ ] . using eqs . ( [ angmom ] ) , ( [ charge ] ) and ( [ entropy ] ) , it is easy to show that the intensive quantities calculated by eqs .
( [ tsmar ] ) , ( [ jsmar ] ) and ( [ usmar ] ) coincide with eqs .
( [ temp ] ) , ( [ om1 ] ) and ( [ pot ] ) .
thus , these thermodynamics quantities satisfy the first law of thermodynamics @xmath158 finally , we investigate the stability of charged rotating black brane solutions of born - infeld - dilaton gravity in the canonical ensemble . in the canonical ensemble ,
the charge and the angular momenta are fixed parameters , and therefore the positivity of the heat capacity or @xmath159 is sufficient to ensure the local stability of the system .
we , first , consider the solutions for @xmath3 .
numerical calculations show that the black brane solutions are stable independent of the values of charge and born - infeld parameters in any dimensions provided @xmath3 . to achieve further understanding for @xmath3
, we perform the stability analysis for the uncharged case @xmath160 .
we obtain @xmath161 } \times \nonumber \\ & & \left [ ( \xi ^{2}-1)(n+1 - 2\alpha ^{2})+\xi ^{2}(1-\alpha ^{2})\right ] .
\label{dmss}\end{aligned}\ ] ] since @xmath162 , one finds that the above expression is positive provided @xmath3 .
thus , the uncharged solutions are stable for @xmath163 .
also it is worth to mention that increasing the electric charge not only does not change a stable solution to unstable one , but it increases the stable phase of the system @xcite .
thus , the charged solutions are also stable for @xmath3 . for @xmath164 ,
numerical calculations show that the solutions may have an unstable phase . in order to investigate the effects of the electric charge @xmath63 , the dilaton coupling constant @xmath2 , and the bi parameter @xmath165 on the unstable phase of the solutions ,
we plot @xmath166 versus @xmath167 , @xmath20 and @xmath102 . figures [ figure1 ] , [ figure2 ] and [ figure3 ] show that the solution is stable only for @xmath168 , where @xmath169 depends on @xmath170 , @xmath63 and @xmath171 .
that is the dilaton field makes the solution unstable .
figure [ figure4 ] shows that as the nonlinearity of the electromagnetic field increases ( @xmath165 decreases ) , the stable phase of the solution decreases .
that is the nonlinearity of the electromagnetic field makes the solutions more unstable .
this fact may be seen in figs .
[ figure4 ] , [ figure5 ] and [ figure6 ] which show that the solutions are unstable for very highly nonlinear field ( small value of the born - infeld parameter ) .
it is notable to mention that @xmath172 decreases with increasing @xmath63 or @xmath170 ( see figs .
[ figure4 ] and [ figure5 ] ) and it increases with increasing @xmath173 ( see fig .
[ figure6 ] ) .
also it is worth to mention that as in the case of rotating solutions in the presence of linear electromagnetic field , the solutions are unstable for small values of electric charge ( see figs .
[ figure7 ] , [ figure8 ] and [ figure9 ] ) .
in this paper , we introduced the @xmath11-dimensional einstein - born - infeld action coupled to a dilaton field and obtained the equations of motion by varying this action with respect to the gravitational field @xmath32 , the dilaton field @xmath16 and the gauge field @xmath43 .
in particular , in the case of the linear electrodynamics with @xmath174 , the system of equations reduce to the well - known equations of emd gravity @xcite .
then , we constructed a new class of charged rotating solutions of @xmath11-dimensional einstein - born - infeld - dilaton theory with cylindrical or toroidal horizons in the presence of liouville - type potentials and investigated their properties .
these solutions are neither asymptotically flat nor ( anti)-de sitter . in the particular case @xmath175 , these solutions reduce to the @xmath11-dimensional charged rotating dilaton black brane solutions constructed in ref .
@xcite , while in the absence of a nontrivial dilaton ( @xmath176 ) , the above solutions reduce to the asymptotically ads charged rotating black brane solutions of einstein - born - infeld theory presented in @xcite . these solutions which exist only for @xmath177
have only a cosmological horizon for ( _ i _ ) @xmath178 despite the sign of @xmath40 , and ( _ ii _ ) positive values of @xmath40 , despite the magnitude of @xmath20 , and therefore they do not represent black branes . for @xmath179 ,
the solutions present black branes with outer and inner horizons if @xmath180 , an extreme black hole if @xmath181 , and a naked singularity if @xmath182 .
we also computed the action and thermodynamic quantities of the @xmath11-dimensional rotating charged black brane such as the temperature , mass , angular momentum , entropy , charge and electric potential , and found that they satisfy the first law of thermodynamics .
it is worth to mention that the conserved quantities computed for when the boundary of spacetimes goes to infinity . finally , we performed a stability analysis in the canonical ensemble by considering @xmath166 for the rotating black brane solutions with infinite boundary in @xmath11-dimensional ebid gravity .
we found that there is no hawking - page phase transition in spite of the presence of nonlinear electromagnetic and dilaton fields for @xmath163 , while the solutions have an unstable phase for @xmath164 .
for @xmath164 , we found that there is always an upper limit @xmath183 for which @xmath166 is positive , provided @xmath184 .
it is worth to note that @xmath185 depend on the born - infeld parameter @xmath102 , the charge @xmath63 and the dimensionality of spacetime .
on the other hand , we found a low limit for born - infeld parameter @xmath172 , for which @xmath186 is positive provided @xmath187 .
it is notable to mention that @xmath172 decreases with increasing @xmath63 or @xmath170 , and it increases with increasing @xmath20 .
indeed , the solution is unstable for highly nonlinear electromagnetic field .
note that the @xmath11-dimensional charged rotating solutions obtained here have flat horizons .
thus , it would be interesting if one can construct charged rotating solutions in @xmath11 dimensions in the presence of dilaton and born - infeld fields with curved horizons .
fradkin e s and tseytlin a a , 1985 _ phys .
_ b * 163 * 123 [ spires ] + metsaev r r , rakhmanov m a and tseytlin a a , 1987 _ phys . lett .
_ b * 193 * 207 [ spires ] + bergshoeff e , sezgin e , pope c and townsend p , 1987 _ phys .
_ b * 188 * 70 [ spires ] garfinkle d , horowitz g t and strominger a , 1991 _ phys .
_ d * 43 * 3140 [ spires ] + gregory r and harvey j a , 1993 _ phys .
_ d * 47 * 2411 [ spires ] [ hep - th/9209070 ] + rakhmanov m a , 1994 _ phys .
_ d * 50 * 5155 [ spires ] [ hep - th/9310174 ] mignemi s and wiltshire d l , 1989 _ class .
quantum grav_. * 6 * 987 [ spires ] + wiltshire d l , 1991 _ phys .
_ d * 44 * 1100 [ spires ] + mignemi s and wiltshire d l , 1992 _ phys .
_ d * 46 * 1475 [ spires ] [ hep - th/9202031 ] horne j h and horowitz g t , 1992 _ phys .
_ d * 46 * 1340 [ spires ] [ hep - th/9203083 ] + shiraishi k , 1992 _ phys .
lett . _ a * 166 * 298 [ spires ] + ghosh t and mitra p , 2003 _ class .
quantum grav . _ * 20 * 1403 [ spires ] [ gr - qc/0212057 ] + sheykhi a and riazi n , 2006 _ int .
* 45 * 2453 [ spires ] [ hep - th/0605072 ] callan c g , lovelace c , nappi c r and yost s a , 1988 _ nucl .
_ b * 308 * 221 [ spires ] + andreev o d and tseytlin a a , 1988 _ nucl .
_ b * 311 * 205 [ spires ] + leigh r g , 1989 _ mod . phys .
_ a * 4 * 2767 [ spires ] garcia a , salazar h and plebanski j f , 1984 _ nuovo .
cim . _ * 84 * 65 [ spires ] + demianski m , 1986 _ found .
* 16 * 187 + breton n , 2003 _ phys . rev .
_ d * 67 * 124004 [ spires ] [ hep - th/0301254 ] + gibbons g w and herdeiro c a r , 2001 _ class .
quantum grav . _ * 18 * 1677 [ spires ] [ hep - th/0101229 ] tamaki t , _ j. cosmol .
_ jcap05(2004)004 [ spires ] [ gr - qc/0310099 ] + fernando s and krug d , 2003 _ gen .
* 35 * 129 [ spires ] [ hep - th/0306120 ] + cataldo m and garcia a , 1999 _ phys .
_ b * 456 * 28 [ spires ] [ hep - th/9903257 ] + fernando s , 2006 _ phys . rev .
_ d * 74 * 104032 [ spires ] [ hep - th/0608040 ] tamaki t and torii t , 2000 _ phys .
_ d * 62 * 061501 [ spires ] [ gr - qc/0004071 ] + tamaki t and torii t , 2001 _ phys .
_ d * 64 * 024027 [ spires ] [ gr - qc/0101083 ] + clement g and galtsov d , 2000 _ phys .
_ d * 62 * 124013 [ spires ] [ hep - th/0007228 ] + yazadjiev s s , fiziev p p , boyadjiev t l and todorov m d , 2001 _ mod .
lett . _ a * 16 * 2143 [ spires ] [ hep - th/0105165 ] maldacena j , 1998 _ adv .
phys . _ * 2 * 231 [ spires ] [ hep - th/9711200 ] + witten e , 1998 _ adv .
_ * 2 * 253 [ spires ] [ hep - th/9802150 ] + aharony o , gubser s s , maldacena j , ooguri h and oz y , 2000 _ phys . rep .
_ * 323 * 183 [ spires ] [ hep - th/9905111 ] + balasubramanian v and kraus p , 1999 _ commun .
phys . _ * 208 * 413 [ spires ] [ hep - th/9902121 ] hunter c j , 1999 _ phys . rev . _ d * 59 * 024009 [ spires ] [ gr - qc/9807010 ] + hawking s w , hunter c j and page d n , 1999 _ phys . rev .
_ d * 59 * 044033 [ spires ] [ hep - th/9809035 ] + mann r b , 1999 _ phys .
_ d * 60 * 104047 [ spires ] [ hep - th/9903229 ] + mann r b , 2000 _ phys .
_ d * 61 * 084013 [ spires ] [ hep - th/9904148 ] | in this paper , we construct a new class of charged , rotating solutions of @xmath0-dimensional einstein - born - infeld - dilaton gravity with liouville - type potentials and investigate their properties .
these solutions are neither asymptotically flat nor ( anti)-de sitter . we find that these solutions can represent black brane , with inner and outer event horizons , an extreme black brane or a naked singularity provided the parameters of the solutions are chosen suitably .
we also compute temperature , entropy , charge , electric potential , mass and angular momentum of the black brane solutions , and show that these quantities satisfy the first law of thermodynamics .
we find that the conserved quantities are independent of the born - infeld parameter @xmath1 , while they depend on the dilaton coupling constant @xmath2 .
we also find the total mass of the black brane with infinite boundary as a function of the entropy , the angular momenta and the charge and perform a stability analysis by computing the heat capacity in the canonical ensemble .
we find that the system is thermally stable for @xmath3 independent of the values of the charge and born - infeld parameters , while for @xmath4 the system has an unstable phase . in the latter case ,
the solutions are stable provided @xmath5 and @xmath6 , where @xmath7 and @xmath8 depend on the charge and the dimensionality of the spacetime .
that is the solutions are unstable for highly nonlinear electromagnetic field or when the dilaton coupling constant is large . |
it is well known that every connected oriented closed 3-manifold @xmath0 contains a fibered knot . in other words
, @xmath0 admits an open book decomposition with connected binding .
the minimum genus @xmath1 of pages of all open book decompositions of @xmath0 is a fundamental invariant of @xmath0 .
for instance , @xmath2 if and only if @xmath3 .
morimoto @xcite started to study genus one fibered knots ( gof - knots ) in lens spaces , and baker ( * ? ? ?
* theorem 4.3 ) completely determined which lens space contains a gof - knot , that is , we already know when @xmath4 holds . however , computation of @xmath1 is difficult in general .
sakasai ( * ? ? ?
* remark 6.10 ) introduced a homological analogue @xmath5 of @xmath1 , which is roughly defined to be the minimum genus of surfaces @xmath6 whose complements are homologically @xmath7 $ ] .
precisely , @xmath5 is defined in terms of homology cobordisms or homologically fibered knots ( see definitions [ def : hc ] ) , and @xmath8 holds by definition .
the author was informed by sakasai the following sufficient condition for @xmath9 when @xmath10 is odd : @xmath11 or @xmath12 is a quadratic residue @xmath13 .
the purpose of this paper is to prove the following theorem which enables us to compute @xmath5 for various 3-manifolds @xmath0 ( see corollary [ cor : list_of_hc ] ) .
[ thm : hc=1 ] @xmath9 holds for any lens space @xmath14 , or equivalently @xmath14 contains a genus one homologically fibered knot .
it is shown in ( * ? ? ?
* remark 6.10 ) that @xmath5 depends only on the isomorphism class of the pair of @xmath15 and the torsion linking form @xmath16 , where @xmath17 is the torsion subgroup of @xmath18 .
the following example shows that @xmath5 is not determined only by the isomorphism class of @xmath15 . for two 3-manifolds @xmath19 ( @xmath20 )
whose torsion linking forms @xmath21 are not isomorphic , we have @xmath22 ( ( * ? ? ?
* remark 6.10 ) ) . in order to prove theorem [ thm : hc=1 ] ,
we should find a surface @xmath23 of genus one whose complement @xmath24)$ ] is a homology cobordism .
( it is easy to see that @xmath25 if and only if @xmath0 is an integral homology 3-sphere . ) the following theorem ( to be proved in section [ sec : primedivmod ] by using the chebotarev density theorem ) and a well - known fact about binary quadratic forms allow us to construct a desired surface @xmath6 . [
thm : primedivmod ] let @xmath26 and @xmath27 be coprime .
then there exist @xmath28 and an odd prime @xmath29 such that the congruent equation @xmath30 is solvable and @xmath31 .
the study mentioned here is related to goda and sakasai s work @xcite on homologically fibered knots .
we discuss a homologically fibered knot ( in a rational homology 3-sphere ) and its alexander polynomial in the last section . in this paper
, @xmath14 denotes the lens space obtained from @xmath32 by dehn surgery on an unknot along the slope @xmath33 , where @xmath34 and @xmath10 are coprime .
the author would like to thank takuya sakasai and gwnal massuyeau for their various discussion .
also , he would like to express his gratitude to mutsuro somekawa and ippei nagamachi for their useful comments to prove theorem [ thm : primedivmod ] .
he wishes to be grateful to institut de recherche mathmatique avance , universit de strasbourg , where most of this paper was written , for the hospitality .
finally , this work was supported by the program for leading graduate schools , mext , japan and jsps kakenhi grant number 16j07859 .
we first review homology cobordisms following garoufalidis and levine ( * ? ? ? * section 2.4 ) .
a _ homology cobordism over @xmath35 _ is a triad @xmath36 , where @xmath37 is an oriented compact 3-manifold and @xmath38 are embeddings satisfying * @xmath39 is orientation - preserving and @xmath40 is orientation - reversing ; * @xmath41 ; * @xmath42 and @xmath43 ; * the induced maps @xmath44 are isomorphisms .
note that the fourth condition is equivalent to the condition that @xmath37 is connected and @xmath45 induce isomorphisms on @xmath46 .
sakasai ( * ? ? ?
* definition 6.9 , remark 6.10 ) introduced the following invariant of 3-manifolds by using homology cobordisms .
[ def : hc ] for a connected oriented closed 3-manifold @xmath0 , @xmath47 is defined by @xmath48 where @xmath49 is the _ closure _ of a homology cobordism @xmath36 over @xmath35 defined by @xmath50 for an embedding @xmath51 , we obtain the triad @xmath52),\iota^-,\iota^+)$ ] , where @xmath53 . whether this triad is a homology cobordism or not depends only on the image of @xmath54 , and we simply say that @xmath55)$ ] is a homology cobordism if the triad is so . and unknot in @xmath32 , where @xmath56 . the box with @xmath57 ( resp .
@xmath58 ) represents suitable @xmath59 positive ( resp .
negative ) full twists on the right . ]
the following key lemma is a corollary of proposition [ prop : seifertmatrix ] , though we give a direct proof in this section .
note that for coprime integers @xmath60 we can choose @xmath61 such that @xmath62 , then the others are written as @xmath63 , @xmath64 for @xmath65 .
[ lem : sabcuv ] the complement @xmath66)$ ] of the surface @xmath67 illustrated in figure [ fig : sabcuv ] is a homology cobordism if and only if there exist @xmath28 and @xmath65 such that @xmath68 it follows from the mayer - vietoris sequence for @xmath69 ) = ( s^3 \setminus { \operatorname{int}}(\sigma\times[-1,1 ] \cup n(u ) ) ) \cup_{\partial n(u ) } n(u)\ ] ] that @xmath70 , where @xmath71 denotes a ( closed ) tubular neighborhood of the unknot @xmath72 and @xmath54 is the map @xmath73 by the definition of dehn surgery , the matrix of @xmath54 with respect to the bases @xmath74 , @xmath75 is @xmath76 where @xmath77 , @xmath78 denote a meridian and longitude on the torus @xmath79 respectively .
suppose that the complement is a homology cobordism , namely @xmath80 is a basis of @xmath81 for both @xmath82 , where @xmath83 , @xmath84 .
then @xmath80 extends to a basis of @xmath85 .
that is , there is a matrix @xmath86 changing the basis @xmath75 to a basis containing @xmath87 .
we may assume that @xmath88 is of the form @xmath89 , where @xmath90 since the @xmath91-matrix at the bottom of the new matrix @xmath92 of @xmath54 must belong to @xmath93 , the absolute value of the @xmath94-entry of @xmath92 equals 1 .
therefore , we may assume @xmath95 then we have @xmath96 it follows from @xmath97 that @xmath98 this completes one direction , and the other is shown by reversing the above argument .
let us prove the main theorem by using lemma [ lem : sabcuv ] , theorem [ thm : primedivmod ] ( to be proved later ) and the following fact ( see , for example , @xcite ) : for @xmath99 , the congruence equation @xmath100 is solvable if and only if there is a binary quadratic form @xmath101 with discriminant @xmath102 such that @xmath103 has a primitive solution .
the case @xmath104 can be seen easily .
suppose @xmath105 . by setting @xmath106 , @xmath107 in theorem [ thm : primedivmod ] ,
we conclude that there are @xmath28 and @xmath108 such that @xmath109 has a solution @xmath110 . here , @xmath111 satisfies @xmath112 . by the above fact
, there is a quadratic form @xmath113 such that @xmath114 and @xmath115 has a solution @xmath116 .
then , integers @xmath117 , @xmath118 and @xmath119 satisfy @xmath120 therefore , we conclude from lemma [ lem : sabcuv ] that @xmath9 .
it follows from the proof of the above fact about quadratic forms that integers @xmath121 are represented by @xmath122 .
however , it seems difficult to represent these integers explicitly by @xmath60 .
we sum up values or estimates of @xmath5 for some @xmath0 as a corollary .
let @xmath123 be the minimum number of generators of a group @xmath124 .
[ cor : list_of_hc ] the following hold for @xmath125 and @xmath126 .
1 . @xmath25 if and only if @xmath127 .
@xmath128 if @xmath15 is isomorphic to @xmath129 or @xmath130 .
3 . @xmath131 if @xmath132 .
4 . if @xmath0 is a rational homology 3-sphere and the subgroup of @xmath15 consisting of 2-torsions is cyclic possibly trivial , then @xmath133 . 5 .
@xmath134 if @xmath135 divides @xmath136 and @xmath137 is not a quadratic residue @xmath138 .
suppose @xmath139 .
then , @xmath131 if there is @xmath140 such that the torsion linking form @xmath141 is isomorphic to @xmath142 and @xmath10 or @xmath143 is a quadratic residue @xmath144
. otherwise , @xmath145 . combining ( 4 ) and well - known inequalities , under the assumption of ( 4 )
, we have @xmath146 where @xmath147 denotes the heegaard genus of @xmath0 . on the other hand , @xmath8 and
@xmath148 hold for any @xmath149 by definition , and @xmath150 is found in ( * ? ? ?
* remark 6.1 ) .
\(1 ) is obvious , ( 2 ) is due to sakasai @xcite , and ( 3 ) is a direct consequence of theorem [ thm : hc=1 ] since @xmath141 is isomorphic to @xmath151 for some @xmath10 .
we next prove ( 4 ) . since @xmath15 is finite , it is isomorphic to @xmath152 for some @xmath153 s with @xmath154 , where @xmath155
. it follows from ( * ? ? ?
* theorem ( 4 ) ) that @xmath141 is isomorphic to @xmath156 for some @xmath157 s , hence @xmath141 is isomorphic to @xmath158 .
therefore , theorem [ thm : hc=1 ] shows @xmath159 .
the proofs of ( 5 ) and ( 6 ) are given in section [ sec : hfk ] .
the goal of this section is to prove theorem [ thm : primedivmod ] which is used in the proof of theorem [ thm : hc=1 ] .
we briefly review the artin symbol of a prime ideal only for the case of abelian extensions following ( * ? ? ?
* chapter x , section 1 ) and ( * ? ? ? * chapter vi , section 7 ) .
let @xmath160 be an algebraic number field , @xmath161 be an abelian extension , namely it is a finite galois extension whose galois group @xmath162 is abelian .
let @xmath163 be an unramified prime ideal of @xmath164 , the ring of integers of @xmath160 , let @xmath165 be a prime ideal of @xmath166 lying above @xmath167 , that is , @xmath168 .
then there exists a unique element of the decomposition group @xmath169 of @xmath165 satisfying @xmath170 for all @xmath171 , where @xmath172 \in { \mathbb{z}}_{\geq1}$ ] .
this automorphism is independent of the choice of @xmath165 , which is denoted by @xmath173 and called the _ artin symbol _ of @xmath167 .
[ ex : artinsymbol ] we review two well - known examples used in this paper .
let @xmath174 be not a perfect square , @xmath175 be the discriminant of @xmath176 , @xmath177 be a prime with @xmath178
. then , @xmath179 is unramified and we deduce @xmath180 & \text{otherwise . }
\end{cases}\ ] ] next , for a primitive @xmath59th root of unity @xmath181 and a prime @xmath177 with @xmath182 , the ideal @xmath179 is unramified and @xmath183 $ ] holds .
[ lem : chebotarev ] let @xmath184 , @xmath185 be abelian extensions , @xmath186 be a subset of @xmath187 with @xmath188 , where @xmath189 denotes @xmath190 and @xmath191 is defined by @xmath192 .
then the set @xmath193 is infinite .
first note that @xmath194 is a galois extension with @xmath195 abelian , and the homomorphism @xmath196 is injective .
we define the set @xmath197 by @xmath198 by the chebotarev density theorem ( see , for example , ( * ? ? ?
* chapter viii , theorem 10 ) , ( * ? ? ?
* chapter vii , theorem 13.4 ) ) , we have @xmath199 , where @xmath200 is the dirichlet density of @xmath197 ( see ( * ? ? ?
* chapter viii , section 4 ) , ( * ? ? ?
* chapter vii , section 13 ) ) .
hence , @xmath197 is infinite . on the other hand , the consistency property ( * ? ? ?
* chapter x , section 1 ) asserts @xmath201 , and thus @xmath202 is also infinite .
[ lem : quadresmod ] for @xmath203 with @xmath204 and @xmath205 , the set @xmath206 is infinite .
put @xmath207 , @xmath208 , @xmath209 and @xmath210)\}$ ] . since @xmath211 ,
the map @xmath54 in lemma [ lem : chebotarev ] is an isomorphism , and thus @xmath188 .
it follows from lemma [ lem : chebotarev ] that @xmath212 $ } \right\}\ ] ] is infinite . here
, example [ ex : artinsymbol ] implies that this set is included in @xmath202 .
[ lem : weintraub ] for a positive integer @xmath213 , @xmath214 or @xmath215 does not belong to the @xmath59th cyclotomic field @xmath216 .
the cases @xmath217 are obvious . for @xmath218 , we consider four cases : ( i ) @xmath219 , ( ii ) @xmath220 , ( iii ) @xmath221 , ( iv ) @xmath222 , where @xmath223 is the 2-adic valuation for @xmath224 . we only discuss ( i ) , and the other cases are shown similarly . in general , @xmath225 if and only if @xmath174 is a product of some integers in @xmath226 ( see , for example , ( * ? ? ?
* corollary 4.5.4 ) ) . since @xmath227
, both @xmath228 and @xmath229 must be perfect squares , though that is impossible except @xmath230 .
it follows from lemmas [ lem : quadresmod ] and [ lem : weintraub ] that there are @xmath28 and an odd prime @xmath29 such that @xmath231 is a quadratic residue @xmath232 and @xmath31 .
therefore , by @xmath233 , @xmath231 is a quadratic residue @xmath232 if and only if @xmath234 is solvable .
moreover , this congruence equation is equivalent to @xmath30 .
the above proof ( and the case @xmath104 in the proof of theorem [ thm : hc=1 ] ) claims that for any @xmath235 there exists @xmath140 such that @xmath14 is homeomorphic to @xmath235 and sakasai s sufficient condition in section [ sec : intro ] is true for @xmath10 . even if @xmath230 , theorem [ thm : primedivmod ] holds for @xmath236
indeed , one can choose @xmath237 respectively .
however , in the case @xmath238 , theorem [ thm : primedivmod ] fails since neither @xmath239 nor @xmath240 is a quadratic residue @xmath241 by the quadratic reciprocity law .
we discuss homologically fibered knots which are defined in terms of homology cobordisms .
seifert matrices of connected oriented surfaces in a rational homology 3-sphere @xmath242 play a key role in this section .
let @xmath243 be an embedding and @xmath244 be the seifert matrix of @xmath6 with respect to @xmath245 ( see figure [ fig : gammaonsigma ] ) , that is , @xmath246 , where @xmath247 .
note that , in the next proposition , @xmath202 can not be regarded as the matrix of @xmath248 unlike in ( * ? ? ?
* proposition 3.2 ) since @xmath249 in general .
suppose that @xmath253 is a homology cobordism , where @xmath254 $ ] .
then one can consider the matrix @xmath255 of the automorphism @xmath256 with respect to @xmath257,\dots,[\gamma_{2g}]\}$ ] and check that @xmath258 similar to the case @xmath259 .
we have @xmath260 hence @xmath251 is equivalent to @xmath261 , and it suffices to show that @xmath262.\ ] ] the inclusion @xmath263 follows from @xmath264 .
next , for any @xmath265 let @xmath266 be an immersed surface satisfying @xmath267=y$ ] and @xmath268 .
then @xmath269 consists of two types of closed curves : curves coming from @xmath270 and ones coming from @xmath271 .
let @xmath272 be the class of the curves coming from @xmath273 .
we have @xmath274 = 0 \in h_1(y\setminus{\operatorname{int}}\widetilde{\sigma}).\ ] ] therefore , @xmath275 .
we next prove the converse .
first note that @xmath276 claims that @xmath277 are injective , and thus there is a unique endomorphism @xmath278 of @xmath279 satisfying @xmath280 .
this equality implies @xmath281 , hence we have the commutative diagram @xmath282 & b \ar[r]\ar[d]_-{(\iota^\pm)_\ast| } & h_1(\sigma_{g,1 } ) \ar[r]\ar[d]^-{(\iota^\pm)_\ast } & h_1(\sigma_{g,1})/b \ar[r]\ar[d]^-{\overline{\iota_\ast } } & 0\ ( \text{exact})\\ 0 \ar[r ] & { \operatorname{im}}\partial_\ast \ar[r]^-{{\operatorname{incl}}}\ar@{-->}[ru ] & h_1(y\setminus{\operatorname{int}}\widetilde{\sigma } ) \ar[r ] & h_1(y ) \ar[r ] & 0\ ( \text{exact } ) . } \ ] ] here , if one shows that the map @xmath283 lifts to a map @xmath284 \dashrightarrow h_1(\sigma_{g,1}),\ ] ] then @xmath285 are isomorphisms , hence @xmath286 is injective .
finally , the assumption @xmath251 says that @xmath286 is an isomorphism , and so are @xmath277 by the five lemma . for any @xmath287 , the excision isomorphism @xmath288 asserts that there is an immersed surface @xmath266 in @xmath289 satisfying @xmath290 and @xmath291=y$ ] .
let @xmath292 ( resp .
@xmath293 ) be the class of the curves @xmath294 ( resp .
@xmath295 ) .
we see that @xmath296 , and therefore @xmath297 .
it is easy to see that the seifert matrix @xmath202 of @xmath67 with respect to @xmath298 is @xmath299 and we have @xmath300 therefore , if the complement is a homology cobordism , then proposition [ prop : seifertmatrix ] shows that there are @xmath301 and @xmath160 in lemma [ lem : sabcuv ] .
conversely , the existence of @xmath301 and @xmath160 implies that @xmath302 and @xmath303 , hence we have @xmath252 .
[ def : hfk ] an oriented knot @xmath305 in a connected oriented closed 3-manifold @xmath0 is called a _ homologically fibered knot of genus @xmath306 _ if there is a seifert surface of @xmath305 such that @xmath307)$ ] is a homology cobordism over @xmath35 .
we next see that the alexander polynomial @xmath310 of an oriented null - homologous knot @xmath305 in a rational homology 3-sphere @xmath242 gives a necessary condition for @xmath305 to be homologically fibered .
recall that @xmath310 is defined by @xmath311,\ ] ] where @xmath202 is a seifert matrix of a seifert surface @xmath6 of @xmath305 . by definition , @xmath310 should be palindromic and satisfy @xmath312 , and its breadth is less than or equal to @xmath313 . [
cor : alexanderpolynomial ] if an oriented knot @xmath305 in @xmath242 is homologically fibered , then the breadth of @xmath310 equals @xmath313 and the coefficient of the highest degree term equals @xmath317 up to sign .
[ rem : crossingchange ] let @xmath321 be a surface in @xmath0 obtained from @xmath322 by two crossing changes between @xmath323 and @xmath324 with opposite sign . then
a mayer - vietoris argument shows that if the complement of @xmath6 is a homology cobordism , then so is the complement of @xmath321 .
\(5 ) for @xmath325 with @xmath326 , by theorem [ thm : hc=1 ] , we know that @xmath5 equals 1 or 2
. remark [ rem : crossingchange ] allows us to find a surface of the form @xmath327 ( see figure [ fig : sabcuuvv ] ) whose complement is a homology cobordism .
the seifert matrix @xmath202 of @xmath327 is given by @xmath328 and it follows from proposition [ prop : seifertmatrix ] that @xmath329 thus , @xmath137 is a quadratic residue @xmath138 .
\(6 ) we may assume that @xmath330 . by theorem [ thm : hc=1 ]
, it suffices to prove that @xmath131 if and only if @xmath10 or @xmath143 is a quadratic residue @xmath144 .
suppose @xmath131 .
there is @xmath331 such that the closure @xmath186 of the homology cobordism @xmath332,{\operatorname{incl}},\widetilde{a})$ ] is borromean surgery equivalent to @xmath0 , where @xmath333 is a homeomorphism of @xmath334 inducing @xmath335 on @xmath336 ( see ( * ? ? ?
* section 6.3 ) ) .
here we have @xmath337 since the mayer - vietoris sequence for @xmath338 \cup \sigma_{1,1}\times[0,1]$ ] shows @xmath339 . therefore , by the proof of ( * ? ? ? * proposition 2 ) , @xmath335 is conjugate to @xmath340 in @xmath341 for some @xmath342 .
now , @xmath343 must be @xmath344 for some @xmath28 , then @xmath333 can be chosen so that @xmath345 .
thus the torsion linking form @xmath142 is isomorphic to @xmath346 .
s. garoufalidis and j. levine .
tree - level invariants of three - manifolds , massey products and the johnson homomorphism . in _ graphs and patterns in mathematics and theoretical physics _ , volume 73 of _ proc . sympos .
pure math .
_ , pages 173203 .
soc . , providence , ri , 2005 .
j. neukirch . , volume 322 of _ grundlehren der mathematischen wissenschaften [ fundamental principles of mathematical sciences]_. springer - verlag , berlin , 1999 . translated from the 1992 german original and with a note by norbert schappacher , with a foreword by g. harder | we prove that every lens space contains a genus one homologically fibered knot , which is contrast to the fact that some lens spaces contain no genus one fibered knot . in the proof , the chebotarev density theorem and binary quadratic forms in number theory play a key role .
we also discuss the alexander polynomial of homologically fibered knots . |
the following is a brief description of fredholm operators . for more details ,
see @xcite .
a bounded operator @xmath0 on a separable hilbert space is _ fredholm _ if there exists a bounded operator @xmath1 such that @xmath2 and @xmath3 are compact .
the fredholm index is defined by @xmath4 the simplest example of a fredholm operator with nonzero index is the unilateral shift operator : let @xmath5 be the canonical basis for the hilbert space @xmath6 , and let the operator @xmath7 act by @xmath8 the reason for denoting the unilateral shift operator by @xmath7 is its similarity to the harmonic oscillator lowering operator .
the adjoint of @xmath7 acts by @xmath9 since @xmath10 , @xmath7 is fredholm .
the kernel of @xmath7 is 1-dimensional and the kernel of @xmath11 is 0-dimensional .
thus @xmath12 and @xmath13 .
although neither the dimension of @xmath14 nor that of @xmath15 is stable under deformations of @xmath0 , the index _ is _ stable . for any compact operator @xmath16 , for any bounded operator @xmath1 , and for @xmath17 sufficiently small , @xcite : @xmath18
the following theorem is standard : if @xmath19 are fredholm operators , then the product @xmath20 is also fredholm , and @xmath21 .
if @xmath0 and @xmath22 are fredholm operators on the same hilbert space , then there is a continuous path of fredholm operators from @xmath0 to @xmath22 if and only if @xmath23 .
( by continuous , we mean relative to the operator norm ) .
put another way , the path components of @xmath24 , the space of fredholm operators on @xmath25 , are indexed by the integers .
the @xmath26-th path component is precisely the set of fredholm operators of index @xmath26 @xcite .
our main concern in this paper is the following problem : suppose one interpolates between fredholm operators with different indices .
what can one say about the way the indices change ?
another way of phrasing this is : what is the phase diagram of fredholm indices ?
the answer to this question depends on the choice of the embedding space . in the space of bounded operators ,
the `` phases''each labeled by its index are open sets .
but the boundary between phases , as we shall explain , is rather wild : a point on the boundary of one phase is also on the boundary of _ every other _ phase .
this behavior is difficult to visualize .
another class of embedding spaces that we consider is associated with toeplitz operators with various regularity assumptions on a class of functions . here , at least if the functions are sufficiently smooth , the boundaries between phases have a simple structure and the phase diagrams satisfy simple rules that have the flavor of gibbs phase rule @xcite .
typical of our results is the statement that under appropriate conditions , phases whose indices differ by one have a common boundary whose codimension is one , and phases whose indices differ by two meet on a set of codimension two etc .
fig.[shifts ] is an example of one of the phase diagrams we obtain .
theories of the quantum hall effect are roughly of two kinds : those that focus on the bulk of the hall and those that focus on the edge @xcite .
it was pointed out by @xcite that the bulk - edge duality is an illustration of the _ holographic principle_. in either approach , the quantized hall conductance can be related to a fredholm index .
it is common knowledge that the hall conductance can be identified with a chern number @xcite . for non - interacting electrons in two dimensions ,
this result is a special case of the fact that the hall conductance is a fredholm index .
since this is not common knowledge , we recall how chern numbers and fredholm indices are related . for non - interacting electrons in two dimensions with the fermi energy in a gap , tkn@xmath27 , showed that the hall conductance for landau hamiltonians with _ periodic _ potential , is related to a chern number @xcite .
the ( magnetic ) brillouin zone associated with the periodicity plays an role in this theory . because of this , the interpretation of the hall conductance as a chern number does not carry over to random or even quasi - periodic potentials nor to `` irrational magnetic fields '' , all of which have no ( classical ) brillouin zone .
although the quantization of the hall conductance can be established in these cases by a limiting argument @xcite , the interpretation as a chern number does not survive .
j. bellissard @xcite , in a work that had impact on non - commutative geometry @xcite , showed that the hall conductance with _
potential , be it periodic , quasi - periodic or random , and real magnetic field , rational or not , is a fredholm index .
this result was derived in @xcite without using non - commutative geometry .
more precisely , consider the ( infinite dimensional ) spectral projection @xmath28 on the states below the fermi energy @xmath29 for the one particle hamiltonian in the plane .
let @xmath30 be the multiplication operator @xmath31 , where @xmath32 is the usual polar angle in the plane .
@xmath30 is a singular gauge transformation that introduces an aharonov - bohm flux tube at the origin of the euclidean plane .
the hall conductance is the fredholm index of @xmath33 thought of as an operator on the range of @xmath28 @xcite .
since the fredholm index does not need a brillouin zone , this approach offers a natural framework that accounts for the quantization and stability of the hall conductance .
finite quantum hall systems have chiral edge currents @xcite . consider the case that the boundary is a circle of circumference @xmath34 .
the dispersion relation of the edge states is approximately linear in a small neighborhood of the fermi energy and the hamiltonian for a single edge channel , with velocity @xmath35 , is @xmath36 now , the projection @xmath28 is associated with the occupied edge states , @xmath37 with @xmath38 . introducing a flux tube into the system
is associated with the unitary @xmath39 and sends @xmath40 .
this leads to the spectral flow of the edge states .
@xmath33 is the unilateral shift operator @xmath7 and the number of edge states that cross the fermi energy is @xmath41 . by an argument of halperin @xcite this is also the hall conductance .
an extension of this idea to harper models with an edge is described in @xcite .
we begin with the space of bounded operators with the topology defined by the operator norm , and we wish to understand the phase diagram of a generic family of such operators . as we shall explain , the phase diagram in the entire space is quite wild : any point on the boundary of the `` index = @xmath42 '' phase is also on the boundary of every other phase . to understand this bizarre behavior ,
recall that the zero operator ( which is _ not _ a fredholm operator ) is on the boundary of every phase : zero is the limit , as @xmath43 , of @xmath44 , with @xmath7 of eq .
( [ shift ] ) , for any @xmath26 .
the point of the theorem is that similar behavior occurs at all boundary points .
let @xmath45 be the set of fredholm operators of index @xmath26 .
every point on the boundary of @xmath45 is also on the boundary of @xmath46 , for every integer @xmath47 .
proof : let @xmath48 be a ( not fredholm ) operator on the boundary of @xmath45 . given @xmath49
, we must find an operator in @xmath46 within a distance @xmath17 of @xmath48 .
suppose that the kernel and cokernel of @xmath48 are infinite dimensional , and that there is a gap in the spectrum of @xmath50 at zero .
( if this is not the case , we may perturb @xmath48 by an arbitrarily small amount to make it so ) .
now let @xmath1 be a unitary map from the kernel of @xmath48 to the cokernel .
let @xmath51 be the orthogonal projection onto @xmath52 , and let @xmath7 be a shift operator on @xmath53 .
for each @xmath54 , @xmath55 has a bounded right inverse @xmath56 it follows that the cokernel of @xmath57 is empty .
it is easy to see that the kernel of @xmath57 is @xmath47 dimensional hence @xmath58 .
similarly , @xmath59 has index @xmath60 . @xmath61
in this section and the next we show that there are interesting and simple `` generic '' phase diagrams of fredholm indices in some finite dimensional spaces , and in some infinite - dimensional spaces with sufficiently fine topologies .
we shall also see also how control is lost as the space is enlarged and the topology is coarsened .
we begin by considering linear combinations of the shift operator @xmath7 and the identity operator 1 .
that is , we consider the operator @xmath62 where @xmath63 and @xmath64 are constants . [ 0 - 1 ] if @xmath65 , then @xmath48 is fredholm .
the index of @xmath48 is 1 if @xmath66 and zero if @xmath67 .
if @xmath68 , then @xmath48 is not fredholm . proof : first suppose @xmath69 . then @xmath48 is invertible : @xmath70 as the sum converges absolutely .
thus @xmath48 has neither kernel nor cokernel , and has index zero . if @xmath71 , then the kernel of @xmath48 is 1-dimensional , namely all multiples of @xmath72 , where @xmath73 . notice how the norm of @xmath74 goes to infinity as @xmath75
however , @xmath76 has no kernel , since for any unit vector @xmath77 , @xmath78 .
thus the index of @xmath48 is 1 . if @xmath79 , then @xmath48 is at the boundary between index 1 and index 0 , and so can not be fredholm.@xmath61 next we consider linear combinations of @xmath80 up to some fixed @xmath81 .
that is , we consider operators of the form @xmath82 this is closely related to the polynomial @xmath83 if none of the roots of @xmath84 lie on the unit circle , then @xmath48 is fredholm , and the index of @xmath48 equals the number of roots of @xmath84 _ inside _ the unit circle , counted with multiplicity . if any of the roots of @xmath84 lie _ on _ the unit circle , then @xmath48 is not fredholm .
proof : the polynomial @xmath85 factorizes as @xmath86 , where @xmath42 is the degree of @xmath84 ( typically @xmath87 , but it may happen that @xmath88 ) .
but then @xmath89 . if none of the roots @xmath90 lie on the unit circle , then each term in the product is fredholm , so the product is fredholm , and the index of the product is the sum of the indices of the factors . by theorem [ 0 - 1 ] , this exactly equals the number of roots @xmath90 inside the unit circle .
if any of the roots lie on the unit circle , then a small perturbation can push those roots in or out , yielding fredholm operators with different indices .
this borderline operator therefore can not be fredholm.@xmath61 the last theorem easily generalizes to linear combination of left - shifts and right - shifts .
the index of an operator @xmath91 equals the number of roots of @xmath92 inside the unit circle , minus the degree of the pole at @xmath93 ( that is @xmath47 , unless @xmath94 ) .
this follow from the fact that @xmath95 since there is no qualitative difference between combinations of left - shifts and combinations of both left- and right - shifts , we restrict our attention to left - shifts only , and consider families of operators of the form ( [ a ] ) .
[ finite - combinations ] in the space of complex linear combinations of 1 , @xmath7 ,
, @xmath81 , almost every operator is fredholm . for every @xmath96 ,
the points where the index can jump by @xmath42 ( by which we mean the common boundaries of regions of fredholm operators whose indices differ by @xmath42 ) is a set of real codimension @xmath42 . in the space of real linear combinations of 1 , @xmath7 ,
, @xmath81 , almost every operator is fredholm . for every @xmath96 , the points where the index jumps by @xmath42 is a stratified space , the largest stratum of which has real codimension @xmath97 , where @xmath98 denotes the integer part of @xmath99 .
proof : our parameter space is the space of coefficients @xmath100 , or equivalently the space of polynomials of degree @xmath101 .
this is either @xmath102 or @xmath103 , depending on whether we allow real or complex coefficients . in either case ,
the set @xmath104 of fredholm operators of index @xmath42 is identical to the set of polynomials with @xmath42 roots inside the unit circle and the remaining @xmath105 roots outside ( if @xmath88 , we say there is a root at infinity ; if @xmath106 , there is a double root at infinity , and so on .
counting these roots at infinity , there are always exactly @xmath26 roots in all . )
the boundary of @xmath104 is the set of polynomials with at most @xmath42 roots inside the unit circle , at most @xmath105 outside the unit circle , and at least one root on the unit circle .
( strictly speaking , the zero polynomial is also on this boundary .
this is of such high codimension that it has no effect on the phase portrait we are developing . ) .
we consider the common boundary of @xmath104 and @xmath107 . if @xmath108 , a nonvanishing polynomial is on the boundary of both @xmath104 and @xmath107 if it has at most @xmath42 roots inside the unit circle and at most @xmath109 roots outside .
it must therefore have at least @xmath110 roots on the unit circle .
if we are working with complex coefficients , this is a set of codimension @xmath110 .
the roots themselves , together with an overall scale @xmath111 , can be used to parametrize the space of polynomials . for each root , being on the unit circle is codimension 1 , while being inside or outside are open conditions .
since the roots are independent , placing @xmath110 roots on the unit circle is codimension @xmath110 .
if we are working with real coefficients , the roots are not independent , as non - real roots come in complex conjugate pairs .
thus , the common boundary of @xmath104 and @xmath107 breaks into several strata , depending on how many real roots and how many complex conjugate pairs lie on the unit circle .
if @xmath110 is even , the biggest stratum consists of having @xmath112 pairs , and has codimension @xmath112 .
if @xmath110 is odd , the biggest stratum consists of having @xmath113 pairs and one real root on the unit circle , and has codimension @xmath114 .
-1 in -1.3in=3.0truein theorem [ finite - combinations ] is illustrated in figure [ shifts ] , where the phase portrait is shown for @xmath115 with real coefficients , with @xmath116 fixed to equal 1 .
the points above the parabola @xmath117 have complex conjugate roots , while points below have real roots .
notice that the transition from index 2 to index 0 occurs at an isolated point when the roots are real , but on an interval when the roots come in complex - conjugate pairs .
it is clear that an almost identical theorem applies to linear combinations of left - shifts up to @xmath81 and right - shifts up to @xmath118 .
the results are essentially independent of @xmath26 and @xmath47 ( their only effect being to limit the size of possible jumps to @xmath119 ) .
we can therefore extend the results to the space of all ( finite ) linear combinations of left- and right - shifts , which is topologized as the union over all @xmath26 and @xmath47 of the spaces considered above .
our result , restated for that space , is [ limit - space ] in the space of finite complex linear combinations of left- and right - shifts of arbitrary degree , almost every operator is fredholm . for every integer @xmath120 ,
the points where the index can jump by @xmath42 ( by which we mean the common boundaries of regions of fredholm operators whose indices differ by @xmath42 ) is a set of real codimension @xmath42 .
if we restrict the coefficients to be real , then , for every @xmath96 , the points where the index jumps by @xmath42 is a stratified space , the largest stratum of which has real codimension @xmath121 .
although theorem [ limit - space ] refers to an infinite - dimensional space , this space is still extremely small each point is a _
finite _ linear combination of shifts . in this section
we consider _ infinite _ linear combinations of shifts .
this is equivalent to studying toeplitz operators .
the _ hardy space _
@xmath25 is the subspace of @xmath122 consisting of functions whose fourier transforms have no negative frequency terms .
equivalently , if we give @xmath122 a basis of fourier modes @xmath123 , where the integer @xmath26 ranges from @xmath124 to @xmath125 , then @xmath25 is the closed linear span of @xmath126 .
we think of @xmath127 as sitting in the complex plane , with @xmath128 .
now let @xmath129 be a bounded , measurable function on @xmath127 , and let @xmath28 be the orthogonal projection from @xmath122 to @xmath25 . if @xmath130 , then @xmath131 ( pointwise product ) is in @xmath122 , and @xmath132 .
we define the operator @xmath133 by @xmath134 an operator of the form ( [ toep ] ) is called a _
toeplitz operator_. we call a toeplitz operator @xmath133 continuous if the underlying function @xmath135 is continuous , and apply the terms differentiable " , smooth " and analytic " similarly . : toeplitz operators can be represented by semi - infinite matrices that have constant entries on diagonals , and the various classes we have defined correspond to the decay away from the main diagonal .
notice that @xmath136 so @xmath137 is simply a shift by @xmath47 , a right shift if @xmath138 and a left - shift if @xmath139 .
all our results about shifts can therefore be understood in the context of toeplitz operators .
theorem [ finite - combinations ] refers to operators @xmath133 , where @xmath135 is a polynomial in @xmath140 of limited degree .
theorem [ limit - space ] considers polynomials or arbitrary degree in @xmath141 and @xmath140 .
we will see that the results carry over to analytic functions on an annulus around @xmath127 , and to a lesser extent to @xmath142 toeplitz operators , but with results that weaken as @xmath42 is decreased . here
are some standard results about toeplitz operators . for details , see @xcite .
a @xmath143 toeplitz operator @xmath133 is fredholm if and only if @xmath135 is everywhere nonzero on the unit circle . in that case
the index of @xmath133 is minus the winding number of @xmath135 around the origin , namely @xmath144 given the first half of the theorem , the equality of index and winding number is easy to understand .
we simply deform @xmath135 to a function of the form @xmath145 , while keeping @xmath135 nonzero on all of @xmath127 throughout the deformation ( this is always possible , see e.g. @xcite ) . in the process of deformation , neither the index of @xmath133 nor the winding number of @xmath135 can change , as they are topological invariants . since the winding number of @xmath146 is @xmath26 , and since @xmath147 ( if @xmath148 , @xmath149 otherwise ) , which has index @xmath150 , the result follows .
we now consider functions @xmath135 on @xmath127 that can be analytically continued ( without singularities ) to an annulus @xmath151 , where the radii @xmath152 and @xmath153 are fixed .
this is equivalent to requiring that the fourier coefficients @xmath154 decay exponentially fast , i.e. that the sum @xmath155 converges .
for now we do not impose any reality constraints or other symmetries on the coefficients @xmath154 .
this space of functions is a banach space , with norm given by the sup norm on the annulus .
this norm is stronger than any sobolev norm on the circle itself .
the analysis of the corresponding toeplitz operators is straightforward and similar to the proof of theorem [ finite - combinations ]
. since @xmath135 has no poles in the annulus , we just have to keep track of the zeroes of @xmath135 . for the index of @xmath133 to change , a zero of @xmath135
must cross the unit circle . for the index to jump from @xmath42 to @xmath156 , @xmath157 zeroes must cross simultaneously . in the absence of symmetry , the locations of the zeroes are independent and
can be freely varied , so this is a codimension-@xmath157 event .
if we impose a reality condition : @xmath158 , then zeroes appear only on the real axis or in complex conjugate pairs .
in that case , changing the index by 2 is merely a codimension-1 event . combining these observations we obtain [ analytic ] in the space of toeplitz operators that are analytic in a ( fixed ) annulus containing @xmath127 , almost every operator is fredholm . for every integer @xmath120 ,
the points where the index can jump by @xmath42 is a set of real codimension @xmath42 . if we impose a reality condition @xmath159 then , for every @xmath96
, the points where the index jumps by @xmath42 is a stratified space , the largest stratum of which has real codimension @xmath97 .
finally we consider toeplitz operators that are not necessarily analytic , but are merely @xmath160 times differentiable , and we use the @xmath161 norm .
our result is [ c^ell ] in the space of toeplitz @xmath161 operators , almost every operator is fredholm . for every integer @xmath42 with @xmath162 ,
the points where the index can jump by @xmath42 is a set of real codimension @xmath42 . for every integer @xmath163 ,
the points where the index can jump by @xmath42 is a set of real codimension @xmath164 . in other words ,
our familiar results hold up to codimension @xmath164 , at which point we lose all control of the change in index . proof : as long as @xmath135 is everywhere nonzero , @xmath133 is fredholm . to get a change in index , therefore , we need one or more points where @xmath135 , and possibly some derivatives of @xmath135 with respect to @xmath32 , vanish .
suppose then that for some angle @xmath165 , @xmath166 for some @xmath167 , but that the @xmath26-th derivative @xmath168 .
this is a codimension @xmath169 event , since we are setting the real and imaginary parts of @xmath26 variables to zero , but have a 1-parameter choice of points where this can occur . without loss of generality
, we suppose that this @xmath26-th derivative is real and positive . by making a @xmath161-small perturbation of @xmath135
, we can make the value of @xmath135 highly oscillatory near @xmath165 , thereby wrapping around the origin a number of times .
however , since a @xmath161-small perturbation does not change the @xmath26-th derivative by much , the sign of the real part of @xmath135 can change at most @xmath26 times near @xmath165 , so the argument of @xmath135 can only increase or decrease by @xmath170 or less .
the difference between these two extremes is @xmath171 , or a change in winding number of @xmath26 . to change the index by an integer @xmath47 , therefore
, we must have the function vanish to various orders at several points , with the sum of the orders of vanishing adding to @xmath47 .
the generic event is for @xmath135 ( but not @xmath172 ) to vanish at @xmath47 different points this is a codimension @xmath47 event , analogous to having @xmath47 zeroes of a polynomial cross the unit circle simultaneously at @xmath47 different points .
all other scenarios have higher codimension and are analogous to having 2 or more zeroes of the @xmath47 zeroes crossing the unit circle at the same point .
the situation is different , however , when the function @xmath135 and the first @xmath160 derivatives all vanish at a point @xmath165 .
then the higher - order derivatives are not protected from @xmath161-small perturbations and , by making such a perturbation , we can change @xmath135 into a function that is identically zero on a small neighborhood of @xmath173 . by making a further small perturbation , we can make @xmath135 wrap around the origin as many times as we like near @xmath173 . more specifically ,
if @xmath135 is zero on an interval of size @xmath174 , then , for small @xmath17 , @xmath175 will wrap around the origin approximately @xmath176 times near @xmath165 . by picking @xmath177 as large ( positive or negative ) as we wish
, we can obtain arbitrarily positive or negative indices . as long
as we take @xmath178 , this perturbation will remain small in the @xmath161 norm .
the results of this section can be extended , with minor modifications , to the algebra of matrix valued toeplitz operators @xcite where the index is related to the winding of the determinant of a matrix .
phase diagrams of quantum hall system describe the dependence of the hall conductance on parameters such as the magnetic field @xmath1 and the fermi energy @xmath179 .
there are three idealized models where the phase diagram can be computed explicitly : the landau hamiltonian in the euclidean plane , whose phase diagram is shown in fig.[landau ] ; the landau hamiltonian for the hyperbolic plane , whose phase diagram is shown in fig.[hyperbolic ] and harper models in the plane @xcite , whose phase diagram is associated with the hofstadter butterfly , shown in fig.[harper ] for the case of a tight binding model on a square lattice .
= 2.5truein -0.8 in -1.2 in = 4.5truein -0.4 in these are not models of toeplitz operators , and none of these models is generic , especially insofar as all of them have symmetries .
however , we consider the extent to which they follow the generic phase rules of ( smooth , complex ) toeplitz operators anyway . where these rules are not followed , we consider how a small generic perturbation might restore the rules . the phase diagram for the euclidean plane , fig .
[ landau ] satisfies the generic phase rules away from the line @xmath180 . on the line @xmath180
, however , the index takes an infinitely large jump , while at the origin infinitely many phases meet .
both are forbidden by the phase rules .
the phase diagram in the hyperbolic plane , fig .
[ hyperbolic ] , satisfies the generic phase rules outside the shaded parabolic region . in the shaded region ,
the operator is not fredholm and the index is not defined .
this is contrary to the phase rules since not being fredholm is expected to be a codimension 1 event . the phase diagram of the harper model , fig .
[ harper ] , is in serious conflict with the phase rule for ( smooth , complex ) toeplitz operators : it is known @xcite , that for a full measure of values of the magnetic field ( irrational , of course ) , the spectrum is a cantor set .
since the boundary between phases is contained in the spectrum , this suggests that any point on the boundary between any two phases can also be on the boundary between infinitely many other phases .
this is the sort of behavior we observed for bounded operators with no restrictions .
however , even in this wildness there is some regularity .
for example , the center of the figure is on the boundary of all phases with _ odd
_ indices while theorem 2 allows for even indices as well . to see how fig .
[ harper ] is obtained , we recall that for a tight - binding model with flux @xmath181 through a unit cell , the hall conductance , @xmath182 associated with the j - th gap , ( provided all gaps below it are open ) satisfies the diophantine equation @xcite @xmath183 a similar equation holds for gaps counted from above . in the harper model
it is known @xcite that all gaps except possibly for the central gap , are open . finally , consider the phase diagram of the harper model with a disordered potential .
this is not soluble in the same sense that the previous models are , but there are numerical results for it .
fig.[tanf ] , which we borrowed from @xcite , shows the phase diagram for a split landau level in the harper model with disorder .
more precisely , the diagram describes a harper model with fractional flux @xmath184 through a unit cell .
= 2.5truein without disorder the conductance @xmath185 of each isolated band satisfies the diophantine equation similar to eq .
[ dio ] , except that for a split landau band @xmath84 and @xmath186 are interchanged . for flux
@xmath184 the diophantine equation fixes the conductances @xmath187 of the bands at the flanks and @xmath188 at the center .
zero disorder is , of course , not generic , and , indeed , there are bands on the @xmath179 axis where the index is not defined , something that the phase rules for toeplitz forbid . under perturbation
the diagram should might so that these bands where the index is not defined disappear .
this is indeed the case .
the diagram in @xcite is obtained by drawing @xmath26 lines emanating from each band where @xmath26 is its hall conductance . in summary ,
the wild character of the phase diagram of the harper model is tamed by disorder and one finds , remarkably , a phase diagram compatible with the phase rules for toeplitz operators .
motivated by the effect of disorder on the harper model phase portrait , we next consider the effect of perturbations on the phase portraits of landau hamiltonians .
such perturbations will modify the phase diagram near phase boundaries . as a consequence one expects a phase diagram to be qualitatively modified near points of accumulation of phases , even if the perturbation is small .
figures 2 and 3 satisfy the phase rules in the region of large magnetic fields , but fail to do so for small magnetic fields .
we now examine how the two figures might be modified to satisfy the phase rules everywhere .
the phase diagram of the landau hamiltonian in the plane , fig .
[ landau ] will be significantly modified near the line @xmath180 which , by symmetry , must lie in a region with index @xmath189 . a schematic phase diagram that is generic and close to the landau phase diagram
is shown in fig .
[ landau+toeplitz ] .
= 2.5truein the phase diagram in fig .
[ hyperbolic ] has a region of full measure , the shaded parabola , where the operator is not fredholm .
this is non - generic , and unstable .
a perturbation might produce a phase diagram like [ hyper - toeplitz ] .
note that the two perturbed diagrams , fig .
[ landau+toeplitz ] and [ hyper - toeplitz ] are topologically identical .
= 2.5truein how do the phase diagrams , figs .
[ landau+toeplitz ] and [ hyper - toeplitz ] , compare with what one finds in experiments on the quantum hall effect ?
for large magnetic fields one finds phase diagrams that resemble both figs .
[ landau ] and [ landau+toeplitz ] and satisfy the phase rules . for weak magnetic fields
one observes a transition to an insulating phase .
the emergence of an insulating phase ( with index 0 ) for small magnetic fields is in agreement with the phase rule and fig .
[ landau+toeplitz ] . however , some experiments @xcite and numerical simulations @xcite have been interpreted as giving evidence to direct transitions from a hall conductance of 2 and 3 to the insulating phase . taken literally
, such transitions would violate the phase rule .
however , these results may merely indicate that , for @xmath1 small , the phase boundaries of fig .
[ landau+toeplitz ] are too closely spaced to be distinguished numerically and experimentally .
the main gap in our analysis is that we have not established a direct relation between the algebra of toeplitz operators , where our phase rules are proven , and the class of operators relevant to ( disordered ) quantum hall systems . at the minimum ,
toeplitz operators serve as a natural mathematical laboratory . however , there is a more direct justification for considering toeplitz operators .
the most elementary paradigm for a quantum hall system is the landau hamiltonian , in which case one has : let @xmath28 be a projection on the lowest landau level in @xmath190 , and let @xmath30 be the gauge transformation associated with an aharonov - bohm flux tube at the origin .
then @xmath33 , acting on the range of @xmath28 , differs from a toeplitz operator by a compact operator .
proof : a basis for the lowest landau level is @xmath191 as a consequence @xmath192 in this case , a compact perturbation of @xmath33 is not only a toeplitz operator ; it is a simple shift . however , if the flux tube is placed at a different point , or if the magnetic field is spread out over a finite region , then we obtain a more general toeplitz operator .
if @xmath28 is a projection on a higher landau level , the same results hold but the calculation is more involved . if @xmath28 is a projection onto multiple landau levels , then @xmath33 is a compact perturbation of a direct sum of toeplitz operators , one for each landau level .
this is not to say that toeplitz operators apply directly to all systems , only that they apply to many .
there are basic models where @xmath33 fails to be toeplitz .
indeed , an elementary model for localization is a random multiplication operator , i.e. @xmath193 on @xmath194 .
this is a caricature of strong disorder .
the eigenfunctions are now concentrated at lattice points .
the projection @xmath28 ( below a fermi energy ) is @xmath195 where the sum is over a random set of lattice points with , @xmath196 , in @xmath197 .
@xmath33 is now a multiplication by a phase .
it is an invertible operator and has fredholm index zero .
it is , however , not toeplitz .
it is tempting to directly study the index of @xmath33 , for spectral projections @xmath28 and unitary operators @xmath30 , rather than rely on generic results based on toeplitz operators .
there are , however , several technical obstacles .
the first is that @xmath33 is thought of as acting on @xmath198 , which is a hilbert space in its own right .
this means that a deformation of the parameters of the system leads to a deformation of the space @xmath199 .
in contrast , our strategy so far is formulated on a fixed space .
the second obstacle is that our results depend on continuity properties while spectral projections tend to have bad continuity properties that come from a discontinuity at the fermi energy . to overcome the first problem one
can replace @xmath33 by an operator @xmath0 defined on the entire hilbert space with coinciding index .
there is large arbitrariness in choosing @xmath0 , but a natural choice is : @xmath200 where @xmath201 . to overcome the second problem one may want to replace @xmath28 by a fermi function .
that is , replace @xmath28 by a smooth version @xmath202 in that case , however , @xmath203 is no longer equal to @xmath28 , and the different expressions for @xmath0 in equation ( [ f ] ) are no longer equivalent . for each choice
, it would be interesting to derive a phase portrait for index@xmath204 as the temperature , fermi energy , magnetic field and degree of randomness are varied .
in this paper we explored what can be said about generic phase diagrams of indices of fredholm operators .
we did not use the fact that the fredholm operators relevant to the quantum hall effect are of the form @xmath33 , with @xmath28 a spectral projection of an ergodic schrdinger operator .
rather , we considered the index of several natural classes ( and algebras ) of operators .
the weakness of this strategy is that we can not say much that is definitive about quantum hall systems . in its defense
, we recall that replacing the particular by the generic proved to be useful in quantum physics in the hands of wigner , von neuman and dyson @xcite .
whether it will turn out to be useful for quantum hall effect remains to be seen .
we thank a. kamenev for drawing our attention to ref @xcite , and e. park , m. reznikov , h. schultz - baldes and e. shimshoni for useful discussions .
this research was supported in part by the israel science foundation , the fund for promotion of research at the technion , the dfg , the national science foundation and the texas advanced research program .
gibbs , _ on the equilibrium of heterogeneous substances _ , the transactions of the connecticut academy , ( 1875 ) ; r. israe , _ convexity in the theory of lattice gases _ , princeton , ( 1979 ) ; b. simon , _ the statistical mechanics of lattice gases _ , princeton ( 1993 ) .
@xmath33 is fredholm provided the integral kernel of the projection , @xmath205 gets large .
such decay is indeed guaranteed by the theory of localization whenever the fermi energy is in the localized regime @xcite .
y. last , _ almost everything about the almost mathieu operator i _ , xi int .
phys . , p. 366372
, d. yagolnitzer , ed .
, int . press ( 1995 ) ; s. jitomirskaya , _ almost everything about the almost mathieu operator ii _ , _ ibid _ 373381 .
sheng and z.y .
weng , phys .
* 78 * , 318 ( 1997 ) ; d.n .
sheng , z.w .
weng and x.g .
wen , cond - mat/0003117 ; b. huckenstein , _ quantum hall effect at low magnetic fields _ , phys .
* 84 * , 3141 ( 2000 ) .
m. aizenman , _ localization at weak disorder : some elementary bounds _ , rev . math . phys . * 6 * , 1163 ( 1994 ) ; m. aizenman and s. molchanov , _ localization at large disorder and extreme energies _ , comm . math
. phys . * 157 * , 245 ( 1993 ) ; m. aizenman and g.m .
graf , _ localization bounds for an electron gas _
, j. phys .
a : math . gen . * 31 * , 67836806 ( 1998 ) , cond - mat/9603116 .
d. khmelnitsky , _ quantisation of hall conductivity _ , jetp lett * 38 * , 552 - 556 ( 1983 ) , and _ quantum hall effect and additional oscillations of conductivity in weak magnetic fields _ , phys .
a * 106 * , 182 ( 1984 ) ; r.b .
laughlin , phys .
lett * 52 * , 2304 ( 1984 ) . | the quantized hall conductance in a plateau is related to the index of a fredholm operator . in this paper
we describe the generic `` phase diagram '' of fredholm indices associated with bounded and toeplitz operators .
we discuss the possible relevance of our results to the phase diagram of disordered integer quantum hall systems .
pacs : 73.40.hm , also 71.10hf , 03.65.db epsf the hall conductance of integer quantum hall systems is described mathematically by the index of fredholm operators .
( for precise definitions , see below ) . in this paper
we investigate the phase diagram of the fredholm index for a few classes of operators . for the algebra of bounded operators , little
can be said beyond the fact that the phase diagrams can be arbitrarily complicated .
but for the algebra of toeplitz operators , and other related classes of operators , we establish a kind of a gibbs phase rule @xcite .
typical of our results is the statement that if the system is governed by two parameters , then one should expect jumps by one at phase boundaries and jumps by up to 2 at triple points , while jumps by more than two should never be observed .
we relate this behavior to experimental results , conjectures and open problems that arise in the context of the quantum hall effect ( qhe ) @xcite . in section 1
we define fredholm operators and their indices , and explore the different sorts of phase diagrams that can arise . in section 2
we recall how fredholm indices are related to the conductance of quantum hall systems . in section 3
we consider phase diagrams for general bounded operators . in section 4
we describe the phase diagram for linear combination of shift operators , and in section 5 we consider general toeplitz operators . in section 6
we discuss the phase diagrams of soluble models related to the quantum hall effect , and how they might be modified by disorder .
we also discuss the relevance of toeplitz operators to the quantum hall effect and present some open problems . |
after the development of molecular biology , the concepts and methods in statistical physics have been widely used to study molecular models of biological evolution @xcite , especially in recent years @xcite .
analytic solutions for models with the infinite genome length may be obtained via several methods , including the maximum principle @xcite , trotter - suzuki method in quantum statistical physics @xcite , quantum field theory @xcite and the hamiltonian - jacobi equation method @xcite .
the infinite genome length limit has been solved fora more complicated ( more nonlinear ) evolution model : the horizontal gene transfer model @xcite .
this model has been solved both for the haploid @xcite , and the hyper - geometric diploid case @xcite .
the study of finite - size corrections is important not only for lattice phase transition models , but also for population genetics @xcite .
here there are situations with few relevant genes for the evolution problem .
although the genome length may be large , e.g. about 10000 in the case of hiv , it is possible to consider only the variable part of the genome which has only @xmath2 bases @xcite ; in such a case , the finite - size effects can not be overlooked .
the study of finite - size corrections allows us to understand as to what extent the results for the infinite genome length can be used to represent those for the finite genome length . in recent works
@xcite the finite genome length corrections have been calculated for the haploid molecular evolution model @xcite .
in @xcite we calculated the finite size correction for the recombination model with single - peak fitness . in the present paper
we calculate the finite genome length corrections for a diploid model with symmetric landscape @xcite as well as for a haploid model with a simple horizontal gene transfer ( hgt ) @xcite for a general symmetric fitness landscape .
the method may be applied to rather general cases of nonlinear probabilistic models @xcite .
actually , we are constructing a perturbation expansion for the eigen - value like variable for the nonlinear operators . in case of linear operators ( like the quantum mechanics ) , to calculate the first order perturbation to the energy we need only the 0-th order expressions of eigen - vectors to calculate the leading corrections to eigen - values due to perturbations . in case of nonlinear system of equations
it is impossible to use the methods of linear algebra .
nevertheless , we can succeed using a trick . to calculate the perturbation expansion of eigen - value like variables ( mean fitness or the asymptotic expression ) , we use equations such as eq.(13 ) , where the contribution of correction terms of the distribution function disappear .
thus the perturbation of mean - fitness ( `` eigen - value '' for a nonlinear problem ) can be calculated using only the zeroth - order expression of eigen - functions ( bulk solutions for probability distribution ) .
the present paper is organized as follows . in section
ii we derive the finite size corrections for a haploid model ( crow - kimura model ) with a new method . in sections
iii and iv the new method is applied to the diploid evolution model and hgt model , respectively . in the appendix
we give finite size corrections for the diploid evolution model with single - peak fitness function .
let us first introduce a new method to calculate finite - size corrections for the crow - kimura model with haploid genotypes @xcite , which is easier to study .
the crow - kimura model is slightly easier to solve than the eigen model @xcite . in the crow - kimura model @xcite ,
any genotype configuration @xmath3 is specified by a sequence of _ l _ two - valued spins @xmath4 for @xmath5 and @xmath6 such that the value @xmath7 represents purines ( a , g ) , and the value @xmath8 represents pyrimidines ( t , c ) , where @xmath9 is the total number of different sequences .
the difference between two configuration @xmath10 and @xmath11 is described by the hamming distance @xmath12 . in other words ,
the hamming distance is the number of different spins between configurations @xmath13 and @xmath11 .
the relative frequency @xmath14 of a given configuration @xmath13 satisfies following equation : @xmath15 here @xmath16 is the replication rate , i.e. the _ fitness _ of an organism with a given genotype and in crow - kimura s model it is specified like a function of the genotype : @xmath17 . in other words
_ fitness _ is the average number of offspring of individual with genotype sequence @xmath13 per unit period of time and it is very meaningful quantity in evolution theory .
@xmath18 is the mutation rate to move from sequence @xmath13 to sequence @xmath11 per unit period of time .
it is important that in crow - kimura s model @xcite only single base mutations are allowed , i. e. @xmath19 for @xmath20 , @xmath21 for @xmath22 and @xmath23 , and @xmath24 for @xmath25 ; the last condition ensures that the time evolution of @xmath14 does not change the normalization condition @xmath26 . in the following , we take @xmath27 . the nonlinear system of eq .
( [ e1 ] ) can be mapped into following linear system of equations , @xmath28 where @xmath29 are related to @xmath14 of eq.([e1 ] ) via @xcite , @xmath30 for the symmetric fitness landscape , where @xmath16 are the same for all the sequences with the same hamming distance from the same reference sequence @xmath31 , it is more convenient to work with _ classes _ , i.e. we classify configurations @xmath13 into the classes according to the value @xmath32 , where @xmath33 is so - called `` magnetization '' of the configuration @xmath13 and is defined as @xmath34 . defining the magnetization @xmath35 for the configurations at the @xmath36-th class as @xmath37 and fitness @xmath38 , we rewrite eq.([e2 ] ) for the probability @xmath39 for a typical configuration in the @xmath36-class as @xmath40 we define the @xmath36-th hamming class as the group of all sequences at the hamming distance @xmath36 from the reference sequences . having the number of configurations at the @xmath36-th class @xmath41 for @xmath42 we have @xmath43 we are interested in calculating the mean fitness @xmath44 and surplus @xmath45 in the steady state . in @xcite , one of us proposed that in the steady state @xmath46 and the mean fitness @xmath47 can be written as @xmath48 where @xmath49 .
then , high order corrections for @xmath47 , and for @xmath50 have been derived using the linear algebra methods and the equations for @xmath51 . in this section
we derive these corrections with an alternative method that can be applied to the strong nonlinear situations . using the ansatz
@xmath52 $ ] and a formula @xmath53 where @xmath54 , one can transform eq .
( [ e5 ] ) into the hamilton - jacobi equation , @xmath55 where @xmath56 is a dummy variable .
we assume an asymptotic @xmath57 and get an ordinary differential equation @xmath58 which corresponds to eq .
( 23 ) in ref .
we take the `` + '' solution for @xmath59 and @xmath60 solution for @xmath61 . in eq .
( [ e8 ] ) , @xmath62 is a function @xmath63 .
the value of @xmath63 , which gives the maximum value of @xmath62 is defined by equation @xmath64 thus the maximum value of @xmath62 can be obtained from @xmath65_{-1\le m\le 1 } , \nonumber\\ v(m)&\equiv & h(m , u_0(m))=f(m)-1+\sqrt{1-m^2},\end{aligned}\ ] ] where the maximum of the first equation is at the point @xmath66 . to calculate higher order corrections ,
we write @xmath50 as @xmath67 equations ( [ e5 ] ) and ( [ e9 ] ) imply that @xcite : @xmath68.\end{aligned}\ ] ] in eq.([e10 ] ) , @xmath62 is the bulk expression of the mean fitness and @xmath69 is the first order correction to it .
having the value of @xmath70 , we can calculate @xmath71 .
@xmath70 can be defined from eq.([e10 ] ) at the point @xmath72 , where the coefficient of @xmath71 becomes zero .
we have an equation @xmath73 equation ( [ e10 ] ) then implies that at @xmath66 @xmath74 let us define @xmath75 .
equation ( [ e8 ] ) can be written as @xmath76 .
expanding eq .
( [ e8 ] ) near @xmath72 with respect to the first and the second arguments of @xmath77 up to the second order , we derive @xmath78 note that there is no @xmath79 term in the last equation .
we can verify this directly : @xmath80 . from eqs .
( [ e7 ] ) , ( [ e8a ] ) and ( [ e8b ] ) , we derive @xmath81 @xmath82 @xmath83 with these expressions in eq.([e13a ] ) , we derive @xmath84 therefore @xmath85 equations ( [ e8b ] ) and ( [ e11 ] ) imply that @xmath86 .
thus eventually , we can use eq .
( [ e13 ] ) to obtain @xcite,@xcite @xmath87.\end{aligned}\ ] ] using the expression for @xmath70 , we can calculate the @xmath71 from eq .
( [ e10 ] ) : @xmath88
\nonumber\\ & = & k_1-(e^{2u'}+e^{-2u ' } ) \nonumber \\ & -&2u''(\frac{1+m}{2}e^{2u'}+\frac{1-m}{2}e^{-2u'}).\end{aligned}\ ] ] therefore , having @xmath89 , eq.([e16 ] ) , we can calculate the @xmath1 accuracy expression for the haploid class probabilities @xmath90 ( we define them as @xmath91 ) , also we can calculate the probabilities @xmath92 , eq.([e9 ] ) .
the comparison of this computations with numerical results are shown in the fig . 1 for the case @xmath93 .
= 0.1 in ( 42,12 ) ( -1.7,0 ) ( 16.5,0 ) ( 7.7,0 ) ( 26,0 ) on the other hand , we can calculate @xmath70 directly as @xmath94 , @xmath95 the bulk surplus term is calculated using the equation @xcite @xmath96 from the definition of @xmath97 , the correction to the surplus can be written as @xmath98 . comparing with eq .
( [ e18 ] ) , we get @xmath99 . using the formula
@xmath100 we derive , assuming @xmath101 : @xmath102 one can define @xmath103 integrating the expression of @xmath104 from eq .
( [ e16 ] ) .
thus we have derived the results of @xcite by an alternative method .
the diploid model has been defined by eq . ( 1 ) in @xcite . there
we derived the analytic solution for the diploid many allele biological evolution models @xcite with general fitness landscapes to the first order using the hamilton - jacobi equation approach . in the present paper
we will find @xmath1 accuracy for mean fitness and genome probabilities of that diploid model . in wbs parallel diploid model @xcite ,
gene probabilities @xmath29 evolve as @xmath105+\sum_{j=1}^m m_{ij}p_j.\ ] ] here @xmath106 is a mutation matrix , defined in section ii .
we have a balance condition @xmath107 .
@xmath108 is the fitness of the genotype @xmath109 , and @xmath110 is the marginal fitness for the sequence @xmath13 .
we assume that the fitness of the configuration @xmath13 is a smooth function of the hamming distance between @xmath13 and the reference configuration @xmath111 .
in such case , it is convenient to work with the overlap @xmath112 instead of the hamming distance @xmath113 . consider the following choice of the matrix @xmath114 : @xmath115 where @xmath116 . in the hypergeometric model @xmath117 and @xmath118
are , respectively , the number of @xmath119 maternal and paternal alleles .
@xmath120 is a smooth analytical function .
we are interested in finding the exact phase structure and the steady state , therefore we can consider only symmetric solutions of @xmath29 . using the expression @xmath121 $ ] in eq .
( 13 ) of @xcite , we obtain coupled equations for @xmath90 that describe the diploid , one locus many allele parallel mutation - selection model @xcite : @xmath122 here we define the marginal fitness for the hamming class @xmath123 . in diploid
case the fitness landscape is defined via a function @xmath120 of two arguments , describing the dominance relations among alleles @xcite .
eq.(21 ) coincides with the crow - kimura model where @xmath124 is the fitness function for the hamming class l. assuming ansatz ( [ e6 ] ) and ( [ e7a ] ) , we derive the bulk expression for steady state solution as @xmath125 where @xmath126 is defined via eqs.(6),(8 ) . the mean fitness per spin @xmath62 and surplus @xmath127
are defined via a system @xcite @xmath128 to find an expression for @xmath124 with @xmath1 accuracy , we calculate the following integral : @xmath129 where @xmath49 .
therefore we derive @xmath130}{l|u''(s)| } + \frac{f''_{bb}[m , s]}{2l|u''(s)| } \nonumber\\ & \equiv & f'_b(m , s)\frac{a}{l}+\frac{f''_{bb}[m , s]}{2l|u''(s)|},\end{aligned}\ ] ] where @xmath131}/{|u''(s)|}$ ] , @xmath97 is the surplus , @xmath132 is the first derivative of the fitness function via the second argument , and @xmath133 is the second derivative of the fitness function via the second argument . for the mean fitness @xmath134 , from eq .
( [ e25 ] ) we have @xmath135}{2|u''(s)|}.\end{aligned}\ ] ] expanding @xmath136 via the first argument in the equation above , we obtain @xmath137\frac{a}{l}+\frac{[f''_{bb}[s , s]+f''_{aa}[s , s]}{2l|u''(s)|}\end{aligned}\ ] ] with @xmath138 being the first derivative of the fitness function via the first argument and @xmath139 being the second derivative of the fitness function via the first argument . on the other hand , we can get an expression for @xmath70 by adding the term @xmath140}{2|u''(s)|}$ ] to the right - hand side of eq.([e16 ] ) .
thus we obtain the following equation for the corrections : @xmath141a+\frac{[f''_{bb}[s , s]+f''_{aa}[s , s]}{2|u''(s)|}\nonumber\\ % % & = & 2u'_1[\sinh(2u')+m\cosh(2u ' ) ] & = & 2u'_1[\frac{1+m}{2}e^{2u'}-\frac{1-m}{2}e^{-2u ' } ] \nonumber\\ & + & f'_b(m , s)a+\frac{f''_{bb}[m , s]}{2|u''(s)|}\nonumber\\ & + & 2u''[\frac{1+m}{2}e^{2u'}+\frac{1-m}{2}e^{-2u'}]+2\cosh(2u').\end{aligned}\ ] ] the first line is derived via direct integration of the fitness function via the steady state distribution , the @xmath71 term in the second line is just the small @xmath71 correction of the right hand side of eq.([e22 ] ) .
the third line corresponds to eq.([e25 ] ) .
the last line corresponds to the correction terms of the haploid model . to calculate @xmath70 we
again consider the point where the coefficient of @xmath71 disappears in the last equation .
we calculate @xmath75 as in the previous section .
we have an equation for @xmath142 @xmath143a\nonumber\\ & = & \frac{f''_{bb}(m_0,s)-f''_{aa}(s , s)-f''_{bb}(s , s)]}{2|u''(s)|}\nonumber\\ & + & \frac{1}{\sqrt{1-m_0 ^ 2}}[1-\sqrt{1-f''_{aa}(m_0,s)(1-m_0 ^ 2)^{3/2}}]\nonumber\\ & + & \frac{f''_{bb}(m_0,s)-f''_{aa}(s , s)-f''_{bb}(s , s)]}{2|u''(s)|},\end{aligned}\ ] ] where @xmath66 is the bulk magnetization . then putting the value of @xmath142 in eq.([e29 ] )
, we get an expression for @xmath70 : @xmath144,\end{aligned}\ ] ] @xmath145,\ ] ] where @xmath146 from eq.([e23 ] ) we calculate first order accuracy expression for the diploid mean fitness @xmath147 and then from eq.([e30 ] ) we compute the @xmath1 accuracy expression for mean fitness @xmath148 . having all these expressions , we compare our analytical results with the direct numerics for different values of @xmath0 and different values of @xmath149 in table i. having @xmath89 by eq.([e28 ] ) , we calculate the @xmath1 accuracy expression for the diploid class probabilities @xmath150 . in fig .
2 we give the comparison of our analytical results with the direct numerics .
.comparison of 0-th order accuracy expression @xmath151 and the 1/l accuracy expression @xmath152 results for the mean fitness of diploid model , for the fitness function @xmath153,for @xmath154 .
@xmath155 is given by direct numerics of eq.(4 ) . [
cols="^,^,^,^,^,^,^",options="header " , ] = 0.1 in ( 42,12 ) ( -1.7,0 ) ( 16.5,0 )
the finite genome length effects are very important for the case of horizontal gene transfer(hgt ) . while the genome length could be large ( about 10000 in case of hiv ) , it is possible to consider only the variable part of genome , @xmath156 @xcite ) .
we consider the model @xcite describing a simple horizontal gene transfer .
the model has been solved @xcite in the large genome limit in @xcite ( mean fitness ) and @xcite ( steady state distribution ) . in this model , besides the mutation process , one spin in the genome is replaced with the spin at the same position from the sequence pool .
we consider the following system of equations @xcite : @xmath157- \nonumber\\ & & p_l(1+\frac{1}{l}\sum_{k}r_kp_k)- \nonumber \\ & & cp_{l}+c(\frac{\bar l}{l } \frac{l}{l } + \frac{{l- l}}{l}\frac{l-\bar l}{l})p_{l}+ \nonumber\\ & & c(1-\frac{{l}-1}{l})\frac{\bar l-1}{l}p_{l-1}+ c\frac{{l}+1}{l}(1-\frac{\bar l+1}{l})p_{l+1}],\end{aligned}\ ] ] where @xmath158 .
the first and second lines correspond to the crow - kimura model @xcite,@xcite .
the third line describes the hgt process with the change of the hamming class .
we cut plus spin with the probability @xmath159 and add plus spin from the sequence pool with the probability @xmath160 , then cut minus spin with probability @xmath161 and add minus spin from the sequence pool with the probability @xmath162 .
the last line corresponds to the process with the change of hamming class .
we cut the plus spin with the probability @xmath163 and add minus spin from sequence pool with the probability @xmath164 , then cut the minus spin with the probability @xmath165 and add plus spin from sequence pool with the plus spin with the probability @xmath166 . considering the ansatz @xmath167,\ ] ] we get the following equation for the bulk terms @xmath168\nonumber\\ & + & \frac{1-m}{2}e^{-2p}[1+c\frac{1+s}{2 } ] + \frac{cms}{2}-\frac{c}{2}-1,\end{aligned}\ ] ] where @xmath169 and @xmath170 is a dummy variable .
the minimum of @xmath171 is at @xmath172 , where @xmath173 gives a potential @xmath174_{u'}$ ] : @xmath175.\end{aligned}\ ] ] we define @xmath176 from the conditions @xmath177 where the derivative is with respect to the first argument .
let us calculate the @xmath75 .
near @xmath66 we have an expansion @xmath178 where @xmath179 is the second derivative of @xmath180 with respect to the first argument , @xmath181 is the second partial derivative of @xmath182 with respect to the second argument of the function @xmath183 . dividing eq.([e38 ] ) by @xmath184
, we derive the equation @xmath185 or @xmath186^{1/4}}.\end{aligned}\ ] ] from eq.(36 ) we have : @xmath187\nonumber\\ & + & [ e^{2u'}(1+c\frac{1-s}{2})+(1+c\frac{1+s}{2})e^{-2u ' } ] \nonumber\\ & + & 2u''[e^{2u'}\frac{1+m}{2}(1+c\frac{1-s}{2 } ) \nonumber \\ & & + \frac{1-m}{2}(1+c\frac{1+s}{2})e^{-2u'}]\nonumber\\ & + & cu_1'(s)\frac{m - e^{2u'}\frac{1+m}{2}+e^{-2u'}\frac{1-m}{2}}{2|u''|}.\end{aligned}\ ] ] the first line is a direct integration of the fitness via steady state distribution , the second line corresponds to the correction of @xmath188 via @xmath189 , the third line corresponds to the corrections in the first line of eq.([e33 ] ) .
the last line corresponds to the @xmath190 corrections from the @xmath191 terms in eq.([e33 ] ) . to calculate @xmath70
, we put the optimal value of @xmath63 and look at the point @xmath72 . then eq.(44 )
is simplified : @xmath192.\end{aligned}\ ] ] deriving @xmath193 from the last equation , we obtain @xmath194 thus @xmath195 @xmath196-\frac{f''(s)}{f'(s)}]\ ] ] for the fitness function @xmath197 with parameters @xmath198 , we derive from the numerics for the mean fitness and order parameters @xmath199 . the analytical formula eq.([e37 ] ) for the infinite genome limit gives @xmath200 , while the numerics for @xmath201 gives @xmath202 and for @xmath203 @xmath204 .
thus we get for @xmath201 , @xmath205 and for @xmath206 , @xmath207 .
we see that our analytical result by eq .
( [ e44 ] ) for the mean fitness expression is well confirmed .
one can define @xmath103 by integrating the expression of @xmath104 from eq .
( [ e42 ] ) .
we calculated the finite genome size corrections to the mean fitness and steady state distribution for strongly nonlinear evolution models : horizontal gene transfer model and diploid parallel evolution model for general fitness function .
the application of the hamilton - jacobi equation for the investigation of the master equation , especially the finite size corrections , is a rather popular method in case of linear master equation for chemistry or ecology @xcite,@xcite . for the models , considered in our article
, the standard methods of linear algebra and the methods of @xcite,@xcite fail .
the key point of our method is that we investigate the point where the coefficient of the correction to the steady state distribution disappears .
then the correction to the fitness is calculated from the smoothness condition .
our method could be applied to other nonlinear probabilistic models as well .
our formulas could be applied for relatively small genome lengths , for the single peak fitness landscape even for @xmath208 , see appendix a. in population genetics usually few allele models are investigated with @xmath209 .
fortunately , for the larger allele numbers we can apply our methods .
our formulas , while a bit involved , are less cumbersome than the formulas for several allele cases in population genetics .
dbs thanks darpa prophecy program and academia sinica for financial support .
consider the following diploid model studied in @xcite : @xmath210 \nonumber \\ & + & \sum_jm_{ij}p_j\end{aligned}\ ] ] with the fitness coefficients @xmath211 here @xmath106 is the mutation matrix , @xmath212 for @xmath20 , and for other cases @xmath213 . for the steady state
we have @xcite @xmath214.\end{aligned}\ ] ] here @xmath215 takes one of the solution @xmath216 defined below @xmath217 where @xmath218 , @xmath219 , and @xmath220 .
we have for the marginal fitness @xmath221 and the fitness @xmath47 @xmath222.\end{aligned}\ ] ] in terms of the marginal fitness , the original equation ( a1 ) can be written in the form @xmath223 where @xmath90 are the class probabilities of the configurations at the hamming distance @xmath36 from the reference sequence , @xmath224 , and @xmath225 .
for the @xmath228 , we have the equation @xmath229 replacing on the right hand side and before the parenthesis on the left hand side @xmath230 and @xmath231 @xmath232 we find a perturbation expression for @xmath233 @xmath234 and for @xmath235 @xmath236}]}.\end{aligned}\ ] ] for @xmath237 we get @xmath238 , while @xmath239 , thus we have @xmath240 accuracy .
d. b. saakian , z. kirakosyan , c .- k .
hu , phys . rev .
e , * 77 * , 061907 ( 2008 ) .
e. munoz , j .-
park , and m. w. deem , _
phys.rev.e _ * 78 * , 061921 ( 2008 ) .
avetisyan , d.b .
saakian , phys.rev.e . *
81 * , 051916 ( 2010 ) .
n. n. v. vijay , vasantika , r. ajmani , a. s. perelson , and n. m. dixit , j. gen .
virology * 89 * , 1467,(2008 ) . c. escudero and a. kamenev , phys.rev .
e * 79 * , 041149 ( 2009 ) .
m. assaf and b. meerson , phys.rev .
e * 81 * , 021116 ( 2010 ) . | using the hamilton - jacobi equation approach to study genomes of length @xmath0 , we obtain @xmath1 corrections for the steady state population distributions and mean fitness functions for horizontal gene transfer model , as well as for the diploid evolution model with general fitness landscapes .
our numerical solutions confirm the obtained analytic equations .
our method could be applied to the general case of nonlinear markov models . |
more than three decades since its original proposal @xcite , inflation has passed one of its most stringent tests with the recent measurements of the temperature anisotropies of the cosmic microwave background ( cmb ) made by the planck satellite @xcite . in particular , models of inflation based on the dynamics of a slowly rolling scalar field generate a primordial spectrum of density perturbations that is essentially adiabatic , gaussian and nearly but not exactly scale - invariant , in agreement with observations . while planck has been able to place strong constraints on several inflationary models , the question remains as to which is the fundamental mechanism driving inflation .
ca - nonical models assume that accelerated expansion quickly erases all traces of any pre - inflationary matter or radiation distributions , so that slow - roll inflation occurs in an almost perfect vacuum state . however , the inflaton field is necessarily coupled to other degrees of freedom in order to dissipate its vacuum energy and reheat the universe , so one may envisage scenarios where dissipative effects become important during and not only after the slow - roll phase .
these are generically known as warm inflation scenarios @xcite , and raise the interesting possibility that the inflationary universe is not in a perfect vacuum , even though vacuum energy is the dominant component for accelerated expansion to take place . in this letter , we show that even when dissipative effects are still small compared to hubble damping , the amplitude of scalar curvature fluctuations can be significantly enhanced , whereas tensor perturbations are generically unaffected due to their weak coupling to matter fields .
this generically reduces the tensor - to - scalar ratio with respect to conventional models and also modifies the tilt of the scalar power spectrum , thereby changing observational predictions considerably .
these effects are particularly significant when the temperature of the radiation bath @xmath1 and also when non - trivial inflaton occupation numbers are sustained during inflation . as an example
, we show that the simplest model of chaotic inflation , @xmath2 , falls well within planck s observational window for a nearly - thermalized state , in a supersymmetric realization of warm inflation with renormalizable interactions , whereas it seems to be ruled out in a cold scenario .
the simplest models of warm inflation consider an adiabatic regime , where dissipative processes occur faster than hubble expansion and the inflaton s slow motion , such that the dissipative part of the inflaton s self - energy from its coupling to other fields leads to an effective friction coefficient in its equation of motion @xcite : @xmath3 which reduces to @xmath4 in the overdamped or slow - roll regime , where @xmath5 .
this implies the generalized conditions @xmath6 on the slow - roll parameters @xmath7 and @xmath8 , which in particular allow for @xmath9 in the strong dissipation regime , @xmath10 , therefore alleviating the need for very flat potentials @xcite .
moreover , the inflaton dissipates its energy through particle production and , if the resulting particles are relativistic and relax to an equilibrium configuration sufficiently fast , a nearly - thermal radiation bath is sourced all through inflation : @xmath11 with @xmath12 for @xmath13 relativistic degrees of freedom . in the slow - roll regime ,
the radiation quickly reaches a quasi - stationary configuration where @xmath14 may be neglected and we obtain @xmath15 .
radiation is then a sub - dominant component , allowing for accelerated expansion , although one may have @xmath0 even for weak dissipation , @xmath16 . at the end of inflation , however , for @xmath17 we have @xmath18 , and radiation may become a relevant component if @xmath10 at this stage .
in fact , in models where dissipation becomes stronger as inflation proceeds , radiation will typically come to dominate once the slow - roll regime has ended , yielding a smooth ` graceful exit ' into a radiation - dominated universe .
a well - studied dissipation mechanism considers a supersymmetric model with a renormalizable superpotential of the form @xcite : @xmath19 where the inflaton corresponds to the scalar component of the chiral multiplet @xmath20 , @xmath21 , with a scalar potential @xmath22 that spontaneously breaks supersymmetry ( susy ) during inflation .
dissipation is sourced by the coupling of the inflaton to the bosonic and fermionic @xmath23 fields and their subsequent decay into @xmath24 scalars and fermions , which form the thermal bath . in the low temperature regime , @xmath25 , which typically corresponds to large field values and where inflaton thermal mass corrections are boltzmann - suppressed @xcite , dissipation is mainly mediated by the decay of virtual scalar modes @xmath26 , yielding @xcite : @xmath27 for @xmath28 and @xmath29 chiral multiplets .
supersymmetry suppresses radiative and thermal corrections to the scalar potential , yielding at 1-loop order : @xmath30 where @xmath31 and @xmath32 is the renormalization scale
. the heavy fields also decay into inflaton particle states through @xmath33 , but this is a sub - leading process , with @xmath34 , where @xmath35 @xcite , with inflaton particles from this process alone typically yielding a negligible component of the radiation bath .
however , dissipative particle production destabilizes the local thermal equilibrium of the plasma , triggering decays , inverse decays and thermal scatterings that redistribute the dissipated energy between all the interacting fields and keep the system close to equilibrium if occurring faster than the hubble rate .
in particular , decays and inverse decays can be efficient thermalization processes @xcite so that , in some parametric regimes , we expect inflaton particles to be sustained in a quasi - thermal state at the ambient temperature @xmath36 by decays and inverse decays of the multiple heavy species in the plasma . although the details of the thermalization process require solving the system of coupled boltzmann equations for all the particle species involved , which may require full numerical simulations and is outside the main scope of this letter , the underlying physical picture can be understood in simple terms . starting from an equilibrium configuration where decay and inverse decay processes are occurring at equal rates , dissipation of the inflaton s energy
will mainly produce an excess of light particles in the @xmath24 sector , which will enhance the rate of inverse decays and consequently increase the @xmath23 sector occupation numbers above their equilibrium value .
this in turn enhances the direct decay rate , producing @xmath24 particles and also an excess of inflaton modes .
this goes on until the balance between decay and inverse decay rates is restored and the system reaches a new equilibrium configuration .
one then expects the energy injected into the system to be distributed amongst all species in the plasma that are produced and annihilated faster than hubble expansion .
the common temperature of these species would decrease due to expansion but this is compensated by the additional energy , keeping the temperature roughly constant in the slow - roll regime .
species that are not created / destroyed sufficiently fast will decouple from the plasma and their effective temperature will be exponentially redshifted away during inflation , quickly reaching a quasi - vacuum state .
a measure of the efficiency of the thermalization processes can then be obtained by comparing the relevant decay rates with the hubble parameter , as we examine in more detail below in the context of chaotic inflation .
fluctuation - dissipation dynamics modifies the evolution of inflaton perturbations , which are sourced by a gaussian white noise term , @xmath37 @xcite : @xmath38 in the slow - roll regime .
we recall that the intensity of the noise is related to the dissipation coefficient through the fluctuation - dissipation theorem and , as shown in the first article in @xcite , dissipation from scalar modes dominates over the fermionic one in the low temperature regime that we are considering here .
spatial correlation properties may , however , be different for fermionic modes @xcite .
also , as mentioned above , dissipative processes may maintain a non - trivial distribution of inflaton particles , @xmath39 , which for sufficiently fast interactions should approach the bose - einstein distribution at the ambient temperature , @xmath40 .
in this case the associated creation and annihilation operators have correlation functions @xmath41(2\pi)^3\delta^3(\mathbf{k}+\mathbf{k'})$ ] and @xmath42 . for a generic inflaton phase - space distribution at the time when observable cmb scales leave the horizon during inflation , @xmath43 ,
one then obtains the dimensionless power spectrum @xcite : @xmath44~,\ ] ] which yields the standard cold inflation result in the limit @xmath45 .
this expression neglects , however , the coupling between inflaton and radiation fluctuations associated with the temperature dependence of the dissipation coefficient in eq .
( [ dissipation_coefficient ] ) , an effect that may significantly enhance the perturbation growth for strong dissipation , @xmath10 @xcite .
since this coupling is negligible if the relevant scales become super - horizon when dissipation is weak , we can obtain an accurate description of the spectrum by taking the limit @xmath46 in the expression above , which yields : @xmath47~.\ ] ] this assumption is justified in models where @xmath48 grows during inflation , such that dissipation has a negligible effect 50 - 60 e - folds before the end of inflation but becomes stronger towards the end , thus helping to prolong the period of accelerated expansion .
we show below that this is indeed the case for our working example of chaotic inflation and discuss the potential effects of strong dissipation at horizon - crossing at the end of this letter .
note that both the second and third terms within the brackets in eq .
( [ scalar_power_spectrum_full ] ) are positive - definite , the former corresponding to non - trivial inflaton occupation numbers and the latter to the leading effect of fluctuation - dissipation dynamics .
hence , the amplitude of the scalar power spectrum always exceeds the vacuum result in warm inflation scenarios . on the other hand , gravity waves
are weakly coupled to the thermal bath and the spectrum of tensor modes retains its vacuum form , @xmath49 . this therefore suppresses the tensor - to - scalar ratio , yielding a modified consistency relation for warm inflation : @xmath50 where @xmath51 is the tensor index .
the primordial tensor spectrum can thus be used to distinguish warm from cold inflation scenarios , the former consequently evading the lyth bound @xcite ( see also @xcite for other scenarios where the lyth bound does not apply ) .
whereas previous studies have focused more on the strong dissipation regime , this result explicitly shows that warm inflation can have a significant observational impact for weak dissipation , where temperatures well above the hubble rate can be sustained .
most importantly , non - trivial inflaton occupation numbers may also generically lower the tensor - to - scalar ratio , which as we illustrate below may have a very significant effect on inflationary predictions .
in the limit where inflaton particle production is inefficient and @xmath52 gives a negligible contribution to the power spectrum , the scalar spectral index is nevertheless modified by the third term in eq .
( [ scalar_power_spectrum_full ] ) , yielding : @xmath53 where @xmath54 and we have used the slow - roll equations , @xmath4 and @xmath55 , to determine the variation of @xmath56 as different scales become super - horizon during inflation .
modifications are , however , more prominent in the opposite limit of nearly - thermal inflaton fluctuations , with @xmath57 . for @xmath58 and @xmath59
we then obtain : @xmath60 which is , in particular , independent of the curvature of the potential , which only determines its running : @xmath61 in this case , a red - tilted spectrum , @xmath62 , corresponds to either potentials with a negative slope , such as hill - top models , or large field models where @xmath63 .
to illustrate the effects of both dissipation and occupation numbers on observational predictions , let us consider the quartic model , @xmath2 , which corresponds to a superpotential @xmath64 and is the canonical model of chaotic inflation @xcite . in this case , we have @xmath65 , which yields @xmath66 for a thermalized inflaton distribution from eq .
( [ scalar_index_thermal ] ) .
this gives a red - tilted spectrum with @xmath67 for @xmath68 , which is super - planckian but smaller than the corresponding field value in the vacuum case , @xmath69 .
this also gives @xmath70 , within the upper bound obtained by planck , @xmath71 ( 95% cl ) , for @xmath72 , as well as a small negative running @xmath73 and a tensor index @xmath74 .
the number of e - folds of inflation can be computed by integrating the slow - roll equations , which may be done analytically for the quartic model @xcite . in particular
, one can use the form of the dissipation coefficient in eq .
( [ dissipation_coefficient ] ) to express the coupled inflaton and radiation equations in the slow - roll regime as a single equation for the dissipative ratio @xmath48 : @xmath75 where @xmath76 for @xmath59 .
this shows explicitly that @xmath48 grows during inflation , justifying our assumption that the system may evolve from the weak to the strong dissipation regime .
inflation ends in this case when @xmath77 , which yields @xmath78 for a thermal spectrum and hence @xmath79 for @xmath80 .
as discussed earlier , the relative abundance of radiation will then also grow towards the end of inflation , with @xmath81 in this case , until it smoothly takes over after slow - roll has ended . integrating eq .
( [ q_eq ] ) from horizon - crossing to the end of the slow - roll regime , we obtain : @xmath82 where @xmath83 .
this yields the required @xmath84 e - folds of inflation with @xmath85 for @xmath86 . for comparison , in the standard cold inflation regime
, one finds @xmath87 , giving @xmath88 for @xmath89 .
this clearly shows that even for weak dissipation at horizon - crossing one may obtain substantially different observational results . for both limits of nearly - thermal and negligible inflaton occupation numbers
, one can use the observed amplitude of curvature perturbations , @xmath90 @xcite and the form of the dissipation coefficient in eq .
( [ dissipation_coefficient ] ) to relate the different quantities at horizon - crossing , with e.g. @xmath91 in the nearly - thermalized regime .
this allows one to express both @xmath92 and @xmath93 in terms of the dissipative ratio or temperature at horizon - crossing for a given number of e - folds of inflation and relativistic degrees of freedom , which is illustrated in figure 1 .
as one can see , observational predictions for the quartic model depend on the distribution of inflaton fluctuations , @xmath52 . for @xmath94 , the spectrum has the same form as in cold inflation , but from eq .
( [ e - folds ] ) one obtains @xmath89 for smaller field values than in cold inflation , yielding a larger tensor fraction and a more red - tilted spectrum . when @xmath95 , however , the spectrum becomes more blue - tilted and @xmath93 is suppressed , although for weak dissipation it remains too large .
[ htbp ] plane for @xmath2 as a function of the dissipative ratio , @xmath96 , 50 - 60 e - folds before the end of inflation , compared with the planck results @xcite , for @xmath97 relativistic degrees of freedom .
the dark green ( light blue ) curves correspond to nearly - thermal ( negligible ) inflaton occupation numbers @xmath52 , with dashed branches for @xmath98 .
note that corresponding curves converge in the cold inflation limit , @xmath99.,title="fig : " ] on the other hand , for nearly - thermal inflaton occupation numbers tensor modes are more strongly suppressed and one obtains a remarkable agreement with the planck results for @xmath58 .
note that for @xmath98 the concept of thermal equilibrium is ill - defined , since the average particle modes have super - horizon wavelengths , so in figure 1 we represent this regime with dashed curves to nevertheless illustrate the transition from a cold to a warm spectrum . also , we take the mssm value @xmath97 only as a reference , with fewer light species further lowering the tensor - to - scalar ratio , since @xmath100 is larger .
this agreement is particularly significant , since the quartic potential is the simplest renormalizable model of chaotic inflation , involving no other scales other than the inflaton field value .
as originally argued by linde @xcite , in large - field models inflation is naturally triggered from a chaotic field distribution following the pre - planckian era , in domains where @xmath101 quickly dominates over gradient and kinetic energy densities .
on the other hand , when inflation only occurs for a @xmath102 plateau , the post - planckian universe must be unnaturally smooth , requiring a fine - tuning of initial conditions that the inflationary paradigm is supposed to solve @xcite .
while other modifications such as a non - minimal coupling to gravity may also bring the quartic model into agreement with observations @xcite , the renormalizable nature of the interactions leading to dissipation is an attractive feature of warm inflation , with only a few controllable parameters .
note , in particular , that interactions with other bosonic and/or fermionic fields are always required since the vacuum energy of the inflaton field must be transferred into light degrees of freedom at the end of inflation to ` reheat ' the universe . in this sense , warm inflation scenarios do not introduce any non - standard modifications to the basic inflationary models but simply correspond to parametric regimes where the universe is kept warm throughout inflation , @xmath1 . for the dissipation coefficient in eq .
( [ dissipation_coefficient ] ) , one obtains in particular @xmath103 , which may be achieved for @xmath104 and @xmath105 , while keeping radiative corrections under control , @xmath106 .
we may express the number of heavy species as : @xmath107 where we have assumed a thermal distribution of inflaton perturbations .
this large multiplicity of @xmath23 species is typical of the form of the dissipation coefficient in eq .
( [ dissipation_coefficient ] ) @xcite , but is expected to be significantly reduced in other regimes , such as for on - shell @xmath23 modes @xcite .
large multiplicities may be obtained in d - brane constructions @xcite , where the @xmath23 fields correspond to strings stretched between brane and antibrane stacks and their number thus grows with the square of the brane multiplicity . due to brane - antibrane annihilation at the end of inflation , these modes will not , however , play a role in the post - inflationary universe .
field multiplicities are also enhanced by the kaluza - klein tower in extra - dimensional scenarios @xcite .
an interesting possibility arises when we consider b- and cp - violating interactions for the @xmath23 fields in eq .
( [ superpotential ] ) , with complex couplings and distinct decay channels . in this case , the out - of - equilibrium nature of dissipation can generate a cosmological baryon asymmetry during inflation @xcite .
the resulting baryon - to - entropy ratio depends on the inflaton field , so that inflaton fluctuations yield both adiabatic and baryon isocurvature ( bi ) perturbations with a nearly - scale invariant spectrum . for the quartic model with @xmath57 ,
bi and adiabatic modes are anti - correlated with relative amplitude @xmath108 and a blue - tilted spectrum @xmath109 @xcite .
this then yields for the relative matter isocurvature spectrum @xmath110 , well within the bound @xmath111 obtained by planck for anti - correlated isocurvature modes with @xmath112 , which is in fact the case that best improves the fit to the data @xcite . the interactions required to produce a baryon asymmetry through dissipation are analogous to those considered in conventional thermal gut baryogenesis or leptogenesis models , with the scalar @xmath23 fields corresponding to e.g. heavy gut bosons or right - handed neutrinos @xcite .
however , since only virtual @xmath23 modes are involved in the dissipative processes , baryogenesis may occur below the gut scale , as opposed to thermal gut baryogenesis models , avoiding the production of dangerous relics such as monopoles .
in particular , we obtain for the temperature at the end of inflation in the quartic model : @xmath113 we note that the effective reheating temperature is roughly an order of magnitude lower since radiation typically takes a few e - folds to take over after the end of slow - roll @xcite .
while gravitinos may still be ubiquitously produced at these temperatures , the inflaton may not decay completely right after inflation if @xmath114 @xcite , as is the case of the quartic model for @xmath96 .
the inflaton may then come to dominate over the radiation bath at a later stage and the entropy produced by its eventual decay may dilute the excess of gravitinos , thus avoiding the potentially associated cosmological problems @xcite .
our results motivate a closer look at thermalization processes and , in particular , we can estimate the total production rate of inflaton particles from the 3-body decay rate of the @xmath115 heavy species in the plasma given above . at horizon - crossing ,
the inflaton is a relativistic degree of freedom , since @xmath116 , and we obtain : @xmath117 where we assumed @xmath118 .
moreover , finite temperature bose factors may considerably enhance this for small couplings @xcite , with e.g. the two - body decay width increasing up to a factor @xmath119 @xcite . also ,
@xmath120 increases during inflation , so that deviations from thermal equilibrium should become less significant .
we then expect inflaton particles to be produced sufficiently fast and remain close to thermal equilibrium with the ambient plasma if the effective couplings @xmath121 are not too small
. both the inflaton and other light fields could actually be in a pre - inflationary thermal state with @xmath1 , with dissipation and the above mentioned processes maintaining a slowly - varying temperature . without dissipation , however , thermal effects would be quickly redshifted away , yielding quite different observational features @xcite .
in this letter , we have shown that the presence of even small dissipative effects at the time when observable scales leave the horizon during inflation may have a significant effect on the spectrum of primordial fluctuations in the warm regime , for @xmath1 .
this generically lowers the tensor - to - scalar ratio and yields a modified consistency relation for warm inflation that may be used to distinguish it in a model - independent way from the standard supercooled scenarios if a tensor component is found and accurately measured .
the main modifications to the scalar spectrum arise from the presence of dissipative noise that sources inflaton fluctuations and from the changes in the phase space distribution of inflaton modes as a consequence of inflaton particle production in the plasma .
we have shown , in particular , that the latter effect may bring the simplest chaotic inflation scenario with a quartic potential into agreement with the planck results for a nearly - thermal distribution .
inflation may thus be triggered from chaotic initial conditions at the planck scale in an observationally consistent way , through simple renormalizable interactions with matter fields that must be present in any inflationary model , as opposed to e.g. a non - minimal coupling to the gravitational sector .
the cosmic baryon asymmetry may also be produced during warm inflation , inducing baryon isocurvature perturbations that are within the current planck bounds for a quartic potential and which may be probed in the near future .
although for the quartic model a nearly - thermal spectrum is observationally preferred , as e.g. for @xmath122 , this is not necessarily true in general .
for example , models such as susy hybrid inflation driven by small radiative corrections @xcite , which follows from eq .
( [ superpotential ] ) , are consistent with the planck data when @xmath52 is negligible and only the fluctuation - dissipation term modifies the spectrum
. in particular , dissipation increases the number of e - folds in this case , whereas in the cold regime only @xmath123 is observationally allowed @xcite .
other low - scale models such as hill - top scenarios are consistent for both the thermal regime @xcite and when the fluctuation - dissipation term in eq .
( [ scalar_power_spectrum_full ] ) is dominant , the same holding for exponential @xcite or inverse power - law potentials , although an alternative reheating mechanism is needed for @xmath59 since dissipation never becomes sufficiently strong in this case .
a detailed analysis of these and other potentials is , however , outside the scope of this letter and will be presented elsewhere .
we have focused on the regime where dissipation is still sub - leading at horizon - crossing , which is simpler to realize since it requires smaller values of @xmath100 and lower field multiplicities .
non - gaussian effects should also be suppressed in this case , with @xmath124 by extrapolating the results in @xcite to weak dissipation , within the bounds obtained by planck @xcite .
the planck collaboration has searched for signals of non - gaussianity in warm inflation models in the strong dissipation regime @xcite , but a dedicated analysis of the bispectrum for @xmath59 is required and further motivated by our results . for strong dissipation the coupling between inflaton and radiation perturbations enhances the growth of fluctuations , making the spectrum more blue - tilted for increasing @xmath48 . to check the validity of our results for the quartic potential , we have evolved the fluctuations numerically extending the analysis in @xcite to weak dissipation , the details of which will be presented in a separate publication .
we find that the primordial spectrum is modified if @xmath125 when the largest scales become super - horizon , since in this case the smallest observable scales leave the horizon 8 - 10 e - folds later when @xmath126 .
shear viscous effects have been shown to suppress the enhanced growth @xcite , but since these imply departures from thermal equilibrium a more detailed analysis is required .
the results presented in figure 1 for @xmath96 are nevertheless in good agreement with the numerical spectrum .
note that even weak dissipative effects can lower the tensor - to - scalar ratio below the expected reach of planck , @xmath127 , or other cmb polarization experiments such as the atacama b - mode search or the south pole telescope , with @xmath128 .
the significant changes to the primordial spectrum from dissipative and thermal effects may also extend beyond the susy realization of warm inflation considered in this letter . in non - susy models , for example , while couplings between the inflaton and other fields must be smaller to prevent large radiative corrections , there may exist parametric regimes where thermalization occurs sufficiently fast to yield near - equilibrium occupation numbers .
one could also envisage alternative models , e.g. the decay of multiple scalar fields that become underdamped at different stages during inflation and produce inflaton particles during the first few oscillations about the minima of their potential .
warm inflation may thus provide a first principles dynamical mechanism to sustain non - trivial occupation numbers for the inflaton and other light particles , based on renormalizable interactions , and it would be interesting to investigate whether phenomenological excited states @xcite could find concrete realizations in this context .
the most important effect of dissipation and/or a non - trivial inflaton particle distribution is the lowering of the tensor - to - scalar ratio in the modified consistency relation in eq .
( [ consistency ] ) , so we expect next year s planck release and future cmb b - mode polarization searches to shed new light on the nature of inflaton fluctuations .
sb and ab are supported by stfc .
mbg and rc are partially supported by micinn ( fis2010 - 17395 ) and `` junta de andaluca '' ( fqm101 ) . ror is partially supported by cnpq and faperj ( brazil ) .
jgr is supported by fct ( sfrh / bpd/85969/2012 ) and partially by the grant ptdc / fis/116625/2010 and the marie curie action nrhep-295189-fp7-people-2011-irses .
a. a. starobinsky , phys .
b * 91 * , 99 ( 1980 ) ; k. sato , mon . not .
soc . * 195 * , 467 ( 1981 ) ; a. h. guth , phys .
* d23 * , 347 ( 1981 ) ; a. albrecht , p. j. steinhardt , phys .
lett . * 48 * , 1220 ( 1982 ) ; a. d. linde , phys .
* b108 * , 389 ( 1982 ) .
p. a. r. ade _ et al .
_ [ planck collaboration ] , arxiv:1303.5082 [ astro-ph.co ] .
a. berera and l. -z .
fang , phys .
lett . * 74 * , 1912 ( 1995 ) [ astro - ph/9501024 ] ; a. berera , phys .
lett . * 75 * , 3218 ( 1995 ) [ astro - ph/9509049 ] .
a. berera , i. g. moss and r. o. ramos , rept .
* 72 * , 026901 ( 2009 ) [ arxiv:0808.1855 [ hep - ph ] ] .
i. g. moss , phys .
b * 154 * , 120 ( 1985 ) .
j. yokoyama and k. -i .
maeda , phys .
b * 207 * , 31 ( 1988 ) .
this model overcomes the concerns raised by thermal corrections when @xmath129 as initially discussed in a. berera , m. gleiser and r. o. ramos , phys .
d * 58 * , 123508 ( 1998 ) [ hep - ph/9803394 ] ; j. yokoyama and a. d. linde , phys .
d * 60 * , 083509 ( 1999 ) [ hep - ph/9809409 ] .
m. bastero - gil , a. berera and r. o. ramos , jcap * 1109 * , 033 ( 2011 ) [ arxiv:1008.1929 [ hep - ph ] ] ; m. bastero - gil , a. berera , r. o. ramos and j. g. rosa , jcap * 1301 * , 016 ( 2013 ) [ arxiv:1207.0445 [ hep - ph ] ] .
h. a. weldon , phys .
d * 28 * , 2007 ( 1983 ) ; a. anisimov , w. buchmuller , m. drewes and s. mendizabal , annals phys . * 324 * , 1234 ( 2009 ) [ arxiv:0812.1934 [ hep - th ] ] .
a. berera , nucl .
b * 585 * , 666 ( 2000 ) [ hep - ph/9904409 ] .
l. m. hall , i. g. moss and a. berera , phys .
d * 69 * , 083525 ( 2004 ) [ astro - ph/0305015 ] .
m. yamaguchi and j. yokoyama , phys .
d * 56 * , 4544 ( 1997 ) [ hep - ph/9707502 ] .
r. o. ramos and l. a. da silva , jcap * 1303 * , 032 ( 2013 ) [ arxiv:1302.3544 [ astro-ph.co ] ] . c. graham and i. g. moss , jcap
* 0907 * , 013 ( 2009 ) [ arxiv:0905.3500 [ astro-ph.co ] ] .
d. h. lyth , phys .
lett . *
78 * , 1861 ( 1997 ) [ hep - ph/9606387 ] .
y. -f .
cai , j. b. dent and d. a. easson , phys . rev .
d * 83 * , 101301 ( 2011 ) [ arxiv:1011.4074 [ hep - th ] ] .
s. hotchkiss , a. mazumdar and s. nadathur , jcap * 1202 * , 008 ( 2012 ) [ arxiv:1110.5389 [ astro-ph.co ] ] .
a. d. linde , phys
. lett .
b * 129 * , 177 ( 1983 ) .
s. bartrum , a. berera and j. g. rosa , phys .
d * 86 * , 123525 ( 2012 ) [ arxiv:1208.4276 [ hep - ph ] ] .
a. ijjas , p. j. steinhardt and a. loeb , phys .
b * 723 * , 261 ( 2013 ) [ arxiv:1304.2785 [ astro-ph.co ] ] .
r. kallosh and a. linde , jcap * 1306 * , 027 ( 2013 ) [ arxiv:1306.3211 [ hep - th ] ] .
m. bastero - gil and a. berera , int .
j. mod .
a * 24 * , 2207 ( 2009 ) [ arxiv:0902.0521 [ hep - ph ] ] .
r. cerezo and j. g. rosa , jhep * 1301 * , 024 ( 2013 ) [ arxiv:1210.7975 [ hep - ph ] ] .
m. bastero - gil , a. berera and j. g. rosa , phys .
d * 84 * , 103503 ( 2011 ) [ arxiv:1103.5623 [ hep - th ] ] .
t. matsuda , phys .
d * 87 * , 026001 ( 2013 ) [ arxiv:1212.3030 [ hep - th ] ] .
m. bastero - gil , a. berera , r. o. ramos and j. g. rosa , phys .
b * 712 * , 425 ( 2012 ) [ arxiv:1110.3971 [ hep - ph ] ] . s. bartrum , a. berera and j. g. rosa , jcap * 1306 * , 025 ( 2013 ) [ arxiv:1303.3508 [ astro-ph.co ] ] .
j. c. bueno sanchez , m. bastero - gil , a. berera , k. dimopoulos and k. kohri , jcap * 1103 * , 020 ( 2011 ) [ arxiv:1011.2398 [ hep - ph ] ] .
k. bhattacharya , s. mohanty and r. rangarajan , phys .
lett . * 96 * , 121302 ( 2006 ) [ hep - ph/0508070 ] .
g. r. dvali , q. shafi and r. k. schaefer , phys .
lett . * 73 * , 1886 ( 1994 ) [ hep - ph/9406319 ] .
j. c. bueno sanchez , m. bastero - gil , a. berera and k. dimopoulos , phys .
d * 77 * , 123527 ( 2008 ) [ arxiv:0802.4354 [ hep - ph ] ] .
i. g. moss and c. xiong , jcap * 0704 * , 007 ( 2007 ) [ astro - ph/0701302 ] .
p. a. r. ade _ et al .
_ [ planck collaboration ] , arxiv:1303.5084 [ astro-ph.co ] . m. bastero - gil , a. berera and r. o. ramos , jcap * 1107 * , 030 ( 2011 ) [ arxiv:1106.0701 [ astro-ph.co ] ] .
s. j. landau , c. g. scoccola and d. sudarsky , phys .
d * 85 * , 123001 ( 2012 ) [ arxiv:1112.1830 [ astro-ph.co ] ] .
a. ashoorioon , k. dimopoulos , m. m. sheikh - jabbari and g. shiu , arxiv:1306.4914 [ hep - th ] . | the amplitude of primordial curvature perturbations is enhanced when a radiation bath at a temperature @xmath0 is sustained during inflation by dissipative particle production , which is particularly significant when a non - trivial statistical ensemble of inflaton fluctuations is also maintained . since
gravitational modes are oblivious to dissipative dynamics , this generically lowers the tensor - to - scalar ratio and yields a modified consistency relation for warm inflation , as well as changing the tilt of the scalar spectrum .
we show that this alters the landscape of observationally allowed inflationary models , with for example the quartic chaotic potential being in very good agreement with the planck results for nearly - thermal inflaton fluctuations , whilst essentially ruled out for an underlying vacuum state .
we also discuss other simple models that are in agreement with the planck data within a renormalizable model of warm inflation .
warm inflation ; chaotic inflation ; cosmic microwave background ; arxiv:1307.5868 [ hep - ph ] |
charm physics plays a unique dual role in the modern investigation of flavor physics .
charm decay and production experiments provide valuable checks and supporting measurements for studies of cp - violation in measurements of ckm parameters in b - physics , as well as outstanding opportunities for searches for new physics .
historically , many methods of heavy quark physics have been first tested in charmed hadrons .
the fact that a b - quark mainly decays into a charm quark makes charm physics an integral part of any b - physics program . as we shall see in many cases ,
direct measurements of charm decay parameters directly affect the studies of fundamental physics in b decays .
any search for new physics falls into two distinct categories : _ direct _ , when new physics particles are directly produced in the experiment and _ indirect _ , when they only affect low energy observables via quantum corrections . in turn
, indirect searches can also be roughly divided into three classes , depending on which observables or processes are used in the search : 1 ) processes forbidden in the standard model ( sm ) , i.e. absent at any order in expansion in any of the coupling constants of sm .
an example of those could be provided by charge - violating transitions @xmath1 or @xmath2 .
2 ) processes forbidden in the standard model at tree level , but possible at higher orders . examples of those processes involve flavor - changing neutral current ( fcnc ) transitions , such as familiar @xmath3 , @xmath4 or @xmath5/@xmath6 mixing .
3 ) processes allowed in the standard model . examples of the tests that belong to this category include examinations of the relations between different observables that are satisfied in the standard model , but not necessary in general .
an example of those are the triangle relations among ckm parameters extracted form several different decays .
the standard model is a very constrained system , which implements a remarkably simple and economic description of all cp - violating processes in the flavor sector by a single cp - violating parameter , the phase of the ckm matrix .
this fact relates all cp - violating observables in bottom , charm and strange systems and provides an excellent opportunity for searches of physics beyond the standard model .
one of the major goals of the contemporary experimental @xmath7 physics program is an accurate determination of the ckm parameters . as we shall see below
, inputs from charm decays are important ingredient in this program , both in the extraction of the angles and sides of the ckm unitarity triangle .
some of the cleanest methods of the determination of the ckm phase @xmath9 $ ] involve the interference of the @xmath10 and @xmath11 quark - level transitions @xcite . a way to arrange for an interference of those seemingly different processes
was first pointed out in @xcite .
it involves interference of two _ hadronic _ decays @xmath12 and @xmath13 , with @xmath14 being any common final state for @xmath15 and @xmath16 decays . since then , many different methods have been proposed , mainly differing by the @xmath17 final state and paths of reaching it : a combination of the cabibbo - favored ( cf ) and doubly - cabibbo suppressed decays ( dcsd ) @xmath18
@xcite , singly - cabibbo suppressed decays ( scsd ) @xmath19 @xcite , cabibbo - favored decays employing large @xmath20 mixing @xmath21 @xcite , as well as multibody versions of these transitions @xcite . for similar methods involving interference of the _ initial _ state , see @xcite .
each of those methods involve observations of several @xmath7 decay modes , so all of the hadronic unknowns , such as the ratio [ bdratio ] a(b^+ d^0 k^+)/a(b^+ k^+)=r_b e^i ( + _ b ) , or @xmath22 , as well as @xmath8 , can be determined from the @xmath7-decay measurements only .
yet , all of those methods will benefit from the separately performed charm decay measurements , the primary reason being the need for large statistics in any experimental determination of @xmath8 .
for example , an ads method @xcite employs a ratio of a dcs to cf decays @xmath23 .
the accuracy of this method will be greatly improved if @xmath24 and @xmath25 are measured separately . while the separate measurement of @xmath24 is already available ,
determination of @xmath25 will became possible at the upcoming tau - charm factories ( tcf ) at cornell ( cleo - c ) and beijing ( bes - iii ) .
the basic idea is to make use of an important experimental advantage of tcf : operation at the @xmath26 threshold , where the @xmath15 and @xmath16 are produced in the quantum - mechanically coherent initial state @xcite .
since the initial @xmath26 state is prepared as @xmath27 where @xmath28 is the relative angular momentum of two @xmath29 mesons , @xmath0 properties of the final states produced in the decay of @xmath30 are anti - correlated , one @xmath29 state decayed into the final state with definite @xmath0 properties immediately identifies or tags @xmath0 properties of the state `` on the other side . ''
that is to say , if one state decayed into , say @xmath31 with @xmath32 , the other state is `` cp - tagged '' as being in the @xmath33 state .
this allows one to measure @xmath34 . in order to see this ,
let us write a triangle relation , @xmath35 which follows from the fact that , in the absence of cp - violation in charm , mass eigenstates of the neutral @xmath29 meson coincide with its cp - eigenstates , [ cpeig ] | d_= | d^0 | .
this implies a relation for the branching ratios , @xmath36 where we used the fact that @xmath37 and neglected @xmath0 violation in mixing , which could undermine the cp - tagging procedure by splitting the cp - tagged state on one side into a linear combination of cp - even and cp - odd states .
its effect , however , is completely negligible here .
now , if both decays of @xmath38 and @xmath39 are measured , @xmath34 can be obtained from the asymmetry @xmath40 both eqs .
( [ cos1 ] ) and ( [ cos2 ] ) can be used to extract @xmath25 at tcf .
similar measurements are possible for other @xmath29 decays @xcite .
in addition , some methods require cross - checks in the charm sector that might restrict the accuracy of these methods . a good example is provided by the original gw method @xcite , which does not take into account the possibility of relatively large @xmath6 mixing @xcite .
this method relies on the simple triangle amplitude relation @xmath41 , again provided by eq .
( [ cpeig ] ) .
an amplitude @xmath42 is measured with @xmath29 decaying to a particular cp - eigenstate .
neglecting @xmath6 mixing , angle @xmath8 can then be extracted , up to a discrete ambiguity , from the measurements of @xmath43 and @xmath44 . in particular , 1 + r_b^2 + 2 r_b c
_ , where @xmath45 and @xmath46 and @xmath47 are defined from the ratio eq .
( [ bdratio ] ) . then , ^2 = .
it is easy to see that @xmath6 mixing , if not properly accounted for , can affect the results of this analysis . indeed , taking @xmath6 mixing into account results in the modification of the definitions of @xmath48 , c_&&(_b ) 2_d + & - & , where @xmath49 with @xmath50 and @xmath51 being the masses and widths of d - meson mass eigenstates @xmath52 , @xmath53 is a cp - parity of @xmath54 , and @xmath55 is a cp - violating phase of @xmath6 mixing .
it is easy to see that @xmath56 can impact the determination of @xmath8 from these modes .
thus , separately constraining @xmath6 mixing parameters will be helpful .
we shall discuss those later .
there are several unitarity triangles that involve charm inputs @xcite . since all cp - violating effects in the flavor sector of the sm are related to the single phase of the ckm matrix , all of the ckm unitarity triangles , including the ones with most charm inputs , have the same area .
this could provides a non - trivial check of the standard model , if measurements of all sides of these triangles are measured with sufficient accuracy .
unfortunately , `` charm triangles '' are `` squashed '' , with one side being much shorter then the other two .
for example , in terms of the wolfenstein parameter @xmath57 , the relation @xmath58 has one side @xmath59 with the other two being @xmath60 , while the relations @xmath61 and @xmath62 have one side @xmath63 with the other two being @xmath64 .
even though the second relation could provide an interesting information about the top quark ckm parameters and the last one can be checked using only the tree - level transitions , it is unlikely that the required accuracy will be achieved in the near future .
in addition , tests similar to the `` first row unitarity '' , @xmath65 ( which at the moment fails at the level of @xmath66 @xcite ) are possible . for example , @xmath67 , it could provide an interesting cross - check on the value of @xmath68 extracted in b - decays , if sufficient accuracy on the experimental measurement of @xmath69 and @xmath70 is achieved . since @xmath71 , both charm and bottom quarks can be regarded as heavy quarks .
naturally , heavy quark symmetry relates observables in b and d transitions . as an example , let us consider measurements in the charm sector affect determinations of the ckm matrix elements relevant to top quark in @xmath5 mixing .
a mass difference of mass eigenstates in @xmath5 system can be written as [ offdiag ] m_d = ^-6/23 |b_d^0|o()|b_d^0 , where @xmath72 ( see ref .
@xcite for complete definitions of the parameters in this expression ) .
the largest uncertainty of about 30% in the theoretical calculation is introduced by the poorly known hadronic matrix element @xmath73 .
evaluation of this matrix element is a genuine non - perturbative task , which can be approached with several different techniques .
the simplest approach ( `` factorization '' ) reduces the matrix element @xmath74 to the product of matrix elements measured in leptonic @xmath7 decays @xmath75 @xmath76 , where we employed the definition of the decay constant @xmath77 , 0||b_l _
d_l|b^0(*p*)= i p_f_b/2 . a deviation from the factorization ansatz is usually described by the parameter @xmath78 defined as @xmath79 ; in factorization @xmath80 .
similar considerations lead to an introduction of the parameter @xmath81 defined for mixing of @xmath82 mesons .
it is important to note that the parameters @xmath83 depend on the chosen renormalization scale and scheme .
it is convenient to introduce renormalization - group invariant parameters @xmath84 [ bhat ] b_b_q = ^-6/23 b_b_q .
we provide averages of @xmath84 , as well as the ratio @xmath85 from the review @xcite as well as from two more recent evaluations @xcite in table [ table1 ] .
thus , at least naively , one can determine ckm matrix element @xmath86 by measuring @xmath77 and @xmath87 and computing @xmath83 .
this direct approach , however , meets several difficulties .
first , leptonic decay constant @xmath77 can in principle be extracted from leptonic decays of charged b mesons .
the corresponding decay width is ( b l ) = f_b^2 |v_ub|^2 m_l^2 m_b ( 1- ) .
this width is seen to be quite small due to the smallness of the ckm factor @xmath88 and helicity suppression factor of @xmath89 .
in addition , experimental difficulties are also expected due to the backgrounds stemming from the presence of a neutrino in the final state .
.[table1 ] renormalization - group independent b - parameters . [ cols="^,^,^",options="header " , ] second , computation of @xmath83 is quite difficult and requires the use of non - perturbative techniques such as lattice or qcd sum rules .
current uncertainties in the determinations of @xmath77 and @xmath83 are quite large .
it turns out that evaluation of the ratio of = ^2 ||^2 , is favored by lattice community , as many systematic errors cancel in this ratio .
this gives a ratio of @xmath90 , which provides a non - trivial constraint on ckm parameters in the @xmath91 plane .
instead , one can make use of ample statistics available in charm production experiments , as heavy quark and @xmath92 flavor symmetries relate the ratio of charm decay constants @xmath93 to beauty decay constants @xmath94 = 1+o(m_s)(1/m_b-1/m_c ) .
note that su(3)-violating corrections can also be evaluated in chiral perturbation theory @xcite .
one still needs to rely on the theoretical determination of @xmath83 .
similar techniques of relating @xmath7 and @xmath29 decays can also be used to extract other ckm matrix elements , like @xmath95 @xcite , studies of lifetime patterns of heavy hadrons @xcite , and tuning lattice qcd calculations @xcite .
another area of modern phenomenology where charm decays play an important role is the indirect search for physics beyond the standard model . indeed , large statistics usually available in charm physics experiment makes it possible to probe small effects that might be generated by the presence of new physics particles and interactions .
a program of searches for new physics in charm is complimentary to the corresponding programs in bottom or strange systems .
this is in part due to the fact loop - dominated processes such as @xmath6 mixing or flavor - changing neutral current ( fcnc ) decays are sensitive to the dynamics of ultra - heavy _ down - type particles_. also , in many dynamical models , including the standard model , the effects in @xmath96 , @xmath97 , and @xmath98 systems are correlated . the low energy effect of new physics particles can be naturally written in terms of a series of local operators of increasing dimension generating @xmath99 ( decays ) or @xmath100 ( mixing ) transitions . for @xmath6 mixing these operators , as well as
the one loop standard model effects , generate contributions to the effective operators that change @xmath15 state into @xmath16 state leading to the mass eigenstates @xmath101 where the complex parameters @xmath102 and @xmath103 are obtained from diagonalizing the @xmath104 mass matrix .
the mass and width splittings between these eigenstates are given in eq .
( [ definition ] ) .
it is known experimentally that @xmath6 mixing proceeds extremely slowly , which in the standard model is usually attributed to the absence of superheavy quarks destroying gim cancellations @xcite .
it is instructive to see how new physics can affect charm mixing . since the lifetime difference @xmath105
is constructed from the decays of @xmath29 into physical states , it should be dominated by the standard model contributions , unless new physics significantly modifies @xmath106 interactions . on the contrary ,
the mass difference @xmath107 can receive contributions from all energy scales .
thus , it is usually conjectured that new physics can significantly modify @xmath107 leading to the inequality @xmath108 , of the order of a percent , is observed @xcite . ] .
the same considerations apply to fcnc decays as well , where new physics could possibly contribute to the decay rates of @xmath109 ( with @xmath110 being exclusive or inclusive final state ) as well as other observables @xcite .
one technical problem here is that in the standard model these decays are overwhelmingly dominated by long - distance effects , which makes them extremely difficult to predict model - independently .
this problem can be turned into a virtue @xcite .
another possible manifestation of new physics interactions in the charm system is associated with the observation of ( large ) cp - violation .
this is due to the fact that all quarks that build up the hadronic states in weak decays of charm mesons belong to the first two generations .
since @xmath111 cabbibo quark mixing matrix is real , no cp - violation is possible in the dominant tree - level diagrams that describe the decay amplitudes . in the standard model
cp - violating amplitudes can be introduced by including penguin or box operators induced by virtual @xmath98-quarks . however , their contributions are strongly suppressed by the small combination of ckm matrix elements @xmath112 .
it is thus widely believed that the observation of ( large ) cp violation in charm decays or mixing would be an unambiguous sign for new physics .
this fact makes charm decays a valuable tool in searching for new physics , since the statistics available in charm physics experiment is usually quite large . as in b - physics ,
cp - violating contributions in charm can be generally classified by three different categories : ( i ) cp violation in the decay amplitudes .
this type of cp violation occurs when the absolute value of the decay amplitude for @xmath29 to decay to a final state @xmath14 ( @xmath113 ) is different from the one of corresponding cp - conjugated amplitude ( `` direct cp - violation '' ) ; ( ii ) cp violation in @xmath6 mixing matrix .
this type of cp violation is manifest when @xmath114 ; and ( iii ) cp violation in the interference of decays with and without mixing .
this type of cp violation is possible for a subset of final states to which both @xmath15 and @xmath16 can decay . for a given final state @xmath14 , cp violating contributions
can be summarized in the parameter _
f = = r_m e^i(+)| | , where @xmath113 and @xmath115 are the amplitudes for @xmath116 and @xmath117 transitions respectively and @xmath118 is the strong phase difference between @xmath113 and @xmath115 .
here @xmath119 represents the convention - independent weak phase difference between the ratio of decay amplitudes and the mixing matrix .
presently , experimental information about the @xmath6 mixing parameters @xmath107 and @xmath105 comes from the time - dependent analyses that can roughly be divided into two categories .
first , more traditional studies look at the time dependence of @xmath120 decays , where @xmath14 is the final state that can be used to tag the flavor of the decayed meson .
the most popular is the non - leptonic doubly cabibbo suppressed decay @xmath121 .
time - dependent studies allow one to separate the dcsd from the mixing contribution @xmath122 , @xmath123 = e^{-\gamma t}|a_{k^-\pi^+}|^2 \qquad\qquad\qquad\qquad\qquad \nonumber
\\ \times ~\left [ r+\sqrt{r}r_m(y'\cos\phi - x'\sin\phi)\gamma t + \frac{r_m^2}{4}(y^2+x^2)(\gamma t)^2 \right],\end{aligned}\ ] ] where @xmath124 is the ratio of dcs and cabibbo favored ( cf ) decay rates .
since @xmath107 and @xmath105 are small , the best constraint comes from the linear terms in @xmath125 that are also _ linear _ in @xmath107 and @xmath105 . a direct extraction of @xmath107 and @xmath105 from eq .
( [ kpi ] ) is not possible due to unknown relative strong phase @xmath25 of dcs and cf amplitudes @xcite , as @xmath126 , @xmath127 .
as discussed above , this phase can be measured independently @xcite .
the corresponding formula can also be written @xcite for @xmath16 decay with @xmath128 and @xmath129 .
second , @xmath15 mixing can be measured by comparing the lifetimes extracted from the analysis of @xmath29 decays into the cp - even and cp - odd final states .
this study is also sensitive to a _ linear _ function of @xmath105 via -1= y - x .
time - integrated studies of the semileptonic transitions are sensitive to the _ quadratic _ form @xmath130 and at the moment are not competitive with the analyses discussed above .
the construction of new tau - charm factories cleo - c and bes - iii will introduce new _ time - independent _ methods that are sensitive to a linear function of @xmath105 .
one can again use the fact that heavy meson pairs produced in the decays of heavy quarkonium resonances have the useful property that the two mesons are in the cp - correlated states @xcite ( see eq .
( [ tcfinitial ] ) ) . by tagging one of the mesons as a cp eigenstate
, a lifetime difference may be determined by measuring the leptonic branching ratio of the other meson .
its semileptonic _ width _ should be independent of the cp quantum number since it is flavor specific , yet its _ branching ratio _ will be inversely proportional to the total width of that meson .
since we know whether this @xmath131 state is tagged as a ( cp - eigenstate ) @xmath132 from the decay of @xmath133 to a final state @xmath134 of definite cp - parity @xmath135 , we can easily determine @xmath105 in terms of the semileptonic branching ratios of @xmath132 .
this can be expressed simply by introducing the ratio [ defcor ] r^l_= , where @xmath136 in @xmath137 stands for an inclusive set of all final states .
a deviation from @xmath138 implies a lifetime difference .
keeping only the leading ( linear ) contributions due to mixing , @xmath105 can be extracted from this experimentally obtained quantity , @xmath139 the current experimental upper bounds on @xmath107 and @xmath105 are on the order of a few times @xmath140 , and are expected to improve significantly in the coming years . to regard a future discovery of nonzero @xmath107 or @xmath105 as a signal for new physics , we would need high confidence that the standard model predictions lie well below the present limits . as was recently shown @xcite , in the standard model , @xmath107 and @xmath105
are generated only at second order in su(3)@xmath141 breaking , x , y ~^2_c ^2 , where @xmath142 is the cabibbo angle .
therefore , predicting the standard model values of @xmath107 and @xmath105 depends crucially on estimating the size of su(3)@xmath141 breaking .
although @xmath105 is expected to be determined by the standard model processes , its value nevertheless affects significantly the sensitivity to new physics of experimental analyses of @xmath29 mixing @xcite .
theoretical predictions of @xmath107 and @xmath105 within and beyond the standard model span several orders of magnitude @xcite .
roughly , there are two approaches , neither of which give very reliable results because @xmath143 is in some sense intermediate between heavy and light . the `` inclusive '' approach is based on the operator product expansion ( ope ) . in the @xmath144 limit , where @xmath145 is a scale characteristic of the strong interactions , @xmath146 and @xmath147 can be expanded in terms of matrix elements of local operators@xcite .
such calculations yield @xmath148 .
the use of the ope relies on local quark - hadron duality , and on @xmath149 being small enough to allow a truncation of the series after the first few terms .
the charm mass may not be large enough for these to be good approximations , especially for nonleptonic @xmath29 decays .
an observation of @xmath105 of order @xmath140 could be ascribed to a breakdown of the ope or of duality , but such a large value of @xmath105 is certainly not a generic prediction of ope analyses .
the `` exclusive '' approach sums over intermediate hadronic states , which may be modeled or fit to experimental data@xcite .
since there are cancellations between states within a given @xmath92 multiplet , one needs to know the contribution of each state with high precision . however , the @xmath29 is not light enough that its decays are dominated by a few final states . in the absence of sufficiently precise data on many decay rates and on strong phases ,
one is forced to use some assumptions . while most studies find @xmath148 , refs .
@xcite obtain @xmath107 and @xmath105 at the @xmath140 level by arguing that su(3)@xmath141 violation is of order unity , but the source of the large su(3)@xmath141 breaking is not made explicit .
it was also shown that phase space effects alone provide enough su(3)@xmath141 violation to induce @xmath150
large effects in @xmath105 appear for decays close to @xmath29 threshold , where an analytic expansion in su(3)@xmath141 violation is no longer possible . thus ,
theoretical calculations of @xmath107 and @xmath105 are quite uncertain , and the values near the current experimental bounds can not be ruled out . therefore , it will be difficult to find a clear indication of physics beyond the standard model in @xmath6 mixing measurements alone
. the only robust potential signal of new physics in charm system at this stage is cp violation .
cp violation in @xmath29 decays and mixing can be searched for by a variety of methods .
for instance , time - dependent decay widths for @xmath151 are sensitive to cp violation in mixing ( see eq.([kpi ] ) ) . provided that the @xmath107 and @xmath105 are comparable to experimental sensitivities , a combined analysis of @xmath151 and @xmath152 can yield interesting constraints on cp - violating parameters @xcite .
most of the techniques that are sensitive to cp violation make use of the decay asymmetry , @xmath153 most of the properties of eq .
( [ acp ] ) , such as dependence on the strong final state phases , are similar to the ones in b - physics @xcite .
current experimental bounds from various experiments , all consistent with zero within experimental uncertainties , can be found in @xcite .
other interesting signals of @xmath0-violation that are being discussed in connection with tau - charm factory measurements are the ones that are using quantum coherence of the initial state .
an example of this type of signal is a decay @xmath154 at @xmath30 with @xmath155 and @xmath156 being the different final cp - eigenstates of the same cp - parity .
this type of signals are very easy to detect experimentally .
the corresponding cp - violating decay rate for the final states @xmath155 and @xmath156 is @xmath157~\gamma_{f_1 } \gamma_{f_2}.\end{aligned}\ ] ] the result of eq .
( [ cprate ] ) represents a generalization of the formula given in ref .
it is clear that both terms in the numerator of eq .
( [ cprate ] ) receive contributions from cp - violation of the type i and iii , while the second term is also sensitive to cp - violation of the type ii .
moreover , for a large set of the final states the first term would be additionally suppressed by su(3)@xmath141 symmetry , as for instance , @xmath158 in the su(3)@xmath141 symmetry limit .
this expression is of the _ second _ order in cp - violating parameters ( it is easy to see that in the approximation where only cp violation in the mixing matrix is retained , @xmath159 ) . as it follows from the existing experimental constraints on rate asymmetries , cp - violating phases
are quite small in charm system , regardless of whether they are produced by the standard model mechanisms or by some new physics contributions . in that respect
, it looks unlikely that the sm signals of cp violation would be observed at cleo - c with this observable . while the searches for direct cp violation via the asymmetry of eq .
( [ acp ] ) can be done with the charged d - mesons ( which are self - tagging ) , investigations of the other two types of cp - violation require flavor tagging of the initial state .
this severely cuts the available dataset .
it is therefore interesting to look for signals of cp violation that do not require identification of the initial state .
one possible cp - violating signal involves the observable obtained by summing over the initial states , @xmath160 for @xmath161 . a cp - odd observable that can be formed out of @xmath162
is an asymmetry @xcite @xmath163 note that this asymmetry does not require quantum coherence of the initial state and therefore is accessible in any d - physics experiment .
the final states must be chosen such that @xmath164 is not trivially zero .
it is easy to see that decays of @xmath29 into the final states that are cp - eigenstates would result in zero asymmetry , while the final states like @xmath165 or @xmath166 would not . a non - zero value of @xmath164 in eq .
( [ totasym ] ) can be generated by both direct and indirect cp - violating contributions .
these can be separated by appropriately choosing the final states . for example , indirect cp violating amplitudes are tightly constrained in the decays dominated by the cabibbo - favored tree level amplitudes , while singly cabibbo suppressed amplitudes also receive contributions from direct cp violating amplitudes .
i would like to thank my collaborators as well as the organizers of this workshop for making it a success .
this work was supported in part by the u.s .
department of energy under grant de - fg02 - 96er41005 .
r. fleischer , these proceedings , arxiv : hep - ph/0306270 . t. j. gershon , these proceedings , arxiv : hep - ex/0307020 . m. gronau and d. wyler , phys . lett .
b * 265 * , 172 ( 1991 ) .
d. atwood , i. dunietz and a. soni , phys .
lett . * 78 * , 3257 ( 1997 ) .
y. grossman , z. ligeti and a. soffer , phys .
d * 67 * , 071301 ( 2003 ) .
r. fleischer , nucl .
b * 659 * , 321 ( 2003 ) .
a. giri , y. grossman , a. soffer and j. zupan , arxiv : hep - ph/0303187 ; j. zupan , theese proceedings ; d. atwood and a. soni , these proceedings , arxiv : hep - ph/0307154 .
a. f. falk and a. a. petrov , phys
. rev .
lett . * 85 * , 252 ( 2000 ) ; a. f. falk , phys .
d * 64 * , 093011 ( 2001 ) .
j. p. silva and a. soffer , phys .
d * 61 * , 112001 ( 2000 ) .
m. gronau , y. grossman and j. l. rosner , phys .
b * 508 * , 37 ( 2001 ) ; e. golowich and s. pakvasa , phys .
b * 505 * , 94 ( 2001 ) .
j. l. rosner and d. a. suprun , arxiv : hep - ph/0303117 . i. i. bigi and a. i. sanda , arxiv : hep - ph/9909479 .
h. abele _ et al .
_ , phys .
lett . * 88 * , 211801 ( 2002 ) .
g. buchalla , a. j. buras and m. e. lautenbacher , rev .
phys . * 68 * , 1125 ( 1996 ) ; a. j. buras , m. jamin and p. h. weisz , nucl .
b * 347 * , 491 ( 1990 ) ; m. ciuchini , e. franco , g. martinelli , l. reina and l. silvestrini , z. phys .
c * 68 * , 239 ( 1995 ) .
m. battaglia _ et al .
_ , arxiv : hep - ph/0304132 . s. aoki _ et al . _
[ jlqcd collaboration ] , arxiv : hep - ph/0307039 .
j. g. korner , a. i. onishchenko , a. a. petrov and a. a. pivovarov , arxiv : hep - ph/0306032 . b. grinstein , phys . rev
. lett . *
71 * , 3067 ( 1993 ) .
z. ligeti , i. w. stewart and m. b. wise , phys .
b * 420 * , 359 ( 1998 ) .
d. pedrini [ the focus collaboration ] , these proceedings , arxiv : hep - ph/0307137 ; m. b. voloshin , phys .
rept . * 320 * , 275 ( 1999 ) ; b. guberina , b. melic and h. stefancic , phys .
b * 484 * , 43 ( 2000 ) .
mackenzie , these proceedings ; a. juttner and j. rolf , these proceedings , arxiv : hep - ph/0306299 ; a. datta , d. kumbhakar , z. phys .
c*27 * , 515 ( 1985 ) ; a. a. petrov , phys .
d*56 * , 1685 ( 1997 ) ; e. golowich and a. a. petrov , phys . lett .
b * 427 * , 172 ( 1998 ) .
s. bergmann , y. grossman , z. ligeti , y. nir , a. petrov , phys .
b * 486 * , 418 ( 2000 ) .
a. f. falk , y. grossman , z. ligeti and a. a. petrov , phys .
d * 65 * , 054034 ( 2002 ) .
s. fajfer , these proceedings , arxiv : hep - ph/0306263 . s. fajfer , s. prelovsek , p. singer and d. wyler , phys .
b * 487 * , 81 ( 2000 ) .
a. f. falk , y. nir and a. a. petrov , jhep * 9912 * , 019 ( 1999 ) .
d. atwood and a. a. petrov , arxiv : hep - ph/0207165 .
h. n. nelson , in _ proc . of the 19th
symp . on photon and lepton interactions at high energy lp99
jaros and m.e .
peskin , arxiv : hep - ex/9908021 .
h. georgi , phys .
b297 , 353 ( 1992 ) ; t. ohl , g. ricciardi and e. simmons , nucl .
b403 , 605 ( 1993 ) ; i. bigi and n. uraltsev , nucl .
b * 592 * , 92 ( 2001 ) , for a recent review see a. a. petrov , _ proc . of 4th workshop on continuous advances in qcd _ , minneapolis , minnesota , 12 - 14 may 2000 , arxiv : hep - ph/0009160 .
j. donoghue , e. golowich , b. holstein and j. trampetic , phys .
d33 , 179 ( 1986 ) ; l. wolfenstein , phys .
b164 , 170 ( 1985 ) ; p. colangelo , g. nardulli and n. paver , phys .
b242 , 71 ( 1990 ) ; t.a .
kaeding , phys .
b357 , 151 ( 1995 ) .
a. a. anselm and y. i. azimov , phys .
b * 85 * , 72 ( 1979 ) ; i. i. bigi and a. i. sanda , _ cp violation _
( cambridge university press , 2000 ) .
d. pedrini , j. phys .
g * 27 * , 1259 ( 2001 ) .
i. i. bigi and a. i. sanda , phys .
b * 171 * , 320 ( 1986 ) .
a. a. petrov , proc . of the _
5th workshop on continuous advances in qcd _ ,
102 - 114 ; arxiv : hep - ph/0209049 . | we review several key aspects of charm phenomenology which are important for the general program of determination of matrix elements of the cabbibo - kobayashi - maskava ( ckm ) matrix .
we argue that charm physics plays an important role in reducing uncertainties in non - perturbative qcd parameters whose values are required for understanding of underlying quark - level processes .
we also review the current status of searches for new physics in mixing and @xmath0 violation in charmed mesons both at the currently operating and proposed facilities . |
the study of stellar populations in galaxies is going through very exciting transformations .
the advent of extensive and high - quality data sets for galaxies in the local and distant universe poses a demand for ever improving stellar population synthesis models . in this paper
, we present a new tool , named ez_ages , that simultaneously address two on - going needs in the field of stellar population synthesis : 1 ) it extracts accurate ages and abundances of several elements from integrated spectra of stellar populations , and 2 ) it performs this task in an automatic fashion , which is well - suited for applications to large data sets .
early stellar population synthesis models for single - burst populations so called `` single stellar populations '' or ssps were based on low resolution stellar spectra .
the models of @xcite and @xcite used optical absorption - line indices with 810 resolution and limited spectral coverage , while @xcite synthesized model spectra with broad wavelength coverage ( ultraviolet through infrared ) and resolution of 1020 in the optical .
all models were for solar - abundance ratios only .
since then , a new generation of ssp models have pushed to higher resolution ( 23 for @xcite and @xcite ) and to variable abundance patterns ( @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite and @xcite ) .
it is well known that many stellar populations , including the galactic halo and bulge , galactic globular clusters , massive early - type galaxies , and the bulges of spiral galaxies , have super - solar [ @xmath2/fe ] . in resolved systems ,
where abundance analysis of individual stars is possible , it has been known for a while that not all @xmath2 elements are equally enhanced . on the other hand , in unresolved systems
it has been only recently that evidence started accumulating to the fact that observed stellar populations have abundance patterns which can not be described by only two parameters ( ie .
[ fe / h ] and [ @xmath2/fe ] ) but rather require more complicated abundance patterns .
in particular , it seems that not all the @xmath2 elements are equally enhanced in early - type galaxies ( e.g. , @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) . in the
best studied , closest , resolved spheroidal system , the galactic bulge , there is also recent evidence that not all @xmath2-elements are equally enhanced @xcite . in particular , @xcite found that , while mg is strongly enhanced relative to fe ( [ mg / fe ] @xmath4 + 0.3 across a range of [ fe / h ] ) , other @xmath2 elements such as si , ca , ti , and o tend to follow a disk - like trend , with [ x / fe ] decreasing with increasing [ fe / h ] .
interestingly , in the case of oxygen arguably the most important among the @xmath2 elements they found [ o / fe ] @xmath5 at [ fe / h ] = 0 .
@xcite show a similar result for milky way disk stars in both the thin and thick disk .
@xcite demonstrates that warm ionized gas in a sample of 7 local early - type galaxies has mean [ o / h ] @xmath6 , while the stellar populations of the galaxies have super - solar [ mg / h ] .
clues to variations of the abundances of other elements , such as c and n , particularly as a function of environment , have been found in a number of studies .
the lick indices c@xmath74668 and cn@xmath7 , which are sensitive to c and o ( and n , in the case of cn@xmath7 ) , have been shown to be weaker in early - type galaxies in dense environments as compared to galaxies in the field , while the strength of the mg _ b _ index remains unchanged across environments @xcite . finally , a number of studies have found evidence that ca seems to behave as an fe - group element in elliptical galaxies , rather than following the trend of other @xmath2 elements ( @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) .
this result has recently been questioned by @xcite and @xcite , who show that the ca4227 index is strongly affected by cn lines , which may be partially responsible for the low [ ca / fe ] suggested by other authors .
although @xmath2-enhanced models are an important improvement over solar - scaled - only , it is fundamental that models with more degrees of freedom be developed , which are able to explore the complexity suggested by the data . by establishing a more detailed snapshot of the abundance patterns of stars in galaxies , these models can pose fundamental constraints on models of galaxy formation .
@xcite presents a new set of ssp models which do exactly that .
these models explicitly include adjustable abundance patterns for multiple elements , allowing the user to separately vary the abundances of c , n , o , mg , ca , na , si , cr , and ti . this paper describes an algorithm for searching through the space of variable abundance models to find the abundance mix which best fits the lick indices measured in a given input stellar population spectrum .
this algorithm has been implemented in an idl code package , called `` ez_ages '' which is publicly available for use .
section [ seq_grid ] describes the sequential grid inversion algorithm . in [ compare_clusters ] , we present ages and abundances estimated by ez_ages for the galactic globular clusters ngc 6121 , 47 tuc , ngc 6441 , and ngc 6528 , as well as the open cluster m67 .
these values are compared to results in the literature determined from isochrone fitting of cluster photometry and from high - resolution spectral abundance analysis for individual stars in the clusters . in [ compare_galaxies ]
we apply ez_ages to galaxy data from the literature , contrasting our results with those by other authors .
section [ conc ] summarizes our conclusions .
the beta - version of ez_ages is available for download , along with instructions on installation and use .
the code ez_ages is designed for use with the single stellar population models described in ( * ? ? ? * hereafter s07 ) .
the details of the model are described in s07 ; here we briefly discuss the aspects of the model relevant to its practical use .
these models include a choice of two sets of isochrones by the padova group : one is the solar - scaled isochrones of @xcite and the other is the @xmath2-enhanced set of isochrones by @xcite with average [ @xmath8fe@xmath9 = + 0.42 $ ] and with individual abundance ratios chosen to match metal - poor field stars
. however , the models based on the @xmath2-enhanced isochrone must be used with caution because @xcite have shown that , due to a problem with the opacity tables adopted , those isochrones predict temperatures that are slightly too high at both the main - sequence turnoff and the giant branch .
the effect is strongest at solar metallicity , where it results in ages that are slightly too old . for details ,
see 4.3 of s07 .
developing isochrones with variable abundance ratios is an active area of current research ( see recent work by @xcite ) and we hope to incorporate a broader range of isochrones into ez_ages in the near future .
in addition to a choice of isochrones , the s07 models allow the user to select an input abundance pattern
. the abundance ratios that can be varied are : [ mg / fe ] , [ c / fe ] , [ n / fe ] , [ o / fe ] , [ ca / fe ] , [ na / fe ] , [ si / fe ] , [ cr / fe ] , and [ ti / fe ] . the s07 models output predicted lick index measurements for a grid of age and [ fe / h ] values , based upon the input abundance ratios .
the models span the age range @xmath10 gyr , with @xmath11 [ fe / h ] @xmath12 for the solar - scaled isochrone and @xmath13 [ fe / h ] @xmath14 for the @xmath2-enhanced isochrone , though only models older than 0.8 gyr are computed with variable abundance ratios .
the lick indices included in the s07 model are h@xmath15 , h@xmath16 , h@xmath17 , h@xmath18 , h@xmath19 , cn@xmath20 , cn@xmath7 , ca4227 , g4300 , fe4383 , c@xmath74668 , fe5015 , mg@xmath7 , mg _ b _ , fe5270 , and fe5335 , as defined in table 1 of @xcite and table 1 of @xcite .
the output model indices can then be compared with lick indices measured in the integrated spectrum of a real stellar population ( i.e. , a stellar cluster or a galaxy ) to determine the best - fitting mean light - weighted age and [ fe / h ] for the stellar population , for a given set of model abundance ratios .
the lick indices included in the s07 model are summarized in table [ indextable ] , along with their main abundance sensitivities .
these are taken from the works of @xcite and @xcite .
the sensitivities from @xcite are taken from their tables 4 and 5 , which give abundance sensitivities for turnoff stars and giant branch stars respectively .
we have chosen to include only those two evolutionary phases in the table , as they dominate the stellar population model spectrum ( accounting for @xmath21% of the integrated light ) , even though all three evolutionary phases are taken into account in the model computations .
in addition to the sensitivities to individual element abundances shown in the table , all indices are sensitive to changes in the total metallicity .
we interpret the effect of [ z / h ] on fe5270 and fe5335 as being due primarily to changes in fe because none of the other elements investigated by kmt05 appear to affect these indices .
table [ indextable ] shows reasonable agreement between the two different sets of sensitivity tables .
abundance variations enter the model in two ways , firstly through the choice of isochrone ( solar - scaled or @xmath2-enhanced ) . at this stage ,
the model takes into account the effect of the abundance mix on stellar evolution specifically , how the chemical mix affects temperatures and luminosities of stars of different masses and at different evolutionary stages .
admittedly , this is done in an approximate fashion , given that unfortunately we only have isochrones computed for two abundance patterns .
secondly , the effect of individual elemental abundances on line and continuum opacities in stellar atmospheres is incorporated into model lick index predictions , using the response functions computed by @xcite . at this second stage , it is possible to vary elemental abundances independently ; at the first stage , the only choices are [ @xmath2/fe ] = 0 and @xmath22 , with total metallicity varying .
it should be stated clearly that these models are cast in terms of [ fe / h ] , and not total metallicity , [ z / h ] .
this reflects our choice to deal explicitly with quantities that can be inferred from measurements taken in the integrated spectra of galaxies total metallicity not being one of them , given our current inability to use integrated spectra of stellar populations to constrain the most abundant of all metals , oxygen ( see discussion in 4.4.1 of s07 ) .
another advantage with casting models in terms of [ fe / h ] is that each elemental abundance can be treated separately , so that the effect of its variation can be studied in isolation from every other elemental abundance ( at the cost of varying the total metallicity ) . in the case of models cast in terms of [ z / h ] , it is impossible to vary the abundance of a single element , because enhancing one element means decreasing the abundances of all other elements to keep total metallicity constant . as an example , in models cast in terms of [ z / h ] , a solar - metallicity ( [ z / h ] = 0 ) @xmath2-enhanced population has only a slightly higher abundance of @xmath2 elements than the sun , while having a much lower iron - abundance .
it is the large iron abundance difference that is responsible for the bulk of the difference between those two hypothetical models , _ not _ the small difference in @xmath2-element abundances . for this reason ,
claims that higher order balmer lines are strongly affected by @xmath2-element abundances @xcite should be taken with caution ( see 4.3 s07 for a thorough discussion of this point ) . in practice
, one would like to use absorption line measurements in the integrated spectrum of a given stellar population to determine not only the age and [ fe / h ] for a fixed set of abundance ratios , but to find the best set of input abundances as well .
the brute - force method of finding the best - fitting abundances would involve generating a set of models that span the available space of age , [ fe / h ] , and abundance patterns , and then identifying the model whose predicted lick indices best match those measured in the spectrum .
however , with nine variable abundance ratios , the search space quickly becomes very large .
if each of those abundance ratios is represented by only four possible values , with 23 model ages and four possible values of [ fe / h ] , that would mean creating more than 24 million models !
instead , we choose to perform a directed search for the best model , taking advantage of the fact that various lick indices are sensitive to only a few , often different , elemental abundances , as indicated in table [ indextable ] .
these sensitivity variations have been modeled by several groups using spectrum synthesis , based on model stellar atmospheres ( see recent papers by @xcite and @xcite ) . a more detailed discussion of the motivation behind this method can be found in s07 .
ez_ages begins by computing a set of models with the chosen isochrone ( either solar - scaled or @xmath2-enhanced ) , using solar abundance ratios for the stellar atmospheres .
it then uses a pair of lines that are sensitive to age and [ fe / h ] , but relatively _ insensitive _ to other elemental abundances a balmer line and an fe line to determine a fiducial age and [ fe / h ] for the stellar population in question . from table 1
we can see that h@xmath19 and the iron lines fe5270 and fe5335 are good choices for determining the fiducial , as they are mostly insensitive to other element abundances . in ez_ages , the default choice of indices for calculating fiducial age and [ fe / h ] are h@xmath19 and @xmath23fe@xmath24 , an average of fe5270 and fe5335 .
the top left panel of figure [ grids ] shows a plot of model grids for h@xmath19 and @xmath23fe@xmath24 .
dotted lines connect constant - age models ( from top to bottom : 1.2 , 2.2 , 3.5 , 7.0 , and 14.1 gyr ) while solid lines connect models with the same [ fe / h ] ( from left to right : -1.3 , -0.7 , -0.4 , 0.0 , and + 0.2 ) .
the square shows a sample data point for a galaxy with age @xmath25 gyrs and [ fe / h ] @xmath26 . in this figure
, it is easy to see which box of the grid encloses the data point in question .
this gives a bounded range in age and [ fe / h ] for the data point .
a two - dimensional linear interpolation within the gridbox is then used to convert the data point in index - index space into fiducial values for the age and [ fe / h ] .
once a fiducial age and [ fe / h ] have been determined from the h@xmath19@xmath23fe@xmath24 model grid , one can then make a similar grid plot with an index which is sensitive to non - solar abundance ratios , for instance a balmer line plotted against mg _
b_. mg _ b _ is strongly affected by the [ mg / fe ] abundance ratio , at fixed [ fe / h ] .
if the chosen s07 model is a good match to the data , then the age and [ fe / h ] estimated by grid inversion from the h@xmath19mg _ b _ plot should match the fiducial age and [ fe / h ] from the h@xmath19@xmath23fe@xmath24 plot . in our example in figure [ grids ] , this is clearly not the case ; the h@xmath19mg _ b _ plot in the top right panel of figure [ grids ] gives a larger value of [ fe / h ] than the h@xmath19@xmath23fe@xmath24 plot in the top left panel .
this is an indication that the example galaxy has a larger [ mg / fe ] ratio than is predicted in a solar - scaled model .
the sequential grid inversion algorithm then increases the input value of [ mg / fe ] and recomputes the s07 model .
increasing [ mg / fe ] will increase the strength of the mg _ b _ index _ at fixed [ fe / h ] _ , which will have the effect of sliding the entire set of model grids to the right in the h@xmath19mg _ b _ plot , while keeping them essentially unchanged in the h@xmath19@xmath23fe@xmath24 plane .
this _ lowers _ the value of [ fe / h ] that is estimated by grid inversion in the h@xmath19mg _ b _ plot , bringing it into agreement with that obtained in the h@xmath19@xmath23fe@xmath24 diagram .
this can been seen in the lower panels of figure [ grids ] , which show h@xmath19@xmath23fe@xmath24 and h@xmath19mg @xmath27 plots for an s07 model computed with [ mg / fe]@xmath28 . for the data shown in this plot ,
the age and [ fe / h ] values indicated by the model grids are consistent for both metal lines . in the sequential grid - inversion algorithm implemented in ez_ages , [ mg / fe ] is increased incrementally until the [ fe / h ] estimated from the h@xmath19mg _ b _ plot matches the fiducial [ fe / h ] from the h@xmath19@xmath23fe@xmath24 plot .
this process is then repeated for other lick indices and other abundance ratios .
the key to the sequential grid inversion algorithm is to proceed with the abundance fitting in such a way as to only adjust one abundance at a time .
once fiducial values for the age and [ fe / h ] of the system have been determined , the goal is to next fit a lick index that introduces only one additional abundance .
mg _ b _ is a good choice because it is dominated by mg and fe ( see table 1 ) , with the latter being determined from the h@xmath19@xmath23fe@xmath24 grid inversion . by adjusting [ mg / fe ] until the h@xmath19mg _
b _ grid inversion matches the fiducial values , ez_ages can find the best - fitting mg abundance .
another index which ez_ages can use to fit a single element is c@xmath74668 , which is dominated by c and o. oxygen is notoriously difficult to estimate in old ( ie .
non - star forming ) unresolved stellar populations , therefore ez_ages does not try to model [ o / fe ] ( but see discussion in [ consist ] ) .
instead , it allows the user to choose an input oxygen abundance .
for consistency , this should be chosen to match the input isochrone : [ o / fe ] = 0.0 for the solar - scaled isochrone , [ o / fe ] = + 0.5 for the @xmath2-enhanced isochrone .
with [ o / fe ] set by the user , this leaves only c as a variable abundance which contibutes strongly to c@xmath74668 .
ez_ages adjusts [ c / fe ] so that the [ fe / h ] estimated in the h@xmath19c@xmath74668 plot matches the fiducial .
the only indices which are strongly affected by nitrogen and which might therefore allow a fit of [ n / fe ] are the two cn lines , cn@xmath20 and cn@xmath7 , both of which respond strongly to changes in c , n , and o. as discussed above , [ o / fe ] is set by the user .
this leaves c and n as variable abundances contributing to the cn indices .
however , if [ c / fe ] has already been determined by fitting c@xmath74668 , then the cn indices can be used to fit for [ n / fe ] at the calculated value of [ c / fe ] .
the computed value of [ n / fe ] is therefore dependent upon the calculated value of [ c / fe ] and a robust error analysis should propagate errors from [ c / fe ] into the the errors in [ n / fe ] .
the s07 models include one lick index dominated by ca and can therefore be used to fit [ ca / fe ] : ca4227 . because the blue continuum region used in measuring ca4227 overlaps with a cn absorption band ,
this index is also very sensitive to [ c / fe ] and [ n / fe ] , so that these abundance ratios must be fit properly before using ca4227 to determine [ ca / fe ] . as with [ n / fe ] , errors in the calculated values of [ c / fe ] and [ n / fe ] contribute to the error in [ ca / fe ] .
again , this should be taken into consideration in a complete error analysis ( see [ errors ] ) .
the set of available lick indices and their element abundance sensitivities thus prescribe an order for abundance fitting : [ c / fe ] , then [ n / fe ] , then [ ca / fe ] .
note that mg _ b _ , the preferred index used to fit [ mg / fe ] , does not depend strongly on these abundances , nor do the other abundances depend on it , therefore it can be fit in any order . based on table [ indextable ] ,
it is clear that some indices are preferable for fitting : those that are dominated by the elemental abundances that we are using them to estimate and no others . because of this , ez_ages has a built - in hierarchy which it uses to choose the indices used in abundance fitting . among the balmer lines ,
the preferred index is h@xmath19 because it is mildly sensitive to total metallicity , but relatively insensitive to individual abundances .
it is followed by h@xmath16 and h@xmath18 , which are somewhat sensitive to individual elements , and
lastly h@xmath15 and h@xmath17 , which are broader indices ( designed to measure balmer strengths in populations with a substantial a star component ) and therefore include a larger number of other absorption features in the index and pseudo - continuum bandpasses . for iron indices ,
the order of preference is @xmath23fe@xmath24 ( an average of fe5270 and fe5335 ) , fe5270 , fe5335 , fe4383 , and fe5015 .
the indices fe5270 and fe5335 are dominated by fe and total metallicity effects , making them relatively clean indicators of [ fe / h ] . of the other two lines ,
both of which are influenced by element abundance ratios , we prefer fe4383 over fe5015 for two reasons : fe5015 falls on a ccd defect in the globular cluster data used in [ compare_clusters ] to test the output of ez_ages and thus can only be tested for m 67 , where it appears to perform similarly well to fe4383 , and the fe5015 bandpass overlaps the [ oiii]@xmath295007 emission line from ionized gas sometimes present in galaxies , thus limiting its usefulness in stellar population modeling of galaxies . for fitting [ mg / fe ] , the preferred index is mg _ b _ followed by mg@xmath7 , which is more sensitive to c than is mg _
b_. for [ c / fe ] , it is c@xmath74668 followed by g4300 , for reasons discussed in detail in [ consist ] . for [ n / fe ] , it is cn@xmath20 followed by cn@xmath7 , although this is fairly arbitrary as there is no strong reason to prefer one cn index over the other . for [ ca / fe ] , the only available index is ca4227 .
it is possible to manually override this order of preference if a given index is unavailable , or to investigate the effects of using other lines in the fit .
note that ez_ages will only fit an abundance ratio if at least one of the indices needed to estimate it is available .
also , abundances which depend on other abundance ratios will not be fit if the prerequisite abundance is not available ( e.g. , ez_ages will not use a cn index to fit [
n / fe ] unless [ c / fe ] has been determined from c@xmath74668 or g4300 ) .
thus far , we have focused only on the dominant abundances affecting each index , however many indices have a weak dependence on other abundances .
particularly if the most desirable lick indices are not available due to limited wavelength coverage , bad skyline subtraction , or nebular emission within the galaxy and other lines have to be used ( such as h@xmath16 in place of h@xmath19 or fe4383 in place of fe5270 or fe5335 ) , the computed fiducial values for age and [ fe / h ] or other abundance values may change once the abundance fitting is complete , because some of the newly determined abundances may affect the indices used to determine those fiducial values . to ensure that the final abundance ratio is consistent for all of the lick indices used in the fit , ez_ages iterates the fitting process up to four times .
it uses the best - fitting abundance pattern from the previous iteration to compute new fiducial values for age and [ fe / h ] , then repeats the fitting process until further iterations do not improve the fit
. the s07 models account for variations in lick index strength due to abundance variations in the elements fe , mg , c , n , ca , o , na , si , cr , and ti .
however , only the first five of these elements can be fit using the algorithm described above .
the elements o , na , si , cr , and ti do not dominate any of the lick indices modeled by s07 and therefore can not be fit in this process .
the ez_ages code allows the user to supply a value for any of these unmodeled elements to use as input to the s07 models . in the absence of user - supplied values , ez_ages has a set of default abundances for these elements . by default ,
it sets [ o / fe ] to match the chosen isochrone ( [ o / fe ] = 0.0 for the solar - scaled isochrones , [ o / fe ] = @xmath30 for the @xmath2-enhanced isochrone ) .
the default for the @xmath2-elements na , si , and ti is that they are set to follow mg , so that [ na / fe ] = [ si / fe ] = [ ti / fe ] = [ mg / fe ] . the iron - peak element cr is set to follow fe , so that [ cr / fe ] = 0.0 .
all of the modeling reported in this analysis was done using the default settings for the unmodeled elements and with the solar - scaled isochrone . to see the effect of using the @xmath2-enhanced isochrone or super - solar [ o / fe ] on abundance fitting , see 4.3 of s07 and figures 12 and 14 in @xcite .
if errors in the lick index data are provided as input , ez_ages uses them to estimate the uncertainties in the ages , [ fe / h ] , and abundance ratios calculated by the sequential grid inversion algorithm .
uncertainties are assumed to be dominated by measurement errors in the line strengths .
systematic uncertainties in the models are ignored , except that the usual caveat applies : comparisons of _ relative _ ages and abundances between multiple objects are more reliable than absolute age or abundance estimates .
uncertainties in age and [ fe / h ] due to line strength measurement errors are computed by displacing the measured data point by the input index error , then redoing the grid inversion on the displaced data point to determine the range in age or [ fe / h ] constrained by 1@xmath31 error in the measured balmer or fe line strength . because the grid spacing is non - linear ( particularly in age ) ,
errors in the positive and negative direction are computed separately .
it has been noted by many previous authors that the errors in ssp age and [ fe / h ] as inferred from index - index plots are not independent due to the fact that the model grids are not orthogonal . as can be seen from an inspection of figure [ grids ] ,
age and [ fe / h ] errors are correlated in the sense that lower inferred ages will lead to higher inferred [ fe / h ] , and vice versa .
ez_ages does not explicitly model this aspect of the inferred errors in age and [ fe / h ] values , but users should keep this effect in mind when interpreting results from ssp index - index fitting . any error in the choice of fiducial age and [ fe / h ] will result in errors in the abundance ratios estimated by fitting to incorrect values of the age or [ fe / h ] .
thus uncertainties in the fiducial age and [ fe / h ] must be propagated through to determine their effect on the abundance ratio estimates .
for example , ez_ages assumes that the uncertainty in [ mg / fe ] is dominated by two sources : the measurement error of the mg - sensitive lick index used in the abundance ratio determination ( e.g. , mg _ b _ ) and the uncertainty in the fiducial [ fe / h ] which ez_ages attempts to match in the h@xmath19mg _ b _ plot . in the following discussion , we will use @xmath32 mg _ b _ and @xmath32[fe / h ] to represent the measurement error in mg _ b _ and the uncertainty in the fiducial [ fe / h ] , respectively .
the uncertainty in [ mg / fe ] due to mg _ b _ measurement error and fiducial uncertainty are denoted in turn by @xmath32[mg / fe]@xmath33 and @xmath32[mg / fe]@xmath34 .
to determine @xmath32[mg / fe]@xmath33 , ez_ages recomputes the [ mg / fe ] abundance for mg _ b_@xmath35 mg _ b _ and mg _
b_@xmath36 mg _ b _ , using the larger of these two uncertainties as the estimated @xmath32[mg / fe]@xmath33 . to determine @xmath32[mg / fe]@xmath34 , ez_ages recomputes [ mg / fe ] using the measured value of mg _ b _ to fit [ fe / h]@xmath35[fe / h ] and [ fe / h]@xmath36[fe / h ] , again choosing the larger uncertainty for estimating @xmath32[mg / fe]@xmath34 .
these two sources of error are treated as independent and the total @xmath32[mg / fe ] is determined by summing in quadrature @xmath32[mg / fe]@xmath33 and @xmath32[mg / fe]@xmath34 .
the process for computing @xmath32[c / fe ] is identical to that for [ mg / fe ] , substituting a c - sensitive lick index for mg _
b_. because [ n / fe ] is determined using a cn - sensitive index such as cn@xmath20 , @xmath32[n / fe ] depends not only on the cn@xmath20 line strength measurement error and @xmath32[fe / h ] , but on @xmath32[c / fe ] as well .
the contributions to @xmath32[n / fe ] due to @xmath32cn@xmath20 and @xmath32[fe / h ] ( i.e. , @xmath32[n / fe]@xmath37 and @xmath32[n / fe]@xmath34 ) are computed as above .
an additional contribution due to @xmath32[c / fe ] is determined by matching the measured cn@xmath20 line strength to the fiducial [ fe / h ] , using models with carbon abundances given by [ c / fe]@xmath35[c / fe ] and [ c / fe]@xmath36[c / fe ] .
these three sources of error are summed in quadrature to obtain @xmath32[n / fe ] .
a similar process is repeated for [ ca / fe ] , with the sole difference that errors in [ n / fe ] also contribute to the [ ca / fe ] error budget . to check that this analysis provides a reasonable estimate of the actual statistical errors , we have used a monte carlo method to generate a large set of artificial `` repeat measurements '' of the lick indices for a test stellar population and compared the resulting distribution of ages , abundances , and abundance ratios to the errors calculated by ez_ages . for our test data , we used lick index values measured in a stack of @xmath410,000 red sequence galaxy spectra taken from the sloan digital sky survey .
as modeled by _
ez_ages _ , these correspond to a stellar population with a mean light - weighted age of 9.1 gyr , with [ fe / h]@xmath38 , [ mg / fe]@xmath39 , [ c / fe]@xmath39 , [ n / fe]@xmath40 , and [ ca / fe]@xmath41 .
we ran four monte carlo simulations for different error ranges ( effectively different @xmath42 spectra ) , in which we assumed the errors were 2% , 5% , 10% , and 20% of the measured index values . for each run
, we simulated 100 repeat measurements of the data by choosing values for each lick index drawn from gaussian distributions centered on the `` true '' data values with widths characterised by the assumed 2% , 5% , 10% , and 20% errors .
this produced a simulation of 100 realizations for each of the four assumed error ranges .
_ ez_ages _ was used to determine the age , [ fe / h ] , and abundance ratios for each simulated measurement .
the distribution of these values for the 100 simulated repeat measurements can be compared with the error estimate made by ez_ages for the original data values and index measurement errors .
this comparison is shown in figure [ errplot ] .
the errors reported by ez_ages are shown on the x - axis , while the y - axis shows the standard deviation @xmath31 of the distributions of ages and abundances produced in each monte carlo simulation . as discussed earlier in this section ,
ez_ages calculates separately errors in the positive and negative direction .
there are thus two error values produced by ez_ages plotted against each monte carlo @xmath31 .
the solid line shows a one - to - one relation .
the dashed lines outline error estimates that are within @xmath43% of the 1@xmath31 monte carlo spread . for the simulations with 2% , 5% , and
10% index measurement errors , the age and abundance ratio error estimates from ez_ages are all within 30% of the results of the monte carlo simulation with the exception of the error in the h@xmath18 age , which is overestimated by about 0.3 gyr in the 2% error case .
although this is a substantial fraction of the simulated h@xmath18 age error , it is well within the uncertainties in the modeling process , which are of order 1 gyr in the @xmath44 gyr age range corresponding to the test data point @xcite . in all other cases , the ez_ages error estimate is an excellent match to the 1@xmath31 errors produced by the full monte carlo realization . for the simulations with 20% index measurement errors ,
ez_ages tends to slightly overestimate the errors as compared to the monte carlo simulation , but in all cases is still within 35% of the simulated error .
the good match shown in this test between the error estimates generated by ez_ages and those produced by a monte carlo simulation of the errors suggests that ez_ages does in fact do a reasonable job of estimating the errors in the age , [ fe / h ] , and abundance ratio values calculated .
this includes the propagation of errors between various stages of the modeling process .
as the data become noisier , ez_ages has a tendency to slightly overestimate the actual uncertainty in the age , [ fe / h ] , and abundance due to measurement errors .
figure [ errplot ] also demonstrates that in order to estimate ages , [ fe / h ] , and abundance ratios within @xmath450.1 dex for a typical early type galaxy , lick indices must be measured with @xmath46% accuracy , which requires spectra with @xmath47 @xmath48 @xcite .
the monte carlo error simulation also allows us to investigate the effect of correlated errors in the ez_ages analysis .
because lines of constant age and [ fe / h ] are not orthogonal in h@xmath19-@xmath23fe@xmath24 space , errors in either index measurement will result in anti - correlated errors in the inferred age and [ fe / h ] , as discussed in detail in @xcite .
for example , an error in h@xmath19 that causes the determined ssp age to be too old will cause the measured [ fe / h ] value to be too low , as can be seen from the grids in figure [ grids ] .
likewise , the sequential grid inversion process for fitting abundance ratios depends on the fiducial age and particularly on the fiducial [ fe / h ] determination .
this will tend to produce an anti - correlation between [ fe / h ] and the measured abundance ratios . as shown in figure [ errplot ] , the simplified error propagation implemented in ez_ages does an excellent job of tracking the total error in each stellar population parameter throughout the fitting process .
however , it does not produce a full error ellipse illustrating correlated errors .
figure [ corr_errs ] uses the results of the monte carlo error simulation to illustrate the correlated errors that result from sequential abundance fitting .
the 100 simulated data realizations for each of the assumed 2% , 5% , 10% , and 20% index measurement errors ( yellow , red , blue , and green points , respectively ) effectively trace out the error ellipses in stellar population parameter space .
figure [ corr_errs]a shows the well - known error anti - correlation between the age and [ fe / h ] measurements , as expected from the non - orthogonality of the age and [ fe / h ] model grids in h@xmath19-@xmath23fe@xmath24 space .
panels b e show the abundance ratio determinations for [ mg / fe ] , [ c / fe ] , [ n / fe ] , and [ ca / fe ] against [ fe / h ] .
the error ellipses for these data are expected to show an anti - correlation of errors because an over - estimate of [ fe / h ] will tend to result in an under - estimate of the other abundances .
for example , if an error in the @xmath23fe@xmath24 index causes the fiducial [ fe / h ] determination to be too high , the measured abundance ratios for other elements will tend to be too low ( at a higher [ fe / h ] , a lower value of [ mg / fe ] is needed to match the same mg @xmath27 index ) .
this anti - correlation is observed between [ fe / h ] and the abundances [ mg / fe ] , [ c / fe ] , and [ ca / fe ] .
interestingly , there appears to be no correlation in the errors in [ fe / h ] and [ n / fe ] . because [ n / fe ] is fit using a cn index , it depends strongly on the value of [ c / fe ] previously determined by fitting either c@xmath74668 or g4300 and thus we expect a anti - correlation of errors between [ c / fe ] and [ n / fe ] .
this is indeed the case , as illustrated in figure [ corr_errs]f , and may explain the lack of correlation in the errors in [ n / fe ] and [ fe / h ] .
as in the case of [ n / fe ] , the fitted values of [ ca / fe ] depend on other abundance fits and thus may produce correlated errors .
these are illustrated in figure [ corr_errs]g , h .
the effect of correlated errors can produce spurious trends in a dataset , thus the correlations illustrated here should be kept in mind when interpreting abundance patterns from ez_ages or a similar abundance fitting process .
of course , the best guard against these effects is to perform abundance fitting only on spectra with @xmath49 @xmath48 , corresponding to 5% index errors ( red points in figure [ corr_errs ] ) or better .
the validation of stellar population synthesis models depends crucially on their ability to match the integrated properties of simple systems , such as stellar clusters , which can be reasonably approximated by single stellar populations .
in fact , this simple truism disguises the considerable complexity of the task , as attested by the vast literature dedicated to using cluster observations to constrain stellar population synthesis models ( s07 ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; and references therein ) .
s07 presented a lengthy discussion of the comparison of his models with integrated data for well - known galactic open and globular clusters .
models were computed for a range of ages and metallicities , adopting the known abundance patterns of m 5 , 47 tuc , ngc 6528 , and m 67 .
these were then compared with lick index measurements taken in the ( very high s / n ) cluster spectra . as a result
, it was found that predictions for balmer and metal lines agreed with the data to within 1-@xmath31 index measurement errors .
also very importantly , the best - matching ages differed by no more than @xmath41 gyr from the value obtained from analysis of cluster color - magnitude diagrams ( cmds ) .
the only exception was m 5 , for which the age found was too low , due to the presence of blue horizontal branch ( hb ) stars , which are not accounted for by the models ( see [ consist ] for a further discussion of this point ) . in this section
we ask the inverse question : how well do the models fare when used to convert a given set of index measurements into ages and elemental abundances ?
again , we rely on data for clusters whose abundances and ages are well constrained from analysis of individual stellar spectra and deep cluster cmds .
two important model checks are performed .
we start by submitting the models to the most challenging of the two tests ( [ real ] ) , which consists of checking whether the results obtained with ez_ages match those from detailed resolved analysis of high - resolution spectra and deep photometry of stellar members . while our _ goal _ is to match ages and the abundances of fe
, mg , c , n , and ca to within 0.1 dex our _ requirement _ is that differences be smaller than 0.2 dex .
moreover , this set of specifications should be met in a quick and non - interactive fashion , suitable for applications to large data sets through use of ez_ages . as a second check ( [ consist ] ) , we verify the internal consistency of ez_ages , by comparing the results obtained for a given set of observations , when using different combinations of lick indices as age and metal - abundance indicators .
the clusters of choice for this exercise are selected on the basis of availability of good quality age and metal abundance determinations ( table [ abuntable ] ) .
they span a wide range of metallicities , from [ fe / h ] @xmath4 1.2 ( ngc 6121 ) to nearly solar metallicity ( m 67 and ngc 6528 ) , and include both old and intermediate - age ( m 67 ) clusters .
lick indices for these clusters are presented in table [ cluster_indices ] . in figure [ cluster_grids ]
cluster data are compared with the s07 solar abundance pattern models in a few index - index plots to illustrate that solar - scaled models _ can not _ match the full set of observed cluster indices .
the indices displayed are the default age and metal - abundance indicators employed by ez_ages .
s07 showed that h@xmath19 , @xmath23fe@xmath24 , mg @xmath27 , c@xmath74668 , cn@xmath20 ( or cn@xmath7 ) , and ca4227 are the most reliable , recommended indices for age , [ fe / h ] , [ mg / fe ] , [ c / fe ] , [ n / fe ] , and [ ca / fe ] determinations .
this will be explored further in
[ consist ] .
henceforth , we will refer to this set of indices as the _
standard set_. the models displayed in figure [ cluster_grids ] were computed for the solar abundance pattern , and the data were measured in the integrated spectra of the above globular clusters from the @xcite spectral library and that of m 67 , obtained by @xcite , and revised by s07 . in the upper left panel , model and data are compared in the @xmath23fe@xmath24 vs. h@xmath19 plane .
s07 showed that these indices are excellent indicators of [ fe / h ] and age , respectively .
while the @xmath23fe@xmath24 average index is relatively insensitive to the abundances of elements other than fe , h@xmath19 is mostly sensitive to age , being only mildly sensitive to the overall metallicity of the stellar population .
therefore , the age and iron abundance of each cluster can be read fairly accurately from the position of the cluster data on the model grid except for the case of ngc 6121 where the presence of blue hb stars affects the h@xmath19-based age ( see discussion in [ consist ] ) .
the remaining panels show plots of other metal - line indices against h@xmath19 , and @xmath23fe@xmath24 against h@xmath16 . in the metal - line vs.h@xmath19 plots
, one can see that the positions on the model grids where the data for most clusters fall are different from those they occupy on the @xmath23fe@xmath24 vs. h@xmath19 plot . as discussed in
[ method ] , this is a clear indication that the abundance pattern of the cluster stars is different from that of the sun .
we apply the method presented in the previous section to determine the abundance pattern of the clusters .
we ran ez_ages on the cluster data presented in table [ cluster_indices ] .
the results are listed in table [ abuntable ] , together with abundance and age determinations from the literature .
the literature values are based on abundance analyses of high - resolution spectra of individual cluster stars and on deep cluster cmds , respectively .
differences between our results and those from the literature are displayed in figure [ comp_lit ] . in that plot ,
the dashed lines indicate the @xmath0 dex goal stated in the ez_ages specifications discussed in [ compare_clusters ] .
average values from the literature are displayed in black and individual determinations are shown in gray . for carbon and nitrogen ,
the values displayed are averages between cn - strong and cn - weak stars , as this best approximates the effect of unresolved spectroscopic observations .
a first inspection of the results shows that most abundance points fall between the two dashed lines , indicating that we are meeting the 0.1-dex goal for most of the abundances and for most clusters .
agreement is in fact particularly good for the abundances of carbon and nitrogen , where the 0.1-dex goal was reached for all clusters for which these elemental abundances are available in the literature . for magnesium and iron
, we meet the 0.1-dex goal for some clusters , and the 0.2-dex requirement for all of them , with one exception : for ngc 6528 , we find [ fe / h ] = @xmath50 , in reasonably good agreement with the value [ fe / h ] = @xmath51 reported by @xcite and @xcite but in significant disagreement with the value [ fe / h ] = @xmath52 reported by @xcite . for calcium
, there is one cluster only for which the 0.2-dex requirement is not met .
for all the others , we meet the 0.1-dex goal .
given that this is the first systematic attempt at deriving abundances for all these five elements from integrated light , these results can be considered very satisfactory .
below we discuss the few exceptions for which agreement is not so good .
although the values of [ fe / h ] determined by ez_ages are almost all within the 0.2-dex requirement , they are consistently slightly lower than the literature abundances determined from invidual stars .
the zeropoint for [ fe / h ] in the models is based upon the [ fe / h ] measurements for the stellar library from which the s07 model is constructed ( the s07 models have not been separately calibrated to the solar spectrum ) and the zeropoint is not known to perfect accuracy .
a systematic 0.1-dex underprediction of [ fe / h ] by the models is possible .
however , this is not necessarily more uncertain than the zeropoints in the cluster abundance determinations from individual stars . in the case of ngc 6528 , for which multiple literature sources are available , the abundance determinations in the literature show substantial scatter : 0.25 dex in [ fe / h ] , 0.07 dex in [ mg / fe ] , and 0.63 in [ ca / fe ] . in this context , it is difficult to address zeropoint issues on the order of 0.1 dex by comparison with existing cluster data .
in general we note that , for some clusters , elemental abundances determined by different groups are in substantial disagreement , which makes the task of interpreting our results a little more complicated .
that is especially the case of ngc 6528 mentioned above , for which iron abundances from two groups differ by @xmath4 0.25 dex , and calcium abundances differ by more than 0.6 dex ! while a definitive statement on our ability to match that cluster s abundances with ez_ages awaits the solution of this discrepancy , we call attention to the fact that agreement with the average values is very good . ez_ages also performs very well in age determinations . for all clusters , with the sole exception of ngc 6121 , the 0.2-dex requirement is met , and for 47 tuc and m 67 , the 0.1-dex goal is also achieved . for ngc 6121 , the age obtained is too young because of the contribution by blue hb stars which tend to boost the strength of balmer lines in the spectra of old stellar populations , mimicking a younger age ( see [ consist ] ) .
the effect of blue hb stars on balmer lines has been pointed out in previous studies ( @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) and it can potentially lead to age underestimates . the cluster ngc 6441 also contains blue hb stars @xcite , but ez_ages determines a relatively old age for the cluster when using the h@xmath19 index in the modelling process , which seems to contradict this interpretation .
unfortunately , we could not find a cmd - based age for this cluster in the literature so we can not make a statement about how the age result from ez_ages compares to an independent age estimate .
however , when the bluer balmer line h@xmath16 is used in the modelling of ngc 6441 , ez_ages finds a younger age than that determined from h@xmath19 , suggesting that the influence of the blue hb stars does show up in h@xmath16 . as discussed by @xcite ,
the signature of blue hb stars is stronger at bluer wavelengths , so that they tend to impact age determinations based on h@xmath53 more strongly than those based on h@xmath19 .
this can be clearly seen in the bottom right panel of figure [ cluster_grids ] , where both ngc 6121 and ngc 6441 look _ younger _ in the @xmath23fe@xmath24 vs. h@xmath16 plot than in that involving h@xmath19 . in ngc 6441 ,
the blue hb stars are greatly outnumbered by red hb stars , unlike ngc 6121 in which the blue hb population is substantial .
they therefore have only a modest influence at the wavelength of h@xmath19 , but a larger influence at the bluer wavelength of h@xmath16 , where the blue hb stars contribute a greater fraction of the total light @xcite .
we further discuss this point in [ consist ] .
finally , we note that the values we found for the abundances of c , n , and ca for ngc 6121 are suspiciously low .
unfortunately , we could not find literature values for c and n abundances to compare with our estimates .
however , the numbers table [ abuntable ] and figure [ comp_lit ] show that our [ ca / fe ] estimate is too low by about 0.3 dex , which confirms our suspicions that there might be a problem with our procedure for [ fe / h ] @xmath54 .
while this clearly deserves further investigation , for the time being _ we strongly discourage users from applying ez_ages in the [ fe / h ] @xmath54 regime_. another check on the accuracy of the stellar population modeling is to compare the index predictions of the best - fitting model to the observed line strengths .
figure [ model_ind ] shows the observed values of the lick indices as measured in the five test cluster spectra , plotted against the predicted index line strengths from the best - fitting s07 model , as determined by ez_ages .
black and gray symbols show indices that are and are not included in the fitting process , respectively .
the indices used in the fitting process are extremely well reproduced by the best - fitting model .
in addition , the indices that are not used in the fitting process are also reasonably well matched by the model predictions , with some scatter but with no indication of systematic problems in the modeling of individual indices . in this test ,
we verify how the results vary when different index sets are adopted as inputs to ez_ages .
this is important , because the full set of indices considered by ez_ages are not always available to observers , due to limitations such as those determined by the instrumental setup adopted , or by the redshift of the sample studied .
therefore , it is important to make sure that age and metal abundances obtained from ez_ages do not depend on which absorption - line indices are employed or , if they do , that the systematics is understood and can be accounted for . to test for this , we ran ez_ages on multiple combinations of line indices as measured for the clusters ngc 6441 and m 67 , and inter - compared the ages and metal abundances obtained .
as a basis for the comparisons , we adopt the results obtained when using the standard set , i.e. , h@xmath19 , @xmath23fe@xmath24 , mg @xmath27 , c@xmath74668 , cn@xmath20 , and ca4227 . the indices used in each separate model fit are listed in table [ diff_ind_tab ] .
we substitute each alternative index separately to asses the effect due to that individual index .
we also examine some combinations of indices , specifically a `` balmer '' model , in which we fit using the average of h@xmath19 , h@xmath18 , and h@xmath16 , an `` all '' model , in which all lines are included in the fit , and a `` high - z '' model , which simulates the effect of applying ez_ages to higher redshift data , where only the bluer indices ( @xmath55 ) may be available because the indices at longer wavelengths are redshift out of the optical spectrum . results from the fitting process are provided in table [ diff_fit_tab ] and displayed in figures [ vary_fit_ngc6441 ] and [ vary_fit_m67 ] , where residuals relative to the age and metal abundances obtained with the standard index set are plotted for the various index combinations .
looking first at ngc 6441 , figure [ vary_fit_ngc6441 ] shows that , for the vast majority of the index combinations , ages and metal abundances vary by less than @xmath0 dex , indicating that ez_ages and the s07 models have attained an outstanding degree of consistency .
there are a few exceptions , though , which occur for the following index combinations : 1 ) when g4300 is used as a carbon abundance indicator in place of c@xmath74668 ( panel c , triangles ) ; 2 ) when higher order balmer lines are used as age indicators ( panel a ) ; 3 ) when higher order balmer lines are used as age indicators and fe4383 is used as the iron - abundance indicator as in the `` high - z '' model ( panel d , triangles ) or when the higher order balmer lines and fe4383 are averaged in with the other lines as in the `` all '' model ( panel d , squares ) . in case 1 , the g4300-based [ c / fe ] and [ n / fe ] values differ from those obtained from the standard indices by 0.14 and + 0.18 , respectively . in case 2 , ages tend to be younger when higher - oder balmer lines are used in place of h@xmath19 , by up to @xmath4 0.2 dex . finally , case 3 is similar to case 2 , except that [ fe / h ] is too high , and consequently all the other abundances are too low .
the difference found in case 1 when replacing c@xmath74668 by the g4300 index is due to the different sensitivities of the two indices to the ratio between the abundances of carbon and oxygen ( c / o ) .
the two indices are sensitive to variations of the oxygen abundance because of details of the molecular dissociation equilibrium in the atmospheres of cool stars . of all molecules in whose formation carbon and oxygen take part , carbon monoxide ( co ) is the hardest one to break , because it has the highest dissociation energy .
therefore , at the temperatures prevalent in the atmospheres of g and k stars , most available free carbon and oxygen atoms are locked in co. as a result , variations in the abundance of oxygen , usually the most abundant of the two species , have a strong influence on the amount of carbon that is free to form other molecules , such as c@xmath7 and ch , which are responsible for the vibrational bands measured by the c@xmath74668 and g4300 indices , respectively .
therefore , an increase in the abundance of oxygen tends to provoke a decrease in the strength of those molecular bands . while the concentration of the c@xmath7 molecule depends quadratically on the abundance of carbon , that of ch depends only linearly on that abundance . as a result
, c@xmath74668 is far more sensitive to carbon than to oxygen , while the the sensitivity of the g4300 index to those two abundances is approximately the same ( though for both indices , the sensitivity to carbon and oxygen abundances have opposite signs ) .
although this makes the g4300 index a more uncertain indicator of carbon abundance , it raises the possibility that a combination of the two indices may be used to constrain the abundance of oxygen .
we verified that possibility by raising the input oxygen abundance in the models from [ o / fe ] = 0 to + 0.3 . as a result
, the carbon and nitrogen abundances inferred from use of c@xmath74668 and g4300 indices agreed to within 0.05 dex .
furthermore , the ages and abundances inferred for the o - enhanced models were not substantially different from the results using the [ o / fe ] = 0 model : [ n / fe ] increased by @xmath56 dex , while the ages inferred from h@xmath18 and h@xmath16 increased by 0.6 gyr .
the values of h@xmath19 age , [ fe / h ] , [ mg / fe ] , and [ ca / fe ] showed negligible change ( @xmath57 gyr or @xmath58 dex ) .
interestingly , the oxygen abundance in ngc 6441 stars ranges from [ o / fe ] = 0.05 to 0.34 @xcite , with stars in the upper end of the interval having `` normal '' oxygen abundances , and likely being more numerous and dominating the cluster light .
therefore , the value we obtained from the cluster s integrated light is in good agreement with the known cluster abundance , which makes it possible that oxygen abundances might be inferred from a combination of the c@xmath74668 and g4300 indices .
this will be further investigated in a forthcoming paper .
the dependence in case 2 of the resulting age on the balmer line adopted is not surprising , given that ngc 6441 is characterized by the presence of blue hb stars ( @xcite , @xcite ) .
because these stars have early - f and a spectral types , they tend to strengthen balmer lines in the integrated spectra .
that effect is obviously more important at bluer wavelengths , where the contribution by early - f and a stars to the integrated light is the greatest .
therefore , higher - order balmer lines tend to be more strongly affected . as a result , when models that do not take into account blue hb stars are compared with the data , they tend to infer systematically younger ages , and more so according to higher order balmer lines ( see discussion in s07 ) .
@xcite have in fact shown that this effect can be used to constrain the morphology of the hb of globular clusters , solely on the basis of balmer line strengths in their integrated spectra . in particular , the ratio between h@xmath16 and h@xmath19 was shown to be very sensitive to the presence of blue hb stars . before discussing case 3 , where fe4383 and h@xmath16
are used in the fitting process , let us look at the results of the consistency test for m 67 . here
again there is agreement to within @xmath0 dex for any combination of indices used , with the exception of combining fe4383 and h@xmath16 .
the three balmer lines all give consistent results for m 67 ( panel a ) , because m 67 does not contain blue hb stars and because the integrated spectrum of m 67 used in the this analysis has been carefully constructed to exclude blue stragglers @xcite .
unfortunately , there is a problem ( as yet not well - understood ) with the g4300 measurements for this cluster @xcite , thus we can not use m 67 to test our hypothesis about the effect of oxygen abundance on g4300 and c@xmath74668 . indeed , the only inconsistency in the abundance determinations for m 67 is in the `` high - z '' model , in which fe4383 and h@xmath16 are used in the fitting process ( case 3 from above ) . here , as with ngc 6441
, we not only derive younger ages , but the resulting iron abundance is higher by @xmath59 dex .
consequently , all the other elemental abundances are found to be somewhat low . for m 67
, we can not invoke blue hb stars to explain this discrepancy .
interestingly , substituting h@xmath16 or fe4383 into the standard index set _ individually _ ( triangles in panel a and squares in panel b ) does not result in large discrepancies from the standard model .
how then can we understand the dramatic differences seen when both h@xmath16 and fe4383 are used in the modelling process ?
figure [ red_blue_grids ] shows solar abundance model grids from the s07 model .
the m 67 data are overplotted as diamonds . in panel
a , we see that h@xmath19 and @xmath23fe@xmath24 produce model grids where lines of constant age and constant [ fe / h ] are nearly perpendicular .
this is one of the reasons this index combination is so useful for age and [ fe / h ] determinations .
when fe4383 is substituted for @xmath23fe@xmath24 ( panel b ) , the lines of constant [ fe / h ] are less nearly vertical , lessening the diagnostic power of the index - index grid .
likewise , when h@xmath16 is substituted for h@xmath19 ( panel c ) , the lines of constant age are less nearly horizontal , again constraining the age and [ fe / h ] determinations .
this is because , in the blue region , line crowding is higher , making h@xmath16 more sensitive to metallicity .
moreover , the blue spectral region is more strongly affected by warm stars than the red , making fe4383 more sensitive to age than its redder counterparts .
notice , however , that there is still enough spread in the grids in panels b and c that small errors or zero - point uncertainties in the data or the models do not result in substatially different age and [ fe / h ] measurements for m 67 . now consider panel d. when h@xmath16 and fe4383 are combined in the fitting process , the model grid lines are very far from perpendicular and the model space collapses down toward the familiar age - metallicity degeneracy . here ,
small errors or zero - point uncertainties result in substantially different age and [ fe / h ] measurements .
even the slight metallicity difference seen between panels a and b becomes a substantial discrepancy in panel d. this raises warning flags for using these indices to determine age and [ fe / h ] .
in particular , we note that fe4383 appears to _ always _ give slightly higher values of [ fe / h ] in the data presented both here and in ongoing work by the authors , suggesting that there is some inaccuracy in the modelling of this index .
the fact that the discrepancy appears in the case of m 67 , whose abundances are known to be solar - scale , suggests that this may be a zeropoint problem in the models , rather than a problem in calculating the abundance sensitivities of this index .
combined with the increased degeneracy of the h@xmath16-fe4383 index - index space , this means that ez_ages should be used cautiously when determining stellar population parameters in higher redshift objects where h@xmath19 is not available , and that the need for high @xmath42 data is particularly acute in this regime .
even when the problem with fe4383 is solved , the degeneracy of h@xmath16-fe4383 index - index space means that small errors or zeropoint uncertainties in data can result in errors in the determination of age and [ fe / h ] from these indices .
this likely applies to all stellar population models , not just those presented in this work and s07 . in summary
, we conclude that ez_ages is characterized by a remarkable degree of consistency , in that ages and metal abundances inferred from different index sets agree in the vast majority of cases to within @xmath0 dex .
two of the inconsistencies we found , namely , that different balmer lines indicate different ages , and different carbon - abundance indicators indicate different carbon abundances , can in fact be potentially explored to extract even more information from the integrated spectra of stellar populations . while blue balmer lines were shown to be useful to indicate the presence of blue hb stars in globular clusters @xcite , the carbon index discrepancy can potentially be used to constrain the abundance of oxygen .
the latter effect will be further investigated in a forthcoming paper .
the third inconsistency is due to the fact that the combined h@xmath16-fe4383 index - index space is substantially more degenerate than h@xmath19-@xmath23fe@xmath24 space , making accurate age and [ fe / h ] determinations from these indices difficult .
this is a matter of concern in integrated studies of globular clusters @xcite and higher redshift galaxies @xcite , for which only these bluer lines may be available .
having performed the reality and consistency checks above , we now turn to comparisons between ez_ages results and those coming from application of other models in the literature .
we focus on the models by ( * ? ? ?
* hereafter tmb03 ) , which also make predictions for variable abundance patterns . the tmb03 models also attempt to match non - solar abundance patterns by fitting lick indices in unresolved stellar population spectra .
these include both solar - scaled and @xmath2-enhanced models , in which the set of @xmath2 elements n , o , mg , ca , na , ne , s , si , and ti are slightly enhanced , while the iron peak elements cr , mn , fe , co , ni , cu , and zn are significantly depressed , in order to vary [ @xmath2/fe ] at fixed total metallicity [ z / h ] . in these models , [ c / fe ]
is fixed at solar .
* hereafter t05 ) use the tmb03 models to estimate ages , [ z / h ] , and [ @xmath2/fe ] for a sample of 124 early - type galaxies , based on measurement of the lick indices h@xmath19 , @xmath23fe@xmath24 , and mg @xmath27 . to compare the stellar population modeling results of using ez_ages and the s07 models to those using tmb03 , we have run ez_ages on the sample of 124 galaxies presented in t05 , using the values of h@xmath19 , @xmath23fe@xmath24 , and mg @xmath27 given in their table 2 . with these indices
, ez_ages can determine the ssp age , [ fe / h ] , and [ mg / fe ] values of the sample galaxies and compare these results with those of the tmb03 models used in t05 .
although the details of the model fitting procedure differ , t05 perform a very similar analysis to that of ez_ages , determining age , [ z / h ] , and [ @xmath2/fe ] from the same lick indices that are the standard set ez_ages uses to calculate age , [ fe / h ] , and [ mg / fe ] .
the models used in t05 use the same @xcite index sensitivity functions as s07 and should therefore show the same index strength variations when the abundances are modified .
one practical difference is that the tmb03 models are cast in terms of [ z / h ] , rather than [ fe / h ] . in order to compare results from the two models , we convert t05 s estimated [ z / h ] into [ fe / h ] using the conversion proposed by tmb03 : [ fe / h ] = [ z / h ] @xmath60 [ @xmath2/fe ] ( tmb03 ) .
the tmb03 models assume that all @xmath2-elements track mg .
ez_ages follows a similar convention , by default setting na , si , and ti to track mg in the modeling process , while cr follows fe .
a significant advantage of ez_ages and the s07 models is that they also track the [ c / fe ] , [ n / fe ] , and [ ca / fe ] abundance ratios using c , n , and ca - sensitive lick indices , rather than making assumptions about how these elements vary .
however , these differences should not affect the comparison of age , [ fe / h ] , and [ @xmath2/fe ] results presented here , as h@xmath19 , @xmath23fe@xmath24 , and mg @xmath27 are predominantly sensitive to age , fe and mg only ( see table [ indextable ] ) , and are thus treated the same by both models through the @xcite sensitivity functions .
the results of fitting the t05 galaxies using ez_ages are shown in figure [ compare_thomas ] . in each panel ,
the dashed line shows a one - to - one relation and the solid line shows the best fitting line to the data while keeping the one - to - one slope fixed .
the resulting zeropoint offsets between the model fits are indicated in the upper left corner of each panel . in general , there is good agreement between the results coming from application of the two models to the same data set .
the scatter is small and the zeropoint offsets are modest .
the values of [ @xmath2/fe ] agree to within a very small offset .
the ages estimated in t05 are typically @xmath61% ( 0.13 dex ) younger than those we find with ez_ages , while the t05 values of [ fe / h ] are @xmath62 dex higher .
in addition to the zeropoint offsets in the age and [ fe / h ] comparisons , the two different analyses do not truly follow a one - to - one slope for either parameter : the [ fe / h ] relation is significantly steeper than the unity relation , while the age comparison shows some curvature . at old ages ( @xmath63 age @xmath64 as estimated by ez_ages )
the slope is very nearly one - to - one , and the agreement between the two age estimates is much better , with a zeropoint offset of 15% ( @xmath65 dex ) between the t05 ages and those from ez_ages . at intermediate ages ( @xmath66 as estimated by ez_ages ) the relation is steeper than unity , while at young ages ( @xmath67 ) the relation flattens out again , although at a substantial zeropoint offset . at intermediate and young ages ,
the t05 ages are 33% ( 0.17 dex in log age ) younger .
the origin of the curvature in the age relation is unclear . in any case ,
age agreement for old galaxies is very good ( within 15% ) and is still within 33% for all galaxies down to the youngest ssp ages estimated in the t05 sample .
based on the cluster comparison of [ real ] and figure [ comp_lit ] , the zeropoint uncertainties in the s07 models are roughly 0.07 dex in @xmath63 age , 0.1 dex in [ fe / h ] , and 0.05 dex in [ mg / fe ] .
the differences between the t05 and the ez_ages results are therefore within the zeropoint uncertainties of the models , with the exception of the age estimates for galaxies younger than 10 gyr . for these galaxies ,
the t05 results are younger by more than twice the indicated zeropoint uncertainties in the s07 and ez_ages models . the best way to resolve
this discrepancy is to compare the models with data for metal - rich , intermediate - age clusters .
while the models by s07 have been shown in the previous section and in schiavon ( 2007 ) to match the data for m 67 ( @xmath4 3.5 gyr - old and solar metallicity ) , more data are clearly needed to better constrain the models in this crucial age / metallicity regime . the non - unity slope of the [ fe / h ] comparison also deserves further investigation .
while the differences are negligibly small in the solar metallicity regime , they climb up to 0.2 dex at the high - metallicity end . for the most metal - rich galaxies , ez_ages obtains [ fe / h ] @xmath4 0.2 , and t05 find galaxies to be more metal - rich by @xmath4 0.2 dex . as in the case of age determinations , we again find ourselves lacking data that could help decide between the two model sets , given the non - existence of integrated spectroscopy for super - metal - rich clusters .
it would be interesting , however , to compare results from the two models by confronting them with cluster data in a regime for which such data are available , such as moderately metal - poor clusters . in figure [ thomas_slope ] , we plot the difference between the values of [ fe / h ] estimated in t05 and by ez_ages as a function of various other properties of the stellar populations .
the top panel shows the [ fe / h ] differences as a function of ssp age and @xmath2-enhancement .
the differences in the [ fe / h ] estimates do not appear to be strongly correlated with either ssp age or @xmath2-enhancement .
in the lower panel , we show the differences in [ fe / h ] as a function of the differences in the estimated ages and @xmath2-enhancements .
the [ fe / h ] differences are correlated with differences in the age estimates between the two different models .
furthermore , the correlation is in the direction expected from the correlated errors in the index - index plots : where t05 obtain substantially younger ages than ez_ages , they also find higher [ fe / h ] .
the difference in the age estimates by the different models therefore naturally explains the observed differences in [ fe / h ] , both the zeropoint shift and the non - unity slope of the [ fe / h ] comparison .
estimates of [ @xmath2/fe ] are substantially less biased by a zeropoint error in the age .
that is because the mg @xmath27 and @xmath23fe@xmath24 indices have very similar age - dependence , so that the zero - point uncertainties in the parameters determined by these two indices ( [ mg / h ] and [ fe / h ] ) cancel out . from this analysis ,
it is clear that the tmb03 models and ez_ages give fairly consistent results , with a relatively small disagreement in the age zeropoints of the models which produces modest disagreements in [ fe / h ] .
abundance ratios such as [ @xmath2/fe ] are much less sensitive to zeropoint offsets .
the comparison of ez_ages results with ages from cmd fitting of galactic clusters ( [ compare_clusters ] ) indicates that the age zeropoint in ez_ages is correct to within 0.07 dex in @xmath63 age .
the [ z / h ] and [ @xmath2/fe ] results from tmb03 models have been compared with cluster data @xcite , but age estimates have not been independently tested , therefore it is difficult to say what the zeropoint uncertainties in the age determinations for these models should be . we therefore conclude that ez_ages and the s07 models give results in good agreement with the tmb03 models ( at least in the high - metallicity regime tested in this analysis of galaxy spectra ) , but that at intermediate ages ( age @xmath68 gyr ) the tmb03 ssp ages may underestimate the true ssp ages by as much as 33% ( 0.17 dex in @xmath63 age ) .
a more general result from this comparison is that the sequential grid inversion algorithm presented in this paper , combined with the ssp models of s07 , does an excellent job of reproducing [ mg / fe ] abundances obtained by other groups with other models .
although the ez_ages abundance results for [ c / fe ] , [ n / fe ] , and [ ca / fe ] in unresolved stellar populations can not be compared to other models ( since no other models exist for the comparison ) , the modeling process for fitting these other abundances is the same as that used to fit [ mg / fe ] and the success of this process for [ mg / fe ] bodes well for other abundances .
the test presented in this section , combined with the stringent tests with cluster data described in [ compare_clusters ] , are reassuring evidence that this modeling process is reasonable and can be used to provide a _ quantitative _ assessment of multiple elemental abundances .
in this paper , we have presented a methodology for measuring ssp ages , [ fe / h ] , and individual abundance ratios [ mg / fe ] , [ c / fe ] , [ n / fe ] , and [ ca / fe ] for unresolved stellar populations .
we do this by exploiting the different sensitivities of various lick indices to a variety of elemental abundances , and the ability of the s07 stellar population models to accurately model the lick indices for a wide range of input abundance patterns .
the algorithm presented here has been implemented in the idl code package ez_ages , which is available for download and general use .
ez_ages age estimates reproduce the results of cluster cmd fitting to within 0.15 dex for all clusters , with the exception of ngc 6121 , whose balmer line strengths are affected by a blue hb . for two out of three of the remaining clusters ,
the ez_ages age is within 0.1 dex of the cmd result . _ [ fe / h ] and [ mg / fe ] :
_ ez_ages estimates match the results from high - resolution spectroscopy of individual cluster members to within 0.1 dex for some of the clusters , and to within 0.2 dex for all clusters ( with the exception of the disagreement between our value of [ fe / h ] for ngc 6528 and the results of @xcite we are in @xmath69 dex agreement with the results of @xcite and @xcite for [ fe / h ] for this cluster ) .
_ [ c / fe ] and [ n / fe ] : _ for the only two clusters where c and n abundances are available ( 47 tuc and m 67 ) , ez_ages results match those from high - resolution spectroscopy of individual cluster members to within 0.1 dex . for ngc 6121 , we find suspiciously low values , that might indicate a problem at the low metallicity end of our models .
therefore , we strongly caution users against using ez_ages to determine these abundances for systems with [ fe / h ] @xmath54 . _
[ ca / fe ] : _ ez_ages results are within @xmath70 dex for all clusters except ngc 6121 . the latter is certainly related to the problem mentioned above regarding [ c / fe ] and [ n / fe ] in the metal - poor regime , so we do not recommend use of ez_ages for [ ca / fe ] determinations for systems with [ fe / h ] @xmath54 . _ consistency test performed on the cluster ngc 6441 : _ using different combinations of lick indices in the fitting process yields results that are consistent to within @xmath71 dex for all stellar population parameters except for the following cases : using g4300 instead of c@xmath74668 to fit [ c / fe ] results in values of [ c / fe ] and [ n / fe ] that differ by @xmath72 and @xmath73 dex , respectively .
this difference is probably due to differences in the index responses to o , which are not explored in this work . using h@xmath18 or h@xmath16 instead of h@xmath19 as an age indicator results in younger ages for the cluster ngc 6441 due to the effect of the cluster s blue hb .
this discrepancy does not exist for m 67 , which does not have a blue hb , and whose integrated spectrum is constructed to exclude blue hb stars . using the bluest balmer and fe lines available , h@xmath16 and fe4383 , can result in age and [ fe / h ] measurements that are substantially different than those determined using the standard set of indices , due to the increased degeneracy of the model space for these indices .
this poses a problem for high redshift studies of stellar populations , where often only the bluest lines are available for analysis .
_ error analysis test : _
ez_ages implements an algorithm that attempts to propagate errors through the modeling process in an efficient way that takes into account the dependence of abundance fitting on other parts of the population fitting process .
monte carlo simulations show that this simplified error estimation does an excellent job of matching the true errors in the population parameters due to measurement errors in the lick indices .
age and [ fe / h ] comparisons show little scatter .
there are zeropoint offsets of 15%33% ( 0.070.17 dex ) in age and 0.08 dex in [ fe / h ] between the two analyses , and a non - unity slope in the [ fe / h ] comparison .
the age differences are small for the oldest ( age @xmath74 gyr ) galaxies and increase for younger galaxies .
the [ fe / h ] effects seem to be a natural result of the age differences .
the s07 and ez_ages age zeropoints are shown here to be good to within 0.07 dex , while the tmb03 age zeropoints are untested .
this suggests that the ages derived in t05 may be too young by up to 33% ( 0.17 dex ) for intermediate age galaxies ( age @xmath68 gyr ) .
the sequential grid inversion algorithm method is proven to work for estimating [ mg / fe ] . since this method is the same as the one used to determine other abundance ratios , those estimates are likely to also be reliable , as demonstrated in the comparison with cluster data .
overall , ez_ages and the s07 ssp models do an excellent job of fitting the ssp age , [ fe / h ] , and abundance ratios [ mg / fe ] , [ c / fe ] , [ n / fe ] , and [ ca / fe ] for unresolved stellar populations .
the small zeropoint uncertainties in age and [ fe / h ] estimates illustrated by comparisons with galactic cluster data demonstrate that absolute estimates can be made for these quantities with high @xmath42 data .
absolute estimates of elemental abundances are robust to the ( small ) zeropoint uncertainties .
ez_ages and the s07 models therefore make it possible , for the first time , to perform a quantitative assessment of multiple individual elemental abundances from medium - resolution spectra of unresolved stellar populations .
furthermore , the abundance fitting process can be run in an automated way on large data sets . with these innovations ,
stellar population analysis is better positioned than ever before to address the task of unravelling the star formation histories of stellar systems .
athey , a. e. 2003 , phd thesis , university of michigan barbuy , b. , et al .
2004 , mmsai , 75 , 398 beasley , m. a. , hoyle , f. & sharples , r. m. 2002 , , 336 , 168 bensby , t. , feltzing , s. , & lundstrm , i. 2004 , , 415 , 155 borges , a.c . , idiart , t.p . , de freitas pacheco , j.a .
& thevenin , f. 1995 , , 110 , 2408 briley , m. m. , harbeck , d. , smith , g. h. , & gebel , e. k. 2004 , , 127 , 1588 bruzual , g. , & charlot , s. 1993 , , 405 , 538 bruzual , g. , barbuy , b. , ortolani , s. , bica , e. , cuisinier , f. , lejeune , t. , schiavon , r.p .
1997 , , 114 , 1531 bruzual , g. , & charlot , s. 2003 , , 344 , 1000 busso , g. , cassisi , s. , piotto , g. , castellani , m. , romaniello , m. , catelan , m. , djorgovski , s.g . , recio blanco , a. , renzini , a. , rich , m.r . ,
sweigart , a. , & zoccali , m. 2007 , , 105 , 119 cardiel , n. , gorgas , j. , cenarro , j. , & gonzlez , j. j. 1998 , , 127 , 597 carretta , e. , cohen , j. g. , gratton , r. g. , & behr , b. 2001 , , 122 , 1469 carretta , e. , gratton , r. g. , bragaglia , a. , bonifacio , p. , & pasquini , l. 2004 , , 416 , 925 coelho , p. , bruzual , g. , charlot , s. , weiss , a. , barbuy , b. , & ferguson , j. w. 2007 , , 382 , 498 dotter , a. , chaboyer , b. , ferguson , j. w. , lee , h .- c . , worthey , g. , jevremovi , d. , & baron , e. 2007 , , 666 , 403 freitas pacheco , j.a . & barbuy b. 1995 , , 302 , 718 fulbright , j. p. , rich , r. m. , & mcwilliam , a. 2005 , nuclear physics a , 758 , 197 fulbright , j. p. , rich , r. m. , & mcwilliam , a. 2007 , , 661 , 1152 gibson , b.k . , madgwick , d.s . ,
jones , l.a .
, da costa , g.s . &
norris , j.e . 1999 , , 118 , 1268 girardi , l. , bressan , a. , bertelli , g. , & chiosi , c. 2000
, , 141 , 371 gratton , r. g. , lucatello , s. , bragaglia , a. , carretta , e. , momany , y. , pancino , e. , & valenti , e. 2006 , , 455 , 271 graves , g. j. , faber , s. m. , schiavon , r. p. , & yan , r. 2007 , , submitted ( astro - ph/0707.1523 ) henry , r.b.c . & worthey , g. 1999 , , 111 , 919 ivans , i. i. , sneden , c. , kraft , r. p. , suntzeff , n. b. , smith , v. v. , langer , g. e. , & fulbright , j. p. 1999
, , 118 , 1273 korn , a. j. , maraston , c. , & thomas , d. 2005 , /aap , 438 , 685 lee , h .- c . & worthey , g. 2005 , , 160 , 176 maraston , c. 2005 , , 362 , 799 maraston , c. & thomas , d. 2000 , 541 , 126 maraston , c. , greggio , l. , renzini , a. , ortolani , s. , saglia , r.p . ,
puzia , t.h . , & kissler - patig , m. 2003 , , 400 , 823 ortolani , s. , barbuy , b. , bica , e. , renzini , a. , zoccali , m. , rich , r. m. , & cassisi , s. 2001 , , 376 , 878 prochaska , l.c . , rose , j.a .
& schiavon , r.p .
2005 , , 130 , 2666 proctor , r.n . ,
forbes , d.a . &
beasley , m.a .
2004 , , 355 1327 rich , r.m . ,
sosin , c. , djorgovski , s.g . ,
piotto , g. , king , i.r . , renzini , a. , phinney , e.s . , dorman , b. , liebert , j. , & meyla , g. 1997 , , 484 , l25 rose , j.a .
1994 , , 107 , 206 salaris , m. , & weiss , a. 2002 , , 388 , 492 salasnich , b. , girardi , l. , weiss , a. , & chiosi , c. 2000 , , 361 , 1023 snchez - blzquez , p. , gorgas , j. , cardiel , n. , cenarro , j. , & gonzlez , j. j. 2003 , , 590 , 91 schiavon , r. p. 2007
, , 171 , 146 schiavon , r.p . & barbuy , b. , 1999 , , 510 , 934 schiavon , r. p. , faber , s. m. , castilho , b.v . , & rose , j. a. 2002a , 580 , 850 schiavon , r. p. , faber , s. m. , rose , j. a. , & castilho , b. v. 2002b , , 580 , 873 schiavon , r.p .
, caldwell , n. & rose , j.a .
2004a , , 127 , 1513 schiavon , r.p . , rose , j.a .
, courteau , s. & macarthur , l. 2004b , , 608 , l33 schiavon , r.p . , rose , j.a . ,
courteau , s. & macarthur , l. 2005 , , 160 , 163 schiavon , r.p .
2006 , , 651 , l93 serven , j. , worthey , g. , & briley , m. m. 2005 , , 627 , 754 shetrone , m. d. , & sanquist , e. l. 2000 , , 120 , 1913 tantalo , r. , chiosi , c. & bressan , a. 1998 , , 333 , 419 tautvaisiene , g. , edvardsson , b. , tuominen , i. , & ilyin , i. 2000 , , 360 , 499 thomas , d. , maraston , c. & bender , r. , 2003 , , 343 , 279 thomas , d. , maraston , c. & korn , a. 2004 , , 351 , 19 thomas , d. , maraston , c. , bender , r. , & de oliveira , c. m. 2005 , , 621 , 673 trager , s. c. , worthey , g. , faber , s. m. , burstein , d. , & gonzlez , j. j. 1998 , , 116 , 1 trager , s. c. , faber , s. m. , worthey , g. , & gonzlez , j. j. 2000 , , 120 , 165 trager , s.c .
, worthey , g. , faber , s.m . & dressler , a. 2005 , , 362 , 2 vazdekis , a. , casuso , e. , peletier , r. f. , & beckman , j. e. 1996 , , 106 , 307 vazdekis , a. , peletier , r.f .
, beckman , j.e . & casuso , e. 1997 , , 111 , 203 vazdekis , a. , salaris , m. , arimoto , n. & rose , j.a .
2001 , , 549 , 274 vazdekis , a. , cenarro , a. j. , gorgas , j. , cardiel , n. , & peletier , r. f. 2003 , , 340 , 1317 weiss , a. , peletier , r.f . & matteucci , f. 1995 , , 296 , 73 weiss , a. , salaris , m. , ferguson , j. w. , & alexander , d. r. 2006 , astro - ph/0605666 ) worthey , g. 1998 , , 110 , 888 worthey , g. , faber , s. m. , & gonzlez , j. j. 1992 , , 398 , 69 worthey , g. , faber , s. m. , gonz ' alez , j. j. , & burstein , d. 1994 , , 94 , 687 worthey , g. & ottaviani , d. l. 1997 , , 111 , 377 zoccali , m. , et al .
2004 , , 423 , 507 + + h@xmath15 & fe , c@xmath75 & fe , c , v + h@xmath16 & fe , mg@xmath76 & fe , c , si + h@xmath17 & c , fe , mg & c , ti , ( mg ) + h@xmath18 & c , fe & c , ( si ) , ( mg ) + h@xmath19 & ... & ( ni ) + + + + fe4383 & fe , mg , c@xmath75 & c , fe + fe5015 & ti , mg , fe & c , fe , ( ti ) + fe5270 & fe & c , ( fe ) + fe5335 & fe & fe , c + + + + mg@xmath7 & mg , c & mg , c , fe + mg _ b _ & mg , fe , c & mg , fe , ( cr ) + + + + cn@xmath20 & c , n , o & c , n , o + cn@xmath7 & c , n , o & c , n , o + g4300 & c , o , fe@xmath76 & c , fe , ti + c@xmath74668 & c , o & c , o , ( si ) + + + + ca4227 & ca , c & ca , o , ( cn ) lccccc|cc h@xmath16 & 2.338 & 0.682 & 0.926 & 0.298 & 0.977 & 0.117 & 0.090 + h@xmath18 & 1.398 & -0.660 & -0.148 & -1.308 & -0.324 & 0.215 & 0.050 + h@xmath19 & 2.280 & 1.602 & 1.804 & 1.572 & 2.246 & 0.083 & 0.090 + fe4383 & 1.267 & 2.491 & 3.048 & 4.666 & 3.841 & 0.267 & 0.200 + fe5270 & 1.211 & 1.913 & 2.128 & 2.799 & 2.647 & 0.142 & 0.090 + fe5335 & 0.984 & 1.685 & 1.785 & 2.417 & 2.332 & 0.079 & 0.080 + mg _ b _ & 1.577 & 2.694 & 2.651 & 3.759 & 2.927 & 0.148 & 0.100 + mg@xmath7 & 0.088 & 0.159 & 0.165 & 0.249 & 0.175 & 0.015 & 0.010 + g4300 & 2.496 & 4.656 & 4.001 & 4.819 & 4.740 & 0.198 & 0.100 + c@xmath74668 & 0.322 & 1.680 & 1.906 & 4.510 & 4.382 & 0.131 & 0.100 + cn@xmath20 & -0.070 & 0.025 & 0.020 & 0.063 & -0.007 & 0.010 & 0.005 + cn@xmath7 & -0.047 & 0.050 & 0.047 & 0.093 & 0.018 & 0.008 & 0.008 + ca4227 & 0.321 & 0.557 & 0.581 & 0.887 & 0.927 & 0.041 & 0.100 + llccccccc ngc 6121 & ez_ages & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 & @xmath82 & + & literature & 13@xmath83 & @xmath84 & @xmath85 & & & @xmath86 & a , b + + 47 tuc & ez_ages & @xmath87 & @xmath88 & @xmath89 & @xmath90 & @xmath91 & @xmath92 & + & literature & 12 & @xmath93 & @xmath94 & @xmath95/@xmath96 & @xmath97/@xmath98 & @xmath99 & c , d , e + + ngc 6441 & ez_ages & @xmath100 & @xmath101 & @xmath102 & @xmath103 & @xmath104 & @xmath105 & + & literature & & @xmath106 & @xmath107 & & & @xmath108 & f , g + + ngc 6528 & ez_ages & @xmath109 & @xmath110 & @xmath111 & @xmath112 & @xmath113 & @xmath114 & + & literature & 11 & @xmath51/@xmath115 & @xmath116/@xmath117 & & & @xmath118/@xmath119 & h , i , j , k + + m 67 & ez_ages & @xmath120 & @xmath121 & @xmath122 & @xmath123 & @xmath124 & @xmath125 & + & literature & 3.5 & @xmath126 & @xmath126 & @xmath126 & @xmath126 & @xmath126 & l , m , n + lcccccc standard & h@xmath19 & @xmath23fe@xmath24 & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + h@xmath18 & * h@xmath18 * & @xmath23fe@xmath24 & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + h@xmath16 & * h@xmath16 * & @xmath23fe@xmath24 & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + balmer & * all * & @xmath23fe@xmath24 & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + fe4383 & h@xmath19 & * fe4383 * & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + fe5015 & h@xmath19 & * fe5015 * & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + fe5270 & h@xmath19 & * fe5270 * & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + fe5335 & h@xmath19 & * fe5335 * & mg _ b _ & c@xmath74668 & cn@xmath20 & ca4227 + mg@xmath7 & h@xmath19 & @xmath23fe@xmath24 & * mg@xmath7 * & c@xmath74668 & cn@xmath20 & ca4227 + g4300 & h@xmath19 & @xmath23fe@xmath24 & mg _ b _ & * g4300 * & cn@xmath20 & ca4227 + cn@xmath7 & h@xmath19 & @xmath23fe@xmath24 & mg _ b _ & c@xmath74668 & * cn@xmath7 * & ca4227 + high - z & * h@xmath16 * & * fe4383 * & ... & * g4300 * & cn@xmath20 & ca4227 + all & * all * & * all * & * all * & * all * & * all * & ca4227 + lccccccccccccc standard & 11.4 & @xmath1270.64 & @xmath1280.17 & @xmath1270.20 & @xmath1280.62 & @xmath1280.04 & & 4.0 & @xmath129 & @xmath130 & @xmath65 & @xmath131 & @xmath132 + h@xmath18 & 6.9 & @xmath1270.54 & @xmath1280.17 & @xmath1270.22 & @xmath1280.60 & @xmath1280.05 & & 3.5 & @xmath65 & @xmath130 & @xmath133 & @xmath131 & @xmath134 + h@xmath16 & 8.2 & @xmath1270.57 & @xmath1280.17 & @xmath1270.22 & @xmath1280.60 & @xmath1280.05 & & 3.7 & @xmath135 & @xmath136 & @xmath133 & @xmath131 & @xmath137 + balmer & 8.3 & @xmath1270.57 & @xmath1280.17 & @xmath1270.21 & @xmath1280.60 & @xmath1280.05 & & 3.7 & @xmath135 & @xmath130 & @xmath133 & @xmath131 & @xmath131 + fe4383 & 11.1 & @xmath1270.55 & @xmath1280.08 & @xmath1270.25 & @xmath1280.58 & @xmath1270.04 & & 3.8 & @xmath138 & @xmath132 & @xmath139 & @xmath132 & @xmath129 + fe5015@xmath140 & ... & ... & ... & ... & ... & ... & & 4.3 & @xmath141 & @xmath142 & @xmath138 & @xmath137 & @xmath134 + fe5270 & 11.3 & @xmath1270.62 & @xmath1280.15 & @xmath1270.21 & @xmath1280.60 & @xmath1280.01 & & 3.9 & @xmath139 & @xmath130 & @xmath65 & @xmath131 & @xmath132 + fe5335 & 11.5 & @xmath1270.66 & @xmath1280.19 & @xmath1270.19 & @xmath1280.64 & @xmath1280.06 & & 4.0 & @xmath129 & @xmath130 & @xmath65 & @xmath131 & @xmath132 + mg@xmath7 & 11.4 & @xmath1270.63 & @xmath1280.22 & @xmath1270.20 & @xmath1280.60 & @xmath1280.02 & & 4.0 & @xmath129 & @xmath143 & @xmath65 & @xmath131 & @xmath132 + g4300@xmath144 & 11.6 & @xmath1270.65 & @xmath1280.15 & @xmath1270.33 & @xmath1280.79 & @xmath1280.02 & & ... & ... & ... & ... & ... & ... + cn@xmath7 & 11.4 & @xmath1270.64 & @xmath1280.16 & @xmath1270.20 & @xmath1280.56 & @xmath1280.02 & & 4.0 & @xmath129 & @xmath130 & @xmath65 & @xmath132 & @xmath137 + high - z@xmath144&5.7 & @xmath1270.34 & @xmath1280.03 & @xmath1270.33 & @xmath1280.50 & @xmath1270.10 & & 2.7 & @xmath145 & ... & @xmath51 & @xmath137 & @xmath146 + all@xmath147&7.5 & @xmath1270.47 & @xmath1280.12 & @xmath1270.26 & @xmath1280.53 & @xmath1270.02 & & 3.7 & @xmath65 & @xmath130 & @xmath133 & @xmath138 & @xmath131 + | we present a method for determining mean light - weighted ages and abundances of fe , mg , c , n , and ca , from medium resolution spectroscopy of unresolved stellar populations .
the method , pioneered by schiavon ( 2007 ) , is implemented in a publicly available code called ez_ages . the method and error estimation
are described , and the results tested for accuracy and consistency , by application to integrated spectra of well - known galactic globular and open clusters . ages and abundances from integrated light analysis agree with studies of resolved stars to within @xmath0dex for most clusters , and to within @xmath1 dex for nearly all cases .
the results are robust to the choice of lick indices used in the fitting to within @xmath0 dex , except for a few systematic deviations which are clearly categorized .
the realism of our error estimates is checked through comparison with detailed monte carlo simulations . finally , we apply ez_ages to the sample of galaxies presented in @xcite and compare our derived values of age , [ fe / h ] , and [ @xmath2/fe ] to their analysis .
we find that [ @xmath2/fe ] is very consistent between the two analyses , that ages are consistent for old ( @xmath3 gyr ) populations , but show modest systematic differences at younger ages , and that [ fe / h ] is fairly consistent , with small systematic differences related to the age systematics .
overall , ez_ages provides accurate estimates of fundamental parameters from medium resolution spectra of unresolved stellar populations in the old and intermediate - age regime , for the first time allowing quantitative estimates of the abundances of c , n , and ca in these unresolved systems . |
star clusters have long been key objects to reconstruct the formation and the dynamical and chemical evolutions of galaxies . as the large magellanic cloud ( lmc ) is considered , the study of its star cluster population has allowed us to learn about its spread in metallicity at the very early epoch @xcite ; the existence of a relatively important gap in its age distribution @xcite ; the evidence of vigorous star cluster formation episodes @xcite ; the complexity of the cluster formation rate during the last million years @xcite , etc .
although somewhat inhomogeneous and clearly still incomplete , the astrophysical properties estimated for a significant number of lmc star clusters have been the starting point to address galactic global issues such as the age - metallicity relationship and the cluster formation rate , among others .
therefore , it results of great importance to enlarge the number of objects confirmed as genuine star clusters and to estimate their fundamental parameters . with the aim of providing mainly with age and metallicity estimates for an increasing number of lmc star clusters
, we have continued a long term observational program carried out at cerro tololo interamerican observatory ( ctio ) using different telescopes in conjunction with ccd cameras and the washington photometric filters @xcite .
a total of 61 clusters have been observed and their fundamental parameters estimated ( see , e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
more recently , we took advantage of a wealth of available images at the national optical astronomy observatory ( noao ) science data management ( sdm ) archives , obtained at the ctio 4-m blanco telescope with the mosaic ii camera attached ( 36@xmath636 arcmin@xmath7 field with a 8k@xmath68k ccd detector array , scale 0.274@xmath1/pixel ) and the washington filters . from the whole volume of observed lmc fields
( see * ? ? ?
* ) we identified 206 clusters previously catalogued by ( * ? ? ? * hereafter b08 ) , and studied 107 of them with some detail @xcite . in this work
, we end up the series of studies of unkown or poorly - known lmc clusters with available washington photometry by analysing the remaining objects in the aforementioned sample .
the paper is organised as follows : washington @xmath8 data are presented in section 2 . the star cluster sample is described in section 3 , while the cleaning of the colour - magnitude diagrams ( cmds ) and the estimation of the cluster ages are presented in sections 4 and 5 , respectively .
we discuss the results and build cluster frequencies in section 6 .
finally , conclusions of this analysis are given in section 7 .
the washington @xmath8 photometric data set used in this work were obtained from an comprehensive process that involved the reduction of the raw images , the determination of the instrumental photometric magnitudes , and the standardisation of the photometry .
we have already described in detail such steps not only in the works cited above , but also in @xcite .
for this reason , we summarise here some specific issues in order to provide the reader with an overview of the photometry quality .
the data come from observations carried out on the nights of dec .
18 - 20 , 2008 through the ctio 2008b-0192 program ( pi : d. geisler ) .
the nights resulted to be of excellent photometric quality -with a typical seeing of @xmath0 1@xmath1 ( see table 1 of * ? ? ? * ) - as judged by the rms errors from the transformation to the standard system ( @xmath0 0.02 mag in each filter ) . from extensive artificial star tests we showed that the 50@xmath3 completeness level is located at @xmath9 @xmath0 23.5 - 24.5 mag and @xmath10 @xmath0 23.0 - 24.0 mag , depending on the crowding and exposure time ( see * ? ? ? * ) .
the photometric errors in the @xmath10 mag increase typically from @xmath11 @xmath4 0.01 mag for the @xmath10 range @xmath0 16.0 - 20.0 mags up to @xmath11 @xmath0 0.20 mag for @xmath10 = 24.0 mag .
the errors for the @xmath12 mags resulted in @xmath13 @xmath4 0.01 for the @xmath10 range @xmath0 16.0 - 19.0 mags and increases exponentially up to @xmath11 @xmath0 0.30 mag for @xmath10 = 24.0 mag .
we here analyse cluster cmds down to @xmath10 @xmath0 22.0 mag , which corresponds to a completeness level nearly 100% and photometric errors @xmath11 @xmath4 0.05 mag .
thus , our photometry yield accurate morphology and position of the main features in the observed cmds of the studied fields .
1 shows a schematic chart with the distribution of the cluster candidates catalogued by b08 ( dots ) and the positions of the mosaic ii fields wherein the objects studied here are located ( boxes ) .
some of the clusters placed within these mosaic ii fields have been previously studied by different authors @xcite , while others have been discarded from this analysis since their photometric data are not complete due to the presence of saturated stars in their images ( bsdl351 , 349 , 1980 , h88 - 267 , 308 , kmhk1274 , hs81 , ogle308 , and sl360 ) .
the final cluster candidate sample studied here contains 90 objects . according to b08 s catalogue
the objects under study are of small angular dimension , typically with radius between @xmath0 8@xmath1 and 19@xmath1 with an average of 11@xmath1 . considering the lmc regions traced by @xcite , some clusters appear projected towards the densest 30 doradus , the bar , and the outer bar regions , while others are distributed within the north - west void and the north - west arm , which are less crowded .
all these substructures are within the so - called inner lmc disc ( r @xmath14 4 kpc ; * ? ? ?
in addition , many of the tiny selected objects are apparently composed by a handful of stars , which makes them of very low contrast over the background stellar density .
the clusters included in this study were first recognised by overplotting the positions of those catalogued by b08 to the deepest mosaic ii @xmath15 images , thus avoiding to mismatch the observed objects and the actual list of catalogued clusters .
we thus assigned to the observed clusters the respective names taken from b08 .
we searched for the same names in the digital sky survey ( dss ) and downloaded 15@xmath16@xmath615@xmath16 @xmath17 images centred on the coordinates matched by the dss .
we used the simbad astronomical database as an additional source for checking the cluster coordinates .
when comparing the dss extracted regions with the observed cluster fields in the @xmath8 images , we could confirm the positions of our selected targets . notice that , since most of the observed objects are of small angular size and many of them are projected towards relatively crowded fields with stellar density fluctuations , the task of distinguishing a star cluster from a chance grouping of stars was not at all straightforward .
table 1 contains a complete list of the recognized objects .
in general terms , the observed cmds of the selected objects are the result of the superposition of different stellar populations distributed along the line - of - sight . for this reason , the use of the observed cmds without subtracting the luminosity function and
the colour distribution of stars belonging to the field might lead to wrong interpretations .
moreover , since the catalogued clusters have been identified as small star concentrations on the basis of a stellar density fluctuation in the sky , their real physical nature require a subsequent confirmation . in most of the cases we probably deal with the presence of a genuine star cluster ,
while in other cases it might constitute a chance grouping of stars or the effect of a non - uniform distribution of the interstellar material in that surveyed region . in order to disentangle cluster stars from field stars , it is required to employ cmds of adjacent fields to subtract the local lmc field luminosity function and colour distribution .
we built cluster cmds from all the measured stars distributed within a circle of radius three times those of the clusters .
for this purpose we used the radii ( r ) listed in table 1 , which were obtained from visual inspection of the objects in the deepest @xmath8 images .
they are large enough as to reach the observed field star region .
notice that our main aim consists in cleaning the cluster cmds from the contamination of field stars within areas around the clusters centres nine times larger than @xmath18r@xmath7 , so that we did not need to trace their radial profiles .
once the cluster areas were delineated , we traced four additional regions with areas equals to the cluster regions and placed more or less equidistant to the clusters centres around the cluster circular areas . from each surrounding region
we built a cmd which shows the features of the local lmc star field towards that particular direction .
we then apply a procedure that compares each one of the field cmds to the cluster cmd and subtract from the latter a representative field cmd in terms of stellar density , luminosity function and colour distribution .
we refer the reader to @xcite for details concerning this decontamination method .
each resultant representative field cmd is built from scanning the individual one using cells that vary in size , thus achieving a better field representation than counting the number of stars in boxes fixed in size .
then , the stars in the cluster cmd that fall within the defined cells and closest to the representative positions are eliminated .
the method allows that cells vary in magnitude and colour separately according to the free path between field stars in the cmd , so that they result bigger in cmd regions with a small number of stars , and vice versa .
the free path is defined as @xmath19(colour)@xmath7 + @xmath19(magnitude)@xmath7 = ( free path)@xmath7 , where @xmath19(colour ) and @xmath19(magnitude ) are the distances from the considered star to the closest one in abscissa and ordinate in the field cmd . in practice , an initial rectangular cell with a dimension of ( @xmath19(colour ) , @xmath19(magnitude ) = ( 0.5 , 1.0 ) is put on a single field star .
then , the method looks for the closest star in magnitude and colour ( it will be placed at a corner of the rectangular cell ) , so that the resulting closest magnitude and colour will be used to define the free path of the considered star .
the task is repeated for every star in the field cmd .
therefore , each field star has associated a different cell .
these cells are then superimposed to the cluster cmd , and the stars closest to their centres are eliminated . in this sense ,
the cell sizes depend not only on the stellar density of that field ( the denser a field the more stars in the field cmd ) , but also on the magnitude and colour distributions of those field stars in the cmd , making some parts of the field cmd more populated than others .
for instance , relatively bright field red giants with small photometric errors usually appear relatively isolated at the top - right zone of the cmd , while faint main sequence ( ms ) stars are more numerous at the bottom part of the cmd . for this reason ,
bigger cells are required to satisfactorily subtract stars from the cluster cmd regions where there is a small number of field stars , while smaller cells are necessary for those cmd regions more populated by field stars . after repeating the procedure of subtracting stars from the cluster cmd using the four selected surrounding fields
, we obtained four cleaned cluster cmds which were compared to each other . as a general rule
, each star in the observed cluster cmd keeps not subtracted different times ; the more the times a star is subtracted , the higher the probability of being a field feature .
thus , we convert the number of times that a star appears in the four cleaned cluster cmds in an estimation of its probability ( p ) of being a fiducial feature in that cluster cmd . for instance , stars that are considered to belong to the field population appear once , or do not appear in the four cleaned cluster cmds ( we refer to them as of p @xmath4 25% ) ; stars that could indistinguishably belong to the field or to the studied object are seen twice ( p = 50@xmath3 ) ; and stars that are predominatly found in the cleaned cluster area rather than in the star field population ( p @xmath20 75% ) are identified more than twice . whenever the cmds for stars with p @xmath4 25% looks like that of p @xmath20 75% , we conclude that both the star field and the cluster are of similar features ( ages ) . to illustrate the bulk of the analysis performed , we produced multiple panel figures for the objects listed in table 1 .
as an example , we include fig .
2 , which depicts three cmds and an enlargement of the @xmath10 image centred on sl390 .
we overplotted to the enlarged image a circle representing the cluster radius used in the analysis and marked the stars that have a chance of being cluster members p @xmath20 75@xmath3 .
the three cmds represent the observed cluster cmd for the stars measured within the cluster radius ( upper - left panel ) ; a single field cmd for an annulus centred on the cluster , with an internal radius 3 times that of the cluster and an equal cluster area ( upper - right panel ) ; and the cleaned cluster cmd ( bottom - left ) .
the colour scale represents stars that statistically belong to the field ( p @xmath4 25% , white ) , stars that might belong either to the field or to the cluster ( p @xmath21 50% , gray ) , and stars that predominantly populate the cluster region ( p @xmath20 75% , black ) .
the multiple panel figures for the entire sample of clusters can be found at the on - line version of the journal as supplementary material .
the distribution of stars with different membership probabilities in the cmd as well as in the cluster field were used at a time to confirm the physical reality of the studied cluster candidates .
we required that the stars with p @xmath20 75% within the cluster radius have as a counterpart a cmd resembling that of a star cluster to conclude that we are dealing with a genuine physical system . in some cases
, we reinforced our conclusion about the nature of an object from the analysis of the distribution of stars with p = 50@xmath3 and @xmath4 25@xmath3 , respectively .
we found that a handful of objects could be possible non - clusters , since either the cmds and/or the distribution of stars in the respective fields do not resemble that of an stellar aggregate ( bsdl218 , 256 , 616 , 661 , 1103 , 2794 , 2824 , 2841 , 2842 , 2891 , 2898 , brht60a , 62b , gkk - o155 , 203 , 205 , h88 - 248 , 318 , kmhk125 , 237 , 258 , 289 , 609 , 897 , ogle291 , 303 , 335 , 344 , and sl371 ) .
3 depicts the results for bsdl616 , where the cleaned cmd shows a ms and a red clump ( rc ) mostly composed by field stars .
other possible non - clusters resulted to be single bright stars surrounded by unresolved faint ones .
fortunately , the mosaic ii images have a better resolution and a fainter magnitude limit than those used for catalogued cluster candidates ( see details in * ? ? ?
* ) , so apparent compact extended objects could be resolved . as an example , fig .
4 compares an enlargement of the @xmath10 image centred on kmhk125 to that obtained from the dss .
we estimated ages for the 61 confirmed star clusters by using the set of theoretical isochrones computed by @xcite for the washington photometric system .
the isochrones were first reddened by the appropriate cluster colour excesses and corrected due to the distance effect .
then , we superimposed those for a metallicity level of z = 0.008 to the cluster cmds and chose the ones which best reproduced the cluster features to assign the cluster ages .
the matching procedure was performed using stars with p @xmath20 75@xmath3 .
the estimation of cluster reddening values was made by interpolating the extinction maps of ( * ? ? ?
* hereafter bh ) .
bh maps were obtained from hi ( 21 cm ) emission data for the southern sky .
they furnish us with foreground @xmath22 colour excesses which depend on the galactic coordinates .
we also took advantages of the magellanic clouds extinction values based on rc stars photometry provided by the ogle collaboration @xcite as described in @xcite .
they resulted in average ( 0.02 @xmath2 0.01 ) mag smaller than those obtained by bh .
as for the ( * ? ? ?
* hereafter sfd ) full - sky maps from 100-@xmath23 m dust emission , we decided not to use them since the authors found that deviations are coherent in the sky and are especially conspicuous in regions of saturation of hi emission towards denser clouds and of formation of h@xmath24 in molecular clouds .
the relatively small angular sizes of the clusters did not allow us to trace reddening variations in any extinction map .
finally , we adopted the values obtained from the bh s maps , which are listed in table 1 .
their accuracy is 0.01 mag in @xmath22 or 10@xmath3 of the reddening , whichever is larger .
notice that a relative @xmath12 colour shift smaller than @xmath0 0.05 mag hardly turns into a meaningful difference when matching isochrones of the same metallicity to the cluster mss .
likewise , the mean @xmath12 colour difference for isochrones of the same age and with the closest metallicity values that the washington photometric system is able to distinguish ( @xmath19([fe / h ] ) = 0.20 dex , @xcite ) results larger than @xmath0 0.20 mag .
consequently , the reddening uncertainties do not affect neither the age nor the metallicity adopted for the clusters .
the lmc is located at a distance of 50 kpc @xcite and according to @xcite it has an average depth of ( 3.44 @xmath2 1.16 ) kpc , which implies a difference in its distance modulus as large as @xmath19(@xmath25 ) @xmath0 0.30 mag . given that the clusters could be placed in from of , or behind the lmc , the simple assumption that all of them are at the same distance would lead to age uncertainties of @xmath26(log(@xmath5 ) ) @xmath4 0.06 .
however , the age difference of the isochrones with the same metallicity bracketing the observed cluster mss resulted to be in average @xmath19(log(@xmath5 ) ) = 0.20 .
therefore , we decided to adopt the value of the lmc distance modulus @xmath27 = 18.493 @xmath2 0.008 reported by @xcite for all the clusters .
indeed , we generally found an excellent match .
as for the cluster metallicity we adopted the same value for all of them .
@xcite showed that during the last 3 gyr ( the age range of our cluster sample ) the lmc cluster population has chemical evolved or spread within the range [ fe / h ] = ( -0.4 @xmath2 0.2 ) dex . on the other hand , according to the washington photometry metallicity sensitivity we should use isochrones for metallicity levels in steps of @xmath19([fe / h ] ) = 0.20 dex . further higher metallicity resolution would lead to negligible changes in the isochrones overplotted on the cluster cmds .
however , we found that only isochrones with the same age and metallicity differences larger than @xmath19([fe / h ] ) @xmath0 0.30 - 0.40 dex can be meaningfully differentiated due to the dispersion of the stars .
for this reasons , we adopted a mean value of [ fe / h ] = -0.40 dex for cluster metallicities , and assumed an error of @xmath26([fe / h ] ) = 0.20 dex . in the matching procedure , which was performed looking at the cluster cmd
, we used a subset of isochrones and superimposed them on the cluster cmds , once they were properly shifted by the corresponding @xmath28 = 1.97@xmath6@xmath22 colour excesses and by the lmc apparent distance modulus ( @xmath29 = @xmath10 + 0.58@xmath22 - @xmath30 ) @xcite .
finally , we adopted as the cluster age the one corresponding to the isochrone which best reproduced the cluster main features in the cmd .
the presence of rcs and/or red giant branch stars in some cluster cmds made the matching procedure easier .
table 1 lists the resulting age estimates , while the bottom left panel in fig .
2 ( likewise the on - line figures ) show the corresponding isochrones superimposed . the observed dispersion seen in the cluster cmds
can be encompassed with a couple of isochrones bracketing the derived mean age by @xmath19(log(@xmath5 ) ) = @xmath20.10 .
we also included these adjacent isochrones in fig .
2 ( bottom - left panel ) .
thirty - seven clusters in the sample have previous age estimates ( see table 1 ) .
@xcite made use of washington @xmath8 photometry obtained at the ctio blanco telescope and the mosaic ii camera and derived an age of log(@xmath5 ) = [email protected] for hs156 , in excellent agreement with our present value ( [email protected] ) .
we also found a tight agreement with @xcite , who derived from ogle data an age of log(@xmath5 ) = 8.2 for hs198 ( [email protected] ) .
a more careful analysis requires the comparison of our ages with those of ( * ? ? ? * hereafter g10 ) and ( * ? ? ?
* hereafter p12 ) for 10 and 25 clusters in common , respectively .
g10 used data from the magellanic clouds photometric survey @xcite of clusters mostly distributed in the main body of the galaxy , which is highly crowded .
although they mention that field contamination is a severe effect in the extracted cluster cmds and therefore influences the age estimates , no decontamination from field cmds was carried out .
it would not be unexpected that some of the studied objects are not real star clusters , particularly those with very uncertain age estimates [ @xmath26(log(@xmath5 ) ) @xmath20 0.5 ] .
indeed , g10 estimated an age value for kmhk125 of log(@xmath5 ) = [email protected] , for which our study suggests it could be a possible non - cluster ( see .
this possibility alerts us to the fact that solely the circular extraction of the observed cmds of clusters located in highly populated star fields is not enough neither for an accurate isochrone fitting to the cluster mss nor for confirming their physical nature .
@xcite found little visible evidence for incompleteness for @xmath31 @xmath14 20 mag , corresponding to a ms turnoff of log(@xmath5 ) @xmath32 8.7 .
the present @xmath33 data set nearly reaches the 100@xmath3 completeness level at @xmath10 @xmath0 22.0 mag , which corresponds to ms turnoffs of log(@xmath5 ) @xmath0 9.6 .
furthermore , the scale of our images is 0.274@xmath1 pix@xmath34 , while that of the magellanic cloud photometric survey is 0.7@xmath1 pix@xmath34 , so that crowding effects at the centre of the objects is more important in their images . p12 derived ages from integrated photometry for the clusters of @xcite .
as is well known , the integrated light brings information about the composite stellar population distributed along the line - of sight . for
clusters of small angular size projected towards relatively crowded fields , the field contamination and stochastic effects , in addition to less deep photometric data could mislead the data analysis .
indeed , @xcite compared the ages coming from @xcite , g10 , and p12 and then , due to the large uncertainties in ages coming from integrated magnitudes and colours , assigned ages to the clusters giving the highest priority to hst ages from @xcite , followed by g10 , the ogle data , and finally p12 .
the best - fitting relation between g10 and p12 over 293 clusters in common resulted in an slope of [email protected] and in a pearson coefficient of 0.77 . from this result @xcite concluded that there exists a good agreement between g10 and p12 .
5 shows the comparison between the ages derived by g10 ( left panel ) and p12 ( right panel ) and our present values .
the error bars correspond to the age uncertainties quoted by the authors , while the thick and thin lines represent the identity relationship and those shifted by @xmath21@xmath26(log(@xmath5)@xmath35 ) , respectively .
black filled squares represent clusters that do not fullfill the requirement @xmath26(log(@xmath5)@xmath35 ) + @xmath26(log(@xmath5)@xmath36 ) @xmath20 @xmath37log(@xmath5)@xmath35 - log(@xmath5)@xmath36 @xmath37 .
as can be seen , there is a reasonable agreement , although several clusters significantly depart from the @xmath21@xmath26 strip .
we have carefully checked our multi - panel fig . 2 ( on - line material ) in order to seek a possible explanation for such deviations and found out that stochastic effects caused by the presence of isolated bright stars , a significant field contamination , and a less faint magnitude limit could be some of the factors that affected the integrated light , and hence the p12 s age estimates .
particularly , the presence of a young field ms could cause the p12 s age for h88 - 270 and hs228 resulted younger than our values .
conversely , an old field population could cause bsdl194 , h88 - 240 , and sl390 appear older at their integrated light .
examples of the presence of isolated bright stars can be bsdl87 , h88 - 238 , 276 , among others .
we conclude that the present results point the need of a better study of the lmc clusters with , e.g. , 8-m class telescopes .
our results suggest that nearly 2/3 of the studied candidate star clusters would appear to be genuine physical systems .
we also show that the ages derived by g10 and p12 can reflect those of the composite stellar populations of the lmc field . as far as we are aware , this is the first time that evidence is presented showing that some lmc candidate star clusters are not possible genuine physical systems . in order to examine whether there exists any dependence of the fraction of possible non - clusters with the position in the lmc , we have made use of the central deprojected galactocentric distances of the nine mosaic ii fields computed by assuming that they are part of a disc having an inclination @xmath38 = 35.8@xmath39 and a position angle of the line of nodes of @xmath40 = 145@xmath39 @xcite .
we refer the reader to table 1 of @xcite which includes a summary of orientation measurements of the lmc disc plane , as well as their analysis of the orientation and other lmc disc quantities , supporting the present adopted values .
notice that most of the clusters catalogued by b08 located within the mosaic fields have already been studied previously -we have made use of them here- @xcite and in the present work , so that there is no incompleteness effects in the estimated fraction of possible non - clusters in these areas .
6 illustrates the behaviour of the proportion of possible non - clusters as a function of the deprojected distance .
as can be seen , there exists no visible trend but an averaged fraction of [email protected] within the lmc inner disc ( distance from the lmc centre @xmath4 4@xmath39 , * ? ? ?
this does not seem to be a significant percentage of the catalogued clusters .
if we assume a similar percentage for the whole catalogued lmc clusters distributed throughout the inner disc , we would expect that some @xmath0 ( 300@xmath2140 ) objects were not real stellar aggregates . in the case that the possible non - clusters had their own spatial distribution ,
it would be possible to recover the expected spatial distribution of genuine star clusters by using the former to correct that of the catalogued clusters .
we finally built the cluster frequencies ( cfs ) -the number of clusters per time unit as a function of age @xcite - for the nine studied lmc regions with the aim of investigating their variations in terms of the position in the galaxy recently reported by ( * ? ? ?
* hereafter p14 ) .
particularly he found that 30 doradus turns out to be the region with the highest relative frequency of the youngest clusters , while the log(@xmath5 ) = 9 - 9.5 ( 1 - 3 gyr ) age range is characterised by cluster formation at a higher rate in the inner regions ( e.g. bar , outer bar ; see * ? ? ?
* ) than in the outer ones ( e.g. , blue arm , constellation iii , south - east arm ) . when building the cfs we took into account the unavoidable complication that the cluster ages have associated uncertainties
indeed , by considering such errors , the interpretation of the resulting cfs can differ appreciably from those obtained using only the measured ages without accounting for their errors .
however , the treatment of age errors in the cf is not a straightforward task .
this happens when an age value does not fall in the bin centre and , due to its errors , has the chance to fall outside it .
note that , since we chose bin dimensions as large as the involved errors , such points should not fall on average far beyond the adjacent bins . however , this does not necessarily happen for all the age values , and we should consider at the same time any other possibility .
we followed the procedure outlined by p14 , which achieves a compromise between the age bin size and the age errors .
particularly , we considered the possibility that the extension covered by an age value ( properly a segment of 2@xmath6@xmath26(age ) long ) may have a dimension smaller , similar or larger than the age bin wherein it is placed .
the assigned weight was computed as the fraction of its age segment that falls in the age bin .
7 shows the resultant cfs built from the catalogue of cluster ages compiled by p14 and from the ages estimated in this work , which encompass 90@xmath3 of the whole sample of catalogued clusters located in the studied lmc regions .
the cfs have been normalized to the total number of clusters used .
we have used the deprojected distances of fig . 6 as the independent spatial variable . at first glance
, we confirm the existence of spatial variations of the lmc cf throughout the inner disc .
however , we do not rule out that some percentage of such a variation comes from the fact that we are dealing with relatively small areas and consequently some age regimes result not well sampled .
nevertheless , fig .
7 suggests that the innermost field contains a handful of clusters older than @xmath0 2 gyr ; the epoch of an important burst of cluster formation that took place after a cluster age gap @xcite .
it also shows that there has been a wider spread between different cfs during the most recent 50 myr of the galaxy lifetime .
the astrophysical properties of lmc star clusters have been the starting point to address galactic global issues such as the age - metallicity relationship and the cluster formation rate , among others . however , the number of clusters with estimated properties is by far less than a half of the catalogued clusters .
therefore , it results of great importance to enlarge the number of objects confirmed as genuine star clusters and to estimate their fundamental parameters .
we have used washington photometry for 90 star cluster candidates of small angular size , typically @xmath0 11@xmath1 in radius , distributed within nine fields ( 36@xmath636 arcmin@xmath7 ) located in the lmc inner disc to estimate their ages and to build their local cluster frequencies .
some clusters are projected towards relatively crowded fields . in order to assure a meaninful interpretation of the observed cluster cmds
, we applied a subtraction procedure to statistically clean them from the field star contamination .
thus , we could disentangle cluster features from those belonging to their surrounding fields .
the employed technique makes use of variable cells in order to reproduce the field cmd as closely as possible . out
of the 90 cluster candidates studied , 61 resulted to be genuine physical systems , whereas the remaining ones were classified as possible non - clusters since either their cmds and/or the distribution of stars in the respective fields do not resemble those of stellar aggregates .
we statistically show that @xmath0 ( 13 @xmath2 6)@xmath3 of the catalogued clusters distributed within the inner disc ( distance from the lmc @xmath4 4@xmath39 ) could be possible non - clusters , independently of their deprojected distances .
we derived the ages for the confirmed clusters from the fit of theoretical isochrones computed for the washingotn system to the cleaned cluster cmds . when adjusting a subset of isochrones we took into account the lmc distance modulus and the individual star cluster colour excesses .
the derived ages resulted to be in the age range 7.8 @xmath4 log(@xmath5 ) @xmath4 9.2 .
finally , we built cfs from ages available in the literature and from the ages estimated in this work , which encompass 90@xmath3 of the whole sample of catalogued clusters located in the studied lmc regions .
we found that there exists some spatial variation of the lmc cf throughout the inner disc .
particularly , the innermost field ( deprojected distance @xmath0 0.56@xmath39 ) contains a handful of clusters older than @xmath0 2 gyr , while the wider spread between different cfs has taken place during the most recent 50 myr of the galaxy lifetime .
we are grateful for the comments and suggestions raised by the anonymous referee which helped us to improve the manuscript .
this research has made use of the simbad database , operated at cds , strasbourg , france , and draws upon data as distributed by the noao science archive .
noao is operated by the association of universities for research in astronomy ( aura ) , inc . under a cooperative agreement with the national science foundation .
this work was partially supported by the argentinian institutions conicet and agencia nacional de promocin cientfica y tecnolgica ( anpcyt ) . | we have used washington photometry for 90 star cluster candidates of small angular size -typically @xmath0 11@xmath1 in radius- distributed within nine selected regions in the inner disc of the large magellanic cloud ( lmc ) to disentangle whether they are genuine physical system , and to estimate the ages for the confirmed clusters . in order to avoid a misleading interpretation of the cluster colour - magnitude diagrams ( cmds )
, we applied a subtraction procedure to statistically clean them from field star contamination .
out of the 90 candidate clusters studied , 61 of them resulted to be genuine physical systems , whereas the remaining ones were classified as possible non - clusters since either their cmds and/or the distribution of stars in the respective fields do not resemble those of stellar aggregates .
we statistically show that @xmath0 ( 13 @xmath2 6)@xmath3 of the catalogued clusters in the inner disc could be possible non - clusters , independently of their deprojected distances .
we derived the ages for the confirmed clusters from the fit of theoretical isochrones to the cleaned cluster cmds .
the derived ages resulted to be in the age range 7.8 @xmath4 log(@xmath5 ) @xmath4 9.2 .
finally , we built cluster frequencies for the different studied regions and found that there exists some spatial variation of the lmc cf throughout the inner disc .
particularly , the innermost field contains a handful of clusters older than @xmath0 2 gyr , while the wider spread between different cfs has taken place during the most recent 50 myr of the galaxy lifetime .
techniques : photometric galaxies : individual : lmc magellanic clouds galaxies : star clusters . |
the term ` geometrical frustration ' denotes a novel class of real systems where the arrangements of magnetic ions in the crystal lattice are such that the spins get frustrated in presence of conflicting exchange interactions resulting a complex magnetic state @xcite . on the other hand , ` spin glass ' ( sg ) denotes a group of materials having metastable magnetic ground state associated with cooperative spin freezing in random fashion below a characteristic temperature @xmath2 .
sg is essentially a manifestation of chemical disorder along with frustration due to random as well as competing magnetic interactions @xcite .
due to such proximity between sg and geometrically frustrated ( gf ) systems , sincere effort has been made in recent years to address an important issue , _
i.e. _ whether an sg like ground state can be realized in gf systems in presence of quenched disorder . as a result of rigorous investigations ,
spin freezing has been revealed in quite a few gf systems , such as srcr@xmath3ga@xmath4o@xmath5 , gd@xmath6ga@xmath7o@xmath8 , y@xmath0mo@xmath0o@xmath9 , zn@xmath10cd@xmath11cr@xmath12o@xmath13 , zncr@xmath14ga@xmath15o@xmath13 and so on @xcite .
notably , some of these compositions show sg freezing even in their stoichiometric form where the amount of quenched disorder is supposed to be negligibly small .
thus , the origin of sg state in a gf material is widely debated , and different models have been proposed to account for that @xcite . apart from the origin , often the nature of the sg state in a gf material
is found to be unusual .
for example , the non linear susceptibility analysis on tb@xmath0mo@xmath0o@xmath9 shows unconventional sg state in this chemically ordered gf compound @xcite . in case of doped gf sample such as srcr@xmath16ga@xmath17o@xmath5 , where sufficient chemical disorder is present ,
neutron scattering studies indicate that the ground state magnetic excitation deviates significantly from that expected for a conventional sg @xcite .
it is therefore appears that the sg state in a gf magnet is quite intriguing and there remains several open questions to be addressed .
chemical substitution in a otherwise stoichiometric gf compound can create random exchange interaction through the breakage or alternation of magnetic bonds .
it can also affect the magnetic interaction through the generation of random strains on substitution of atoms of different ionic radii @xcite .
the situation can be more complicated if the substituting atoms are magnetic in nature , as they can introduce additional magnetic interaction in the system . in this work we have investigated the role of chemical substitution ( both magnetic and non - magnetic ) on the magnetic ground state of licro@xmath0 .
licro@xmath0 is a well known gf magnet having layered hexagonal stacking along the @xmath18 axis @xcite .
it is a quasi two dimensional magnetic oxide with extremely weak inter - layer interaction .
recently our group has reported the development of large electric polarization on cu substitution at the li site @xcite which is associated with enormous magneto - structural transition .
the cr@xmath19 ions in licro@xmath0 remain within edge - sharing distorted cro@xmath20 octahedra and it results a layered arrangements of cr ions separated by li@xmath21 and o@xmath22 .
the magnetic frustration in this compound arises due to triangular arrangements of cr@xmath19 ( @xmath23 = 3/2 ) ions in the basal plane coupled with each other through afm type interaction .
the compound shows a complex double-@xmath24 120@xmath1 afm ordering below @xmath25 = 62 k @xcite , although the curie - weiss temperature is substantially high ( @xmath26 = -700 k ) @xcite resulting a large frustration factor @xmath27 = @xmath28/@xmath25 = 11.3 .
the 2d triangular network in licro@xmath0 provides us a good opportunity to study the effect of chemical substitution on the magnetic ground state .
polycrystalline samples of licr@xmath10mn@xmath15o@xmath0 ( @xmath29 = 0.0 , 0.02 , 0.05 , and 0.1 , hereafter denoted by undoped , mn2 , mn5 , and mn10 respectively ) and licr@xmath10ga@xmath15o@xmath0 ( @xmath29 = 0.1 , hereafter denoted by ga10 ) were prepared as described elsewhere @xcite .
powder x - ray diffraction ( cu k@xmath30 ) was carried out on all the samples which are found to be single phase in nature having rhombohedral crystal structure ( space group : @xmath31 ) .
the estimated lattice parameters of the undoped sample are @xmath32 = 2.92 and @xmath18 = 14.43 , which are very close to the previously reported values .
dc magnetization ( @xmath33 ) measurements were performed between 2 and 300 k using a quantum design squid magnetometer and a cryogen free high magnetic field system from cryogenic ltd . ,
the heat capacity was measured on quantum design physical properties measurement system using relaxation technique .
element mapping was performed in a high resolution transmission electron microscope ( tem ) from jeol .
.variation of the afm onset temperature ( @xmath34 ) , zfc - fc bifurcation temperature ( @xmath35 ) , coercive field ( @xmath36 ) at 3 k , and difference between fc and zfc magnetization ( @xmath37 ) at 3 k for licr@xmath10a@xmath15o@xmath12 ( a = ga , mn ) . [ cols="^,^,^,^,^ " , ] [ values ]
the @xmath38 variation of @xmath33 for doped and undoped samples is depicted in fig .
1 . measurements were performed in zero - field - cooled ( zfc ) and field - cooled ( fc ) conditions under an applied magnetic field of 100 oe .
both zfc and fc curves of licro@xmath0 increase gradually below 300 k and show a peak around @xmath34 = 75 k which signifies the onset of afm transition as reported earlier @xcite .
below 15 k , @xmath33 shows a sharp upturn with decreasing @xmath38 .
such upturn in low dimensional magnetic systems is often attributed to paramagnetic impurities and/or broken chain effect @xcite .
a bifurcation between zfc and fc curves appears below @xmath35 @xmath39 250 k for the pure sample . in licro@xmath0 , such irreversibility
was reported earlier and it was attributed to the formation of _ frozen magnetic clusters _ having short range magnetic correlations @xcite . on ga doping ( see the @xmath33-@xmath38 curve of ga10 ) , the overall behavior of the thermomagnetic curve remains almost unaltered .
although , the peak becomes broader with a lowering of @xmath34 ( 65 k ) .
the magnitude of @xmath33 decreases slightly in case of ga - doped sample . in contrary ,
mn doping at the cr site causes considerable change in the magnetic properties of licro@xmath0 .
1(b)-1(d ) show zfc and fc @xmath33(@xmath38 ) curves of mn2 , mn5 , and mn10 samples respectively , between 2 k and 300 k. a relatively weak hump like feature in all the doped samples can be observed in the zfc or fc curves signifying the onset of the afm transition .
the anomaly is found to be around @xmath34 @xmath39 54 k , 51 k and 45 k for mn2 , mn5 , and mn10 samples respectively . in mn5 and mn10 , the low-@xmath38 _ curie tail_-like rise is completely absent as well .
the bifurcation point between fc and zfc data gradually shifts to lower @xmath38 with increasing mn concentration ( see table i ) .
we have plotted @xmath40 in fig .
1(e ) to have a quantitative idea of the irreversibility in @xmath33 . for undoped , ga10 and
mn2 samples @xmath41 becomes non - zero below about 250 k , 285 k and 280 k respectively , whereas for mn5 and mn10 samples , it is negligibly small down to 50 k and 20 k respectively .
however , below this temperatures @xmath41 rises sharply for mn5 and mn10 indicating large thermomagnetic irreversibility ( see table i ) . such low temperature rise in @xmath42
is absent in other samples and it signifies a different mechanism for the observed irreversibility in these two compositions .
2(a ) depicts isothermal field dependence of @xmath33 at 3 k for all the samples . for the undoped and ga10 samples ,
the @xmath33-@xmath43 data is linear and does not show any hysteresis .
this signifies strong afm character .
@xmath33(@xmath43 ) of the mn - doped samples show slight non - linearity especially at the high field region with curvature increasing with increasing mn concentration . the magnitude of @xmath33 is found to rise with mn doping , while for 10% ga doped sample it is almost same as of the undoped sample .
fig.2(b ) emphasizes the low field region of @xmath33(@xmath43 ) curves of the mn doped samples .
interestingly , the presence of hysteresis can be observed in mn5 and mn10 compositions with a systematic enhancement of coercivity with mn content ( see table i ) .
an sg like sate is generally characterized by a free energy landscape with innumerable numbers of nearly degenerate ground state configurations separated from each other by potential barriers of random height ( landscape of random potential wells ) @xcite .
a signature of such scenario is the presence of magnetic relaxation , which occurs due to the passage from one metastable state to another with time .
fig.2(c ) depicts normalized @xmath33 _ vs. _ time(@xmath44 ) data for the mn and ga doped samples at 3 k. to record this data , samples were first zero field cooled from 300 k , and a field of 100 oe was switched on .
all the mn - doped samples show the presence of finite relaxation with the magnitude getting higher with mn concentration . however , we failed to observe any measurable relaxation in undoped and ga10 sample .
these relaxation data in mn - doped samples are found to be best described by a modified stretched exponential law @xcite , @xmath45 \label{expo}\ ] ] where , @xmath46 is the initial magnetization , @xmath47 is the amplitude of the metastable part , @xmath48 is the time constant and the parameter @xmath49 signifies the distribution of local energy minima .
it approaches unity for a system with long range magnetic order .
fitting to the @xmath33(@xmath44 ) data with this equation results @xmath50 = 0.64 , 0.63 , 0.60 for mn2 , mn5 , and mn10 respectively .
the gradual decrease of the value of @xmath50 signifies that the system passes through increased number of local minima in the energy landscape during relaxation .
systems having spin glass like ground states are found to show @xmath50 ranging between @xmath390.2 - 0.6 @xcite .
the estimated @xmath50 values of the present compositions practically fall in this regime . evidently , the signature of metastability in mn -doped samples are not visible in the undoped or ga - doped samples .
it thus seems that the magnetic ground state in nonmagnetic ga - doped sample is different from that of magnetic mn - doped counterparts . a possible way to distinguish a glassy metastable state from a long range
ordered one is through magnetic memory measurements .
we performed field stop field cool ( fsfc ) memory in @xmath33(@xmath38 ) measurement on the mn and ga doped samples ( see fig.3 ) @xcite . in this protocol ,
sample was first field cooled from 300 k down to 5 k with intermediate stops of 3600 s each with the magnetic field being reduced to zero during each stop ( denoted as @xmath51 ) . after cooling
, the samples were subsequently heated back to 300 k without any stop in presence of @xmath43 ( denoted as @xmath52 ) .
a system with frozen / blocked spins ( or spin clusters ) is expected to show anomalies in the heating curve at the same temperatures where the sample was allowed to age during cooling .
we do not observe any such anomalies in ga10 and mn2 samples ( figs .
3 ( a ) and ( b ) respectively ) at the stopping temperatures . on the other hand @xmath52 curves for mn5 and mn10 ( figs .
3(c ) and 3(d ) respectively ) show anomalies in @xmath52 with clear changes in slope at the stopping positions .
the anomalies associated with the memory effect is found to be much stronger in mn10 sample .
although the existence of fsfc memory effect is a convincing signature of a frozen or blocked magnetic state , it can not distinguish between a spin glass and superparamagnet . in order to resolve this issue , memory measurement in zfc condition
was performed on mn10 . here
the protocol remains almost the same as of fssc baring the fact that the cooling was performed in zero field with a stop for 18000 s at 12 k. fig .
3 ( e ) shows the subsequent heating curves ( @xmath52 ) along with the reference curves ( @xmath53 ) , which is actually simple zfc heating curve without any stop during cooling .
the difference curve @xmath54 = @xmath52 - @xmath53 is plotted for mn10 in the insets of fig .
interestingly , @xmath52 and @xmath53 curves for mn10 follows different path around the stopping temperature of 12 k. this feature gets much more prominent in the difference curve as shown in the inset of fig .
3(e ) , where a dip can be seen with its minimum at @xmath55 = 12 k. the appearance of zfc memory in mn10 signifies that the substitution of cr ions with mn disrupts the afm order and turns on a spin glass like magnetic ground state .
it is now pertinent to know the nature and origin of sg state in mn10 sample , particularly to what extent it corresponds to a conventional sg state .
we addressed this point through heat capacity ( @xmath56 ) measurements on both mn10 and ga10 samples as depicted in fig .
4 . in the @xmath56 versus @xmath38 data
[ see fig .
4(a ) ] , a clear anomaly around 45 k is observed in both the samples which signifies the long range afm transition .
apparently the @xmath56 vs. @xmath38 data of two samples look identical , however a careful look brings out some subtle differences which are important to characterize the magnetic ground state of these compositions .
the magnetic contribution to the specific heat ( @xmath57 ) has been calculated by subtracting the lattice contribution ( @xmath58 ) from total heat capacity @xmath56 . here
@xmath58 has been estimated from the heat capacity data of isostructural non - magnetic compound licoo@xmath0 taken from reference @xcite followed by proper scaling as prescribed by bouvier _
the main panel of fig .
4(a ) ( left axis ) shows the @xmath59 versus @xmath38 plot .
@xmath25 in both the samples are seen as a broad peak around 45 k ( this resembles well with the reported @xmath59 versus @xmath38 data of pure licro@xmath0 @xcite ) .
a marked difference is observed between @xmath59 data of mn10 and ga10 below @xmath25 . for ga10 , @xmath59 smoothly decreases with decreasing @xmath38 , whilst @xmath59 shows a shoulder like feature below about 14 k in mn10 .
this feature matches well with the fc - zfc bifurcation temperature @xmath35 in the @xmath60data of the same sample .
we have carefully examined the low temperature part of @xmath57 of both the samples as depicted in fig .
. a higher value of @xmath57 is observed for the mn10 sample . in mn10 ,
the larger value of the magnetic entropy ( @xmath61 ) calculated from @xmath57 ( see inset of fig .
4(b ) ) is related to the excess magnetic disorder . for an afm sample with linear @xmath62 dispersion relation of magnons ,
one expects @xmath57 to vary as @xmath63 at low-@xmath38 .
a predominant @xmath63 dependent @xmath57 is found for ga10 , which signifies an ordered afm state .
we have fitted the observed @xmath57 data of mn10 and ga10 samples using the relevant algebraic expressions and it brings out some novel information regarding their respective magnetic ground states . for ga10 , it is found that a simple @xmath63 is unable to provide a good fit to @xmath57 below 14 k. rather an equation @xmath64 provides the best fit to the data with @xmath65 = 3.6 mj(mol)@xmath66k@xmath67 and @xmath68 = 9.77@xmath6910@xmath70 mj(mol)@xmath66k@xmath71 .
since these samples are highly insulating , an electronic origin of linear @xmath38 term in @xmath57 can be ruled out .
such linear @xmath38 term is likely to have magnetic origin and probably it denotes some disordered magnetic state . now turning to mn10
, @xmath57 shows a linear behavior below 14 k , which corresponds an sg state ( since the sample is an insulator ) @xcite .
an upward curvature is observed below about 3 k in the @xmath72 data of mn10 . the linear part between 12 and 3 k ,
if extrapolated , hits the abscissa at a positive finite value of @xmath38 . such finite intercept and low-@xmath38 curvature
are previously observed in several canonical sgs @xcite .
@xmath57 between 12 and 3 k was fitted by an algebraic formula @xmath73 with @xmath74 = -127.66 mj(mol)@xmath66k@xmath66 and @xmath75 = 55.25 mj(mol)@xmath66k@xmath70 .
the sg state in mn10 sample is of reentrant type , which develops out of an afm ordered state .
occurrence of reentrant spin glass ( rsg ) in afm system , albeit fewer in number than its fm counterpart , are well documented in the literature @xcite .
it has been argued that mn@xmath19-cr@xmath19 interaction can be fm type particularly when the metal atoms are in octahedral oxygen environment @xcite .
substitution of mn at the cr site can give rise to random fm bonds in the otherwise 120@xmath1 non - co - linear afm structure of licro@xmath0 ( see fig .
5 ) . in the random field model of rsg as proposed by aeppli _
et al . _ , @xcite presence of spatially uncorrelated conflicting magnetic bonds can destroy the long range fm ordering through the emergence of random microscopic field .
the system then attains a frozen non - ergodic state similar to conventional spin glasses .
an equivalent scenario can occur when an afm ordering is destroyed resulting a reentrant glassy state , which would be appropriate for the present mn - doped licro@xmath0 sample .
similar argument was actually provided in case of afm spin glass observed in cr doped manganites @xcite .
a cartoon of the 120@xmath1 afm spin arrangement on a 2d hexagonal lattice is depicted in fig .
substitution of one magnetic impurity ( dashed spin in fig .
5(d ) ) at the center of the hexagon with nearest neighbor fm interaction ( similar to cr - mn fm bond in mn doped licro@xmath0 ) can destroy the ordering arrangement through the development of conflicting sense of interactions .
if such substitutions are spatially random , the system will never attain an order arrangements of spins ( like the regular gf system without quenched disorder ) and may lead to a glassy state . it has long been argued whether the ground state of an rsg system is truly spin glass like or sg state coexists with the long range ordered state @xcite . unlike ga10 , we do not observe any @xmath63 term in the @xmath57 versus @xmath38 ( expected for an afm state ) plot below 14 k of mn10 .
this indicates that the ground state of mn10 is truly sg type without any coexisting afm component .
doping of magnetic atoms often results superparamagnetic like phase due to the clustering of dopant atoms ( as in cu - mn or au - fe alloys @xcite ) . in that case
one would expect a @xmath76 term in the @xmath57 versus @xmath38 plot due to intra - cluster fm coupling .
@xcite a clear linear variation of @xmath57 in mn10 ruled out the possibility of any superparamagnetic clustering effect .
we physically examined the sample using high resolution transmission electron microscope for element mapping , where the spatial distribution of the elements present in the sample can be observed .
our data ( see fig . 5 ) does not indicate any clustering of mn atoms even down to the nanometer scale .
it is generally believed that a true rsg state exists only in a magnetic system below three dimension @xcite and thus mn10 can be regarded as a novel example of such complete reentry for its layered quasi-2d structure .
we would like to conclude by comparing our result with the recently published work on doped zncr@xmath0o@xmath4 , where a quasi - spin glass state is observed on small amount ( 1% to 5% ) of ga doping @xcite .
the quasi spin glass state is characterized by thermomagnetic irreversibility and a linear term in the @xmath56 versus @xmath38 data .
the authors argued that the substitution of ga at the cr site creates _ quasi - spins _ which give rise to some degree of non - ergodicity . for our case even 10%
ga doping does not give rise to a true spin glass state .
however , it should be noted that a linear-@xmath38 part is indeed present on top of the cubic part in @xmath57 of ga10 .
the coefficient of linear term is 3.6 mj(mol)@xmath66k@xmath70 which is substantially smaller than the linear term in ga doped zncr@xmath0o@xmath4 ( 9.6 and 47.6 mj(mol)@xmath66k@xmath70 on 1% and 5% ga doping respectively ) .
ga10 is certainly not a spin glass , but it resembles to some extent with the quasi spin glass state reported in doped zncr@xmath0o@xmath4 .
both licro@xmath0 and zncr@xmath0o@xmath4 are gf magnetic systems , although they are quite different as far as the crystal structure ( layered and cubic spinel respectively ) and dimension of magnetic interaction ( quasi-2d and 3d respectively ) are concerned .
zncr@xmath0o@xmath4 undergoes first order phase transition paving the path for disorder through strain field distribution @xcite .
these factor may enhances the chances of added metastability in case of ga doped zncr@xmath0o@xmath4 . summarizing
, a true sg state is only observed in licro@xmath0 on mn doping at cr site , whilst similar ga doping does not lead to an sg state .
the 10% mn - doped sample shows a true reentrance of sg phase in an otherwise afm state .
it appears that random magnetic impurities are essential for the development of an glassy magnetic state in this gf compound .
this is quite different from the case of other gf systems where very small non - magnetic impurity can turn the system to a glassy phase .
therefore , it seems that there is no generalized rule for the development of glassy state in a gf system .
sc wishes to thank csir ( india ) for his research studentship .
the funding from csir ( grant number : 03(1209)/12/emr - ii ) for the present work is thankfully acknowledged .
the authors also like to thank the low temperature & high magnetic field ( lthm ) facilities at ugc - dae csr , indore ( sponsored by dst ) for heat capacity measurements .
j. e. greedan , j. mater . chem . * 11 * , 37 ( 2001 ) .
a. p. ramirez , czech .
* 46 * , 3247 ( 1996 ) .
a. harrison , j. phys . :
matter * 16 * , s553 ( 2004 ) .
a. p. ramirez , annu .
sci . * 24*,453 ( 1994 ) .
a. p. ramirez , g. p. espinosa , and a. s. cooper , phys .
lett . * 64 * , 2070 ( 1990 ) .
p. schiffer , a. p. ramirez , d. a. huse , p. l. gammel , u. yaron , d. j. bishop , and a. j. valentino , phys .
lett . * 74 * , 2379 ( 1995 ) .
m. j. p. gingras , c. v. stager , n. p. raju , b. d. gaulin , and j. e. greedan , phys .
lett . * 78 * , 947 ( 1997 ) .
d. fiorani , s. viticoli , j. l. dormann , j. l. tholence , and a. p. murani , phys .
b * 30 * , 2776 ( 1984 ) .
w. ratcliff ii , s. -h .
lee , c. broholm , s. -w .
cheong , and q. huang , phys .
b * 65 * , 220406 ( 2002 ) .
a. andreanov , j. t. chalker , t. e. saunders , and d. sherrington , phys .
b * 81 * , 014406 ( 2010 ) .
d. k. singh and y. s. lee , phys .
lett . * 109 * , 247201 ( 2012 ) . s .-
lee , c. broholm1 , g. aeppli , a. p. ramirez , t. g. perring c. j. carlile , m. adams , t. j. l. jones and b. hessen , europhys . lett.*35 * , 127 ( 1996 ) .
l. k. alexander , n. bttgen , r. nath , a. v. mahajan , and a. loidl , phys .
b * 76 * , 064429 ( 2007 ) . j. sugiyama , m. mansson , y. ikedo , t. goko , k. mukai , d. andreica , a. amato , k. ariyoshi , and t. ohzuku , phys .
b * 79 * , 184411 ( 2009 ) .
s. seki , y. onose , and y. tokura , phys .
* 101 * , 067204 ( 2008 ) .
a. olariu , p. mendels , f. bert , l. k. alexander , a. v. mahajan , a. d. hillier , and a. amato , phys .
b * 79 * , 224401 ( 2009 ) .
k. dey , s. majumdar and s. giri , phys .
b * 87 * , 094403 ( 2013 ) .
h. kadowaki , h. takei , and k. motoya , j. phys .
: condens .
matter * 7 * , 6869 ( 1995 ) .
y. motoyama , h. eisaki , and s. uchida , phys .
* 76 * , 3212 ( 1996 ) .
s. chattopadhyay , s. giri , and s. majumdar , j. phys . : condens .
matter * 23 * , 216006 ( 2011 ) .
k. binder and a. p. young , rev .
* 58 * , 801 ( 1986 ) .
j. a. mydosh , _ spin glasses : an experimental introduction _
( taylor & francis , 1993 ) . j. c. phillips , rep .
59 * , 1133 ( 1996 ) .
r. v. chamberlin , g. mozurkewich , and r. orbach , phys .
52 * , 867 ( 1984 ) .
r. s. freitas , l. ghivelder , f. damay , f. dias , and l. f. cohen , phys .
b * 64 * , 144404 ( 2001 ) .
f. wang , j. zhang , y. -f .
chen , g. -j .
wang , j. -r .
sun , s. -y .
zhang , and b. -g .
shen , phys .
b * 69 * , 094424 ( 2004 ) .
d. chu , g. g. kenning , and r. orbach , phys .
72 * , 3270 ( 1994 ) .
a. bhattacharyya , s. giri , and s. majumdar , phys .
b * 83 * , 134427 ( 2011 ) .
s. chattopadhyay , s. jana , s. giri , and s. majumdar , j. phys . : condens .
matter * 24 * , 436005 ( 2012 ) .
m. sasaki , p. e. jonsson , h. takayama , and h. mamiya , phys .
b * 71 * , 104405 ( 2005 ) . y. sun , m. b. salamon , k. garnier , and r. s. averback , phys .
* 91 * , 167206 ( 2003 ) .
h. kawaji , m. takematsu , t. tojo , t. atake , a. hirano and r. kanno , j. thermal analysis and calorimetry * 68 * , 833 ( 2002 ) .
m. bouvier , p. lethuillier , and d. schmitt , phys .
b * 43 * , 13137 ( 1991 ) .
d. l. martin , phys . rev .
b * 20 * , 368 ( 1979 ) .
j. dho , w. s. kim , and n. h. hur , phys .
lett . * 89 * , 027202 ( 2002 ) .
w. r. chen , f. c. zhang , j. miao , b. xu , x. l. dong , l. x. cao , x. g. qiu , b. r. zhao , and p. dai , appl .
. lett . * 87 * , 042508 ( 2005 ) .
x. zhang , s. tang , y. li , and y. du , phys .
a * 374 * , 2175 ( 2010 ) .
s. nimori and d. li , j. phys .
. jpn . * 75 * , 195 ( 2006 ) .
g. h. jonker , physica * 22 * , 707 ( 1956 ) .
y. sun , w. tong , x. xu , and y. zhang , phys .
b * 63 * , 174438 ( 2001 ) .
x. xiao , s. l. yuan , y. q. wang , g. m. ren , j. h. miao , g. q. yu , z. m. tian , l. liu , l. chen , and s. y. yin , solid state commun .
* 141 * , 348 ( 2007 ) .
j. deisenhofer , m. paraskevopoulos , h. -a .
k. von nidda , and a. loidl , phys .
b * 66 * , 054414 ( 2002 ) .
r. ganguly , i. k. gopalakrishnan , and j. v. yakhmi , physica b * 275 * , 308 ( 2000 ) .
g. aeppli , s. m. shapiro , r. j. birgeneau , and h. s. chen , phys .
b * 28 * , 5160 ( 1983 ) .
k. jonason , j. mattson and p. nordblad , phys .
b * 53 * , 6507 ( 1996 ) . i. mirebeau , s. itoh , s. mitsuda , t. watanabe , y. endoh , m. hennion , and r. papoular , phys . rev .
b * 41 * , 11405 ( 1990 ) .
t. h. kim , m. c. cadeville , a. dinia , and h. rakoto , phys .
b * 53 * , 221 ( 1996 ) .
s. niidera and f. matsubara , phys .
b * 75 * , 144413 ( 2007 ) .
j. o. thomson , and j. r. thompson , j. phys .
f : metal phys . * 11 * , 247 ( 1981 ) .
a. d. laforge , s. h. pulido , r. j. cava , b. c. chan , and a. p. ramirez , phys .
110 * , 017203 ( 2013 ) .
i. kagomiya , h. sawa , k. siratori , k. kohn , m. toki , y. hata , and e. kita , ferroelectrics * 268 * , 327 ( 2002 ) .
lee , c. broholm , t. h. kim , w. ratcliff ii , and s -w .
cheong , phys .
lett . * 84 * , 3718 ( 2000 ) . | the development of spin glass like state in a geometrically frustrated ( gf ) magnet is a matter of great debate .
we investigated the effect of magnetic ( mn ) and nonmagnetic ( ga ) doping at the cr site of the layered gf antiferromagnetic compound licro@xmath0 .
10% ga doping at the cr site does not invoke any metastability typical of a glassy magnetic state .
however , similar amount of mn doping certainly drives the system to a spin glass state which is particularly evident from the relaxation , magnetic memory and heat capacity studies .
the onset of glassy state in 10% mn doped sample is of reentrant type developing out of higher temperature antiferromagnetic state .
the spin glass state in the mn - doped sample shows a true reentry with the complete disappearance of the antiferromagnetic phase below the spin glass transition .
mn doping at the cr site can invoke random ferromagnetic cr - mn bonds in the otherwise 120@xmath1 antiferromagnetic triangular lattice leading to the non - ergodic spin frozen state .
the lack of spin glass state on ga doping indicates the importance of random ferromagnetic / antiferromagnetic bonds for the glassy ground state in licro@xmath0 .
spin glass state in gf system has been earlier observed even for small non - magnetic disorder , and our result indicates that the issue is quite nontrivial and depends strongly on the material system concerned . |
the random walk model of price changes in financial time series has been so durable because it is nearly correct . at first approximation the difference between real price changes and the random walk model is too small to be detected using traditional time series analysis @xcite .
more precisely , when looking at large samples of data , some features appear that break the random walk approximation .
for example , the statistics of price distribution at small time scales is not gaussian but governed by non - extensive statistics @xcite .
we can also detect large range correlation in the absolute returns , which mean that persistent behaviors exist that are not embedded in the random walk model @xcite , which can be seen as a consequence of the non - extensive statistics @xcite .
more explicitly , the non - extensive formalism provides an expression for the probability density function @xmath0 of price returns @xmath1 at a given time scale @xmath2 @xcite : @xmath3\right\}_+{^\frac{1}{1-q}}\ ] ] where @xmath4 is a real parameter representing the degree of non - extensivity ( @xmath5 in the gaussian limit ) , @xmath6 is a normalization constant and @xmath7 is a scale parameter .
also , @xmath8 is proportional to the variance of the distribution . in the expression of @xmath0
, the subindex + indicates that @xmath9 if the expression inside the brackets is non - positive . in general , for real markets , using large samples of data
, the @xmath4 index can be found to range from @xmath10 to @xmath11 .
this represents intuitively the degree of the resulting anomalous diffusion from the underlying interaction among financial trades . under certain approximations , regarding a free diffusion process , we can write @xcite : @xmath12 super diffusion occurs for @xmath13 . in fig .
[ 5minret ] , we illustrate this behavior on the return distributions of the euro future contract , sampled in 5 minutes units .
the similarity in shape for two different years , 2002 where the euro contract was moving up and 2005 where it was globally down is observed .
these shapes correspond effectively to @xmath13 .
however , any deviations from the gaussian limit can not be detected at a local level , when we observe the market in a short window of time .
our first statement is still valid . at first order of observation , prices in financial markets
behave randomly and it remains impossible to predict whether the next price movement will be up or down . in the following , we show that this difference between real markets and random walks , as small as it is , is detectable using modern statistical analysis with hypothesis testing , even when we observe the market locally .
in particular , it is detectable once we wan build a trading systems on the basis of multivariate analysis and hypothesis testing @xcite .
indeed , tools of statistical physics have been proven to be efficient in many areas , like extracting the average properties of a macroscopic system from its microscopic dynamics , even if approximately known .
the same holds for financial systems . even though it is difficult or almost impossible to write down the microscopic equation of motion that drives prices at each instant
, it is possible to extract a relevant statistical information , that makes sense to take decisions at a local level . in a first part
, we exemplify this issue on the behavior with time of the euro future contract ( ec ) at high frequency .
we show that we can infer the non - random content of the ec erratic behavior using a multivariate analysis embedded in a trading algorithm . in a second and third parts ,
we examine different markets , largely uncorrelated to the euro future , namely the dax and cacao future contracts ( labeled as fdax and cc ) .
the same procedure is followed using a trading system , based on the same ingredients .
for the system running on the cc , we use the same system as built on the euro future .
we show that similar and good results can be obtained on this variety of markets and we conclude that this is an evidence that some invariants , as encoded in our system , have been identified .
we use five minutes sampling of the ec time series , from january 2000 till august 2011 , which makes 839k quotes that we use to build the trading system .
we conserve only the close of each quote .
this large sample of data points is necessary to infer statistical properties with a high confidence level , as shown in the following .
also , in the context of this analysis , the fine tuning of the time series with a five minutes resolution is useful to focus on possible intermittent behavior of the series at small scales ( five minutes ) , that could disappear at larger scales .
a typical quote of the ec is like @xmath14 .
the unit of the last digit is what we call a basis point .
for example , we consider that a price movement from @xmath14 to @xmath15 corresponds to a price change of @xmath16 basis points .
more precisely , if we buy the contract at time t1 ( on the quote q1 ) at @xmath14 and sell this contract at time t2@xmath17t1 ( on the quote q2 ) at @xmath15 , then this trade corresponds to a gain of @xmath16 basis points ( without fees ) . to keep the procedure as close to reality as possible we consider fees of two times the slippage , which means that this trade is counted in our approach as a trade of @xmath18 basis point ( net of fees ) . a fundamental issue in the analysis is to break the data samples in three parts , that we call in - sample , out - sample and live - sample .
the decomposition is done as follows : * 2000 - 2007 : in - sample * 2008 - 2009 : out - sample * 2010 - 2011 : live - sample what is the interest of this decomposition of the data series ?
the idea is that we intend to build a trading system on this series .
this means that we intent to design an algorithm that will take decisions like buy or sell 1 ec contract at a given quote .
this decision at a given quote will be based on multivariate analysis , as mentioned in the beginning . in order to process this way
, we need a data sample on which the algorithm is built and all parameters of the algorithm are fitted .
therefore , this data sample needs to be large in order to be relevant statistically .
this sample is called in - sample ( i ) and is defined as the period 2000 - 2007 .
the second sample , called out - sample ( ii ) , is used as a validation stage .
all algorithms built on ( i ) are expected obviously to give satisfactory results on ( i ) . however , as parameters of the model are fitted on the sample ( i ) , there is no guarantee that the model could behave properly on another data sample .
if it does so , this means that the algorithm is not a pure artifact and contains a part of the real dynamics of the market .
this is the purpose of the sample ( ii ) , defined as the period 2008 - 2009 .
if the trading system built on ( i ) fails on ( ii ) , it is rejected and another algorithm is designed .
note that we have other intermediate validation stage to make the full process more robust : we come back on this point later in the article .
also , note that there is no guarantee at this level that what we describe in this paragraph is possible .
finally , once we have obtained an algorithm that works on ( ii ) and satisfies our robustness tests , if any , we observe it on what we call the live - sample ( iii ) , defined as the period 2010 - 2011 .
our building process is made to guarantee at this step the good functioning of the trading system and that s what we show in the following .
note that if we can drive the analysis to this last step and if it works , it is a clear proof of our claim of the previous part on a specific example ( ec ) : the difference between real markets and random walks , as small as it is , is detectable using modern statistical analysis in multivariate analysis .
the multivariate approach refers to the number of parameters introduced in the definition of triggers for trades decisions along the ec series .
the basic elements of the algorithm design can be much simple .
the gross featurse of a trend - following strategy are exposed in @xcite .
let us note @xmath19 the value of the price series at time @xmath20 .
an exponential moving average @xmath21 of length memory @xmath2 can be defined on this time series , as : @xmath22 from being initially with no position , a trend - following system buys one share when @xmath23 reaches a given value @xmath24 and stays long until @xmath23 hits the value @xmath25 , at which point the system sells back and takes the opposite position , and so on .
a complexity can easily be added to this mechanism by defining intermediate thresholds to break positions taken by the system @xcite .
the trade distribution for this simple theoretical system is given in fig .
[ theo ] . obviously , there is no possibility with such a simple algorithm to reconstruct a profitable strategy over ten years of high frequency data .
however , we can use the ground idea of this mechanism , namely trend - following , in building a more complex architecture .
we use four different memory lengths and consider crossing of this exponential moving averages , as potential triggers for trade decisions .
not all moving averages are used for each decision .
the choice is based on ranges of volatilities .
indeed , we have observed that there are some transition domains in volatilities of prices where it is preferable not to trade or to branch more stringent triggers .
an important idea in our structure is also some exit conditions based on extreme conditions in profit , either on positive or negative .
these last conditions depend also on the time window on which the non nominal cumulative profit is realized . on a fundamental ground ,
our system architecture is just a refinement of a basic trend - following strategy .
in addition , the system has learned how to play with volatility ranges to trigger decisions and how to protect over - profits realized for example in high volatile periods .
therefore , if we can show that this strategy leads to profitable results ( net of fees ) , it will be a proof of the validity of the trend - following hypothesis on the market , taking into account multivariate tests to activate the trend follower .
let us note that the use of moving averages is a powerful experimental method to access to the non - trivial statistical texture of a time series .
if we consider 2 standard moving averages of lengths @xmath26 and @xmath27 , with @xmath28 , then the density of crossing points of the 2 averages is given by : @xmath29^{h-1}\ ] ] where @xmath30 is the hurst exponent , that characterizes the persistence or anti - persistent of the data series @xcite .
we have @xmath31 parameters optimized on the in - sample ( i ) .
the optimization is performed in order to achieve the best sharpe ratio .
results are shown in fig . [ in ] .
we present the behavior with time of the ec contract itself as well as the cumulative equity of the designed trading system ( expressed in basis points , net of fees ) .
we observe the nice behavior of the equity , increasing with time , which shows that the strategy is profitable and coherent with respect to different market regimes .
the bottom plot in fig .
[ in ] corresponds also to the running of the trading system , but this time on the randomized in - sample .
exactly , we have added to each quote value of the data series ( i ) a random number that ranges between @xmath32 and @xmath33 times the slippage of the ec contract .
and we run the trading system on this series , which leads to the bottom plot of fig .
this randomization is necessary as we do not want the trading system to be dependent on the point - to - point correlation and also the model must be flexible to absorb distortion of the data series .
this is what we observe in fig .
[ in ] ( bottom ) : the system is robust against randomization of the data series .
the degradation observed on the overall profit is not dramatic and the equity is still much reasonable .
note that all systems designed that have failed at this stage have been rejected .
before considering the out - sample stage , we have an intermediate essential step of validation of the trading system . to ensure that the system is robust , we need more that the randomization of the data series . we need to distort the strategy itself in many ways : for example , force an exit of given trade at a given time , do not execute randomly some trades , delay the execution of orders by several quotes , execute an order but at a wrong price , with a prejudice for the trading system , multiply the fees ( slippage ) by a factor 2,3 or 4 etc .
thus , we have a list of @xmath34 stress tests and for each case , the trading system is run and a result is obtained .
all this must be done on the original data series ( i ) and on its randomized version . in all cases
, we must observe that the system is stable and robust .
this is shown in fig .
[ stress ] , where we present the sharpe ratio for all @xmath34 stress tests considered .
we do not provide the equity in each configuration .
we summarize each case by one entry in fig .
[ stress ] , as a value of the sharpe ratio for the case under study .
the idea is that the robustness is ensured if we do not observe pathological values in the sharpe ratios , even for the more extreme stress tests .
this is what we observe in fig .
[ stress ] , with an average value of @xmath35 and a rms of @xmath36 . in all configurations
, the model stays reasonable . by this method
, we have also shown that the trading system does not depend on the fine tuning of any of the fitted parameters .
otherwise , a few stress tests would have failed deeply . on the contrary ,
our strategy depends weakly on any of its inputs , which gives a lot of flexibility on all variables of the system with always a profitable result obtained . at this step ,
it is not unreasonable to claim that we have designed a robust algorithm .
however , a new validation stage is determinant using the out - sample ( ii ) .
this is a decisive test as we are running on new data , that the system does not know , in the sense that parameters have been fitted on another set of data . in principle parameters
are robust as we have already explored many configurations for the data series and the system .
however , the out - sample test will kill all systems that still have some elements of over - fit in their construction . indeed ,
such systems fail to give good results when running on the sample ( ii ) and are rejected .
this is what happens for most of the systems that can be designed if the input ideas are not carrying decisive features of the inside dynamics of the time series .
that s why it is not an easy task and many attempts are needed before converging towards acceptable solutions . in fig .
[ out ] we show the result for the trading system described above .
we observe a correct behavior of the equity , which qualifies definitely this system .
we interpret this as a clear evidence that the time series of the ec exhibits features of trend - following , under certain conditions , as encoded in our trading system . finally , in fig .
[ live ] , we check the result on the live - sample period ( iii ) , in 2010 - 2011 . here
, we do not expect any failure , otherwise the full process described above must be rejected .
effectively , we observe a nice behavior of the cumulative equity ( net of fees ) , much compatible with what has been designed on the in - sample .
this confirms our statement above on the dynamical content of the trading algorithm we have presented here . in order to illustrate very simply the gross feature of the model
, we present two distributions in fig .
we show the trade return spectrum ( fig .
[ dist]-left ) , in which we recognize a typical trend - following system , reminiscent from the standard behavior plotted in fig .
we observe also in fig .
[ dist ] ( right ) that the system is effectively working at high frequency with an average duration of trades of @xmath37 minutes . from the above discussion , we know that our system is robust against a variation of the sampling of the data , for example from 5 minutes to 10 minutes . also , as the average duration of trades is @xmath37 minutes for the nominal system , it makes sense to move the system from 5 minutes to 10 minutes data sampling and check the results . in order to make the change in an optimal way , we have rescaled some parameters such that we have a perfect homotopy between the construction at 5 and 10 minutes sampling .
results on the live sample of data ( iii ) is shown in fig .
[ livefinal ] .
we observe the good behavior , in accordance ( homotopy ) with fig.[live ] , as expected .
a final comment is in order concerning the stress tests and robustness analysis .
this study ensures statistically the reasonable functioning of the system whatever market regimes and trading conditions .
we can see the system as an unfolding procedure , transforming the price series to a trade series .
if we do not control obviously the price series , we have a control on how the trade series develop , regardless the behavior of the price series . that s what we have shown above
this is a decisive element in the construction of the model .
the feed back from the cumulative equity itself is an element of the strategy reconstruction , involved with the basic conditions of the model as described earlier .
let us add that this is a clear advantage we get using this kind of approach for decoding the markets to a certain extend .
the idea is to use the same system , as built in section [ ec ] , on the dax future ( fdax ) series .
if we use exactly the same system as defined for euro future , we obtain a sharpe ratio of @xmath38 over 10 years of data .
this is reasonable but it makes sense to re - optimize some parameters on the fdax series . in the updating process , some conditions are also re - examined and modified according to what data ( in - sample ) requires .
we use again a five minutes sampling for the fdax time series , from january 1999 till august 2011 .
this which makes 460k quotes of data . we conserve only the close of each quote . as in section [ ec ] , this large sample of data points is necessary to infer statistical properties with a high confidence level .
a typical quote of the fdax is like @xmath39 .
the unit of the last digit is what we call a basis point .
for example , we consider that a price movement from @xmath39 to @xmath40 corresponds to a price change of @xmath41 basis points .
more precisely , if we buy the contract at time t1 ( on the quote q1 ) at @xmath39 and sell this contract at time t2@xmath17t1 ( on the quote q2 ) at @xmath40 , then this trade corresponds to a gain of @xmath41 basis points ( without fees ) . to keep the procedure as close to reality as possible we consider fees of two times the slippage , which means that this trade is counted in our approach as a trade of @xmath42 basis point ( net of fees ) .
we follow the analysis process detailed previously .
then , 3 samples of data are defined : * 1999 - 2006 : in - sample * 2007 - 2009 : out - sample * 2010 - 2011 : live - sample as we start the analysis of the fdax in 1999 , we end up the in - sample in 2006 and not 2007 as was done for the ec time series .
as mentioned above , the basic elements of the algorithm design are the same as the ones used for the ec time series in section [ ec ] .
the main differences concern the treatment of volatility ranges .
there is a stronger focus on this issue for the fdax model .
we have always @xmath31 parameters optimized on the in - sample ( i ) , the system is then validated on ( ii ) and observed on ( iii ) . the optimization is performed in order to achieve the best sharpe ratio on ( i ) .
results are summarized in fig .
we present the behavior with time of the fdax contract itself as well as the cumulative equity of the designed trading system ( expressed in basis points , net of fees ) for the three data samples defined above .
we observe the nice behavior of the equity , increasing with time , which shows that the strategy is profitable and coherent with respect to different market regimes .
we observe also that the out - sample ( ii ) validates the good behavior of the strategy , which is confirmed on the recent period 2010 - 2011 .
as explained in section [ ec ] , we have also guaranteed the robustness of the algorithm using a battery of stress tests . in all cases
, we must observe that the system is stable and robust .
this is shown in fig .
[ stress2 ] , where we present the sharpe ratio for all stress tests considered . as in section [ ec ]
, the idea is that the robustness is ensured if we do not observe pathological values in the sharpe ratios , even for the more extreme stress tests .
we interpret the good result obtained in fig .
[ in2 ] as a clear evidence that the time series of the fdax exhibits features of trend - following , under certain other conditions , as encoded in our trading system and as already observed on the ec time series .
this will be confirmed again on the cacao future in the next section . finally , in fig .
[ live2 ] , we illustrate explicitly the result on the live - sample period ( iii ) , in 2010 - 2011 .
effectively , we identify a nice behavior of the cumulative equity ( net of fees ) , much compatible with what has been designed on the in - sample .
this confirms our statement above on the dynamical content of the trading algorithm we have presented here . in order to illustrate very simply the gross feature of the model
, we present two distributions in fig .
[ dist2 ] .
we show the trade return spectrum ( fig .
[ dist2]-left ) , in which we recognize a typical trend - following shape .
we observe also in fig .
[ dist2 ] ( right ) that the system is effectively working at high frequency with an average duration of trades of less that @xmath43 minutes .
for the cacao future ( cc ) , we have at our disposal the data series ranging only from 2003 till mid-2010 .
then , we have chosen to run exactly the ec system on this index , with fees for each trade always equal to 2 times the slippage .
we do not produce any further optimization .
results are presented in fig .
[ cacao ] .
we observe a reasonable equity curve , which proves that the system is functioning correctly on this index , uncorrelated to ec and fdax .
this confirms the message of this article that the trading system defined above contains elements of invariants of financial markets .
tools of statistical physics @xcite have been proven to be efficient in many scientific areas . in a similar way for financial time series , knowing that the difference between real markets and random walks is very small , a modern statistical multivariate analysis can help to extract this difference .
this is what is encoded in trading systems .
we have shown how to achieve the construction of such a system on the euro future contract at high frequency .
a typical element of the dynamics of this system is then accessible , namely the trend - following content involved in a more complex architecture on volatilities .
then , we have produced other examples on completely different markets , largely uncorrelated to the euro future , the dax and cacao future contracts .
the same procedure is followed using similar seed ideas and technical inputs .
we have shown that similar results can then be obtained and we conclude that this is an evidence that some invariants , as encoded in our system , have been identified , on very different markets explored over a 10 years period of time . one essential point in our process is that trading models , like the one used in our approach , are highly sensitive to non - linear relations in price series .
this comes with the multivariate data analysis . in this article
, we have also focused the discussion on the necessity of a deep robustness analysis to ensure the validity of the overall construction .
an immediate question can be raised concerning the rationale behind this content .
our observation is universal in the sense that the same algorithm , for example on ec , is running on more that 10 years of data , where the monetary policy has changed several times .
then , our approach is not attached to a particular regime of interest rates .
there are certainly herding behaviors at the origin of the values of parameters encoded in our system
. these herding phases may appear with strengths governed by certain fear levels , corresponding to volatility domains .
also , in some circumstances , nothing special can be said .
finally , a global rationale explanation of a given trading system is very complex and probably not unique .
this is beyond the scope of this article .
see ideas in @xcite . in this article
, we have completed a pure experimental analysis .
the concept of invariance comes with the observation that we extract similar seed features from largely uncorrelated financial time series .
this is a first step , rooted exclusively on data . c. tsallis , braz .
j. phys .
* 29 * ( 1999 ) 1 .
d. prato and c. tsallis , phys .
e * 60 * ( 1999 ) 2398 .
a. rapisarda , a. pluchino and c. tsallis , arxiv : cond - mat/0601409 .
c. amsler _ et al .
_ [ particle data group ] , _ phys .
* b667 * ( 2008 ) 1 ; available at pdg.lbl.gov . | for the pedestrian observer , financial markets look completely random with erratic and uncontrollable behavior . to a large extend ,
this is correct . at first approximation
the difference between real price changes and the random walk model is too small to be detected using traditional time series analysis .
however , we show in the following that this difference between real financial time series and random walks , as small as it is , is detectable using modern statistical multivariate analysis , with several triggers encoded in trading systems .
this kind of analysis are based on methods widely used in nuclear physics , with large samples of data and advanced statistical inference .
considering the movements of the euro future contract at high frequency , we show that a part of the non - random content of this series can be inferred , namely the trend - following content depending on volatility ranges .
of course , this is not a general proof of statistical inference , as we focus on one particular example and the generality of the process can not be claimed .
therefore , we produce other examples on a completely different markets , largely uncorrelated to the euro future , namely the dax and cacao future contracts .
the same procedure is followed using a trading system , based on the same ingredients .
we show that similar results can be obtained and we conclude that this is an evidence that some invariants , as encoded in our system , have been identified .
they provide a kind of quantification of the non - random content of the financial markets explored over a 10 years period of time .
# 1#2#3#4#1 * # 2 * ( # 3 ) # 4 * about the non - random content of financial markets * laurent schoeffel + + cea saclay , irfu / spp , 91191 gif / yvette cedex , + france |
exactly solvable models in physics have been generating a renewed interest , because of the fact that they can be solved in terms of creation and annihilation operators by means of factorization method [ 1 - 5 ] . in the year 2002 , this present method was proposed in the two related works : the ladder operators for the modified pschl - teller potential and the morse potential were obtained @xcite . with factorization method , the ladder operators of a quantum mechanical system with some important potentials like morse potential , pschl - teller potential , the pseudoharmonic potential , the infinitely square - well potential and other quantum systems have been established @xcite . for comprehensive review ( see dong and references therein @xcite ) .
algebraic methods underlying lie symmetry and its associated algebra have been widely used to study many of these exactly solvable potentials . in 2007 , rasinariu et al .
@xcite gave a review of the progress made so far in solving exactly solvable problems in quantum mechanics , by connecting supersymmetry and spectrum generating algebras through the property of shape invariance .
the solutions of exactly solvable models can be achieved by dynamical algebraic approaches [ 1 - 11 ] . in 1991 ,
de lange and raab @xcite presented operator methods with shift operators ( that is , raising and lowering operators ) for the hamiltonian of exactly solvable models . for the factorization method , we are going to adopt dong s approach @xcite by finding the ladder operators @xmath10 with the following properties : @xmath11 in this case , we are seeking for the ladder operators of the form @xmath12 these ladder operators depend on the physical variable @xmath13 , which is different for different quantum systems . with equation([m1 ] ) , the following expressions are obtained : @xmath14 from which we find that the products of the operators @xmath15 and @xmath16 acting on the given wavefunctions @xmath17 and @xmath18 , respectively , have the same constant @xmath19 the @xmath0 algebra has useful applications in physics @xcite .
the algebra of the group @xmath0 is used to generate the energy spectra while the representation matrices of the group could be used to calculate time dependent excitations of the bound states and the scattering states respectively . for diatomic molecules , the generalised kratzer potentials is considered as one of the molecular potentials .
these potentials have the general features of the true interaction energy , interatomic , inter - molecular and dynamical properties , their wavefunctions do vanish at the origin like morse potential .
this added advantage make these potentials important in molecular physics , chemical physics , solid state physics .
these potentials ( kratzer and modified kratzer molecular potentials ) that is being considered in the present work are important molecular potentials which describe the interaction between two atoms .
these potentials have been used extensively to describe the molecular structure and interactions and have been receiving much attention in the history of quantum chemistry for some decades ago @xcite .
it is the purpose of this paper to study the dynamical symmetries of the kratzer - type molecular potentials ( generalized kratzer molecular potentials ) and to establish that the generators obtained are an @xmath0 dynamical group , and to obtain some numerical results for the energy eigenvalues and matrix elements for some diatomic molecules .
the paper is organized as follows . in section @xmath20
, we study the exact solutions of the kratzer - type molecular potentials .
section @xmath21 contained the construction of the creation and annihilation operators , the commutation relations of some of these operators and the matrix elements of some related functions @xmath1 and @xmath2 .
also , in section @xmath22 , the numerical calculations of the energy levels and matrix elements for some diatomic molecules for the kratzer and modified kratzer potentials are given .
we conclude with section @xmath23 .
consider the molecular potential ( the generalized kratzer potential ) of the form @xcite @xmath24 this potential is of great interest because , it is a generalised form of the molecular potential which can be used to generate other forms of the kratzer - type molecular potentials : * standard morse or kratzer - fues potential @xmath25 where @xmath26 is the dissociation energy between two atoms in a solid and @xmath27 is the equilibrium intermolecular separation . on comparing with equation ( 5 ) , we have @xmath28 and this potential has minimum value to be @xmath29 at @xmath30 [ 2 , 12 - 14 , 17 - 23 ] .
* modified kratzer potential @xmath31 where , @xmath32 , @xmath33 and @xmath34 .
modified kratzer potential is kratzer potential shifted by the amount @xmath26 @xcite .
the graphs of the two potentials ( equations @xmath35 and @xmath36 ) are shown below in * figure 1 . * * figure 1 .
* shapes of the kratzer and modified kratzer potentials for @xmath8 and @xmath9 diatomic molecules . consider the motion of a particle in a spherically symmetric potential @xmath37 @xmath38\psi(r,\theta , \phi ) \label{m8}\ ] ] where the potential @xmath37 is taken as in equation ( [ m5 ] ) .
we seek for the wavefunctions of the form @xmath39 which reduces equation ( [ m8 ] ) into the radial and angular wave functions as : @xmath40 - \frac{\ell~(\ell + 1)}{r^{2 } } \right\}r_{n , \ell}~(r)= 0 \label{m10}\ ] ] and @xmath41 where @xmath42 and @xmath43 are the radial and angular solutions of equations ( [ m10 ] and [ m11 ] ) respectively . for @xmath44
to be finite everywhere , @xmath45 must vanish at @xmath46 , that is , @xmath47 , then @xmath45 is a real function . for the bound state energy eigenvalues for this quantum system ,
the following dimensionless abbreviations are introduced : @xmath48^{\frac{1}{2}}r;~~\alpha = \left ( \frac{- \mu}{2 \hbar^{2}(e - c ) } \right)^{\frac{1}{2}}a;~~\beta_{\ell}~(\beta_{\ell } + 1 ) = \frac{2 \mu b}{\hbar^{2 } } + ~ \ell~(\ell + 1 ) \label{m12}.\ ] ] the above equations ( [ m12]a ) and ( [ m12]b ) are chosen in order that the acceptable bounds state solutions are obtained , this can only be possible if @xmath49 ( otherwise , continuum state solutions will be obtained ) .
@xmath50 in equation ( [ m12]c ) represents the usual centrifugal term , that is , @xmath51 , where @xmath7 is the angular momentum quantum number .
these substitutions allow us to obtain the following hypergeometric - type equation .
@xmath52r_{n , \ell}~(\rho ) = 0
. \label{m13}\ ] ] this differential equation has an irregular singularity as @xmath53 , where its normalized solutions in bound states behave like @xmath54 .
it further has a singularity at @xmath55 , where @xmath56 .
then , the ansatz for the wave functions which is a physically acceptable solution for @xmath57 can be expressed in the form @xmath58 and therefore , equation ( [ m13 ] ) becomes @xmath59 \frac{d g(\rho ) } { d \rho } - [ \beta_{\ell } + 1 - \alpha ] g ( \rho ) = 0 . \label{m15}\ ] ] this is the associated laguerre differential equation ( kummer equation)@xcite , the solution of equation ( [ m15 ] ) which is regular at origin ( regular at @xmath46 or @xmath60 ) is the degenerate hypergeometric function @xmath61 for large values of @xmath62 , this solution diverges as @xmath63 , thus preventing normalization , except for when @xmath64 becomes a polynomial .
therefore , the solution for the radial equation for this generalized kratzer - type molecular potential is @xmath65 where the normalization @xmath66 is determined from the requirement that @xmath67 by using the expression that relates the associated laguerre functions with the confluent hypergeometric functions @xmath68 together with the following important formular @xcite @xmath69 and by substituting @xmath70 , the normalized radial wave function is obtained as @xmath71 where @xmath72^{\frac{1}{2}}. \label{m23}\ ] ] the corresponding eigenvalues are : @xmath73^{2 } } + c , \label{m24}\ ] ] where @xmath50 is obtained as the positive root of equation ( [ m12]c ) given as @xmath74 .
\label{m25}\ ] ]
the ladder operators can be generated directly from the eigenfunction with the factorization method as shown in dong @xcite and references therein .
we shall find the diferential operators @xmath75 with the following property : @xmath76 the operators of the form @xmath77 which depend only on the physical variable @xmath1 are to be obtained .
the action of the differential operator @xmath78 on wave functions ( [ m22 ] ) gives : @xmath79 the expression above is used to construct the ladder operators @xmath75 by using the recurrence relations of the associated laguerre functions in order to find the relation between @xmath80 and @xmath81 .
to find these , the following recurrence relations of the associated laguerre functions are used @xcite : @xmath82 and the creation and annihilation operators are obtained as : @xmath83 where @xmath84 is the number operator with the property @xmath85 the action of the creation and annihilation operators on the radial wavefunctions @xmath86 gives the following properties @xmath87 where @xmath88 on studying the dynamical group associated to the annihilation and creation operators @xmath89 and @xmath90 and based on the results of equations ( [ m32 ] ) and ( [ m33 ] ) , we can evaluate the commutator @xmath91 $ ] as : + @xmath92r_{n , \ell}~(r ) % = ( 2n_{r } + 2\beta_{\ell } + 2 ) r_{n_{r } , \ell}(r ) = 2 \ell_{0}r_{n , \ell}~(r ) , \label{m34 } % \end{array}\ ] ] where @xmath93 and @xmath94 is defined as @xmath95 thus , operators @xmath96 and @xmath94 satisfy the following commutation relations : @xmath97r_{n , \ell}~(r ) = \mp\hat{\cal{l}}_{\mp } r_{n \mp 1 , \ell}~(r)}. \label{m37}\ ] ] the action of @xmath90 on the radial wavefunction @xmath86 gives @xmath98 if @xmath99 is non - zero , then , it is an eigenfunction of @xmath94 with eigenvalue @xmath100 .
thus , the effect of @xmath90 is to raise the eigenvalue by one unit .
similarly , the action of @xmath89 on the radial wavefunction @xmath86 gives @xmath101 and if @xmath99 is non - zero , then , it is an eigenfunction of @xmath94 with eigenvalue @xmath102 ( this is the reason for calling @xmath90 and @xmath89 raising and lowering operators respectively ) . for the hermitian operators , we define the operators as follows : @xmath103 we obtained the following commutation relations @xmath104 = - i \hat{\cal{l}}_{z}},~~ \displaystyle{[\hat{\cal{l}}_{y},~\hat{\cal{l}}_{z } ]
= i \hat{\cal{l}}_{x}},~~ \displaystyle{[\hat{\cal{l}}_{z},~\hat{\cal{l}}_{x } ] = i \hat{\cal{l}}_{y}}. \label{m41 } % \end{array}\ ] ] the casimir operator @xcite can be expressed as @xmath105 the casimir operator @xmath106 now satisfies @xmath107 = [ \hat{\cal{c } } , \hat{\cal{l}}_{x } ] = [ \hat{\cal{c } } , \hat{\cal{l}}_{y } ] = [ \hat{\cal{c } } , \hat{\cal{l}}_{z } ] = 0 , \label{m43 } % \end{array}\ ] ] the operators @xmath108 , @xmath109 , @xmath110 , @xmath111 and @xmath112 satisfy the commutation relations of the dynamical group @xmath113 algebra , which is isomorphic to an @xmath114 algebra ( i. e. @xmath113 @xmath115 @xmath114 .
the commutation rules are valid for the infinitesimal operators of the non - compact group @xmath0 @xcite .
these relations coincide with the formulas that define the action of the infinitesimal operators @xmath108 and @xmath112 of the dynamical group @xmath0 on a basis @xmath116 of the irreducible representation @xmath117 belonging to the discrete positive series in an abstract hilbert space @xcite .
the eigenvalues have the ground state and therefore , the representation of the dynamical group @xmath0 belongs to @xmath117 , using the dirac notation @xmath118 : @xmath119^{\frac{1}{2 } } \mid j , k \pm 1 \rangle } \\
\nonumber & k= - j + n , ~~n= 1 , 2 , ~~\ldots , j<0 . \label{m44 } \end{array}\ ] ] furthermore , the following expressions can be easily obtained from the operators @xmath96 and @xmath94 as follows : @xmath120 } \\
\nonumber & \displaystyle{r \frac{d}{dr}= \frac{1}{2}(\hat{\cal{l}}_{+ } - \hat{\cal{l}}_{- } ) - 1}. \label{m45 } \end{array}\ ] ] with these , the matrix elements for @xmath1 and @xmath2 are obtained as follows : @xmath121 } \label{m46 } % \end{array}\ ] ] and @xmath122 from equations ( [ m46 ] ) and ( [ m47 ] ) , we can deduce the following relations : @xmath123 and @xmath124 these relations form a useful link for finding the matrix elements from ladder operators .
the energy eigenvalues for the kratzer and modified kratzer potentials are obtained respectively , as : @xmath125^{2 } } \label{m50}\ ] ] and @xmath126^{2 } } + d_{0 } , \label{m51}\ ] ] where @xmath127 .
\label{m52}\ ] ] .reduced masses and spectroscopic properties of the @xmath8 and @xmath9 diatomic molecules in the ground electronic state .
the data listed in this table are taken from @xcite . [
cols="^,^,^ " , ] [ tab3 ] the matrix elements for @xmath1 and @xmath2 are given as : @xmath128 } \label{m53 } \end{array}\ ] ] and @xmath129 where @xmath130 with equations ( [ m53 ] ) and ( [ m54 ] ) , we can deduce the following relations : @xmath131 and @xmath132 in this case , the matrix elements for the two potentials ( kratzer and modified kratzer potentials ) give the same results , since these matrix elements depend on @xmath133 and @xmath134 only , and not on @xmath135 .
hence , when the kratzer potential is shifted by the amount @xmath26 , ( that is , the modified krazter potential ) , it has no effect on the matrix elements . for the numerical results , see tables @xmath22 and @xmath23 .
@xmath6 & @xmath7 & @xmath136 & @xmath137 & @xmath138 & @xmath4 + @xmath139 & @xmath140 & -0.4761490924054464 & 16.97323506451677 & 195.3895273866551 & 161.4430572576216 + @xmath139 & @xmath139 & -0.4761710273791221 & 16.97343118040934 & 195.3940309777938&161.4471686169751 + @xmath20 & @xmath140 & -0.4107558210705384 & 9.473123928272610 & 162.6714457007481&143.7251978442029 + @xmath20 & @xmath139 & -0.4107755614421622 & 9.473236107448747 & 162.6755771602085&143.7291049453111 + @xmath20 & @xmath20 & -0.4108150436563324 & 9.473460462197114 & 162.6838400978704&143.7369191734761 + @xmath21 & @xmath140 & -0.3789129917968390 & 7.666855095230551 & 148.3356842660929&133.0019740756318 + @xmath21 & @xmath139 & -0.3789316534237137 & 7.666946649417545 & 148.3396482090482&133.0057549102131 + @xmath21 & @xmath20 & -0.3789689781099853 & 7.667129754891770 & 148.3475761191884&133.0133166094049 + @xmath21 & @xmath21 & -0.3790249687207100 & 7.667404405854176 & 148.3594680449690&133.0246592332607 + @xmath22 & @xmath140 & -0.3533372017343955 & 6.615793772861618 & 137.1863080576614&123.9547205119381 + @xmath22 & @xmath139 & -0.3533549972979473 & 6.615873160774083 & 137.1901403213880&123.9583939998398 + @xmath22 & @xmath20 & -0.3533905898264022 & 6.616031934119029 & 137.1978048773512&123.9657410091131 + @xmath22 & @xmath21 & -0.3534439821224761 & 6.616270087936851 & 137.2093017825661&123.9767616066924 + @xmath22 & @xmath22 & -0.3535151205980674 & 6.616587357061658 & 137.2246186798360&123.9914439657127 + @xmath23 & @xmath140 & -0.3312181486228352 & 5.899103372782779 & 127.7353299114129&115.9371231658474 + @xmath23 & @xmath139 & -0.3312351970446675 & 5.899174359294310 & 127.7390497423004&115.9407010237118 + @xmath23 & @xmath20 & -0.3312692952626132 & 5.899316330129842 & 127.7464894362121&115.9478567759524 + @xmath23 & @xmath21 & -0.3313204460252490 & 5.899529280914656 & 127.7576490574162&115.9585904955868 + @xmath23 & @xmath22 & -0.3313885980893305 & 5.899812974634305 & 127.7725166245981&115.9728906753295 + @xmath23 & @xmath23 & -0.3314738123034079 & 5.900167633018892 & 127.7911043436078&115.9907690775700 + [ tab4 ] @xmath6 & @xmath7 & @xmath136 & @xmath141 & @xmath138 & @xmath4 + @xmath139 & @xmath140 & -0.4819553974463722 & 16.21611432627930 & 178.3853413621973 & 145.9531127096387 + @xmath139 & @xmath139 & -0.4819818755489971 & 16.21633768030118 & 178.3902449613004 & 145.9575696006980 + @xmath20 & @xmath140 & -0.4124035377598499 & 9.040224937878095 & 147.1042625831576 & 129.0238127074014 + @xmath20 & @xmath139 & -0.4124272376527343 & 9.040352590403065 & 147.1087421492765 & 129.0280369684704 + @xmath20 & @xmath20 & -0.4124746395526131 & 9.040607890585319 & 147.1177013069172 & 129.0364855257466 + @xmath21 & @xmath140 & -0.3785773170199991 & 7.313655510582683 & 133.4152241892145 & 118.7879131680491 + @xmath21 & @xmath139 & -0.3785996534470747 & 7.313759628367830 & 133.4195129653269 & 118.7919937085912 + @xmath21 & @xmath20 & -0.3786443283574571 & 7.313967860029029 & 133.4280905503876 & 118.8001548303296 + @xmath21 & @xmath21 & -0.3787112793749515 & 7.314279887895779 & 133.4409442456685 & 118.8123844698770 + @xmath22 & @xmath140 & -0.3514058390444385 & 6.309618628486895 & 122.7741267465619 & 110.1548894895881 + @xmath22 & @xmath139 & -0.3514270813879861 & 6.309708856787779 & 122.7782656025786 & 110.1588478890030 + @xmath22 & @xmath20 & -0.3514695680843574 & 6.309889310052519 & 122.7865433532380 & 110.1667647331330 + @xmath22 & @xmath21 & -0.3515332399199089 & 6.310159713089281 & 122.7989477575504 & 110.1786283313719 + @xmath22 & @xmath22 & -0.3516182293790348 & 6.310520592939504 & 122.8155035676796 & 110.1944623818006 + @xmath23 & @xmath140 & -0.3278978812775772 & 5.625407532078025 & 113.7570644421939 & 102.5062493780379 + @xmath23 & @xmath139 & -0.3279181800627655 & 5.625488165029850 & 113.7610753342060 & 102.5100990041463 + @xmath23 & @xmath20 & -0.3279587796016631 & 5.625649427995388 & 113.7690971617573 & 102.5177983057665 + @xmath23 & @xmath21 & -0.3280196234075503 & 5.625891075135918 & 113.7811180744773 & 102.5293359242055 + @xmath23 & @xmath22 & -0.3281008382258643 & 5.626213577576859 & 113.7971620775729 & 102.5447349224192 + @xmath23 & @xmath23 & -0.3282023110716663 & 5.626616443662535 & 113.8172054699764 & 102.5639725826513 + [ tab5 ]
in this paper , i have studied the eigenvalues , the eigenfunctions and the matrix elements of the kratzer - type molecular potentials ( generalized kratzer molecular potentials ) .
the ladder ( creation and annihilation ) operators for the radial wavefunctions are established .
also , the hermitian operators of these ladder operators are obtained .
these operators satisfy the commutation relations of an @xmath0 dynamical group , the action of @xmath142 on the wavefunctions that reveals the fact about the raising and lowering effect are established .
the matrix elements of the different functions @xmath1 and @xmath143 are also obtained from the ladder operators in a closed form .
the solutions of the generalized kratzer molecular potentials are obtained via an @xmath0 algebraic approach and the results can be generalized to other form of potentials in equation ( [ m5 ] ) .
the generalized kratzer potential model proposed in this work @xcite , allows one to obtain the eigenvalues , the matrix elements and the radial eigenfunctions for the two important molecular potentials simultaneously ( that is , the kratzer and modified kratzer potentials ) .
this generalization covers the descriptions about the two potentials , the shapes of the two potentials for @xmath8 and @xmath9 diatomic molecules are also shown in figure @xmath139 .
more importantly , the generalized kratzer potential model used in this work generates results for the two molecular potentials .
in addition , the numerical results are obtained for the eigenvalues for @xmath8 and @xmath9 diatomic molecules for the two molecular potentials .
the results obtained are compared with other existing results ( eqr , aim and nu ) , and the results agree favourably with other results , see tables @xmath20 and @xmath21 .
tables @xmath22 and @xmath23 show the matrix elements of @xmath1 , @xmath2 , their sum ( @xmath3 ) and their difference ( @xmath4 ) obtained for @xmath8 and @xmath9 diatomic molecules , the results obtained are the same results for the two molecular potentials .
this is because , the matrix elements depend on @xmath133 and @xmath134 only , and not on @xmath135 . hence , the shifted amount @xmath26 , ( in the case of the modified krazter potential ) has no effect on the matrix elements .
the advantage of the present approach is that it enables one to find the energy eigenvalues , eigenfunctions and matrix elements in a simple way .
the approach presented in this study is efficient and is a very useful link for finding the matrix elements from ladder operators .
this approach can be used to find the energy eigenvalues , eigenfunctions and radial matrix elements of the schrdinger equation with a given exactly solvable molecular potentials for various diatomic molecules for any values of @xmath6 and @xmath7 quantum numbers .
the ladder operators constructed in this study , are very useful tools in quantum - mechanical calculations of the various matrix elements based on the kratzer molecular basis function .
furthermore , these operators can be used in constructing coherent states .
i am very grateful to the ictp for my postdoctoral position at the icmpa - unesco chair under the prj -@xmath144 ictp project , this work is a product of the motivations i received from prof .
m. n. hounkonnou ( the chair , icmpa , universite dabomey - calavi , benin ) and other icmpa - unesco chair visiting professors .
i appreciate the efforts of profs .
dong , s. h. ; sen , k. d. ; l@xmath145pez - bonila , j. l. and emeritus prof . k. t. hecht for sending valuable materials to me .
i thank the anonymous referees for the careful reading of the manuscript and the suggestions that improved the paper .
99 schrdinger , e. ( 1940 ) : proc .
r. irish acad .
a * 46 * , 183 ; + schrdinger , e. ( 1940 ) : proc .
r. irish acad .
a * 47 * , 53 ; + infeld , l. and hull , t. e. ( 1951 ) : rev .
* 23 * , 21 .
dong , s. h. ( 2007 ) : _ factorization method in quantum mechanics ( fundamental theories of physics ) _ , vol .
150 , springer , netherlands .
dong , s. h. , lemus , r. and frank , a. ( 2002 ) : int .
j. quantum chem . * 86 * , 433 ; + dong , s. h. and lemus , r. ( 2002 ) : int .
j. quantum chem . * 86 * , 265 .
dong , s. h. and ma , z. q. ( 2002 ) : am . j. phys . * 70 * , 520 ; + dong , s. h. and ma , z. q. ( 2002 ) : int . j. mod .
e * 11 * , 155 ; + dong , s. h. and dong , s. s. ( 2003 ) : int .
e * 12 * , 809 ; + dong , s. s. and dong , s. h. ( 2002 ) : czech .
* 52 * , 753 ; + dong , s. h. ( 2002 ) : can . j. phys .
* 80 * , 129 ; + dong , s. h. ( 2002 ) : z. phys
. chem . * 216 * , 103 ; + dong , s. h. ( 2003 ) : appl .
* 16 * ( 2 ) , 199 ; + dong , s. h. ( 2004 ) : comput . math .
appl . * 47 * ( 6 - 7 ) , 1035 .
rasinariu , c. , mallow , j. v. and gangopadhyaya , a. ( 2007 ) : cent .
* 5*(2 ) , 111 . de lange , o. l. and raab , r. e. ( 1991 ) : _ operator methods in quantum mechanics _
, clarendon press , oxford .
hecht , k. t. ( 2000 ) : _ quantum mechanics _ ( graduate texts in contemporary physics ) , springer - verlag , new york .
lorente , m. ( 2001 ) : j. phys .
a : math . gen .
* 34 * , 569 .
sodoga , k. , hounkonnou , m. n. and debiais , g. ( 2005 ) : proceedings of the fourth international workshop on contemporary problems in mathematical physics , cotonou , benin , eds .
j. govaerts , m. n. hounkonnou and msezane , a. z. ( world scientific , singapore , 2005 ) , 315 .
barut , a. o. and raczka , r. ( 2000 ) : _ theory of group representations and applications _ , second edition , world scientific , singapore .
adams , b. g. ( 1994 ) : _ algebraic approach to simple quantum systems ( with applications to perturbation theory ) _ , springer - verlag , new york .
bayrak , o. , boztosun , i. and iftci , i. ( 2007 ) : int .
. chem . * 107 * , 540 .
patil , s. h. and sen , k. d. ( 2007 ) : phys .
a * 362 * , 109 .
flgge , s. ( 1994 ) : _ practical quantum mechanics _ , springer - verlag , berlin , 178 .
berkdemir , c. , berkdemir , a and han , j. ( 2006 ) : chem .
lett . * 417 * , 326 .
oyewumi , k. j. , akinpelu , f. o. and agboola , a. d. ( 2008 ) : int .
phys . * 47 * , 1039 .
oyewumi , k. j. ( 2008 ) : `` exactly analytical solutions of the schrdinger equation for some molecular potentials in @xmath146-dimensions '' .
proceedings of the fifth international workshop on contemporary problems in mathematical physics copromaph 5 held on the @xmath147 nov .
2007 under the auspices of the international chair in mathematical physics and applications icmpa , university of abomey - calavi , cotonou , benin republic .
eds : profs .
j. govaerts and m. n. hounkonnu , 193 .
ikhdair , s. and sever , r. ( 2007 ) : `` on the solutions of the schrdinger equation with some molecular potentials : wave function ansatz '' @xmath148 .
ikhdair , s. and sever , r. ( 2007 ) : cent .
* 5(4 ) * , 516 .
ikhdair , s. and sever , r. ( 2008 ) : `` exact quantization rule to the kratzer - type potentials : an application to the diatomic molecules '' @xmath149 .
oyewumi , k. j. ( 2005 ) : foundations phys . lett .
* 18 * , 75 .
qiang , w. c. ( 2003 ) : chin .
* 12 * , 1054 .
kratzer , a.(1920 ) : z. physik . * 3 * , 289 ; + fues , e. ( 1926 ) : ann . physik . * 80 * , 367 .
gradshteyn , i. s. and ryzhik , i. m. ( 1994 ) : _ tables of integrals , series , and products _ , fifth ed .
, academic press , new york .
casimir , h. ( 1931 ): proceedings roy . acad . * 34 * , 844 .
nikiforov , a. f. , suslov , s. k. and uvarov , v. b. ( 1991 ) : _ classical orthogonal polynomials of a discrete variable _ , springer - verlag , berlin .
karplus , m. and porter , r. n. ( 1970 ) : _ atoms and molecules _ , wa benjamin : menlo park , ca .
brewer , l. and rosenblatt , g. m. ( 1969 ) : adv .
high temp .
* 2 * , 1 . | the dynamical symmetries of the kratzer - type molecular potentials ( generalized kratzer molecular potentials ) are studied by using the factorization method .
the creation and annihilation ( ladder ) operators for the radial eigenfunctions satisfying quantum dynamical algebra @xmath0 are established .
factorization method is a very simple method of calculating the matrix elements from these ladder operators .
the matrix elements of different functions of @xmath1 , @xmath2 , their sum @xmath3 and difference @xmath4 are evaluated in a closed form .
the exact bound state energy eigenvalues @xmath5 and matrix elements of @xmath1 , @xmath2 , their sum @xmath3 and difference @xmath4 are calculated for various values of @xmath6 and @xmath7 quantum numbers for @xmath8 and @xmath9 diatomic molecules for the two potentials .
the results obtained are in very good agreement with those obtained by other methods .
* key words * : schrdinger equation , exact solutions , wavefunction , ansatz , ladder operators , su(1 , 1 ) , kratzer potential , ( modified ) kratzer potential , diatomic molecules .
= 0.28 in : 03.65.fd , 03.65.ge , 03.65.ca , 03.65-w |
let @xmath2 be a polytope in @xmath3 .
euler - maclaurin formulas are expressions which may be used to approximate riemann sums of smooth functions @xmath4 such as @xmath5 in terms of integrals involving @xmath0 .
such formulas may be traced back to the work of l. euler and c. maclaurin in the first half of the xix century . partly due to its connections to other areas of mathematics ,
this topic has attracted great interest recently , see for instance berline - vergne @xcite , brion - vergne @xcite , cappell - shaneson @xcite , guillemin - sternberg @xcite , guillemin - sternberg - weitsman @xcite , guillemin - stroock @xcite , guillemin - wang @xcite , karshon - sternberg - weitsman @xcite , shaneson @xcite , tate @xcite , and vergne @xcite .
let us also mention that in ( * ? ? ?
* section 5 , theorem 5.9 ) charles and v ngoc obtained an asymptotic expansion for a sum over integral points of a convex polytope as a consequence of a trace formula , and computed explicitly the first term of this expansion .
our interest in euler - maclaurin formulas originates mainly in the applications they have in spectral theory of differential operators , see for instance the recent articles of burns - guillemin - wang @xcite and zelditch @xcite .
euler - maclaurin formulas are also connected to major problems in number theory , see eg .
lagarias @xcite . at least in dimension one , these formulas have many other applications in other areas , for instance in numerical analysis and computational mathematics , see e.g. berrut @xcite and tausch @xcite , in actuarial sciences , for instance simple approximations in terms of annuities with annual payments are available based on euler - maclaurin formulas ( see @xcite and the references therein ) , and in the study of the casimir effect problem , see for instance @xcite . in @xcite , costin and garoufalidis describe applications of euler - maclaurin formulas to linear difference equations with a small parameter , and to the study of knotted 3-dimensional objects .
recently , tate @xcite obtained concrete euler - maclaurin formulas to approximate riemann sums of the form ( [ eq : sum ] ) for smooth functions on lattice polytopes in any dimension , building primarily on the work of berline and vergne @xcite .
as a consequence of this general result and its rather involved proof , tate obtained a fully explicit formula for riemann sums over delzant polygons ( figure [ fig : polytopes ] ) , that is delzant polytopes in dimension 2 .
this case is already of interest because such delzant polygons arise as images of the moment maps of 4-dimensional toric manifolds .
let us briefly explain how toric manifolds enter the picture .
let @xmath6 be a compact connected symplectic toric manifold , @xmath7 the underlying moment map , and @xmath8 the associated delzant polytope .
assume that there exists a prequantum line bundle @xmath9 , that is a holomorphic , hermitian line bundle which chern connection has curvature @xmath10 , and consider the space @xmath11 of holomorphic sections of @xmath12 ; @xmath13 is the hilbert space of geometric quantization .
a standard result states that the dimension of @xmath13 is equal to the number @xmath14 of integer points in the polytope @xmath15 ( see @xcite for a nice exposition in the case @xmath16 . ] ) , that is the so - called ehrhart polynomial @xcite of this polytope .
an analogous result for toric varieties was one of the motivations behind works by khovanski @xcite , followed by khovanski and pukhlikov @xcite ( see also kantor and khovanski @xcite ) , providing formulas for sums of values of @xmath0 on lattice points of a lattice polytope , where @xmath0 is of the form polynomial times exponential . + more generally , let @xmath17 be commuting self - adjoint berezin - toeplitz operators ( each @xmath18 acts on the space @xmath13 defined above ; for more details , see for instance @xcite and references therein ) whose principal symbols are the components of the moment map @xmath7 .
a result recently obtained by charles , pelayo and v ngoc ( * ? ? ?
* theorem @xmath19 ) states that the joint spectrum @xmath20 ( consisting of joint eigenvalues ) of @xmath21 coincides , up to @xmath22 , with the set @xmath23 where @xmath24 is any vertex of the delzant polytope @xmath8 and the sequence of functions @xmath25 admits a @xmath26-asymptotic expansion of the form @xmath27 , where each @xmath28 is smooth .
thus , given any smooth function @xmath29 , the trace of the operator @xmath30 is equal to @xmath31 where @xmath32 is defined by the formula @xmath33 ( the particular case @xmath34 corresponds to the computation of the dimension of @xmath13 , see discussion above ) .
the right - hand side of this equality is of the form ( [ eq : sum ] ) , up to a multiplicative factor @xmath35 , and can thus be evaluated thanks to euler - maclaurin formulas .
the goal of this paper is double .
firstly , we recover some of tate s results by elementary methods different from those used in @xcite starting from a result due to guillemin and sternberg @xcite , which was itself derived by elementary means .
more precisely , we obtain explicit euler - maclaurin formulas to approximate ( [ eq : sum ] ) for compactly supported functions @xmath0 on wedges in @xmath3 , intervals , polygons , and polytopes @xmath2 in @xmath36 , where the coefficients in the asymptotic expansion are sums of differential operators involving only derivatives of @xmath0 in directions normal to the faces of @xmath2 .
secondly , we provide explicit examples in dimension 2 ; the first one is fully explicit , in the sense that the asymptotic expansion of ( [ eq : sum ] ) is finite and that we provide all of its coefficients , while for the second one we provide some numerical evidence of the correctness of this asymptotic expansion .
the only reason to restrict ourselves to intervals , polygons and @xmath1-polytopes , is because of technical difficulties due to our method of proof .
we hope to be able to extend our techniques to higher dimensions in future works . however , we believe that the methods and results presented here are themselves already interesting , understandable in the sense that they only involve elementary computations , and may serve as a comprehensive introduction to the more general topic of euler - maclaurin formulas over lattice polytopes in any dimension as treated in @xcite . furthermore ,
despite the simplicity of the computations involved in the cases that we treat , we have not seen them explicitly written anywhere , which constitutes another reason to think that this proof is worth presenting on its own .
the structure of the paper is as follows : in section [ mainm ] we will state our main result , and the remaining sections of the paper are devoted to its proof and further refinements . in the case of polygons
@xmath2 , our approach leads to formulas in which the coefficients in the asymptotic expansion can be explicitly calculated , and we do so in section [ n2case ] . in the appendix , we recall some of tate s notation and results and provide a comparison with ours .
let @xmath37 and let @xmath38 be the integer lattice in @xmath39 .
let @xmath40 and @xmath41 be the corresponding dual spaces .
let @xmath42 denote the pairing of @xmath41 with @xmath39 .
the subset @xmath43 given by the inequalities @xmath44 is called an _ integer @xmath45-wedge _ if for every @xmath46 the constant @xmath47 is an integer and the vector @xmath48 is a primitive lattice vector in @xmath41 .
let @xmath49 be the subspace of @xmath50 spanned by the @xmath48 s .
we say that @xmath51 is _ regular _ if @xmath52 is a lattice basis of the lattice @xmath53 . when @xmath54 , define the diffeomorphism @xmath55 , such that @xmath56 has coordinates @xmath57 , @xmath58 , in the standard orthonormal basis of @xmath3 .
let @xmath59 be the facet of @xmath51 defined by @xmath60 for @xmath61 , set @xmath62 and let the constant @xmath63 be defined as the jacobian of the diffeomorphism @xmath64 , and write @xmath65 let @xmath66 be the dual basis of @xmath67 .
the constant @xmath63 is the inverse of the volume of the parallelotope generated by the vectors @xmath68 for @xmath69 such that @xmath70 , that is , the primitive outwards vectors defining the face @xmath71 ( see figure [ fig : kalpha ] ) .
we introduce the following notation : * @xmath72 ; * @xmath73 is given by @xmath74 if @xmath75 , @xmath76 if @xmath70 ; * if @xmath77 , @xmath78 is the @xmath79-tuple of vectors @xmath80 * @xmath81 stands for the number of indices @xmath69 such that @xmath75 .
we define @xmath82 , @xmath83 as follows : @xmath84 , @xmath85 , @xmath86 if @xmath87 and @xmath88 with @xmath89 the @xmath90-th bernoulli number .
let @xmath91 a regular @xmath92-wedge is an example of an @xmath92-dimensional delzant polytope .
let @xmath93 be an @xmath92-dimensional polytope .
we say that @xmath2 is _ delzant _ if it is a simple and regular polytope is _ simple _ if there are exactly @xmath92 edges meeting at each vertex of @xmath2 ; it is _ regular _ if the primitive vectors in the direction of the edges span a basis of @xmath94 , i.e. for each vertex @xmath24 of @xmath2 , the edges of @xmath2 which intersect at @xmath24 lie on rays @xmath95 , @xmath96 , where @xmath97 is a lattice basis of @xmath38 . ] .
suppose that @xmath2 has @xmath98 facets .
then @xmath2 is defined by @xmath98 equations : @xmath99 where @xmath100 . for @xmath101 , we denote by @xmath102 the set of faces of codimension @xmath103 of @xmath2 .
the following theorem gives asymptotic euler - maclaurin formulas for riemann sums .
it holds in any dimension for wedges , and in dimensions @xmath104 , @xmath105 , and @xmath1 for polytopes .
the case of @xmath106-dimensional polytopes is more complicated to handle with our techniques , and we leave it to future works ( see remark [ xx ] ) .
the first assertion of the theorem is the same as ( * ? ? ?
* proposition 3.1 ) ( see the appendix for a comparison ) , and the second one is similar to ( * ? ? ? * theorem 5.1 ) .
in fact , the uniqueness result ( * ? ? ?
* theorem 5.3 ) implies that the operators that we construct are the same , but what differs is the way we obtain them .
[ keyt ] let @xmath107 .
then the following hold . * if @xmath51 if a regular @xmath92-dimensional wedge in @xmath3 , @xmath108 where the integral is taken over @xmath51 if the intersection is empty , and the integral over a single point means evaluation at this point .
the sign @xmath109 indicates equality modulo @xmath110 . *
if @xmath93 , @xmath111 , is an @xmath92-dimensional delzant polytope with vertices in @xmath94 , for every @xmath112 and every face @xmath113 with @xmath114 , there exists a linear differential operator @xmath115 of degree @xmath116 depending only on @xmath71 and involving only derivatives of @xmath0 in directions normal to @xmath71 such that @xmath117 theorem [ keyt ] will follow from combining several upcoming results : * proposition [ keylemma ] ; * theorem [ theo : w ] ; * theorem [ h2 ] .
some of the results of the paper are more general than theorem [ keyt ] , but we leave them to later sections for simplicity .
our proof of theorem [ keyt ] is different from the proof of tate s general result in ( * ? ? ?
* theorem 5.1 ) , self - contained , and elementary ( in the sense that it relies only on freshman calculus ) .
we expect to extend part ( ii ) of theorem [ keyt ] to higher dimensions in future works . in the case where @xmath2 is a polygon , we give concrete expressions for the coefficients in the formula in theorem [ keyt ] , see theorem [ lemma : polydim2 ] .
-dimensional wedge.,width=151 ]
the following approximation result was recently proven by guillemin - sternberg @xcite .
let @xmath51 be a regular integer @xmath45-wedge defined by .
let @xmath118 be the perturbed set defined by @xmath119 , @xmath46 , where @xmath120 . then ,
if @xmath107 , we have that @xmath121 where @xmath122 and @xmath123 is the todd function , on the variable @xmath124 , for every @xmath125 .
the sign @xmath109 indicates equality modulo @xmath110 .
[ lm : wedgegs ] now let @xmath126 be an @xmath92-dimensional delzant polytope with vertices in @xmath38 and exactly @xmath98 facets .
then @xmath2 is defined by the inequalities @xmath127 where @xmath47 is an integer and @xmath128 is a primitive vector perpendicular to the @xmath129-facet of @xmath2 , and pointing outwards from @xmath2 .
because @xmath2 is simple by assumption , every codimension @xmath130 face of @xmath2 is defined by a collection of equalities @xmath131 , @xmath132 , where @xmath71 is a subset of @xmath130 elements of the set @xmath133 .
let @xmath134 denote the @xmath130-wedge @xmath135 because @xmath2 is regular , each @xmath130-wedge @xmath134 is regular . guillemin and sternberg have recently shown @xcite the following euler - maclaurin formula .
[ regsim : gs ] let @xmath2 be a delzant polytope with vertices in @xmath94 .
let @xmath136 be the perturbed polytope defined by the equations @xmath137 , @xmath100 . then , if @xmath107 , we have @xmath138 where @xmath139 and @xmath123 is the todd function on the variable @xmath124 , for every @xmath140 .
a general asymptotic euler - maclaurin formula for riemann sums over lattice polytopes ( simple or not ) was given by tate @xcite .
as far as we know , theorem [ regsim : gs ] does not follow from tate s formula .
other euler - maclaurin formulas have been obtained by berline - vergne @xcite , brion - vergne @xcite , cappell - shaneson @xcite , karshon - sternberg - weitsman @xcite , and zelditch @xcite among other authors .
let @xmath51 be a regular integer @xmath92-wedge defined by . recall that @xmath66 is the dual basis of @xmath67 and @xmath59 is the facet of @xmath51 defined by @xmath141 then @xmath142 generates the edge @xmath143 .
for any integer @xmath103 , we introduce the operator @xmath144 defined by @xmath145 with the convention that the integral is taken over @xmath51 if the intersection is empty , and that integrating a function over the vertex means evaluating it at the vertex .
@xmath63 is a constant depending only on the face @xmath146 , of which we gave an interpretation earlier .
[ keylemma ] if @xmath107 , we have that @xmath147 the idea of the proof is the following : first , we compute the full asymptotic expansion given by lemma [ lm : wedgegs ] in the case of the standard @xmath92-wedge @xmath148 . then , we perform a change of variables to deal with the case of a general regular @xmath92-wedge .
we start by writing the expansion of @xmath149 ; we recall that @xmath150 write @xmath151 where @xmath152 .
the coefficient @xmath153 of @xmath154 in @xmath155 is equal to @xmath156 with @xmath157 hence @xmath158 from this result , we deduce a formula for the case of the standard wedge .
remember that @xmath159 thus @xmath160 where @xmath161 and @xmath162 . from this
we obtain a general formula when @xmath163 : @xmath164 this finally yields @xmath165 this gives the desired formula in the case of the standard wedge .
let us now turn to the general case .
let @xmath51 be the regular @xmath92-wedge defined by ( [ for:1 ] ) .
define the diffeomorphism @xmath166 , such that @xmath56 has the following coordinates in the standard orthonormal basis of @xmath3 : @xmath167 then @xmath168 is the standard wedge ; moreover , @xmath169 is a diffeomorphism from @xmath170 to @xmath171 for each subset @xmath172 of @xmath173 .
we have that @xmath174 with @xmath175 .
but we know from the previous case that @xmath176 therefore , it only remains to prove that for every @xmath177 , @xmath178 first , we have to express the quantity @xmath179 in terms of @xmath0 . since @xmath180 , we have @xmath181 but @xmath182
so @xmath183 .
it follows that @xmath184 by induction , we find @xmath185 now , we have to understand integrals of the form @xmath186 from the previous discussion @xmath187 hence @xmath188 where @xmath63 is the jacobian of the diffeomorphism @xmath189 , which was to be proved .
another way to write this result is the following : we can eliminate the constants @xmath63 by normalizing the measure on the face @xmath190 ; we set @xmath191
in order to deduce a formula for polygons and @xmath1-dimensional polytopes , we need to rewrite the asymptotic expansion in proposition [ keylemma ] in a suitable form .
some of the results of this section apply for general @xmath37 , and when this is the case we present the most general version of the result .
the final results require that we restrict to @xmath192 , or @xmath1 .
recall that for @xmath193 , @xmath194 denotes the set of faces of codimension @xmath98 of @xmath51 . in order to simplify the notation , we set @xmath195 for @xmath196 .
our objective is to write a formula of the following kind : @xmath197 where @xmath198 is a differential operator associated to the face @xmath71 , with good properties in a sense that we will precise later .
let @xmath199 .
there exists a subset @xmath200 of @xmath173 such that @xmath201 ; the family @xmath202 is a basis of the linear subspace spanned by @xmath71 . a first expression for @xmath198 is @xmath203 observe that if @xmath204 , then @xmath205 . if @xmath206 , then @xmath198 involves the integral @xmath207 when @xmath208 , the situation is a little bit more complicated , because we integrate directional derivatives of @xmath0 involving the vectors @xmath209 .
but we would like to keep only quantities that depend on the face @xmath71 and nothing else ; this is why we decompose the vectors @xmath209 as follows : @xmath210 where @xmath211 is the outward primitive normal to the facet @xmath212 .
next , we expand the quantity @xmath213 as a linear combination of @xmath214 terms involving the vectors @xmath215 and @xmath216 .
more precisely , write equation ( [ eq : vi ] ) as @xmath217 where @xmath218 , @xmath219 if @xmath220 and @xmath221 , @xmath222 if @xmath223 . moreover , set @xmath224 and define integers @xmath225 , @xmath226 as follows : @xmath227 if @xmath228 , , @xmath229 if @xmath230
. then @xmath231 which can be written by multilinearity as @xmath232 we now want to get rid of the vectors @xmath233 , @xmath220 when they appear in the quantity @xmath234 if @xmath235 all belong to @xmath172 , then only the normal vectors @xmath236 appear , and we have nothing to do ; this constitutes @xmath237 favorable cases . in unfavorable cases , we use the following lemma . [ lemma : int ] for any function @xmath238 and for every @xmath220 , we have @xmath239 where @xmath240 is the face of codimension @xmath241 defined by @xmath242 .
let the elements of @xmath243 be denoted by @xmath244 .
one has @xmath245 when @xmath246 , applying lemma [ lemma : int ] to functions of the form @xmath247 and repeating this as many times as necessary , we can get rid of all the vectors @xmath233 , @xmath220 in the expression of @xmath248 and keep only integrals over faces of codimension greater than @xmath98 of derivatives of @xmath0 applied to vectors that are normal to the hyperplanes defining the faces . before we state and prove our result , let us introduce some useful notation . let @xmath249 be the cone generated by the set @xmath250 if @xmath251 is a non - empty subset of @xmath39 , let @xmath252 be the vector subspace generated by the elements of the form @xmath253 , @xmath254 .
the following two lemmas hold in any dimension ( not just @xmath105 and @xmath1 ) .
[ lemma : depend ] decompose the vectors @xmath142 , @xmath255 , as in equation ( [ eq : vi ] ) .
choose another set of vectors @xmath256 such that * @xmath257 , @xmath258 belongs to @xmath259 * the family @xmath260 is a primitive lattice basis , and write , for @xmath255 , @xmath261 then for @xmath223 , we have @xmath262 . in other words , this means that the scalars @xmath263 only depend on the face @xmath71 ; when @xmath71 will be considered as a face of a polytope instead of a wedge , then the contribution coming from each wedge will display the same coefficient . by definition of the vectors @xmath233 , @xmath236 , @xmath264 , one has @xmath265 hence ,
the @xmath98 coefficients @xmath263 , @xmath223 , are obtained by solving the linear system ( [ eq : zeta ] ) of @xmath98 equations .
now , we express the vector @xmath258 in the basis @xmath266 : @xmath267 since the vector @xmath258 belongs to @xmath268 , all the scalar products @xmath269 , @xmath270 , vanish .
this implies that for every @xmath270 , @xmath271 .
thus , the matrix @xmath272 of change of basis from @xmath273 to @xmath274 is of the form @xmath275 where @xmath276 since @xmath273 and @xmath277 are primitive lattice bases , we have @xmath278 , and hence for every @xmath255 , @xmath279 . but
@xmath142 and @xmath258 belong to @xmath249 , so @xmath280 .
this yields that for @xmath281 , we have @xmath282 .
[ lemma : dependmu ] decompose the vectors @xmath142 , @xmath255 , as in equation ( [ eq : vi ] ) , and fix @xmath220 . set @xmath283 , and choose another set of vectors @xmath284 such that
* @xmath285 , @xmath258 belongs to @xmath286 * the family @xmath287 is a primitive lattice basis . and
write , for @xmath255 , @xmath288 then we have @xmath289 .
in other words , this means that the scalar @xmath290 only depends on the face @xmath291 . by definition of the vectors @xmath292 , @xmath293 , @xmath294 , one has @xmath295 which can be written in matrix form @xmath296 where @xmath297 are the column vectors given by @xmath298 and @xmath299
is the symmetric matrix whose generic coefficient is @xmath300 , @xmath301 .
similarly , the constants @xmath302 satisfy the system of equations @xmath303 and @xmath304 thanks to the proof of the previous lemma , we know that there exists scalars @xmath305 , @xmath306 , @xmath307 such that @xmath308 hence we have @xmath309 as well as @xmath310 and @xmath311 using these relations , equations ( [ eq : ek ] ) become @xmath312 while equation ( [ eq : ei ] ) becomes @xmath313 considering the linear combination ( [ eq : ei ] ) - @xmath314([eq : ek ] ) , we replace equation ( [ eq : ei ] ) by the new equation ( we do no write the details of the computations ) @xmath315 together with equations ( [ eq : ek ] ) , this means that the coefficients @xmath302 are solutions of the system @xmath316 where @xmath299 and @xmath317 are as before , @xmath318 is defined as @xmath319 but with the coefficients @xmath302 instead of @xmath320 , and @xmath49 is the column vector whose entries are @xmath321 , @xmath322 comparing this to ( [ eq : s1 ] ) yields @xmath323 , and in particular @xmath324 .
[ theo : w ] assume that @xmath111 . for every @xmath112 and every face @xmath199 with @xmath325 ,
there exists a linear differential operator @xmath115 of degree @xmath326 depending only on @xmath71 ( in the sense introduced in the previous lemmas ) and involving only derivatives of @xmath0 in directions normal to the face @xmath71 such that @xmath327 to compute @xmath115 , we apply the previous technique to faces of codimension smaller than @xmath98 and gather their contribution as integrals over @xmath71 .
it follows from lemma [ lemma : depend ] and lemma [ lemma : dependmu ] that @xmath115 depends only on the face @xmath71 ; let us briefly explain how .
if @xmath328 , we have to handle two types of faces : the two edges ( @xmath329 ) and the vertex ( @xmath330 ) of the wedge .
there is not much to say about the case of the vertex .
when we integrate over an edge , and we apply our technique , we will find * constants @xmath331 in front of derivatives of @xmath0 evaluated at the vertex , and there is nothing to prove , * constants @xmath332 in front of integrals of derivatives of @xmath0 on @xmath71 , and lemma [ lemma : depend ] ensures that it only depends on the face @xmath71 .
if @xmath333 , we have three types of faces , namely planes ( @xmath329 ) , edges ( @xmath330 ) and the vertex ( @xmath334 ) .
the difference with the previous case is that when we consider integrals over a plane , we obtain integrals over edges belonging to this plane , each one displaying a factor @xmath331 ; lemma [ lemma : dependmu ] ensures that it only depends on the given edge .
[ xx ] 1 . in dimension @xmath106 and higher ,
the situation is more complicated , and lemma [ lemma : depend ] and lemma [ lemma : dependmu ] are not enough to obtain a similar theorem . indeed
, think of the following situation : we take @xmath335 and want to evaluate @xmath336 we start by expanding @xmath337 and we apply lemma [ lemma : int ] to obtain @xmath338 we have to apply the method one more time for each of these integrals .
for instance , we put @xmath339 and to compute this integral , we write @xmath340 which yields , again thanks to lemma [ lemma : int ] @xmath341 + hence , in the expression of @xmath172 , we obtain the term @xmath342 does the factor only depend on the face @xmath343 ? we think that our previous lemmas are not enough to give an answer to this question .
2 . in principle , one should be able to obtain explicit expressions for the operators @xmath344 , but this leads to computations involving a very large number of constants , so it is not reasonable to try to write such an expression .
however , if we restrict ourselves to dimension @xmath105 , we can be fully explicit , as we will see later .
as in section [ sec : intro ] , let @xmath93 be a delzant polytope with vertices in @xmath94 in dimension @xmath111 with equations @xmath345 , @xmath100 .
recall that for @xmath346 , @xmath347 denotes the set of faces of codimension @xmath45 of @xmath2 .
we introduce as in theorem [ theo : w ] the operators @xmath115 associated to a face @xmath71 of the polytope ( remembering that it only depends on the face as part of the polytope ) . for any integer @xmath103
, we define the operator @xmath348 by @xmath349 [ h2 ] if @xmath107 , we have that @xmath350 [ thm : main ] notice first that @xmath351 where @xmath90 is the number of vertices of @xmath2 and @xmath352 is the regular wedge which is the intersection of the @xmath92 facets @xmath353 , @xmath264 , intersecting at the vertex @xmath68 .
cover @xmath2 by open sets @xmath354 , @xmath355 , such that @xmath354 contains the vertex @xmath68 and does not intersect any other facet than the @xmath353 , @xmath264 .
choose a partition of unity associated to this open covering and write @xmath356 where @xmath357 has support included in @xmath354 . then @xmath358
now , from formula ( [ eq : expansionwedge ] ) , we know that for @xmath359 @xmath360 hence it is enough to check that for all @xmath103 @xmath361 this amounts to show that for each face @xmath71 of the polytope @xmath362 but this is clear because @xmath115 is linear and because @xmath363 .
we would like to compute explicitly the operators @xmath364 ; unfortunately , as already said , this seems to be quite complicated in all generality .
however , we can give nice formulas in dimension @xmath105 .
let @xmath2 be a regular integer polygon defined by . in this case
, we only have two types of faces : vertices ( codimension @xmath105 ) and edges ( codimension @xmath104 ) . let @xmath365 ( resp .
@xmath317 ) be the set of edges ( resp .
vertices ) of @xmath2 .
if @xmath366 belongs to @xmath365 , let @xmath367 be the associated outward primitive normal vector ; if @xmath24 belongs to @xmath317 , let @xmath368 be the integral basis of @xmath369 such that the two edges meeting at @xmath24 are contained in the half - lines @xmath370 , @xmath371 ; we denote by @xmath372 the edge generated by @xmath373 .
define the quantities @xmath374 and @xmath375 .
now , let @xmath366 be an edge , and let @xmath376 be the cone generated by the set @xmath377 . given a generator @xmath378 of @xmath379 ,
there exists a vector @xmath380 such that @xmath381 is a primitive lattice basis of @xmath382 .
the quantity @xmath383 does not depend on the choice of @xmath380 .
the lemma follows from lemma [ lemma : depend ] , but its proof is very simple , so we present it next .
choose another vector @xmath384 such that @xmath385 is a primitive lattice basis of @xmath382 .
write @xmath386 ; then , one has @xmath387 .
the matrix of the change of the basis is of the form @xmath388 because its determinant must be @xmath389 , we have @xmath390 .
since both @xmath391 and @xmath392 belong to @xmath376 , the only possibility is @xmath393 .
[ lemma : polydim2 ] in theorem [ thm : main ] the operators @xmath348 are given by : * @xmath394 ; + * @xmath395 ; + * @xmath396 ; + * if @xmath397 , then @xmath398 where @xmath399 and @xmath400 is equal to @xmath401 * if @xmath397 , @xmath402 is equal to @xmath403 theorem [ lemma : polydim2 ] recovers the formula in ( * ? ? ?
* corollary @xmath404 ) . to compare the two formulas ,
one may notice that tate does not separate the even and odd cases , and that in the odd case nearly every coefficient in tate s formula vanishes because of the properties of the bernoulli numbers .
we have to compute the operators @xmath115 as in section [ subsection : expwedge ] .
we start by the case @xmath405 .
let @xmath24 be a vertex and let @xmath51 be the wedge formed by this vertex and the two incident edges .
define the vectors @xmath406 as before , and let @xmath407 ( resp .
@xmath408 ) be the edge generated by @xmath409 ( resp .
@xmath410 ) .
we have @xmath411 if @xmath412 is the outward primitive vector normal to the edge @xmath372 , we write @xmath413 where @xmath414 ( resp .
1 ) if @xmath415 ( resp .
2 ) . taking the scalar product with @xmath236 and @xmath416
, we find @xmath417 now , thanks to lemma [ lemma : int ] , we have ( being careful that the vector @xmath418 is the opposite of the vector @xmath419 in this lemma ) @xmath420 adding the contributions from each vertex , we obtain the desired formula . now , let @xmath421 ; then @xmath422 we write @xmath423 by lemma [ lemma : int ] , we have @xmath424 and hence we obtain @xmath425 by a straightforward induction , this yields @xmath426 the case @xmath427 works in a similar way .
let us describe two examples in the @xmath105-dimensional case .
the first one is fully explicit , in the sense that we choose a function for which the asymptotic expansion given in theorem [ thm : main ] is finite , and we provide every coefficient of the latter .
the second one is more involved and we compute the first three terms of the asymptotic expansion and give numerical computations of the remainder .
let @xmath2 be the triangle with vertices @xmath428 , @xmath429 and @xmath430 and let @xmath431 be the function defined by @xmath432 ( multiplied by a cutoff function so that it is compactly supported ) .
then @xmath433 therefore , we have @xmath434 using standard formulas for sums of integers and squares of integers , one can check that @xmath435 let us compare this with theorem [ lemma : polydim2 ] . with the notation of this lemma , we have @xmath436 so the zeroth order terms agree .
let us give names to the vertices and edges of @xmath2 as follows : we put @xmath437 , @xmath438 and @xmath439 , and we let @xmath372 denote the edge joining the vertices @xmath440 ( or @xmath441 ) and @xmath442 ( or @xmath443 ) .
then we have @xmath444 and hence @xmath445 furthermore , we have @xmath446 with @xmath447 and @xmath448 we have @xmath449 , @xmath450 , @xmath451
. moreover @xmath452 and thus @xmath453 .
this yields @xmath454 now , one can check that @xmath455 and @xmath456 .
we obtain @xmath457 and therefore @xmath458 finally , we have @xmath459 because the derivatives of @xmath0 of order greater than @xmath105 vanish .
let us present another example .
we choose @xmath2 as in the previous paragraph , and consider the function @xmath460 we want to evaluate the sum @xmath461 let us compute the coefficients @xmath462 , @xmath463 , of the asymptotic expansion given in theorem [ thm : main ] .
firstly , we have @xmath464 secondly , notice that @xmath465 and hence @xmath466 finally , since @xmath467 and @xmath468 we obtain @xmath469 ( we deliberately keep this fraction as it is , which will make sense in view of the next result ) .
moreover , we have @xmath470 and @xmath471 thus @xmath472 we deduce from the previous results that @xmath473 we can now compute @xmath474 in figure [ fig : comparison ] , we compare the value of @xmath475 to the numerical value of @xmath476 , for @xmath477 . in figure
[ fig : error ] , we numerically evaluate the quantity @xmath478 more precisely , we plot the value of @xmath479 as a function of @xmath480 .
the remainder @xmath481 displays a behavior in @xmath482 , as expected . and @xmath476 as functions of @xmath483 . ] as a function of @xmath480 .
notice the behavior in @xmath482 of @xmath481 . ]
we believe that one should be able to prove item ( ii ) in theorem [ keyt ] in dimensions @xmath484 with the same elementary method we use in this paper ( since a similar result was already stated in tate s article in any dimension ) . however , in dimensions @xmath484 the computations appear to be complicated ( but still elementary ) and will be the object of future works ( see also remark [ xx ] ) .
we devote this appendix to comparing tate s formula for riemann sums over regular wedges ( * ? ? ? * proposition @xmath485 ) and the first assertion of theorem [ keyt ] of the present paper .
firstly , we need to recall some of the notation and definitions used in tate s article @xcite ; most of them come from section @xmath1 of the latter .
+ let @xmath251 be a finite dimensional real vector space , of dimension @xmath486 , and @xmath487 be a lattice in @xmath251 .
let @xmath488 be a unimodular cone ( a wedge ) in @xmath251 and @xmath365 be the integral basis of @xmath487 generating @xmath488 .
then @xmath489 is the set of families of non - negative integers with @xmath490 elements ; an element @xmath491 is of the form @xmath492 .
if @xmath366 belongs to @xmath365 , define @xmath493 in such a way that @xmath494 and @xmath495 if @xmath496 . for @xmath497 , let @xmath498 be the set of @xmath491 such that for all @xmath499 , @xmath500 ; then @xmath491 can be written @xmath501 . if @xmath502 is a subset of @xmath365 , set @xmath503 moreover , set @xmath504 . for @xmath505 , set @xmath506 and @xmath507 . for @xmath508 and for @xmath497 nonempty , set @xmath509 with @xmath510 , where @xmath82 is the @xmath92-th bernoulli number .
moreover , for @xmath511 , set @xmath512 finally , define the differential operator @xmath513 as @xmath514 with @xmath515 . using the previous comparison ,
@xmath519 becomes with our notation : @xmath520 hence , formula ( [ formula ] ) becomes @xmath521 which coincides with our formula for wedges ( first assertion of our main theorem [ keyt ] ) . _
acknowledgements_. ap was supported by nsf career grant dms-1055897 and nsf grant dms-0635607 .
work on the topic of this paper started when ap was a postdoctoral fellow at mit ( 2007 - 2008 ) and during this period he was supported by an nsf postdoctoral fellowship .
he would like to thank victor guillemin for many useful discussions on this paper , and on many other topics related to symplectic geometry and spectral theory , during his stay at mit . finally , we thank michle vergne for comments on a preliminary version of this manuscript .
m. d. hamilton : the quantization of a toric manifold is given by the integer lattice points in the moment polytope .
_ toric topology _
, 131140 , contemp .
, 460 , amer .
soc . , providence , ri , 2008 .
kantor and a.g .
khovanski : une application du thorme de riemann - roch combinatoire au polynme dehrhart des polytopes entiers de @xmath522 .
_ comptes rendus de lacadmie des sciences .
srie i. mathmatique _ ,
* 317 * , no . 5 , 501507 , 1993 .
a. g. khovanski and a. v. pukhlikov : the riemann - roch theorem for integrals and sums of quasipolynomials on virtual polytopes .
_ rossiskaya akademiya nauk .
algebra i analiz _ , * 4 * , no .
4 , 188216 , 1992 . * yohann le floch * + school of mathematical sciences + tel aviv university + ramat aviv + tel aviv 6997801 , israel .
+ _ e - mail : _ [email protected] + * lvaro pelayo * + university of california , san diego + department of mathematics + 9500 gilman dr , # 0112 + la jolla , ca 92093 + united states of america + _ e - mail _ : [email protected] | recently there has been a renewed interest in asymptotic euler - maclaurin formulas , partly due to applications to spectral theory of differential operators .
using elementary means , we recover such formulas for compactly supported smooth functions @xmath0 on intervals , polygons , and @xmath1-dimensional polytopes @xmath2 , where the coefficients in the asymptotic expansion are sums of differential operators involving only derivatives of @xmath0 in directions normal to the faces of @xmath2 .
our formulas apply to wedges of any dimension .
this paper builds on , and is motivated by , works of guillemin , sternberg , and others , in the past ten years . |
the creation , detection , and study of exciton condensates in semiconductors constitute interesting problems in condensed matter physics.@xcite exciton populations are usually created by photo - exciting a semiconductor , so that the electron - hole pairs consist of conduction band electrons and valence band holes .
then , optical techniques are employed to detect and study the excitonic system.@xcite recently , there has also been interest in the possibility of _ electrical detection _ , and the realization _ without photoexcitation _ , of such exotic states in a bilayer quantum hall system , where the excitons result from the pairing of electrons in one quantum hall layer with holes in the neighboring layer.@xcite this report examines a hybrid scenario where _ photoexcitation _ serves to create inter - landau level , instead of inter - band , excitations in a quantum hall system , as _ electrical detection _ is employed to detect the system response.@xcite in particular , we study _ transport _ in a semiconductor single - layer two - dimensional electron system , where _ microwave photo - excitation _ produces a steady state density of inter - landau level excitations ( `` excitons '' ) .
surprisingly , such a system exhibits novel zero - resistance states without concomitant hall quantization , at low temperatures , @xmath3 , about magnetic fields , @xmath4 = @xmath5 and @xmath4 = @xmath6 , where @xmath7 = @xmath8 , @xmath9 is an effective mass , @xmath10 is the electron charge , and @xmath11 is the radiation frequency.@xcite the experimental observations have led to a broad theoretical study of this situation.@xcite at the present , it is uncertain whether the observed zero - resistance states should be viewed as a novel condensate with a role for excitons , although we had suggested this possibility.@xcite theoretical studies have realized radiation - induced magneto - resistivity oscillations through various approaches.@xcite the observed zero - resistance states have then been attributed to a negative resistivity instability , and current - domain formation.@xcite as there remain many open questions from both the experimental and theoretical perspectives , we examine here just the experimental situation , and refer the reader to the literature for further theory .
experiments were carried out on standard devices fabricated from gaas / algaas heterostructures .
the best material was typically characterized by an electron density , @xmath12(4.2 k ) @xmath13 @xmath14 @xmath15 @xmath16 @xmath17 , and an electron mobility , @xmath18 , up to @xmath19 @xmath15 @xmath20 @xmath21/vs .
standard four - terminal lock - in based electrical measurements were performed with the sample mounted inside a waveguide and immersed in pumped liquid he-3 or he-4 .
electromagnetic ( em ) waves in the microwave part of the spectrum were generated using tunable sources .
the em waves were directed to the sample through the waveguide .
further details appear elsewhere.@xcite the data of fig .
1(a ) illustrate the radiation - induced effect on the diagonal ( @xmath22 ) and hall resistances ( @xmath23 ) , when the specimen is excited with microwaves at @xmath11 = 108 ghz . in fig .
1(a ) , the application of microwaves induces giant oscillations in @xmath22 , which are characterized by the property that the @xmath22 under photoexcitation drops below the dark @xmath22 over wide @xmath4 intervals .
thus , at approximately @xmath24 = @xmath25 and @xmath24 = @xmath26 , with @xmath25 = 1,2,3 ... , the @xmath22 under photoexcitation is equal to the dark @xmath22 . for @xmath27 , the photo - excited @xmath22
is reduced with respect to the dark @xmath22 . on the other hand , for @xmath28 , the irradiated @xmath22
is enhanced with respect to the dark @xmath22 .
a remarkable feature in the data are the zero - resistance states which are manifested about @xmath29 , i.e. , in the vicinity of @xmath30 and @xmath31 .
these @xmath22 minima exhibit activated transport vs. the temperature , which is suggestive , from the experimental perspective , of a possible radiation - induced gap in the electronic spectrum.@xcite the data of fig .
1a also illustrate the influence of radiation on the hall effect . here
, a comparison of the dark @xmath23 with the photo - excited @xmath23 reveals subtle microwave induced oscillations , which are period commensurate with the radiation induced @xmath22 oscillations.@xcite yet , there do not occur microwave - induced plateaus in the hall effect that coincide with the radiation - induced vanishing @xmath22 in the vicinity of b = @xmath5 and @xmath32 . in order to highlight these radiation - induced changes in @xmath23 ,
@xmath33 is shown along with @xmath22 in fig .
1(b ) . a comparison of fig .
1(a ) and fig .
1(b ) shows clearly that @xmath23 is reduced in magnitude over the b - intervals , where @xmath22 is enhanced by the radiation . on the other hand , as the diagonal resistance vanishes upon microwave excitation , as in the vicinity of @xmath30 , for example , the correction @xmath34 also vanishes . as the period of the radiation induced oscillations ( see fig .
1 ) are mostly sensitive to the @xmath4-component that is perpendicular to the plane of the 2des , they tend to span a wider @xmath4-interval , at a finite tilt angle of the specimen .
thus , under tilt , the characteristic field scale becomes @xmath35 , and the dark @xmath22 intercepts the photo - excited @xmath22 in the vicinity of @xmath36/j$ ] and @xmath36/(j+1/2)$ ] , with @xmath25 = 1,2,3 ... these features can be detected in fig . 2 the phase of the radiation induced @xmath22 oscillations is an important parameter because it can be used to compare experiment to theory.@xcite to determine this observable , @xmath22 maxima and minima of 182 ghz data ( see fig .
3 ) are initially labelled by integers and half - integers , respectively , assuming a cosine waveform .
the extrema labels , denoted by the index @xmath37 in fig . 3 ,
are then plotted vs. extremal values of @xmath38 , and a linear fit , @xmath39 , is applied to determine the slope , @xmath40 = @xmath7 , and the ordinate intercept . in fig .
3 , the ordinate intercept for the fit of extremal data points at 1/4 demonstrates a 1/4-cycle phase shift of the extrema with respect to the assumed ( cosine ) waveform.@xcite thus , the plot ( fig .
3(a ) ) demonstrates that the phase is broadly consistent with a lineshape @xmath41 , i.e. , resistance minima about @xmath42 , and resistance maxima about @xmath43 .
a similar analysis of the nodes or fixed points suggests that they are characterized by @xmath44 , or @xmath45 , see fig .
3 . in prior work
, we have shown that the radiation induced resistance _ minima _ exhibit activated transport , i.e. , @xmath46.@xcite as a supplement to that previous report , we examine the @xmath3-dependence of the resistance _ maxima _ in fig .
thus , the inset in fig .
4 shows @xmath22 under microwave excitation , with a pair of marked maxima , whose @xmath3-evolution is followed in the main part of the figure .
4 shows that the @xmath22 maxima increase rapidly with decreasing temperature , and the variation can be empirically represented by an exponential function .
the @xmath3-variation of the maxima have been attributed by dmitriev et al .
, to the @xmath3-dependence of the inelastic scattering rate.@xcite measurements examining possible reflection / emission and transmission characteristics of the high mobility 2des under microwave excitation are illustrated in fig .
5 and fig .
6 . for these experiments
[ see inset , figs . 5 and 6 ] , a resistance sensor was placed immediately above or below the sample.@xcite fig .
5 exhibits the dark , and irradiated at 30 ghz , @xmath22 , along with the sensor @xmath47 response under photoexcitation . the irradiated @xmath22 shows giant microwave - induced oscillations .
note that , in this low-@xmath11 case , the irradiated @xmath22 crosses the dark @xmath22 also near @xmath48 . also shown in the figure
is @xmath49 , which serves to bring out weak features in the @xmath47 .
the @xmath47 signal shows a pair of significant features between @xmath7 and 2@xmath7 . here , the largest feature might originate from a two - photon process . in the derivative signal ,
there is a very weak feature near -@xmath7 ( but not at + @xmath7 ) , which could signify electrically detected cyclotron resonance .
the origin of other structure in the sensor signal is not fully understood .
some transmission measurements are shown in fig .
, the optimal radiation induced @xmath22 response was observed in the vicinity of -9 db . a further increase of the radiation intensity ( db @xmath50 0 ) produced a reduction in the peak height along with an increase in the resistance at the minima ( see fig .
6(a ) ) . at the same time , the transmission sensor ( see fig .
6(b ) ) indicated structure at @xmath4 @xmath51 @xmath7 , which correlated with a strong radiation - induced decrease in @xmath22 just above 0.3 tesla .
observed oscillations in @xmath22 below @xmath7 ( fig .
6(a ) ) were imperceptible , however , in the detector response ( see fig .
6(b ) ) .
it is also interesting to examine the effect of simultaneous excitation of the specimen at two different microwave frequencies.@xcite thus , fig .
7 exhibits measurements with excitation of the specimen at 15 ghz only , and at 50 ghz only . here , at 50 ghz
, there occur a number of radiation - induced oscillations and one broad zero - resistance state about @xmath52 . on the other hand , at 15 ghz , the radiation seems to produce a non - monotonic variation in @xmath22 , mostly above @xmath7 .
simultaneous excitation at two - frequencies produces a @xmath22 trace that looks very much like the trace at the higher @xmath11 , i.e. , 50 ghz . an interesting feature is observed here by comparing the @xmath53 and @xmath54 signals at @xmath4 @xmath13 0.07 t and @xmath4 @xmath13 0.14 t. at both these @xmath4 , the peak amplitude of the @xmath53 and @xmath54 signal is reduced with respect to the @xmath54 signal . yet
, one observes that in the @xmath4 @xmath13 0.07 t case , only @xmath53 produces an enhancement , while in the @xmath4 @xmath13 0.14 t case , it produces a resistance suppression , with respect to the dark signal .
thus , perhaps , one should attribute the small decrease in the amplitude of the @xmath53 and @xmath54 signal to a `` breakdown '' type heating effect , due to the increased radiation intensity resulting from the combined excitation .
in summary , radiation induced zero - resistance states and associated oscillations in the high mobility 2des exhibit a rich phenomenology . here , we have shown that ( a ) @xmath22 exhibits saturation at zero - resistance under microwave excitation in the vicinity of @xmath55 b_{f}$ ] , with concomitant vanishing hall resistance correction , @xmath56 , see fig . 1 , ( b ) the period of the radiation induced oscillations are mostly sensitive of the normal component of the magnetic field , see fig .
2 , ( c ) the oscillatory minima are shifted by 1/4 cycle with respect to the @xmath57 = @xmath25 condition , see fig . 3 , ( d ) the oscillatory resistance maxima increase strongly with decreasing temperature , see fig . 4 , ( e ) a response is observable in the transmission and reflection signals , mostly above @xmath7 , see figs . 5 and 6 , and ( e ) simultaneous photoexcitation at two radiation frequencies seems not to produce simple superposition , see fig . 7 .
we acknowledge discussions with k. von klitzing , v. narayanamurti , j. h. smet , and w. b. johnson , and the receipt of high quality mbe material from v. umansky .
r. g. mani et al . , _ nature ( london ) _ * 420 * , 646 ( 2002 ) ; _ phys .
b _ * 69 * , 193304 ( 2004 ) ; cond - mat/0305507 ; cond - mat/0306388 ; _ phys . rev .
b _ * 69 * , 161306 ( 2004 ) ; _ phys . rev
. lett . _ * 92 * , 146801 ( 2004 ) ; r. g. mani , _ physica e _ * 22 * , 1 ( 2004 ) ; cond - mat/0407143 . | the microwave - excited high mobility two - dimensional electron system exhibits , at liquid helium temperatures , vanishing resistance in the vicinity of @xmath0 b_{f}$ ] , where @xmath1 , m@xmath2 is an effective mass , e is the charge , and _
f _ is the microwave frequency . here ,
we summarize some experimental results .
journal - reference : international journal of modern physics b * 18 * , 3473 ( 2004 ) |
plenty of neutrino experiments performed in the last two decades confirmed that the three types of neutrinos in the standard model are massive particles , and the mass eigenstates @xmath4 do not coincide with the flavor eigenstates @xmath5 which enter into the charged current interactions .
the mixing and flavor oscillation phenomena of neutrinos can be described by the mass - squared differences @xmath6 ( where @xmath7 is the mass of @xmath8 state ) , and the so - called pmns unitary mixing matrix parametrized by the three mixing angles @xmath9 and one cp - violating phase @xmath10 .
the analysis of the solar and long baseline reactor neutrinos led to best - fit values @xmath11 and @xmath12 ; while the data from the atmospheric and long baseline accelerator experiments result in @xmath13 and @xmath14 @xcite .
the last mixing angle has been measured recently with the new generation of medium baseline reactor experiments , including double chooz @xcite , daya bay @xcite and reno @xcite , with the best - fit value @xmath15 .
double chooz , daya bay and reno experiments with @xmath16 , where @xmath17 and @xmath18 are the baseline and neutrino energy respectively , are sensitive to @xmath19-induced flavor oscillation with the amplitude @xmath20 .
in fact , the measurement of the small mixing angle @xmath0 in these experiments was achieved thanks to the highly controlled systematic errors and efficient background rejection down to @xmath21 of signal .
thus , in principle , the data of these experiments can be also used to discover / constrain new physics in the neutrino sector . the set up and baseline to energy ratio of double chooz , daya bay and reno experiments make them sensitive to small admixture of a new sterile neutrino state with electron anti - neutrinos , with mass - squared difference @xmath22 .
the existence of a sterile neutrino state with mass @xmath23 , the so - called @xmath24 model , was motivated by lsnd @xcite , miniboone @xcite and reactor anomalies @xcite .
recent global analyses of data for @xmath24 model can be found in @xcite .
also , several experiments have been proposed to check this scenario ( see @xcite and references therein ; see also @xcite ) .
however , from the phenomenological point of view , it is worthwhile to probe the existence of a _ light
_ sterile neutrino , a model which we call it @xmath25 . in this regard ,
we perform a detailed analysis of double chooz , daya bay and reno data for @xmath25 model . in this model
, in addition to the @xmath0 and @xmath19 parameters , the active - sterile mixing parameters @xmath26 contribute to @xmath27 oscillation probability .
we discuss the correlation among these parameters and the constraints that can be derived on them .
also , we discuss the robustness of @xmath0 determination in the presence of the sterile neutrinos . we show that the reported value of @xmath0 holds also in the presence of sterile neutrinos and the data of daya bay and reno play a crucial role in this robustness .
the prospect of sterile neutrino search in medium baseline reactor experiments has been studied in a number of papers . the possibility to search for light sterile neutrinos in reactor experiments
have been mentioned in @xcite . in @xcite , by calculating the sensitivity of double chooz to sterile neutrinos with @xmath28 , in the @xmath29 model , it has been concluded that with only far detector data the @xmath0 angle can be confused with active - sterile neutrino mixing angles . in
@xcite the interplay between a sterile neutrino with @xmath30 and @xmath0 determination has been studied by computing sensitivity of double chooz and daya bay ; with the conclusion that disentangling these parameters requires information about the positron recoil energy distortions .
a simulation of medium baseline experiments in the search of light sterile neutrinos has been performed in @xcite and the dependence of limits on systematic errors has been studied .
also , in @xcite , the correlation of @xmath0 with active - sterile mixing parameters has been studied with an emphasize on the reactor anomaly and its connection to the cosmological data .
the effect of sterile neutrinos on @xmath0 determination in both medium and long baseline experiments has been studied in @xcite . in @xcite
the data of daya bay and reno has been analyzed in @xmath24 framework ; the obtained limits are consistent with the limits of this paper in the same range of @xmath31 . constraining the sterile neutrino scenario with the solar and kamland data has been studied in @xcite which we will discuss it in section [ combined ] . in this paper
we extend the previous searches to @xmath32 and perform an analysis of the available data from the medium baseline experiments .
we will show that due to the slight mismatch of data and @xmath33 prediction at double chooz at @xmath34 mev , the @xmath25 model is favored by @xmath35 significance ; however , incorporating daya bay and reno data diminish the significance to @xmath36 c.l .. the paper is organized as follows : in section [ sec : standard3nu ] , we analyze the double chooz , daya bay and reno data in the standard @xmath33 framework .
we will show that our results are consistent with theirs within the error budget . in section [ sec : lightsterile ]
, we discuss the phenomenology of light sterile neutrinos and derive the @xmath37 survival probability in @xmath25 model .
section [ sec:3p1lightsterile ] is devoted to the analysis of data in @xmath25 model . in section [ analysis3p1dc ]
we analyze the data of double chooz , and in section [ combined ] we present the results of combined analysis of double chooz , daya bay and reno data .
our conclusions are summarized in section [ sec : con ] .
in this section we reproduce the results of the medium baseline reactor experiments ( double chooz @xcite , daya bay @xcite and reno @xcite ) in the @xmath33 framework . due to the moderately short baseline of these experiments
, the 12-induced oscillation can be ignored and the @xmath37 survival oscillation probability can be casted in the following form : @xmath38 where @xmath39 is the distance of the detector from the source , and @xmath40 is the energy of the reactor neutrinos . as discussed in @xcite , in the extraction of @xmath0 value
the effect of uncertainty in the value of @xmath19 is quite small and we fix it to the best - fit value @xmath41 measured by minos experiment @xcite .
the detection of the reactor antineutrinos is through the inverse beta decay ( ibd ) process , @xmath42 .
neutrino energy can be reconstructed by measuring the prompt positron energy @xmath43 ( neglecting the neutron recoil energy ) .
thus , by reconstructing the spectrum of the observed @xmath37 events , a fit of eq .
( [ psur ] ) to the data can give information about the value of @xmath0 .
in the following we discuss each of these experiments in detail .
the double chooz experiment has detected 8,249 candidates of electron antineutrino events with 33.71 gw.ton.years exposure using a @xmath44 detector which is located at @xmath45 m far from the reactor cores .
the total livetime of experiment is @xmath46 days .
the expected number of events in the case of no - oscillation ( _ i.e. _ , @xmath47 ) are 8,937 ( including background events ) . from a rate plus spectral shape analysis they found @xmath48 , which excludes the no - oscillation hypothesis at @xmath49 c.l .
( @xmath50 ) @xcite . to reproduce the results of double chooz
we follow the method described in @xcite .
the observed events in double chooz is separated in 18 prompt energy bins , between 0.7 mev and 12.2 mev ( fig .
14 of @xcite ) .
the analysis is performed by defining two different data - taking periods : _ i _ ) both reactors on ( 139.27 days ) with 6,088 total ibd candidates ; _ ii _ ) one reactor with less than @xmath51 of nominal power ( 88.66 days ) with 2,161 total ibd candidates .
the numbers of expected events without oscillation , background events and observed events in each bin of energy is published in @xcite .
the expected number of events for @xmath52 in the @xmath53-th bin of energy can be calculated by @xmath54 where @xmath55 is the expected number of events for @xmath47 in the @xmath53-th energy bin ( obtained from fig .
14 of @xcite , after subtracting background events ) , and @xmath56 is the averaged @xmath37 survival probability in eq .
( [ psur ] ) in the @xmath53-th energy bin . to analyze the data of double chooz we define the following @xmath57 function @xmath58\right)^2}{(\sigma^{{\rm obs}}_{i})^2+(\sigma^{{\rm osc}}_i)^{2 } } + \frac{\alpha^2}{\sigma^2_{\alpha}}+\frac{b^2}{\sigma^2_b},\end{aligned}\ ] ] where @xmath53 runs over the 36 bins of energy ( 18 for each period of data - taking ) ; @xmath59 , @xmath60 and @xmath61
are the observed , background and expected number of events in the @xmath53-th bin , respectively .
the @xmath62 and @xmath63 represent the statistical errors of the observed events and expected events with oscillation , respectively .
the systematic uncertainties in the normalization of the reactor neutrino flux and background events are taken into account by the @xmath64 and @xmath65 pull terms , with @xmath66 and @xmath67 .
.comparison between our analysis and the analyses reported by double chooz , daya bay and reno experiments .
the quantity @xmath68 shows the improvement in the fit of data due to nonzero @xmath0 . [
cols="^,^,^,^,^",options="header " , ] by marginalizing @xmath69 with respect to @xmath64 and @xmath65 , we obtain the best - fit value of the mixing angle @xmath70 , with the normalized ( to the number of degrees of freedom ) @xmath57 value of @xmath71 , which excludes the no - oscillation hypothesis at 2.7@xmath72 c.l .
( compare to the reported 2.9@xmath72 c.l . in @xcite ) .
our @xmath73 range of mixing parameter @xmath20 from double chooz is given in the first row of table [ table1 ] , which is similar to the range reported by double chooz collaboration . also , in fig .
[ fig : dchis3 ] we compare the best - fit energy spectrum of events from our analysis in the @xmath33 framework ( including the background events ) with double chooz data and the spectrum for @xmath47 , for each integration period .
the contribution of the background events is @xmath74 events in the first couple of energy bins and reduces to @xmath75 events for higher energy bins .
the left and right panels of fig .
[ fig : dchis3 ] are for the cases where both reactors are on and one reactor runs with less than @xmath51 of power ; which clearly the former plays the main role in the analysis due to the higher statistics .
the preference to the nonzero @xmath0 can be easily recognized by comparing the red solid curve for the best - fit value of @xmath0 with the blue dashed curve which shows the distribution for vanishing @xmath0 .
the daya bay reactor neutrino experiment measured the best - fit value @xmath76 , excluding the zero value at 5.2@xmath72 c.l .
@xcite . in 55 days of livetime ,
10,416 ( 80,376 ) electron anti - neutrino candidates have been detected in the far hall ( near halls ) .
the ratio of observed to expected number of events is @xmath77 . from this deficit , @xmath78 has been determined , based on a rate - only analysis .
the daya bay experiment consists of three underground experimental halls ( eh1 , eh2 and eh3 ) , where two antineutrino detectors ( ad ) are located in eh1 and one ad in eh2 ( near halls ) .
three ads are located at the far hall ( eh3 ) at a distance where @xmath37 survival oscillation probability deviates maximally from one .
daya bay collaboration have published the ratio of observed to expected number of events ( assuming no - oscillation ) for each ad ( fig . 4 of @xcite ) ; _ i.e. _ , @xmath79 where @xmath80 . using @xmath81s and the number of observed ibd candidates in each ad ( given in table ii of @xcite )
, the expected number of events without oscillation can be calculated by : @xmath82 .
the averaged oscillation probability for each ad of daya bay can be written as ( see @xcite ) @xmath83 where @xmath84 is the weighted distance of the @xmath53-th ad from the reactors ( fig . 4 of @xcite ) and @xmath85 is the fractional energy distribution of neutrinos . using the top panel of fig . 5 in @xcite for the energy distribution of events ,
the expected number of events including the oscillation for the @xmath53-th ad becomes @xmath86 we analyze the daya bay data by defining the following @xmath57 function ( which includes only the rate information ) : @xmath87 ^ 2 } { \left(\sigma_i^{{\rm stat}}\right)^2 } \nonumber \\ & + & \sum_r\frac{\alpha_r^2}{\sigma_r^2}+\sum_{i=1}^6\left(\frac{\epsilon^2_i}{\sigma^2_d}+\frac{\eta^2_i}{(\sigma^2_b)_i}\right)~ , \label{chidn } \end{aligned}\ ] ] where @xmath88 and @xmath89 are respectively the total number of observed and expected ibd candidate events in the @xmath53-th ad ; with @xmath90 representing statistical error of observed number of events which is defined as @xmath91 , where @xmath60 is the number of background events in the @xmath53-th ad .
the @xmath92 is the fraction of neutrino flux from @xmath93-th reactor at the @xmath53-th ad .
the systematic uncertainties of reactor flux , detection efficiency and background events are taken into account by the pull terms with the pull parameters @xmath94 , @xmath95 and @xmath96 respectively ; with @xmath97 , @xmath98 and @xmath99 presented in table ii of @xcite .
the parameter @xmath100 accommodates the uncorrelated flux normalization uncertainty which we marginalize without any pull term compensation .
all the values used in @xmath101 function of eq .
( [ chidn ] ) are listed in table ii of @xcite . after minimizing @xmath101 with respect to all parameters
, we find the best - fit value of the mixing parameter @xmath102 , which is consistent with the daya bay result .
the @xmath57 value and @xmath73 range of @xmath20 are shown in the second row of table [ table1 ] .
the reno experiment observed disappearance of reactor @xmath37 with @xmath103 of significance . in 229 days of data - taking period
, the number of observed neutrinos at far ( near ) detector is 17,102 ( 154,088 ) .
the ratio of the number of observed neutrinos to the number of expected neutrinos ( for @xmath47 ) is @xmath104 . from this deficit , reno collaboration obtained @xmath105 , based on a rate - only analysis @xcite .
@xmath121 versus @xmath20 , for double chooz ( red solid curve ) , reno ( blue dot - dashed curve ) and daya bay ( black dashed curve ) experiments.,scaledwidth=60.0% ] minimizing @xmath122 of eq .
( [ chireno ] ) with respect to all the pull parameters , we find the best - fit value @xmath123 , which is consistent with the reno result reported in @xcite .
the summary of all our results about reno can be seen in the third row of table [ table1 ] . in fig .
[ fig : chi2,all,3nu ] we show @xmath124 versus @xmath20 for the three discussed experiments . comparing our results with the results of the double chooz , daya bay and reno collaborations in table [ table1 ]
show that : @xmath53 ) our best - fit values for @xmath20 are fairly close to the reported values by collaborations ; @xmath125 ) our exclusion of @xmath47 ( quantified by @xmath126 ) and also @xmath73 allowed interval of @xmath20 are compatible with the corresponding values reported by the collaborations .
it worths to mention that the errors in the value of @xmath127 obtained in our analyses are moderately smaller than the errors obtained by the collaborations ( see table [ table1 ] ) .
this can be due to some optimistic assumptions and/or lack of information about the systematics .
however , the closeness of our best - fit values and fair matching of error intervals to the collaboration values , make our analysis viable . in the rest of this paper , we will use the data of these experiments to probe the existence of sterile neutrino state with mass - squared difference @xmath128 .
although the majority of data from the neutrino oscillation experiments can be interpreted consistently in the @xmath33 framework , persisting anomalies ( including miniboone , lsnd , reactor and gallium anomalies ) motivate the existence of a sterile neutrino state with mass @xmath129 ev .
although several experiments have been proposed to check the existence of @xmath130 ev sterile neutrinos ( see @xcite ) , the possibility of the existence of a light sterile neutrino ( @xmath131ev@xmath132 ) is neither excluded strongly nor planned to be explored substantially ) is studied in . ] . in this paper
we probe this possibility in the light of double chooz , daya bay and reno published data . for simplicity , we extend the neutrino sector of the standard model by adding one light sterile neutrino : the @xmath25 model . in this model the mass spectrum of neutrino sector consists of three mostly active neutrino mass eigenstates with masses @xmath133 and one mostly sterile neutrino mass eigenstate with mass @xmath134 such that : @xmath135 the mixing of these states can be described by generalizing the pmns matrix to a @xmath136 unitary matrix , @xmath137 , which can be parametrized by six mixing angles ( we assume cp - conservation in neutrino sector ) .
specifically , we adopt the following parametrization of @xmath137 : @xmath138 where @xmath139 ( i , j = 1, ... ,4 and @xmath140 ) is the @xmath136 rotation matrix in the @xmath141-plane with angle @xmath142 .
totally , four new mixing parameters are introduced in @xmath25 model : three mixing angles @xmath143 which quantify the @xmath144 , @xmath145 and @xmath146 mixings , respectively ; and one new mass - squared difference which we choose as @xmath147 . the two extra mass - squared differences @xmath148 and
@xmath149 are not independent and can be written as : @xmath150 and @xmath151 . in the medium baseline reactor experiments ( double chooz , daya bay and reno ) which we are considering in this paper , the distances of near and far detectors to the sources are a few hundred meters and @xmath152 km , respectively ; and the energy of neutrinos emitted from reactors is @xmath153 a few mev . in this
energy and baseline ranges the oscillation phase induced by @xmath154 can be ignored and the @xmath37 survival probability is given by : @xmath155 -4|u_{e3}|^2|u_{e4}|^2\sin^2\left(\frac{\delta m^2_{43}l}{4e}\right)~,\end{aligned}\ ] ] where @xmath17 and @xmath18 are the baseline and energy respectively . in terms of the parametrization of eq .
( [ uei ] ) , the oscillation probability is @xmath156 in the analysis of @xmath25 scenario , we treat the @xmath0 , @xmath157 and @xmath31 in eq .
( [ 3 + 1probability ] ) as free parameters and fix @xmath19 to its best - fit value from minos experiment @xcite . in fig .
[ fig : prob,3 + 1 ] we show the @xmath37 survival probability for different values of @xmath158 , @xmath20 and @xmath31 , versus @xmath159 .
the blue box shows the relevant values of @xmath159 for the medium baseline experiments for both near and far detectors .
the upper left panel shows the probability for the best - fit value of @xmath0 in the @xmath33 framework .
the upper right panel depicts the probability for nonzero active - sterile mixing parameters shown in the legend ; and as can be seen , for relatively large @xmath160 in this panel the effect of sterile neutrino is averaged out especially for the far detector .
the two lower panels show the probability for the best - fit values of the mixing parameters obtained in our analysis of @xmath25 ( see section [ sec:3p1lightsterile ] ) . some symmetries in eq .
( [ 3 + 1probability ] ) can be recognized .
for example , there is a degeneracy in eq .
( [ 3 + 1probability ] ) where in the limit @xmath161 ( or @xmath162 ) , when @xmath47 , the angle @xmath157 imitates the role of @xmath0 when @xmath163 .
thus , it is always possible to obtain a fit to the data in @xmath25 model as good as the fit in @xmath33 framework by setting @xmath164 and exchanging @xmath157 with @xmath0 .
however , it would be possible that deviation of @xmath31 from @xmath19 ( that is , nonzero @xmath149 ) , which leads to a shift of extrema positions in fig . [
fig : prob,3 + 1 ] , results in a better fit than the @xmath33 fit .
notice that varying @xmath0 in the @xmath33 scheme leads to a change in the depth of minima while leaving the positions of minima unchanged .
so , generally we expect to obtain better fits by extending the @xmath33 framework to @xmath25 model .
the possibility of having two independent measurements in near and far detectors ( as in daya bay and reno ) helps in lifting this degeneracy ( see section [ combined ] ) , while with one detector only ( double chooz ) the degeneracy manifests ( see section [ analysis3p1dc ] ) .
in this section we confront the data from double chooz , daya bay and reno experiments with the prediction of @xmath25 model . in section [ analysis3p1dc ] , we consider the double chooz data and in section [ combined ] , we perform the combined analysis including the data of the other two experiments . for the analysis of the double chooz data for @xmath25 model , we follow the method of section [ dc3analysis ] , with the modification that the number of expected events in the @xmath25 model is given by @xmath165 where @xmath166 is the average of survival probability in eq .
( [ 3 + 1probability ] ) in the @xmath53-th energy bin . using the same @xmath69 as in eq .
( [ chidc3 ] ) , we find the following best - fit values : @xmath167 with the minimum value @xmath168 , which shows improvement with respect to @xmath33 analysis with minimum of @xmath169 ( see the first row of table [ table1 ] ) .
the main feature of the @xmath25 analysis with double chooz data is the significantly different best - fit value for @xmath20 in eq .
( [ 3p1bf ] ) with respect to the @xmath33 best - fit value @xmath70 . in fig .
[ fig : chi2,dc3 + 1,q13 ] , we compare the @xmath69 as a function of @xmath20 for the @xmath33 and @xmath25 models .
the red dashed curve for @xmath25 is calculated by marginalizing @xmath69 with respect to @xmath31 and @xmath158 .
the shift in the best - fit value of @xmath20 in @xmath25 is clear from fig .
[ fig : chi2,dc3 + 1,q13 ] .
also , from the best - fit value @xmath170 down to zero , the @xmath57 value is nearly constant which shows the negative impact of @xmath157 on establishing a nonzero value for @xmath0 from double chooz data .
thus , as we expect , for the double chooz data active - sterile mixing parameters can mimic the effect of @xmath0 ( see section [ sec : lightsterile ] ) .
@xmath171 versus @xmath20 for @xmath25 model ( red dashed curve ) and @xmath33 framework ( blue solid curve ) , for double chooz data.,title="fig:",scaledwidth=55.0% ] [ fig : prob1 ] in the left and right panels of fig .
[ fig : chi2,dc3 + 1 ] we show the @xmath69 values as a function of @xmath158 and @xmath31 , respectively ; where the @xmath69 values are obtained by marginalization over other parameters . as can be seen from fig .
[ fig : chi2,dc3 + 1,1 ] , with the double chooz data , zero value of @xmath158 can be excluded by more than @xmath172 2.2@xmath72 .
comparing fig .
[ fig : chi2,dc3 + 1,1 ] with the red curve in fig .
[ fig : chi2,all,3nu ] shows that , as we anticipated , for double chooz data extending @xmath33 framework to @xmath25 , effectively corresponds to exchanging @xmath0 with @xmath157 . in fig .
[ fig : chi2,dc3 + 1,2 ] the best - fit value @xmath173 , shows up as a minimum in @xmath69 . the @xmath37 survival probability for the best - fit values in eq .
( [ 3p1bf ] ) , is shown in fig . [
fig : prob3 ] . to understand qualitatively the choice of best - fit values , in fig .
[ fig : his , dc3 + 1 ] we plot the energy distribution of events at double chooz detector . in this figure
the red and blue dashed curves correspond respectively to distribution of events for the best - fit values in @xmath33 and @xmath25 models .
the improvement of the fit to data for @xmath25 is fairly visible , specially for prompt energies @xmath174 mev ( due to the higher statistics , the main contribution comes from the left panel of fig . [ fig : his , dc3 + 1 ] ) .
let us take a closer look at this energy range .
for example , counting the energy bins from left side , in the seventh bin ( @xmath175 mev ) clearly the @xmath25 model matches the data better .
this energy bin corresponds to @xmath176 m / mev . comparing figs .
[ fig : prob11 ] and [ fig : prob3 ] in this range of @xmath159 shows that the @xmath37 survival probability decreases in @xmath25 model with respect to @xmath33 framework .
numerically , for the average of probability in this bin we obtain : @xmath177 and @xmath178 ; where the ratio of these two ( @xmath179 ) coincides with the ratio of red to blue dashed curves in fig .
[ fig : his , dc3 + 1,1 ] for the 7th bin of energy .
conversely , for the fifth bin of energy in fig .
[ fig : his , dc3 + 1,1 ] , we obtain @xmath180 and @xmath181 which leads to an increase in the number of events in @xmath25 model and again better fit to data .
the same improvement occurs for the other neighbor energy bins which eventually leads to a better fit in @xmath25 model with @xmath182 . by changing the value of @xmath31
, this distortion in the number of events moves to higher or lower energies where already the prediction of @xmath33 model matched with data points and consequently the fit deteriorates .
notice that although a significant difference exists between figs .
[ fig : prob11 ] and [ fig : prob3 ] in the range @xmath183 m / mev , since @xmath45 m in double chooz experiment , contribution of this range of @xmath159 , corresponding to @xmath184 mev , is quite negligible . in figs .
[ fig : allowed , dc3 + 1,2 ] and [ fig : allowed , dc3 + 1,3 ] we show the 2-dimensional allowed regions from double chooz data in @xmath185 and @xmath186 planes , respectively . as can be seen from fig .
[ fig : allowed , dc3 + 1,2 ] for @xmath187 even vanishing @xmath0 is allowed at 95% c.l .. for lower values of @xmath31 ( @xmath188 ) also @xmath47 is allowed at @xmath189 c.l . which is a consequence of @xmath190 degeneracy mentioned in section [ sec : lightsterile ] .
on the other hand , in the same range of @xmath31 larger values of @xmath191 are allowed ( see fig . [
fig : allowed , dc3 + 1,3 ] ) . at lower confidence levels ,
the closed allowed regions appear in both figs .
[ fig : allowed , dc3 + 1,2 ] and [ fig : allowed , dc3 + 1,3 ] for @xmath192 which is a result of the minimum in fig .
[ fig : chi2,dc3 + 1,2 ] .
to summarize , extending the standard @xmath33 framework by adding one light sterile neutrino state improves the fit to double chooz data and weakens the lower limit on @xmath20 such that the zero value is allowed at less than @xmath193 c.l .. also , in @xmath25 model , the allowed region and the best - fit of @xmath20 shift to smaller values . from double chooz data ,
the @xmath25 model is favored at @xmath153 2.2@xmath72 c.l . with respect to @xmath33 framework .
however , it should be noticed that even in the @xmath33 framework , the signal of nonzero @xmath20 from the double chooz experiment is rather weak ( @xmath194 ) , and as we saw in section [ sec : standard3nu ] , the inclusion of daya bay and reno data significantly enhance the signal . in this regard
, we perform a combined analysis of all data for @xmath25 model in the next section . in this section
we probe the @xmath25 model with the combined analysis of double chooz ( shape and rate ) , daya bay and reno ( rate only ) data .
the global @xmath195 is defined by : @xmath196 as described in section [ sec : standard3nu ] in eqs .
( [ chidc3 ] ) , ( [ chidn ] ) and ( [ chireno ] ) . after minimizing @xmath197 with respect to all the pull parameters , we find the following best - fit values @xmath198 with the minimum value @xmath199 . comparing with the @xmath33 framework @xmath200 , the @xmath25 leads to improvement of the fit to global data ; however , the significance of the improvement is small .
the following comments are in order about the combined analysis : * in fig .
[ fig : chi2,combined , q13 ] we show @xmath195 as a function of @xmath20 for @xmath25 model ( blue solid curve ) and @xmath33 framework ( red dashed curve ) , after marginalizing over @xmath31 and @xmath158 .
the @xmath73 range of 13-mixing angle is @xmath201 .
as can be seen , inclusion of daya bay and reno data increases the best - fit value of @xmath20 , such that the zero value of @xmath0 can be excluded by @xmath202 .
the best - fit value of @xmath0 and exclusion of @xmath47 in the combined analysis of @xmath25 model is comparable with the case of @xmath33 framework , although a bit smaller ( see table [ table1 ] ) . *
[ fig : chi2,combined , q14 ] shows @xmath195 versus @xmath158 .
comparison with fig .
[ fig : chi2,dc3 + 1,1 ] shows that inclusion of daya bay and reno data shifts the best - fit value of @xmath158 to lower values , and also the significance of nonzero @xmath157 decreases to @xmath203 .
the @xmath73 range is @xmath204 . + the @xmath57 versus @xmath31 for the @xmath25 model , from the combined data of double chooz , daya bay and reno experiments ( blue curve ) . for comparison the curve for only double chooz data
is shown ( red dashed curve ) . ]
+ + * for the mass - squared difference @xmath205 , the best - fit value of double chooz and combined analysis is the same ( see eqs .
( [ 3p1bf ] ) and ( [ 3p1allexpts ] ) , see also fig .
[ fig : combinedw ] ) .
the best - fit value @xmath182 for combined analysis originates from the double chooz data for the same reason discussed in section [ analysis3p1dc ] about fig .
[ fig : his , dc3 + 1 ] ; namely since the position of extrema in @xmath37 oscillation probability depends only on @xmath31 value . it is straightforward to check that for mixing parameters of eq .
( [ 3p1allexpts ] ) still the improvement of the fit to double chooz data in @xmath25 holds to some extend in the range @xmath34 mev .
* in figs .
[ fig : allowed , combined , q13dms14 ] and [ fig : allowed , combined , q14dms14 ] we show the allowed region in @xmath206 and @xmath207 planes respectively . from fig .
[ fig : allowed , combined , q13dms14 ] we see that the global analysis of double chooz , daya bay and reno data in the @xmath25 model excludes @xmath47 with high confidence level ( @xmath208 ) .
also , the best - fit value of @xmath209 is fairly close to the value in @xmath33 framework .
thus , the measured value of @xmath20 is robust with respect to the existence of a light sterile neutrino ; and the daya bay and reno data play an important role in this robustness ( compare fig . [
fig : allowed , dc3 + 1,2 ] and [ fig : allowed , combined , q13dms14 ] ) . in both figs .
[ fig : allowed , combined , q13dms14 ] and [ fig : allowed , combined , q14dms14 ] a closed allowed region appears at @xmath192 in low confidence levels which stem from the double chooz data .
particularly , in fig .
[ fig : allowed , combined , q14dms14 ] the closed region indicates that @xmath163 can be excluded at @xmath210 c.l . ; but , however , by increasing the significance the closed region transforms to upper limit and @xmath163 is allowed . * in fig .
[ fig : allowed , q13q14 ] we show the allowed region in @xmath211 plane for double chooz only ( left panel ) and for the combined analysis ( right panel ) . in fig . [
fig : allowed , dc , q13q14 ] ( which is for double chooz ) , there is an anti - correlation between the allowed values of @xmath0 and @xmath157 : for smaller values of @xmath0 larger values of @xmath157 are favored and vice - versa .
this anti - correlation is a manifestation of @xmath212 degeneracy discussed in section [ sec : lightsterile ] .
the break of degeneracy in low confidence levels ( red and blue curves in fig .
[ fig : allowed , dc , q13q14 ] ) is due to the mismatch of double chooz data and @xmath33 prediction in @xmath213 mev which favors larger @xmath157 . however , by including the daya bay and reno data in fig .
[ fig : allowed , combined , q13q14 ] ( with the advantage of making two independent measurements in near and far detectors of each experiment ) the anti - correlation and degeneracy break and end up in two islands with the characteristic that both of them favor nonzero @xmath0 . for the first island : @xmath214 , while for the second one : @xmath215 ; where the latter originates from double chooz contribution . in both panels of fig .
[ fig : allowed , q13q14 ] the green dashed curve represents the limit from the solar and kamland data @xcite . *
finally we would mention that our results are robust with respect to the uncertainty in the value of @xmath216 , since varying this parameters will change just the positions of the extrema in fig .
[ fig : prob11 ] while the improvement to the data requires additional extrema as in fig .
[ fig : prob4 ] . as a cross check
, we have tested the stability of our results by varying @xmath216 within their @xmath73 uncertainty ranges from the minos experiment @xcite .
the negligible change in our results justifies the initial assumption of fixed @xmath19 .
searches for the sterile neutrinos and investigating its impact on experimental results obtained , or planned to be obtained , is one of the cutting edge questions in neutrino physics .
although the initial motivation was interpretation of lsnd anomaly by sterile neutrinos with mass @xmath217 ev , further anomalies such as cosmological hints and low energy solar data stimulated the existence of lighter sterile neutrinos . in this paper
we investigated the impact of existence of a light sterile neutrino , @xmath25 model , on the medium baseline reactor experiments : double chooz , daya bay and reno .
the baseline and energy of these experiments provide the opportunity to probe active - sterile oscillation with @xmath218 . among these three experiments ,
the double chooz consists of one detector and present both rate and shape ( in energy ) information of observed events ; while the daya bay and reno experiments , each equipped by near and far detectors , provide the deficit in the total number of events in far detector(s ) with respect to near detector(s ) .
the energy distribution of events in double chooz is in good agreement with @xmath33 framework prediction except in the range @xmath219 mev .
this discrepancy leads to a better fit in @xmath25 such that we obtained the best - fit values @xmath220 , @xmath221 and @xmath222 for mixing parameters . with the double chooz data the @xmath33 framework can be excluded at @xmath35 c.l ..
also , the best - fit value of @xmath0 angle is significantly different than the reported value in @xmath33 framework and @xmath47 is allowed in less than @xmath73 c.l .. inclusion of the rate information from daya bay and reno experiments alters the conclusion such that with the combined data of double chooz , daya bay and reno we obtain the best - fit values @xmath223 with the combined data the @xmath25 model is favored at @xmath36 c.l .. the value of @xmath0 angle is close to the reported value in @xmath33 framework and so robustness of @xmath0 determination can be claimed .
the persisting @xmath212 degeneracy in @xmath25 , which exists in the limit @xmath162 , can be lifted mainly from the data of daya bay and reno . despite the preference for @xmath25 model , a large part of the parameter space of this model
is excluded in our analysis , better than the other constraints by a factor of 2 .
the planned near future data from these experiments can significantly exclude / strengthen the favored nonzero active - sterile mixing parameters found in this paper .
especially , the energy spectrum of data in daya bay and reno experiments can decisively rule out / confirm it .
also , installation of near detector in the double chooz experiment can provide valuable information about the observed anomaly in @xmath219 mev , wether supporting it or contradicting it .
_ note added 1 : _ after completion of this work , the energy spectrum of daya bay data have been reported @xcite .
interestingly , an anomalous behavior , similar to the double chooz data , in the energy range @xmath224 mev can be identified . however , a detailed analysis taking into account both daya bay and double chooz data is worthwhile .
_ note added 2 : _ before submitting this paper , the article @xcite appeared on arxiv with results close to the results in this paper .
however , in @xcite just the rate analysis of the double chooz data is considered and consequently the @xmath224 mev anomaly mentioned in this paper does not show up in @xcite .
o. l. g. p. thanks the ictp and the financial support from the funding grant 2012/16389 - 1 , so paulo research foundation ( fapesp ) .
a. e. thanks the financial support from the funding grant 2009/17924 - 5 , so paulo research foundation ( fapesp ) .
thanks support from cnpq ( grant 311102/2011 - 5 ) and fapesp ( grant 2010/07359 - 6 ) .
we would like to thank antonio palazzo for providing us the information about kamland+solar limit .
a. aguilar - arevalo _ et al . _
[ lsnd collaboration ] , phys .
d * 64 * , 112007 ( 2001 ) [ hep - ex/0104049 ] .
a. a. aguilar - arevalo _ et al .
_ [ miniboone collaboration ] , phys .
lett . * 110 * , 161801 ( 2013 ) [ arxiv:1207.4809 [ hep - ex ] , arxiv:1303.2588 [ hep - ex ] ] .
g. mention , m. fechner , t. lasserre , t. a. mueller , d. lhuillier , m. cribier and a. letourneau , phys .
d * 83 * , 073006 ( 2011 ) [ arxiv:1101.2755 [ hep - ex ] ] .
j. kopp , p. a. n. machado , m. maltoni and t. schwetz , jhep * 1305 * , 050 ( 2013 ) [ arxiv:1303.3011 [ hep - ph ] ] . c. giunti , m. laveder , y. f. li and h. w. long , arxiv:1308.5288 [ hep - ph ] .
k. n. abazajian , m. a. acero , s. k. agarwalla , a. a. aguilar - arevalo , c. h. albright , s. antusch , c. a. arguelles and a. b. balantekin _ et al .
_ , arxiv:1204.5379 [ hep - ph ] .
a. esmaili and a. yu . smirnov , arxiv:1307.6824 [ hep - ph ] ; a. esmaili , f. halzen and o. l. g. peres , jcap * 1307 * , 048 ( 2013 ) [ arxiv:1303.3294 [ hep - ph ] ] ; a. esmaili , f. halzen and o. l. g. peres , jcap * 1211 * , 041 ( 2012 ) [ arxiv:1206.6903 [ hep - ph ] ] ; a. esmaili and o. l. g. peres , phys . rev .
d * 85 * , 117301 ( 2012 ) [ arxiv:1203.2632 [ hep - ph ] ] ; s. razzaque and a. yu . smirnov , jhep * 1107 * , 084 ( 2011 ) [ arxiv:1104.1390 [ hep - ph ] ] .
p. huber and t. schwetz , phys .
d * 70 * , 053011 ( 2004 ) [ hep - ph/0407026 ] .
a. bandyopadhyay and s. choubey , arxiv:0707.2481 [ hep - ph ] .
a. de gouvea and t. wytock , phys .
d * 79 * , 073005 ( 2009 ) [ arxiv:0809.5076 [ hep - ph ] ] .
k. bora , d. dutta and p. ghoshal , jhep * 1212 * , 025 ( 2012 ) [ arxiv:1206.2172 [ hep - ph ] ] .
e. ciuffoli , j. evslin and h. li , jhep * 1212 * , 110 ( 2012 ) [ arxiv:1205.5499 [ hep - ph ] ] .
b. bhattacharya , a. m. thalapillil and c. e. m. wagner , phys .
d * 85 * , 073004 ( 2012 ) [ arxiv:1111.4225 [ hep - ph ] ] .
s. k. kang , y. d. kim , y. ko and k. siyeon , arxiv:1303.6173 [ hep - ph ] .
a. palazzo , phys .
d * 83 * , 113013 ( 2011 ) [ arxiv:1105.1705 [ hep - ph ] ] ; a. palazzo , phys .
d * 85 * , 077301 ( 2012 ) [ arxiv:1201.4280 [ hep - ph ] ] .
p. bakhti and y. farzan , arxiv:1308.2823 [ hep - ph ] .
talk by j. p. ochao at xxiv workshop on weak interactions and neutrinos " , accessible from : http://indico.cern.ch/getfile.py/access?contribid=55&sessionid=1&resid=0&materialid=slides&confid=234536 | medium baseline reactor experiments ( double chooz , daya bay and reno ) provide a unique opportunity to test the presence of light sterile neutrinos .
we analyze the data of these experiments in the search of sterile neutrinos and also test the robustness of @xmath0 determination in the presence of sterile neutrinos .
we show that existence of a light sterile neutrino state improves the fit to these data moderately .
we also show that the measured value of @xmath0 by these experiments is reliable even in the presence of sterile neutrinos , and the reliability owes significantly to the daya bay and reno data . from the combined analysis of the data of these experiments
we constrain the mixing of a sterile neutrino with @xmath1 to @xmath2 at @xmath3 c.l .. |
the discovery of pulsed emission from the representative ultra - luminous x - ray sources ( ulx ) , m82 x-2 , by @xcite was really sensational , since it for the first time provided good piece of information regarding the mass of the central objects in the ulxs .
the ulxs are off - nuclear , compact , very luminous x - ray sources with x - ray luminosities exceeding @xmath5 erg s@xmath6 .
there has been a long debate regarding the central engine of the ulxs : it could be either sub - critical accretion onto an intermediate - mass black hole ( see e.g. , * ? ? ?
* ; * ? ? ?
* ) , or super - critical accretion onto a normal stellar - mass black hole ( see e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , or the combination of these two .
although the direct mass measurement is the most effective way to settle down this issue , it is not always feasible due mainly to technical difficulties .
in fact , ulxs are mostly very faint because of large distances .
that is , there had been no direct evidences in favor of or against the super - critical accretion scenario for ulxs until the discovery of the ulx - pulsar .
since it is obvious that the mass of a neutron star ( ns ) is less than @xmath7 2 m@xmath8 , the discovery of the ulx - pulsar robustly points to the occurrence of super - critical accretion , at least , in some ulxs .
( we can not ruled out a possibility that the ulxs may be a heterogeneous group . )
the discovery of the ulx - pulsar has re - opened the issue of super - critical accretion onto a magnetized ns .
this is not totally a new issue but was already intensively discussed in the 1970 s , being pioneered by ( * ? ? ?
* hereafter , bs76 ) .
they solved one - dimensional structure of an accretion column , finding that the column is composed of two regions : an upper , nearly free - fall region and a lower settling region .
they also claimed luminosities can naturally exceed the eddington luminosity , since copious photons produced via dissipation of kinetic energy of accreting gas propagate towards side boundaries and can freely escape .
large amounts of photons can thus leak away from the side of the column .
bs76 mainly addressed the super - critical accretion problem assuming hollow columns , and the limiting luminosity sensitively depends on the geometry of the accretion funnel , i.e. , the thickness parameter of the wall of the column .
the two dimensional gas motion was not taken into account , and the super - critical column accretion without assuming the hollow structure is not well understood .
in addition , they assumed a radiative diffusion approximation to evaluate radiative flux at the side surface of the accretion column , which may underestimate the amount of the photons escaping from the side wall of the column . in this _
, we mainly address the following questions : ( 1 ) how is super - critical accretion through a narrow accretion column on a star feasible ? ( 2 ) can the luminosity of the ulx - pulsar be explained ?
( 3 ) what observational signatures are expected for such a source ? ( 4 ) are there any substantial differences between the accretion processes onto black holes and those onto nss ? in relation to these questions , it is of great importance to note that @xcite solved the problem of super - critical accretion disk onto non - magnetized nss for the first time , by means of the two - dimensional radiation - hydrodynamic ( rhd ) simulations .
his main conclusions are that ( 1 ) super - critical accretion is indeed feasible and that ( 2 ) a large fraction of the liberated energy goes to kinetic energy in ns accretion .
by contrast , more fraction of energy goes to radiation in black hole accretion . in the case of accretion onto strongly magnetized nss , however , flow structure is distinct from that onto non - magnetized nss : the strong magnetic fields of the nss penetrating the accretion disks can disrupt the disk accretion and polar accretion flows aligned with the b - fields of the nss can be formed inside alfv@xmath9n radii . to examine whether the situation changes or not in the case of column accretion ,
we solve the gas dynamics and radiation transfer inside the funnel by two - dimensional rhd simulations .
we expect much larger radiation output from the side wall than that from the top of accretion column , thus very high accretion luminosities being expected .
such geometrical differences will be essential from the viewpoints of radiation hydrodynamics , as will be demonstrated in this _
we apply the simulation by @xcite to the case of ns accretion through a narrow accretion column ( or funnel ) .
we set a calculation box with a cone shape having a half opening angle of @xmath10 on the ns surface and fill it with uniform tenuous gas with density of @xmath11 and high temperature of @xmath12 k. we further assume that the initial radiation temperature surrounding the column is low , @xmath13 k , and that gas has no angular momentum , for simplicity .
the mass of a ns is set to be @xmath14 .
since the gas layer within the column is not in hydrostatic balance , gas starts to fall onto the lower boundary after the start of a simulation .
we calculate two dimensional gas motion and radiation transfer by solving a set of rhd equations in the spherical coordinates ( @xmath4 , @xmath15 , @xmath16 ) , assuming axisymmetry .
the basic equations are @xmath17 @xmath18 + \frac{\chi}{c}f^{r}_{0},\ ] ] @xmath19 @xmath20 @xmath21 here , @xmath22 is the mass density , @xmath23 is the velocity ( we set @xmath24 , assuming no angular momentum of gas ) , @xmath25 is the gas pressure , @xmath26 is the internal energy density of gas , @xmath27 is the blackbody intensity , @xmath28 is the radiation energy density of gas , @xmath29 is the radiative flux , @xmath30 is the radiation pressure tensor ( all of the quantities with suffix @xmath31 are measured in the comoving frame of the fluid ) , @xmath32 is the absorption opacity , and @xmath33 is the total opacity , where @xmath34 is the cross section of thomson scattering and @xmath35 is the proton mass .
we do not take into account thermal compton cooling / heating , which , if included , might change the temperature structure near the ns surface , though the overall structure should not alter significantly because of a very large optical depth of the column .
general relativistic effects are incorporated by a pseudo - newtonian potential prescribed by , @xmath36 . here
, @xmath37 is the schwarzschild radius , @xmath38 is the gravitational constant , and @xmath39 is the speed of light .
simulations are carried out using a rhd code @xcite , which solves the radiative transfer by adopting the flux - limited diffusion ( fld ) approximation @xcite and hydrodynamic equations by using virginia hydrodynamics one ( vh-1 ) based on lagrange - remap version of the piecewise parabolic method ( ppm , * ? ? ?
* ) , which is the third order godunov s scheme .
the size of the computational domain is @xmath40 @xmath41 @xmath4 @xmath41 @xmath42 and @xmath43 @xmath41 @xmath44 @xmath41 @xmath45 , where @xmath46 and @xmath47 ; that is , we set @xmath48 km , typical ns radius .
we set a rather long accretion column ( with a height of 500 @xmath49 ) so as to include the trapping radius , which is expected to be at @xmath50 @xcite , within the calculation box .
( we will show later @xmath51 . )
however , we focus our discussion in a narrow region , within 40 km ( @xmath52 ) . the number of grid cells is @xmath53 .
the grid points are distributed such that @xmath54 and @xmath55 . at @xmath56
the accreting gas is absorbed , however , we assume that the energy carried by the gas is immediately converted to the radiation and is released as outgoing radiation flux @xcite .
the top of the computation box at @xmath57 is set to be a free boundary , where both inflow and outflow of gas is allowed , i.e. , the fluid quantities with no gradient are assigned to the ghost cells at every time step .
we impose the mirror boundary condition both at @xmath58 and @xmath45 for the gas .
as for the radiation , we set the mirror boundary at @xmath58 , while we let radiation go out through the boundary at @xmath59 with radiation flux , @xmath60 .
we set the initial mass density to be @xmath61 so as to reproduce the luminosity of m82 x-2 , @xmath62 for neutron stars with @xmath63 , i.e. , @xmath64 . here
, we have assumed the radiative efficiency @xmath65 . by using the relation @xmath66
, we can obtain @xmath67 @xmath68 for a radius inside which matter can accrete enough by the end of our simulation @xmath69 , free - fall velocity there @xmath70 , and the solid angle subtended by the accretion column @xmath71 .
in fact , we confirm that the simulated luminosity is @xmath72 , which is consistent with the estimated luminosity of m82 x-2 .
as was mentioned before , gas starts to freely fall towards the ns immediately after a simulation starts , since there is no sufficient pressure force preventing the gas fall in the initial state .
the mass accretion rate onto the ns shows @xmath73 , as expected above .
[ ! th ] we first show snapshots displaying gas and radiation properties over the two - dimensional plane in figure [ fig1 ] .
these are color contours of matter density overlaid with gas velocity [ panels ( a ) and ( c ) ] and those of radiation energy density overlaid with comoving radiation flux [ panels ( b ) and ( d ) ] , respectively .
the right two panels are magnified views of the innermost parts of the corresponding left panels .
two - zone structure is very clear in these plots .
the interface between the two zones is located at around @xmath74 , where a shock structure is observed there ; velocity vectors are uniformly downward in the upper free - fall region , whereas they show significant deviations from the uniform down - flow below the interface .
even circular motion is observed there .
another important feature found in this figure is that most of radiation originates from the side wall of the lower settling region of the accretion column , not from the top of the column .
interestingly , the luminosity of the ns surface is kept barely below @xmath75 so that mass accretion onto the ns can proceed , liberating accretion energy sideway .
the inset figure in panel ( d ) of figure [ fig1 ] clearly shows that the radiation flux profile @xmath76 is nearly flat along the side wall below the interface .
we see that the fractional luminosity @xmath77 exceeds the eddington luminosity ( @xmath78 erg s@xmath6 ) at @xmath79 , where @xmath80 with @xmath81 being the mesh spacing ( @xmath82 km near the ns surface ) .
the total luminosity amounts to @xmath83 erg s@xmath6 @xmath84 , as will be explicitly shown later . in order to check how physical quantities vary as matter falls in the settling region , we plot the azimuthally averaged gas density and velocity in figure [ fig2 ] .
we find that mater density gradually increases downward by about two orders of magnitudes , while the absolute value of gas velocity decreases downward by roughly the same orders of magnitudes .
these gradual changes arise , since we azimuthally averaged the physical quantities , and since the shock surface is not parallel to ns surface but is tilted ( see figure [ fig1 ] ) . note also that the radiation entropy displayed in the bottom panel shows a hump with a plateau shape below the shock region , indicating that energy dissipation really occurs there and that dissipated energy directly goes to radiation with going very little to gas .
( note that the gas entropy is very small , compared with the radiation entropy . )
the plateaus shape indicates that the radiation gains entropy from gas and loses entropy by outgoing radiation at similar rates . [ !
t ] sec .
each quantity is averaged in the @xmath15 direction , e.g. , @xmath85 .
a weak shock formed at @xmath86 km separates the upper , nearly free - fall region ( shaded region ) and the settling region , in which gas is drastically decelerated by the radiation pressure and matter is thus being accumulated .
( bottom ) same as the top panel but for gas entropy and radiation entropy . ]
we have so far confirmed that super - critical accretion actually occurs through the accretion column .
the next issue is to clarify how it occurs despite huge radiation energy density accumulating on the ns surface . for this purpose ,
we examine how the energy conversion occurs from gas to radiation by using figure [ fig3 ] . here , the plotted quantities are either of volume - integrated energy density or surface - integrated energy flux density defined as : @xmath87 , and @xmath88 , where @xmath89 is energy density , @xmath90 is energy flux density and integration is made from radius @xmath4 to @xmath91 km .
the selected quantities are gravitational energy ( @xmath92 ) , advection energy of gas ( @xmath93 ) , time derivative of gas energy density ( @xmath94 ) , radiation energy density converted from gas ( with two different expressions , @xmath95 and @xmath96 ) , time derivative of radiation energy density ( @xmath97 ) , advection energy of radiation ( @xmath98 ) , and radiation energy flux density emitted from the unit area of the side wall ( @xmath99 ) , respectively , where @xmath100 , @xmath101 , @xmath102 , @xmath103 , @xmath104 , and @xmath105 , respectively . [ !
t ] sec . ]
let us first see how gas energy changes its form in the upper panel of figure [ fig3 ] . as gas falls down ,
the gravitational energy is gradually released ( see the line labeled with @xmath106 ) .
however , the sum of the liberated energy and the advection energy is kept nearly constant until gas reaches the shock surface at @xmath107 km , meaning that all the liberated energy is carried downward by the falling gas . since photons near the ns
push the plasma outward , falling gas is suddenly decelerated at the shock surface . below the shock surface , a part of the liberated energy is converted to radiation energy ( see the line labeled with @xmath108 ) . as a result
, the gas energy @xmath109 is kept nearly constant in time , which guarantees the existence of stationary flow .
let us next see how radiation energy changes its form in the lower panel of figure [ fig3 ] . as mentioned above , gas energy is converted to radiation energy below @xmath110 ( see the line labeled with @xmath111 ) .
nearly the same amount of energy is radiated away through the side boundary ( see the line labeled with @xmath112 ) .
that is , the converted energy to radiation is carried to the side boundary and is radiated away at a similar height .
we note , however , that the time derivative of the radiation energy density ( @xmath97 ) is non - zero below @xmath110 and is nearly balanced with the advection of radiation energy .
this occurs because of convective motion of radiation bubbles [ see circulating vectors in fig . 1 ( c ) and ( d ) ] .
the matter under the shock slowly accretes onto the ns , and emit the photons , whose luminosity is nearly eddington luminosity .
we note that isotropic luminosity at the ns surface is @xmath113 , which can be obtained by multiplying the luminosity at the ns surface within the accretion column @xmath114 by the correction factor @xmath115 .
this isotropic luminosity is consistent with that of @xcite .
in this @xmath116 we study the two dimensional structure of a narrow super - critical accretion column with radiation transfer by means of the axisymmetric radiation hydrodynamic simulation .
the accretion column is composed of an upper , free - fall region and a lower settling region with a shock being formed at the interface .
gas energy is converted to radiation only below the shock surface , and it is finally radiated away from the side surface .
super - eddington luminosity is produced there , while nearly eddington luminosity is emitted from the ns surface .
such basic energy flow is summarized in figure [ fig4 ] .
there are similarities and differences among three super - critical accretion systems ( see table 1 ) .
the most distinctive features are found in the final form of energy ; that is , it differs among them how excess energy can be taken away from the central engine at a rate exceeding the classical eddington limit .
excess energy is carried by trapped photons in the black hole disk accretion , by outflow material in the form of its kinetic energy in disk accretion onto a non - magnetized ns @xcite , and by outgoing radiation emitted from the side boundary in column accretion onto a magnetized ns . in any case the outward flux measured in the laboratory frame ( @xmath117 ) is much reduced in the innermost part of accretion flow .
this condition makes super - critical accretion feasible .
since most of excess energy is finally radiated away from the side boundary in the magnetized ns case , strong electromagnetic radiation should be finally observed , unlike other two cases .
the most unique feature of the ulx - pulsar is highly anisotropic radiation fields .
this occurs , since large luminosity radiation goes out through the side boundary and not from the top of the accretion column .
if the rotation axis and the magnetic axis of the ns do not coincide , we thus expect pulsed emission to be observed . in the present study , we observe circulating gas motion in the settling region ( see figure [ fig1 ] ) .
such motion affects energy transport processes and is described by the photon advection term in our analysis ( see the bottom panel of figure [ fig3 ] ) .
such gas motion , known as photon bubbles , was intensively studied by @xcite by means of the rhd simulations .
they carefully solve the super - eddington atmosphere on the magnetized ns and found strong radiation - driven , optically thin outflow embedded in optically thick , inflowing plasma .
they estimated that the growth timescale of such photon bubbles are is on the order of a millisecond .
they also observe rapid time variability on millisecond timescales .
@xcite made the linear analysis , finding that the growth time is inversely proportional to the wavenumber , thereby concluding that the layer is eventually dominated by a few large bubbles .
these features are in good agreement with our results . despite the existence of internal gas circulation
, the fractional luminosity @xmath118 is remarkably uniform except in the innermost region and thus seems not to be affected by photon bubbles [ see the inserted figure next to figure 1(d ) ] .
the photon bubbles are time - dependent , thus photon advection directly leads to time changes of radiation energy density ( see the bottom panel of figure [ fig3 ] ) . such time dependence is smoothed on long terms and , hence , does not affect the fractional luminosity profile . in the present study , we assume that magnetic fields are not so strong so that the two - dimensional gas motion can be allowed within the column
. that is , in our model , both of radiation - pressure and ram - pressure forces are comparable to the lorentz force and transport the radiative energy by the advection , while the accretion column is possibly sustained due to marginally strong lorentz force at the side surface of the column . by using our simulation data ,
this marginal b - field strength inside the column can be estimated to be @xmath119 g , which is not anomalous in pulsars .
the perfect confinement of the column may look slightly difficult to be sustained , since the radiative force can dominate the lorentz force with the marginal magnetic fields at the side wall .
however , the column accretions as shown in our simulation should be still feasible for the following two reasons : ( 1 ) the magnetic fields inside the column will be pushed towards the side wall due to the fluid motion with the frozen - in magnetic fields , leading to accumulation of the b - field at the column side . then the enhanced magnetic tension force and magnetic pressure gradient force can overcome the radiative force .
( 2 ) even if the perfect confinement of the column is not sustained , the mass loss due to radiative force in the @xmath15-direction at the side wall can not significantly change our result , since the sound speed of the radiation dominated gas at the side surface is @xmath120% of the accretion velocity , indicating that the mass loss rate through the column side will only be @xmath121% , at maximum , of mass accretion rate .
we plan to explore as future work the cases of much stronger magnetic fields which forbid the fluid motions perpendicular to the magnetic field ( i.e. , the @xmath15-direction ) . here
, we discuss the location of the photosphere . in the @xmath4-direction ,
the scattering photosphere and the effective photosphere appear at @xmath122 km and @xmath123 km , respectively , inside from the outer boundary ( @xmath124 km ) .
both of them are thus located far outside the shock surface ( @xmath107 km ) . in the @xmath15-direction ,
the scattering photosphere is at around @xmath125 km from the side boundary at @xmath126 .
although the column is effectively optically thin ( @xmath127 ) at @xmath128 km if the gas temperature is kept high , the effective optical depth is more likely to be @xmath129 if comptonization is properly taken into account ; that is , the matter temperature will then be much cooler @xmath130 k ( i.e. , gas temperature will decrease to radiation temperature via the comptonization ) .
that is , the effective photosphere will appear around 0.5 km from the side wall , if the gas is strongly comptonized .
the comptonization timescale is , in fact , much shorter than the accretion timescale as briefly discussed below . in the present study
, we rather simplified the physical situation to focus on the basic observational features of the super - critical accretion column , thus omitting some potentially important physical processes .
for example , we ignored thermal compton scattering , which could be of great importance when considering detailed radiation spectra @xcite because the plasma is scattering dominant and gas temperature is much greater than radiation temperature under the shock surface .
( we note that the fluid dynamics found in this _ letter _ will not be significantly affected by the comptonization as described below . ) at first
, we show the importance of the comptonization to determine gas temperature distribution , and then , we present the reason why the dynamics will not be significantly modified by the comptonization . under the shock surface ,
gas temperature is orders of magnitude higher than radiation temperature .
the comptonization , which relaxes the gas and radiation temperature , will sufficiently change the gas temperature distribution , because the estimated comptonization timescale ( @xmath131s ) is by seven orders of magnitude shorter than the accretion timescale(@xmath125 s ) . here
, the comptonization timescale is estimated by using the compton cooling rate based on the kompaneets equation ( see , e.g. the equation ( 3 ) in * ? ? ?
* ) , substituting gas temperature @xmath119k , radiation temperature @xmath130k , and mass density @xmath132 .
thus , we can expect that the gas temperature will significantly decrease due to the comptonization . despite the importance of the compton scattering
, we can expect that the dynamics of the flow will not be significantly changed by the comptonization .
this is because the flow dynamics is governed by the radiative force and the radiation energy density is 1 - 3 orders of magnitude larger than the gas internal energy density below the shock surface , which indicates that the increment of the radiation energy converted from the gas internal energy via the comptonization is too small to significantly change the dynamics of the accretion column .
the comptonization may be , therefore , important only when we calculate the spectra of the accreting column : we expect that a modified - blackbody spectrum with a wien peak will be formed , since the accretion column is scattering dominant and the compton _
y_-parameter is much greater than unity under the shock surface . in this _
letter _ , three dimensional gas motion ( due to the ns rotation , etc . ) is also ignored .
further , we used the fld approximation for solving radiation transfer , which need to be improved .
we also simplify the plasma processes near the ns ; that is , we have not considered neutrino emission nor included corrections to hydrodynamics and radiative transport due to strong magnetic fields @xcite .
these effects will be explored in a next paper .
we thank h.r .
takahashi for useful discussion .
the numerical simulations were carried out on the xc30 at the center for computational astrophysics , national astronomical observatory of japan .
this work is supported in part by grants - in - aid of the ministry of education , culture , sports , science and technology ( mext ) ( 26400229 , s.m . ;
15k05036 , k.o . ) .
this work was supported in part by mext hpci strategic program and the center for the promotion of integrated sciences ( cpis ) of sokendai , and mext as a priority issue ( elucidation of the fundamental laws and evolution of the universe ) to be tackled by using post - k computer and jicfus . | prompted by the recent discovery of pulsed emission from an ultra - luminous x - ray source , m82 x-2 ( ulx - pulsar ) , we perform a two - dimensional radiation - hydrodynamic simulation of a super - critical accretion flow onto a neutron star through a narrow accretion column .
we set an accretion column with a cone shape filled with tenuous gas with density of @xmath0 above a neutron star and solve the two dimensional gas motion and radiative transfer within the column .
the side boundaries are set such that radiation can freely escape , while gas can not .
since the initial gas layer is not in a hydrostatic balance , the column gas falls onto the neutron - star surface , thereby a shock being generated . as a result
, the accretion column is composed of two regions : an upper , nearly free - fall region and a lower settling region , as was noted by basko & sunyaev ( 1976 ) .
the average accretion rate is very high ; @xmath1 ( with @xmath2 being the eddington luminosity ) , and so radiation energy dominates over gas internal energy entirely within the column . despite the high accretion rate , the radiation flux in the laboratory frame is kept barely below @xmath3 at a distance @xmath4 in the settling region so that matter can slowly accrete .
this adjustment is made possible , since large amount of photons produced via dissipation of kinetic energy of matter can escape through the side boundaries .
the total luminosity can greatly exceed @xmath2 by several orders of magnitude , whereas the apparent luminosity observed from the top of the column is much less .
due to such highly anisotropic radiation fields , observed flux should exhibit periodic variations with the rotation period , provided that the rotation and magnetic axes are misaligned . |
quasi one dimensional two - leg ladders sr@xmath3ca@xmath4cu@xmath5o@xmath6 ( ca@xmath7 ) have been intensively investigated over the last years due to their fascinating physical properties.@xcite spin ladders sr@xmath3ca@xmath4cu@xmath5o@xmath6 have a spin - liquid ground state with a finite spin - gap@xcite and antiferromagnetic ( af ) long - range order@xcite coexisting with spin singlets and hole pairs at low @xmath11 .
furthermore , for samples with the largest ca concentration , an applied pressure in the range 30 - 80 kbar stabilizes a superconducting ground state in the ladder planes with a transition temperature passing through a maximum at 10 k around 40 kbar .
@xcite consequently , clarifying the interplay between spin and charge degrees of freedom in doped antiferromagnets is a very important matter for understanding the onset of a superconducting state in spin ladders as well as in high-@xmath12 superconductors .
the sr@xmath3ca@xmath4cu@xmath5o@xmath6 system is intrinsically hole doped as the stoichiometry implies an average copper valency of 2.25 .
the redistribution of the preexisting holes among chains and ladders in this composite system , depending on the ca content , applied pressure and temperature is one of the most important factor which controls the physical properties ( in particular superconductivity ) of these spin - ladders .
optical measurements@xcite and the calculation of the madelung potentials @xcite have shown that for the @xmath13 compound holes are staying essentially in the chains providing a valency of 2.06 for the ladder cu sites .
their localization in the chains leads to an insulating behavior .
in addition , upon ca substitution holes are transferred from the chains to the ladders .
@xcite at the same time , the longitudinal conductivity ( along the c - axis ) increases , leading to quasi one dimensional ( q-1-d ) metallic properties .
@xcite this hole transfer can be caused by the reduction of the distance between chains and ( sr , ca ) layers , @xcite which results in an enhancement of the electrostatic potentials in the chains.@xcite experimentally , however , the precise amount of hole transfer is still under discussion .
osafune _ et al_.@xcite have found that the copper valency in the ladders increases from @xmath14 up to @xmath15 ( @xmath16 holes per cu ) when @xmath7 increases from @xmath17 to @xmath18 .
the systematic measurements of the @xmath0cu nuclear spin - spin relaxation time @xmath19 ( ref . ) probing the spin correlation length has shown that holes were doped into cu@xmath8o@xmath9 ladder with a content of @xmath20 , 0.22 and 0.25 holes per cu for @xmath7=6 , 9 and 11.5 , respectively .
recently , isobe _
et al_.@xcite have investigated the temperature dependence of the crystal structure of sr@xmath21ca@xmath22cu@xmath5o@xmath6 ladders by the rietveld analysis of neutron diffraction data . using the bond - valence sum calculation they have estimated the hole concentration in ladder planes as @xmath23 per cu1 , _ i.e. _ about three times smaller than what was reported in ref .
. moreover , isobe _
et al_.@xcite found that holes located in the ladders tend to move back into the chains at low temperature and that almost all holes were located in the chains near the liquid - he temperature .
a similar estimate for the hole content @xmath24 for the ca12 compound has been recently deduced from polarization - dependent @xmath25-ray absorption data by ncker _
et al_.@xcite with a decreasing @xmath7 , @xmath25-ray data indicate only a marginal decrease of @xmath26 for ca0 .
more recently , thurber _
et al_.@xcite have reported nmr measurements of the @xmath1o nuclear quadrupole interactions @xmath27 on the ca@xmath7 ( @xmath7=0 , 3 , 8) .
they have related the changes in @xmath27 to the variation in the effective hole concentration @xmath28 in the cu@xmath8o@xmath9 ladder layer and have found a decrease of the hole number in cu@xmath8o@xmath9 going to low temperature for all @xmath7 values .
in addition , they have noticed only a small increase of the hole density of @xmath29 in the ladders layer going from ca0 to ca8 .
the main effect of ca doping on the structure of sr@xmath3ca@xmath4cu@xmath5o@xmath6 is a reduction of the @xmath30 parameter .
a similar conclusion is reached by analyzing the effect of hydrostatic pressure.@xcite it is the @xmath30 parameter which is most sensitive to pressure .
hence , we can expect an increase of the ladder hole density to take place as pressure is increased .
this assumption is in agreement with the decrease of the electrical resistivity observed in refs . under high pressure .
experimentally , however , quantitative estimates for the hole number transferred into the ladder subsystem under high pressure are still missing . in order to address such a question the present work attempts to provide an experimental determination of the hole distribution among ladder and chain subsystems as a function of temperature , doping , and pressure .
the latter point is crucial since pressure is a prerequisite for the stabilization of superconductivity in sr@xmath3ca@xmath4cu@xmath5o@xmath6 .
we present detailed copper and oxygen nmr studies of the nuclear quadrupole interaction tensor , @xmath2 , in ca0 and ca12 single crystal samples at ambient pressure and under the pressure of 32 kbar which can drive the ca12 system to the situation where superconductivity is stabilized at low temperature .
the measurements of @xmath2 were performed at the different ion sites cu1 , o1 and o2 ( see fig .
1 ) in the cu@xmath8o@xmath9 ladder plane .
the part concerning the ambient pressure data is consistent with the work of previous workers .
@xcite our results emphasize an additional transfer of holes from chains to ladders when pressure is applied in highly ca - substituted samples .
o nmr spectrum in sr@xmath31cu@xmath5o@xmath6 for @xmath32 measured at @xmath11=45 k under ambient pressure ( solid line ) and the results its simulation ( dotted line ) . ]
nmr studies were carried out on monocrystalline samples of sr@xmath31cu@xmath5o@xmath6 ( ca0 ) and sr@xmath8ca@xmath33cu@xmath5o@xmath6 ( ca12 ) grown by the traveling solvent floating zone method .
@xcite the samples were enriched with @xmath1o isotope as described elsewhere.@xcite @xmath0cu(@xmath34 ) and @xmath1o(@xmath35 ) nmr measurements have been performed at a field of @xmath36 tesla with the usual fourier transform method .
the components of the quadrupole interaction tensor , @xmath37 were determined from @xmath1o and @xmath0cu nmr spectra recorded for the different orientations of the single crystals in the external magnetic field using a simulation software taking into account quadrupolar corrections to the zeeman frequency in a second order perturbation theory . a typical @xmath1o nmr spectrum obtained for the ca0 sample at ambient pressure ( solid line ) and the result of its simulation ( dotted line )
are displayed in fig .
[ spectrum ] .
a precise orientation of the single crystal in the magnetic field along the corresponding axis has been reached through a fine adjustment of the angular position of the pressure cell in the magnetic field .
as far as the cu nucleus is concerned we used the second order shift of the central line as explained below but we also checked on the sample ca0 that the value of @xmath38 thus obtained was in very good agreement with the first order quadrupolar shift derived from the frequency difference between the central and the satellite lines .
nmr frequencies up to the second order quadrupolar contribution are given by the equations , @xmath39 for @xmath40-axis where @xmath41 , @xmath42 , @xmath43 are the knight shift , the larmor frequency in a diamagnetic substance and the quadrupolar tensor components , respectively . for @xmath44-axis , @xmath45 can be expressed as , @xmath46 when the magnetic field is aligned along the @xmath47-axis , the quadrupolar contribution to the nmr shift is about ten times larger than for the case @xmath48 according to equations 1 and 2 and consequently must be treated on equal footings with the magnetic shifts .
cu1 @xmath49 versus @xmath41 plots with temperature as the implicit parameter at @xmath50bar and 28kbar in ca12 .
straight lines concern the low temperature domain . ]
we have determined @xmath49 ( eq .
( 2 ) ) using the values @xmath51 , @xmath52 , @xmath53 estimated by magishi _
et - al_. @xcite for the neighboring composition @xmath54 at low @xmath11 .
assuming that @xmath55 in the spin gap regime at ambient pressure we obtain : @xmath56 . since the spin parts @xmath57 and @xmath58 are both proportional to the spin susceptibility @xmath59 and given @xmath60 which is temperature independent , the plot @xmath49 versus @xmath41 with temperature as the implicit parameter must be linear , @xmath61 , \ ] ] where @xmath62 are the uniform hyperfine fields .
@xmath49 versus @xmath41 plots at 1bar and 28kbar are shown in figure .
[ fig10p1.eps ] .
as seen in the figure , these plots are not fitted by a straight line over the whole @xmath11 range .
we can ascribe the deviation of the data from straight lines to the temperature dependence of @xmath63 , namely , it is due to @xmath64 increasing with the temperature .
such an increase of @xmath64 was reported in several nmr studies of @xmath65 @xcite . in those works
, it was indicated that the variation of @xmath64 is negligible in the low @xmath11 regime but becomes quite appreciable above @xmath66 .
therefore , @xmath67 versus @xmath68 plots were fitted by straight lines in the temperature range below @xmath69 . as a result of the fit
, @xmath70 is independent of the applied pressure .
this finding agrees well with @xmath71 derived in reference @xcite . using the ratio @xmath72 and @xmath73 data @xmath74 temperature one can derive with a little bit of algebra the @xmath11 dependence of @xmath64 displayed on figure.[cu1nqvst ] .
temperature dependences of the @xmath75- and @xmath47-axis components of the oxygen and copper nuclear quadrupole frequences , @xmath76 , measured for the cu1 , o1 and o2 ladder sites in ca0 and ca12 samples under ambient and high pressures are presented in figures [ cu1nqvst ] , [ o1nqvst ] and [ o2nqvst ] . for the ca- free sample , @xmath77 , for the cu1 , o1 , and o2 sites
depends on temperature only weakly below 150 k , however , increases steeply above 150 k. as to the ca12 compound , @xmath78(cu1,o1,o2 ) shows a dramatic change in the temperature range @xmath79 k and a @xmath11 independent behavior outside this region . at the ladder cu1 sites in ca0 and ca12 single crystals under ambient and high pressures . ]
as is seen in the figures , the quadrupole frequencies @xmath80(cu1,o1,o2 ) and @xmath81(o1 ) reveal an increase with both the ca content and an applied pressure while on the contrary @xmath81(o2 ) is reduced .
this fact is indicative of an increase of the asymmetry parameter for efg tensor at the o2 site , @xmath82(o2 ) , with increasing ca content and pressure .
let us now consider possible reasons for the evolution of @xmath83 varying doping , pressure and temperature . in a semiempirical approach
the quadrupole frequency can be written as a sum of two contributions,@xcite @xmath84 where the first term describes the contribution from neighboring ions in a point - charge model enhanced by the sternheimer antishielding factor @xmath85 and @xmath86 is the term arising from holes in the orbitals of the ion itself . for the cu 3d and o 2p unfilled shells the hole contributions @xmath87 and @xmath88 can be expressed as follows@xcite @xmath89 @xmath90\nonumber\\= \nu_{2p,0}\left[n_{2p , \alpha}-\frac{n_{2p , \beta}}{2}-\frac{n_{2p , \gamma}}{2}\right ] ( \alpha , \beta , \gamma = a , b , c).\quad \label{oxhole}\end{aligned}\ ] ] here , @xmath91 ( @xmath92 ) is the quadrupole frequency when there is one hole in the 3d ( 2p ) orbital and @xmath93 ( @xmath94 ) is the hole number in the cu 3d ( o 2p,@xmath95 ) shell . if we take @xmath96 @xmath97 , @xmath98 @xmath97 , @xmath99 a.u.@xcite , @xmath100 a.u.@xcite we get , @xmath101 mhz , @xmath102 mhz . at the ladder leg o1 site in ca0 and ca12 for h@xmath103 ( a ) and for h@xmath104 ( b ) under ambient and high pressures . ] at the ladder rung o2 site in ca0 and ca12 for h@xmath103 ( a ) and for h@xmath104 ( b ) under ambient and high pressures . ]
quite generally , the lattice contribution to the quadrupole frequency reads , @xmath105 where @xmath106 is the effective charge of the @xmath107-th ion and @xmath108 is its position .
the sum goes over all crystal sites except for the one considered . in our calculations
we used the same values for @xmath109 and @xmath110 as were found in high-@xmath12 cuprates.@xcite then , for @xmath95-component of the quadrupole frequency at the oxygen @xmath111 sites ( @xmath111=o1,o2 ) , @xmath112 reads ( in mhz ) , @xmath113+\nonumber\\2.80\times 10^{-9 } \sum_{j}l_{\alpha}^{k}(j)e_{j}. \label{oxholelat}\end{aligned}\ ] ] here , @xmath114 is the contribution of the ions located at one of the position @xmath115 ( @xmath115 = o1 , o2 , o3 , cu1 , cu2 , sr / ca sites ) to the @xmath95-component of the lattice sum for the @xmath111-th site ( @xmath111=o1 , o2 ) .
cu2 and o3 are copper and oxygen sites in the cuo@xmath8 chains . similarly , for @xmath116(cu1 )
we get , @xmath117 it is important to notice that there is a rather large uncertainty in the lattice contribution ( eq.(7 ) ) to the quadrupole frequency calculated by using a point charge model , which is an oversimplified approximation .
furthermore , the sternheimer factors @xmath118 and @xmath119 are only crudely determined in the ladders .
this uncertainty conveys an important uncertainty for the actual determination of the absolute hole concentration in the ladder cu@xmath8o@xmath9 layers .
fortunately , the change in the lattice contribution caused by a variation of the temperature , ca doping and pressure are negligible .
hence , the modification of the quadrupole frequency @xmath120 observed in the experiment is dominated by the second term @xmath121 in eq .
( 4 ) . in this case , the uncertainty in the lattice contribution @xmath122 can be neglected since it is a negligible contribution altogether .
the change of quadrupole frequencies @xmath123 and @xmath124 originating from the hole transfer between chains and ladders can be expressed as follows , @xmath125+\nonumber\\2.8\times 10^{-9 } \sum_{j}l_{\alpha}^{k}(j)\delta n_{j}+2.8\times 10^{-9 } \sum_{j}\delta l_{\alpha}^{k}(j ) e_{j}. \label{chox}\end{aligned}\ ] ] @xmath126 where @xmath127 ( @xmath128 ) is the variation of the hole number in the cu 3d ( o 2p,@xmath95 ) orbitals and @xmath129 is the change of the hole density in the entire o 2p ( or cu 3d ) orbital of the ion at the site @xmath115 . the third term in eqs .
( [ chox ] ) and ( [ chcu ] ) takes into account a change of the lattice contribution due to a variation of lattice parameters as a function of temperature , pressure and doping .
these small corrections have been calculated for each cases of the @xmath130 variation using the results of refs . .
next , we have assumed that the holes occupy only the o @xmath131 orbitals , _
i. e. _ @xmath132 .
hence , @xmath133 in addition we have assumed that the holes transferred into the ladders are removed evenly from all atoms of the chains , _
i. e_. @xmath134(o3)@xmath135(cu2)@xmath136 , leading to , @xmath137 this latter assumption has practically no effect on the final result .
for instance , the result of our calculations is hardly modified if we assume that all holes in the chains are located either in oxygen o 2p or copper cu 3d orbitals .
we have also ignored any effect of covalent bonding between copper and ligand oxygens . since the transferred hyperfine fields at the o1 and o2 sites are not affected by pressure , doping and temperature@xcite a variation of @xmath130 due to the change of covalent effects exceeding 1 - 2 % change in @xmath130 seems very unlikely .
the solution of the system of equations eqs .
( [ chox ] ) - ( [ chdn ] ) allows us to derive the change in the hole concentration , @xmath138 , for each orbitals .
the hole distribution in ca0 and ca12 ladder compounds under ambient and high pressures is summarized in figs .
[ orbhvst ] and [ holevst ] .
dependence of the hole number per ladder cu in the cu@xmath8o@xmath9 ladder transferred from the cuo@xmath8 chains ( left vertical axis ) and of the total hole concentration after calibration with the nexafs data , _ see text _ ( right vertical axis ) .
the distribution of transferred holes among the cu1 , o1 ( b ) and o2 ( c ) sites . ]
[ orbhvst](a ) shows for compositions ca0 and ca12 at ambient and high pressures the temperature dependences of the total hole number in the ladders ( counted per cu1 ) which are transferred from the cuo@xmath8 chains ( left vertical axis ) .
this is the only quantity our measurements can determine accurately . in fig .
[ orbhvst](a ) the number of holes @xmath28 transferred into the ladder plane of sr@xmath31cu@xmath5o@xmath6 was assumed arbitrarily to be zero at low @xmath11 .
the overall tendency for the hole density in the ladders is a decrease at low temperature .
such a decrease can happen for the following reason : the backtransfer of holes from ladders to chains . in this respect
, we could ascribe the temperature of @xmath139 below which on fig.[o2nqvst]a @xmath140 becomes temperature independent to the onset of a cdw state@xcite ( or a hole crystallization@xcite ) in the ladders of ca0 .
however , such an effect is not so clearly observed in the hole number on the cu1 sites .
in addition we can rule out such an interpretation for the sudden drop of the hole number in ca12 around @xmath141 since for this compound the cdw ground state is no longer stable at low temperature@xcite .
the backtransfer of charges from the ladders to the chains as temperature is lowered , when the distance between the chains and the ladders is shortened seems to contradict the fact that a shrinkage of the intersheet cuo@xmath8-cu@xmath8o@xmath9 distance leads to a hole transfer towards the ladders .
this discrepancy has been recently resolved by isobe _
et al_. @xcite these authors have shown that it is the minimum o3-cu1 distance of the modulated structure which governs the hole density in the ladders .
moreover , in spite of the average o3-cu1 distance becoming shorter at low temperature the minimum does expands .
this expansion leads in turn to a reverse hole transfer from ladders to chains . using the bond - valence sum calculation , isobe _
et al_.@xcite have obtained that the ladder layer of the ca13.6 compound contains 1.3 holes per formula unit ( _ f.u_. ) ( _ i.e. _ 0.092 hole per cu1 ) at room temperature .
in addition , according to this calculation , as the temperature is lowered down to 5 k nearly all holes ( @xmath1421 hole per _ f.u_. or 0.07 hole per cu1 ) move back to the chains .
we note that our results give for ca12 a much smaller fraction ( 0.28 holes per _ f.u_. or 0.02 hole per cu1 ) of the holes moving back into the chains with decreasing temperature . furthermore , since we only see in the nmr / nqr spectrum the contribution from evenly distributed holes@xcite , our results indicate that a rather substantial number of holes in ca12 is remaining on the ladders ( possibly delocalized ) even down to the lowest temperature .
supporting the possibility that not all holes in the ca - rich samples are backtransferred from the ladders to the chains at low @xmath11 is the observation of an antiferromagnetic long - range ordering taking place in the chains of ca - rich ( @xmath143 ) spin ladders . in the case
when six holes per _
f.u_. are located in chains , spin - singlet dimers form and antiferromagnetic long - range order does not take place in chains .
it is the situation which prevails in ca - poor compounds with @xmath144 ( ref . ) .
the * presence * of af order in the chains of the ca - rich ladders can serve as an indication that the hole number in the chains is less than 6 holes per _ f.u_. in that event , `` lone '' spins and/or `` spin - trimers '' can form in the chains inducing staggered spin modulations on the remaining spin singlets and af long - range order in the spin liquid state is established .
we have obtained that an increase of the ca content up to 12 atoms per _ f.u_. leads to only a small increase of the hole density in the ladder layer ( @xmath145 hole / cu1 ) .
this value is much smaller than the 0.16 hole / cu1 found in optical conductivity measurements of osafune _ et al . _ for ca12@xcite but
is quite close to the value 0.02 hole / cu1 derived by in nexafs measurements of ncker _
et al_.@xcite interestingly , although the results of ncker _ et al_. and those of osafune _ et al_. relative to the hole number in cu@xmath8o@xmath9 planes of ca - rich samples are quite different both investigations report the same absolute value @xmath146 hole / cu1 for the ca0 compound at room temperature . consequently ,
if the hole density in the ladder of ca0 is really 0.06 holes per cu1 at room temperature then our results imply that 0.04 hole / cu1 still persist in the ladder plane of the ca - free sample at low @xmath11 .
hence , it is necessary to add 0.04 hole / cu1 to the amount @xmath28 of holes transferred to the ladders , see fig .
[ orbhvst](a ) ( left axis ) , in order to obtain the total hole number , @xmath147 , located in ladder layers as shown on the right axis of the figure .
then , @xmath147(ca12 , 300k ) becomes 0.09 and 0.12 hole / cu1 at 1 bar and 32 kbar respectively . at this stage
it is interesting to compare our findings for the amount of holes sitting on the ladders in ca0 with the predictions derived from an _ ab initio _
calculation by gell and lepetit.@xcite following their calculation , the number of holes on the ladders should be much smaller than 1 per _ f.u_. in order to lead to a localization of magnetic electrons in agreement with the experimental observation of dimeric units separated by zhang - rice singlets . according to fig .
[ orbhvst](a ) our experiment gives 0.04 per cu1 or ( 0.56 holes per _ f.u _ ) on the ladders at low temperature .
as far as the ca12 compound is concerned , our experiment gives under ambient pressure a number of 1 hole per _ f.u .
_ transferred into the ladders at low temperature .
this figure compares fairly well with the _ ab - initio _ calculation made for the ca13.6 system @xcite predicting a small hole transfer in - between 1 and 2 per _ f.u . _ in order to explain the existence of an antiferromagnetic ground state at low temperature in the heavily ca - substituted samples .
an other important result of our study is the observation of an additional increase of @xmath147 in the ladder layer of the ca - rich compound under high pressure .
the amount of increase is comparable to what is obtained at ambient pressure going from ca0 to ca12 .
it is apparent , therefore , that the important role of high pressure in achieving the conditions for the stabilization of superconductivity in ca12 is an increase of the hole doping of the conducting cu@xmath8o@xmath9 planes . on the other hand
, high pressure leads to only a slight transfer of holes from chains to ladders in ca - free compound ca0 .
note , however , that the @xmath148(ca0 , 32 kbar , 300k)@xmath1490.0262 hole / cu1 , obtained in our experiments , is quite close to the @xmath28(ca3 , 1 bar , 300k)@xmath1490.0270 hole / cu1 reported by thurber _
et al_.@xcite it is well known that the main effect of ca doping and of a hydrostatic pressure on the structure of sr@xmath3ca@xmath4cu@xmath5o@xmath6 is a reduction of the * b * parameter.@xcite the contraction of lattice parameters induced by a pressure of 30 kbar corresponds to a substitution of three sr by three ca , _
@xmath150kbar@xmath151bar@xmath152.@xcite thus , the equality between @xmath148(ca0 , 32 kbar ) and @xmath28(ca3 , 1 bar ) suggests that the transfer of holes from chains to ladders under pressure and upon ca doping is caused , at least for the ca - poor compounds , by the same reason namely , a shrinkage of the intersheet cuo@xmath8-cu@xmath8o@xmath9 distance . the hole distribution among cu1 , o1 and o2 sites of the ladders as a function of temperature at different pressure and ca content is presented in figs .
[ orbhvst](b ) and [ orbhvst](c ) .
these figures show that the holes transferred from the chains to the ladders with increasing temperature sit mainly on the o1 and cu1 sites .
ca substitution leads to the transfer of holes which reside on the cu1 and o2 rung sites .
finally , nearly all holes transferred at room temperature into the ladders when pressure is applied occupy the o2 sites .
figure [ holevst ] shows the distribution of holes located in the 2p orbitals of the oxygen rung site o2 in ca12 between the 2p , a and 2p , c orbitals .
as is seen in the figure , these holes go primarily in the 2p orbitals oriented along the @xmath153-axis of the rungs at increasing ca content and pressure .
this is in agreement with the conclusions of ref . where an enhancement of the hole density in the o 2p , a orbitals in going from ca0 to ca12 was found .
it has been shown in a number of theoretical studies that the spin gap value @xmath154 in the spectrum of spin excitations of two - leg ladders is very sensitive to the degree of carrier doping of the cu@xmath8o@xmath9 layer.@xcite namely , @xmath154 decreases as the hole number in the ladder plane is increased .
our previous nmr studies@xcite have revealed a substantial reduction of the spin gap under pressure . since the applied pressure causes the transfer of holes from the chains to the ladders it is reasonable to explain the spin gap reduction under pressure by an increase of the hole density in the ladder layer . as mentioned above ,
the ladder layers of ca12 contain a finite amount of nonlocalized holes even at low @xmath11 the number of which increasing under pressure .
therefore , the growth of the residual spin susceptibility at low @xmath11 at increasing pressure observed in our previous works@xcite can be related to the increase of single mobile holes in the ladders at low @xmath11 .
such a situation leads in turn to a finite density of states at the fermi level which contributes to the spin susceptibility as seen from the korringa contribution to the nmr relaxation .
note that the achievement of an appropriate carrier doping level due to the @xmath75-axis shrinkage under pressure is not a sufficient condition for a stabilization of superconductivity in ca@xmath7 compounds .
otherwise , even a less substituted ca@xmath7 sample should show superconductivity applying a high pressure .
experimentally , sr@xmath155ca@xmath156cu@xmath5o@xmath6 did not reveal superconductivity up to 85 kbar.@xcite ca substitution has an effect on the lattice similar to high pressure although not equivalent.@xcite the lattice parameter @xmath157 in the ca@xmath7 ladders shows hardly any change under pressure but on the other hand is very sensitive to ca doping .
for instance , an hydrostatic compression of a ca8 sample up to 110 kbar can not reduce the @xmath157 parameter down to its value in ca12 under 30 kbar.@xcite since a reduction of the @xmath157 parameter enhances the interladder interaction along the @xmath153-axis it leads to a more two - dimensional ladder system .
we suggest that prerequisites for the stabilization of superconductivity in a spin - ladder are an adequate `` two - dimensional degree '' and a large enough carrier doping level .
ca substitution is essential to achieve a two dimensional character while the necessary carrier density can only be attained by pressure .
in summary , we have reported a @xmath0cu and @xmath1o nmr investigation of the nuclear quadrupole interaction tensor , @xmath2 , in the hole doped spin ladder system sr@xmath3ca@xmath4cu@xmath5o@xmath6 ( @xmath7=0 and 12 ) at ambient pressure and under a pressure of 32 kbar when ca12 ladders are superconducting .
we have related the changes in @xmath2 to a variation in the effective hole concentration of the ladder layer and have determined the hole distribution among ladder and chain systems as a function of temperature , ca content and an applied hydrostatic pressure .
the results are summarized in figs .
[ orbhvst ] and [ holevst ] .
the overall tendency of the temperature dependence for the hole density in the ladder layer is a reduction as temperature is lowered .
the effective hole concentration in the cu@xmath8o@xmath9 ladder plane of ca0 compounds depends only weakly on the temperature below 150 k but increases steeply above 150 k. the additional hole number in ladder layers of ca12 shows a dramatic change in the temperature range @xmath79 k and a @xmath11-independent behavior outside this region .
holes transferred from the chains to the ladders at increasing temperature occupy mainly the cu @xmath158 and the o 2p , c orbitals at the cu1 and o1 sites of ladders .
an increase of the ca content and of the applied pressure leads to an additional doping of the cu@xmath8o@xmath9 ladder units .
ca substitution leads to a transfer of holes in the orbitals of cu1 and rung oxygen o2 ions whereas under pressure nearly all transferred holes occupy the o 2p , a orbitals of o2 oxygens .
a comparison between the effect of pressure and ca substitution on the hole number in the ladder layers and of the lattice parameters suggests that the hole transfer from chains to ladders under pressure and upon ca doping is caused , at least for ca poor compounds by a shrinkage of the intersheet cuo@xmath8-cu@xmath8o@xmath9 distance .
furthermore , our results indicate that a rather substantial number of holes in the ladder planes of ca12 is remaining delocalized down to the lowest temperature .
we suggest that the important role of high pressure for realizing superconductivity in ca12 is an increase of the hole density in the ladder layers . however , the stabilization of a superconducting ground state in the spin ladder sr@xmath3ca@xmath4cu@xmath5o@xmath6 is achievable only at a rather high `` 2d degree '' in the ladder plane .
this two - dimensionality is obtained more by ca substitution than by high pressure .
finally , it is worth putting the behavior of superconducting spin ladders in the context of high @xmath159 ( htsc ) superconductors .
once enough 2-d character has been achieved by ca doping , superconductivity can then be stabilized at a strongly pressure dependent critical temperature .
@xmath159 amounts to 5 k in ca12 under 30 kbar@xcite and passes through a maximum at 10k in ca11.5@xcite and ca13.6@xcite under 45 and 35 kbar respectively .
as suggested by the present nmr / nqr study , pressure enhances the carrier concentration in the ladder planes .
the pressure ( hole concentration ) dependence of @xmath159 is also reminiscent of the parabolic dependence found for @xmath159 _ versus _ the carrier concentration in many htsc @xcite .
a. revcolevschi , u. ammerhal and g. dhalenne are gratefully acknowledged for the samples preparation .
we thank a. inyushkin for helping us with the @xmath160 isotope exchange process .
y.p . would like to thank s. verkhovskii for useful discussions .
y.p . acknowledges the cnrs and `` russian science support foundation '' for a financial support .
the work was supported by the ministry of sciences and technology of the russian federation under contract no 25/03 and the russian foundation for basic researches ( project n 05 - 02 - 17846 ) .
n. ncker , m. merz , c. a. kuntscher , s. gerhold , s. schuppler , r. neudert , m. s. golden , j. fink , d. schild , s. stadler , v. chakarian , j. freeland , y. u. idzerda , k. conder , m. uehara , t. nagata , j. goto , j. akimitsu , n. motoyama , h. eisaki , s. uchida , u. ammerahl , and a. revcolevschi . , 62:14384 , 2000 . | we report the results of a @xmath0cu and @xmath1o nmr study of the nuclear quadrupole interaction tensor , @xmath2 , in the hole doped spin ladder system sr@xmath3ca@xmath4cu@xmath5o@xmath6 ( @xmath7=0 and 12 ) performed under ambient and high pressures .
nmr data show that the hole density in the cu@xmath8o@xmath9 ladder layer grows with temperature , ca content and an applied pressure .
we have derived the hole occupation of cu 3d and o 2p orbitals at the different ion sites in the cu@xmath8o@xmath9 ladder as a function of the temperature , ca substitution and pressure .
we also suggest that the most important role of high pressure for the stabilization of a superconducting ground state in ca - rich two - leg ladders is an increase of the hole concentration in the conducting cu@xmath8o@xmath9 planes .
we have obtained an estimate of 0.10 hole per cu1 for the hole concentration at low temperature in ca12 under 32 kbar when this compound undergoes a superconducting transition at @xmath10 .
such a value fits fairly well with the doping phase diagram of cuprate superconductors . |
time - fractional equations of the form @xmath8for @xmath9 have been studied by a number of authors since the 1980s : see , for example , wyss ( @xcite ) , nigmatullin ( @xcite ) , schneider and wyss ( @xcite ) , mainardi ( @xcite , @xcite ) and , more recently , nigmatullin ( @xcite ) , angulo et al .
( @xcite , @xcite ) .
hyperbolic fractional equations similar to ( [ equation1 ] ) have been analyzed , for example , by engler ( @xcite ) . for exhaustive reviews on this topic ,
also consult samko , kilbas and marichev ( @xcite ) and podlubny ( @xcite ) . for interesting applications of fractional equations to physical problems
see , for example , saichev and zaslavsky ( @xcite ) , nigmatullin et al .
( @xcite ) , angulo et al .
( @xcite ) .
fractional diffusion equations of order @xmath10 emerge in the study of the distribution of the local time of pseudoprocesses related to higher - order heat - type equations ; see beghin and orsingher ( @xcite ) . the time - fractional derivative appearing in ( [ equation1 ] ) must be understood in the sense of dzerbayshan caputo , that is@xmath11 where @xmath12 .
considering the derivative in the sense of dzerbayshan
caputo permits us to study initial value problems for ( [ equation1 ] ) with initial data represented by derivatives of integer order ; on this topic , consult mainardi ( @xcite ) . we assume , in particular ,
the following initial condition : @xmath13and@xmath14 the general solution to equation ( [ equation1 ] ) subject to ( condition 1 ) or ( [ cond1 ] ) is well known [ see podlubny ( @xcite ) , formula ( 4.22 ) , page 142 ] and reads @xmath15\\[-8pt ] & = & \frac{1}{2\lambda
t^{\nu /2}}w_{-\nu /2,1-\nu /2}\biggl ( -% \frac{|x|}{\lambda t^{\nu /2}}\biggr ) \nonumber\end{aligned}\]]where @xmath16 in ( [ equation2 ] ) denotes the so - called wright function , whose general form is @xmath17 some properties of the wright function are investigated in mainardi and tomirotti ( @xcite ) and in gorenflo , mainardi and srivastava ( @xcite ) .
initial value problems ( as well as problems on half - lines with boundary conditions ) for equations like ( [ equation1 ] ) are extensively treated and solved in mainardi ( @xcite , @xcite , @xcite ) , gorenflo and mainardi ( @xcite ) and buckwar and luchko ( @xcite ) .
it has been proved also that @xmath7 is nonnegative and integrates to one for all @xmath1 ; see , for example , orsingher and beghin ( @xcite ) .
we present here some alternative forms of the solution @xmath7 of ( equation1 ) , either as integral functions like@xmath18 or in terms of stable densities @xmath19 as@xmath20 in orsingher and beghin ( @xcite ) , we proved that in the special case @xmath21 the solution ( [ equation2 ] ) coincides with the distribution of the process@xmath22called the iterated brownian motion , which consists of a brownian motion @xmath23 whose `` time '' is an independent reflecting brownian motion .
in beghin and orsingher ( @xcite ) we have generalized this result to the case where @xmath24 @xmath25 . in this case
, for @xmath26 , the solution ( [ equation2 ] ) coincides with the distribution of the process @xmath27where the vector process @xmath28 has the following joint distribution : @xmath29{n^{n}t}}w_{2}\cdots w_{n-1}^{n-2},\nonumber\\[-8pt]\\[-8pt ] \eqntext{w_{j}\geq 0 , 1\leq j\leq n-1,}\end{aligned}\]]for @xmath30 . in ( [ iterated brownian motion ] ) the role of `` time ''
is played by the product of independent , positive - valued r.v.s , which can not be identified with well - known distributions as in the special case ( [ iterated ] ) . in the special case @xmath31
, we note that @xmath32 , because ( [ mult ] ) becomes the distribution of a reflecting brownian motion .
we are now able to prove a much stronger result for the case @xmath33 , @xmath25 , and for @xmath34 , which has a number of interesting consequences
. we will show below that ( [ equation2 ] ) for @xmath33 can be written down as @xmath35and this coincides with the distribution of @xmath36where the @xmath37 s are independent brownian motions . the iterated brownian motion _ _ @xmath38__@xmath39 has been actively investigated and many of its properties have been obtained by khoshnevisan and lewis ( @xcite ) , burdzy and san martn ( @xcite ) , allouba ( @xcite ) . the connection between fractional generators of order @xmath40 and the iterated brownian motion @xmath41 has been studied in allouba and zheng ( @xcite ) and baeumer , meerschaert and nane ( @xcite ) .
this connection was obtained in orsingher and beghin ( @xcite ) as a particular case of the analysis of the fractional telegraph equation .
the identity @xmath42\\[-8pt ] & & \qquad=\frac{1}{(2t)^{1/2^{n+1}}}\sum_{k=0}^{\infty } \frac { ( - 2|x|/(2t)^{1/2^{n+1 } } ) ^{k}}{k!\gamma ( -k/2^{n+1}+1 - 1/2^{n+1 } ) } \nonumber\end{aligned}\]]shows that there is a deep connection between wright functions and gaussian distributions . for the @xmath5-times
iterated brownian motion @xmath43 , @xmath44 , we obtain the distributions of the maximum and the sojourn time ( together with the expression of moments ) and we work out in detail an explicit form of them for the case of the classical iterated brownian motion @xmath45 @xmath46 we note that @xmath43 converges in distribution , for @xmath47 , to a gauss laplace ( or bilateral exponential ) random variable , independent from @xmath44 .
in orsingher and beghin ( @xcite ) we have seen that for the fractional telegraph - type equation@xmath48the general solution coincides with the distribution of the telegraph process @xmath49 whose time is an independent reflecting brownian motion @xmath50 we remark that process ( [ iterated telegraph ] ) converges to ( iterated brownian motion ) in the kac sense ( i.e. , for @xmath51 , @xmath52 , in such a way that @xmath53 .
related interpretations of the solutions to @xmath54are discussed in beghin and orsingher ( @xcite ) and orsingher and beghin ( @xcite ) .
generalized forms of the fractional telegraph equation ( equation3 ) and of its solutions can be found in saxena , mathai and haubold ( @xcite ) .
we obtain here various types of relationships between the solutions @xmath55 for different values of @xmath56 the first one we present is the following:@xmath57(valid for any
@xmath58 , where @xmath59 is the solution of ( equation1 ) with order @xmath60 instead of @xmath56 formula ( [ intro.1 ] ) leads , for @xmath61 to the @xmath5-times iterated brownian motion defined in ( [ iperiterated ] ) , since it permits us to obtain , in an alternative way , the relationship ( [ multiple integral ] ) . in the general case , ( [ intro.1 ] )
shows that the process related to the equation ( [ equation1 ] ) of order @xmath62 can be interpreted as the composition of a process governed by the same equation , but with order @xmath63 with a gaussian - distributed time .
we also derive the analogous relationship@xmath64where@xmath65 here the roles of space and time are interchanged with respect to ( intro.1 ) .
therefore from ( [ intro.2 ] ) a further interpretation of the solution emerges , because it coincides with the density of the process@xmath66 where @xmath67 is a brownian motion and @xmath68 is a process independent from @xmath67 with a distribution for each @xmath69 given in ( intro.3 ) .
a relationship similar to ( [ intro.1 ] ) and connecting @xmath7 with @xmath70 is established ( by applying the multiplication formula of gamma function ) for @xmath71 and @xmath72 .
substantially different situations are encountered for the special cases @xmath73 @xmath74 and @xmath75 in particular for @xmath74 we show that the solution to ( [ equation1 ] ) possesses the following simple form:@xmath76{3t}}ai\biggl ( \frac{% airy function .
the latter emerges as a solution to third - order heat - type equations of the form@xmath77 by using again the relationship ( [ intro.1 ] ) we get , for the case @xmath78 , the following result:@xmath79{3z}}ai\biggl ( \frac{|x|}{\lambda \sqrt[3]{3z}}\biggr ) \,dz.\ ] ] this suggests that we should interpret @xmath80 as the distribution of @xmath81 where @xmath82 is a process whose one - dimensional distribution is given in ( intro.4 ) , which coincides with the symmetric stable process of order @xmath83 .
similar relationships seem not to hold for the solutions to fractional equations of order @xmath84 , because the fundamental solutions to @xmath85 @xmath86 , are sign - varying functions on the whole @xmath87-axis ( while , for @xmath88 only on the negative half - line ) , as shown in detail in lachal ( @xcite ) . therefore they can not be used to construct the functions @xmath7 emerging from ( [ equation1 ] ) , which , for @xmath1 , are nonnegative and integrate to one .
we note that the solutions to ( [ n - order equation ] ) themselves have been represented as distributions of compositions of artificial processes , which do not display a probabilistic structure [ see funaki ( @xcite ) , hochberg and orsingher ( @xcite ) , benanchour , roynette and vallois ( @xcite ) ] .
finally the previous results permit us to establish connections between the solutions @xmath89 and @xmath90
. moreover the explicit form ( [ intro.4 ] ) of @xmath91 suggests that we should interpret them as distributions of processes similar to the @xmath5-times iterated brownian motion , but with the role of @xmath67 replaced by @xmath82 and the time represented by nested products of the random variables @xmath92 defined in ( [ iterated brownian motion ] ) .
in this section we examine in detail various relationships between solutions to diffusion equations like ( [ equation1 ] ) and processes involving brownian motion . all results of this section
refer to equations of order @xmath93 we start with the following general theorem : [ thm1 ] the solution to @xmath94for @xmath95 , can be represented as @xmath96where @xmath59 is the solution to @xmath97or @xmath98 by applying the duplication formula of the gamma function we have that @xmath99by plugging ( [ gamma ] ) into ( [ equation2 ] ) we get that @xmath100 \\ & = & \frac{1}{\sqrt{\pi } 2^{\nu } t^{\nu /2}}\int_{0}^{\infty } e^{-w}w^{-\nu /2 - 1/2}\bigl(2\sqrt{tw}\bigr)^{\nu } u_{2\nu } \bigl(x,2\sqrt{%
tw}\bigr)\,dw \\ & = & \frac{1}{\sqrt{\pi } } \int_{0}^{\infty } e^{-w}w^{-1/2}u_{2\nu } \bigl(x,2% \sqrt{tw}\bigr)\,dw \\ & = & \bigl [ 2\sqrt{tw}=z\bigr ] \\ & = & \frac{1}{\sqrt{\pi t}}\int_{0}^{\infty } e^{-z^{2}/(4t)}u_{2\nu } ( x , z)\,dz\end{aligned}\]]and this concludes the proof .
an alternative proof of the relationship ( [ vologda ] ) is based on the fourier transforms , since for @xmath7 the following result is known:@xmath101 where @xmath102 is the mittag leffler function .
taking the fourier transform of ( [ vologda ] ) we get that @xmath103 \\ & & \qquad=\sum_{k=0}^{\infty } \frac{(-\beta ^{2}\lambda ^{2})^{k}}{\gamma ( 2k\nu + 1)% } \frac{(2\sqrt{t})^{2k\nu + 1}}{2\sqrt{\pi t}}\gamma \biggl ( \nu k+\frac{1}{2}% \biggr ) \\ & & \qquad=\sum_{k=0}^{\infty } \frac{(-\beta ^{2}\lambda ^{2})^{k}}{\gamma ( 2k\nu + 1)% } \frac{(2\sqrt{t})^{2k\nu + 1}}{2\sqrt{\pi t}}\sqrt{\pi } 2^{1 - 2\nu k}\frac{% \gamma ( 2\nu k ) } { \gamma ( \nu k ) } \\ & & \qquad=\sum_{k=0}^{\infty } \frac{(-\beta ^{2}\lambda ^{2}t^{\nu } ) ^{k}}{\gamma ( k\nu + 1)}=\int_{-\infty } ^{+\infty } e^{i\beta x}u_{\nu } ( x , t)\,dx.\end{aligned}\ ] ] [ rem2.1 ] in the special case where @xmath104 , formula ( [ vologda ] ) yields @xmath105 \\ & = & \frac{1}{\sqrt{\pi t}}\int_{0}^{\infty } \frac{e^{-x^{2}/(2y)}}{% \sqrt{2\pi y}}\frac{e^{-y^{2}/(4t(2\lambda ^{2})^{2})}}{2\lambda ^{2}% } \,dy .
\nonumber\end{aligned}\ ] ] particularly interesting is the case where @xmath106 that is , when @xmath107 , because ( [ omsk ] ) reduces to @xmath108which permits us to conclude that , in this case , the solution coincides with the probability density of the iterated brownian motion ( [ iterated ] ) .
[ rem2.2 ] if we generalize our analysis to the @xmath5-dimensional case and take @xmath109 we can show that the process related to a fractional equation of the form@xmath110with initial condition@xmath111 has components represented by iterated brownian motions with a common random time . in other words , the solution to ( [ dim ] ) coincides with the distribution of the vector process@xmath112 where @xmath113 @xmath114 , are mutually independent brownian motions and also independent from @xmath67 . to check this result we evaluate the fourier transform of the solution to ( [ dim ] ) as follows:@xmath115 from ( [ dim.1 ] ) we get the inverse fourier transform in the following form:@xmath116
the main difference with respect to the case of the usual multivariate heat equation is that the components of the iterated brownian motions are no longer independent because they are related to each other by the common random time @xmath67 ( with infinitesimal variance @xmath117 we pass now to our second theorem , which is related to the case @xmath118 @xmath25 .
[ thm2.2 ] for @xmath119 the solution to equation ( [ equation1 ] ) under the initial condition ( [ condition 1 ] ) can be written as @xmath120 in view of the duplication formula for the gamma function we can write @xmath121\\[-8pt ] & & \qquad=\sqrt{\pi } 2^{% 1/2^{n}+k/2^{n}}\frac{\gamma ( 1-k/2^{n}-1/2^{n } ) } { \gamma ( 1/2-k/2^{n+1}-1/2^{n+1 } ) } \nonumber\end{aligned}\]]so that the first member of ( [ equation2 ] ) becomes , for @xmath122 and @xmath123 , @xmath124at this point we can use the reflection formula for the gamma function @xmath125\\[-8pt ] & = & \frac{\pi } { \cos k\pi /2\gamma ( ( 1+k)/2 ) } \nonumber\end{aligned}\]]and this shows that only even terms of ( [ ujasnoe otnoshenie ] ) must be retained
. we can therefore write that @xmath126^{-1 } \nonumber \\ & & \qquad=\frac{2}{(2t)^{1/2^{n+1}}}\sum_{k=0}^{\infty } \biggl ( -\frac{2|x|}{% ( 2t)^{1/2^{n+1}}}\biggr ) ^{k}\nonumber\\ & & \quad\qquad { } \times\frac { ( \int_{0}^{\infty } \cdots \int_{0}^{\infty } e^{-\sum_{j=1}^{n}w_{j}}\prod_{j=1}^{n}w_{j}^{-% ( k+1)/2^{n+2-j}-1/2}\,dw_{j } ) \cos k\pi /2\gamma ( k ) } { k!(\sqrt{\pi } ) ^{n+1}2^{\sum_{j=0}^{n-1}1/2^{n - j } } 2^{k\sum_{j=0}^{n}1/2^{n - j}}\gamma ( k/2 ) } \\ & & \qquad=\frac{2}{(2t)^{1/2^{n+1}}}\nonumber\\ & & \quad\qquad{}\times\sum_{r=0}^{\infty } \biggl ( -\frac{2|x|}{% ( 2t)^{1/2^{n+1}}}\biggr ) ^{2r}\nonumber\\ & & \quad\qquad { } \times\frac { ( \int_{0}^{\infty } \cdots \int_{0}^{\infty } e^{-\sum_{j=1}^{n}w_{j}}\prod_{j=1}^{n}w_{j}^{-% ( 2r+1)/2^{n+2-j}-1/2}\,dw_{j } ) ( -1)^{r}}{(\sqrt{\pi } % ) ^{n+1}2^{\sum_{j=0}^{n}1/2^{n - j}}2^{2r\sum_{j=0}^{n}1/2^{n - j}}r ! } \nonumber\\ & & \qquad=\frac{2}{(2t)^{1/2^{n+1}}2^{2 ( 1 - 1/2^{n+1 } ) } ( % \sqrt{\pi } ) ^{n+1}}\nonumber\\ & & \quad\qquad{}\times\sum_{r=0}^{\infty } \frac{(-1)^{r}}{r!}\biggl ( \frac{% x^{2}}{t^{1/2^{n}}}2^{2 - 1/2^{n}}\biggr ) ^{r}\bigl ( 2^{-2 ( 2 - 1/2^{n } ) } \bigr ) ^{r}\nonumber\\ & & \quad\qquad { } \times \biggl ( \int_{0}^{\infty } \cdots \int_{0}^{\infty } e^{-\sum_{j=1}^{n}w_{j}}\prod_{j=1}^{n}w_{j}^{-(2r+1)/2^{n+2-j}- 1/2}\,dw_{j}\biggr ) .\nonumber\end{aligned}\ ] ] by considering that @xmath127 ^{r } = e^{-x^{2}/2^{2}(2/t ) ^{1/2^{n } } \prod_{j=1}^{n}w_{j}^{-1/2^{n+1-j}}},\ ] ] we can write ( [ due.8 ] ) as follows : @xmath128 in order to calculate the integrals let us write@xmath129
so that@xmath130 and@xmath131 therefore we get @xmath132 now we make the similar substitution@xmath133 so that we get again@xmath134 and @xmath135 in view of these substitutions , formula ( [ due.9 ] ) is transformed into@xmath136 by similar transformations , after @xmath137 additional steps , we arrive at@xmath138 ) } e^{-w_{1}}w_{1}^{-1/2^{2}- 1/2}\,dw_{1}.\end{aligned}\ ] ] by means of the position@xmath139 we get that@xmath140 and @xmath141 we arrive at the final expression @xmath142which coincides with ( [ due.6 ] ) .
[ rem2.3 ] it is well known that the laplace
fourier transform of the solution to ( equation1 ) with initial conditions ( [ condition 1 ] ) or ( [ cond1 ] ) is equal , for @xmath9 to@xmath143 we check that the laplace
fourier transform of ( [ due.6 ] ) reduces to ( lf ) for @xmath2 and @xmath144 : @xmath145 by applying the duplication formula we get that@xmath146 and thus@xmath147 by taking the laplace transform of ( [ due.11 ] ) we get@xmath148 which coincides with ( [ lf ] ) , for @xmath2 and @xmath34 .
the form ( [ due.6 ] ) of the solution @xmath149 shows that it coincides with the distribution of the @xmath5-times iterated brownian motion defined in ( [ iperiterated ] ) .
another representation of the solution to the fractional equation ( equation1 ) can be inferred from the following result : [ thm2.3 ] the solution @xmath150 to the initial value problem ( [ eq4 ] ) , for @xmath95 , can be written as @xmath151where @xmath152 and @xmath59 is the solution of ( [ init1 ] ) or ( init2 ) .
we first note that for the solutions to ( [ init1 ] ) or ( [ init2 ] ) the following result holds:@xmath153as can be obtained by taking the laplace transform of @xmath154 the solution to the corresponding equation @xmath155 coincides with the solution to@xmath156 and easily yields ( [ lt ] ) ; see also ( 3.3 ) of orsingher and beghin ( @xcite ) .
therefore , by taking the laplace transform of ( [ new.1 ] ) , we get@xmath157 \\ & & \qquad=\frac{s^{\nu -1}}{2\lambda } \int_{0}^{\infty } \frac{1}{\sqrt{2\pi z\lambda } } e^{-x^{2}/(2z\lambda)}e^{-s^{\nu } /\lambda
z/2}\,dz \\ & & \qquad=\frac{s^{\nu /2 - 1}}{2\lambda } e^{-|x|s^{\nu /2 } /\lambda } \end{aligned}\]]and this coincides with the laplace transform of @xmath158 .
[ rem2.4 ] formula ( [ new.1 ] ) suggests that we should represent the solution of ( [ eq4 ] ) as the distribution of the process@xmath159where @xmath67 is a brownian motion with infinitesimal variance @xmath160 and @xmath161 , @xmath44 , is a process , independent from @xmath162 with law equal to ( [ new.2 ] ) .
it is straightforward that , for @xmath163 , the process ( [ new.3 ] ) coincides with the iterated brownian motion _ _ @xmath38__@xmath164 ; see ( [ iterated ] ) . by comparing the relationship ( [ new.1 ] ) with ( [ vologda ] )
we note also that , in the composition of processes , brownian motion plays in the second case the role of `` time , '' while in the first one it represents `` space . ''
some properties of the classical iterated brownian motion have been obtained by several authors and include the law of iterated logarithm [ burdzy and san martn ( @xcite ) ] and the modulus of continuity [ khoshnevisan and lewis ( @xcite ) ] .
applications of the iterated brownian motion to diffusion in cracks are dealt with in de blassie ( @xcite ) .
we start by presenting the distribution of the maximum of the @xmath5-times iterated brownian motion and , in an explicit form , for the usual iterated brownian motion .
[ thm3.1 ] for the @xmath5-times iterated brownian motion @xmath165 where @xmath37 , @xmath166 , are independent brownian motions , we have for @xmath167 that @xmath168 for @xmath169 we can write that@xmath170 for @xmath171 @xmath172 , we have analogously that @xmath173\\[-8pt ] & & \qquad=2\int_{0}^{+\infty } \pr \ { b_{1}(y)\in d\beta \ } \pr \biggl\ { \max_{0\leq w\leq t}|\mathcal{i}_{n-1}(w)|\in dy\biggr\ } \nonumber\end{aligned}\ ] ] and , by induction , we obtain ( [ max.1 ] ) . [ rem3.1 ] in the case @xmath174 we can give an explicit expression for ( [ max.1 ] ) as follows:@xmath175\\[-8pt ] & & \qquad\quad\hspace*{43pt}\phantom{{}\times\biggl\ { \sum_{k=-\infty } ^{+\infty } ( -1)^{k}\biggl [ } { } + ( 1 - 2k)\frac{e^{-w^{2}/(2 t ) ( 1 - 2k)^{2}}}{\sqrt{2\pi t}}\biggr ] \biggr\ } \,dw \nonumber \\ & & \qquad=2\sum_{k=-\infty } ^{+\infty } ( -1)^{k}\biggl [ \pr \biggl\ { \mathcal{i}% _ { 1}\biggl ( \frac{t}{(1 + 2k)^{2}}\biggr ) \in d\beta \biggr\ } + \pr \biggl\ { \mathcal{i}_{1}\biggl ( \frac{t}{(1 - 2k)^{2}}\biggr ) \in d\beta \biggr\ } \biggr ] \nonumber \\ & & \qquad=2d\beta \sum_{k=-\infty } ^{+\infty } ( -1)^{k}\biggl [ u_{1/2}\biggl ( \beta , \frac{t}{(1 + 2k)^{2}}\biggr ) + u_{1/2}\biggl ( \beta , \frac{t}{% ( 1 - 2k)^{2}}\biggr ) \biggr ] , \nonumber\end{aligned}\]]where @xmath176 is given in ( [ mezzo ] ) and in the first step we applied the well - known result for the maximal distribution of the absolute value of brownian motion [ see shorack and wellner ( @xcite ) , page 34 ] .
the last term of ( [ max.4 ] ) shows that the distribution of the maximum of the iterated brownian motion can be expressed in terms of its probability law @xmath177 , as in the case of the classical brownian motion .
in principle we could write explicitly the distribution of the maximum of @xmath43 in terms of @xmath149 , but this produces a sum of @xmath178 terms , each of which has a very entangled structure . on the basis of the same principles it is possible to write down the distribution of the sojourn time on the positive half - line of the process @xmath179 @xmath180 @xmath3 defined as @xmath181 this random variable takes values in @xmath182 , because during the interval @xmath183 the process @xmath184 ( which plays the role of time for @xmath185 ) can span the whole positive real axes . for the process @xmath43 ,
@xmath44 , the distribution of @xmath186 reads @xmath187\\[-10pt ] \eqntext{0\leq s<\infty .}\end{aligned}\ ] ] the definition of @xmath186 given in ( [ soj.1 ] ) implies that @xmath188 \in ds\biggm\vert \max_{0\leq w\leq t}|\mathcal{i}_{n-1}(w)|\biggr\ } \biggr\ } \!\!\!\!\!\!\!\ ! \\ & & \qquad=\int_{s}^{+\infty } \pr \ { \gamma _ { z}\in ds\ } \pr \biggl\ { \max_{0\leq w\leq t}|\mathcal{i}_{n-1}(w)|\in dz\biggr\ } .
\nonumber\end{aligned}\ ] ] by inserting the arc - sine law in ( [ soj.2 ] ) we get ( [ sogg ] ) .
we can check that ( [ sogg ] ) integrates to one@xmath189 [ rem3.2 ] for the iterated brownian motion @xmath190 the distribution of @xmath186 can be written explicitly as follows:@xmath191 by the transformation @xmath192 the integrals in ( [ soj.3 ] ) are converted [ for @xmath193 into@xmath194where , in the last step , we have applied formula 3.469.1 of gradshteyn and rhyzik ( @xcite ) and @xmath195 $ ] [ by formula 8.485 of gradshteyn and rhyzik ( @xcite ) ] . by @xmath196 we denote the bessel function of imaginary argument of order @xmath62 , that is , @xmath197 .
therefore we get @xmath198 we now derive the explicit form of the moments of even order of _ _ @xmath199__@xmath200 .
[ thm3.3 ] for the process @xmath43 ,
@xmath44 , the moments of order @xmath201 are given by @xmath202\\[-8pt ] & = & \frac{2^{k/2^{n}}}{2^{2k}}\frac{(2k)!}{\gamma ( k/2^{n}+1 ) } t^{k/2^{n}}. \nonumber\end{aligned}\ ] ] the first expression in ( [ due.14 ] ) can be proved by observing that , for , @xmath203 \nonumber \\ & = & \frac{(2k)!}{k!}\frac{1}{2^{k}}e\vert b_{2}(|b_{3}(\cdots|b_{n+1}(t)|\cdots)|)\vert ^{k } \nonumber\\[-8pt]\\[-8pt ] & = & \frac{(2k)!}{k!}\frac{1}{2^{k}}2\int_{0}^{+\infty } x^{k}\pr \ { b_{2}(|b_{3}(\cdots|b_{n+1}(t)|\cdots)|)\in dx\ } \nonumber \\ & = & \frac{(2k)!}{k!}\frac{1}{2^{k}}2\int_{0}^{+\infty } x^{k}\pr \ { \mathcal{i}_{n-1}(t)\in dx\ } , \nonumber\end{aligned}\]]which coincides with the second line of ( [ due.14 ] ) . by performing the integrations in ( [ due.15 ] ) we get the explicit expression of the moments of order @xmath201:@xmath204 \\ & = &
\sqrt{\pi ^{n}}2^{n - k - k/2-\cdots - k/2^{n-1}}\frac{\gamma ( k)}{% \gamma ( k/2^{n } ) } \frac{2^{k/2+\cdots+k/2^{n}% } t^{k/2^{n}}}{2^{k}\sqrt{\pi ^{n}}}\frac{(2k)!}{k ! } \nonumber \\ & = & t^{k/2^{n}}2^{n-2k+k/2^{n}}\frac{(2k)!}{k\gamma ( k/2^{n } ) } .
\nonumber\end{aligned}\ ] ] [ rem3.3 ] for @xmath205 formula ( [ due.14 ] ) coincides with the moments @xmath206 , which is as it should be , since _ _
@xmath38__@xmath207 for @xmath174 , the moments of the iterated brownian motion _
_ @xmath38__@xmath208 can be evaluated directly as follows:@xmath209which coincides with ( [ due.14 ] ) for @xmath210 for any @xmath172 and @xmath211 , we obtain the explicit form of the variance@xmath212 while , for @xmath213 it is @xmath214 , as expected .
[ rem3.4 ] for all @xmath44 , the sequence _
_ @xmath38__@xmath200 converges in distribution@xmath215 for @xmath216 , to the gauss
laplace exponential random variable and its density is independent from @xmath217 from ( [ multiple integral bis ] ) we get that@xmath218 by working on the fourier transform ( [ due.11 ] ) of @xmath149 we have the following alternative proof:@xmath219 formula ( [ lim.4 ] ) coincides with the characteristic function of ( lim.1 ) . loosely speaking
, this shows that the composition of infinite brownian motions produces the bilateral exponential distribution . in view of ( [ multiple integral ] )
we have also the identity@xmath220\\[-8pt ] & & \qquad = e^{-2|x|},\nonumber\end{aligned}\ ] ] which is a rather striking result .
furthermore , if we assume that@xmath221 the fractional equation ( [ equation1 ] ) is converted into@xmath222 subject to@xmath223 which is satisfied by ( [ lim.1 ] ) for all @xmath224 .
[ rem3.5 ] for the random process @xmath225 where @xmath49 is a telegraph process ( with parameters @xmath51 and @xmath226 ) independent from the brownian motions @xmath113 @xmath227 , we have a similar result .
the distribution @xmath149 of ( [ tel.1 ] ) is a solution to @xmath228 and its characteristic function is equal to @xmath229 , \nonumber\end{aligned}\]]where @xmath230 and @xmath231 [ see orsingher and beghin ( @xcite ) , formula ( 2.7 ) , for @xmath232 . for @xmath216
we get from ( [ tel.2 ] ) that@xmath233which is the characteristic function of the bilateral exponential random variable , with density@xmath234 clearly , for @xmath235 and @xmath236 , ( [ tel.3 ] ) reduces to ( [ lim.1 ] ) and ( [ tel.4 ] ) coincides with ( [ lim.4 ] ) .
in some special cases it is possible to present the solutions of the fractional equations ( [ equation1 ] ) in a more attractive fashion . this is the case for @xmath74 .
the explicit form of @xmath240 is given in the next theorem , in terms of airy functions . by combining this result with the relationship given in theorem [ thm1 ]
, @xmath241 can be represented consequently in an interesting form .
[ thm4.1 ] the solution to @xmath242can be represented as @xmath243{3t}}ai\biggl ( \frac{% \label{tre.5 } ai(w ) & = & \frac{1}{\pi } \int_{0}^{+\infty } \cos \biggl ( \alpha w+\frac{\alpha ^{3}}{3}\biggr ) \,d\alpha \nonumber\\[-8pt]\\[-8pt ] & = & \frac{w^{1/2}}{3}\biggl [ i_{-1/3}\biggl ( \frac{2w^{3/2}}{3}\biggr ) -i_{1/3}\biggl ( \frac{2w^{3/2}}{3}\biggr ) \biggr ] \nonumber\end{aligned}\]]is the airy function and @xmath196 denotes the bessel function of imaginary argument of order @xmath62 . from ( [ equation2 ] )
we readily have that @xmath244\\[-8pt ] & = & \frac{1}{2\pi \lambda t^{1/3}}\sum_{k=0}^{\infty } \frac { ( -|x|% /(\lambda t^{1/3 } ) ) ^{k}\gamma ( ( k+1)/3)\sin ( \pi ( k+1)/3 ) } { k!}. \nonumber\end{aligned}\ ] ] by direct inspection the following identity is proven to hold:@xmath245 and , by inserting this into ( [ tre.6 ] ) , we get that@xmath246 we note that , from ( [ tre.5 ] ) , for all @xmath247,@xmath248 \nonumber \\ & = & \frac{w^{1/2}}{3}\biggl [ \sum_{k=0}^{\infty } \biggl ( \frac{w^{3/2}}{3}% \biggr ) ^{2k-1/3}\frac{1}{k!\gamma ( k-1/3 + 1)}\nonumber\\ & & \hspace*{28pt } { } -\sum_{k=0}^{\infty } \biggl ( \frac{w^{3/2}}{3}\biggr ) ^{2k+1/3}\frac{1% } { k!\gamma ( k+1/3 + 1)}\biggr ] \\ & = & \sum_{k=0}^{\infty } \frac{w^{3k}}{3^{2k+2/3}}\frac{1}{k!\gamma ( k+% 2/3)}-\sum_{k=0}^{\infty } \frac{w^{3k}}{3^{2k+4/3}}\frac{1}{% k!\gamma ( k+4/3 ) } \nonumber \\ & = & \frac{2}{3^{7/6}}\sum_{k=0}^{\infty } \biggl ( \frac{w}{3^{2/3}}% \biggr ) ^{k}\frac{\sin ( 2\pi ( k+1)/3)}{\gamma ( ( k+2)/3)\gamma ( % ( k+3)/3)}. \nonumber\end{aligned}\ ] ] the last step can be justified by taking @xmath249 and @xmath250 . while for @xmath251 the last term in ( [ tre.9 ] ) is equal to zero , in the other two cases the two series are obtained . the triplication formula of the gamma function [ see lebedev ( @xcite ) , page 14 ] , that is,@xmath252for @xmath253 yields@xmath254 from ( [ tre.10 ] ) we have that@xmath255and ( [ tre.4 ] ) easily follows by comparing ( [ tre.11 ] ) and ( tre.8 ) .
[ rem4.1 ] the expression of @xmath240 obtained in the previous theorem can be recognized ( up to the factor @xmath256 ) as the solution of the third - order heat - type equation @xmath257evaluated at @xmath258 .
since @xmath259 , for @xmath260 , is positive - valued [ see figure [ fig1](a ) ] and the function ( [ tre.4 ] ) integrates to one ( as we show below ) , @xmath240 is a true probability distribution:@xmath261{3t}}% ai\biggl ( \frac{x}{\lambda \sqrt[3]{3t}}\biggr)\ , dx+\int_{-\infty } ^{0}\frac{1% } { \lambda \sqrt[3]{3t}}ai\biggl ( -\frac{x}{\lambda \sqrt[3]{3t}}\biggr ) \,dx% \biggr ] \\ & & \qquad=2\frac{3}{2}\int_{0}^{+\infty } ai ( y ) \,dy=1,\end{aligned}\]]where the last step follows by noting that @xmath262 ; see nikitin and orsingher ( @xcite ) .
therefore we can think of @xmath240 as the probability law of a process @xmath263 whose distribution at time @xmath69 is obtained from the solution @xmath264 of equation ( [ tre.13 ] ) , as follows:@xmath265 @l@ + ( a)(b ) [ rem4.2 ] for the case @xmath266 the solution @xmath267 to ( equation1 ) can be written , thanks to the relationship ( [ vologda ] ) , as@xmath268\\[-8pt ] & = & \frac{1}{\sqrt{\pi t}}\int_{0}^{\infty } e^{-z^{2}/(4t)}\frac{3^{2/3}% } { 2\lambda z^{1/3}}ai\biggl ( \frac{|x|}{\lambda \sqrt[3]{3z}}\biggr ) \,dz . \nonumber\end{aligned}\ ] ] we can represent ( [ tre.14 ] ) as the distribution of the process@xmath269 with @xmath82 and @xmath67 independent .
the results ( [ tre.4 ] ) and ( [ tre.14 ] ) show that the solutions @xmath240 and @xmath267 are both unimodal with maximum at @xmath270 ; see figure [ fig1](b ) .
this is in accordance with the general result that , for @xmath95 , the solutions to the fractional equation ( [ equation1 ] ) have a unique maximal point at @xmath271 .
we consider now the case @xmath239 , which is qualitatively different from those dealt with so far , because the solutions of fractional equations of order @xmath272 display a substantially different behavior .
[ thm4.2 ] the solution to @xmath273is given by @xmath274 from ( [ equation2 ] ) we have that @xmath275\\[-8pt ] & = & \frac{1}{2\lambda \pi t^{2/3}}\sum_{k=0}^{\infty } \biggl ( -\frac{|x|}{% \lambda t^{2/3}}\biggr ) ^{k}\frac{\gamma ( 2/3(k+1))\sin ( 2\pi ( k+1)/3)}{k!}. \nonumber\end{aligned}\ ] ] by means of the duplication formula for the gamma function we have that@xmath276 and therefore @xmath277 can be rewritten as@xmath278 \nonumber\\ & = & \frac{3^{2/3}}{2\lambda \sqrt{\pi } 2^{1/3}t^{2/3}}\int_{0}^{+\infty } e^{-w}w^{-1/6}ai\biggl ( -\frac{|x|}{\lambda } \biggl ( \frac{2}{t}\sqrt{% \frac{w}{3}}\biggr ) ^{2/3}\biggr ) \,dw.\nonumber\end{aligned}\ ] ] we can show that @xmath279 .
indeed , from ( [ qua3.bis ] ) we have that@xmath280 \\ & & \qquad=\frac{2}{\lambda \sqrt{\pi } } \biggl ( \frac{3}{4t}\biggr ) ^{2/3}\lambda \biggl ( \frac{2}{t}\sqrt{\frac{1}{3}}\biggr ) ^{-2/3}\int_{0}^{+\infty } e^{-w}w^{-1/2}\int_{-\infty } ^{0}ai ( y ) \,dy\,dw \\ & & \qquad=\frac{1}{\sqrt{\pi } } \frac{2}{3}2\biggl ( \frac{3}{4}\biggr ) ^{2/3}% \frac{2^{-2/3}}{3^{-1/3}}\int_{0}^{+\infty } e^{-w}w^{-1/2}\,dw=1.\end{aligned}\ ] ] [ rem4.3 ] in view of theorem [ thm1 ] we have the following representation for @xmath240 , which is alternative to ( [ tre.4]):@xmath281 by inserting ( [ tre.4 ] ) into the left - hand side of ( [ qua.6 ] ) we obtain that@xmath282{3t}}ai\biggl ( \frac{|x|}{\lambda \sqrt[3]{3t}}\biggr)\nonumber \\ & & \qquad=\frac{3^{2/3}}{2\lambda \sqrt{\pi } 2^{1/3}}\int_{0}^{+\infty } \frac{e^{-% z^{2}/(4t)}}{\sqrt{\pi t}z^{2/3}}\,dz\nonumber\\ & & \quad\qquad{}\times\int_{0}^{+\infty } e^{-w}w^{-1/6}ai\biggl ( -\frac{|x|}{\lambda } \biggl
( \frac{2}{z}\sqrt{\frac{w}{3}}% \biggr ) ^{2/3}\biggr)\ , dw \nonumber \\ & & \qquad=\bigl [ \mbox{by the substitution } s=\sqrt[3]{2^{2}twz^{-2}}\bigr ] \nonumber \\ & & \qquad=\frac{3^{2/3}}{2\lambda \sqrt{\pi } 2^{1/3}}\int_{0}^{+\infty } \frac{e^{-% z^{2}/(4t)}}{\sqrt{\pi t}z^{2/3}}\frac{3z^{5/3}}{2^{5/3 } t^{5/6}}s^{3/2}\,dz\\ & & \quad\qquad{}\times\int_{0}^{+\infty } e^{-z^{2}s^{3}/(4t)}ai\biggl ( -\frac{|x|s}{\lambda \sqrt[3]{3t}}\biggr ) \,ds
\nonumber \\ & & \qquad=\frac{3^{5/3}}{2^{3}\lambda \pi t^{4/3}}\int_{0}^{+\infty } ai\biggl ( -% \frac{|x|s}{\lambda \sqrt[3]{3t}}\biggr ) s^{3/2}\,ds\int_{0}^{+\infty } ze^{-z^{2}(1+s^{3})/(4t)}\,dz \nonumber \\ & & \qquad=\frac{3^{5/3}}{2^{2}\lambda \pi t^{1/3}}\int_{0}^{+\infty } \frac{s^ { 3/2}}{1+s^{3}}ai\biggl ( -\frac{|x|s}{\lambda \sqrt[3]{3t}}\biggr )
\,ds \nonumber \\ & & \qquad=\frac{3^{2/3}}{2\lambda t^{1/3}}\int_{0}^{+\infty } \pr \ { |b ( t_{0 } ) |\in ds\ } ai\biggl ( -\frac{|x|s}{\lambda \sqrt[3]{3t}}% \biggr ) , \nonumber\end{aligned}\]]where @xmath283 is the mckean law representing the distribution of the position of a brownian motion @xmath67 at the instant @xmath284 see mckean ( @xcite ) . by setting @xmath285{3t}}$ ] in ( [ qua.ai ] ) and performing some simplifications we get@xmath286
formula ( [ airy6 ] ) shows an interesting property of airy functions : the value of the exponentially decreasing part of @xmath287 can be obtained by averaging its oscillating component @xmath288 with the well - known density of @xmath289 ( see figure [ fig1 ] ) .
[ rem4.4 ] the solution @xmath277 can also be expressed in terms of a stable density of order @xmath290 indeed , by using the representation of the stable density below@xmath291we know that for @xmath292 , @xmath293 and for @xmath294 the following series representation holds true:@xmath295see formula ( 6.9 ) , page 583 of feller ( @xcite ) ( up to some corrections ) and lukacs ( @xcite ) . for @xmath296 and @xmath297 formula ( [ qua.7 ] ) reads@xmath298\\[-8pt ] & = & \frac{2}{3}\frac{1}{\pi } \sum_{r=0}^{\infty } ( -x ) ^{r}\frac{% \sin \ { ( r+1)2/3\pi \ } } { r!}\gamma \biggl ( \frac{2}{3}% ( r+1)\biggr ) . \nonumber\end{aligned}\ ] ]
if we compare ( [ qua.4 ] ) with ( [ qua.8 ] ) we get that@xmath299 a different proof of the relationship between stable laws and the solutions of fractional diffusion equations , based on the inversion of the fourier transform , can be found in fujita ( @xcite ) .
formula ( [ qua.9 ] ) proves the nonnegativity of the expression ( qua.5 ) , as a function of @xmath87 .
in this section we present some generalizations of the results of sections [ s2 ] and [ s4 ] .
we start by giving a relationship between the solutions @xmath7 and @xmath70 , @xmath71 , and obtain some explicit expressions for @xmath300 . in this case
the interpretation of @xmath89 as the distribution of compositions of different types of processes is possible .
also in this case we encounter processes with a random time which possesses a branching structure ( depending on @xmath5 ) .
we now state a general result which is alternative to ( [ vologda ] ) and permits us to exploit the explicit expression of @xmath158 .
[ thm5.1 ] the solution to the initial value problem ( [ equation1])([condition 1 ] ) , for @xmath301 , can be represented as @xmath302 where @xmath303 is the solution to @xmath304and@xmath305 in view of the triplication formula ( [ tripl ] ) , for @xmath306 , we have that @xmath307{3^{3}wzt})^{3\nu /2}}\biggr ) ^{k}\frac{dw\,dz}{k!\gamma ( 1-% 3\nu ( k+1)/2 ) } \nonumber \\ & = & \frac{\sqrt{3}}{2\pi 3^{3/2\nu } t^{\nu /2}}\int_{0}^{+\infty } \int_{0}^{+\infty } e^{-w - z}w^{-\nu /2 - 1/3}z^{-\nu /2 -2/3}\nonumber\\ & & \phantom{\frac{\sqrt{3}}{2\pi 3^{3/2\nu } t^{\nu /2}}\int_{0}^{+\infty } \int_{0}^{+\infty } } { } \times\bigl(\sqrt[3]{3^{3}wzt}\bigr)^{3\nu /2}u_{3\nu } \bigl(x,\sqrt[3]{% 3^{3}wzt}\bigr)\,dw\,dz \\ & = & \frac{\sqrt{3}(3^{3}t)^{\nu /2}}{2\pi 3^{3/2\nu } t^{\nu /2}}\int_{0}^{+\infty } \int_{0}^{+\infty } e^{-w - z}w^{-\nu /2 - 1% /3}z^{-\nu /2 - 2/3}\nonumber\\ & & \phantom{\frac{\sqrt{3}(3^{3}t)^{\nu /2}}{2\pi 3^{3/2\nu } t^{\nu /2}}\int_{0}^{+\infty } \int_{0}^{+\infty } } { } \times(wz)^{\nu /2}u_{3\nu } \bigl(x,\sqrt[3]% { 3^{3}wzt}\bigr)\,dw\,dz \nonumber \\ & = & \frac{\sqrt{3}}{2\pi } \int_{0}^{+\infty } \int_{0}^{+\infty } e^{-w - z}w^{-% 1/3}z^{-2/3}u_{3\nu } \bigl(x,3\sqrt[3]{wzt}\bigr)\,dw\,dz , \nonumber\end{aligned}\]]which reduces to ( [ cin.2 ] ) , after the change of variables@xmath308{w}\sqrt[3]{t^{1/2 } } , \cr v=\sqrt{3}\sqrt[3]{z}\sqrt[3]{t^{1/2}}. } \ ] ] it can be easily checked that , also in this form , the solution integrates to one . by using the last expression in ( [ cin.4 ] ) we get@xmath309{% wzt}\bigr)\,dx\,dw\,dz \\ & & \qquad=\frac{\sqrt{3}}{2\pi } \int_{0}^{+\infty } e^{-w}w^{-1 + 2/3 } \,dw\int_{0}^{+\infty } e^{-z}z^{-1 + 1/3}\,dz \\ & & \qquad=\frac{\sqrt{3}}{2\pi } \gamma \biggl ( \frac{2}{3}\biggr ) \gamma \biggl ( \frac{1}{3}\biggr ) = 1,\end{aligned}\]]since , by the triplication formula for @xmath310 , it is @xmath311 [ rem5.1 ] by using the previous result it is possible to obtain alternative forms for the solution to the initial value problem for @xmath312 and for @xmath313 indeed , in the first case it is@xmath314\\[-8pt ] & = & \frac{3}{2\pi \sqrt{t}}\int_{0}^{+\infty } \int_{0}^{+\infty } se^{- ( s^{3}+v^{3})/(3\sqrt{3t})}\frac{e^{-x^{2}/(4\lambda ^{2}(sv))}}{% 2\lambda \sqrt{\pi sv}}\,ds\,dv .
\nonumber\end{aligned}\ ] ] the relationship ( [ cin.5 ] ) shows that @xmath80 can be interpreted as the distribution of a brownian motion ( with infinitesimal variance @xmath315 ) at a random time @xmath316 , that is,@xmath317 , \label{cin.5bis}\]]where @xmath318 possesses joint density@xmath319 this result corresponds to ( [ iterated brownian motion ] ) , for @xmath312 and it represents a counterpart of result ( [ iterated ] ) with the reflecting brownian motion replaced by the product @xmath320 with joint distribution given in ( [ cin.6 ] ) . in the case
@xmath321 , from ( [ cin.2 ] ) we have that@xmath322and this suggests that we interpret @xmath323 as the distribution of the process@xmath324 .
\label{airy}\ ] ] the process ( [ airy ] ) is analogous to ( [ cin.5bis ] ) with the role of brownian motion played by the process @xmath82 .
analogously to ( [ airy ] ) , for @xmath325 , we get@xmath326 \cdot g_{2 } [ g_{1}(t)\cdot g_{2}(t ) ] \}\ ] ] which has distribution coinciding with@xmath327as an application of ( [ cin.2 ] ) and ( [ nuova ] ) shows .
the results of theorem [ thm1 ] and [ thm5.1 ] can be furthermore generalized in order to relate the solutions @xmath158 with @xmath328 .
[ thm5.2 ] the solution to equation ( [ eq4 ] ) , for @xmath329 , @xmath330 can be represented as @xmath331{m^{m}t}}w_{2}\cdots w_{m-1}^{m-2}\\ & & \hspace*{77pt } { } \times u_{m\nu } ( x , w_{1}w_{2}\cdots w_{m-1 } ) \,dw_{1}\cdots dw_{m-1}. \nonumber\end{aligned}\ ] ] from ( [ equation2 ] ) , by using the multiplication formula of the gamma function [ see magnus and oberhettinger ( @xcite ) ] , that is,@xmath332for @xmath333 , we get that@xmath334^{-1 } \\ & & \qquad=\frac{\sqrt{m}}{2\lambda t^{\nu /2}(2\pi ) ^{(m-1)/2}m^{m/2\nu } } \\ & & \quad\qquad { } \times\int_{0}^{+\infty } \cdots\int_{0}^{+\infty } e^{-w_{1}-\cdots - w_{m-1}}w_{1}^{-% 1/m}w_{2}^{-2/m}\cdots w_{m-1}^{-(m-1)/m } \\ & & \quad\qquad\phantom{{}\times\int_{0}^{+\infty } \cdots\int_{0}^{+\infty } } { } \times u_{m\nu } \bigl ( x , m\sqrt[m]{w_{1}\cdots w_{m-1}t}\bigr)\ , dw_{1}\cdots dw_{m-1}.\end{aligned}\ ] ] by means of the transformation@xmath335{m}\sqrt[m]{w_{j}}\sqrt[m]{t^{1/(m-1)}},\ ] ] we finally get ( [ sei.1 ] ) .
we prove now a general result , valid for any @xmath336 , which gives another representation for the solution @xmath0 , alternative to those presented in the previous sections .
[ thm5.3 ] the solution to ( [ equation1 ] ) with initial condition ( [ condition 1 ] ) or ( [ cond1 ] ) has the following form : @xmath337for @xmath336 . by applying the reflection property of the gamma function
we rewrite ( [ equation2 ] ) as@xmath338\\[-8pt ] & & \phantom{\frac{1}{2\pi \lambda t^{\nu /2}}\int_{0}^{+\infty } } { } \times\sum_{k=0}^{\infty } \biggl [ \biggl ( -\frac{|x|w^{\nu /2}}{\lambda t^{\nu /2}}\biggr ) ^{k}\frac{1}{k!}\frac{e^{i\nu \pi ( k+1)/2}-e^{-% i\nu \pi ( k+1)/2}}{2i}\biggr ] \,dw \nonumber \\ & = & \frac{1}{2\pi \lambda t^{\nu /2}}\nonumber\\ & & { } \times\int_{0}^{+\infty } e^{-w}\frac{w^ { \nu /2 - 1}}{2i}\nonumber\\ & & \phantom{{}\times\int_{0}^{+\infty } } { } \times\bigl [ e^{-|x|w^{\nu /2}e^{i\nu /2\pi } / ( \lambda t^{\nu /2})}e^{i\nu /2\pi } \nonumber\\ & & \hspace*{54pt } { } -e^{-|x|w^{\nu /2 } e^{-i\nu /2\pi } /(\lambda t^{\nu /2})}e^{-i\nu /2\pi } \bigr ] \,dw , \nonumber\end{aligned}\]]which coincides with the first form of ( [ sin ] )
. the second line can be obtained by the change of variable @xmath339 .
[ rem5.2 ] we can check that , for @xmath340 ( i.e. , for the heat equation ) , the first expression in ( [ sin ] ) reduces to the gaussian density : @xmath341\\[-8pt ] & = & [ w = y^{2 } ] \nonumber \\ & = & \frac{1}{\sqrt{4\pi t\lambda
^{2}}}e^{-x^{2}/(4t\lambda ^{2})}. \nonumber\end{aligned}\]]in the last step we used formula 3.896.4 , page 514 , of gradshteyn and ryzhik ( @xcite ) . the same check can be done for the second expression in ( [ sin ] ) .
an alternative form of ( [ sin ] ) can be obtained by means of a double integration by parts , as follows : [ coro5.1 ] the solution to ( [ equation1 ] ) with initial condition ( [ condition 1 ] ) or ( [ cond1 ] ) can be rewritten as @xmath342\\[-8pt ] & & \phantom{\frac{1}{\pi \nu |x|}\int_{0}^{+\infty } } { } \times\sin \biggl ( \frac{% \nonumber\end{aligned}\]]for @xmath336 .
the first integration in ( [ sin ] ) gives @xmath343 therefore , from ( [ sin ] ) we have that@xmath344\ , dw \\ & & \qquad=\frac{1}{\pi \nu |x|\sin ( \nu \pi /2)}\int_{0}^{+\infty } e^{-w}e^{-|x|w^{\nu /2}/(\lambda
( \nu \pi /2 ) } \\ & & \quad\qquad{}\times \biggl\{\biggl [ \cos \biggl(\frac{\nu \pi } { 2}\biggr)\cos \biggl ( \frac{|x|w^{\nu /2}}{% \lambda t^{\nu /2}}\sin \biggl(\frac{\nu \pi } { 2}\biggr)\biggr)\\ & & \hspace*{56pt } { } + \sin \biggl(\frac{\nu \pi } { 2}% \biggr)\sin \biggl ( \frac{|x|w^{\nu /2}}{\lambda t^{\nu /2}}\sin \biggl(\frac{\nu \pi } { 2}% \biggr)\biggr ) \biggr ] \\ & & \quad\qquad\hspace*{17pt}{}-\cot \biggl(\frac{\nu \pi } { 2}\biggr)\biggl [ \sin \biggl(\frac{\nu \pi } { 2}\biggr)\cos \biggl ( \frac{|x|w^{\nu /2}}{\lambda t^{\nu /2}}\sin \biggl(\frac{\nu \pi } { 2}\biggr)\biggr)\\ & & \hspace*{110pt } { } -\cos \biggl(\frac{\nu \pi } { 2}\biggr)\sin \biggl ( \frac{|x|w^{\nu /2}}{\lambda t^{\nu /2}% } \sin \biggl(\frac{\nu \pi } { 2}\biggr)\biggr ) \biggr ] , \end{aligned}\]]which easily gives ( [ sin.3 ] ) . [ rem5.3 ]
we can check that , for @xmath340 , ( [ sin.3 ] ) reduces again to the gaussian density : @xmath345 \\ & = & \frac{\lambda ^{2}t}{\pi |x|^{3}}\int_{0}^{+\infty } ye^{-y^{2}/2 \lambda ^{2}t/|x|^{2}}\sin \frac{y}{\sqrt{2}}\,dy \\ & = & \frac{1}{\sqrt{2}\pi |x|}\int_{0}^{+\infty } e^{-y^{2}/2 \lambda ^{2}t/|x|^{2}}\cos \frac{y}{\sqrt{2}}\,dy \\ & = & \frac{1}{2\sqrt{\pi t\lambda ^{2}}}e^{-x^{2}/(4t\lambda ^{2})}\end{aligned}\]]as in ( [ gaus ] ) . with respect to ( [ sin ] ) , formula ( [ sin.3 ] ) is more appealing as it allows an easier analysis of the limit for @xmath346:@xmath347 for @xmath348 , ( [ lim ] ) decreases for all values of @xmath349 .$ ] moreover in the case @xmath340 , formula ( [ lim ] ) gives the maximum value of the brownian density . for @xmath350 ( [ lim ] ) is zero for all @xmath44 , because in this case ( [ equation1 ] ) becomes the wave equation and its solution has the form of the sum of dirac s impulse functions travelling in opposite directions . by means of the following formula@xmath351
[ gradshteyn and ryzhik ( @xcite ) , formula 3.941.1 , page 523 ] we can check that ( [ sin.3 ] ) integrates to one , as follows:@xmath352 , for @xmath353 .
( ) the solution @xmath354 , for @xmath355 . ] finally it is interesting to analyze the behavior of the solution ( for @xmath87 varying and @xmath69 fixed ) , which is substantially different in the two intervals @xmath95 and @xmath356 ( see figure [ fig2 ] above ) .
we rewrite formula ( [ sin.3 ] ) as follows : for @xmath357@xmath358 where @xmath359 and @xmath360 .
the first derivative of @xmath361 with respect to @xmath87 is equal to zero if@xmath362where@xmath363 the solution to ( [ eq ] ) is@xmath364 or , otherwise , @xmath365 by choosing @xmath366 , we obtain that @xmath158 attains its maximum on the positive half - line if @xmath367 for @xmath356 there exists only one value of @xmath87 which verifies the condition ( [ max ] ) and this is in accordance with the behavior of the solutions @xmath7 presented in fujita ( @xcite ) , where the relationship with stable laws is exploited . on the other hand , for @xmath368 no positive value satisfies ( [ max ] ) and
therefore the maximum is in the origin .
the previous results are confirmed by the following theorems .
we now present the general results concerning the relationship between the solution @xmath158 and the stable densities .
we need to analyze the two intervals @xmath95 and @xmath356 separately .
[ thm5.4 ] for @xmath95 , the solution to @xmath369can be represented as @xmath370\\[-8pt ] & = & \frac{1}{\nu |x|^{2/\nu + 1}}p_{\nu /2}\biggl ( \frac{1}{% \biggr ) \nonumber\end{aligned}\]]where @xmath371 is the density of a stable distribution of parameters @xmath372 and @xmath293 ; see ( [ stab ] ) . from ( [ equation2 ] ) , by using the reflection formula for the gamma function we have that@xmath373\\[-8pt ] & = & \frac{1}{2\lambda t^{\nu /2}}\sum_{k=0}^{\infty } \biggl ( -\frac{|x|}{% \lambda t^{\nu /2}}\biggr ) ^{k}\frac{1}{k!}\frac{\sin ( \pi \nu ( k+1)/2)% } { \pi } \gamma \biggl ( \frac{\nu ( k+1)}{2}\biggr ) .
\nonumber\end{aligned}\ ] ] in view of the series representation of stable functions , which for @xmath374 reads@xmath375\ ] ] [ see feller ( @xcite ) , formula ( 6.10 ) , page 583 , with some corrections , lukacs ( @xcite ) and zolotarev ( @xcite ) ] , we can obtain the first expression in ( [ stable.1 ] ) .
the second expression can be derived by applying the self - similarity property of the stable random variables .
finally we consider the case @xmath376 and we state the following result : [ thm5.5 ] the solution to @xmath377for @xmath376 , can be represented as @xmath378\\[-8pt ] & = & \frac{1}{\nu } p_{2/\nu } \biggl ( |x|;\frac{2}{\nu } ( \nu -1),\lambda ^{2/\nu } t\biggr ) , \nonumber\end{aligned}\ ] ] where @xmath379 is the density of a stable distribution of parameters @xmath380 and @xmath293 . by following the same steps as in the previous theorem we can recognize in ( [ stable.2 ] ) , up to the normalizing constant , the series representation of the stable law @xmath381 of order @xmath382 [ see ( [ qua.7 ] ) ] , so that we get ( [ gen.2 ] ) .
[ rem5.4 ] in view of theorems [ thm2.3 ] and [ thm5.5 ] and by considering the property of self - similarity of the stable laws , we can write that@xmath383\\[-8pt ] & = & \frac{1}{\nu } \int_{0}^{\infty } \frac{e^{-x^{2}/(4w\lambda)}}{% \sqrt{4\pi w\lambda } } p_{1/\nu } \biggl ( |w|,\frac{1}{\nu } ( 2\nu -1),\lambda ^{1/\nu } t\biggr)\ , dw . \nonumber\end{aligned}\ ] ] formula ( [ stable.7 ] ) shows that the solution @xmath384 for @xmath385 can be interpreted as the distribution of the process @xmath386 @xmath44 , where @xmath387 is the stable process with density @xmath388 .
moreover , as a consequence of theorems 2.1 and 5.5 , the solution of our problem ( [ equation1])([condition 1 ] ) , for @xmath389 , can be written in an alternative to the form ( [ vologda ] ) also as a stable law evaluated at a brownian time:@xmath390 [ rem5.5 ] we check that , for @xmath340 , both expressions ( [ stable.1 ] ) and ( gen.2 ) yield the gaussian density@xmath391 we start by considering the last expression in ( [ stable.1])@xmath392where [ from ( [ stab ] ) ] , for @xmath393@xmath394\\[-8pt ] & = & \frac{1}{\sqrt{2}\lambda t^{1/2}}\frac{e^{-1/(2y(\sqrt{2}\lambda t^{1/2})^{2})}}{\sqrt{2\pi y^{3}}}. \nonumber\end{aligned}\ ] ] by taking in ( [ gen.8 ] ) @xmath395 we get from ( [ gen.7 ] ) the gaussian density ( [ gen.6 ] ) . formula ( [ gen.2 ] ) immediately supplies ( [ gen.6 ] ) for @xmath396
the authors thank one anonymous referee for bringing their attention to some relevant papers on fractional equations .
thanks are also due for his accurate check of the text and of the calculations . | in this paper the solutions @xmath0 to fractional diffusion equations of order @xmath1 are analyzed and interpreted as densities of the composition of various types of stochastic processes .
for the fractional equations of order @xmath2 , @xmath3 we show that the solutions @xmath4 correspond to the distribution of the @xmath5-times iterated brownian motion . for these processes the distributions of the maximum and of the sojourn time
are explicitly given .
the case of fractional equations of order @xmath6 , @xmath3 is also investigated and related to brownian motion and processes with densities expressed in terms of airy functions . in the general case we show that @xmath7 coincides with the distribution of brownian motion with random time or of different processes with a brownian time .
the interplay between the solutions @xmath7 and stable distributions is also explored .
interesting cases involving the bilateral exponential distribution are obtained in the limit . and . . |
let @xmath1 be a coxeter system .
@xmath2 is called coxeter group of _ type @xmath3 _
if @xmath2 has an effective , proper and cocompact action on some contractible manifold of dimension @xmath4 , in this case , @xmath4 is called the _ dimension _ of @xmath2 .
we say a coxeter group of type @xmath0 if it is of type @xmath3 for some @xmath4 .
examples of coxeter groups of type @xmath0 are compact reflection groups on @xmath5 . in @xcite potyagailo and vinberg
give the bound for the dimension of right - angled coxeter groups which have fundamental domain of finite volume and construct examples of those groups up to dimension 8 . from those examples if one truncate the vertices at @xmath6 and introduces new reflecting faces , corresponding to those truncated vertices , one can obtain new coxeter groups of type hm . by caprace s criterion @xcite
these groups are relatively hyperbolic relative to their affine special subgroups of codimension 1 ( if all maximal affine subgroups of a coxeter group are of codimesion 1 , we say that the group has maximal flats of codimesion 1 ) .
a question we can ask about relatively hyperbolic coxeter groups of type @xmath0 with maximal flats of codimension 1 is that [ question : dimension ] is there an upper bound on the dimension @xmath4 of relatively hyperbolic coxeter groups of type @xmath3 with maximal affine subgroups of codimension 1 ?
this question is inspired by the example of relatively hyperbolic coxeter groups of type @xmath0 described above and the result of the paper @xcite .
in addition , the boundedness of dimension is also proven for reflection groups in hyperbolic space which have fundamental domain of finite volume ( @xcite ) . in this paper
we prove some boundedness properties for relatively hyperbolic coxeter groups of type @xmath0 with maximal flats of codimension 1 .
in particular , we will prove the two following theorems : [ theorem : rightangled_case ] let @xmath2 be a right - angled coxeter group of type @xmath0 with maximal euclidean coxeter subgroups of codimension 1 .
then the dimension of @xmath2 does not exceed 14 .
a similar result is also hold for general coxeter groups of type @xmath0 .
[ theorem : general_case ] let @xmath2 be a relatively hyperbolic coxeter group of type hm with flats of codimension 1 .
then the dimension of @xmath2 is less than 996 .
notice that a similar result of nonexistence of gromov hyperbolic coxeter groups of type @xmath0 is proven by januszkiewicz and swiatkowski in @xcite , which is an extension of a result of e. vinberg .
the paper is structured as follows .
section [ sec : basic ] contains basic definitions and notations .
we prove theorem [ theorem : rightangled_case ] in section [ sec : rightangled_case ] and theorem [ theorem : general_case ] in section [ sec : general_case ] .
section [ sec : technical ] contains some technical points needed for the proofs and the last section contains some examples of relatively hyperbolic coxeter groups .
i would like to thank my advisor , mike davis , for insightful discussions .
recall that a coxeter group is a group @xmath2 that has a presentation of the form : @xmath7 where @xmath8 and @xmath9 are two sets , @xmath10 = 1 and @xmath11 for @xmath12 ( @xmath13 means there is no relation between @xmath14 and @xmath15 ) .
@xmath1 is called a _ coxeter system_. + given a coxeter system @xmath1 the caley 2-complex of @xmath2 can be completed to a complex @xmath16 , called the _
davis complex _ ( see @xcite ) .
the link @xmath17 of each vertex in @xmath16 is a simplical complex which has exactly @xmath18 vertices , labeled by elements of @xmath8 and any @xmath19 vertices @xmath20 spans a @xmath21simplex if and only if the subgroup generated by @xmath20 of @xmath2 is finite .
the matrix @xmath22 is called _
coxeter matrix_. there is a method , due to coxeter , of encoding the information in a coxeter matrix @xmath23 into a graph @xmath24 with edges labeled by integers greater than 3 or the symbol @xmath6 .
this graph is called _ coxeter graph_.
the vertex set of @xmath24 is @xmath9 .
two vertices @xmath25 and @xmath26 are connected by an edge if and only if @xmath27 . the edge @xmath28 is labeled by @xmath29 if @xmath30 .
the graph @xmath24 together with the labeling of its edges is called the _ coxeter diagram _ associated to @xmath23 .
the diagram is called _ positive definite ( elliptic ) , semidefinite ( parabolic ) , _ or _
indefinite _ if the corresponding matrix @xmath23 has this property .
the distance @xmath31 between vertices @xmath32 and @xmath33 of a diagram @xmath8 is defined to be the length ( the number of edges ) of the shortest path joining @xmath32 and @xmath33 . if @xmath32 and @xmath33 are not joined by any path , then @xmath34 .
recall that an indefinite diagram whose proper sub - diagrams are all either elliptic or parabolic is called a _ quasi - lanner diagram_. the list of all quasi - lanner diagrams can be found in @xcite .
it is important for the proof of theorem [ theorem : general_case ] that the distance of any two vertices in a quasi - lanner diagram does not exceed 8 , and there is at most one pair of vertices which has distance equal to 8 .
another way to characterize coxeter group of type @xmath0 is via the link @xmath17 of @xmath2 or via the quotient of @xmath16 by @xmath2 .
first , recall that a space @xmath35 is a _ generalized homology
n - sphere _ if it is a homology n - manifold with the same homology as @xmath36 .
a pair @xmath37 is a _ generalized homology @xmath4-disk _ if it is a homology n - manifold with boundary and if it has the same homology as @xmath38 .
when the coxeter group is type @xmath3 , the link @xmath39 is a generalized homology @xmath40-sphere and the fundamental domain @xmath41 is a so - called _ generalized simple polytope _ ( or a generalized polytope for short ) , i.e. @xmath42 is a generalized homology @xmath4-disk and for each @xmath43 , @xmath44 is a generalized homology disk of dimension @xmath45 , where @xmath46 is a subset of the set of generators of @xmath2 such that the subgroup @xmath47 generated by elements in @xmath46 is a finite group and @xmath48 is the fundamental domain for @xmath47 ( for proof see @xcite , chapter 10 ) . in this paper , we will use these characteristics of coxeter group of type @xmath0 to prove our theorems .
a convex bounded @xmath4-dimensional polyhedron @xmath49 is called _ simple _
( respectively , _ almost simple _ ) if the links of its vertices are simplices ( respectively , direct products of simplices ) . a polyhedron @xmath41 is called _ edge - simple _ if the links of its vertices are simple polyhedrons ( or equivalently , the links of its edges are simplices ) .
note that an almost simple polyhedron is edge - simple and if @xmath41 is one of the above three types then its faces are polyhedra of the same type as @xmath41 .
denote @xmath50 the numbers of faces of dimensions @xmath51 of @xmath41 , respectively , and @xmath52 denotes the average number of @xmath25-faces of a @xmath19-face of @xmath41 ( here @xmath53 is the number of @xmath25-faces of a face @xmath24 ) .
the following theorem is due to nikulin @xcite .
[ theorem : nikulin ] for every simple convex bounded polyhedron @xmath49 for @xmath54 $ ] the following estimate holds : @xmath55 } + \binom{i}{[(n+1)/2]}}{\binom{k}{[n/2 ] } + \binom{k}{[(n+1)/2]}}\ ] ] khovanskii extends this result to edge - simple polyhedron @xcite .
[ theorem : khovanskii ] the estimate from theorem [ theorem : nikulin ] holds for an edge - simple polyhedron @xmath49 .
we will use these inequalities for our fundamental domain @xmath41 of the coxeter group @xmath2 of type @xmath0 . here
@xmath41 is generally not a polyhedron in @xmath56 , but the links of its vertices are simplices ( faces in @xmath17 ) and we can still apply these inequalities to @xmath41 since the proofs of theorems [ theorem : nikulin ] and [ theorem : khovanskii ] use only combinatorial properties of a polyhedron .
in theorem [ theorem : rightangled_case ] and theorem [ theorem : general_case ] we are considering coxeter group of type @xmath0 with maximal flats ( maximal euclidean subgroups ) of codimension 1 .
type @xmath0 means that the nerve @xmath17 is a generalized homology sphere of dimension @xmath57 and the fundamental domain @xmath41 is a generalized simple polytope ( of dimension @xmath4 ) .
the condition that all maximal flats are of codimension 1 guarantees that the nerves of those flats are of correct dimension ( i.e , of dimension @xmath58 ) and they are also of type @xmath0 .
it is easy to see that for any nerve @xmath59 of an euclidean subgroup of @xmath2 of codimension 1 , if we cut @xmath17 along @xmath59 , we get two connected components ( alexander duality ) .
we say that a nerve @xmath59 of an euclidean subgroup of @xmath2 ( of codimension 1 ) lies in the boundary of @xmath17 if at least one of those connected components is a cone on @xmath59 . to prove theorem [ theorem : rightangled_case ] and theorem [ theorem : general_case ] it is crucial that the nerves of all maximal euclidean subgroups lie in the boundary of @xmath17 .
this can be assumed since if @xmath59 is not in the boundary of @xmath17 we can cut @xmath17 along @xmath59 and cone off @xmath59 in each connected component .
each new piece we get from this cutting process is again a generalized homology sphere ( a simple mayer - vietoris sequence argument ) and has @xmath59 in the boundary .
the opposite process of cutting is gluing .
if two nerves @xmath59 and @xmath60 of relatively hyperbolic coxeter groups of type @xmath3 with isolated flats of codimension 1 have the same flat , then we can glue @xmath59 and @xmath60 along the flat and get a new nerve of the same type .
in this section we prove theorem [ theorem : rightangled_case ] .
let @xmath2 be a right - angled coxeter group of type @xmath0 with maximal euclidean subgroups of codimension 1 . applying the cutting process to the nerve @xmath17 of @xmath2
, we can assume that the nerves of euclidean subgroups of codimension 1 of @xmath2 lie in the boundary of @xmath17 .
we follow the proof of the main theorem in potyagailo - vinberg s paper @xcite . in this paper
the authors prove the boundedness for right - angled reflection groups in hyperbolic spaces of finite volume , i.e. , the fundamental domain of the action of a right - angled reflection group on hyperbolic space , which is a polyhedron , has finite volume .
our case is similar to this one in the sense that the cubes of codimension 1 ( these are the fundamental domain for euclidean subgroups ) can be thought as cusps for the polyhedron . to use their proof we collapse each cube in the ( generalized ) polytope @xmath41 to a point to get another polytope @xmath61 .
the condition that cubes are isolated in this situation guarantees that @xmath61 is actually a generalized polytope .
we call a vertex in @xmath61 a _ cusp _ if it is the vertex gotten from collapsing a cube
. two codimension 1 faces of @xmath61 are _ parallel _ if they meet at a cusp but do not otherwise intersect .
we use @xmath62 to denote a face of @xmath61 gotten from face @xmath63 of @xmath41 by collapsing cubes in @xmath41 . observe that 1 .
@xmath61 is not a simple generalized polytope as @xmath41 , but it is an edge - simple generalized polytope because the links of edges in @xmath61 are the same as the links of the same edges in @xmath41 . and
we can still apply the nikulin - khovanskii inequality for @xmath61 .
2 . since @xmath41 is simple , any face of @xmath41 is also simple . and
if a face @xmath63 has non - empty intersection with a cube of codimension 1 of @xmath41 then the intersection is a a cube of codimension 1 of @xmath63 ( so the above definition of cusp is still valid for faces of @xmath61 ) .
let @xmath64 be faces of codimension 1 of @xmath61
. then 1 .
[ first ] if @xmath65 are pairwise mutually adjacent , then they meet at a @xmath66-dimensional face ; 2 .
[ second ] if @xmath67 and @xmath68 are parallel and @xmath69 is adjacent to them ( @xmath67 and @xmath68 are not adjacent ) , then @xmath70 meet at a cusp ; 3 . [ third ] if @xmath67 and @xmath68 are parallel and @xmath69 and @xmath71 are adjacent to them , then @xmath72 meet at a cusp .
part [ first ] follows from the property of the nerve @xmath17 that @xmath17 is a _ flag complex _
( recall that a flag complex is a simplicial complex in which any finite collection of vertices is a simplex if there are edges connecting any two vertices from the collection , and since @xmath2 is right - angled coxeter group the nerve @xmath17 is a flag complex ) . if @xmath65 are pairwise mutually adjacent , then the same is true for @xmath73
. then the vertices @xmath74 in the nerve @xmath17 corresponding to @xmath73 are connected by edges .
since @xmath17 is a flag complex , @xmath74 form a triangle in @xmath17 .
this means @xmath73 meet at a @xmath75dimensional face .
part [ second ] : 4 vertices @xmath76 in the nerve @xmath17 corresponding to @xmath73 and the cusp form a square .
if this square is an empty square , it must be part of an octahedron face in @xmath17 ( dual of the cube face of @xmath41 ) because of the isolated cube condition of @xmath41 . in this case
all the faces @xmath73 have intersection with the same cube in @xmath41 , so they meet at a cusp
. if the square is not empty , there must be an edge connecting two opposite vertices . and
it s not in the case of part [ first ] ( no edge between @xmath77 and @xmath78 ) then there must be an edge connecting @xmath79 and @xmath80 .
part [ third ] : let @xmath76 are vertices in @xmath17 corresponding to @xmath81 , and @xmath82 corresponding to the two cubes ( cusps ) which @xmath65 and @xmath65 meet following part [ second ] . if @xmath83 and @xmath84 are the same , then nothing needs to prove . if not , then @xmath82 are cone points over the nerves @xmath85 of euclidean subgroups of @xmath2 .
now @xmath59 and @xmath60 have two common vertices @xmath86 which are not connected by an edge , they must be the same nerve of an euclidean subgroup of @xmath2 by the criterion of isolated flat for relatively hyperbolic coxter group ( @xcite ) .
then in this case @xmath17 is the suspension of @xmath59 and thus @xmath2 has dimension @xmath57 which is a contradiction .
now we will consider faces of dimensions 2 , 3 , 4 and 5 of @xmath61 .
denote @xmath87 the number of @xmath19-faces of @xmath61 .
in particular , @xmath88 is the total number of ordinary vertices and cusps of @xmath61 .
the number of cusps will be denoted by @xmath89 . for each case below , the number of faces is for the corresponding face of @xmath61 .
faces of dimension 2 : let @xmath62 be any face of dimension 2 of @xmath61 .
we will prove that @xmath90 . since @xmath41 is a right - angled polytope , the two dimensional polyhedron @xmath63 of @xmath41 has at least 5 edges .
if @xmath62 does not contain any cusps , then @xmath91 , and the inequality is true .
if @xmath62 has cusps , it means some edges of @xmath63 are also edges of squares in cubes , those edges will be collapsed to cusps in @xmath62 , those edges of @xmath63 are not edges in cubes will be remained edges in @xmath62 , so the inequality is still true for this case . the difference @xmath92 will be called the _ excess _ of @xmath93 and denoted by @xmath94 .
faces of dimension 3 : for each 2-dimensional face @xmath95 of some 3-dimensional face @xmath62 of @xmath61 we have @xmath96 summing over all @xmath95 and taking into account that each edge of @xmath62 belongs to 2 faces and each cusp belongs to 4 faces , we get @xmath97 on the other hand , eliminating @xmath98 from the euler equation @xmath99 and the obvious equation @xmath100 ( counting the number of vertices in edges in two ways , counting by edges , counting by vertices - since @xmath63 is simple each original vertex belongs to 3 edges ) gives @xmath101 substituting this into , we finally obtain @xmath102 want to prove the inequality @xmath103 here the argument from @xcite can be used because for the 3 dimension case @xmath41 can be thought of hyperbolic polytope , this follows from steinitz theorem ( theorem 4.1 in @xcite ) .
it follows from and that @xmath104 faces of dimension 4 : let @xmath95 be some 3 dimensional face of 4-dimensional face @xmath62 of @xmath61 .
there are @xmath105 3-dimensional faces adjacent to @xmath95 and , for each cusp of @xmath95 , there is an extra 3-dimensional face having only this cusp in common with @xmath95 . together with @xmath95 ,
this gives at least @xmath106 3-dimensional faces of @xmath62 .
so implies @xmath107 we need a more subtle inequality @xmath108 to prove it , take again any hyperface @xmath95 of @xmath62 .
there are at least @xmath109 hyperfaces meeting @xmath95 and at least @xmath110 cusps , so @xmath111 . if @xmath112 , then follows .
by we have @xmath113 .
consider two cases .
let @xmath114 .
then implies that all but one 2-faces of @xmath95 have zero excess .
let @xmath115 be a 2-face of @xmath95 with zero excess , i.e. @xmath116 .
since @xmath117 , we have @xmath118 let @xmath119 be the hyperface of @xmath61 adjacent to @xmath95 along @xmath115 .
by we have @xmath120 . comparing this with , we have @xmath121 we see that @xmath119 must have either a 2-face @xmath122 not intersecting @xmath95 , or a cusp beyond @xmath95 ( 2-face of @xmath119 if intersects @xmath95 then it has common edge ( so @xmath123 ) , 1 is contributed to @xmath115 , @xmath124 is contributed to the number of 2-faces of @xmath119 which are parallel to @xmath115 ) . in the first case
the hyperface adjacent to @xmath119 along @xmath122 does not intersect @xmath95 by proposition [ first ] , [ second ] .
so in both case holds .
let @xmath125 .
then implies that all 2-faces of @xmath95 have zero excess .
let @xmath115 be any of them .
then @xmath115 is a triangle with two cusps , or a quadrilateral with one cusp , or else a pentagon without cusps .
if @xmath115 is not a triangle , then @xmath126 consider the hyperface @xmath119 of @xmath61 adjacent to @xmath95 along @xmath115 .
then implies that @xmath119 has at least two 2-faces not intersecting @xmath95 or cusps beyond @xmath95 , whence again follows .
let finally all 2-faces of @xmath95 be triangles with two cusps .
take any parallel 2-faces @xmath122 and @xmath127 of @xmath95 , and let @xmath119 and @xmath128 be the hyperfaces of @xmath61 adjacent to @xmath95 along @xmath122 and @xmath127 respectively . by the above each of them must have either a 2-face not intersecting @xmath95 or a cusp beyond @xmath129 . if these
are two 2-faces , then the hyperfaces of @xmath61 adjacent to @xmath119 and @xmath128 along them , can not coincide by proposition [ third ] .
if these are two cusps , then they can not coincide as @xmath119 and @xmath128 are parallel at a cusp of @xmath95 .
so in all the cases holds .
faces of dimension 5 : take any hyperface @xmath95 of 5-dimensional face @xmath62 of @xmath61
. there are @xmath130 hyperfaces adjacent to @xmath95 and , for each cusp of @xmath95 , there is an extra hyperface having only this cusp in common with @xmath95 . together with @xmath95 ,
this gives at least @xmath131 hyperfaces .
so implies @xmath132 using the nikulin - khovanskii inequality ( theorems [ theorem : nikulin ] , [ theorem : khovanskii ] ) gives for the average number @xmath133 of 4-faces of a 5-face of @xmath61 @xmath134 on the other hand , it follows from that @xmath135 . in both case
this means that @xmath136 .
in the following sections we will prove theorem [ theorem : general_case ] using the argument in prokhorov s paper @xcite . applying the cutting process in section 3 , we can assume that the nerve @xmath17 of the coxeter group @xmath2 has all nerves of ( maximal ) euclidean subgroups of codimension one on its boundary .
let recall the notations we use in this section .
let @xmath2 be a coxeter group and @xmath17 is its nerve which has the type as described above , @xmath61 is the fundamental domain of @xmath2 action and it is the dual of @xmath17 .
@xmath41 is gotten from @xmath41 by collapsing all the @xmath40-faces which are corresponding to the cone points one the nerves of euclidean subgroups of codimension 1 to points .
all dihedral angles of @xmath41 are of the form @xmath137 , @xmath138 .
there are two types of vertices in @xmath41 : _ finite vertices _ and _ infinite vertices_. finite vertices are actual vertices in @xmath41 ( and @xmath61 ) , infinite vertices are the ones which corresponding with those euclidean subgroups .
the link of a finite vertex is a simplex ( this is the nerve of a finite subgroup and is stabilizer of the vertex ) , the link of infinite vertex is a product of simplexes ( and it is the dual of the nerve of euclidean subgroup , which is a joint of simplices ) .
since the links of vertices of @xmath41 are either simplices or products of simplices , @xmath41 is almost simple .
a 3 dimensional face @xmath63 of @xmath41 is called _ bad _ if it has the type of a triangular bi - pyramid ( see figure 1 ) . in this polyhedron @xmath63 , vertices @xmath139 and @xmath80 are at infinity because their links are squares which are products of simplices .
these squares are the nerves of euclidean subgroups so they are euclidean squares .
thus , dihedral angles of @xmath63 at vertices @xmath140 are equal to @xmath141 .
then , the links of vertices @xmath77 and @xmath142 are triangles with angles equal to @xmath141 , this means they are spherical triangles and so vertices @xmath77 and @xmath142 are finite vertices .
3-faces of @xmath41 different from the bad ones are called _
good_. the following lemmas are from @xcite .
although they are stated for polyhedron of finite volume for a reflection group in hyperbolic space , their proof uses only combinatorial properties of those polyhedra , which are satisfied in our case .
[ lemma : badgood3face ] suppose that @xmath143 is the number of all 3-faces of @xmath41 , @xmath144 is the number of bad 3-faces , and @xmath145 is the number of good 3-faces of @xmath41 . then for @xmath146 @xmath147 the following definition is needed for the next lemma : a dihedral angle @xmath148 of a 3 dimensional face @xmath63 of @xmath41 is defined to be a pair consisting of a 3-face @xmath63 and a set of two 2-faces @xmath149 and @xmath150 of @xmath63 which intersect in an edge .
[ lemma : inequality ] let @xmath41 be an n - dimensional almost simple polyhedron , and let @xmath151 be a positive number .
suppose that dihedral angles ( edges ) of 3-faces of @xmath41 can be assigned nonnegative weighs so that : 1 .
[ condition1 ] the sum @xmath152 of weights over 3-faces at every edge @xmath153 does not exceed c(n-1 ) , and 2 .
[ condition2 ] the sum @xmath154 of weights over all edges of any good 3-face @xmath155 is not less than @xmath156 , where k is the number of 2-faces of @xmath155 . then @xmath157 . to prove theorem [ theorem : general_case ]
we will use lemma [ lemma : inequality ] . in this section
, we will assign dihedral angles of 3-faces of @xmath41 with weights so that the conditions of lemma [ lemma : inequality ] hold for @xmath158 . by lemma [ lemma : inequality ] : @xmath159 , which is the statement of theorem [ theorem : general_case ] .
+ with every unordered set of 2-faces @xmath160 of a fixed 3-face @xmath63 of a fundamental polyhedron @xmath41 we associate a sub - diagram @xmath161 of the diagram @xmath8 of @xmath41 which is generated by the vertices corresponding to @xmath40-faces of @xmath41 which either contain @xmath63 or intersect @xmath63 in a 2-face from k. the vertices of @xmath161 corresponding to the faces of the second type are called _
marked_. a diagram of a dihedral angle @xmath148 of a 3-face @xmath63 of @xmath162 is defined to be the sub - diagram @xmath161 , where @xmath163 .
this is an elliptic sub - diagram of rank @xmath57 , since it is diagram of the stabilizer of the common edge of @xmath149 and @xmath150 .
to each dihedral angle @xmath148 of @xmath41 assign weight 1 if the distance between marked vertices in its diagram is at most 7 , 1/3 if the inequality @xmath164 holds for marked vertices @xmath32 and @xmath33 in @xmath148 , and 0 otherwise .
the following proposition is a corollary from the classification of euclidean coxeter diagrams ( see , for example , @xcite ) .
[ prop : euclideancoxeter ] for a connected euclidean coxeter diagram @xmath46 of rank @xmath4 the number of unordered pairs of vertices with distance at most @xmath151 between them does not exceed @xmath165 .
a _ complete diagram _ of a 3-face @xmath63 is defined to be the diagram @xmath161 , where @xmath166 is the set of all 2-faces of @xmath63 . among all possible sub - diagrams @xmath167 , where @xmath168 , consider the following types * _ type 1 : _
@xmath169 contains four 2-faces .
every indefinite sub - diagram of @xmath170 contains all marked vertices ( since any sub - diagram of @xmath170 containing at most 3 marked vertices lies in stabilizer of one of the vertices , thus it is finite if the vertex is finite or it is euclidean if the vertex is infinite ) .
a diagram of a dihedral angle of @xmath63 is obtained from @xmath170 excluded two marked vertices . *
_ type 2 : _
@xmath169 contains three 2-faces , any two of which intersect at an edge .
the diagram @xmath170 is indefinite ( since it does not fix any vertex , it can not be finite or euclidean ) .
a diagram of a dihedral angle of @xmath63 is obtained from @xmath170 excluded one marked vertex . * _ type 3 : _
@xmath169 contains three 2-faces , two of which intersect at a vertex at infinite , the other two pair of 2-faces intersect at edges .
the diagram @xmath170 is indefinite ( since it does not fix any vertex , it can not be finite or euclidean ) .
a diagram of a dihedral angle of @xmath63 is obtained from @xmath170 excluded one from two marked vertices corresponding to two 2-face intersecting at infinity .
if the third marked vertex is excluded from @xmath170 , we obtain a diagram of a dihedral angle of @xmath63 . * _ type 4 : _
@xmath169 contains four 2-faces which can be divided into two groups @xmath171 and @xmath172 .
each of @xmath171 and @xmath172 contains two opposite 2-faces .
it is clear that @xmath173 and @xmath174 are indefinite sub - diagrams , and any indefinite sub - diagram of @xmath170 must contain both marked vertices from some pair . a diagram of a dihedral angle of @xmath63 is obtained from @xmath170 excluded two marked vertices from different pairs @xmath171 and @xmath172 .
the sum @xmath175 of weights on a face @xmath63 is defined to be the sum of weights of all sub - digrams of dihedral angles of @xmath63 contained in @xmath161 .
similarly , we can define the sum of weights on a sub - diagram @xmath167 .
[ lemma : weightof3face ] the following statements are true : 1 .
[ type1 ] if a complete diagram of a face @xmath63 contains a sub - diagram @xmath170 of type 1 , then @xmath176 .
[ type2 ] if a complete diagram of a face @xmath63 contains a sub - diagram @xmath170 of type 2 , then @xmath177 .
[ type3 ] the sum of weights on sub - diagram of type 3 is not less than 1 .
[ type4]the sum of weights on sub - diagram of type 4 is not less than 1/3 . for [ type1 ] , [ type2 ] , [ type3 ] :
any indefinite sub - diagram of @xmath170 contains all marked vertices .
consider the minimum indefinite sub - diagram @xmath178 .
if a non - marked vertex is excluded from @xmath23 , we can not get an infinite diagram since @xmath23 is minimal .
thus we get a finite or euclidean sub - diagram .
obviously , if any marked vertex is excluded from @xmath23 , one obtains a finite or an euclidean sub - diagram .
that means @xmath23 is a quasi - lanner diagram . and
since @xmath23 is minimal , it is connected .
notice that since @xmath23 is connected graph , there are at least three pair of marked vertices in @xmath170 of type 1 which are joined in @xmath23 by a path containing no other marked vertices from @xmath170 .
thus , by the classification of quasi - lanner diagram ( see @xcite ) for all such pairs @xmath179 we have @xmath180 . in case [ type2 ]
there exist two pairs satisfying this property , and in case [ type3 ] at least one such pair exists .
for case [ type4 ] , by similar argument as above we have each pair @xmath181 of marked vertices is contained in a quasi - lanner diagram which has no marked vertices from another pair .
let @xmath23 and @xmath182 be such quasi - lanner diagrams . if @xmath23 and @xmath182 are orthogonal then four marked vertices from type 4 would form subgroup @xmath183 which is euclidean .
this means the 4 2-faces of @xmath63 in type 4 meet at vertex at infinity .
but this is not the case we consider in type 4 .
thus , @xmath23 and @xmath182 are joined by an edge or they have common vertex .
therefore there exist marked vertices @xmath32 and @xmath33 from different pairs which can be joined by a path not containing other mark vertices . by the classification of quasi - lanner diagram
we see that @xmath184 . from the definition of weights of dihedral angle the sum of weights on sub - diagram of type 4
is not less than 1/3 .
we finish the proof of the theorem [ theorem : general_case ] by proving the following statement : [ prop : checkcondition ] the fundamental domain @xmath41 satisfies the conditions of lemma [ lemma : inequality ] for @xmath158 . we need to check conditions [ condition1 ] and [ condition2 ] from lemma [ lemma : inequality ] .
the diagram of an edge @xmath153 of @xmath41 is elliptic of order @xmath57 ( an edge in @xmath41 is an intersection of @xmath57 codimention 1 faces of @xmath41 ) .
the diagram of a dihedral angle @xmath185 of a 3-face @xmath63 such that @xmath186 is completely determined by two marked vertices in the diagram of @xmath153 . by proposition [ prop : euclideancoxeter ]
the number of pairs of vertices in the diagram of the edge which are at distance at most 15 ( respectively , at most 7 ) can not exceed @xmath187 ( respectively , @xmath188 ) .
we see that the sum of wights at the edge is at most @xmath189 , and so condition [ condition1 ] of lemma [ lemma : inequality ] holds .
it is sufficient to verify condition [ condition2 ] of lemma [ lemma : inequality ] for 3-faces which contain no more than six 2-faces .
it is not difficult to find all of them ( it was done in prokhorov s paper @xcite ) .
figure 3 includes complete list , except a cone over a pentagon , which is not a direct product of simplexes and so can not be a 3-face for @xmath41 .
we will check condition [ condition2 ] of lemma [ lemma : inequality ] for each case listed above . *
the complete diagram of this face is a sub - diagram of type 1 , and @xmath176 by lemma [ lemma : weightof3face ] . *
the sub - diagram @xmath190 is of type 2 , and @xmath191 .
* two sub - diagrams of type 3 are @xmath192 and @xmath177 .
* there are three sub - diagrams of type 4 , each consists of 4 consecutive 2-faces of the cube , and @xmath193 . * for this case there is a sub - diagram of type 2 corresponding to 2-faces which intersect the shaded triangle , and @xmath177 .
* this case is similar to e. * the sub - diagram @xmath194 is of type 3 , and @xmath195 . *
the sub - diagram @xmath196 is of type 3 , and @xmath195 .
thus , proposition [ prop : checkcondition ] and so theorem [ theorem : general_case ] is proved .
so far we consider relatively hyperbolic coxeter groups of type @xmath0 with maximal flats of codimension 1 .
naturally we can ask if a coxeter group @xmath2 of type @xmath0 is relatively hyperbolic relative to its maximal affine subgroups , then must it have a special subgroup which is affine of codimension 1 ?
the answer for this question is no .
for @xmath197 there is an example in @xcite of 4-dimensional right - angled coxeter group which is relatively hyperbolic and has maximal abelian subgroup of dimension 2 . here in this section
we present an example of relatively hyperbolic coxeter group with fundamental domain a 12-cube with maximal flats of dimension 6.the idea is that from the diagram for the right - angled cube consisting of @xmath198 vertical edges each labeled @xmath6 we add edges between some vertices so that the link @xmath17 of new coxeter group is the same type as for the right - angled cube .
in the following example there are 8 vertical edges labeled @xmath6 dividing into 4 groups @xmath199 , each of @xmath200 and @xmath201contains 1 edge , @xmath202 and @xmath203 contain 3 edges .
the upper vertex of @xmath200 is connected to all three upper vertices of @xmath202 , the lower vertice of @xmath200 is connected to all three lower vertices of @xmath203 .
the upper vertice of @xmath201 is connected to all three upper vertices of @xmath203 , and the lower vertice of @xmath201 is connected to all three lower vertices of @xmath202 .
using criteria from @xcite we can check that this group is relatively hyperbolic coxeter group with a maximal flat of codimension 6 and a maximal flat of codimension 2 ( sub - cubes generated by @xmath204 or @xmath205 ) it is not known whether the above group acts on hyperbolic space or not . even in the non - relative case there are not many examples of ( word ) hyperbolic coxeter groups which are not actual hyperbolic groups ( i.e do not act on hyperbolic space ) .
one class of examples was introduced recently by r. greene in @xcite .
his technique can be applied to produce similar example for relatively hyperbolic coxeter groups .
here is a rough idea how to produce such an example .
let @xmath35 be an octahedron , @xmath206 be the product of @xmath35 crossed an interval . if we cone off the two copies @xmath207 of @xmath35 in @xmath208 we get a 3-sphere .
we will triangulate this sphere so that one octahedron , @xmath209 , is remained ( this will be the nerve of a codimension 1 flat ) and there is no other empty square except for the ones in @xmath209 . for
the cone on @xmath210 , apply przytycki - swiatkowski @xcite subdivision process .
the triangles in @xmath210 are subdivided and there are new vertices in @xmath210 , we connect the vertices in @xmath209 with appropriate new vertices in @xmath210 to have a triangulation of 3-sphere .
you can check that this triangulation does nt have any empty squares rather than the ones in @xmath209 .
and so this is the nerve of a relatively hyperbolic right - angled coxeter group of type @xmath0 with one codimension 1 flat .
the fundamental domain of this coxeter group is not embedded into 4 dimensional hyperbolic space by greene s argument @xcite .
99 n. bourbaki , _ groupes et algebres de lie , _ chaps . 4 - 6 , actualites , sci .
1337 , hermenn , paris 1968 .
caprace , _ buildings with isolated subspaces and relatively hyperbolic coxeter groups_. innov .
incidence geom . 10
( 2009 ) , pp/ 15 - 31 .
m. davis , _ the geometry and topology of coxeter groups _ ,
london mathematical society monographs , 2008 .
m. davis , t. januszkiewicz , j. lafont , _
4-dimensional cat(0 ) manifolds with no riemannian smoothings _ , duke math .
j , 161 ( 2012 ) .
r. greene , _ hyperbolicity and word hyperbolicity for rarg in 4 dimension _
, in progress
. t. januszkiewicz , j. swiatkowski , _ hyperbolic coxeter groups of large dimension _
, comment .
78 ( 2003 ) 555 - 583 .
a. f. khovanskii , _ hyperplane sections of polyhedra , toric varieties , and discrete groups in lobachevsky space _ , funktsional .
i prilozhen . 20 ( 1986 ) ,
no . 1 , 50 - 61 ; english trnasl . in functional anal .
appl 20(1986 ) .
v. v. nikulin , _ on the classification of arithmetic groups generated by reflections in lobachevsky spaces , _ izv akad .
nauk sssr ser .
45 ( 1981 ) , 113 - 143 ; english transl . in math ussr izv . 18 ( 1982 ) .
potyagailo , vinberg _ right angled coxeter groups in hyperbolic spaces _ , commentarii mathematici helvetici , vol 80 , 2005 . m. n. prokhorov , _ the absence of discrete reflection groups with noncompact fundamental polyhedron of finite volume in lobachevsky space of large dimension , _ math .
ussr izvestiya vol .
28(1987 ) , no .
j. ratcliffe , _ foundations of hyperbolic manifolds _ , graduate texts in mathematics , second edition , springer .
e. vinberg , _ the absence of crystallographic groups of reflections in lobachevsky spaces of large dimension , _ trudy moskov .
47 ( 1984 ) , 68 - 102 ; english transl . in trans .
moscow math .
1985 ( 47 ) . | we study relatively hyperbolic coxeter groups of type @xmath0 with maximal euclidean coxeter subgroups of codimension 1 .
our main result in this paper is that the dimension of these groups is bounded above . |
cycle covers of graphs form a prominent topic in graph theory which is closely related to several deep and open problems .
a _ cycle _ in a graph is a subgraph with all degrees even .
cycle cover _ is a collection of cycles such that each edge is contained in at least one of the cycles ; we say that each edge is _
covered_. the cycle double cover conjecture of seymour @xcite and szekeres @xcite asserts that every bridgeless graph @xmath4 has a collection of cycles containing each edge of @xmath4 exactly twice which is called a _
cycle double cover_. in fact , it was conjectured by celmins @xcite and preissmann @xcite that every graph has a cycle cover comprised of five cycles .
the cycle double cover conjecture is related to other deep conjectures in graph theory . in particular , it is known to be implied by the shortest cycle cover conjecture of alon and tarsi @xcite which asserts that every bridgeless graph with @xmath0 edges has a cycle cover of total length at most @xmath5 . recall that the _ length _ of a cycle is the number of edges contained in it and the length of a cycle cover is the sum of the lengths of its cycles .
the reduction of the cycle double cover conjecture to the shortest cycle cover conjecture can be found in the paper of jamshy and tarsi @xcite .
the best known general result on short cycle covers is due to alon and tarsi @xcite and bermond , jackson and jaeger @xcite : every bridgeless graph with @xmath0 edges has a cycle cover of total length at most @xmath6 .
there are numerous results on short cycle covers for special classes of graphs , e.g. , graphs with no short cycles , highly connected graphs or graphs admitting a nowhere - zero @xmath7-/@xmath8-flow , see e.g. @xcite .
the reader is referred to the monograph of zhang @xcite for further exposition of such results where an entire chapter is devoted to results on the shortest cycle cover conjecture .
the least restrictive of refinements of the general bound of alon and tarsi @xcite and bermond , jackson and jaeger @xcite are the bounds for cubic graphs .
jackson @xcite showed that every cubic bridgeless graph with @xmath0 edges has a cycle cover of total length at most @xmath9 and fan @xcite later showed that every such graph has a cycle cover of total length at most @xmath3 . in this paper , we strengthen this bound in two different ways .
first , we improve it to @xmath2 for _ cubic _ bridgeless graphs .
and second , we show that the bound @xmath10 also holds for @xmath0-edge bridgeless graphs with _ minimum degree three _ , i.e. , we extend the result from @xcite on cubic graphs to all graphs with minimum degree three .
as in @xcite , the cycle covers that we construct consist of at most three cycles . though the improvements of the original bound of @xmath11 on the length of a shortest cycle cover of an @xmath0-edge bridgeless graph can seem to be rather minor , obtaining a bound below @xmath12 for a significant class of graphs might be quite challenging since the bound of @xmath13 is implied by tutte s @xmath8-flow conjecture @xcite .
the paper is organized as follows . in section [ sec - prelim ]
, we introduce auxiliary notions and results needed for our main results .
we demonstrate the introduced notation in section [ sec - intermezzo ] where we reprove the bound @xmath14 of alon and tarsi @xcite and bermond , jackson and jaeger @xcite using our terminology . in sections [ sec - cubic ] and [ sec - mindegree ]
, we prove our bounds for cubic graphs and graphs with minimum degree three .
in this section , we introduce notation and auxiliary concepts used throughout the paper .
we focus on those terms where confusion could arise and refer the reader to standard graph theory textbooks , e.g. @xcite , for exposition of other notions .
graphs considered in this paper can have loops and multiple ( parallel ) edges . if @xmath15 is a set of edges of a graph @xmath4
, @xmath16 denotes the graph with the same vertex set with the edges of @xmath15 removed .
if @xmath17 , we simply write @xmath18 instead of @xmath19 . for an edge @xmath20 of @xmath4 , @xmath21 is the graph obtained by contracting the edge @xmath20 , i.e. , @xmath21 is the graph with the end - vertices of @xmath20 identified , the edge @xmath20 removed and all the other edges , including new loops and parallel edges , preserved .
note that if @xmath20 is a loop , then @xmath22 .
finally , for a set @xmath15 of edges of a graph @xmath4 , @xmath23 denotes the graph obtained by contracting all edges contained in @xmath15 .
if @xmath4 is a graph and @xmath24 is a vertex of @xmath4 of degree two , then the graph obtained from @xmath4 by _ suppressing _ the vertex @xmath24 is the graph obtained from @xmath4 by contracting one of the edges incident with @xmath24 , i.e. , the graph obtained by replacing the two - edge path with the inner vertex @xmath24 by a single edge .
an _ edge - cut _ in a graph @xmath4 is a set @xmath15 of edges such that the vertices of @xmath4 can be partitioned into two sets @xmath25 and @xmath26 such that @xmath15 contains exactly the edges with one end - vertex in @xmath25 and the other in @xmath26 .
such an edge - cut is also denoted by @xmath27 .
note that edge - cuts need not be minimal sets of edges whose removal increases the number of components of @xmath4 .
an edge forming an edge - cut of size one is called a _
bridge _ and graphs with no edge - cuts of size one are said to be _
bridgeless_. note that we do not require bridgeless graphs to be connected .
the quantity @xmath28 denotes the number of edges with one end - vertex in @xmath29 and the other vertex in @xmath30 ; in particular , if @xmath27 is an edge - cut of @xmath31 , then @xmath32 is its size .
a graph @xmath4 with no edge - cuts of odd size less than @xmath33 is said to be _ @xmath33-odd - connected_. for every set @xmath34 of edges of @xmath4 , edge - cuts in @xmath35 correspond to edge - cuts ( of the same size ) in @xmath4 .
therefore , if @xmath4 has no edge - cuts of size @xmath33 , then @xmath35 also has no edge - cuts of size @xmath33 .
as said before , a _ cycle _ of a graph @xmath4 is a subgraph of @xmath4 with all vertices of even degree .
a _ circuit _ is a connected subgraph with all vertices of degree two and a _
@xmath36-factor _ is a spanning subgraph with all vertices of degree two . in the proof of our main result
, we will need to construct a nowhere - zero @xmath37-flow of a special type . in order to exclude some `` bad '' nowhere - zero flows
, we will first modify the graph under consideration in such a way that some of its edges must get the same flow value .
this goal will be achieved by splitting some of the vertices of the graph .
let @xmath4 be a graph , @xmath24 a vertex of @xmath4 and @xmath38 and @xmath39 some of the neighbors of @xmath24 in @xmath4 .
the graph @xmath40 that is obtained by removing the edges @xmath41 and @xmath42 from @xmath4 and adding a two - edge path between @xmath38 and @xmath39 ( see figure [ fig - splitting ] ) is said to be obtained by _ splitting the vertices @xmath38 and @xmath39 from the vertex @xmath24_. note that if @xmath43 , i.e. , the edge @xmath41 is a loop , the graph @xmath40 is the graph obtained from @xmath4 by removing the loop @xmath41 and subdividing the edge @xmath42 .
similarly , if @xmath44 , @xmath40 is obtained by removing the loop @xmath42 and subdividing the edge @xmath41 . finally ,
if @xmath45 , then the graph @xmath40 is obtained from @xmath4 by removing the loops @xmath41 and @xmath42 and introducing a new vertex joined by two parallel edges to @xmath24 .
classical ( and deep ) results of fleischner @xcite , mader @xcite and lovsz @xcite assert that it is possible to split vertices without creating new small edge - cuts .
let us now formulate one of the corollaries of their results .
[ lm - split - orig ] let @xmath4 be a @xmath8-odd - connected graph .
for every vertex @xmath24 of @xmath4 of degree four , six or more , there exist two neighbors @xmath38 and @xmath39 of the vertex @xmath24 such that the graph @xmath40 is also @xmath8-odd - connected .
zhang @xcite proved a version of lemma [ lm - split - orig ] where only some pairs of vertices are allowed to be split off ( also see @xcite for related results ) .
[ lm - split - cyclic ] let @xmath4 be an @xmath46-odd - connected graph for an odd integer @xmath46 .
for every vertex @xmath24 of @xmath4 with neighbors @xmath47 , @xmath48 , there exist two neighbors @xmath49 and @xmath50 such that the graph @xmath51 is also @xmath46-odd - connected ( indices are modulo @xmath33 )
. however , none of these results is sufficient for our purposes since we need to specify more precisely which pair of the neighbors of @xmath24 should be split from @xmath24 .
this is guaranteed by the lemmas we establish in the rest of this section .
let us remark that lemma [ lm - split - cyclic ] can be obtained as a consequence of our results .
we start with a modification of the well - known fact that `` minimal odd cuts do not cross '' for the situation where small even cuts can exist .
[ lm - uncross ] let @xmath4 be an @xmath46-odd - connected graph ( for some odd integer @xmath46 ) and let @xmath52 and @xmath53 be two cuts of @xmath4 of size @xmath46 .
further , let @xmath54 for @xmath55 .
if the sets @xmath56 are non - empty for all @xmath55 , then there exist integers @xmath57 , @xmath58 and @xmath59 such that @xmath60 and one of the following holds : * @xmath61 , @xmath62 , @xmath63 and @xmath64 , or * @xmath65 , @xmath66 , @xmath67 and @xmath68
. see figure [ fig - uncross ] for an illustration of the two possibilities .
let @xmath69 be the number of edges with exactly one end - vertex in @xmath56 .
observe that @xmath70 @xmath71 in particular , one of the numbers @xmath72 and @xmath73 is even and the other is odd .
assume that @xmath72 is odd .
hence , @xmath74 is also odd . since @xmath4 is @xmath46-odd - connected , both @xmath72 and @xmath74 are at least @xmath46 . if @xmath75 , then @xmath76 the equation ( [ eq - uncross1 ] ) , the inequality ( [ eq - uncross2 ] ) and the inequality @xmath77 imply that @xmath78 .
hence , the inequality ( [ eq - uncross2 ] ) is an equality ; in particular , @xmath79 and @xmath64 .
if @xmath80 , we obtain the same conclusion .
since the sizes of the cuts @xmath52 and @xmath53 are the same , it follows that @xmath81 .
we conclude that the graph @xmath4 and the cuts have the structure as described in the first part of the lemma .
the case that @xmath72 is even ( and thus @xmath73 is odd ) leads to the other configuration described in the statement of the lemma .
next , we use lemma [ lm - uncross ] to characterize graphs where some splittings of neighbors of a given vertex decrease the odd - connectivity .
[ lm - split - gener ] let @xmath4 be an @xmath46-odd - connected graph for an odd integer @xmath82 , @xmath24 a vertex of @xmath4 and @xmath47 some neighbors of @xmath24 .
if every graph @xmath51 , @xmath83 , contains an edge - cut of odd size smaller than @xmath46 , the vertex set @xmath84 can be partitioned into two sets @xmath85 and @xmath86 such that @xmath87 , @xmath88 for @xmath89 and the size of the edge - cut @xmath90 is @xmath46 .
first observe that it is enough to prove the statement of the lemma for simple graphs ; if @xmath4 is not simple , then subdivide each edge of @xmath4 and apply the lemma to the resulting graph in the natural way .
the assumption that @xmath4 is simple allows us to avoid unpleasant technical complications in the proof .
the proof proceeds by induction on @xmath33 .
the base case of the induction is that @xmath91 .
let @xmath90 be an edge - cut of @xmath40 of odd size less than @xmath46 and let @xmath92 be the new degree - two vertex of @xmath40 . by symmetry , we can assume that @xmath87 .
we can also assume that @xmath38 and @xmath92 are in the same set @xmath93 ; otherwise , moving @xmath92 to the set @xmath93 containing @xmath38 either preserves the size of the cut or decreases the size by two . if both @xmath94 and @xmath95 , then @xmath96 as an edge - cut of @xmath4 has the same size as in @xmath40 which contradicts the assumption that @xmath4 is @xmath46-odd - connected . if @xmath94 and @xmath97 , then @xmath96 is also an edge - cut of @xmath4 of the same size as in @xmath40 which is again impossible .
hence , both @xmath38 and @xmath39 must be contained in @xmath86 , and the size of the edge - cut @xmath98 in @xmath4 is larger by two compared to the size of @xmath90 in @xmath40 .
since @xmath4 has no edge - cuts of size @xmath99 , the size of the edge - cut @xmath90 in @xmath40 is @xmath99 and the size of @xmath98 in @xmath4 is @xmath46 .
hence , @xmath85 and @xmath100 form the partition of the vertices as in the statement of the lemma .
we now consider the case that @xmath101 . by the induction assumption
, @xmath4 contains a cut @xmath102 of size @xmath46 such that @xmath103 and @xmath104 .
similarly , there is a cut @xmath105 of size @xmath46 such that @xmath106 and @xmath107 .
let @xmath108 for @xmath55 ( see figure [ fig - split - gener ] ) .
apply lemma [ lm - uncross ] for the graph @xmath4 with @xmath109 and @xmath110 . since @xmath111 , the first case described in lemma [ lm - uncross ] applies and the size of the cut @xmath112 is @xmath46 .
hence , the cut @xmath112 is a cut of size @xmath46 in @xmath4 .
we now present a series of lemmas that we need later in the paper .
all these lemmas are simple corollaries of lemma [ lm - split - gener ] .
[ lm - split4 ] let @xmath4 be a @xmath8-odd - connected graph , and let @xmath24 be a vertex of degree four and @xmath38 , @xmath39 , @xmath113 and @xmath114 its four neighbors .
then , the graph @xmath40 or the graph @xmath115 is also @xmath8-odd - connected graph . observe that the graph @xmath40 is @xmath8-odd - connected if and only if the graph @xmath116 is @xmath8-odd - connected .
lemma [ lm - split - gener ] applied for @xmath117 , the vertex @xmath24 , @xmath118 and the vertices @xmath38 , @xmath39 , @xmath113 and @xmath114 yields that the vertices of @xmath4 can be partitioned into two sets @xmath85 and @xmath86 such that @xmath87 , @xmath119 and @xmath120 . hence , @xmath121 this contradicts our assumption that @xmath4 has no edge - cuts of size one .
[ lm - split6 ] let @xmath4 be a @xmath8-odd - connected graph , and let @xmath24 be a vertex of degree six and @xmath122 its neighbors .
at least one of the graphs @xmath40 , @xmath115 and @xmath116 is also @xmath8-odd - connected .
lemma [ lm - split - gener ] applied for @xmath117 , the vertex @xmath24 , @xmath118 and @xmath38 , @xmath39 , @xmath113 and @xmath114 yields that the vertices of @xmath4 can be partitioned into two sets @xmath85 and @xmath86 such that @xmath87 , @xmath119 and @xmath120 .
if @xmath86 contains @xmath123 neighbors of @xmath24 ( note that @xmath124 ) , then @xmath125 which is equal to @xmath126 or @xmath127 contradicting the fact that @xmath4 is @xmath8-odd - connected .
[ lm - split8 ] let @xmath4 be a @xmath8-odd - connected graph , and let @xmath24 be a vertex of degree @xmath128 and @xmath129 its neighbors .
at least one of the graphs @xmath51 , @xmath130 , is also @xmath8-odd - connected . since there is no partition of the vertices of @xmath4 into two parts @xmath85 and @xmath86 such that @xmath87 , @xmath88 for @xmath131 , and @xmath120 , lemma [ lm - split - gener ] applied for @xmath117 , the vertex @xmath24 , @xmath132 and the vertices @xmath49 , @xmath131 yields the statement of the lemma .
we will also need the following corollary .
[ lm - split8spc ] let @xmath4 be a @xmath8-odd - connected graph , and let @xmath24 be a vertex of degree eight and @xmath133 its neighbors .
suppose that @xmath40 is @xmath8-odd - connected .
at least one of the following graphs is also @xmath8-odd - connected : @xmath134 , @xmath135 , @xmath136 and @xmath137 .
the degree of the vertex @xmath24 in @xmath40 is six . by lemma [ lm - split6 ] ,
at least one of the graphs @xmath134 , @xmath138 and @xmath135 is @xmath8-odd - connected . if @xmath138 is @xmath8-odd - connected , we apply lemma [ lm - split4 ] for the vertex @xmath24 and its neighbors @xmath139 , @xmath114 , @xmath140 and @xmath141 ( in this order ) .
hence , the graph @xmath136 ( which is isomorphic to @xmath142 ) or the graph @xmath137 is @xmath8-odd - connected .
we need one more vertex operation in our arguments
vertex expansion .
if @xmath4 is a graph , @xmath24 a vertex of @xmath4 and @xmath85 a subset of its neighbors , then the graph @xmath143 is the graph obtained from @xmath4 by removing the vertex @xmath24 and introducing two new vertices @xmath38 and @xmath39 , joining @xmath38 to the vertices of @xmath85 , @xmath39 to the neighbors of @xmath24 not contained in @xmath85 , and adding an edge @xmath144 .
we say that @xmath143 is obtained by _ expanding the vertex @xmath24 with respect to the set @xmath85_. see figure [ fig - expand ] for an example .
let us remark that this operation will be applied only to vertices @xmath24 incident with no parallel edges .
the following auxiliary lemma directly follows from results of fleischner @xcite : [ lm - expand ] let @xmath4 be a bridgeless graph and @xmath24 a vertex of degree four in @xmath4 incident with no parallel edges .
further , let @xmath38 , @xmath39 , @xmath113 and @xmath114 be the four neighbors of @xmath24 .
the graph @xmath145 or the graph @xmath146 is also bridgeless .
we start with recalling a classical result of jaeger on the existence of nowhere - zero @xmath37-flows .
[ thm-4flow ] if @xmath4 is a graph that contains no edge - cuts of size one or three , then @xmath4 has a nowhere - zero @xmath37-flow . in our arguments , we will need to construct @xmath37-flows that have some additional properties .
some of these properties will be guaranteed by vertex splittings introduced earlier . in this section
, we establish an auxiliary lemma that guarantees the existence of a nowhere - zero @xmath37-flow which we were not able to prove with vertex splittings only .
we use the following notation in lemma [ lm - flow ] : if @xmath147 is a flow on a graph @xmath4 and @xmath15 is a subset of the edges of @xmath31 , then @xmath148 is the set of the values of @xmath147 on @xmath15 .
[ lm - flow ] let @xmath4 be a bridgeless graph admitting a nowhere - zero @xmath37-flow .
assume that * for every vertex @xmath24 of degree five , there are given two multisets @xmath149 and @xmath150 of three edges incident with @xmath24 such that @xmath151 ( a loop incident with @xmath24 can appear twice in the same set ) , and * for every vertex @xmath24 of degree six , the incident edges are partitioned into three multisets @xmath149 , @xmath150 and @xmath152 of size two each ( every loop incident with @xmath24 appears twice in the sets @xmath149 , @xmath150 and @xmath152 , possibly in the same set ) .
the graph @xmath4 has a nowhere - zero @xmath37-flow @xmath147 such that * for every vertex @xmath24 of degree five , @xmath153 and @xmath154 , and * for every vertex @xmath24 of degree six , @xmath155 , i.e. , the edges incident with @xmath24 have all the three possible flow values , or @xmath156 , or @xmath157 , or @xmath158 . for simplicity
, we refer to edges with the flow value @xmath159 red , @xmath160 green and @xmath161 blue .
note that each vertex of odd degree is incident with odd numbers of red , green and blue edges and each vertex of even degree is incident with even numbers of red , green and blue edges ( counting loops twice ) .
we say that a vertex @xmath24 of degree five is _ bad _ if @xmath156 or @xmath157 , and it is _ good _ , otherwise .
similarly , a vertex @xmath24 of degree six is _ bad _ if @xmath147 has only two possible flow values at @xmath24 , @xmath162 , @xmath163 and @xmath164 ; otherwise , @xmath24 is good .
choose a @xmath37-flow @xmath147 of @xmath31 with the least number of bad vertices .
if there are no bad vertices , then there is nothing to prove .
assume that there is a bad vertex @xmath24 .
let us first analyze the case that the degree of @xmath24 is five .
let @xmath165 be the edges incident with @xmath24 . by symmetry
, we can assume that @xmath166 , @xmath167 , the edges @xmath168 , @xmath169 and @xmath170 are red , the edge @xmath171 is green and the edge @xmath172 is blue ( see figure [ fig - flow - bad ] ) .
we now define a closed trail @xmath173 in @xmath31 formed by red and blue edges .
the first edge of @xmath173 is @xmath168 .
let @xmath174 be the last edge of @xmath173 defined so far . if @xmath175 , then @xmath176 is one of the edges @xmath169 , @xmath170 and @xmath172 and the definition of @xmath173 is finished .
assume that @xmath177 .
if @xmath178 is not a vertex of degree five or six or @xmath178 is a bad vertex , add to the trail @xmath173 any red or blue edge incident with @xmath178 that is not already contained in @xmath173 .
if @xmath178 is a good vertex of degree five , let @xmath179 be the edges incident with @xmath178 , @xmath180 and @xmath181 .
if @xmath178 is incident with a single red and a single blue edge , leave @xmath178 through the other edge that is red or blue . otherwise , there are three red edges and one blue edge or vice versa .
the next edge @xmath182 of the trail @xmath173 is determined as follows ( note that the role of red and blue can be swapped ) : [ cols="^,^,^,^,^,^,^",options="header " , ] clearly , we have obtained a cycle cover of @xmath4 with three cycles .
the length of the cycles is increased at most by @xmath183 and thus it is at most @xmath184 [ lm - parallel - suppress ] let @xmath4 be an @xmath0-edge bridgeless graph with vertices @xmath38 and @xmath39 joined by @xmath185 parallel edges .
if the degrees of @xmath38 and @xmath39 are @xmath183 and the graph @xmath186 obtained by contracting all the edges between @xmath38 and @xmath39 and suppressing the resulting vertex of degree two has a cycle cover with three cycles of length at most @xmath187 , then @xmath4 has a cycle cover with three cycles of length at most @xmath10 .
let @xmath188 , @xmath189 and @xmath190 be the cycles of total length at most @xmath187 covering the edges of @xmath186 , let @xmath191 be the @xmath33 parallel edges between the vertices @xmath38 and @xmath39 , and let @xmath192 be the other neighbor of @xmath49 , @xmath193 .
add the edges incident with @xmath194 and @xmath195 to those cycles @xmath188 , @xmath189 and @xmath190 that contain the edge @xmath196 and then proceed as in the proof of lemma [ lm - parallel - inner ] .
the length of the cycles is increased by at most @xmath197 and thus it is at most @xmath198 where the last inequality holds unless @xmath199 .
if @xmath199 and the edge @xmath196 is contained in at most two of the cycles , the length is increased by at most @xmath200 .
if @xmath199 and the edge @xmath196 is contained in three of the cycles , each of the parallel edges is added to exactly one of the cycles and thus the length is increased by at most @xmath201 . in both cases ,
the length of the new cycle cover can be estimated as follows : @xmath202 we are now ready to prove our bound for graphs with minimum degree three .
[ thm - main ] let @xmath4 be a bridgeless graph with @xmath0 edges and with minimum degree three or more .
the graph @xmath4 has a cycle cover of total length at most @xmath10 that is comprised of at most three cycles . by lemmas [ lm - parallel - leaf][lm - parallel - suppress ] , we can assume without loss of generality that if vertices @xmath38 and @xmath39 of @xmath4 are joined by @xmath33 parallel edges , then either @xmath91 and the degrees of both @xmath38 and @xmath39 are equal to @xmath203 , or @xmath199 , the degree of @xmath38 is @xmath199 and the degree of @xmath39 is @xmath204 ( in particular , both @xmath38 and @xmath39 have odd degrees ) .
note that the graphs @xmath186 from the statement of lemmas [ lm - parallel - leaf][lm - parallel - suppress ] are also bridgeless graphs with minimum degree three and have fewer edges than @xmath4 which implies that the reduction process described in lemmas [ lm - parallel - leaf][lm - parallel - suppress ] eventually finishes . as the first step , we modify the graph @xmath4 into bridgeless graphs @xmath205 eventually obtaining a bridgeless graph @xmath186 with vertices of degree two , three and four .
set @xmath206 .
if @xmath207 has no vertices of degree five or more , let @xmath208 .
if @xmath207 has a vertex @xmath24 of degree five or more , then lemma [ lm - split - cyclic ] yields that there are two neighbors , say @xmath38 and @xmath39 , of @xmath24 such that the graph @xmath209 is also bridgeless .
we set @xmath210 to be the graph @xmath209 .
we continue while the graph @xmath207 has vertices of degree five or more .
clearly , the final graph @xmath186 has the same number of edges as the graph @xmath4 and every cycle of @xmath186 corresponds to a cycle of @xmath4 .
next , each edge of @xmath186 is assigned weight one , each vertex of degree four is expanded to two vertices of degree three as described in lemma [ lm - expand ] and the edge between the two new vertices of degree three is assigned weight zero ( note that the vertex splitting preserves the parity of the degree of the split vertex and thus no vertex of degree four is incident with parallel edges ) .
the resulting graph is denoted by @xmath211 .
note that every cycle @xmath212 of @xmath211 corresponds to a cycle @xmath213 of @xmath4 and the length of @xmath213 in @xmath4 is equal to the sum of the weights of the edges of @xmath212 .
next , the vertices of degree two in @xmath211 are suppressed and each edge @xmath20 is assigned the weight equal to the sum of the weights of edges of the path of @xmath211 corresponding to @xmath20 .
the resulting graph is denoted by @xmath214 .
clearly , @xmath214 is a cubic bridgeless graph .
also note that all the edges of weight zero in @xmath211 are also contained in @xmath214 , no vertex of @xmath214 is incident with two edges of weight zero , and the total weight of the edges of @xmath214 is equal to @xmath0 .
we apply lemma [ lm - rainbow - mindegree ] to the cubic graph @xmath214 .
let @xmath215 be the rainbow @xmath36-factor of @xmath214 and let @xmath216 be the cycle of @xmath211 corresponding to the @xmath36-factor of @xmath215 .
note that @xmath216 is a union of vertex - disjoint circuits .
let @xmath217 , @xmath218 and @xmath219 be the sets of edges of @xmath211 contained in paths corresponding to red , green and blue edges in @xmath214 .
let @xmath220 be the weight of the red edges in @xmath221 , @xmath222 the weight of green edges and @xmath223 the weight of blue edges .
lemma [ lm - rainbow - mindegree ] yields @xmath224 .
we construct two different cycle covers , each comprised of three cycles , and eventually combine the bounds on their lengths to obtain the bound claimed in the statement of the theorem .
* the first cycle cover . *
the first cycle cover that we construct is a cycle cover of the graph @xmath211 ( which yields a cycle cover of @xmath4 of the same length as explained earlier ) .
let @xmath225 be the number of circuits of @xmath216 of weight @xmath46 .
note that @xmath226 can be non - zero since a circuit of weight three need not have length three in @xmath211 .
recall now the notation @xmath227 and @xmath228 used in the proof of theorem [ thm - intermezzo ] for circuits @xmath212 and set @xmath15 of edges that are incident with even number of vertices of @xmath212 .
in addition , @xmath229 denotes the edges of @xmath227 with weight one and @xmath230 denotes such edges of @xmath228 . in the rest of the construction of the first cycle cover , we always assume that @xmath231 .
the sets @xmath188 , @xmath189 and @xmath190 are completed to cycles in a way similar to that used in the proof of theorem [ thm - intermezzo ] . as in the proof of theorem [ thm - cubic ]
, we define subsets @xmath232 , @xmath233 and @xmath234 of the edges of a circuit @xmath212 of @xmath216 .
the subset @xmath232 are just the edges @xmath235 .
the subset @xmath233 are either the edges of @xmath236 or @xmath237we choose the set with fewer edges with weight one in common with @xmath238 . finally , @xmath239 .
the cycle @xmath188 consists of all the red and green edges , i.e. , the edges contained in @xmath240 , and the edges of @xmath232 for every circuit @xmath212 of @xmath216 .
the cycle @xmath189 consists of the red and blue edges and the edges of @xmath233 . finally , the cycle @xmath190 consists of the green and blue edges and the edges of @xmath234 .
we now estimate the number of the edges of @xmath212 of weight one contained in @xmath188 , @xmath189 and @xmath190 .
let @xmath241 be the edges of weight one contained in the circuit @xmath212 , @xmath242 and @xmath243 for @xmath244 . by the choice of @xmath233 , the number of edges of weight one in @xmath245 is @xmath246 .
consequently , the number of edges of @xmath212 of weight one contained in the cycles @xmath188 , @xmath189 and @xmath190 is : @xmath247 , the number of edges contained in the set @xmath248 is at most @xmath249 . by the choice of @xmath233 , @xmath250 .
consequently , it holds that @xmath251 and eventually conclude that the sets @xmath188 , @xmath189 and @xmath190 contain at most @xmath252 edges of the circuit @xmath212 with weight one .
if @xmath253 , the estimate given in ( [ eq - mindegree-1 ] ) can be further refined . let @xmath213 be the circuit of @xmath254 corresponding to @xmath212 . clearly , @xmath213 is a circuit of length four .
color a vertex @xmath24 of the circuit @xmath213 red : : if @xmath24 has degree three and is incident with a red edge , or @xmath24 has degree four and is incident with green and blue edges , green : : if @xmath24 has degree three and is incident with a green edge , or @xmath24 has degree four and is incident with red and blue edges , blue : : if @xmath24 has degree three and is incident with a blue edge , or @xmath24 has degree four and is incident with red and green edges , and white : : otherwise .
observe that either @xmath213 contains a white vertex or it contains an even number of red vertices , an even number of green vertices and an even number of blue vertices .
if @xmath213 contains a white vertex , it is easy to verify that @xmath255 for a suitable permutation of red , green and blue colors ; the inequality ( [ eq - mindegree - pat ] ) also holds if @xmath213 contains two adjacent vertices of the same color ( see figure [ fig - mindegree - pat ] ) . next , we show that every such circuit contains a white vertex or two adjacent vertices with the same color ( and thus ( [ eq - mindegree - pat ] ) always holds ) .
if the circuit of @xmath215 corresponding to @xmath212 in @xmath214 contains an edge of weight two or more , then @xmath212 contains a white vertex and the estimate ( [ eq - mindegree - pat ] ) holds .
otherwise , all vertices of @xmath212 have degree three in @xmath211 and thus the circuit @xmath212 is also contained in @xmath215 .
since the edges of @xmath212 have weight zero and one only , the pattern of @xmath212 is one of the patterns listed in lemma [ lm - rainbow - mindegree ] . a close inspection of possible patterns of @xmath213 yields that the cycle @xmath213 contains a white vertex or it contains two adjacent vertices with the same color .
we conclude that the estimate ( [ eq - mindegree - pat ] ) applies . hence , if @xmath253 , the estimate ( [ eq - mindegree-1 ] ) can be improved to @xmath256 .
we now estimate the length of the cycle cover of @xmath211 formed by the cycles @xmath188 , @xmath189 and @xmath190 .
since each red , green and blue edge is covered by exactly two of the cycles , we conclude that : @xmath257 note that we have used the fact that the sum @xmath258 is equal to the number of the edges of @xmath4 . * the second cycle cover . *
the second cycle cover is constructed in an auxiliary graph @xmath259 which we now describe .
every vertex @xmath24 of @xmath4 is eventually split to a vertex of degree three or four in @xmath186 .
the vertex of degree four is then expanded .
let @xmath260 be the vertex of degree three obtained from @xmath24 or one of the two vertices obtained by the expansion of the vertex of degree four obtained from @xmath24 . by the construction of @xmath216 , each @xmath260 is contained in a circuit of @xmath216 .
the graph @xmath259 is constructed from the graph @xmath211 as follows : every vertex of @xmath211 of degree two not contained in @xmath216 that is obtained by splitting from a vertex @xmath24 is identified with the vertex @xmath260 .
the edges of weight zero contained in a cycle of @xmath216 are then contracted .
let @xmath34 be the cycle of @xmath259 corresponding to the cycle @xmath216 of @xmath211 .
note that @xmath34 is formed by disjoint circuits and it contains @xmath225 circuits of weight / length @xmath46 .
observe that @xmath259 can be obtained from @xmath4 by splitting some of its vertices ( perform exactly those splittings yielding vertices of degree two contained in the circuits of @xmath216 ) and then expanding some vertices .
in particular , every cycle of @xmath259 is also a cycle of @xmath4 .
edges of weight one of @xmath259 one - to - one correspond to edges of weight one of @xmath211 , and edges of weight zero of @xmath259 correspond to edges of weight zero of @xmath211 not contained in @xmath216 .
hence , the weight of a cycle in @xmath259 is the length of the corresponding cycle in @xmath4 .
the edges not contained in @xmath34 are red , green and blue ( as in @xmath211 ) .
each circuit of @xmath34 is incident either with an odd number of red edges , an odd number of green edges and an odd number of blue edges , or with an even number of red edges , an even number of green edges and an even number of blue edges ( chords are counted twice ) .
let @xmath261 . if @xmath262 contains a red circuit ( which can be a loop ) , recolor such a circuit to blue .
similarly , recolor green circuits to blue .
let @xmath263 , @xmath264 and @xmath265 be the resulting sets of red , green and blue edges and @xmath266 , @xmath267 and @xmath58 their weights . clearly , @xmath268 . also note that each circuit of @xmath34 is still incident either with an odd number of red edges , an odd number of green edges and an odd number of blue edges , or with an even number of red edges , an even number of green edges and an even number of blue edges .
since the red edges form an acyclic subgraph of @xmath261 , there are at most @xmath269 red edges and thus the total weight @xmath266 of red edges is at most @xmath270 ( we forget `` @xmath271 '' since it is not important for our further estimates ) .
a symmetric argument yields that @xmath272 .
let us have a closer look at circuits of @xmath34 with weight two .
such circuits correspond to pairs of parallel edges of @xmath259 ( and thus of @xmath4 ) . by our assumption ,
the only parallel edges contained in @xmath4 are pairs of edges between two vertices @xmath38 and @xmath39 of degree three and triples of edges between vertices @xmath38 and @xmath39 of degree three and five . in the former case ,
both @xmath38 and @xmath39 have degree three in @xmath259 .
consequently , each of them is incident with a single colored edge . by the assumption on the edge - coloring ,
the two edges have the same color . in the latter case , the third edge @xmath144 which corresponds to a loop in @xmath273
hence , the other two edges incident with @xmath39 must have the same color , which is red , green or blue . in both cases ,
the vertex of @xmath262 corresponding to the circuit @xmath144 is an isolated vertex in the subgraph of @xmath262 formed by red edges or in the subgraph formed by green edges ( or both ) .
it follows we can improve the estimate on @xmath266 and @xmath267 : @xmath274 we are now ready to construct the cycle cover of the graph @xmath259 .
its construction closely follows the one presented in the proof of theorem [ thm - intermezzo ] .
the cycle cover is formed by three cycles @xmath188 , @xmath189 and @xmath190 . the cycle @xmath188 consists of all red and green edges and the edges of @xmath275 for every circuit @xmath212 of @xmath34 .
@xmath189 consists of all red and green edges and the edges of @xmath276 for every circuit @xmath212 . finally , the cycle @xmath190 contains all red and blue edges and the edges of @xmath277 for every circuit @xmath212 . clearly , the sets @xmath188 , @xmath189 and @xmath190 are cycles of @xmath259 and correspond to cycles of @xmath4 whose length is equal to the the weight of the cycles @xmath188 , @xmath189 and @xmath190 in @xmath259 .
we now estimate the total weight of the cycles @xmath188 , @xmath189 and @xmath190 . each red edge
is covered three times , each green edge twice and each blue edge once . each edge of @xmath34
is contained in either @xmath188 or @xmath189 and for every circuit @xmath212 of @xmath34 at most half of its edges are also contained in @xmath190 .
we conclude that the total length of the cycles @xmath188 , @xmath189 and @xmath190 can be bounded as follows ( note that we apply ( [ eq - mindegree-2 ] ) to estimate the sum @xmath278 and we also use the fact that the number of the edges of @xmath34 is at most @xmath279 by lemma [ lm - rainbow - mindegree ] ) : @xmath280 the last inequality follows from the fact that @xmath281 for @xmath282 . the length of the shortest cycle cover of @xmath4 with three cycles exceeds neither the bound given in ( [ eq - mindegree - sc ] ) nor the bound given in ( [ eq - mindegree - lc ] ) .
hence , the length of the shortest cycle cover of @xmath4 is bounded by any convex combination of the two bounds , in particular , by the following : @xmath283 @xmath284 @xmath285 the proof of theorem [ thm - main ] is now completed .
this research was started while the second and the fifth authors attended pims workshop on cycle double cover conjecture held at university of british columbia , vancouver , canada .
support of pacific institute for mathematical sciences ( pims ) during the workshop is gratefully acknowledged .
99 n. alon , m. tarsi : covering multigraphs by simple circuits , siam j. algebraic discrete methods * 6 * ( 1985 ) , 345350
. j. c. bermond , b. jackson , f. jaeger : shortest coverings of graphs with cycles , j. combin .
theory ser .
b * 35 * ( 1983 ) , 297308 .
u. a. celmins : on cubic graphs that do not have an edge @xmath127-coloring , ph .
d. thesis , university of waterloo , waterloo , canada , 1984 .
r. diestel : graph theory , graduate texts in mathematics vol .
173 , springer - verlag , new york , 2000 .
g. fan : covering weighted graphs by even subgraphs , j. combin .
theory ser .
b * 49 * ( 1990 ) , 137141 .
g. fan : integer flows and cycle covers , j. combin .
theory ser .
b * 54 * ( 1992 ) , 113122 .
g. fan : shortest cycle covers of cubic graphs , j. graph theory * 18 * ( 1994 ) , 131141 .
g. fan , a. raspaud : fulkerson s conjecture and circuit covers , j. comb .
theory ser .
b * 61 * ( 1994 ) , 133138 .
h. fleischner : eine gemeinsame basis fr die theorie der eulerschen graphen und den satz von petersen , monatsh . math .
* 81 * ( 1976 ) , 267278 .
d. r. fulkerson : blocking and antiblocking pairs of polyhedra , math .
programming * 1 * ( 1971 ) , 168194 .
l. a. goddyn : cycle covers of graphs , ph . d. thesis , university of waterloo , waterloo , canada , 1988 .
b. jackson : shortest circuit covers and postman tours of graphs with a nowhere - zero @xmath7-flow , siam j. comput .
* 19 * ( 1990 ) , 659660 .
b. jackson : shortest circuit covers of cubic graphs , j. combin .
theory ser .
b * 60 * ( 1994 ) , 299307 .
f. jaeger : flows and generalized coloring theorems in graphs , j. combin .
theory ser .
b * 26 * ( 1979 ) , 205216 .
u. jamshy , m. tarsi : shortest cycle covers and the cycle double cover conjecture , j. combin .theory ser .
b * 56 * ( 1992 ) , 197204 .
u. jamshy , a. raspaud , m. tarsi : short circuit covers for regular matroids with nowhere - zero @xmath8-flow , j. combin .
theory ser .
b * 43 * ( 1987 ) , 354357 .
t. kaiser , d. kr ' al , s. norine : unions of perfect matchings in cubic graphs , in m. klazar , j. kratochvl , j. matouek , r. thomas , p. valtr ( eds . ) : topics in discrete mathematics , springer , 2006 , 225230 .
d. krl , e. majov , o. pangrc , a. raspaud , j .- s .
sereni , m. koviera : projective , affine , and abelian colorings of cubic graphs , to appear in european journal on combinatorics .
l. lovsz : on some connectivity properties of eulerian graphs , acta math
. acad .
* 28 * ( 1976 ) , 129138 .
e. majov , m. koviera : fano colourings of cubic graphs and the fulkerson conjecture , theoret .
* 349 * ( 2005 ) , 112120 .
w. mader : a reduction method for edge - connectivity in graphs , annals of discrete math .
* 3 * ( 1978 ) , 145164 .
m. preissmann : sur les colorations des ar^ etes des graphes cubiques , th ` ese de doctorat de 3eme , grenoble , 1981 .
a. raspaud : cycle covers of graphs with a nowhere - zero @xmath7-flow , j. graph theory * 15 * ( 1991 ) , 649654 .
d. seymour : sums of circuits , in : grapht theory and related topics ( j. a. bondy and u. s. r. murty , eds . ) , academic press , new york ( 1979 ) , 342355 .
g. szekeres : polyhedral decompositions of cubic graphs , bull .
* 8 * ( 1973 ) , 367387 .
z. szigeti : edge - connectivity augmentation of graphs over symmetric parity families , discrete math .
* 308 * ( 2008 ) , 65276532 . c. q. zhang
: integer flows and cycle covers of graphs , crc , 1997 . c. q. zhang : circular flows of nearly eulerian graphs and vertex - splitting , j. graph theory * 40 * ( 2002 ) , 147161 . | the shortest cycle cover conjecture of alon and tarsi asserts that the edges of every bridgeless graph with @xmath0 edges can be covered by cycles of total length at most @xmath1 .
we show that every cubic bridgeless graph has a cycle cover of total length at most @xmath2 and every bridgeless graph with minimum degree three has a cycle cover of total length at most @xmath3 . |
the history of computing tables of elliptic curves over the rational numbers extends as far back as some of the earliest machine - aided computations in number theory .
the first of these tables appeared in the proceedings of the 1972 conference held in antwerp @xcite .
vast tables of elliptic curves now exist , as computed by cremona @xcite and stein
watkins @xcite .
these tables have been used extensively in research in arithmetic geometry to test and formulate conjectures , and have thereby motivated many important advances . in this article
, we continue this tradition by constructing a table of genus @xmath0 curves over the rational numbers .
we find a total of isomorphism classes of curves with absolute discriminant at most @xmath2 ; for each curve , we compute an array of geometric and arithmetic invariants of the curve and its jacobian , as well as information about rational points and the associated @xmath1-function .
this data significantly extends previously existing tables ( see section [ sec : compl ] for a comparison and discussion of completeness ) , and it is available online at the _ @xmath1-functions and modular forms database ( lmfdb ) _
we anticipate that the list of curves and the auxiliary data we have computed will be useful in ongoing investigations in genus @xmath0 , including the sato
tate conjecture as formulated by fit
rotger sutherland @xcite and the paramodular conjecture of brumer
kramer @xcite .
this article is organized as follows . in section [ sec : notation ] , we set up basic notation and background ; in section [ sec : boxsearch ] , we describe our search procedure ; in sections [ sec : invs][sec : end ] , we outline the computation of various invariants associated to the curve and its jacobian ; and in the appendix , we provide tables exhibiting some interesting aspects of the data .
in this section , we set up basic notation and recall some relevant background material . for a survey of computational aspects of genus 2 curves ,
see the article by poonen @xcite .
let @xmath3 be an abelian variety over a field @xmath4 .
for a field extension @xmath5 , we denote by @xmath6 the abelian variety over @xmath7 obtained from @xmath3 by base extension from @xmath4 to @xmath7 .
we write @xmath8 to denote the ring of endomorphisms of @xmath3 _ defined over @xmath4 _ ; we write @xmath9 for endomorphisms of @xmath3 defined over an extension @xmath7 . for a @xmath10-algebra @xmath11
we use @xmath12 to denote the @xmath11-endomorphism algebra of @xmath3 . finally , by a over a field @xmath4 we mean a smooth , projective , geometrically integral scheme of finite type over @xmath4 of dimension @xmath13 .
let @xmath4 be a perfect field and let @xmath14 be a curve of genus 2 over @xmath4 .
then @xmath14 has a ( or ) @xmath15 where @xmath16 $ ] have degrees bounded by @xmath17 and @xmath18 , respectively .
it follows that @xmath14 is hyperelliptic over @xmath4 , with the map @xmath19 of degree @xmath0 .
now suppose that @xmath20 . completing the square yields a weierstrass model @xmath21 for @xmath14 .
for @xmath22 , let @xmath23[x]$ ] and define @xmath24.\ ] ] for a polynomial @xmath25 $ ] with @xmath26 , define @xmath27 by specialization .
following liu @xcite , the of the weierstrass model @xmath28 is defined by @xmath29 for @xmath30 and @xmath31 in @xmath32[x]$ ] , the greatest common divisor of the coefficients of @xmath33 is @xmath34 , so @xmath35,\ ] ] and in particular , if @xmath36 $ ] , then @xmath37 for a domain @xmath38 .
for any @xmath39 $ ] with @xmath40 and @xmath41 , we have @xmath42 if and only if @xmath43 is a weierstrass model for a genus 2 curve over @xmath4 . in the special case that @xmath44 is monic with @xmath45 and @xmath46 ( which can be achieved if and only if @xmath14 has a weierstrass point defined over @xmath4 ) , this agrees with the definition given by lockhart @xcite . if @xmath47 is another weierstrass model for the curve @xmath14 , then it is related to via a change of variables of the form @xmath48,\ \deg j \le 3,\ ] ] and we have @xmath49 ; see liu @xcite .
thus , as an element of @xmath50 , the discriminant @xmath51 is an invariant of @xmath14 , and if @xmath52 is a discrete valuation on @xmath4 , then @xmath53 is also an invariant .
now suppose that @xmath4 is a local field or a ( global ) number field , and let @xmath11 be the ring of integers of @xmath4 .
an of @xmath14 is a weierstrass model with @xmath36 $ ] , which implies @xmath37 .
we may rescale any given weierstrass model to obtain an integral model after clearing denominators .
if @xmath4 is a non - archimedean local field with discrete valuation @xmath52 , then a for @xmath14 is an integral model that minimizes @xmath54 , and the @xmath55 is the @xmath11-ideal generated by @xmath51 for any minimal model @xmath43 . if instead @xmath4 is a number field , then the minimal discriminant of @xmath14 is defined by @xmath56 where @xmath57 ranges over the non - zero prime ideals of @xmath11 and @xmath58 is the completion of @xmath4 at @xmath57 , and we say that an integral model for @xmath14 is a if @xmath59 .
global minimal models do not always exist , but if @xmath4 has class number one , then @xmath14 has a global minimal model ( * ? ? ?
* remarque 6 ) .
minimal models for genus 2 curves over @xmath60 can be computed in magma @xcite via the function ` minimalweierstrassmodel ` , implemented by michael stoll .
when @xmath61 , the discriminant @xmath62 of every minimal model is the same , since it is determined up to sign by the principal ideal @xmath55 and @xmath63 ; we then view @xmath55 as the integer @xmath62 and define the of @xmath14 to be @xmath64 .
we wish to enumerate integral models @xmath43 of genus 2 curves @xmath14 over @xmath60 with the goal of finding as many non - isomorphic curves as possible with absolute discriminant @xmath65 , where @xmath66 is our chosen discriminant bound .
computing the minimal discriminant of a given curve @xmath14 is time - consuming , so we instead just search for integral models with @xmath67 .
we then minimize these models if necessary , but this is rarely required ; absolute discriminants bounded by @xmath2 will necessarily have minimal valuation at all primes @xmath68 , since @xmath69 .
our approach is to enumerate a large finite set @xmath70 of integral models from which we will select those with @xmath71 .
the set @xmath70 is defined by constraints on the coefficients of @xmath72 $ ] that are chosen so that @xmath70 contains many minimal models .
we can assume without loss of generality that the coefficients @xmath73 of @xmath74 lie in the set @xmath75 : indeed , if @xmath76 , applying with @xmath77 and @xmath78 and @xmath79 yields an integral model @xmath80 with the same discriminant .
we can not , however , assume @xmath81 ; indeed , only a small fraction of the curves we eventually found have minimal models with @xmath81 .
this leaves @xmath82 possible @xmath83 $ ] with @xmath41 to consider .
one can reduce this to 6 via the substitutions @xmath84 and @xmath85 which leave @xmath51 fixed ; however , we choose not to do this as these transformations may move a model outside the set @xmath70 we are searching ( we do use this symmetry to optimize our search ) . for the coefficients of @xmath86 we impose constraints of the form @xmath87 that determine the shape of a box ( orthogonal parallelepiped ) in which we hope to find integral models with small discriminants .
we experimented with various shapes and eventually settled on three : @xmath88 the `` flat '' box @xmath89 is a hypercube ( a ball in the @xmath90-metric ) .
the `` weighted '' box @xmath91 complements @xmath89 and tries to find curves with small discriminant but large `` flat '' height , with weighting designed to give each term in the discriminant roughly the same size .
the `` hill '' box @xmath92 complements @xmath93 and tries to find curves with small discriminant but large `` flat '' and `` weighted '' height : it allows larger middle coefficients with smaller end coefficients .
taken together , the boxes @xmath94 are a crude attempt to take the oddly shaped region defined by @xmath95 and approximate it with boxes that have little overlap .
table [ tab : boxsearch ] lists the number of isomorphism classes with @xmath96 for @xmath97 , using boxes of cardinality @xmath98 with @xmath99 for these three shapes . to squeeze out a bit more , we also considered sets @xmath100 which can be viewed as a union of @xmath101 boxes . for our final search we settled on @xmath102 with cardinalities @xmath103 and @xmath104 ; see table [ tab : boxsearchfinal ] for more details .
the discriminant polynomial @xmath105 $ ] is a homogeneous polynomial of degree 10 with 246 terms . directly computing @xmath106 for a single model @xmath107
involves at least 1000 arithmetic operations , but to compute @xmath51 for all @xmath107 in a box we instead use an algorithm that efficiently enumerates values of a multivariate polynomial @xmath108 over integer points in a box @xmath109 of @xmath110 . with this approach
we require less than 6 clock cycles , on average , to compute each discriminant @xmath51 for @xmath111 .
the algorithm first creates a with @xmath112 levels .
a node at level @xmath113 in the tree corresponds to a monomial in @xmath114 , identified by a vector of @xmath115 non - negative exponents , and contains storage for an integer coefficient .
the nodes at level @xmath112 ( the leaves ) correspond to the terms of @xmath116 .
the nodes at levels @xmath117 correspond to the terms of a partially instantiated polynomial @xmath118 , for some @xmath119 ; the monomial exponents depend only on @xmath116 , but the coefficients vary with @xmath120 .
each node at level @xmath121 has an edge leading to the node at level @xmath115 corresponding to the substitution @xmath122 . for @xmath123 , the polynomial @xmath124 is univariate , and we may efficiently enumerate its values at integers in the interval @xmath125 using finite differences , as in kedlaya sutherland @xcite .
each step requires just @xmath126 additions , where @xmath127 .
for @xmath128 we reduce to the case @xmath129 as follows .
we begin with @xmath130 and compute the powers @xmath131 for @xmath132 , where @xmath133 .
we then use these powers to instantiate the monomials at level @xmath112 of the tree at @xmath134 , using one multiplication per node , thereby obtaining a coefficient value for each instantiated monomial .
we then set the coefficient value of each node at level @xmath129 to the sum of the coefficient values of its instantiated children .
the nodes at level @xmath129 now represent the terms of the polynomial @xmath135 in @xmath129 variables whose values we may recursively enumerate over integer points in @xmath136 .
we then increment @xmath137 and repeat until @xmath138 . the time to enumerate the values of @xmath108 over @xmath139
is dominated by the time spent in the bottom layer of the recursion ( @xmath123 ) , which involves @xmath140 integer additions .
we treat each of the 16 possible values of @xmath141 as fixed , and for each @xmath141 view @xmath142 as a polynomial in the @xmath143 variables @xmath144 . using a monomial tree to enumerate values of @xmath145 $ ] yields an algorithm that essentially uses just @xmath146 additions per discriminant . as a further optimization , we work modulo @xmath147 so that each addition can be accomplished in a single clock cycle .
the monomial tree for @xmath116 contains 703 nodes and requires less than 8 kilobytes of storage , which easily fits in the l1 data cache of a modern cpu .
to find integral models @xmath43 with small discriminant , we look for pairs @xmath107 for which @xmath148 for some integer @xmath149 . for the relatively few pairs that pass this test ,
we recompute @xmath51 over @xmath10 to check if in fact @xmath150 ; this recomputation occurs so rarely that it has a negligible impact on the total running time .
our search can be trivially parallelized by partitioning each box into products of the form @xmath109 and applying the algorithm described in the previous section . using the compute engine on google s cloud platform @xcite
, we ran a parallel computation to enumerate all pairs @xmath107 with @xmath151 in the sets @xmath152 defined above , which together contain over @xmath153 pairs @xmath107 .
we allocated high - cpu compute nodes distributed across six of google s compute engine regions ( two in the central u.s .
, three in the eastern u.s . , and one in western europe ) .
each high - cpu node comes equipped with 32 intel haswell cores and 28.8 gb of memory , yielding a total of cores .
we performed approximately 50 core - years of computation over the course of less than 8 hours of elapsed time .
to reduce the cost we used preemptible nodes . on average , roughly 3 percent of our nodes were preempted each hour ; these were automatically restarted from checkpoints that were created every few minutes .
we now compare our results with existing tables of genus 2 curves over @xmath60 . in every case
we find that our tables include all previously known examples that fall within our discriminant range ( as well as many that are completely new ) .
* the most comprehensive list of genus 2 curves over @xmath60 previously compiled is due to stoll @xcite , who searched for curves with odd absolute discriminants bounded by @xmath154 , finding distinct @xmath60-isomorphism classes with distinct discriminants .
our search found all of these curves , as well as many more : in total , we found @xmath60-isomorphism classes of genus 2 curves over @xmath60 with distinct odd discriminants of absolute value up to @xmath154 .
our tables differ already at the third smallest absolute minimal discriminant @xmath155 , where we find two non - isomorphic curves @xmath156 only one of which ( the first ) was found by stoll @xcite .
the smallest odd absolute discriminant found in our search that is new is , which arises for the curve @xmath157 * merriman
smart @xcite determined all genus 2 curves over @xmath60 with good reduction away from @xmath0 .
up to @xmath60-isomorphism there are 427 such curves , of which http://www.lmfdb.org/genus2curve/q/?abs_disc=128,256,512,1024,4096,8192,16384,32768,65536,131072,262144,524288[29 curves ] have absolute discriminant bounded by @xmath2 , all found by our search .
( the models given by smart ( * ? ? ?
* table ) are not necessarily minimal ; we used magma @xcite to compute minimal models . ) * baker
gonzlez - jimnez
poonen @xcite and gonzlez gonzlez - jimnez @xcite determined a complete list of new modular curves of genus @xmath0 , i.e. , those curves @xmath14 that admit a dominant morphism @xmath158 from the modular curve @xmath159 ( for which @xmath160 ) for some level @xmath161 .
up to @xmath60-isomorphism there are 213 such curves : 149 have @xmath60-simple jacobians @xcite and 64 have split jacobians @xcite . of these 213 curves ,
all 61 with absolute discriminant bounded by @xmath2 were found in our search .
* gonzlez gonzlez - jimnez
gurdia @xcite ( see also gonzlez gurdia rotger @xcite ) computed equations of genus @xmath0 curves whose jacobians are modular abelian varieties , i.e. , they are subabelian varieties of the jacobian @xmath162 for some @xmath163 . among the 75 curves they list , the 3 with absolute discriminant bounded by @xmath2
were found in our search . *
rotger @xcite computed explicit models for some shimura curves of genus 2 ; of these , the @xmath17 curves with absolute discriminant bounded by @xmath2 were also found in our search . * in pursuit of the paramodular conjecture , brumer
kramer @xcite list 38 genus 2 curves @xmath14 whose jacobians @xmath164 have @xmath165 of odd conductor less than @xmath166 . of these
, @xmath167 have absolute discriminant bounded by @xmath2 , all of which were found in our search . in total
we found 56 @xmath60-isomorphism classes of genus 2 curves @xmath14 whose jacobians have odd conductor less than @xmath166 , of which http://www.lmfdb.org/genus2curve/q/?cond=169,249,277,295,349,353,363,389,427,461,523,529,555,587,597,603,691,709,713,731,741,743,745,763,797,807,841,847,893,909,925,953,961,971,975,997&disc=&num_rat_wpts=&torsion=&torsion_order=&two_selmer_rank=&is_gl2_type=false&st_group=&real_geom_end_alg=&aut_grp=&geom_aut_grp=[43 curves ] have @xmath168 .
these all appear to lie in the @xmath167 isogeny classes identified by brumer
kramer using jacobians of curves with absolute discriminants below @xmath2 .
* in a similar vein , farmer
koutsoliotas
lemurell @xcite used analytic methods to determine a complete list of the integers @xmath169 that can arise as the conductor of degree-4 @xmath1-functions under a standard set of hypotheses .
excluding @xmath1-functions arising from products of elliptic curves or classical modular forms with real quadratic character , they list 12 values of @xmath169 that may arise for the @xmath1-function of a genus 2 curve ( * ? ? ?
* theorem 2.1 ) ( the 9 odd values match the brumer kramer list up to @xmath170 ) .
we found http://www.lmfdb.org/genus2curve/q/?cond=249,277,295,349,353,388,389,394,427,461,464,472[21 curves ] matching these 12 values ( and at least one for each @xmath171 ) , including http://www.lmfdb.org/genus2curve/q/?cond=388,464,472[6 curves ] that match the even values @xmath172 , @xmath173 , @xmath174 on their list . despite our attempts at exhaustiveness ,
the data in table [ tab : boxsearch ] suggest that our tables are still incomplete .
indeed , there are genus 2 curves with absolute minimal discriminant bounded by @xmath2 that do not have an integral model in any of the sets @xmath152 that we searched : we found 28 such curves while experimenting with boxes of different shapes .
one way to get a sense for the completeness of our tables is to consider an analogous search among genus 1 curves .
the discriminant polynomial then has degree 6 , rather than 10 , so we use the discriminant bound @xmath175 .
the jacobian of such a genus 1 curve is isomorphic to an elliptic curve whose conductor must also be bounded by @xmath176 and can therefore be found in cremona s tables @xcite ( by ogg s formula , the conductor of an elliptic curve over @xmath60 divides its minimal discriminant ) . in total
there are 1154 @xmath60-isomorphism classes of elliptic curves with absolute discriminant bounded by @xmath176 .
a search for genus 1 curves @xmath43 with @xmath46 and @xmath177 using the same coefficient bounds as in the box @xmath178 already hits 1132 of these @xmath60-isomorphism classes , or about 98 percent .
the isomorphism class of a genus @xmath0 curve over an algebraically closed field can be characterized by a number of _
invariants_. for elliptic curves , only the elliptic @xmath179-invariant is required . in genus
@xmath0 , there are different choices of invariants available , all with their own distinct advantages and drawbacks .
we refer to seminal work by igusa @xcite and later work by mestre @xcite for some theoretical expositions .
the details of the implementation in magma can be found in its handbook @xcite ; the package was written by everett howe , based on routines developed by fernando rodriguez - villegas , and with additional functionality added by reynald lercier and christophe ritzenthaler .
let @xmath4 be a field with @xmath180 and let @xmath14 be a genus @xmath0 curve over @xmath4 .
then @xmath14 has a simplified weierstrass model @xmath181 where the @xmath182 are the distinct roots of @xmath44 in an algebraic closure @xmath183 of the base field @xmath4 . given two such roots @xmath184 , we denote the difference @xmath185 by @xmath186 .
we then form the expressions @xmath187 where each sum and product runs over the distinct expressions obtained by permuting the index set @xmath188 .
the invariants @xmath189 , defined by igusa @xcite by modifying a set of invariants due to clebsch , are known as the .
two curves @xmath14 and @xmath190 over @xmath4 are isomorphic over the algebraic closure @xmath183 if and only if @xmath191 describe the same point in weighted projective @xmath192-space ; that is , if and only if there exists a @xmath193 such that @xmath194 . the igusa clebsch invariants break down in characteristic @xmath0 , and to deal with this problem , igusa @xcite found a common normal form for hyperelliptic curves in arbitrary characteristic . motivated by this , he defined the invariants @xcite @xmath195 now called the , with @xmath196 .
moreover , these invariants are compatible with specialization of the model , for example , reducing modulo a prime .
while the igusa invariants are in a sense the best available with respect to arithmetic questions , we need @xmath197 of them to describe a moduli space of dimension @xmath18 . however , the moduli space of genus @xmath0 curves over a given field admits a simpler description : the corresponding moduli space can always be described by @xmath18 affine coordinates .
such coordinates were given by cardona
quer nart
pujolas @xcite . for @xmath180 ,
these ( sometimes also known as ) can be defined by @xmath198 the g2-invariants are _ absolute _ , with @xmath199 in affine space .
this time , @xmath14 has potentially good reduction at a prime if and only if none of the g2-invariants reduce to @xmath200 modulo that prime .
conversely , given three g2-invariants in a field @xmath201 , there always exists a curve with these invariants .
however , writing down such a curve may require a quadratic extension of @xmath201 .
this obstruction was studied by mestre @xcite and cardona quer @xcite .
the @xmath1-series of a genus @xmath0 curve @xmath14 over @xmath60 is a dirichlet series with an euler product @xmath202 such that for primes @xmath203 of good reduction we have @xmath204 and @xmath205 $ ] has degree @xmath206 .
efficient methods for computing the coefficients of @xmath207 for all good primes @xmath203 up to a given bound are discussed in kedlaya
sutherland @xcite and harvey
sutherland @xcite and available in ` smalljac ` @xcite .
if abelian varieties @xmath208 over @xmath60 are isogenous over @xmath60 , then @xmath209 for all primes @xmath203 . to facilitate the efficient grouping of curves into candidate isogeny classes ( of their jacobians ) , we define the of an abelian surface @xmath3 to be the unique integer @xmath210 $ ] for which @xmath211 where @xmath212 and @xmath213 , with @xmath214 and @xmath215 ( so the @xmath216 correspond to successive digits in the base-@xmath217 expansion of @xmath218 ) .
isogenous abelian surfaces necessarily have the same hash , and while the converse need not hold in general , it almost always does
. hashes also provide a way to provisionally identify automorphic objects ( such as modular forms ) that have the same @xmath1-function . for an abelian surface @xmath3 over a number field @xmath4
, there are 52 distinct sato
tate groups @xmath219 that arise , of which 34 occur
when @xmath61 ( * ? ? ?
* theorem 1.5 ) .
a coarser classification is obtained if we restrict our attention to the identity component @xmath220 , equivalently , the sato tate group @xmath221 of the base extension of @xmath3 to the minimal field @xmath7 over which all its endomorphisms are defined . in this case
there are just 6 possibilities , corresponding to 6 types of real endomorphism algebras @xmath222 , all of which can arise when @xmath61 .
these are listed in table [ table : connectedst ] , along with the possible types of abelian varieties @xmath6 ( up to isogeny ) .
the fact that the average analytic rank in our database is significantly greater than the expected average value of @xmath223 is likely due to the fact that having small coefficients made it more likely that @xmath224 will be a sextic whose leading coefficient is a square , in which case the difference of the two points at infinity is a rational divisor that typically has infinite order ( it is torsion if and only if the continued fraction expansion of @xmath225 in the hyperelliptic function field @xmath226 is periodic @xcite ) .
when this occurs the mordell - weil rank of the jacobian is @xmath13 greater than it would be generically . | we describe the construction of a database of genus @xmath0 curves of small discriminant that includes geometric and arithmetic invariants of each curve , its jacobian , and the associated @xmath1-function . this data has been incorporated into the @xmath1-functions and modular forms database ( lmfdb ) . |
the question of an experimental evidence of physics beyond the standard model ( sm ) has been and still is an issue in particle physics .
this topic is especially popular nowadays , when new high - energy experiments are being constructed and put in work . regarding the sm as a low - energy approximation of a more general theory
, one can expect the new physics to manifest itself in two ways : important effects in measurements at high energies or small effects at relatively low energies .
the latter is the case of the muon magnetic anomaly , often referred to as the muon @xmath1 problem . to answer the question ,
whether some new physic plays a role in this case , one needs to achieve a high precision in both , theoretical prediction and experimental measurement .
the muon gyromagnetic ratio relates the muon spin and magnetic momentum @xmath2 and the magnetic anomaly is defined by @xmath3 the higher muon mass makes its magnetic momentum , in comparison to the magnetic momentum of the electron , more sensitive to the non - perturbative hadronic effects and thus can the @xmath4 determination serve as a good test field of the theory of strong interactions .
a high - precision experimental result is available , the anomaly @xmath4 was measured by the @xmath1 collaboration in the e821 experiment at bnl with unprecedent precision of 0.7 ppm @xcite .
a high - precision theoretical prediction is available also , but only in the electromagnetic and electroweak sector the accuracy is satisfactory . in case of the hadronic contribution @xmath5 more precision is desirable , because this contribution is the dominant source of the total uncertainty of the theoretical prediction , which does not allow to conclude on possible physics beyond the sm .
depending on the author and the analyzed data one can cite different numbers . if one takes the values from the summary publication @xcite @xmath6 and @xmath7 one observes @xmath8
, @xmath9 .
the hadronic part is dominated by the lowest order and this can be calculated using the formula@xmath10 where @xmath11/[x^{2}+(1-x)\frac{s}{m_{\mu}^{2}}]$ ] and @xmath12 .
the most important contribution comes from the pion channel @xmath13 , which we study in this work .
the data are used to evaluate the integral up to few gev , above a perturbative calculation is possible . in order to compare to other authors we chose three different limits for the upper integration limit ( @xmath14 , @xmath15 and @xmath16 ) .
the expression ( [ eq : a_mi_had_lo ] ) was up to now @xcite always evaluated only by the direct integration of the @xmath17 data ( with small exceptions , see @xcite ) .
our motivation and our aim is to demonstrate , that the use of an appropriate model allows for a dramatic error reduction . the error reduction in this domain is of a crucial importance , since there is an effort to decrease even more the experimental error @xcite and so , if one wants to solve the @xmath1 puzzle , the theoretical precision will have to follow .
we base our evaluation of @xmath18 on the unitary and analytic ( u&a ) model of the pion electromagnetic ( em ) form factor and relate the form factor to the cross section @xmath19 , so as to obtain @xmath20 . for a reliable prediction one needs to construct a model that incorporates all known properties of the pion em form factor .
the form factor is charge - normalized @xmath21 and pqcd predicts its asymptotic behavior @xmath22 , where @xmath23 .
it is known , that @xmath24 is an analytic function in the whole complex @xmath25 plane , besides the branch points and the cut on the positive real axis , the cut going from the lowest branch point @xmath26 up to @xmath27 .
the pion form factor has also cut on the second , unphysical riemann sheet for @xmath28 and satisfies the unitarity condition @xmath29=\left[a_{1}^{1}(t)\right]^{*}f_{\pi}(t)+\sigma(t)$ ] , where @xmath30 is the _ p_-wave isovector transition amplitude for elastic @xmath31 scattering and @xmath32 approximates all higher contributions . for @xmath33 one
has @xmath34 and arrives to the so - called elastic unitarity condition .
the construction of the u&a model starts from the vector meson dominance ( vmd ) picture in which@xmath35 where @xmath36 is the vector meson - pion coupling constant , @xmath37 is the universal vector meson coupling constant and three resonances are taken into account .
we proceed to a non - linear transformation @xmath38^{2}}\ ] ] in order to build in the cut , the lowest branch point @xmath39 and an effective branch point @xmath40 , the latter meant to approximate all higher branch points .
the model then becomes@xmath41 @xmath42 @xmath43,\ ] ] where @xmath44 and @xmath45 .
the expression can be modified further ; it can be shown that depending on the relative positions of thresholds and masses one has @xmath46 and @xmath47 .
the masses and @xmath40 are considered as free parameters of the model and so their relative positions are not fixed . however , in order to provide explicit formulas in the following text , we will here suppose @xmath48 , what is actually being confirmed by the result of the data fitting .
the next step in the model construction is incorporation of the cut on the second riemann sheet .
we use a succession of poles and zeros generated by a rational function to approximate this cut ( pad approximation ) .
we do it by adding to the model a multiplicative term @xmath49 , where @xmath50 and @xmath51 are free parameters from the interval @xmath52 ( corresponds to @xmath28 ) . the construction of the model is finally achieved by giving the vector mesons non - zero decay widths @xmath53 .
the model gets form@xmath54=\left(\frac{1-w^{2}}{1-w_{n}^{2}}\right)^{2}\frac{\left(w - w_{z}\right)\left(w_{n}-w_{p}\right)}{\left(w_{n}-w_{z}\right)\left(w - w_{p}\right)}\ ] ] @xmath55 @xmath56 @xmath57 @xmath58.\ ] ] the ratio of couplings @xmath59 is a free parameter of the model , the two remaining ratios , @xmath60 and @xmath61 , can be related to the first one by using the normalization condition @xmath21 and by considering the behavior of the imaginary part of the form factor at @xmath62 .
the @xmath63 interference is taken into account when fitting the experimental data by a breit - wigner term , the fitting function is@xmath64+re^{i\phi}\frac{m_{\omega}^{2}}{m_{\omega}^{2}-t - im_{\omega}\gamma_{\omega}},\ ] ] where @xmath65 and the amplitude @xmath66 is considered as an additional free parameter .
the fitting function thus has 11 free parameters in total ( @xmath67 three masses , three widths , ratio @xmath59 , @xmath68 , @xmath51 and @xmath66 ) and is used to fit 523 experimental points measured in experiments _ cleo _ , _ na7 _ , _ olya _ , _ cmd _ , _ cmd-2 _ , _ snd _ , _ kloe _ , in _ jinr _ dubna and in @xmath69 collaboration at jefferson lab @xcite .
the best fit result is shown in figure [ fig : fit ] .
fit of the data with the u&a model . ]
two approaches were used for the error evaluation .
the first one was a `` monte carlo '' method based on a random number generator with the assumption of the gaussian distribution for the uncertainties of the published data points .
for each point a new one was randomly generated using the gaussian probability density function with the mean identical to the original point and @xmath70 equal to the published error .
doing this for each point , new `` random '' data set was obtained .
this data set was then fitted and the values of the parameters of the model @xmath71 as well as the value of @xmath18 were extracted . repeating the whole procedure 4000 times , we reached statistics high enough to allow us for a reliable error calculation .
the mean @xmath72 and the @xmath70 were calculated from the 4000 values and since the mean is not , in general , identical with the optimal - fit value @xmath73 we present asymmetric uncertainties @xmath74 , where @xmath75 and @xmath76 . in the second approach
the program minuit was used to establish the uncertainties of the model parameters .
then , taking the numerical derivatives for @xmath77 , the uncertainty was propagated to @xmath18 using the covariance matrix . in this method
the errors are symmetric .
in addition to the integration of the model , we also performed a direct integration of the data points based on the trapezoidal rule , so as to cross - check our compatibility with other authors .
our results and some results from other authors @xcite are summarized in table [ tab : results ] .
llll & interval @xmath78 $ ] & @xmath79 & @xmath80 & @xmath81 model , method 1 & @xmath82 & @xmath83 & @xmath84 model , method 2 & @xmath85 & @xmath86 & @xmath87 data & @xmath88 & @xmath89 & @xmath90 _ davier _ & @xmath91 & & _ hagiwara et al . _
& & @xmath92 & _ yndurin et al .
_ & & & @xmath93
the use of the u&a model dramatically reduces the error on @xmath18 .
this is not an arbitrary feature of the model but originates from model - independent information which is additional to the data in the integration region and which can be taken into account when the model is used .
the most important sources contributing to error reduction are * expected smoothness of the @xmath24 at small scale @xmath94 : the model provides a function behaving smoothly at small @xmath94 . * experimental data outside the integration region :
the fit is done not only to the data inside , but also to the data outside the integration region .
* theoretical knowledge on @xmath24 : the model respects all known properties of the pion em form factor .
especially the first point plays an important role .
the new precise data tend to lie above older , less precise data and , in some regions , the vertical spread of the data is very important , at the limit of inconsistency .
if the calculation of the integral is based directly on data , then less precise data shift the mean value of the integral and enlarge the uncertainty .
when the model is used , the predicted behavior of @xmath24 as given by the result of the fit is mostly determined by precisely measured points and is only little influenced by data with important uncertainties .
this leads to more appropriate mean value and smaller errors . in consequence ,
one arrives to the mean value of @xmath18 which is higher then what is obtained by the direct data integration and to much reduced uncertainty .
the shift in the mean value goes in the `` right '' direction and brings the theoretical value closer to the experimental one .
the error estimates from the two used methods are not fully compatible , the first method gives smaller errors .
this might be related to statistical fluctuations ( 1st method ) and to approximations - numerical derivatives and linearization ( 2nd method ) . in this article we presented the calculation of @xmath18 based on the u&a model .
this approach allows for important error reduction and we plan to use it also for evaluating the contributions of other channels to @xmath95 .
the work was partly supported by slovak grant agency for sciences vega , grant no . 2/7116/29 . | we use the unitary and analytic model of the pion electromagnetic form factor in order to evaluate in the lowest order the @xmath0 contribution into the muon magnetic anomaly .
we demonstrate , that this technique enables us to reduce the uncertainty of the theoretical prediction importantly in comparison to usual approaches , where the measured data are integrated directly . |
the polynomial remainder sequence ( prs ) is one of fundamental tools in computer algebra .
although the euclidean algorithm ( see knuth ( @xcite ) for example ) for calculating prs is simple , coefficient growth in prs makes the euclidean algorithm often very inefficient . to overcome this problem ,
the mechanism of coefficient growth has been extensively studied through the theory of subresultants ; see collins ( @xcite ) , brown and traub ( @xcite ) , loos ( @xcite ) , etc . by the theory of subresultant
, we can remove extraneous factors of the elements of prs systematically . in this paper
, we consider a variation of the subresultant .
when we calculate prs for polynomials which have a nontrivial gcd , we usually stop the calculation with the gcd .
however , it is sometimes useful to continue the calculation by calculating the prs for the gcd and its derivative ; this is necessary for calculating the number of real zeros including their multiplicities .
we call such a prs a `` recursive prs . ''
although the theory of subresultants has been developed widely , the corresponding theory for recursive prs is still unknown within the author s knowledge : this is the main problem which we investigate in this paper . by `` recursive subresultants ,
'' we denote determinants which represent elements of recursive prs by the coefficients of initial polynomials .
this paper is organized as follows . in sect .
[ sec : recprs ] , we introduce the concept of recursive prs . in sect .
[ sec : subresrecprs ] , we define recursive subresultant and show its relationship to recursive prs . in sect . [ sec : disc ] , we discuss briefly using recursive subresultants in approximate algebraic computation .
first , we review the prs , then define the recursive prs . at
last , we show recursive sturm sequence as an example of recursive prs .
let @xmath0 be an integral domain and polynomials @xmath1 and @xmath2 be in @xmath3 $ ] .
we define a polynomial remainder sequence as follows .
[ def : prs ] let @xmath1 and @xmath2 be polynomials in @xmath3 $ ] of degree @xmath4 and @xmath5 ( @xmath6 ) , respectively . a sequence @xmath7 of nonzero polynomials
is called a _ polynomial remainder sequence ( prs ) _ for @xmath1 and @xmath2 , abbreviated to @xmath8 , if it satisfies @xmath9 for @xmath10 , where @xmath11 @xmath12 are elements of @xmath0 and @xmath13 .
a sequence @xmath14 @xmath15 is called a _ division rule _ for @xmath8 ( see von zur gathen and lcking ) .
if @xmath16 is a constant , then the prs is called _ complete_. if @xmath1 and @xmath2 are coprime , the last element in the complete prs for @xmath1 and @xmath2 is a constant .
otherwise , it equals the gcd of @xmath1 and @xmath2 up to a constant : we have @xmath17 for some @xmath18 .
then , we can calculate new prs , @xmath19 , and if this prs ends with a non - constant polynomial , then calculate another prs for the last element , and so on . by repeating this calculation , we can calculate several prss `` recursively '' such that the last polynomial in the last sequence is a constant .
thus , we define `` recursive prs '' as follows .
[ def : recprs ] let @xmath1 and @xmath2 be the same as in definition [ def : prs ] .
then , a sequence @xmath20 of nonzero polynomials is called a _ recursive polynomial remainder sequence _ ( recursive prs ) for @xmath1 and @xmath2 , abbreviated to @xmath21 , if it satisfies @xmath22 for @xmath23 .
if @xmath24 , @xmath25 satisfy @xmath26 for @xmath27 and @xmath28 , then a sequence @xmath29 is called a _ division rule _ for @xmath21 . furthermore ,
if @xmath30 is a constant , then the recursive prs is called complete .
sturm sequence is a variant of prs , which is used in sturm s method , for calculating the number of real zeros of univariate polynomial ( for detail , see cohen ( @xcite ) for example ) .
note that sturm s theorem is valid for not only polynomials having simple zeros but also those having multiple zeros ( see bochnak , coste and roy ( @xcite ) for example ) .
here , we define `` recursive sturm sequence '' to calculate the number of real zeros including multiplicities , as follows .
let @xmath31 be a real polynomial of degree @xmath4 .
let a sequence of nonzero polynomials be defined by a recursive prs in definition [ def : recprs ] , calculated as @xmath32 with division rule given by @xmath33 for @xmath27 and @xmath28 .
then , the sequence is called the _ recursive sturm sequence _ of @xmath31 .
[ ex : recsturmseq ] let @xmath34 , and calculate the recursive sturm sequence of @xmath31 . the first sequence @xmath35 has the following elements : @xmath36 the second sequence @xmath37 has the following elements : @xmath38 the last sequence @xmath39 has the following elements : @xmath40 for prs @xmath41 , @xmath42 , define sequences of nonzero real numbers @xmath43 and @xmath44 as @xmath45 where @xmath46 denotes the degree of @xmath47 and @xmath48 denotes the leading coefficients of @xmath47 .
then , @xmath43 and @xmath44 for @xmath42 are @xmath49 for a sequence of nonzero real numbers @xmath50 , let @xmath51 be the number of sign variations of the elements of @xmath52 .
then , we calculate the number of the real zeros of @xmath31 , including multiplicity , as @xmath53
let @xmath1 and @xmath2 be polynomials in @xmath3 $ ] such that @xmath54 to make this paper self - contained and to use notations for our definitions , we first review the fundamental theorem of subresultants , then discuss subresultants for recursive prs . there exist several different definitions of subresultants . here , we follow definitions and notations basically by von zur gathen and lcking ( @xcite ) .
let @xmath1 and @xmath2 be as in .
the _ sylvester matrix _ of @xmath1 and @xmath2 , denoted by @xmath55 , is an @xmath56 matrix constructed from the coefficients of @xmath1 and @xmath2 , such that @xmath57 next , we define the `` subresultant matrix '' to derive polynomial subresultants .
let @xmath1 and @xmath2 be defined as in . for @xmath58 ,
the _ @xmath59-th subresultant matrix _ of @xmath1 and @xmath2 , denoted by @xmath60 , is an @xmath61 sub - matrix of @xmath55 obtained by taking the left @xmath62 columns of coefficients of @xmath1 and the left @xmath63 columns of coefficients of @xmath2 , such that @xmath64 let @xmath1 and @xmath2 be defined as in . for @xmath58 and @xmath65 , let @xmath66 be a sub - matrix of @xmath60 obtained by taking the top @xmath67 rows and the @xmath68-th row ( note that @xmath69 is a square matrix ) .
then , the polynomial @xmath70 is called the _ @xmath59-th subresultant _ of @xmath1 and @xmath2 .
[ th : fundsubres ] let @xmath1 and @xmath2 be defined as in , @xmath71 be complete prs , and @xmath14 @xmath72 be its division rule .
let @xmath73 and @xmath74 for @xmath75 , and @xmath76 for @xmath77 .
then , we have @xmath78 for @xmath79 . by the above theorem
, we can express coefficients of prs by determinants of matrices whose elements are the coefficients of initial polynomials .
we construct `` recursive subresultant matrix '' whose determinants represent elements of recursive prs by the coefficients of initial polynomials . to make help for readers ,
we first show an example of recursive subresultant matrix for recursive sturm sequence in example [ ex : recsturmseq ] . [ ex : recsubresmat ] we express @xmath31 and @xmath80 in example [ ex : recsturmseq ] by @xmath81 let @xmath82 , then the matrices @xmath83 , @xmath84 and @xmath85 are given as @xmath86 where horizontal lines in matrices divide them into the upper and the lower components .
note that the matrix @xmath87 is derived from @xmath84 by multiplying the @xmath88-th row by @xmath89 for @xmath90 and deleting the lowest row .
similarly , the @xmath91-th recursive subresultant matrix @xmath92 is constructed as @xmath93 [ def : recsubresmat ] let @xmath1 and @xmath2 be defined as in , and let @xmath94 be complete recursive prs for @xmath1 and @xmath2 as in definition [ def : recprs ] .
let @xmath95 and @xmath96 for @xmath27 .
then , for each tuple of numbers @xmath97 with @xmath27 and @xmath98 , define matrix @xmath99 as follows . 1 . for @xmath100 ,
let @xmath101 .
2 . for @xmath102 ,
let @xmath99 consist of the upper block and the lower block , defined as follows : 1 .
the upper block is partitioned into @xmath103 blocks with diagonal blocks filled with @xmath104 , where @xmath104 is a sub - matrix of @xmath105 obtained by deleting the bottom @xmath106 rows .
let @xmath107 be a sub - matrix of @xmath105 obtained by taking the bottom @xmath106 rows , and let @xmath108 be a sub - matrix of @xmath107 by multiplying the @xmath109-th rows by @xmath110 for @xmath111 , then by deleting the bottom row .
then , the lower block consists of @xmath112 blocks of @xmath107 such that the leftmost block is placed at the top row of the container block and the right - side block is placed down by 1 row from the left - side block , then followed by @xmath113 blocks of @xmath108 placed by the same manner as @xmath107 .
@xmath114 as a result , @xmath99 becomes as shown in fig .
[ fig : recsubresmat ] .
then , @xmath99 is called the _ @xmath97-th recursive subresultant matrix _ of @xmath1 and @xmath2 .
[ prop : recsubresmat ] for @xmath27 and @xmath115 , the numbers of rows and columns of @xmath99 , the @xmath97-th recursive subresultant matrix of @xmath1 and @xmath2 are @xmath116 and @xmath117 respectively . by induction on @xmath118 . for @xmath100 , by definition of the subresultant matrix , we see that @xmath119 is a @xmath61 matrix .
let us assume that equations and are valid for @xmath120 .
then , we calculate the numbers of the rows and columns of @xmath99 as follows . 1 . the numbers of rows of @xmath107 and @xmath108 are equal to @xmath106 and @xmath121 , respectively , hence the number of rows a block which consists of @xmath107 and @xmath108 in @xmath99 equals @xmath122 on the other hand , the number of rows of @xmath104 is equal to @xmath123 , hence the number of rows of diagonal blocks in @xmath124 is equal to @xmath125 by adding formulas ( [ eq : recsubresmatrow - low ] ) and ( [ eq : recsubresmatrow - up ] ) , we obtain ( [ eq : recsubresmatrow ] ) .
2 . the number of columns of @xmath105 is equal to @xmath126 , hence the number of columns of @xmath99 is equal to ( [ eq : recsubresmatcol ] ) .
this proves the proposition .
now , we define recursive subresultants by recursive subresultant matrices .
[ def : recsubres ] let @xmath1 and @xmath2 be defined as in , and let @xmath127 @xmath128 be complete recursive prs for @xmath1 and @xmath2 as in definition [ def : recprs ] .
let @xmath95 and @xmath96 for @xmath27 .
for @xmath98 and @xmath129 , let @xmath130 be a sub - matrix of the @xmath97-th recursive subresultant matrix @xmath99 obtained by taking the top @xmath131 rows and the @xmath132-th row ( note that @xmath133 is a square matrix ) .
then , the polynomial @xmath134 is called the _ @xmath97-th recursive subresultant _ of @xmath1 and @xmath2 .
finally , we derive the relation between recursive subresultants and coefficients in recursive prs .
[ lem : recsubres ] let @xmath1 and @xmath2 be defined as in , and let @xmath94 be complete recursive prs for @xmath1 and @xmath2 as in definition [ def : recprs ] .
let @xmath135 , @xmath46 , @xmath95 and @xmath96 for @xmath27 and @xmath136 , and let @xmath137 for @xmath27 and @xmath138 .
furthermore , for @xmath139 and @xmath98 , define @xmath140 and let @xmath141 .
for @xmath23 and @xmath98 , define @xmath142 and let @xmath143 and @xmath144 .
then , for the @xmath97-th recursive subresultant of @xmath1 and @xmath2 with @xmath27 and @xmath98 , we have @xmath145 where @xmath146 and @xmath147 for @xmath102 . for a univariate polynomial @xmath148 with @xmath149 for @xmath150 ,
let us denote the coefficient vector for @xmath31 by @xmath151 .
we prove the lemma by induction on @xmath118 . for @xmath100
, it follows immediately from the fundamental theorem of subresultants ( theorem [ th : fundsubres ] ) .
let us assume that ( [ eq : recsubreslem ] ) is valid for @xmath120 .
then , we have @xmath152 hence we have @xmath153 for @xmath154 .
therefore , there exists a matrix @xmath155 such that @xmath156 and that we can transform @xmath105 to @xmath157 by eliminations on columns .
furthermore , by eliminations and exchanges on columns in the block @xmath158 as shown in brown and traub ( @xcite ) , we can transform @xmath159 to @xmath160 such that @xmath161 is a lower triangular matrix with all diagonal elements equal to 1and that , where @xmath162and @xmath163 are sub - matrices of @xmath159 and @xmath164 , respectively , obtained by the same manner as we have obtained @xmath165 .
therefore , we have @xmath166 similarly , let @xmath167 . then , by the same transformations in the above , we can transform @xmath168 to @xmath169 satisfying @xmath170 where @xmath171 and @xmath172 are sub - matrices of @xmath168 and @xmath173 , respectively , obtained by taking the top @xmath174 rows and the @xmath175-th row for @xmath176 .
therefore , for , we can transform @xmath99 to @xmath177 as shown in fig .
[ fig : recsumbresmatbar ] by eliminations and exchanges on columns in each column block , and let @xmath178 be sub - matrix of @xmath177 obtained by the same manner as we have obtained @xmath179 .
@xmath180 then , we have @xmath181 from ( [ eq : mbardetk ] ) and ( [ eq : mbardetk-1 ] ) and since there exist @xmath182 blocks of @xmath183 and @xmath184 in @xmath177 with each of which divided into the upper and the lower block . furthermore , by exchanges on columns
, we can transform @xmath177 to @xmath185 as shown in fig .
[ fig : recsubresmathat ] , and let @xmath186 be sub - matrix of @xmath185 obtained by the same manner as we have obtained @xmath179 . @xmath187 then , we have @xmath188 because the @xmath189-th column in @xmath177 was moved to the @xmath190-th column in @xmath185 for .
furthermore , we have @xmath191 because the lower - right block of @xmath192 and @xmath193 in @xmath185 is equal to @xmath194 . finally , from ( [ eq : rel - m - mbar ] ) , ( [ eq : rel - mbar - mhat ] ) and ( [ eq : rel - mhat - subres ] ) , we have @xmath195 therefore , by the definitions of recursive subresultant , we obtain ( [ eq : recsubreslem ] ) .
this proves the lemma . with the same conditions as in lemma [ lem : recsubres ] , and for @xmath27 and @xmath196
, we have @xmath197 by substituting @xmath198 in ( [ eq : recsubreslem ] ) by ( [ eq : subresthm1])([eq : subresthm4 ] ) , we obtain ( [ eq : recsubresthm1])([eq : recsubresthm4 ] ) , respectively .
we show an example of the proof of lemma [ lem : recsubres ] for recursive subresultant matrix in example [ ex : recsubresmat ] .
let us express @xmath47 in example [ ex : recsturmseq ] by @xmath199 with @xmath46 . by eliminations and exchanges of columns as shown in brown and traub ( @xcite ) , we can transform @xmath200 and @xmath201 in ( [ eq : recsubresmatex ] ) to @xmath202 and @xmath203 , respectively , as @xmath204 where @xmath205 .
furthermore , we have @xmath206 with @xmath207 where @xmath208 and @xmath209 are sub - matrices of @xmath210 and @xmath87 , respectively , obtained by taking the top 4 rows and the @xmath211-th row . therefore , by eliminations and exchanges on columns , we can transform @xmath92 in ( [ eq : recsubresmatex ] ) to @xmath212 as @xmath213 satisfying @xmath214 .
furthermore , by exchanges on columns , we can transform @xmath212 to @xmath215 as @xmath216 satisfying @xmath217 .
therefore , we have @xmath218 for @xmath219 , and we have @xmath220
in this paper , we have defined recursive prs as well as recursive subresultants , and proved a similar theorem as the fundamental theorem of subresultant .
the concept of recursive subresultant is inspired , in approximate algebraic computation , by representing coefficients in recursive prs by the coefficients of initial polynomials .
for example , consider calculating recursive sturm sequence of a polynomial with floating - point number coefficients by floating - point arithmetic . in the case
the initial polynomial has multiple or close zeros , there may exist a polynomial in the sequence such that it is difficult to decide whether the polynomial becomes zero or not .
also , zero recognition of very small leading coefficient is another important problem because it plays crucial role in calculating the number of real zeros . for the problem of zero recognition of very small leading coefficients , the present author and sasaki ( @xcite )
have proposed a criterion for calculating the number of real zeros correctly by sturm s method : if the sturm sequence satisfy certain condition on sylvester matrix , then we can neglect the small leading coefficient which makes computation of the sturm sequence more stable .
we expect that the recursive subresultant ( matrix ) will be useful for zero recognition of a polynomial in recursive sturm sequence , by representing its coefficients by the coefficients of initial polynomials then analyzing it by numerical methods ; this is the problem on which we are working now .
the author thank prof .
tateaki sasaki very much for revising the original manuscript , and the referees for their helpful suggestions .
von zur gathen , j. , lcking , t. : subresultants revisited ( extended abstract ) . in gonnet , g.h .
, panario , d. , viola , a. , eds . : latin 2000 : theoretical informatics .
volume 1776 of lecture notes in computer science .
springer ( 2000 ) 318342 | we introduce concepts of `` recursive polynomial remainder sequence ( prs ) '' and `` recursive subresultant , '' and investigate their properties . in calculating prs , if there exists the gcd ( greatest common divisor ) of initial polynomials , we calculate `` recursively '' with new prs for the gcd and its derivative , until a constant is derived .
we call such a prs a _
recursive prs_. we define _ recursive subresultants _ to be determinants representing the coefficients in recursive prs by coefficients of initial polynomials .
finally , we discuss usage of recursive subresultants in approximate algebraic computation , which motivates the present work . |
as one of the three clock - synchronization algorithms studied for wireless sensor networks ( wsns ) under unknown delay @xcite , leng and wu proposed the generalization of the maximum - likelihood - like estimator ( mlle ) proposed by noh _ et al_. @xcite . to overcome the drawback of the mlle that it can utilize only the time stamps in the first and the last of @xmath0 message exchanges , they extend the gap @xmath1 between two subtracting time stamps from @xmath2 to a range of @xmath3 $ ] so that the generalized mlle can take into more time stamps in estimating clock skew . specifically , the time stamps in two - way message exchanges are modeled as ( * ? ? ?
( 1 ) and ( 2 ) ) @xmath4 where @xmath5 and @xmath6 denote the clock offset and clock skew of the child node @xmath7 with respect to the parent node @xmath8 , respectively ; @xmath9 represents the fixed portion of one - way propagation delay , while @xmath10 and @xmath11 are its variable portions ( see fig . 1 of @xcite ) .
based on and , they construct new sequences @xmath12 ( @xmath13 and @xmath14 ) and model them as follows ( * ? ? ? * eqs .
( 10 ) and ( 11 ) ) : @xmath15 for @xmath13 . noting that @xmath16 and @xmath17 because @xmath18 and @xmath19 are i.i.d .
zero - mean gaussian random variables with variance @xmath20 , they obtain the maximum - likelihood estimator ( mle ) for @xmath6 given by ( * ? ? ?
* eq . ( 13 ) )
@xmath21 the major problem in the derivation of the mle for @xmath6 given in is that , even though @xmath18 and @xmath19 are i.i.d .
gaussian random variables , the noise components @xmath22 and @xmath23 are not in general : for @xmath24 and @xmath25 , @xmath26 \notag \\ & = -\operatorname{e}\left[x_{\alpha+m}x_{n}\right ] - \operatorname{e}\left[x_{m}x_{\alpha+n}\right ] \notag \\ & = \begin{cases } -\sigma^{2 } , & \mbox{if } \alpha=\left|m - n\right| \\ 0 , & \mbox{otherwise } \end{cases } , \end{aligned}\ ] ] and
the same goes for @xmath23 .
note that , if the noise components are independent one another as claimed in @xcite , the expectation in must be zero .
the consequence of is that @xmath1 should be greater than @xmath27 , i.e. , @xmath28 in order to maintain the validity of the derivation of the mle for @xmath6 ( * ? ? ?
* eq . ( 13 ) ) and its performance bound ( * ? ? ?
* eq . ( 29 ) ) : if @xmath29 , there exists at least one pair of @xmath30 and @xmath31 satisfying @xmath32 so that the noise components are no longer independent one another .
for example , let @xmath31 be 1
. then @xmath33 satisfies the said condition . because @xmath29 and @xmath34 @xmath30 belongs to @xmath35 .
[ fig : noise_correlation_effect ] clearly shows the effect of the noise correlation on the mean square error ( mse ) of estimation of clock skew and the relationship between @xmath1 and @xmath0 when snr=30 db and @xmath36 . in the figure
, ge1 denotes the simulation results of the generalized mlle for time stamps and resulting sequences generated according to the original models of through ; ge2 , on the other hand , denotes the results for the time sequences in and with the noise components @xmath37 and @xmath38 replaced by two newly - generated i.i.d .
zero - mean gaussian random variables with variance @xmath39 .
if @xmath1 is greater than @xmath27 , we can see that the results of ge1 closely match with the performance bounds ( i.e. , pb@xmath40 ) because there is no issue of noise correlation ; for example , when @xmath1 is 10 , the results of ge1 match with the performance bounds for @xmath0 up to 20 .
compared to the results for ge1 , the results for ge2 of a fictitious model show that they can attain the performance bounds irrespective of the value of @xmath1 because there is no issue of noise correlation at all .
it is interesting , though , that the results of ge1 for @xmath41 show even better performance than the performance bounds . with the valid range of @xmath1 given by , the selection of the optimal @xmath1 given in eqs .
( 32 ) and ( 33 ) of @xcite should be modified accordingly . because @xmath42 in eq .
( 32 ) of @xcite is concave downward for the whole range of real - valued @xmath43 $ ] , @xmath44 in eq .
( 33 ) of @xcite is now simplified as follows : @xmath45 k .- l .
noh , q. m. chaudhari , e. serpedin , and b. w. suter , `` novel clock phase offset and skew estimation using two - way timing message exchanges for wireless sensor networks , '' _ ieee trans .
_ , vol .
55 , no . 4 ,
pp . 766777 , apr . | the generalization of the maximum - likelihood - like estimator for clock skew by leng and wu in the above paper is erroneous because the correlation of the noise components in the model is not taken into account in the derivation of the maximum likelihood estimator , its performance bound , and the optimal selection of the gap between two subtracting time stamps .
this comment investigates the issue of noise correlation in the model and provides the range of the gap for which the maximum likelihood estimator and its performance bound are valid and corrects the optimal selection of the gap based on the provided range .
clock synchronization , two - way message exchanges , maximum likelihood estimation . |
considerable observational evidence now exists for supermassive black holes ( smbhs ) in the centers of giant elliptical galaxies .
these mass concentrations , long suspected to be the origin of non - thermal energy in active galaxies , are now thought to be responsible for large gas or stellar velocities in galactic cores ( ho 1998 ) . we discuss here the interaction of smbhs with hot ( @xmath3 k ) interstellar gas . the local compression and heating of interstellar gas in the smbh potential
can be detected with the _ chandra _ x - ray telescope .
indeed , the inflow of interstellar gas may explain the growth of smbhs in galactic cores
. however , x - ray detection of smbhs may not be possible if the non - thermal interstellar pressure is comparable to the pressure of hot thermal gas .
as we were completing this work , a similar study of the smbh - gas interaction by quataert & narayan ( 1999 ) ( qn ) appeared on the astro - ph preprint server .
while some of the conclusions of qn are supported by our flow calculations , our results are qualitatively different in a number of important ways .
the steady state flows considered by qn are constrained to pass through a sonic point , becoming supersonic near the smbh . by comparison , our ( non - steady ) computational gas flows near the smbh adapt naturally to the interstellar cooling flow in the surrounding galaxy and , as a result , our flow velocity is fully subsonic near the smbh .
we also consider several additional physical effects not discussed by qn : time - dependent ejection of mass by evolving stars , heating by type ia supernovae , non - thermal pressure , and the gravitational field of ( dark ) cooled interstellar gas which in some observations may be confused with the mass of the smbh .
although we use observations of the bright virgo elliptical ngc 4472 as a guide for our models , we do not seek here a set of parameters for an optimal fit to this galaxy .
instead , our objective is to illustrate some of the physical effects that must be considered in interpreting x - ray observations of the cores of bright ellipticals with the _ chandra _ telescope . at a distance
@xmath4 mpc appropriate for ngc 4472 , _
chandra s _ resolution ( @xmath5 arcseconds ) is @xmath6 pc .
the computational procedure we follow here is identical to that described in our recent papers ( e.g. brighenti & mathews 1999a ) . with ngc 4472 in mind
, we construct a galaxy in a cosmological environment beginning with an overdensity perturbation designed to form a galaxy group . stars of total mass @xmath7 @xmath8 ( appropriate to ngc 4472 with @xmath9 ; van der marel 1991 ) , are formed from the baryons at an early time ( @xmath10 gyr ) and later ( @xmath11 gyrs ) merge into a de vaucouleurs profile .
an evolving nfw dark halo develops around the galaxy over time , accompanied by secondary infall , although within an effective radius ( @xmath12 kpc for ngc 4472 ) stellar mass loss provides about half of the hot interstellar gas .
a cooling flow develops in the hot interstellar gas at early times and slowly evolves toward its present configuration at @xmath13 gyrs . over time a large mass of interstellar gas cools in ngc 4472 , @xmath14 @xmath8 . since this mass exceeds the mass observed in the central black hole in ngc 4472 , @xmath15 @xmath8 ( magorrian et al .
1998 ) , interstellar gas must cool in localized thermal instabilities distributed throughout an extended galactic volume within @xmath16 .
this distributed interstellar cooling is modeled by adding a sink term to the equation of continuity : @xmath17 where @xmath18 is the time for gas to cool locally by radiative losses at constant pressure ( e.g. sarazin & ashe 1989 ) .
we find that taking @xmath19 provides a satisfactory global interstellar flow solution for ngc 4472 that matches currently observed density , temperature and metallicity profiles reasonably well ( brighenti & mathews 1999a ; 1999b ) ; we adopt @xmath19 here . to resolve the small scale flow near the smbh we have refined our logarithmic spatial grid in the galactic core to an innermost zone of radius 12 pc .
interstellar gas is influenced by the black hole within a radius @xmath20 pc ; @xmath21 is the sound speed in the hot gas .
we ignore the effects of galactic rotation .
although the mass - losing stellar systems in large ellipticals like ngc 4472 are slowly rotating , their x - ray images do not indicate the flattening expected if the hot interstellar gas shared this rotation ( brighenti & mathews 1996 ) .
evidently , efficient cooling dropout or turbulent viscosity removes angular momentum from the hot gas .
figure 1 shows the interstellar density , temperature and ( rosat band ) x - ray surface brightness evolved to the present time , @xmath13 gyrs . during the evolution , new gas and energy
are supplied from stellar sources : ( 1 ) the specific stellar mass loss rate is @xmath22^{-1.3}$ ] sec@xmath23 and ( 2 ) gas is heated by type ia supernovae at a rate snu@xmath24snu@xmath25 , where snu@xmath26 snia per 100 yrs per @xmath27 @xmath28 ( cappellaro et al .
the interstellar gas temperature is nearly virial , @xmath29 k. the stellar density has a de vaucouleurs profile but flattens to @xmath30 within a `` break '' radius @xmath31 pc ( faber et al .
1997 ) .
dashed lines in figure 1 refer to solutions without central black hole and the solid lines refer to solutions with a central black hole of mass @xmath32 @xmath8 introduced at @xmath11 gyrs . in this @xmath19 solution , an additional @xmath33 @xmath8 flows inside @xmath34 pc by @xmath13 gyrs , increasing the central dark mass to @xmath35 @xmath8 . the gas density ( with smbh )
varies approximately as @xmath36 over most of the flow . for @xmath37 pc , the computed density appears to exceed the innermost observations ; we discuss this apparent discrepancy in more detail below .
light curves in figure 1 for @xmath38 refer to the background flow ; heavy curves show the projected temperature including emission from distributed cooling regions and correspond to the temperature observed . for constant @xmath39 , cooling dropout is important throughout the flow and extends to the region adjacent to the smbh .
the gas temperature rises dramatically within about 50 pc from the smbh detection of this thermal peak by _
chandra _ would provide additional evidence for a smbh .
this temperature spike should be visible even when viewed with _
chandra _ in projection against the entire cooling flow : the mean projected thermal temperatures within apertures of @xmath40 , 100 and 500 pc are 2.95 , 2.70 and 2.00 @xmath41 k ( with smbh ) and 1.04 , 1.05 and 1.09 @xmath41 k ( without smbh ) respectively .
the upper panel in figure 2 shows the current flow velocity @xmath42 and the local mass flow @xmath43 . a small influence due to the smbh potential at times
@xmath44 can be seen in the distant flow @xmath45 kpc where the flow time @xmath46 .
overall , the flow is very subsonic even in the vicinity of the smbh where @xmath47 km s@xmath23 .
this occurs for two reasons : ( 1 ) the high pressure in the central thermal spike ( e.g. holzer & axford 1970 ) , enhanced by the smbh , and ( 2 ) mass dropout which tends to make subsonic flow even slower . by contrast , the solutions of qn ( with @xmath48 ) are constrained to pass through a sonic transition , causing the gas to accelerate through the thermal spike .
[ without mass dropout ( @xmath49 ) we also find a transition to supersonic flow at @xmath50 pc . ]
our calculations shown in figures 1 and 2 were done with boundary condition @xmath51 ; accumulated gas in the central zone cools there . to encourage the onset of supersonic flow near the smbh
, we repeated the calculation allowing the gas to flow freely through a small radius ( as small as @xmath52 pc with a very high resolution grid ) and found that the flow was essentially unchanged from that shown in the figures .
the fully subsonic character of our solutions is therefore quite robust and independent of our boundary condition at @xmath53 .
[ in principle our flow should become supersonic very near the schwarzschild radius ( @xmath54 pc ) , but this can not be resolved by our spatial grid and it is likely that additional physical effects become important as @xmath55 approaches @xmath56 . ] because our flow in the thermal spike near the smbh is subsonic , the cooling time @xmath57 remains comparable to the flow time @xmath58 throughout this region and cooling dropout continues unabated at small radii , unlike the qn solutions .
this concentrated mass dropout has two important astrophysical consequences : ( 1 ) the mass deposited by interstellar cooling near the smbh may be comparable to that of the smbh itself , and ( 2 ) the rate that mass flows into the smbh at the origin is significantly reduced ( by about @xmath59 ) below the qn solutions .
the mass of cooled interstellar gas in our model with smbh in figure 2 is large : @xmath60 , @xmath61 and @xmath62 @xmath8 within @xmath63 , 100 and 300 parsecs respectively .
these masses are comparable to smbh masses found by magorrian et al .
( 1998 ) and they occur within the small region ( @xmath64 arcseconds or 330 pc for ngc 4472 ) where stellar velocities are enhanced by the smbh . the gravitational acceleration due to the smbh , the dropout mass , old stars and dark halo are also shown in figure 2 ;
note that the dropout mass dominates gravitational forces at @xmath65 pc . in view of the possibility of contamination with dropout mass ,
the true mass of the central smbh may be lower than those quoted by magorrian et al .
these considerations depend on our assumption that interstellar mass dropout continues to the smallest galactic radii ; at the present time there is no reason to question this assumption . in figure 3
we illustrate a solution identical to that in figure 2 but with an initial smbh mass of only @xmath66 @xmath8 ; the visibility of the thermal spike is noticeably reduced .
we now return to the apparent disagreement in figure 1 between our models and the innermost observed @xmath67 and deprojected electron density
. can the apparent flattening of observed @xmath67 at small radii be due to absorption of x - rays by cold gas in @xmath37 pc ?
the hii gas that produces optical emission lines in ngc 4472 ( macchetto et al . 1996 ) can not account for the x - ray absorption since h@xmath68 emission from such hii clouds in pressure equilibrium would be @xmath69 times more intense than @xmath70 observed in this region of ngc 4472 .
suppose instead that the absorbing gas is cold hi or h@xmath71 and is distributed in a uniform disk extending to @xmath72 pc oriented perpendicular to the line of sight . if the disk column density is @xmath73 @xmath74 ( i.e. @xmath75 ) , the total disk mass , @xmath76 @xmath8 , would be consistent with upper limits on observed hi and h@xmath71 in ngc 4472 , @xmath77 @xmath8 ( bregman , roberts & giovanelli 1988 ; braine , henkel & wiklind 1997 ) . however , in pressure equilibrium with the hot cooling flow gas , the density of such cold gas ( at @xmath78 k mathews & brighenti 1999 ) is very high , @xmath79 @xmath80 , and the disk thickness is incredibly small , @xmath81 pc .
whether such a peculiar and fragile disk could exist in ngc 4472 ( and other virgo ellipticals with similar cores ; brighenti & mathews 1997 ) seems unlikely . however , optical line emission from the surface of the disk ionized by stellar uv ( mathews & brighenti 1999 ) would exceed observed limits . for hii gas at @xmath82
k the pressure equilibrium density at @xmath83 pc is @xmath84 @xmath80 and the stromgren radius is @xmath85 cm .
the total h@xmath68 energy emitted by the disk , @xmath86 ergs s@xmath23 , is @xmath87 times larger than the total h@xmath68 luminosity from ngc 4472 , @xmath88 ergs s@xmath23 .
this inconsistency can not be removed with other geometrical arrangements of the absorbing cold gas within 300 pc , so the apparent inner flattening in figure 1 is probably not due to absorption .
we have also explored the possibility that the @xmath67 observations shown in figure 1 were centrally flattened by the 4 arcsecond half power gaussian prf ( giaconni et al .
1979 ) of the _ einstein _ hri .
however , such a convolution of our computed @xmath67 can not duplicate the core observed in ngc 4472 .
x - ray cores in ngc 4472 , ngc 4636 and ngc 4649 ( trincheri , fabbiano & canizares 1986 ) are not artifacts of limited observational resolution .
these difficulties led us to explore two additional `` non - standard '' models for ngc 4472 : flows ( 1 ) with extreme mass dropout inside 300 pc or ( 2 ) with additional non - thermal pressure in this same region .
it is often claimed that mass dropout reduces the x - ray surface brightness in the central regions of cooling flows , but this is not true in general . for flows
in which most of the gas in the central region has advected inward from large radii , cooling dropout at large radii can reduce the amount of gas and associated emission @xmath67 in the central core .
however , this does not apply to galactic cooling flows in which much of the gas in the central flow originates from stars within or near this region .
in addition , the x - ray emissivity and @xmath67 are locally enhanced at ( denser ) cooling sites where mass dropout occurs . to illustrate the futility of lowering the central @xmath67 in ngc 4472 with increased local cooling
, we show in figure 3 density and temperature profiles for the standard flow ( at @xmath89 and with smbh ) but with a cooling dropout that truncates the background flow density : @xmath90 with @xmath91 @xmath80 .
the thermal spike around the smbh is still present in this solution , but the apparent @xmath92 ( and @xmath67 ) within several 100 pc is only slightly reduced .
the mean projected temperature within 100 pc is @xmath93 k. finally , we consider cooling flow models in which @xmath92 and @xmath67 in the inner flow are lowered by an additional non - thermal pressure due to relativistic particles or magnetic fields , as suggested by radio emission observed in most elliptical cores . to explore this possibility
we show in figure 3 the standard flow ( at @xmath13 gyrs with smbh ) in which an additional pressure is introduced in the galactic core : @xmath94 , where @xmath95 is the thermal gas pressure .
the density ( and @xmath96 ) in the cooling flow within @xmath2 pc are reduced by this additional pressure and parameters for @xmath97 could be adjusted in an _ ad hoc _ fashion to achieve a more perfect fit with the observations .
however , the apparent temperature of the thermal gas is dramatically lowered since the external flow is now supported mostly by non - thermal pressure .
the projected apparent temperature within @xmath98 pc is only @xmath99 k , assuming this gas is not heated by collective interactions with the relativistic gas .
the discernibility of the thermal spike near the smbh is now greatly reduced .
in this study we have examined central cooling flows in a luminous elliptical similar to ngc 4472 . we have emphasized the evolution of quiescent flows that arise naturally from gas and energy supplied by galactic stars and by secondary infall in the outer regions .
we have not explored possible transient interactions between the cooling flow and an intermittently active galactic core ( e.g. ciotti & ostriker 1997 ) .
our objective here has been to anticipate some of the interpretive issues that may arise when _
observations become available .
\(1 ) some dark matter currently attributed to the central supermassive black hole ( smbh ) may be due to a population of non - luminous stars created from cooled interstellar gas having a bottom - heavy imf . in the standard cooling flow for ngc 4472 with smbh , neutral clouds become gravitationally unstable ( and truncate the imf ) at masses @xmath100 , 0.3 and 0.8 @xmath8 at @xmath63 , 100 and 300
pc respectively ( mathews & brighenti 1999 ) . such dark stars , possibly having approximately radial orbits , can also produce a radial variation in the global baryonic @xmath101 .
\(2 ) a successful observation of a high temperature peak in @xmath1 pc by _ chandra _ in a nearby bright elliptical would provide additional strong evidence for the existence of smbhs .
however , if additional non - thermal pressure is present in elliptical cores the smbh thermal peak may no longer be discernible .
\(3 ) our central flows ( @xmath37 pc ) differ from the bondi - type transonic flow described by quataert & narayan ( 1999 ) since we do not insist on a sonic transition .
the mass flow rate into the smbh is considerably reduced in our models .
\(4 ) as long as our standard cooling flow assumptions hold , we expect the gas density to rise steadily toward the smbh approximately as @xmath36 .
but current observations suggest that such a gas density peak is not present .
about half of the gas within 300 pc in ngc 4472 is supplied by mass loss from stars in this region . in our standard cooling flow model
we have assumed that all stellar ejecta rapidly enters the hot interstellar phase ( in @xmath102 yrs ) .
it has occasionally been suggested ( e.g. thomas 1986 ) that stellar ejecta cools before entering the hot gas , effectively reducing @xmath103 .
however , we find that the central peak in @xmath92 and @xmath67 persists even if the stellar mass loss rate @xmath103 is reduced at small @xmath55 .
moreover , dynamical arguments and gas temperature and metallicity gradients observed at larger radii ( @xmath104 ) strongly indicate that cooler , enriched stellar ejecta is conductively melting into the hot phase .
studies of the evolution of hot gas in elliptical galaxies at uc santa cruz are supported by nasa grant nag 5 - 3060 and nsf grant ast-9802994 for which we are very grateful .
fb is supported in part by grant murst - cofin 98 .
braine , j. , henkel , c. & wiklind , t. 1997 , a&a 321 , 765 bregman , j. n. , roberts , m. s. & giovanelli , r. 1988 , apj 330 , l93 brighenti , f. & mathews , w. g. 1996 , apj 470 , 747 brighenti , f. & mathews , w. g. 1997 , apj 486 , l83 brighenti , f. & mathews , w. g. 1999a , apj 512 .
65 brighenti , f. & mathews , w. g. 1998b , apj ( in press ) cappellaro , e. , turatto , m , tsvetkov , d. yu . , bartunov , o. s. , pollas , c. , evans , r. & hamuy , m. 1997 , 1997 , astron & ap .
322 , 431 ciotti , l. & ostriker , j. p. 1997 , apj 487 , l105 faber , s. m. et al . 1997 , aj , 114 , 1771 giacconi , r. et al . 1979 ,
apj , 230 , 540 ho , l. c. 1999 , in _ observational evidence for black holes in the universe _ , ed . s. k. chakrabarti ( dordrecht : kluwer ) , p 000 holzer , t. e. & axford , w. i. 1970 , ara&a 8 , 31 irwin , j. a. & sarazin , c. l. 1996 , apj 471 , 683 macchetto , f. , et al .
1996 , a&as 120 , 463 magorrian , j. , et al .
1998 , apj 115 , 2285 mathews , w. g. & brighenti , f. 1999 , apj ( in press ) ( astro - ph/9907364 ) quataert , e. & narayan , r. 1999 , apj ( in press ) ( astro - ph/9908199 ) sarazin , c. l. & ashe , g. a. 1989 , apj , 345 , 22 thomas , p. a. 1986 , mnras 220 , 949 trinchieri , g. , fabbiano , g. , & canizares , c. r. 1986 , apj , 310 , 637 van der marel , r. p. 1991 , mnras , 253 , 710 | anticipating forthcoming _
chandra _ x - ray observations , we describe the continuation of interstellar cooling flows deep into the cores of elliptical galaxies .
interstellar gas heated to @xmath0 kev in the potential of massive black holes ( @xmath1 pc ) should be visible unless thermal heating is diluted by non - thermal pressure .
since our flows are subsonic near the massive black holes , distributed cooling continues within @xmath2 pc .
dark , low mass stars formed in this region may be responsible for some of the mass attributed to central black holes .
.2 in |
for large classes of dynamical systems with good mixing properties , it is possible to obtain strong statistical limit laws such as the central limit theorems and its refinements including the almost sure invariance principle ( asip ) @xcite .
an immediate consequence of the asip is the weak invariance principle ( wip ) which is the focus of this paper .
thus the standard wip ( weak convergence to brownian motion ) holds for general axiom a diffeomorphisms and flows , and also for nonuniformly hyperbolic maps and flows modelled by young towers @xcite with square integrable return time function ( including hnon - like attractors @xcite , finite horizon lorentz gases @xcite , and the lorenz attractor @xcite ) .
recently , there has been interest in statistical limit laws for dynamical systems with weaker mixing properties such as those modelled by a young tower where the return time function is not square integrable . in the borderline case where the return time lies in @xmath2 for all @xmath3 ,
it is often possible to prove a central limit theorem with nonstandard norming ( nonstandard domain of attraction of the normal distribution ) .
this includes important examples such as the infinite horizon lorentz gas @xcite , the bunimovich stadium @xcite and billiards with cusps @xcite . in such cases , it is also possible to obtain the corresponding wip ( see for example @xcite ) . for young towers with non - square - integrable return time function , the central limit theorem generally fails .
gouzel @xcite ( see also zweimller @xcite ) obtained definitive results on convergence in distribution to stable laws .
the only available results on the corresponding wip are due to tyran - kamiska @xcite who gives necessary and sufficient conditions for weak convergence to the appropriate stable lvy process in the standard skorohod @xmath0 topology @xcite .
however in the situations we are interested in , the @xmath0 topology is too strong and the results in @xcite prove that weak convergence fails in this topology . in this paper
, we repair the situation by working with the @xmath1 topology ( also introduced by skorohod @xcite ) .
in particular , we give general conditions for systems modelled by a young tower , whereby convergence in distribution to a stable law can be improved to weak convergence in the @xmath1 topology to the corresponding lvy process .
the proof is by inducing ( see @xcite for proofs by inducing of convergence in distribution ) .
young towers by definition have a good inducing system , namely a gibbs - markov map ( a markov map with bounded distortion and big images @xcite ) .
the results of tyran - kamiska @xcite often apply positively for such induced maps ( see for example the proof of theorem [ thm - const ] below ) and yield weak convergence in the @xmath0 topology , and hence the @xmath1 topology , for the induced system .
the main theoretical result of the present paper discusses how @xmath1 convergence in an induced system lifts to the original system ( even when convergence in the @xmath0 topology does not lift ) . as a special case ,
we recover the aforementioned results @xcite on the wip in the nonstandard domain of attraction of the normal distribution . in the remainder of the introduction ,
we describe how our results apply to pomeau - manneville intermittency maps @xcite .
in particular , we consider the family of maps @xmath4 , @xmath5 $ ] , studied by @xcite , given by @xmath6 \\ 2x-1 , & x\in(\frac12,1 ] \end{cases}\end{aligned}\ ] ] for @xmath7 , there is a unique absolutely continuous ergodic invariant probability measure @xmath8 .
suppose that @xmath9 is a hlder observable with @xmath10 .
let @xmath11 .
for the map in , our main result implies the following : [ thm - lsv ] let @xmath12\to[0,1]$ ] be the map with @xmath13 and set @xmath14 .
let @xmath15\to{{\mathbb r}}$ ] be a mean zero hlder observable and suppose that @xmath16 .
define @xmath17 .
then @xmath18 converges weakly in the skorohod @xmath1 topology to an @xmath19-stable lvy process .
( the specific lvy process is described below . )
[ rmk - jm ] the @xmath0 and @xmath1 topologies are reviewed in section [ sec - jm ] .
roughly speaking , the difference is that the @xmath1 topology allows numerous small jumps for @xmath18 to accumulate into a large jump for @xmath20 , whereas the @xmath0 topology would require a large jump for @xmath20 to be approximated by a single large jump for @xmath18 .
since the jumps in @xmath18 are bounded by @xmath21 , it is evident that in theorem [ thm - lsv ] convergence can not hold in the @xmath0 topology .
situations in the probability theory literature where convergence holds in the @xmath1 topology but not the @xmath0 topology include @xcite .
theorem [ thm - lsv ] completes the study of weak convergence for the intermittency map with @xmath7 and typical hlder observables .
we recall the previous results in this direction .
if @xmath22 then it is well - known that @xmath23 satisfies a central limit theorem , so @xmath24 converges in distribution to a normal distribution with mean zero and variance @xmath25 , where @xmath25 is typically positive .
moreover , @xcite proved the asip .
an immediate consequence is the wip : @xmath26 converges weakly to brownian motion . if @xmath27 and @xmath16 , then gouzel @xcite proved that @xmath23 is in the nonstandard domain of attraction of the normal distribution : @xmath28 converges in distribution to a normal distribution with mean zero and variance @xmath29 .
dedecker & merlevede @xcite obtained the corresponding wip in this situation ( with @xmath30 ) .
finally , if @xmath13 and @xmath16 , then gouzel @xcite proved that @xmath31 converges in distribution to a one - sided stable law @xmath32 with exponent @xmath33 .
the stable law in question has characteristic function @xmath34 where @xmath35 and @xmath36 is the invariant density .
let @xmath37 denote the corresponding @xmath19-stable lvy process ( so @xmath38 has independent and stationary increments with cadlag sample paths and @xmath39 ) .
tyran - kamiska @xcite verified that @xmath17 does not converge weakly to @xmath20 in the @xmath0 topology .
in contrast , theorem [ thm - lsv ] shows that @xmath18 converges weakly to @xmath20 in the @xmath1 topology .
the remainder of this paper is organised as follows .
in section [ sec - therewasanoldmanfromjapan_whowroteversethatneverwouldscan_whenaskedwhyitwas_hesaiditsbecause_iliketogetasmanywordsintothelastlineasipossiblycan ] we state our main abstract result , theorem [ t_inducingwip ] , on inducing the wip . in section [ sec - proof ]
we prove theorem [ t_inducingwip ] . in section [ sec - ex ]
we consider some examples which include theorem [ thm - lsv ] as a special case .
in this section , we formulate our main abstract result theorem [ t_inducingwip ] .
the result is stated in subsection [ sec_theinducingthm ] after some preliminaries in subsection [ sec - jm ] .
* distributional convergence .
* to fix notations , let @xmath40 be a probability space and @xmath41 a sequence of borel measurable maps @xmath42 , where @xmath43 is a separable metric space
. then distributional convergence of @xmath44 w.r.t.@xmath45 to some random element @xmath46 of @xmath47 will be denoted by @xmath48 . _ _ strong distributional convergence _ _ @xmath49 on a measure space @xmath50 means that @xmath51 for all probability measures @xmath52 . * skorohod spaces .
* we briefly review the required background material on the skorohod @xmath53 and @xmath54 topologies @xcite on the linear spaces @xmath55 $ ] , @xmath56 $ ] , and @xmath57 of real - valued _ cadlag _ functions ( right - continuous @xmath58 with left - hand limits @xmath59 ) on the respective interval , referring to @xcite for proofs and further information .
both topologies are polish , with @xmath53 stronger than @xmath54 .
it is customary to first deal with bounded time intervals .
we thus fix some @xmath60 and focus on @xmath61 $ ] .
( everything carries over to @xmath56 $ ] in an obvious fashion . ) throughout , @xmath62 will denote the uniform norm .
two functions @xmath63 are close in the @xmath53__-topology _ _ if they are uniformly close after a small distortion of the domain .
formally , let @xmath64 be the set of increasing homeomorphisms @xmath65\rightarrow \lbrack0,t]$ ] , and let @xmath66 denote the identity
. then @xmath67 defines a metric on @xmath68 which induces the @xmath53-topology .
while its restriction to @xmath69 $ ] coincides with the uniform topology , discontinuous functions are @xmath53-close to each other if they have jumps of similar size at similar positions . in contrast , the @xmath54__-topology _ _ allows a function @xmath70 with a jump at @xmath71 to be approximated arbitrarily well by some continuous @xmath72 ( with large slope near @xmath71 ) . for convenience , we let @xmath73 $ ] denote the ( possibly degenerate ) closed interval with endpoints @xmath74 , irrespective of their order .
let @xmath75\times \mathbb{r}:x\in\lbrack g(t^{-}),g(t)]\}$ ] denote the completed graph of @xmath76 , and let @xmath77 be the set of all its parametrizations , that is , all continuous @xmath78\rightarrow\gamma(g)$ ] such that @xmath79 implies either @xmath80 or @xmath81 plus @xmath82 .
then @xmath83 gives a metric inducing @xmath54 . on the space @xmath57 the @xmath84-topology , @xmath85 , is defined by the metric @xmath86 .
convergence @xmath87 in @xmath88 means that @xmath89 for every continuity point @xmath90 of @xmath76 .
for either topology , the corresponding borel @xmath91-field @xmath92 on @xmath68 , generated by the @xmath93-open sets , coincides with the usual @xmath91-field @xmath94 generated by the canonical projections @xmath95 .
therefore , any family @xmath96}$ ] or @xmath97 of real random variables @xmath98 such that each path @xmath99 is cadlag , can be regarded as a random element of @xmath68 , equipped with @xmath100 or @xmath54 .
recall that for any ergodic measure preserving transformation ( _ m.p.t .
_ ) @xmath101 on a probability space @xmath50 , and any @xmath102 with @xmath103 , the _ return time _
function @xmath104 given by @xmath105 is integrable with mean @xmath106 ( kac formula ) , where @xmath107 .
moreover , the _ first return map _ or _ induced map _
@xmath108 is an ergodic m.p.t . on the probability space @xmath109
this is widely used as a tool in the study of complicated systems , where @xmath110 is chosen in such a way that @xmath111 is more convenient than @xmath101 . in particular , given an _
( i.e. a measurable function ) @xmath112 , it may be easier to first consider its _ induced version _
@xmath113 on @xmath110 , given by @xmath114 . by standard arguments , if @xmath115 then @xmath116 and @xmath117 . in this setup
, we will denote the corresponding ergodic sums by @xmath118 and @xmath119 , respectively .
our core result allows us to pass from a weak invariance principle for the induced version to one for the original observable .
such a step requires some a priori control of the behaviour of ergodic sums @xmath120 during an excursion from @xmath110 .
we shall express this in terms of the function @xmath121 $ ] given by @xmath122 note that @xmath123 vanishes if and only if the ergodic sums @xmath120 grow monotonically ( nonincreasing or nondecreasing ) during each excursion . hence bounding @xmath123 means limiting the growth of @xmath124 until the first return to @xmath110 in at least one direction .
the expression @xmath125 can be understood also in terms of the maximal and minimal processes @xmath126 defined during each excursion @xmath127 by @xmath128 [ prop - predominant ] ( i ) in the `` predominantly increasing '' case @xmath129 , we have @xmath130 .
\(ii ) in the `` predominantly decreasing '' case @xmath131 , we have @xmath132 .
this is immediate from the definition of @xmath133 and @xmath134 .
we use @xmath123 to impose a weak monotonicity condition for @xmath124 during excursions .
[ * inducing a weak invariance principle*][t_inducingwip]let @xmath101 be an ergodic m.p.t . on the probability space @xmath50 , and let @xmath102 be a subset of positive measure with return time @xmath135 and first return map @xmath111 .
suppose that the observable @xmath136 is such that its induced version @xmath137 satisfies a wip on @xmath109 in that @xmath138 where @xmath139 is regularly varying of index @xmath140 , and @xmath141 is a process with cadlag paths .
moreover , assume that @xmath142 then @xmath23 satisfies a wip on @xmath50 in that @xmath143
if the process @xmath20 in for the induced system is an @xmath19-stable lvy process , then the limiting process in is @xmath144 . in general , the convergence from ( [ eq_inducedwipinm1 ] ) fails in @xmath145 , even if ( [ eq_wipinduced ] ) holds in the @xmath53-topology . that this is the case for the intermittent maps
was pointed out in ( * ? ? ?
* example 2.1 ) .
if the process @xmath20 in for the induced system has continuous sample paths , then the statement and proof of theorem [ t_inducingwip ] is greatly simplified and the uniform topology ( corresponding to the uniform norm @xmath146 ) can be used throughout . in particular , the function @xmath123 is replaced by @xmath147 . in the case of normal diffusion @xmath148 , condition is then satisfied if @xmath149 . a simplified proof based on the one presented here is written out in ( * ?
* appendix ) . in the applications that we principally have in mind ( including the maps ) , the observable @xmath9 is integrable , and hence so is its induced version @xmath150 .
in particular , if @xmath23 has mean zero , then @xmath137 has mean zero and we are in a situation to apply theorem [ t_inducingwip ] . from this
, it follows easily that if condition holds with @xmath151 and condition holds with @xmath23 replaced throughout by @xmath152 in the definition of @xmath123 , then conclusion is valid with @xmath153 with a little more effort it is also possible to handle more general centering sequences where the process @xmath154 in condition takes the form @xmath155 for real sequences @xmath156 , @xmath157 with @xmath158 .
the monotonicity condition ( [ eq_essmonotonicity ] ) will be shown to hold , for example , if we have sufficiently good pointwise control for single excursions : [ * pointwise weak monotonicity*][p_pointwisemonotonicity]let @xmath101 be an ergodic m.p.t .
on the probability space @xmath50 , and let @xmath102 be a subset of positive measure with return time @xmath135 .
let @xmath139 be regularly varying of index @xmath140 .
suppose that for the observable @xmath136 there is some @xmath159 such that for a.e .
@xmath160 , @xmath161 then the weak monotonicity condition ( [ eq_essmonotonicity ] ) holds .
the proofs of theorem [ t_inducingwip ] and proposition [ p_pointwisemonotonicity ] are given in section [ sec - proof ] .
assuming strong distributional convergence @xmath162 in ( [ eq_wipinduced ] ) , rather than @xmath163 , is not a restriction , as an application of the following result to the induced system @xmath164 shows .
[ * automatic strong distributional convergence*][p_automaticstrongdistrcge]let @xmath101 be an ergodic m.p.t . on a @xmath165-finite space @xmath50 .
let @xmath100 or @xmath1 and let @xmath166 be real sequences with @xmath158 , assume that @xmath112 is measurable , and that there is some probability measure @xmath52 and some random element @xmath46 of @xmath57 such that @xmath167 then , @xmath49 in @xmath88 .
this is based on ideas in @xcite .
according to zweimller ( * ? ? ?
* theorem 1 ) , it suffices to check that @xmath168 .
the proof of ( * ? ? ?
* corollary 3 ) shows that @xmath169 alone ( that is , even without ( [ eq_abcdefg ] ) ) implies @xmath170 .
since @xmath171 ( see ( * ? ? ? * theorem 12.3.2 ) ) , the case @xmath172 then is a trivial consequence .
there is a systematic typographical error in @xcite in that the factor @xmath71 in the centering process @xmath173 is missing , but the arguments there work , without any change , for the correct centering .
in this section , we give the proof of theorem [ t_inducingwip ] and also proposition [ p_pointwisemonotonicity ] . throughout
, we assume the setting of theorem [ t_inducingwip ] .
in particular , we suppose that @xmath101 is an ergodic m.p.t . on the probability space @xmath50 , and that @xmath102 is a subset of positive measure with return time @xmath135 and first return map @xmath111 . when @xmath110 is chosen appropriately , many features of @xmath101 are reflected in the behaviour of the ergodic sums @xmath174 , i.e. the times at which orbits return to @xmath110 . these are intimately related to the _ occupation times _ or _ lap numbers _
@xmath175 the visits to @xmath110 counted by the @xmath176 separate the consecutive _ excursions from _ @xmath110 , that is , the intervals @xmath177 , @xmath178 . decomposing the @xmath101-orbit of @xmath179 into these excursions
, we can represent the ergodic sums of @xmath23 as @xmath180 with remainder term @xmath181 encoding the contribution of the incomplete last excursion ( if any ) .
next , decompose the rescaled processes accordingly , writing @xmath182 with @xmath183 , and @xmath184 . on the time scale of @xmath185 ,
the excursions correspond to the intervals @xmath186 , @xmath178 , where @xmath187 is given by @xmath188 .
note that the interval containing a given point @xmath189 is that with @xmath190 . hence @xmath191 in this subsection , we record some consequences of the ergodic theorem which we will use below . but first an elementary observation , the proof of which we omit . [ l_historyofsequence]let @xmath192 be a sequence in @xmath193 such that @xmath194 .
define a sequence of functions @xmath195 by letting @xmath196 .
then , for any @xmath60 , @xmath197 converges to @xmath198 uniformly on @xmath199 $ ] . for the occupation times of @xmath110
, we then obtain : [ * strong law of large numbers for occupation times*][l_lapnumberetc ] the occupation times @xmath176 satisfy * a ) * @xmath200 . *
b ) * moreover , for any @xmath60 , @xmath201}\left\vert k^{-1}n_{\left\lfloor tk\right\rfloor } -t\mu(y)\right\vert \longrightarrow0\text{\quad a.e . on } x\text{\quad as } k\rightarrow\infty\text{.}\ ] ] the first statement is immediate from the ergodic theorem . the second then follows by the preceding lemma .
[ l_returntimecontrol ] for any @xmath60 , @xmath202 . applying the ergodic theorem to @xmath111 and the integrable function @xmath135
, we get @xmath203 , and hence also @xmath204 a.e . on @xmath110 .
the result follows from lemma [ l_historyofsequence ] . * pointwise control of monotonicity behaviour .
* we conclude this subsection by establishing proposition [ p_pointwisemonotonicity ] .
[ * proof of proposition [ p_pointwisemonotonicity ] * ] we may suppose without loss that the sequence @xmath157 is nondecreasing
. since this sequence is regularly varying , @xmath205 for all @xmath206 .
hence for @xmath207 fixed , there are @xmath206 and @xmath208 s.t .
@xmath209 whenever @xmath210 and @xmath211 . as a consequence of lemma [ l_returntimecontrol ] ,
there is some @xmath212 such that @xmath213 satisfies @xmath214 for @xmath215 . in view of ( [ eq_pointwisemonotonicityctrl ] ) we then see ( using monotonicity of @xmath139 again ) that @xmath216 which proves ( [ eq_essmonotonicity ] ) . as a first step towards theorem [ t_inducingwip ] , we prove that switching from @xmath218 to @xmath219 preserves convergence in the skorohod space .
[ * convergence of * @xmath217][l_unctyatt]under the assumptions of theorem [ t_inducingwip ] , @xmath221 for @xmath222 and @xmath223 , we let @xmath224 . since @xmath225
, we have @xmath226 we regard @xmath185 , @xmath227 , @xmath20 , and @xmath228 as random elements of @xmath229 .
note that @xmath230 @xmath231 and @xmath76 non - decreasing@xmath232 .
let @xmath233 denote the constant random element of @xmath68 given by @xmath234 , @xmath235 .
recalling lemma [ l_lapnumberetc ] b ) , we see that for @xmath236-a.e .
@xmath237 we have @xmath238 uniformly on compact subsets of @xmath239 . hence , @xmath240 in @xmath241 holds @xmath236-a.e .
in particular , @xmath242 by assumption ( [ eq_wipinduced ] ) we also have @xmath243 in @xmath241 .
but then we automatically get @xmath244 since the limit @xmath233 of the second component is deterministic .
the composition map @xmath245 , @xmath246 , is continuous at every pair @xmath247 with @xmath248 @xmath231 and @xmath76 strictly increasing and continuous@xmath232 , cf .
* theorem 13.2.3 ) . as the limit @xmath249 in ( [ eq_jointj1cgepnun ] ) satisfies @xmath250 , the standard mapping theorem for distributional convergence ( cf .
* theorem 3.4.3 ) ) applies to @xmath251 , showing that @xmath252 in view of ( [ eq_timechange ] ) , this is what was to be proved . passing from convergence of @xmath253 to convergence of @xmath254 requires a little preparation .
[ lem - m1b ] ( i ) let @xmath255 $ ] and @xmath256
. then @xmath257 } ( g \vert _ { [ t_{j-1},t_j]},g ' \vert _ { [ t_{j-1},t_j]}).\ ] ] ( ii ) let @xmath258 $ ] and @xmath259 .
then @xmath260 } ( g_j,\bar{g}_j ) \leq 2 g_j^{\ast } + ( t_{j } -t_{j-1 } ) , \ ] ] where @xmath261 the first assertion is obvious . to validate the second ,
assume without loss that @xmath262 and that @xmath263 is predominantly increasing in that @xmath264 .
in this case , @xmath265 for the nondecreasing function @xmath266 .
therefore , @xmath267 } ( g_1,g_1^{\uparrow } ) \leq \| g_1 - g_1^{\uparrow } \| = g_1^{\ast}.\ ] ] letting @xmath268 , it is clear that @xmath267 } ( \bar{g}_1^{\uparrow},\bar{g}_1 ) \leq
\| \bar{g}_1^{\uparrow } - \bar{g}_1 \| = \vert g_1^{\uparrow}(t_1 ) - g_1(t_1 ) \vert \leq g_1^{\ast}.\ ] ] finally , we check that @xmath267 } ( g_1^{\uparrow},\bar{g}_1^{\uparrow } ) \leq t_1 -t_0.\ ] ] to this end , we refer to figure [ fig - m1 ] where @xmath269 and @xmath270 represent the completed graphs of @xmath271 and @xmath272 respectively . here
@xmath273 consists of one horizontal line segment followed by one vertical segment .
the picture of @xmath274 is schematic , it may also contain horizontal and vertical line segments .
choose @xmath275 on the graph of @xmath274 that is equidistant from @xmath276 and @xmath277 and let @xmath278 be the point on @xmath277 that is the same height as @xmath275 .
choose parametrizations @xmath279 of @xmath280 , @xmath281 , satisfying * @xmath282 , * @xmath283 , * @xmath284 for all @xmath285 $ ] .
automatically @xmath286 and by construction @xmath287 , as required .
+ as a consequence , we obtain : [ lem - m1 ] @xmath288 .
let @xmath160 and decompose @xmath199 $ ] according to the consecutive excursions , letting @xmath289 , @xmath290 .
consider @xmath291 , @xmath292 $ ] .
if we set @xmath293 } $ ] , then @xmath294 as defined in lemma [ lem - m1b ] coincides with @xmath295 } $ ] , so that @xmath296}(g_j,\bar{g}_j).\ ] ] but @xmath297 , and since @xmath298 , for suitable @xmath299 , we see that lemma [ lem - m1b ] gives @xmath260}(g_j,\bar{g}_j ) \leq ( n^{-1}r + 2 b(n)^{-1}\phi^\ast ) \circ f^j,\ ] ] as required .
* proof of theorem [ t_inducingwip ] .
* fix any @xmath60 . by lemma [ l_returntimecontrol ] , @xmath300 a.e . and by assumption @xmath301 .
hence lemma [ lem - m1 ] guarantees that @xmath302 recall also from lemma [ l_unctyatt ] that @xmath303,\mathcal{m}_{1})\text{.}\label{eq_m1cgewnfinitet}\ ] ] it follows ( see ( * ? ? ?
* theorem 3.1 ) ) from and that @xmath304,\mathcal{m}_{1})\text{.}\end{aligned}\ ] ] this immediately gives @xmath305 in @xmath306 .
strong distributional convergence as asserted in ( [ eq_inducedwipinm1 ] ) follows via proposition [ p_automaticstrongdistrcge ] .
we continue to suppose that @xmath101 is an ergodic m.p.t . on a probability space @xmath50 with first return map @xmath308 where @xmath103 .
suppose further that the induced map @xmath309 is gibbs - markov with ergodic invariant probability measure @xmath310 and partition @xmath311 , and that @xmath312 is constant for each @xmath313 .
let @xmath9 be an @xmath314 mean zero observable , with induced observable @xmath150 .
[ thm - const ] suppose that @xmath23 is constant on @xmath315 for every @xmath313 and @xmath316 .
if @xmath137 lies in the domain of an @xmath19-stable law , then @xmath137 satisfies the wip in @xmath317 with @xmath318 .
if in addition condition holds , then @xmath23 satisfies the wip in @xmath319 . by @xcite
, @xmath320 converges in distribution to the given stable law .
the assumptions guarantee that the induced observable @xmath137 is constant on each @xmath321 .
hence we can apply tyran - kamiska ( * ? ? ?
* corollary 4.1 ) to deduce that @xmath137 satisfies the corresponding @xmath19-stable wip in @xmath317 .
in particular , condition is satisfied .
the final statement follows from theorem [ t_inducingwip ] . for certain examples , including pomeau - manneville intermittency maps
, we can work with general hlder observables , thus improving upon ( * ? ? ?
* example 4.1 ) .
the idea is to decompose the observable @xmath23 into a piecewise constant observable @xmath322 and a hlder observable @xmath323 in such a way that only @xmath322 `` sees '' the source of the anomalous behaviour . in the remainder of this section ,
we carry out this procedure for the maps and thereby prove theorem [ thm - lsv ] .
( lemma [ lem - ly ] and proposition [ prop - gmclt ] below hold in the general context of induced gibbs - markov maps . ) fix @xmath324 and let @xmath325 denote the symbolic metric on @xmath110 , so @xmath326 where @xmath327 is the least integer @xmath328 such that @xmath329 lie in distinct elements of @xmath311 .
an observable @xmath150 is _ piecewise lipschitz _ if @xmath330 for each @xmath313 , and _ lipschitz _ if @xmath331 . the space @xmath332 of lipschitz observables @xmath150 is a banach space .
note that @xmath137 is integrable with @xmath333 if and only if @xmath334 .
suppose first that @xmath111 is weak mixing ( this assumption is removed below ) .
then @xmath336 has no eigenvalues on the unit circle except for the simple eigenvalue at @xmath342 ( corresponding to constant functions ) . by lemma [ lem - ly](a )
, there exists @xmath343 such that the remainder of the spectrum of @xmath335 lies strictly inside the ball of radius @xmath93 .
in particular , there is a constant @xmath344 such that @xmath345 for all @xmath346 , @xmath347 . by lemma [ lem - ly](b ) , @xmath338 .
hence @xmath348 .
following gordin @xcite , write @xmath349
. then @xmath350 ( since @xmath351 and @xmath352 ) .
applying @xmath335 to both sides and noting that @xmath353 , we obtain that @xmath354 .
it follows that the sequence @xmath355 defines a reverse martingale sequence . by burkholder s inequality ( *
* theorem 3.2 ) , @xmath356 by doob s inequality @xcite ( see also ( * ? ? ?
* equation ( 1.4 ) , p. 20 ) , @xmath357 . by markov s inequality , for @xmath358 fixed ,
@xmath359 as @xmath360 . hence @xmath361 . since @xmath137 and @xmath362 differ by a bounded coboundary , @xmath340 as required .
it remains to remove the assumption about eigenvalues ( other than @xmath342 ) for @xmath335 on the unit circle .
suppose that there are @xmath363 such eigenvalues @xmath364 , @xmath365 , @xmath366 ( including multiplicities ) .
then we can write @xmath367 where @xmath368 and @xmath369 .
in particular , the above argument applies to @xmath370 , while @xmath371 , @xmath372 .
we verify the hypotheses of theorem [ t_inducingwip ] . a convenient inducing set for the maps
is @xmath376 $ ] .
let @xmath377 .
( the first term is the important one , and the second term is simply an arbitrary choice that ensures that @xmath322 has mean zero while preserving the piecewise constant requirement in theorem [ thm - const ] . )
write @xmath378 and note that @xmath323 is a mean zero piecewise hlder observable vanishing at @xmath198 .
we have the corresponding decomposition @xmath379 for the induced observables . by theorem [ thm - const ]
, @xmath380 satisfies the wip ( in the @xmath0 topology ) .
let @xmath381 denote the hlder exponent of @xmath23 . by the proof of (
* theorem 1.3 ) , @xmath382 induces to a piecewise lipschitz mean zero observable @xmath383 satisfying @xmath384 for suitably chosen @xmath337 .
moreover @xcite shows that @xmath385 lies in @xmath386 provided that @xmath387 .
exactly the same argument shows that @xmath383 lies in @xmath2 provided @xmath388 .
in particular , for any @xmath389 , there exists @xmath390 such that @xmath391 .
since we are normalising by @xmath392 , it follows from proposition [ prop - gmclt ] that @xmath383 does not contribute to the wip .
it remains to verify condition .
in fact , we show that @xmath125 is bounded .
suppose that @xmath393 ( the case @xmath394 is treated similarly ) .
choose @xmath358 such that @xmath395 on @xmath396 $ ] .
define the decreasing sequence @xmath397 where @xmath398 , @xmath399 , and let @xmath363 be such that @xmath400 for all @xmath401 . then for @xmath160 , @xmath402 $ ] for @xmath403 .
the arguments in the proof of theorem [ thm - lsv ] apply to a much wider class of examples , including intermittent maps with neutral periodic points or with multiple neutral fixed / periodic points . in such cases ,
condition is again automatically satisfied .
the research of i m was supported in part by epsrc grant ep / f031807/1 held at the university of surrey .
we are very grateful to the referee for very helpful suggestions that led to a significantly simplified proof of the main result in this paper .
m. j. field , i. melbourne and a. t " or " ok . decay of correlations , central limit theorems and approximation by brownian motion for compact lie group extensions .
_ ergodic theory dynam .
systems _ * 23 * ( 2003 ) 87110 . | we consider weak invariance principles ( functional limit theorems ) in the domain of a stable law .
a general result is obtained on lifting such limit laws from an induced dynamical system to the original system .
an important class of examples covered by our result are pomeau - manneville intermittency maps , where convergence for the induced system is in the standard skorohod @xmath0 topology . for the full system ,
convergence in the @xmath0 topology fails , but we prove convergence in the @xmath1 topology . *
keywords : * nonuniformly hyperbolic systems , functional limit theorems , lvy processes , induced dynamical systems * msc codes : * 37d25 ; 28d05 , 37a50 , 60f17 |
assume that @xmath2 is a complex separable hilbert space and denote by @xmath5 the algebra of all bounded linear operators on @xmath2 .
we also use the notations @xmath6 and @xmath7 .
the spectrum of @xmath8 is denoted by @xmath9 , while the _ numerical range _ of @xmath1 is defined by @xmath10 in this paper we are only concerned with contractions , that is , operators of norm at most 1 .
an arbitrary contraction can be decomposed as a direct sum of a unitary operator and a completely nonunitary contraction .
the following basic properties of the numerical range of a contraction can be found , for instance , in ( * ? ? ?
[ pr : basic ] let @xmath8 , @xmath11
. then @xmath12 is a convex subset of @xmath13 which satisfies : 1 .
@xmath14 and @xmath15 if @xmath1 is completely nonunitary .
2 . @xmath16 ; 3 .
@xmath17 .
we denote @xmath18 and @xmath19 ; these are called the _ defect operator _ and the _ defect space _ of @xmath1 respectively .
the dimensions of @xmath20 and @xmath21 are called the _ defect indices _ of @xmath1 .
it is well - known that @xmath1 admits unitary dilations ; that is , there exist a space @xmath22 and a unitary operator @xmath23 such that @xmath24 , where @xmath25 denotes the orthogonal projection in @xmath26 onto @xmath2 .
one can always take @xmath26 to be @xmath27 ; however , this is not the optimal choice when the defect spaces of @xmath1 are of equal finite dimension @xmath0 . indeed , in this case there exist unitary dilations acting on @xmath28 , and @xmath0 is the smallest possible value of @xmath29 .
we call such dilations _
unitary @xmath0-dilations_. it is obvious that @xmath30 for any unitary dilation @xmath31 of @xmath1 .
choi and li @xcite showed that , in fact , @xmath32 thus answering a question raised by halmos ( see , for example , @xcite ) .
we note that when @xmath2 is @xmath33-dimensional , the construction in @xcite produces dilations which act on a space of dimension @xmath34 .
assume now that @xmath35 .
we prove in section [ se : unitary dilations ] the stronger result @xmath36 that is , we use only the most `` economical '' unitary dilations of @xmath1 .
this relation was proved earlier in special cases , namely for @xmath37 and @xmath4 in @xcite , for @xmath37 and general @xmath0 in @xcite , and for particular cases with @xmath38 in @xcite and @xcite . the remainder of the paper deals with the special case when @xmath39 .
partly as a consequence of the results in section [ se : unitary dilations ] , we can investigate in detail the geometric properties of @xmath12 .
the study in section [ se : general defect 1 ] does not use the functional model of a contraction , but this model is necessary for the results proved in later sections .
some properties are similar to those proved for finite dimensional contractions , but new phenomena appear for which we provide some illustrative examples .
since the numerical range of a direct sum is the convex hull of the union of the numerical ranges of the summands , and the numerical range of a unitary operator is rather well understood , we always assume in the sequel that @xmath1 is completely nonunitary . also , if @xmath1 has different defect indices , then @xmath40 ; so
we restrict ourselves to studying completely nonunitary contractions with equal defect indices .
let @xmath1 be a completely nonunitary contraction with @xmath41 .
the operator @xmath42 is a partial isometry on @xmath43 and @xmath44 .
both @xmath45 and @xmath46 have dimension @xmath0 , and so any unitary operator @xmath47 determines a unitary @xmath0-dilation @xmath48 of @xmath1 by the formula @xmath49 conversely , any unitary @xmath0-dilation of @xmath1 is unitarily equivalent to some @xmath48 .
the set @xmath50 consists of isolated fredholm eigenvalues of @xmath48 .
[ le : extending partial isometry ] assume that @xmath51 is a hilbert space and @xmath52 is a partial isometry such that @xmath53 .
if @xmath54 is such that @xmath55 is fredholm and @xmath56 , then there exists a partial isometry @xmath57 with @xmath58 , @xmath59 , such that @xmath60 .
observe that the fredholm index of @xmath55 is equal to zero since @xmath61 is in the closure of @xmath62 and so codim@xmath63 .
also , @xmath64 .
if @xmath65 , we pick @xmath66 of norm 1 , and define the desired @xmath57 by @xmath59 , @xmath67 , and @xmath68 for @xmath69 . if @xmath70 , then @xmath71 , whence there exist a nonzero vector @xmath72 and @xmath73 such that @xmath74
write @xmath75 , with @xmath76 and @xmath77 .
we have then @xmath78 or , equivalently , @xmath79 on each side of this equality the two terms are mutually orthogonal . equating the norms and noting that @xmath80 , we obtain @xmath81 , the required @xmath57 is then obtained by adding to the condition @xmath59 the relation @xmath82 and @xmath68 for @xmath83 .
indeed , we have then by @xmath84 since @xmath85 , @xmath86 , and thus @xmath61 is an eigenvalue of @xmath57 with multiplicity @xmath87 .
[ pr : choose eigenvalues ] assume that @xmath88 are distinct points in @xmath89 and @xmath90 are positive integers satisfying @xmath91 .
there exists a unitary @xmath0-dilation @xmath31 of @xmath1 such that @xmath92 is an eigenvalue of @xmath31 with multiplicity @xmath93 , @xmath94 .
consider the partial isometry @xmath95 defined by .
we construct inductively partial isometries @xmath96 such that 1 .
@xmath97 ; 2 .
@xmath98 for @xmath99 ; 3 .
@xmath100 ; 4 .
@xmath101 .
assume that @xmath102 and @xmath103 has been constructed .
if the inequality in ( 4 ) is strict , we choose @xmath104 to be an arbitrary partial isometry satisfying ( 2 ) and ( 3 ) with @xmath105 in place of @xmath33 .
if ( 4 ) is an equality , then there exists @xmath106 such that @xmath107 . since @xmath108 is a finite rank perturbation of the invertible operator @xmath109 , it is fredholm .
lemma [ le : extending partial isometry ] applied to @xmath110 and @xmath111 , produces a partial isometry @xmath57 ; we define @xmath112 .
this completes the induction process ; the operator @xmath113 satisfies the conclusion of the proposition .
proposition [ pr : choose eigenvalues ] is stated without proof in @xcite .
[ le : intersection_num_ranges ] assume that @xmath31 is a unitary @xmath0-dilation of @xmath1 , @xmath114 are eigenvalues of @xmath31 ( not necessarily distinct , but with repeated values being multiple eigenvalues ) .
then the convex hull of @xmath114 contains a point in @xmath12 .
if @xmath115 are corresponding eigenvectors of @xmath31 , then their linear span @xmath116 has dimension @xmath117 , and thus its intersection with @xmath2 contains a vector @xmath118 of norm 1
. then @xmath119 belongs to the numerical range of the restriction of @xmath31 to @xmath116 , which is precisely the convex hull of @xmath114 .
[ th : intersection result ] assume that @xmath1 is a contraction with @xmath41 . 1 .
if @xmath120 is a support line for @xmath121 , then there exists a unitary @xmath0-dilation @xmath31 of @xmath1 such that @xmath120 is a support line for @xmath122 .
2 . @xmath123 .
obviously ( 2 ) follows from ( 1 ) , so we only have to prove ( 1 ) .
assertion ( 1 ) is immediate if @xmath120 is tangent to @xmath124 , so we may assume that @xmath120 intersects @xmath124 in two points .
these two points must belong to the closure of one of the component arcs @xmath125 of @xmath126 .
consider then , for @xmath127 , the line @xmath128 parallel to @xmath120 at a distance @xmath129 and that does not meet @xmath12 .
for @xmath129 sufficiently small , @xmath128 intersects @xmath125 in two points @xmath130 . by proposition [ pr : choose eigenvalues ] there exists a unitary @xmath0-dilation @xmath131 of @xmath1 ( acting on @xmath43 ) which has @xmath132 as an eigenvalue of multiplicity @xmath0 .
we assert that @xmath120 is a support line for @xmath133 . indeed ,
if @xmath31 had an eigenvalue @xmath134 on @xmath135 , we could apply lemma [ le : intersection_num_ranges ] to the set formed by @xmath61 repeated @xmath0 times together with @xmath136 to obtain an element in @xmath12 on the segment @xmath137 $ ] which is at distance @xmath129 from @xmath12 . finally , noting that the set of unitary @xmath0-dilations of @xmath1 on @xmath43 is compact , a limit point of the unitary dilations @xmath131 will itself have @xmath120 as a support line .
note that based on the sz .- nagy foias functional model @xcite one can give a function theoretical description of all unitary @xmath0-dilations of @xmath1 ; this description , originating in @xcite , is actually used in @xcite .
as seen above , it appears that direct arguments are simpler and more transparent .
the functional model does however offer some advantages which we start to exploit in section [ se : functional model ] .
the remainder of the paper deals with the particular case when @xmath39 and @xmath1 is completely nonunitary ; to avoid trivialities , we will assume in the sequel that @xmath138 .
the first condition is of course the simplest case of finite and equal defect numbers , and this allows us to obtain several geometrical properties of @xmath12 , extending results obtained for finite dimensional matrices in @xcite and for certain infinite dimensional operators in @xcite .
recall that a point of a convex set @xmath139 in the plane is said to be _ exposed _ if there is a support line for @xmath139 that intersects it only in that point ; such a line is called an _ exposing _ line .
a _ corner _ is an exposed point for which the exposing line is not unique .
if @xmath1 is a completely nonunitary contraction , then all points in @xmath140 are obviously exposed .
we describe next some other properties of the boundary @xmath141 .
[ th : exposed and corners ] let @xmath8 be a completely nonunitary contraction with @xmath142 . 1 .
if @xmath143 is an exposed point of @xmath121 , then @xmath144 .
any eigenvalue of @xmath1 is in the interior of @xmath12 .
@xmath121 has no corners in @xmath145 .
\(1 ) assume that @xmath120 is an exposing line for @xmath146 .
if @xmath147 then @xmath148 , since otherwise we would have @xmath149\subset\overline{w(t)}$ ] .
thus @xmath120 intersects @xmath124 in two points @xmath150 , neither of which belongs to @xmath9 .
theorem [ th : intersection result ] yields a unitary 1-dilation @xmath31 of @xmath1 whose numerical range also has @xmath120 as a support line . then @xmath151 and @xmath152 are eigenvalues of @xmath31 , and lemma [ le : intersection_num_ranges ] implies that @xmath144 .
\(2 ) it was shown in @xcite that if an eigenvalue @xmath151 is in the boundary of the numerical range , then the corresponding eigenspace reduces @xmath1 .
thus @xmath153 , and @xmath154 .
since , similarly , @xmath155 , @xmath156 is both unitary and completely nonunitary , which is possible only when @xmath156 acts on the trivial space @xmath157 .
this contradicts the assumption @xmath138 .
\(3 ) if @xmath146 is a corner of @xmath121 and @xmath158 , then there exist a support line of @xmath12 that passes through @xmath146 and intersects the unit circle in two points in @xmath89 . as shown above , it follows then that @xmath146 actually belongs to @xmath12 . by a result in @xcite
, @xmath146 has to be an eigenvalue of @xmath1 , contradicting ( 2 ) .
assertions ( 2 ) and ( 3 ) of theorem [ th : exposed and corners ] were proved in @xcite in case the characteristic function of @xmath1 is a blaschke product .
it was shown in @xcite that , when @xmath1 acts on a finite dimensional space and @xmath31 is a unitary 1-dilation of @xmath1 , the boundary of the numerical range of @xmath31 is a polygon whose @xmath117 sides are tangent to @xmath12 .
the theorem below is an infinite dimensional version of this result ( see also @xcite for a more restricted infinite dimensional extension ) .
the following sections show that additional information can be obtained under certain conditions .
[ th : polygon ] let @xmath1 be a completely nonunitary contraction with @xmath159 , @xmath31 a unitary @xmath160-dilation of @xmath1 , and let the arc @xmath161 be a connected component of @xmath89 .
then : 1 .
if the segment @xmath162 $ ] is a support line for @xmath12 , then @xmath163 contains at most one eigenvalue of @xmath31 .
2 . if the segment @xmath162 $ ] is not a support line for @xmath12 , then 1 .
the only possible limit points of @xmath163 are @xmath164 and @xmath165 ; 2 .
each segment determined by consecutive values of @xmath163 on @xmath124 is a support line for @xmath12 , and its intersection with @xmath121 is actually in @xmath12 .
if the segment @xmath162 $ ] is a support line for @xmath12 , then it separates @xmath12 from @xmath166 .
if @xmath163 contained two eigenvalues , lemma [ le : intersection_num_ranges ] would imply the existence of a point in @xmath12 on the segment joining them , which is a contradiction .
assume then that the segment @xmath162 $ ] is not a support line for @xmath12 .
it is immediate that the eigenvalues of @xmath31 can not have a limit point in the interior of the arc @xmath166 , since that point would also be a limit point of @xmath12 .
since @xmath30 , there exist points in @xmath167 contained in the open arc @xmath168 , and they form a set of eigenvalues of @xmath31 which is at most countable . if , for two such eigenvalues @xmath169 , @xmath170 , then the segment @xmath171 $ ] is a support line for @xmath12 .
indeed , it must have a common point with @xmath12 by lemma [ le : intersection_num_ranges ] . on the other hand ,
if there are points in the circular segment determined by @xmath61 and @xmath136 that belong to the boundary of @xmath12 , those points should belong to @xmath167 . since @xmath167 is the closed convex hull of @xmath172
, one must have points in @xmath173a contradiction .
if @xmath2 is finite dimensional , then @xmath174 , so @xmath175 .
the analogue of theorem [ th : polygon](2b ) for this case is contained in ( * ? ? ?
* theorem 6.3 ) .
we end this section with a preliminary result that will be used in section [ se : segments ] . [
le : eigenvalues ] let @xmath1 be a completely nonunitary contraction with @xmath39 and @xmath31 a unitary @xmath160-dilation of @xmath1 .
assume that @xmath120 is a support line for @xmath12 as well as for @xmath167 , and @xmath176 , with @xmath177 .
then @xmath178 is nonempty if and only if @xmath179 and @xmath180 are eigenvalues of @xmath31 .
when this happens , @xmath178 contains exactly one point , and the set @xmath181 is a subspace of dimension one .
suppose @xmath182 ; then @xmath183 for some @xmath118 with @xmath184 . using the spectral theorem for @xmath31 , it is easy to see that @xmath185 must be eigenvalues of @xmath31 , and @xmath118 is a linear combination of corresponding eigenvectors @xmath186 and @xmath187
we claim that @xmath186 and @xmath188 are uniquely determined ( up to a scalar ) . indeed ,
if @xmath189 and @xmath190 with @xmath191 , then lemma [ le : intersection_num_ranges ] implies that @xmath192 , contrary to proposition [ pr : basic](1 ) .
assume that there exist two distinct points @xmath193 ; then @xmath194 and @xmath195 , and @xmath196 are two linearly independent vectors in the linear span of @xmath186 and @xmath187 .
it would follow that @xmath197 are in the linear span of @xmath196 , and thus in @xmath2 .
this contradicts again proposition [ pr : basic](1 ) .
a similar argument shows that @xmath198 can not contain two linearly independent vectors .
the theory of the characteristic function of a completely nonunitary contraction can be used in order to obtain more precise results about its numerical range .
we recall briefly the main elements of this theory in the scalar case ; the general situation can be found in @xcite .
we denote by @xmath199 and @xmath200 the usual hardy spaces in the unit disc , while @xmath201 .
given @xmath202 with @xmath203 , denote @xmath204 .
the _ model space _ associated to @xmath205 is defined by @xmath206 and the model operator @xmath207 is the compression to @xmath208 of multiplication by @xmath209 on @xmath210 . this _ functional model _ becomes simpler for inner functions @xmath205 , in which case @xmath211 and @xmath212 .
it is known ( see @xcite for @xmath205 inner and @xcite for the general case ) that @xmath213 consists of the zeros of @xmath205 and they are all eigenvalues , while the set @xmath89 is the union of the subarcs @xmath166 of @xmath124 accross which @xmath205 can be extended as an analytic function with @xmath214 on @xmath166 .
we use the notation @xmath215 ; this is also an analytic function in the neighborhood of each of the connected components of @xmath216 . the theory of sz - nagy
foias @xcite says that @xmath207 is a completely nonunitary contraction , and any completely nonunitary contraction @xmath1 with @xmath39 is unitarily equivalent to some @xmath207 . in this case
@xmath205 is called the _ characteristic function _ of @xmath1 ; it is inner precisely when @xmath217 strongly .
since the numerical range is a unitary invariant , in the sequel we study the operator @xmath207 for a given function @xmath205 .
[ th : facts theta ] all unitary dilations of @xmath207 to @xmath218 can be indexed by a parameter @xmath219 , in such a way that a point @xmath220 is an eigenvalue of @xmath221 if and only if @xmath222 has an angular derivative in the sense of carathodory at @xmath151 and @xmath223 . in particular , @xmath224 consists precisely of the solutions in @xmath216 of the equation @xmath225 .
we have then @xmath226 most of the assertions of the theorem are proved in @xcite for @xmath205 inner and in @xcite for general @xmath205 .
the only exception is formula , which is obtained by taking @xmath4 in theorem [ th : intersection result](2 ) .
note that @xmath225 might have no solutions in some ( or any ) component of @xmath216 ; it follows that @xmath221 has no eigenvalues therein .
if there is @xmath54 such that the equation @xmath225 has no solutions in the whole set @xmath216 , then @xmath227 is precisely the closed convex hull of @xmath228 .
the space @xmath229 is finite dimensional if and only if @xmath205 is a finite blaschke product .
this is the case studied extensively by gau and wu ( see @xcite ) and we exclude it in the sequel .
then @xmath207 acts on an infinite dimensional space , and @xmath230 . in the remainder of this section
we discuss the behavior of @xmath222 at an endpoint of a component of @xmath216 .
if @xmath231 is such a connected component , then @xmath222 can be extended as an analytic function across @xmath166 to an open set @xmath232 ; we still denote by @xmath222 the extended function .
we can define a continuous function @xmath233 by the formula @xmath234 , where the choice of the branch of the logarithm is such as to insure the continuity of @xmath235 .
we have also @xmath236 the following two propositions state essentially that the nontangential behavior of @xmath237 is closely related to its behavior along the arc @xmath166 .
they are certainly known , but we have not found an appropriate reference , so we include the proofs for completeness . in a different context , a discussion of the behavior of an analytic function in the neighborhood of an isolated singularity can be found in @xcite .
[ pr : limit of the function ] with the above notations , the following are equivalent : 1 .
@xmath237 has a nontangential limit at @xmath238 .
2 . @xmath239 .
there exists @xmath240 such that @xmath235 is one - to - one on @xmath241 , and @xmath235 can be continuously extended to @xmath242 . when these conditions are not satisfied , the equation @xmath225 has , for each @xmath54 , an infinite sequence of roots in @xmath166 converging to @xmath238 .
it is immediate that ( 2 ) is equivalent to ( 3 ) . on the other hand , ( 2 ) states the existence of the limit of @xmath222 at @xmath238 along the arc @xmath166 .
the equivalence between ( 2 ) and ( 1 ) is then a consequence of lindelf s theorem @xcite ( see ( * ? ? ?
* theorem 2.3.1 ) for a more recent presentation ) applied to @xmath222 defined on @xmath243 . finally , since @xmath235 is increasing , if ( 2 ) is not satisfied , then @xmath244 .
the last assertion is then immediate .
[ pr : limit of the derivative ] assume _
( 1)(3 ) _ in proposition _ [ pr : limit of the function ] _ are true
. then the following are equivalent : 1 .
@xmath237 has an angular derivative in the sense of carathodory at @xmath238 .
@xmath235 is right differentiable at @xmath242 .
when these conditions are not satisfied , @xmath235 has an infinite right derivative at @xmath242 .
write @xmath245 where @xmath246 is a blaschke product with zeros @xmath247 and @xmath248 is a positive measure . according to a result of ahern and clark @xcite
, @xmath222 has an angular derivative at @xmath249 if and only if @xmath250 both conditions ( 1 ) and ( 2 ) are true for a product if and only if they are true for each of the factors , so we discuss separately the two factors in .
the same argument implies that we may assume that @xmath251 , with @xmath252 , @xmath253 , and @xmath254 .
for the blaschke product @xmath246 , yields @xmath255 in particular , if @xmath256 , @xmath257 and the value of each integral is the angle at @xmath258 subtended by the arc @xmath259 @xcite .
this angle is bounded below if @xmath260 ; as @xmath235 is bounded on @xmath166 , we may assume ( discarding some of the zeros , if necessary ) that @xmath261 for all @xmath0 .
then , for @xmath262 and sufficiently small , @xmath263 comparing then ( for @xmath264 ) and , the equivalence of ( 1 ) and ( 2 ) follows in this case ; moreover @xmath265 if @xmath222 is the singular factor in , then yields @xmath266 for @xmath262 sufficiently small , @xmath267 and again , comparing ( for @xmath264 ) and , the equivalence of ( 1 ) and ( 2 ) follows
. moreover , @xmath268 the last assertion of the proposition follows from and .
we have stated propositions [ pr : limit of the function ] and [ pr : limit of the derivative ] for the left endpoint of @xmath269 ; similar statements are valid for @xmath270 .
note that if @xmath238 is an isolated point of @xmath271 and is thus the endpoint of two components of @xmath272 , then the equivalent assertions of proposition [ pr : limit of the function ] can not be true on both sides of @xmath273 .
finally , let @xmath231 be , as above , a connected component of @xmath274 .
the set @xmath275 is either empty or equal to an open interval @xmath276 .
we define the function @xmath277 by @xmath278 then @xmath279 is strictly increasing and differentiable on @xmath276 .
it follows from theorem [ th : polygon ] that for any @xmath280 the line determined by @xmath209 and @xmath281 is a support line for @xmath282 , and that any support line for @xmath282 that intersects @xmath124 in two points in @xmath216 is of this form for some @xmath166 and @xmath283 .
in this section we fix a function @xmath202 with @xmath284 . in order to investigate the boundary of @xmath12 , we discuss separately its intersection with each of the circular segments that correspond to connected components of @xmath216 .
theorem [ th : facts theta ] shows that this study depends on the behavior of @xmath237 on such an interval .
we already know by theorem [ th : exposed and corners ] that the boundary of @xmath282 has no corners in the open disk @xmath145 and contains no eigenvalues of @xmath207 .
it was proved in @xcite , for @xmath205 a finite blaschke product , that the boundary does not contain straight line segments . as we show in this section ,
this is not true for all inner functions . for @xmath285 , we call @xmath286 $ ] a _ maximal _ segment of @xmath287 if it is not contained strictly in a larger segment included in @xmath287 .
[ th : segments in d ] there are no maximal segments of @xmath287 contained in @xmath145 .
assume @xmath288 $ ] is a maximal segment of @xmath287 contained in @xmath145 .
denote by @xmath289 and @xmath290 , @xmath291 the intersections of the support line @xmath120 of this segment with @xmath124 ( with the points @xmath292 in this order ) .
then @xmath289 and @xmath290 belong to the same connected component @xmath166 of @xmath272 ; the function @xmath279 is defined in @xmath293 and @xmath294 .
a geometrical argument shows that the left hand derivative of @xmath279 in @xmath293 may be estimated by using that , for values of @xmath295 slightly smaller than @xmath293 , the support line of @xmath282 which passes through @xmath296 is close to the line @xmath297 . a computation involving similar triangles ( see figure 1 ) yields
then @xmath298 a similar computation , using the fact that the right hand derivative of @xmath299 in @xmath293 can be estimated by using , for @xmath295 slightly larger than @xmath293 , the line @xmath300 , yields @xmath301 since @xmath299 is differentiable at @xmath293 , we obtain a contradiction if @xmath302 .
next we discuss the possible segments contained in @xmath121 which have an endpoint on @xmath124 . by proposition [ pr : basic](3 )
this endpoint , which we denote by @xmath179 , must belong to @xmath303 ; moreover , it has actually to be in the boundary ( relative to @xmath124 ) of this last set .
we consider then an arc @xmath304 that is a connected component of @xmath305 .
one case can be settled easily .
[ th : one - to - one on the arc ] @xmath306\subset\partial w(s_\theta)$ ] if and only if @xmath237 is one - to - one on @xmath307 .
when these conditions are satisfied , the intersection @xmath306\cap w(s_\theta)$ ] is either empty or contains one point .
the second case happens when @xmath222 has an angular derivative in the sense of carathodory at both @xmath179 and @xmath180 . if @xmath308 for some @xmath309 , then @xmath310 and @xmath311 are eigenvalues of @xmath221 , and therefore there exists an element of @xmath282 on @xmath312 by lemma [ le : intersection_num_ranges ]
. therefore @xmath306 $ ] canot be in the boundary of @xmath282 .
conversely , suppose @xmath237 is one - to - one on @xmath307 .
if @xmath313 , then @xmath151 is the only solution on that arc of the equation @xmath314 , and thus it is the only eigenvalue therein of the unitary dilation @xmath315 .
the intersection of @xmath316 with the circular segment determined by @xmath307 is therefore the triangle @xmath317 .
the intersection of all these triangles for @xmath313 does not contain any point from the ( open ) circular segment determined by @xmath179 and @xmath180 , and therefore theorem [ th : facts theta ] implies that @xmath306\subset\partial w(s_\theta)$ ] . by theorem
[ th : intersection result](1 ) , there exists a unitary dilation @xmath221 of @xmath207 that has @xmath306 $ ] as a support line , and lemma [ le : eigenvalues ] implies that @xmath306\cap w(s_\theta)\not=\emptyset$ ] precisely when @xmath179 and @xmath180 are both eigenvalues of @xmath221 . by theorem
[ th : facts theta ] , this happens when @xmath222 has an angular derivative in the sense of carathodory at @xmath179 and @xmath180 .
we are then left with the case when precisely one of the endpoints of the segment belongs to @xmath124 .
a preliminary result discusses corners of @xmath318 on @xmath124 .
[ th : angle at end point ] if @xmath304 is a connected component of @xmath305 , then the following are equivalent : 1 .
there exists a point @xmath319 such that @xmath237 is one - to - one on @xmath320 .
there exists a line segment @xmath321 $ ] , with @xmath322 , which is a support line for @xmath287 .
assume that ( 1 ) is true .
it is enough to consider the case when @xmath237 is not one - to - one on the whole arc @xmath320 , since otherwise we can apply theorem [ th : one - to - one on the arc ] .
the assumptions of proposition [ pr : limit of the function ] are then satisfied .
the function @xmath323 has a limit @xmath324 when @xmath325 on the arc @xmath320 , and there exists another point @xmath326 such that @xmath327 .
moreover , if @xmath279 is defined as in , then each of the segments @xmath328 $ ] is a support line for @xmath282 , and thus the same is true for their limit , which is @xmath329 $ ] . thus ( 2 ) is true . on the other hand ,
if ( 1 ) is not true , it follows that for any @xmath54 the equation @xmath225 has an infinite number of solutions in any interval @xmath320
. then theorem [ th : facts theta ] implies that any interval @xmath320 contains segments which are support lines for @xmath282 , so ( 2 ) can not be true . as noted above
, if @xmath330 and is an isolated singularity , then it is not possible for @xmath237 to be one to one on a left as well as on a right neighborhood of @xmath151 ( on @xmath124 ) .
thus , if a point of @xmath124 is a corner of @xmath121 , then one of the support lines is tangent to @xmath124 .
we can say more about the situation in theorem [ th : angle at end point ] .
[ th : end points ] suppose there exists a point @xmath319 such that @xmath222 is one - to - one on @xmath320 .
the following are equivalent : 1 .
there exists a segment @xmath321\subset \partial w(s_\theta)$ ] ( @xmath322 ) .
@xmath237 has an angular derivative in the sense of carathodory at @xmath179 .
when these equivalent conditions are satisfied , the intersection @xmath321\cap w(s_\theta)$ ] contains exactly one point .
condition ( 2 ) above means that @xmath222 satisfies the assumptions of proposition [ pr : limit of the derivative ] .
using the notations therein ( in particular , @xmath331 ) , it is easy to see that the required differentiability of @xmath235 at @xmath242 is equivalent to the differentiability of @xmath279 , or to the boundedness of the derivative of @xmath279 on a right neighborhood of @xmath242 . to show that @xmath287 contains a segment with an end at @xmath151
, one has , by theorem [ th : facts theta ] , to prove that the intersection of all half - planes determined by @xmath332 $ ] , with @xmath333 , and containing @xmath282 contains such a segment ; a moment s reflection shows that it is actually enough to consider @xmath334 in a neighborhood of @xmath179 .
the argument now is close to the one in theorem [ th : segments in d ] .
remembering that @xmath335 satisfies @xmath336 , take @xmath337 , and denote by @xmath338 the intersection between the segments @xmath329 $ ] and @xmath328 $ ] .
again by considering similar triangles ( see figure 2 ) we obtain that , when @xmath334 is close to @xmath179 , we have @xmath339 condition ( 1 ) is equivalent to the fact that @xmath340 is bounded below in a neighborhood of @xmath242 .
the above estimate shows that this happens if and only if the quotient @xmath341 is bounded .
the last assertion is obviously equivalent to the boundedness of the derivative of @xmath299 or , as noted above , to condition ( 2 ) .
assume now that the equivalent assertions of the theorem are satisfied .
if @xmath342 , then @xmath343 , and it follows from theorem [ th : facts theta ] that @xmath179 and @xmath310 are eigenvalues of @xmath344 . applying then lemma [ le : eigenvalues ]
shows that @xmath329\cap w(s_\theta)$ ] contains exactly one point . with some supplementary effort
one can show that , when it exists , the point in @xmath321\cap w(s_\theta)$ ] is precisely @xmath146 , the endpoint of the maximal segment contained in @xmath141 .
we already know from theorem [ th : exposed and corners ] that the exposed points of @xmath318 which are in @xmath145 actually belong to @xmath282 .
more can be said about boundary points of @xmath282 when @xmath205 is inner . in this case
one can describe all invariant subspaces of @xmath207 as having the form @xmath345 for some inner divisor @xmath346 of @xmath205 ( * ? ? ?
* proposition x.2.5 ) .
remember also that , by theorem [ th : exposed and corners](1 ) , an exposed point of @xmath318 belongs actually to @xmath282 .
[ th : exposed1 ] assume @xmath205 is an inner function , @xmath347 , and @xmath348 is a proper invariant subspace for @xmath207 . then @xmath349 ;
if , moreover , @xmath146 is an exposed point of @xmath318 , then @xmath350 .
the set @xmath198 is a subspace of dimension one ; if @xmath351 generates it , then @xmath351 is cyclic for both @xmath207 and @xmath352 .
let @xmath120 be a support line for @xmath146 which intersects @xmath124 in @xmath150 . from lemma [ le : eigenvalues ] and
theorem [ th : facts theta ] it follows that @xmath205 has an angular derivative in the sense of carathodory at @xmath151 and @xmath152 , and @xmath150 are consecutive solutions of the equation @xmath225 for some @xmath54 .
thus , if we move from @xmath151 to @xmath152 counterclockwise , the argument of @xmath323 increases by @xmath353 . if @xmath354 is an invariant subspace of @xmath207 , there exists a decomposition @xmath355 , with @xmath356 inner , such that @xmath357 ; @xmath358 is then unitarily equivalent to @xmath359 .
if @xmath360 , then @xmath120 is also a support line for @xmath146 with respect to this last set .
a similar reasoning tells us that the argument of @xmath361 has to increase by @xmath353 if we move from @xmath151 to @xmath152 counterclockwise .
but , since @xmath362 , it would follow that the argument of @xmath346 stays constant , and thus @xmath346 is a constant .
this contradicts the assumption that @xmath354 is a proper invariant subspace ; thus @xmath349 .
if @xmath146 is an exposed point of @xmath318 and belongs to @xmath363 , then it is also an exposed point of @xmath363 . by theorem
[ th : exposed and corners](1 ) , it must be actually in @xmath364 , and the preceding argument applies .
the subspace @xmath365 has dimension 1 by lemma [ le : eigenvalues ] . if @xmath351 generates @xmath116 , and @xmath354 is the invariant subspace of @xmath207 generated by @xmath351 , then ( 1 ) implies that @xmath366 .
the same argument can be applied to @xmath367 , which is an exposed point in @xmath368 , if we note that @xmath352 is unitarily equivalent to @xmath369 , where @xmath370 .
it follows from theorem [ th : exposed1](1 ) that , if @xmath348 is a proper invariant subspace for @xmath207 , then @xmath364 is contained in the interior of @xmath282 . for @xmath205 a finite blaschke product ,
this is proved in @xcite .
note that the assertion is not necessarily true for @xmath363 , as shown by example [ ex : interior ] .
we end this section with the analogue of a finite dimensional result proved in @xcite .
recall that a completely nonunitary contraction is said to be of class @xmath371 if @xmath372 for some nonzero function @xmath373 .
[ th : exposed direct summand ] assume that @xmath202 is an inner function and @xmath8 is a contraction of class @xmath371 , such that @xmath374 .
then we have @xmath375 .
moreover , if @xmath12 contains a point in @xmath376 , then @xmath1 has an orthogonal summand which is unitarily equivalent to @xmath207 .
it is well - known that if @xmath205 is the minimal function of @xmath1 , then @xmath1 is unitarily equivalent to the restriction of @xmath377 ( for some hilbert space @xmath26 ) to an invariant subspace @xmath354 .
we can assume that @xmath378 and @xmath1 is equal to this restriction .
the inclusion @xmath375 is obvious .
let @xmath379 be the given point and take @xmath380 with @xmath381 .
we have @xmath382 for some unit vectors @xmath383 , an orthonormal sequence @xmath384 , and @xmath385 with @xmath386 .
since @xmath387 is an exposed point , it follows that we must have @xmath388 for all @xmath389 . according to theorem [ th : exposed1](2 )
, we must have @xmath390 for some @xmath391 with @xmath392 .
then @xmath393 where @xmath394 is a unit vector .
again by theorem [ th : exposed1](2 ) , @xmath351 is cyclic for @xmath207 and @xmath352 , so @xmath395 is a reducing subspace of @xmath396 contained in @xmath2 which provides the required orthogonal summand .
we consider some examples of functions @xmath205 in the unit ball of @xmath200 and discuss the numerical ranges of the corresponding model operators @xmath207 .
in particular , they show that all cases described in theorems [ th : one - to - one on the arc ] , [ th : angle at end point ] , and [ th : end points ] actually occur .
also , it was proved for finite blaschke products in @xcite and for inner functions with one singularity on the unit circle in @xcite that @xmath318 determines the function @xmath205 .
examples [ ex : conformal ] and [ ex : interior ] show that this is not true in general .
[ ex : conformal ] let @xmath205 be a conformal mapping of the disc @xmath145 onto the region @xmath397 bounded by an arc @xmath168 on @xmath124 and a smooth curve that joins @xmath164 and @xmath165 and is contained in @xmath145 .
if @xmath398 , then @xmath399 is the closure of the arc @xmath400 , where @xmath401 , while it follows from theorem [ th : one - to - one on the arc ] that @xmath121 is precisely the closed convex hull of this set ; that is , the circular segment determined by @xmath400 .
note that this numerical range is depends only on @xmath402 and @xmath403 , and not on the precise shape of @xmath397 .
the example above has the advantage of simplicity , but one can obtain interesting examples of numerical ranges by looking only at inner functions @xmath205 . we may thus discuss the different cases possible cases for a point in the boundary of @xmath140 .
[ ex : infinite solutions ] let @xmath404 be the decomposition into blaschke product @xmath246 and singular inner part with measure @xmath405 , and let @xmath249 be an endpoint of a connected component of @xmath216 .
it is shown in ( * ? ? ? * lemma 3.3 ) that if there exists a sequence of zeros of @xmath246 that converge to @xmath249 in a stolz angle , or if @xmath406 , then @xmath237 is not one - to - one on any arc @xmath407 .
the assumptions of proposition [ pr : limit of the function ] are not satisfied , and @xmath141 is smooth at @xmath334 by theorem [ th : angle at end point ] .
[ ex : corners with segment ] to obtain examples of corners of @xmath141 , we use the above quoted fact that if @xmath222 is written under the form , then is necessary and sufficient for the existence of the angular derivative .
choose then a function @xmath205 such that @xmath334 is an endpoint of a component of @xmath274 and is satisfied . according to theorem [ th : end points ]
, @xmath141 contains then a segment with one endpoint at @xmath334 .
[ ex : corners no segment ] another result of ahern and clark @xcite says that if an inner function is written as in with @xmath248 singular , @xmath408 , and @xmath409 then @xmath205 has a nontangential limit at @xmath334 .
choose the zeros of the blaschke product and the singular measure such that @xmath205 is an endpoint of a connected component of @xmath274 , is satisfied , but is not .
then , according to theorem [ th : end points ] , @xmath334 is a corner of @xmath141 , but @xmath141 does not contain a line segment ending at @xmath334 .
[ ex : interior ] the last example is related to theorem [ th : exposed1 ] .
consider a blaschke product @xmath205 whose zeros @xmath258 accumulate on @xmath410 .
this can be done such that the left hand side of is as small as we like , both for @xmath411 and for @xmath412 .
if this quantity is sufficiently small , @xmath205 is one - to - one on @xmath413 , and it follows then from theorem [ th : one - to - one on the arc ] that @xmath414 . we can then take a proper subsequence of @xmath258 with the same properties . if @xmath415 is the corresponding blaschke product , then @xmath416 is a proper invariant subspace of @xmath207 , and its restriction to this subspace is unitarily equivalent to @xmath417
. we know by theorem [ th : exposed1 ] that @xmath418 is contained in the interior of @xmath419 . however , @xmath420
. we can be more careful with the choice of the zeros of @xmath205 , such that @xmath222 has angular derivatives in the sense of carathodory at 1 and @xmath421 , and @xmath422 .
then the argument of @xmath222 increases by @xmath353 on the upper semicircle ; also , this is not possible for the argument of @xmath423 , which is a proper divisor of @xmath222 .
it follows from lemma [ le : eigenvalues ] and theorem [ th : facts theta ] that @xmath282 contains precisely one point @xmath424 $ ] , while @xmath425 . | an @xmath0-dilation of a contraction @xmath1 acting on a hilbert space @xmath2 is a unitary dilation acting on @xmath3 .
we show that if both defect numbers of @xmath1 are equal to @xmath0 , then the closure of the numerical range of @xmath1 is the intersection of the closures of the numerical ranges of its @xmath0-dilations .
we also obtain detailed information about the geometrical properties of the numerical range of @xmath1 in case @xmath4 .
= 1 |
x - ray observations of radio galaxies are of value for several reasons .
low brightness radio lobes may be thermally confined by the hot , extended atmospheres in elliptical galaxies or clusters of galaxies and x - ray observations can provide the density , temperature and pressure of this gas , thus constraining the pressures of the relativistic gas and magnetic field in a lobe ( e.g. worrall & birkinshaw 2000 ) .
high brightness regions , such as the hot spots in fanaroff - riley class ii ( frii ) radio galaxies , can not be statically confined by thermal pressure and must be expanding and/or advancing away from the nucleus .
if the motion is supersonic , such a feature must be preceded by a bow shock in the surrounding medium .
the effect of bow shocks on the ambient gas may be identifiable in thermal x - ray emission , depending on the run of temperature and density in the post - shock gas ( e.g. clarke , harris & carilli 1997 ) .
thermal x - rays may also be detectable from radio jets and lobes themselves if they entrain and shock surrounding gas ( e.g. de young 1986 ) .
the relativistic particles responsible for the synchrotron radio emission must also radiate through inverse compton scattering of ambient photons ( such as the microwave background and the starlight of the galaxy ) or the synchrotron photons themselves .
this inverse compton radiation is expected to be spread over a wide range of frequencies , including the x - ray band .
detection of both synchrotron and inverse compton radiation provides a direct measurement of the magnetic field in the emitting region ( e.g. harris , carilli & perley 1994 ) , allowing a check on the common assumption of equipartition of energy between cosmic rays and magnetic fields .
lastly , synchrotron x - ray emission may be observable if electrons and/or positrons of sufficiently high energy are present , a finding which would provide important clues about the origin of the relativistic particles in these objects .
for these reasons , we have begun a program of imaging and spectroscopy of radio galaxies with the chandra x - ray observatory .
pictor a is the seventh brightest extragalactic radio source in the sky at 408 mhz ( robertson 1973 ) , and the most powerful radio galaxy ( p@xmath7 = 6.7 @xmath8 10@xmath9 w hz@xmath10 for z = 0.035 , q@xmath11 = 0 and h@xmath11 = 50 km s@xmath10 mpc@xmath10 ) with z @xmath12 0.04 . the best radio maps ( perley , rser & meisenheimer 1997 , hereafter prm ) reveal two round , diffuse lobes with very much brighter radio hot spots on the sides away from the nucleus , corresponding to an frii classification .
the total angular diameter of the radio emission is about 76 ( 430 kpc ) .
the western hot spot , some 42 ( 240 kpc ) from the nucleus , is a remarkable object , being amongst the brightest of radio hot spots in radio galaxies and quasars .
prm have found a very faint radio jet extending from the compact , flat spectrum ( cf .
jones , mcadam & reynolds 1994 ) nuclear core to the western hot spot . recently ,
vlbi observations ( tingay et al . 2000 ) have revealed a milli arc second- ( pc- ) scale nuclear radio jet which is approximately aligned with the large - scale radio jet .
simkin et al .
( 1999 ) have reported an optical continuum extension @xmath2 01 ( 95 pc ) to the w of the nucleus and the direction to this extension is also roughly aligned with the larger scale radio jet .
rser & meisenheimer ( 1987 ) discovered optical emission from the western hot spot .
this emission has a featureless continuum , is strongly polarized and thus of synchrotron origin .
these authors also found a marginal detection of the hot spot at x - ray wavelengths in an einstein ipc observation .
further observations of the hot spot in the infrared have been reported by meisenheimer , yates & rser ( 1997 ) .
the overall spectrum of the western hot spot conforms to a power law with index @xmath13 = 0.74 ( s @xmath14 @xmath15 ) from 327 mhz to @xmath16 10@xmath17 hz , where the spectrum turns down ( see fig .
2i of meisenheimer , yates & rser 1997 ) .
the morphologies of the hot spot are very similar at radio and optical wavelengths , with a bright , compact , leading `` core '' and a fainter , following `` filament '' , which extends @xmath2 15@xmath18 ( 14 kpc ) more or less perpendicular to a line joining the hot spot to the nucleus ( prm ) .
fainter radio emission associated with the hot spot is also seen to the e and ne of the optical and radio filament , the overall angular extent of the radio hot spot being @xmath2 25@xmath18 ( 24 kpc ) .
the radio to optical spectrum of the `` filament '' is somewhat steeper ( @xmath19 = 0.98 ) than that of the bright , compact region ( @xmath19 = 0.87 ) .
optical imaging and polarimetry of the bright part of the hot spot with the hubble space telescope ( thomson , crane & mackay 1995 ) confirms the high polarization ( @xmath20 50% ) and resolves the hot spot into highly polarized `` wisps '' elongated nearly perpendicular to the nucleus - hot spot line .
the nucleus of pictor a has long ( schmidt 1965 ; danziger , fosbury & penston 1977 ) been known for its strong emission - line spectrum , with broad wings to both forbidden and permitted lines .
double - peaked , very broad balmer lines were discovered in 1993/1994 ( halpern & eracleous 1994 ; sulentic et al .
1995 ) ; these double - peaked lines were not present in 1983 .
such line profiles are commonly interpreted as emission from an accretion disk ( e.g. storchi - bergmann et al .
the nucleus of pictor a is a strong x - ray source ; the spatially integrated emission observed with asca has a power - law continuous spectrum and no evidence for an fe k@xmath13 line ( eracleous & halpern 1998 ) .
further observations of the x - ray spectrum of pictor a have been reported recently by padovani et al .
( 1999 ) and eracleous , sambruna & mushotzky ( 2000 ) .
we selected pictor a for observation with chandra because of the x - ray emission from the western hot spot , tentatively detected with the einstein ipc and confirmed by us through an inspection of an archival rosat pspc observation .
further , the optical synchrotron emission from the hot spot is of sufficient extent ( @xmath16 15@xmath18 ) to be well resolved by chandra , making this object a prime candidate for the study of high energy processes in the hot spots of powerful radio galaxies .
in addition to the chandra observations , we also report analysis of archival hubble space telescope ( hst ) optical and near uv imaging observations of the western hot spot .
section 2 describes the observations and their reduction while section 3 presents the results . in section 4
, we discuss the orientation of the radio source and the processes responsible for the x - ray emission of the western hot spot and jet . concluding remarks are given in section 5 .
pictor a was observed by the chandra x - ray observatory on jan 18 2000 ( sequence number 700018 , obsid 346 ) using the advanced ccd imaging spectrometer ( acis ) spectroscopic array , which provides excellent spatial resolution ( @xmath21 1@xmath18 ) with medium spectroscopic resolution ( @xmath2 130 ev fwhm at 1 kev ) at the aim point on chip s3 ( back illuminated ) . in order to obtain good spatial resolution on the western hot spot while retaining excellent resolution on the nucleus
, the latter was placed about 1@xmath22 from the aim point in the direction towards the s3/s2 chip boundary ( the y direction ) .
the nucleus was thus about 1@xmath22 from the s3/s2 boundary .
the western hot spot was @xmath2 35 from the aim point and also on chip s3 .
the spatial resolution at the nucleus is barely different from the optimal , while that at the hot spot is 12 @xmath8 19 fwhm .
the eastern radio lobe is partly on s3 and partly on s2 ( front illuminated ) . the total integration time ( `` good time intervals '' ) was 26.177 ksecs , taken with the default frame time of 3.2 sec .
there were no flares in the background count rate , which remained constant at 1.4 counts per second over the s3 chip .
the data were processed using version r4cu4upd4.2 of the pipeline software , and we have used the updated gain file acisd1999-09-16gainn0003.fits . the data extraction and analysis
have been performed using version 1.1 of the ciao software and version 11.0 of xspec . in order to create an x - ray image covering the whole radio source ( shown as fig .
1 ) , it was necessary to remove ccd artifacts that take the form of broad , linear features along the readout direction in chips s2 and s3 .
this was accomplished by first copying a spatial subsection of the level 2 events file and rotating it by -37.6@xmath6 , so that the axis of the s array ran horizontally .
the image was then smoothed with a gaussian with @xmath23 = 10 and the average of the rows computed , including rejection of those pixels in each column which were more than 2@xmath23 from the mean for that column .
the average of the rows was then subtracted from each row .
finally , the image was rotated by + 37.6@xmath6 , so the rows and columns corresponded to the cardinal directions .
this process not only removed most of the linear artifacts , but also the ` streak ' from the very strong nuclear source that results from its readout and much of the differences in background level between the two chips .
to derive the x - ray spectrum of the hot spot , counts were extracted from a rectangular region @xmath24 in extent which includes all of the x - ray emission from the hot spot .
background counts were taken from an area of identical shape offset to the west .
other extraction regions were tried for the background but no significant changes were seen .
we obtain a total of 3413 counts from the hot spot and 125 from the background .
response and auxiliary response files were created using the standard ciao tools .
the source counts were grouped so that each channel has at least 25 counts so that the @xmath25 fitting in xspec is meaningful .
we followed a similar procedure to obtain the x - ray spectrum of the jet .
counts were taken from a rectangular region 103@xmath18 @xmath8 24 that contains the whole jet and minimises background ( the @xmath16 10@xmath18 gap between the nucleus and the eastern end of the jet was excluded ) .
the background was taken from adjacent regions , and no significant difference was found for the jet s spectrum when using different regions for the background .
we obtain a total of 481 counts from the jet and 240 from the background .
the x - ray response of the ccds depends on the location of the source on the chip . to account for this
, the back - illuminated s3 chip has 1024 individual calibration measurements ( 256 per node ) arranged in a @xmath26 grid covering the face of the chip .
this oversamples the variation in the response of the ccd , and a single calibration measurement may be used for a source extending up to @xmath27 pixels ( approximately 32@xmath18 @xmath8 32@xmath18 ) .
the x - ray jet extends 228 pixels ( @xmath28 ) from the nucleus , so we divided it into four sections , each of which has a separate response .
we also note that the jet crosses a ` node boundary ' ( the division between sections of the chip that have different read - out amplifiers ) , and that the response may vary sharply across this boundary . to avoid any problem with this , we also divided the jet at the node boundary so as not to mix events collected by different nodes .
each of the four spectra were grouped to give each channel at least 20 counts so that @xmath25 fitting is valid .
these spectra were fitted simultaneously in xspec with each spectrum having its own background , response and auxiliary response file .
the spectral properties are assumed to be constant along the length of the jet so the parameters in the fit are tied together with the exception of their normalizations .
one drawback of treating the jet in this way is that the individual spectra will have extremely low count rates , and fitting may only be valid over a limited range in energy .
it is also difficult to produce clear figures of the spectral fits since there are four almost coincident data sets . to overcome these problems the entire jet spectrum
may be treated as a single data set , using an ` averaged ' response function that is the photon weighted average of the four individual responses .
we use pi spectra , which are gain corrected , to minimize the effect of collecting photons from two different nodes .
the results we obtain using this technique are equivalent , within the errors , to those obtained through fitting the spectra individually .
six images of the western hot spot were obtained on august 19 , 1995 as part of go program # 5931 ( pi meisenheimer ) , three each in the f300w and f622w broadband filters .
the images in each filter were dithered by several arc seconds in order to improve the effective spatial resolution of the images on the pc chip . however , the lack of duplicate exposures at each pointing , i.e. , no cr - split , and the paucity of sources in the field greatly complicated attempts at alignment of the images . instead , a number of cosmic rays were cleaned from each image by hand , and then the total number of sky - subtracted counts in the hot spot was determined for each frame . after averaging the results for each filter
, the counts were converted to units of flux using the synphot package in iraf .
the derived values , as plotted in fig .
6 , are 30 @xmath29jy and 104 @xmath29jy for the f300w and f622w filters , respectively .
the errors in these values have been estimated at 10% .
a low resolution x - ray image , covering all of the radio emitting regions and with selected vla radio contours ( from prm ) overlaid , is shown in fig . 1 , while a full resolution x - ray image of the nucleus and structure to the west of the nucleus is shown in fig .
the brightest x - ray source is associated with the nucleus of the galaxy .
the position of this source , as measured with the chandra aspect system , is within 2@xmath18 of the radio and optical nuclear positions given by prm .
the nucleus is so bright in x - rays that it suffers from strong `` pile up '' , so no useful spatial or spectroscopic information can be obtained from this exposure ; a chandra observation with a shorter frame time is scheduled to obtain a reliable nuclear spectrum .
a linear feature , visible in both figures , extends 19 ( 110 kpc ) westward from the nucleus and is spatially coincident with the faint radio jet ( prm ) ; we shall refer to this feature as the `` x - ray jet '' .
the x - ray jet `` points '' at the western hot spot , which is 42 ( 240 kpc ) from the nucleus . when the nuclear x - ray source is shifted slightly to coincide with the nuclear radio source , the peak of the x - ray emission from the hot spot is within 1@xmath18 of its radio peak .
there is also evidence for faint , diffuse x - ray emission around the nucleus and extending @xmath16 1@xmath22 from it , but more sensitive observations are needed to confirm this component .
faint x - ray emission extends eastwards from the nucleus to the two eastern radio hot spots , 31 and 34 from the nucleus ( fig .
this faint x - ray emission is seen to brighten somewhat within @xmath16 1@xmath22 of these hot spots .
there are also four compact x - ray sources within the brightest radio contour around the eastern hot spots ; none of these x - ray sources coincides with either radio hot spot .
lastly , we point out a faint ( @xmath30 counts above the background ) , compact x - ray source 323@xmath18 e of the nucleus at @xmath13 = 05@xmath31 20@xmath32 [email protected] , @xmath34 = 45@xmath6 47@xmath22 44@xmath18 ( j2000 ) .
this x - ray source is beyond the eastern radio lobe , but it lies in p.a .
101@xmath6 from the nucleus , a direction which is precisely opposite to the direction to the western hot spot ( p.a .
281@xmath6 ) .
examination of the digital sky survey reveals a faint point source within @xmath16 1@xmath18 of this position , almost at the limit of the serc - j survey ( we estimate @xmath35 ) , and marginally detected in the second epoch southern ( uk schmidt ) survey .
it is unclear whether this compact x - ray source is related to pictor a. images of the western hot spot at radio ( prm ) , optical ( rser 1989 ; prm ) and x - ray wavelengths with similar resolutions are shown in fig .
3 . as described in the introduction ,
the radio and optical hot spots have a very similar morphology . in a higher resolution ( 045 @xmath8 009 ) radio image ( fig .
19 of prm ) , the bright `` core '' is found to have a sharp leading ( i.e. away from the nucleus ) edge , with a slower decline on the side towards the nucleus . behind this bright region
is a linear feature elongated in p.a .
32@xmath6 ( fig .
3 ) , referred to as the `` filament '' by rser & meisenheimer ( 1987 ) and the `` plateau '' by prm .
the overall x - ray morphology is remarkably similar to the radio and optical , with both the `` core '' and the `` filament '' detected .
much of the apparent elongation of the brighter part of the `` core '' of the x - ray hot spot ( fig .
3 ) reflects the shape of the point spread function ( psf ) .
4 compares profiles along the major and minor axes of the hot spot through both the hot spot and the model psf at the location of the hot spot . at half maximum , the hot spot is somewhat wider than the model , but the model psf does not include broadening by imperfections in the aspect solution . in order to assess the reliability of the model psf , we have compared a compact x - ray source ( visible in fig .
1 ) some 152@xmath18 from the nucleus in p.a .
40@xmath6 with the chandra model psf at this location .
the observed profile of this source ( which is at a similar distance from the aim point as the hot spot ) is in good agreement with the model psf , but broader by 02 at the half power point .
this number is similar to the excess broadening of the hot spot over the model psf along its minor axis ( fig .
4 ) , suggesting that the brighter part of the hot spot is unresolved in this direction .
the excess fwhm of the hot spot over the psf along its major axis ( fig .
4 ) is somewhat larger .
nevertheless , we feel that caution is necessary , since other effects ( e.g. the motion of the source on the detector during an observation ) could lead to further broadening compared with the model psf .
the conservative conclusion is that the brightest part of the `` core '' is marginally resolved or unresolved in x - rays .
there are , however , clear indications of structure in the hot spot at lower brightness levels . as in the radio ,
the x - ray core has a sharper edge away from the nucleus than towards it .
a narrow `` ridge '' extends @xmath2 35 ( 3.4 kpc ) in p.a .
108@xmath6 ( the general direction towards the nucleus ) from the hot spot peak . a weaker feature some 6@xmath18 ( 5.7 kpc ) ssw of the core and elongated towards p.a .
279@xmath6 , is seen in both the radio and x - ray images . in conclusion ,
when the small differences in resolution of the x - ray and radio images ( fig .
3 ) are taken into account , the x - ray and radio morphologies are remarkably similar .
this morphological similarity indicates that the x - ray , optical and radio - emitting regions are physically co - spatial and argues that the x - ray emission does not originate from a bow shock ahead of the radio hot spot .
the x - ray spectrum of the hot spot has been modelled in various ways .
a hot plasma with solar abundances and in collisional equilibrium ( `` raymond - smith '' model ) provides a poor fit to the data ( table 1 ) and can be ruled out . if the abundances of all elements heavier than helium are allowed to vary , we find that an abundance of @xmath36 10 per cent solar is needed for this thermal model to provide an acceptable description of the data
. however , this model ( table 1 ) requires a hydrogen column well below the galactic value of n@xmath37 = 4.2 @xmath8 10@xmath38 atoms @xmath39 towards pictor a ( heiles & cleary 1979 ) , and is thus implausible .
fixing the column at the galactic value while allowing the abundances to vary leads to a poor fit ( table 1 ) .
further , we shall argue in section 3.2.3 that a thermal model can be excluded on grounds that the required gas density is so high that the radio emission would be faraday depolarized , contrary to observation .
lastly , a power law of photon index @xmath0 = 2.07@xmath40 absorbed by a column density of solar abundance gas n@xmath37 = ( 7.1@xmath41 ) @xmath8 10@xmath38 atoms @xmath39 provides an excellent description of the spectrum ( table 1 , fig . 5 )
this column density is only slightly greater than the galactic column density towards pictor a , suggesting that the interstellar medium of our galaxy is responsible for most of the observed absorption .
the unabsorbed flux and luminosity of the western hot spot in the 2 - 10 kev band are 3.1 @xmath8 10@xmath42 erg @xmath39 s@xmath10 and 1.7 @xmath8 10@xmath43 erg s@xmath10 , respectively .
we first evaluate the thermal model .
if we assume a representative intrinsic dimension of 1.9 @xmath8 0.9 @xmath8 0.9 kpc ( i.e. 2@xmath18 @xmath8 1@xmath18 @xmath8 1@xmath18 ) , as suggested for the bright `` core '' by the x - ray image ( figs 3 and 4 ) , the required gas density for the best fitting raymond - smith model is n@xmath44 @xmath2 3 @xmath45 . because the x - rays come from the same general region of space as the radio and optical emission ( fig .
3 ) , we may estimate the magnetic field in this putative thermal gas from the synchrotron emission . for an equipartition magnetic field of 4.7 @xmath8 10@xmath46 gauss ( obtained assuming a total cosmic ray energy equal to twice that of the electrons , @xmath13 = 0.74 [ meisenheimer , yates & rser 1997 ] , lower and upper cut - off frequencies 10@xmath47 and 10@xmath17 hz , respectively , and an emitting volume corresponding to the size of the radio core - 760 @xmath8 190 @xmath8 190 pc [ 08 @xmath8 02 @xmath8 02 ] ) , the rotation measure through the hot spot is expected to be @xmath2 6 @xmath8 10@xmath48 rad m@xmath49 .
if the larger volume adopted above for the x - ray source is used to calculate the equipartition field , a value of 1.4 @xmath8 10@xmath46 gauss and a rotation measure through the hot spot of @xmath2 2 @xmath8 10@xmath48 rad m@xmath49 are found .
in contrast , the high polarization observed at 6 cm implies a rotation measure internal to the hot spot @xmath12 900 rad m@xmath49 ( n@xmath44 @xmath12 0.014 @xmath45 ) , a factor of @xmath2 200 lower than required for thermal emission .
if the 20 cm polarization is used , the upper limit on rotation measure and density are an order of magnitude smaller , but the hot spot is not well resolved at this wavelength .
we thus conclude that our upper limit to the gas density renders a thermal model untenable
. this conclusion could be wrong if either a ) the magnetic field has a preferred direction but many reversals ; such a field structure can provide the high observed synchrotron polarization but gives little faraday rotation , or b ) the thermal gas is actually in small , dense clumps with a small covering factor .
we favor neither of these : the field structure of a ) is physically implausible , and it is very improbable that thermal gas of density 3 @xmath45 , let alone the higher density required if the gas is clumped , exists 240 kpc from the pictor a galaxy .
another argument for a low gas density in the hot spot comes from the photoelectrically absorbing column .
most or all of this column comes from gas in our galaxy ; any contribution from other regions is @xmath12 3 @xmath8 10@xmath38 atoms @xmath39 ( table 1 ) . if spread uniformly throughout the hot spot `` core '' of average diameter @xmath16 1 kpc , the density is @xmath12 0.1 @xmath45 ( as long as electrons are not stripped from the k shells of the relevant elements ) , a factor of 30 below the density needed for thermal emission . in summary ,
our finding ( section 3.2.2 ) that a thermal model provides a poor description of the x - ray spectrum , and the upper limits to the gas density from the limits on faraday depolarization and the absorbing column , rule out a thermal model of the x - rays from the hot spot we conclude that the x - rays are non - thermal in origin and discuss relevant models in section 4.2.2 .
the x - ray jet extends 19 ( 110 kpc ) westward from the nucleus in p.a .
281@xmath6 ( fig .
this direction is only 1@xmath6 away from that quoted for the radio jet out to 3@xmath22 ( 170 kpc ) from the nucleus ( prm ) .
such a difference in p.a .
is within the errors of measurement , so we can conclude that the x - ray and radio jets are coincident .
the profile of x - ray emission along the jet , shown in fig . 6 , consists of a number of ` knots ' , the brightest of which is @xmath50 from the nucleus . in section 4.1
, we shall argue that the western radio lobe is the nearer .
the ratio of the x - ray fluxes of the western and eastern jets , using the brightest part of the western jet , is more than a factor of 10 ( fig .
6 ) , suggesting relativistic boosting , as discussed in section 4.3.2 .
we have compared the transverse profiles across the jet with the expected psf .
after deconvolution from the psf , the jet is found to be transversely extended .
the width of the jet varies somewhat along its length , but its typical fwhm = 20 ( 1.9 kpc ) .
we are confident of this transverse extent because : a ) the profile of the compact source 152@xmath18 from the nucleus in p.a .
40@xmath6 agrees well with the expected psf at its location ( section 3.2.1 ) , and b ) the linear feature that results from the ccd readout of the strong nuclear source gives a measure of the 1d psf at the nucleus ; this feature is much narrower than the jet .
these arguments show that the measured jet width is not a result of poor aspect solution or focus .
the jet is so faint in the radio that it is difficult to measure its width .
prm state `` no accurate estimate of the jet s width is possible , but it is clearly not greatly resolved by the 75 beam '' .
thus the upper limit to the radio width is consistent with the x - ray width . given the low count rate , the background begins to dominate above 2.5 kev .
the same models as used for the hot spot were tried for the jet ( table 2 ) .
the raymond - smith thermal models provide a poor fit to the spectrum , even when the metal abundance is a free parameter .
an absorbed power - law model provides an excellent description of the spectrum of the jet ( fig .
the absorbing column density ( @xmath51 = 5.8@xmath52 @xmath8 10@xmath38 atoms @xmath39 ) and photon index ( @xmath0 = 1.94@xmath3 ) are similar to those found for the hot spot .
again we conclude that most of the absorption occurs in the interstellar medium of our galaxy .
the unabsorbed 2 - 10 kev flux and luminosity of the jet are 2.5 @xmath8 10@xmath53 erg @xmath39 s@xmath10 and 1.4 @xmath8 10@xmath54 erg s@xmath10 , respectively .
even if the metal abundance is left as a free parameter , a raymond - smith thermal plasma model provides a poor fit to the jet s spectrum ( section 3.3.2 , table 2 ) .
since there is no information on the radio polarization of the jet , we can not argue against a thermal model on grounds of internal faraday depolarization , as we did for the western hot spot .
the x - ray absorbing column within the jet must be @xmath36 8.4 @xmath8 10@xmath38 atoms @xmath39 ( the maximum allowed column [ table 2 ] minus the galactic column ) , which implies a density n@xmath55 @xmath12 0.15 @xmath45 .
for the observed x - ray emission to be thermal , the average density of hot gas in the jet ( taken to be a cylinder of radius 1@xmath18 and length 19 ) is n@xmath55 @xmath2 0.05 @xmath45 , compatible with the limits from the absorbing column .
thus the only argument against thermal emission is the poor description of the spectrum by a raymond - smith model .
we will discuss non - thermal models for the jet in section 4.3.2 .
the projected distance , d@xmath56 , of the w radio hot spot of pictor a from the nucleus is larger than that , d@xmath55 , of the e radio hot spots .
this observation is consistent with the w radio lobe being the nearer one , since the extra light travel time to the more distant lobe implies that it is seen at an earlier phase in its motion away from the nucleus and thus appears closer to the nucleus .
further , the east lobe is more depolarized than the west at long radio wavelengths ( prm ) , again suggesting that the w lobe is the nearer ( laing 1988 ) . both the radio and x - ray jets of pictor a are seen on the w side of the nucleus .
these considerations strongly suggest that the jet sidedness is a result of relativistic boosting .
the angle of the source axis to the line of sight is , of course , unknown , but can be estimated from d@xmath56 and d@xmath55 if the source is assumed to be intrinsically symmetric , since @xmath57 where @xmath58 = v / c , v is the velocity with which the hot spots are receding from the nucleus and @xmath5 is the angle between the radio axis and our line of sight . a recent study ( arshakian & longair 2000 ) has found the mean value of @xmath58 in a sample of frii radio galaxies to be 0.11 @xmath1 0.013 . for pictor a , we may then write @xmath59 so @xmath5 = 23@xmath6 if v = 0.11 c. in view of its sensitivity to v and the assumption of symmetry , this value of @xmath5 should not be taken too seriously .
the radio lobes are remarkably round and are well separated in the vla images ( prm ) , although joined by a ` waist ' near the nucleus .
the overall radio extent in the plane of the sky is 430 kpc and the outward motions of the vlbi jet components ( tingay et al .
2000 ) are subluminal .
all these properties indicate that @xmath5 is not small . a large value of @xmath5 is suggested by the fact that the hot spots project further away from the nucleus than the outer edges of the lobes themselves .
tingay et al .
( 2000 ) have estimated that the vlbi - observed jet has @xmath5
@xmath12 51@xmath6 based on a jet - deflection model .
8 shows the broad band spectrum of the western hot spot .
all of the radio points and most of the optical and near infrared ones were taken from meisenheimer , yates & rser ( 1997 ) . to these we added the flux densities at @xmath2 2900 and @xmath2 6200 obtained from our analysis of archival hst data ( section 2.2 ) .
these hst values may underestimate the true flux densities as some of the emission is resolved out .
also included is the original estimate of the x - ray flux density from the einstein observatory ( rser & meisenheimer 1987 ) and the chandra - measured intrinsic spectrum .
the einstein point is a factor of @xmath16 2 higher than the chandra measurement ; this may be a result of partial blending of the hot spot with the much stronger nuclear source in the low spatial resolution einstein observation .
the radio spectrum of the hot spot is well described by a power law with @xmath13 = 0.740 @xmath1 0.015 ( meisenheimer , yates & rser 1997 ) , but there must be a break or turnover in the spectrum at 10@xmath4 hz to accommodate the near infrared and optical measurements .
it is also apparent that the x - ray spectrum is not a simple extension of the radio and optical measurements to higher frequencies .
the luminosities of the radio to optical and the x - ray components are given in table 3 .
it has long been realised that the outward motion of hot spots in frii radio galaxies is sub - relativistic .
the most recent work ( arshakian & longair 2000 ) obtains a mean outward velocity of 0.11c @xmath1 0.013c . while the velocity of a hot spot in any individual object is uncertain , we shall neglect boosting or diminution of the hot spot fluxes by bulk relativistic motion .
the broad - band spectrum ( fig .
8) shows that the x - ray emission of the western hot spot can not be synchrotron radiation from a high energy extension of the population of electrons responsible for the radio to optical synchrotron radiation .
it is , therefore , natural to consider inverse compton scattering for the x - ray emission . in the bright core of the hot spot ,
the radiant energy density ( @xmath60 ) of the synchrotron radiation dominates that of the microwave background radiation and the galaxy starlight ( table 3 ) , so it is appropriate to consider a synchrotron self - compton model .
when the scattering electrons follow a power - law distribution in @xmath61 (= e / mc@xmath62 ) , n(@xmath61 ) = n@xmath63 over all @xmath61 , then the inverse compton spectrum is a power law with a spectral index @xmath64 , the same index as the synchrotron spectrum . , where @xmath65 is the frequency of the synchrotron photon and @xmath66 is the rest mass energy of the electron .
] the fact that the radio ( @xmath67 = 0.740 @xmath1 0.015 ) and x - ray ( @xmath68 = 1.07@xmath40 ) spectral indices are different then suggests difficulties with an inverse compton model . however , in reality , the electron energy spectrum is a power law over only a certain range of @xmath61 ( from @xmath69 to @xmath70 ) , and the synchrotron spectrum is a power law over only a certain range of @xmath71 ( from @xmath72 to @xmath73 ) .
these limited ranges yield `` end effects '' in synchrotron self - compton spectra : the inverse compton scattered spectra must turn down below @xmath74 and above @xmath75 ( e.g. blumenthal & gould 1970 ; rybicki & lightman 1979 ) .
further , the synchrotron self - compton spectrum is no longer an exact power law between these limits . in order to provide a more realistic evaluation of inverse compton and synchrotron self - compton models ,
we have performed numerical calculations of spectra in spherical geometries using the computer code of band & grindlay ( 1985 , 1986 ) , which was kindly provided by dan harris . given the observed radio
optical spectrum ( fig .
8) , we have assumed a power law electron spectrum with @xmath76 = 2.48 from @xmath69 to @xmath70 .
the magnetic field is treated as a variable .
we first computed a synchrotron self - compton spectrum which passes through the chandra - measured flux ( model 1 ) .
this was achieved for a magnetic field strength of @xmath77 gauss , a factor of 14 below the equipartition field in the radio `` core '' of the hot spot ( table 4 ) .
however , the predicted spectrum is similar to that of the radio source and does not match the measured x - ray spectrum ( fig . 8) . as an alternative to reducing the field strength , the power radiated in inverse compton radiation
could be increased by increasing the radiation density .
the nucleus of pictor a might emit a narrow , collimated beam of radiation ( like those inferred to be present in bl lac objects ) along the axis of the jet .
this beam could illuminate the hot spot , but be invisible to us .
however , an ( isotropic ) nuclear luminosity of 1.6 @xmath8 10@xmath78 erg s@xmath10 is needed merely to _ equal _ the radiant energy density in synchrotron radiation in the core of the hot spot .
the ( isotropic ) nuclear luminosity would have to be @xmath2 1.5 @xmath8 10@xmath79 erg s@xmath10 to provide sufficient radiation for the hot spot to radiate the observed x - ray flux by inverse compton scattering if the field has its equipartition value .
this luminosity is unreasonably large .
an alternative is that the x - rays are a combination of synchrotron and inverse compton radiation . in model 2
, we assumed that the turnover in the synchrotron spectrum at 10@xmath4 hz is an effect of synchrotron losses on a continuously injected electron spectrum with energy index @xmath76 = 2.48 .
the assumed electron spectrum is thus a broken power law ( kardashev 1962 ) : @xmath80 the synchrotron spectrum then has an index @xmath13 = 0.74 well below the break and @xmath13 = 0.74 + 0.5 = 1.24 is expected well above it ( dotted line in fig . 8) . by adding this spectrum to the predicted synchrotron self - compton component for a magnetic field of @xmath81 gauss ( dot - dashed line in fig .
8) , which is again well below equipartition , we obtain the solid line in fig .
8 , which is a good description of the chandra spectrum .
the parameters of model 2 are given in table 4 . because the energy density in magnetic field is @xmath20 an order of magnitude larger than the synchrotron radiation density , the power in the first order synchrotron self - compton component is less than that in the synchrotron radiation ( e. g. rees 1967 )
further , the second order scattered component is even weaker ( 3 orders of magnitude below the first order radiation in the chandra band ) . assuming that the electrons are continuously accelerated in the hot spot
, we may interpret the turnover frequency as the frequency at which the `` half - life '' to synchrotron losses of the radiating electrons is equal to their escape time from the hot spot .
for the field of @xmath81 gauss needed to match the x - ray spectrum , electrons radiating at 10@xmath17 hz have a synchrotron `` half - life '' of @xmath2 10@xmath82 yrs .
for a hot spot radius of @xmath2 250 pc , the corresponding streaming velocity of the relativistic electrons is @xmath2 0.1c .
the problem with model 2 is the fine tuning needed : synchrotron and synchrotron self - compton emission must be present in the chandra band with comparable fluxes . in the previous subsection
, we showed that the x - ray emission is so strong that the magnetic field in the hot spot must be a factor of @xmath16 14 below equipartition if the x - rays are produced by inverse compton scattering from the electrons that generate the _ observed _ radio emission .
even then , the observed x - ray spectrum is different to the predicted one ( fig .
8) , and so this model ( model 1 , table 4 ) may be ruled out .
the other process capable of generating the observed x - rays is synchrotron radiation .
however , the x - ray emission of the hot spot is not a smooth continuation of the radio optical synchrotron spectrum . in model 2 , we contrived to fit the x - ray spectrum by a combination of synchrotron emission from an extension of the radio - optical spectrum and synchrotron self - compton from the radio - optical synchrotron - emitting electrons .
in view of the unsatisfactory nature of these models , we now consider inverse compton and synchrotron models involving hypothetical new electron populations . _
i ) x - rays as inverse compton radiation _ for a successful inverse compton scattering model , a low energy population of relativistic electrons with an index of the energy spectrum p @xmath2 2@xmath83 + 1 is needed .
the required value of p is thus
p@xmath84 @xmath2 3.14@xmath85 .
9 shows a model ( model 3 , table 4 ) for the hot spot in which a synchrotron - emitting component with a larger spectral index than that observed in the radio has been added at low radio frequencies .
the spectrum of the hot spot has not been measured below 327 mhz , but the extension of this hypothetical spectrum down to lower frequencies does not exceed the measured _ integrated _ flux density of pictor a ( the lowest such measurements are at @xmath16 80 mhz , see prm ) .
it is envisaged that the population of electrons which emits the _ observed _ synchrotron radio emission is in a relatively strong ( e.g. equipartition , table 3 ) magnetic field , so that the synchrotron self- compton emission from this _ electron _ population is negligible , as shown above . however , the observed radio - optical synchrotron _ radiation _ , as well as the radio synchrotron radiation of the hypothesised component , is available for scattering by the hypothetical low energy electron population . as can be seen from fig .
9 , the low energy electrons must radiate predominantly through the inverse compton channel , requiring a very weak magnetic field ( model 3 , table 4 ) which is @xmath16 100 times weaker than the equipartition field in the core of the western hot spot .
the calculation shows that the second order compton scattered component is @xmath20 2 orders of magnitude weaker than the first order compton scattered component in the chandra band .
we find that the spectrum of the low energy electrons has to be steep ( p = 3.3 ) as shallower spectra ( p @xmath12 3.3 ) cause the inverse compton spectrum to increase above @xmath86 hz , contrary to our x - ray observations .
taking this model to its extreme , the hypothetical electron population could be in a region free of magnetic field , and thus radiate no radio emission at all .
this requirement for a weak or absent magnetic field associated with this population is difficult to understand in view of the finding that the x - ray , optical and radio emitting regions of the western hot spot are co - spatial ( section 3.2.1 and fig .
high resolution radio observations at lower frequencies would provide stronger constraints on this model .
_ ii ) x - rays as synchrotron radiation _ alternatively , the x - ray emission of the hot spot could be synchrotron radiation from a separate population of electrons . in such models ( models 4a and b ) , the energy index of the x - ray emitting electrons would be p@xmath84 @xmath2 3.14@xmath85 , similar to the above inverse compton model . in view of the short synchrotron loss times
( a few years for a 5 kev - emitting electron in the equipartition field ) , the energy index at injection would be p@xmath87 @xmath2 2.14@xmath85 in a steady state , continuous injection model
. the synchrotron break frequency must thus be below the x - ray band but above 10@xmath88 hz ( otherwise the emission would exceed the observed flux at the latter frequency ) .
models 4b and 4a ( table 4 ) , in which we have adopted the equipartition field , represent these two extremes for the synchrotron break frequency , respectively . thus , in both these models , the extrapolation of the x - ray spectrum to lower frequencies does not exceed the observed radio , infrared or optical flux .
interestingly , the required value of p@xmath87 is the same to within the errors as the canonical index for particle acceleration by a strong shock ( p = 2 , e.g. bell 1978 ; blandford & ostriker 1978 ) . in this picture
, electrons would have to be continuously reaccelerated by such shocks on pc scales .
unfortunately , there is no information on the radio spectrum of the jet .
we have only prm s statement that the average excess brightness of the jet above the background lobe emission is estimated to be about 10 mjy ( 75 beam)@xmath10 at @xmath89 20 cm . over the length of the detected x - ray jet
, we then estimate a total 20 cm flux density of 152 mjy , with large uncertainties .
the spectral index between 20 cm and 1 kev is @xmath90 = 0.87 , which is similar to @xmath91 = 0.85 , the spectral index for the _ entire _ radio source for @xmath71 @xmath92 400 mhz ( prm ) .
this value of @xmath90 may also be compared with the index within the chandra band of @xmath93 = 0.94@xmath3 .
given the errors , it is not out of the question that the jet has a constant spectral index from ghz frequencies to 10 kev , but the radio spectrum is needed to check this . simply joining the radio and x - ray flux densities with @xmath94 = 0.87 implies a total optical magnitude for the jet of v @xmath2 23 mag , and an optical surface brightness of v @xmath2 29 mag ( arc sec)@xmath49 .
the fact that the jet has not been detected in the optical is consistent with these numbers .
some estimated parameters of the jet are given in table 5 .
it is apparent that the microwave background radiation dominates the radiant energy within the jet , unless the radius of the radio synchrotron - emitting region is much smaller than the observed radius of the x - ray jet .
the x - rays from the jet could be either inverse compton or synchrotron radiation , which we discuss in turn .
evaluation of these models is limited by our ignorance of the jet s radio spectrum .
for concreteness , we shall adopt a jet radius of 1@xmath18 ( 950 pc ) for both the radio and x - ray emission . _
i ) x - rays as inverse compton radiation _
we consider first inverse compton models in which doppler boosting is unimportant , and show that such models of the jet s x - ray emission are implausible .
we then consider models including doppler boosting . _ a ) insignificant doppler boosting by the jet
_ adopting @xmath95 = 0.87 , s@xmath96 = 152 mjy for the part of the jet detected in x - rays ( section 4.3.1 ) , and the equipartition field h@xmath97 = 2.3 @xmath8 10@xmath98 gauss , the predicted inverse compton scattered x - ray flux falls a factor of @xmath2 500 below that observed . in order to match the observed x - ray flux ,
we need to reduce the magnetic field to h@xmath97 @xmath2 7 @xmath8 10@xmath99 gauss , a factor of @xmath2 30 below equipartition . alternatively , as for the hot spot , we could invoke a radiation beam from the nucleus to boost the radiation density in the jet . to obtain the observed x - ray flux from the jet while retaining the equipartition field
requires an ( isotropic ) nuclear luminosity of @xmath16 2 @xmath8 10@xmath78 erg s@xmath10 , which is implausibly high .
we could also suppose , as we did for the hot spot , that the _ observed _ radio jet contains a stronger ( e.g. equipartition ) magnetic field and there is , in addition , a population of electrons in a weak or absent field . because we do not know the radio spectrum of the jet , and have thus not been able to demonstrate that the radio and x - ray spectral indices are different
, we can not prove that the radio- and x - ray - emitting electrons represent different populations , as we could for the hot spot .
the following simple fact argues against inverse compton scattering for the jet s x - ray emission _
if doppler boosting is unimportant _ : in the radio band , the western lobe dominates the jet , while in x - rays the converse is true .
the ratio of the rates of energy loss to inverse compton scattering and synchrotron radiation is @xmath100 the radiant energy density in both the jet and western lobe is dominated by the microwave background radiation .
the equipartition magnetic field in the jet is @xmath2 2.3 @xmath8 10@xmath98 gauss ( for r@xmath101 = 1@xmath18 , see table 5 ) and that in the lobe is @xmath16 5 @xmath8 10@xmath102 gauss .
thus one expects that the ratio ( de / dt)@xmath103/(de / dt)@xmath104 should be larger for the lobe than the jet .
given that the lobe s radio synchrotron radiation overwhelmingly dominates that of the jet , the lobe s inverse compton emission should dominate the jet by an even larger factor , contrary to observation .
however , the jet s inverse compton emission would be larger if either a ) its magnetic field is well below equipartition , or b ) a narrow beam of radiation is emitted by the nucleus along the jet , providing a larger @xmath60 than the microwave background , as discussed above .
both of these possibilities are _ ad hoc _ and so we consider the prominence of the jet compared to the lobe in the chandra image as an argument against inverse compton scattering in the absence of doppler boosting .
_ b ) significant doppler boosting by the jet _
the above discussion neglects relativistic boosting or diminution by possible bulk relativistic motion of the jet .
there is now a strong case for energy transport at bulk relativistic velocities to the hot spots in frii sources ( e.g. bridle 1996 ) . for emission which is isotropic in the rest frame and has a power law spectrum , the observed flux density , @xmath105 ,
is related to the flux density in the rest frame , @xmath106 , by @xmath107 where @xmath108^{-1}$ ] , @xmath0 is the lorentz factor of the bulk flow , @xmath58 is the bulk velocity in units of the speed of light , and @xmath5 is the angle between the velocity vector and the line of sight ( e.g. urry & padovani 1995 ) .
this equation assumes that the emission comes from a discrete , moving source . for a smooth , continuous jet
, the exponent 3+@xmath13 becomes 2+@xmath13 ( begelman , blandford & rees 1984 ) .
equation ( 6 ) will describe relativistic boosting or diminution of the jet s synchrotron radiation , as long as that radiation is isotropic in the jet s rest frame .
as noted above ( table 5 ) , the radiation density in the jet is dominated by the microwave background radiation , which is isotropic in the observer s frame and anisotropic in the rest frame of the jet . in this case , the principal dependence of the inverse compton scattered radiation on @xmath5 is given by ( dermer 1995 ) @xmath109 where an additional term which depends slowly on @xmath110 has been omitted ( begelman & sikora 1987 ; dermer 1995 ) .
retaining the discrete source model , the ratio of inverse compton scattered to synchrotron spectral flux density is then @xmath111 thus the ratio of inverse compton to synchrotron flux is increased by doppler boosting ( @xmath34 @xmath92 1 ) .
our goal is to develop successful models for the jet s x - ray emission for various assumed angles @xmath5 .
selection of a given @xmath5 defines a maximum value of @xmath34 ( @xmath112 @xmath2 1/@xmath5 ) , which is achieved when @xmath0 @xmath2 1/@xmath5 .
assuming @xmath112 ( which is the most optimistic choice , since it minimises the required reduction of h below equipartition at a given @xmath5 ) , we have computed the value of h required to reproduce the observed x - ray to radio flux ratio , and compared it with the equipartition field which would be inferred from the synchrotron emission by an observer in the rest frame of the jet ( estimated by `` deboosting '' the observed radio flux ; h@xmath22 @xmath16 h@xmath113 ) .
the results are given in table 6 .
for our canonical @xmath5 = 23@xmath6 ( section 4.1 ) , the field must be at least a factor of 6 below equipartition .
an equipartition model demands @xmath5 = 8@xmath6 , which is surely too small for a lobe - dominated frii radio galaxy like pictor a ( see discussion in section 4.1 ) .
such an angle would make pictor a one of the largest known radio galaxies ( length 3 mpc ) .
we conclude that inverse compton scattering of the microwave background is a viable model for the x - ray emission of the jet of pictor a , but requires a magnetic field which is substantially below equipartition .
_ ii ) x - rays as synchrotron radiation _ as already noted ,
the available data are consistent with a single power law with @xmath90 = 0.87 from 1.4 ghz to 10 kev .
this is , however , an unlikely physical situation for a purely synchrotron spectrum in view of the short energy loss times of the x - ray emitting electrons to synchrotron and inverse compton radiation . for electrons emitting synchrotron radiation at 1.4 ghz and 1 kev in the equipartition field of 2.3 @xmath8 10@xmath98 gauss , the times to lose half their energy are @xmath2 4 @xmath8 10@xmath114 and 300 yrs , respectively .
if the energies of the radio - emitting electrons are not significantly reduced by synchrotron and inverse compton losses , we would expect the radio spectrum to be flatter than the x - ray spectrum ; in the simplest case involving continuous injection , @xmath95 = @xmath93 0.5 @xmath2 0.4 . it would thus be valuable to measure @xmath95 . for a synchrotron model
in which the radiation is isotropic in the frame of bulk jet motion , the observed ratio of flux in the approaching side of the jet to the receding side is given by @xmath115 from x - ray observations @xmath116 and @xmath117 at the brightest part of the x - ray jet . assuming @xmath5
@xmath92 23@xmath6 ( section 4.1 ) , then @xmath58 @xmath92 0.3 .
if the x - rays are the result of inverse compton scattering of the microwave background radiation , the exponent 3+@xmath13 in equation ( 9 ) becomes 4 + 2@xmath13 ( cf .
equation 7 ) and a somewhat smaller lower limit to @xmath58 is obtained .
our chandra study of pictor a has shown that the x - ray emissions from the jet and the western hot spot are non - thermal .
the spectra of both are well described by an absorbed power law with ( flux density ) spectral index @xmath13 @xmath2 1.0 .
the x - ray spectrum of the hot spot is not a smooth extension of the radio to optical synchrotron spectrum , which turns down or cuts off near @xmath16 10@xmath17 hz . inverse compton scattering of the synchrotron radio photons by the relativistic electrons responsible for the radio emission ( i.e. a synchrotron self - compton model ) may be ruled out for the hot spot s x - ray emission , as the predicted spectrum differs from that observed .
we considered the possible existence of a population of relativistic electrons in the hot spot that radiates synchrotron emission at frequencies below that at which its spectrum has been measured ( i.e. @xmath12 327 mhz ) . by choosing an appropriate index for the energy spectrum of these electrons , we constructed a successful synchrotron self - compton model for the x - rays at the price of reducing the magnetic field to @xmath16 1% of equipartition ( see fig . 9 and model 3 in table 4 ) .
more generally , relativistic electrons could exist in regions with weak or absent magnetic fields .
since the properties of such electron populations are unconstrained by radio observations , one can always create inverse compton models that match the x - ray spectra .
it is , however , then difficult to understand why the x - rays from both the jet and the hot spot should correlate so well with synchrotron radio emission , which must arise in a relatively strong magnetic field .
the x - ray spectrum of the hot spot may be reproduced in a composite synchrotron plus synchrotron self - compton model . in this picture , the spectral index of the synchrotron radiation is supposed to increase by 0.5 above the break near 10@xmath17 hz , as would be expected in a continuous injection model .
addition of this synchrotron emission to the synchrotron self - compton emission expected for a magnetic field a factor of 9 below equipartition reproduces the observed spectrum ( see fig . 8 and model 2 ,
table 4 ) .
the model is contrived , requiring similar fluxes from the two components in the chandra band , but can not be ruled out .
if the jet is non - relativistic , inverse compton scattering is an implausible model for its x - ray emission , for it requires a magnetic field a factor of 30 below equipartition .
further , it is hard to understand why the jet is brighter than the lobe ( the opposite of the situation in the radio ) in such a model . if the jet is relativistic , these difficulties are eased , and we consider inverse compton scattering by such a jet off the microwave background a viable mechanism . however , the magnetic field must still be well below equipartition for plausible angles between the jet and the line of sight in this lobe - dominated , frii radio galaxy .
synchrotron radiation is a plausible x - ray emission process for both jet and hot spot .
strong , non - relativistic shocks are believed to accelerate relativistic particles yielding an energy spectrum n(e ) @xmath14 e@xmath118 with p = 2 at injection .
since the half lives of x - ray emitting electrons to synchrotron losses are very short ( @xmath16 years ) , the spectrum steepens and a synchrotron spectral index of @xmath13 = 1.0 is expected , in excellent accord with observations ( models 4a , b , table 4 ) .
the separate population of radio - optical synchrotron emitting electrons remains unexplained ; the radio spectral index - @xmath119 = 0.740 @xmath1 0.015 - is close to the average for non - thermal radio sources .
various processes , including acceleration in weak shocks , synchrotron losses and effects of tangled fields ( see summary in longair 1994 , ch .
21 ) , have been invoked to account for the difference between the typical index seen in radio sources and the value @xmath13 = 0.5 expected in the canonical model .
hot spots are , of course , associated with two shocks - an internal , mildly relativistic shock in the jet and a non - relativistic bow shock in the intergalactic medium .
both may reasonably be expected to accelerate cosmic rays and the resulting fluxes and energy spectra may differ . as well as being directly accelerated in shocks , high @xmath61 relativistic electrons may result from a ` proton induced cascade ' initiated by photopion production ( e.g. sikora et al .
1987 ; biermann & strittmatter 1987 ; mannheim & biermann 1989 ; mannheim , krlls & biermann 1991 ) . in this process ,
relativistic proton - photon collisions create @xmath120 , @xmath121 and @xmath122 . the last two decay into relativistic electrons , positrons , neutrinos and antineutrinos ( e.g. biermann & strittmatter 1987 ) .
the process thus provides a supply of synchrotron x- and @xmath61-ray emitting electrons and positrons from high energy protons accelerated by shocks .
mannheim , krlls & biermann ( 1991 ) considered this process for production of synchrotron x - ray emission in radio galaxy hot spots .
their calculations suggest that the proton induced cascade produces x - ray emission about an order of magnitude weaker and with a harder spectrum than is observed for the western hot spot of pictor a. as they note , the predicted luminosity can be increased by increasing the number of relativistic protons or the number of photons ( the latter might occur , for example , if the hot spot is illuminated by a beam of radiation from the galaxy nucleus ) .
this process seems promising and further calculations of the expected x - ray luminosity and spectrum of a proton induced cascade would be worthwhile .
our discussion of the western hot spot has assumed it moves outwards non - relativistically .
while this is established for the population of radio hot spots in general , it is not necessarily true for pictor a. the fact that the western hot spot is on the near side and is much brighter than the eastern hot spots at both radio and x - ray wavelengths raises the possibility of relativistic outflow .
some of the problems with inverse compton models of the hot spot might then be eased ( as they are for the jet ) , but the difference between the radio and x - ray spectral indices remains a difficulty .
this research was supported by nasa through grant nag 81027 and by the graduate school of the university of maryland through a research fellowship to asw .
we are extremely grateful to rick perley for providing the radio images published by prm in numerical form .
we also wish to thank the staff of the chandra science center , especially dan harris and shanil virani , for their help .
bridle , a. h. 1996 , in energy transport in radio galaxies and quasars , asp conference series vol .
100 , eds .
e. hardee , a. h. bridle and j. a. zensus , p 383 ( astronomical society of the pacific : san francisco ) raymond - smith & @xmath123 & @xmath124 & solar & @xmath125 & & & 166 / 97 + raymond - smith & 4.2 ( frozen ) & @xmath126 & solar & @xmath127 & & & 232 / 99 + raymond - smith & @xmath128 & @xmath129 & @xmath130 solar & @xmath131 & & & 101 / 96 + raymond - smith & 4.2 ( frozen ) & @xmath132 & @xmath133 solar & @xmath134 & & & 122 / 98 + power law & 7.1@xmath41 & & & & 2.07@xmath40 & ( 1.34@xmath135 ) @xmath8 10@xmath46 & 95 / 98 + lcccccccc raymond - smith & 4.2 ( frozen ) & @xmath136 & solar & @xmath137 & & & 27 / 15 + raymond - smith & 4.2 ( frozen ) & @xmath138 & @xmath139 solar & @xmath140 & & & 19 / 14 + power law & 5.8@xmath52 & & & & 1.94@xmath3 & ( 1.45@xmath141 ) @xmath8 10@xmath98 & 15 / 14 + lc l(327 mhz - optical ) & @xmath2 3 @xmath8 10@xmath142 erg s@xmath10 + l(0.1 - 10 kev ) & 4.9 @xmath8 10@xmath43 erg s@xmath10 + h@xmath143 ( bright core ) & 4.7 @xmath8 10@xmath46 gauss + h@xmath143 ( filament ) & 9.1 @xmath8 10@xmath98 gauss + @xmath60 ( core , synchrotron radn ) & @xmath16 8 @xmath8 10@xmath144 erg @xmath45 + @xmath60 ( filament , synchrotron radn ) & @xmath16 1 @xmath8 10@xmath42 erg @xmath45 + @xmath60 ( microwave background ) & 4 @xmath8 10@xmath42 erg @xmath45 + @xmath60 ( galaxy optical light ) & 4 @xmath8 10@xmath145 erg @xmath45 + 1 & 5.3 & 2.48 & @xmath146 & & & @xmath147 & @xmath77 & 0.07 + 2 & 2.2 & 2.48 & @xmath146 & @xmath148 & & @xmath147 & @xmath81 & 0.11 + 3 & 485.0 & 3.30 & @xmath149 & & @xmath150 & @xmath147 & @xmath151 & 0.01 + 4a & @xmath152 & 2.14 & & @xmath153 & & @xmath147 & @xmath154 & 1 + 4b & @xmath155 & 2.14 & & @xmath156 & & @xmath147 & @xmath154 & 1 + lc h@xmath143 & 2.3 @xmath8 10@xmath98 ( r@xmath157/1 arc sec)@xmath158 gauss + @xmath60 ( synchrotron radn)@xmath159 & @xmath16 1 @xmath8 10@xmath160 ( r@xmath157/1 arc sec)@xmath10 erg @xmath45 + @xmath60 ( microwave background ) & 4 @xmath8 10@xmath42 erg @xmath45 + @xmath60 ( galaxy optical light ) & @xmath16
8 @xmath8 10@xmath160 ( r/1 arc min)@xmath49 erg @xmath45 + lcccccccc any value@xmath161 & 1 & 1 & 7@xmath810@xmath99 & 2.3@xmath810@xmath98 & 0.03 + 23@xmath6 & 2.6 & 2.6 & 2@xmath810@xmath102 & 1.2@xmath810@xmath98 & 0.16 + 8@xmath6 & 7.2 & 7.2 & 6@xmath810@xmath102 & 6@xmath810@xmath102 & 1 + | we report x - ray observations of the nearby , powerful radio galaxy pictor a with the chandra observatory and optical and near uv observations of its western radio hot spot with the hubble space telescope .
x - ray emission is detected from the nucleus , a 19 ( 110 kpc ) long jet to the west of the nucleus , the western radio hot spot some 42 ( 240 kpc ) from the nucleus , and the eastern radio lobe .
the morphology of the western hot spot is remarkably similar to that seen at radio and optical wavelengths , where the emission is known to be synchrotron radiation .
the x - ray spectrum of the hot spot is well described by an absorbed power law with photon index @xmath0 = 2.07 @xmath1 0.11 .
the x - ray jet coincides with a weak radio jet and is laterally extended by @xmath2 20 ( 1.9 kpc ) .
the observed jet is up to @xmath2 15 times brighter in x - rays than any counter jet , a difference ascribed to relativistic boosting as the western radio lobe is probably the closer .
the jet s spectrum is well modelled by an absorbed power law with @xmath0 = 1.94@xmath3 and poorly fitted by a raymond - smith thermal plasma model .
the emission processes responsible for the x - rays are discussed in detail .
the radio - optical spectrum of the hot spot breaks or turns down at 10@xmath4 hz , and its x - ray spectrum is not a simple extension of the radio - optical spectrum to higher frequencies .
thermal models for the hot spot s x - ray emission are ruled out .
synchrotron self - compton models involving scattering from the _ known _ population of electrons give the wrong spectral index for the hot spot s x - ray emission and are also excluded .
a composite synchrotron plus synchrotron self - compton model can match the x - ray observations but requires similar contributions from the two components in the chandra band .
we show that the hot spot s x - ray emission could be synchrotron self - compton emission from a hitherto unobserved population of electrons emitting at low radio frequencies , but do not favor this model in view of the very weak magnetic field required .
an inverse compton model of the jet , in which it scatters microwave background photons but moves non - relativistically , requires a magnetic field a factor of @xmath2 30 below equipartition , and ad hoc conditions to explain why the radio lobes are fainter than the jet in x - rays but brighter in the radio .
these problems are alleviated if the jet moves relativistically , but models with an equipartition field require an implausibly small angle ( @xmath5 ) between the jet and the line of sight .
a model with @xmath5 @xmath2 23@xmath6 and a field a factor of 6 below equipartition seems viable .
synchrotron radiation is an alternative process for the x - ray emission .
the expected synchrotron spectrum from relativistic electrons accelerated by strong shocks and subject to synchrotron radiation losses is in very good agreement with that observed for both the hot spot and jet .
the possibility that the relativistic electrons result via photo - pion production by high energy protons accelerated in shocks ( a ` proton induced cascade ' ) is briefly discussed . |
ultraluminous x - ray sources are point - like x - ray sources in external galaxies with positions off the nucleus and apparent isotropic luminosities in excess of the eddington luminosity of a 10@xmath7 black hole ( @xmath8 ergs s@xmath1 ; for review see * ? ? ?
* ; * ? ? ?
ulxs are often found in actively starforming galaxies , e.g. m82 @xcite and the antennae galaxies @xcite .
there are statistically more ulxs in starburst galaxies compared to spiral galaxies @xcite .
@xcite investigated three star burst galaxies and found x - ray sources were preferentially near star clusters but displaced with higher luminosity sources nearer to clusters .
all these suggest that ulxs trace young stellar objects .
some surveys have reported higher numbers of ulxs in elliptical galaxies than in spirals @xcite , but the elliptical galaxies suffer more contamination from background sources @xcite . @xcite found that bright x - ray sources in elliptical galaxies have spectra consistent with galactic x - ray binaries in the high state , but that the number of sources in the ulx range is consistent with the number of expected background sources . however , some elliptical galaxies contain ulxs associated with globular clusters , e.g. ngc 1399 @xcite and ngc 720 @xcite , which strongly suggests an association with the elliptical galaxy .
lllcll[t ] & ( a ) & 320 & 3.5 & acis - i / none & 1999 oct 18 + & ( b ) & 319 & 55.9 & acis - s / none & 2000 jan 18 + & ( c ) & 239 & 3.6 & acis - i / none & 2000 jan 19 + ngc 1399 & ( d ) & 49898 & 13.1 & acis - s / hetg & 2000 jun 15 + & ( e ) & 240 & 43.5 & acis - s / hetg & 2000 jun 16 + & ( f ) & 2389 & 14.7 & acis - s / hetg & 2001 may 08 + & ( g ) & 4172 & 44.5 & acis - i / none & 2003 may 26 + & ( i ) & 315 & 72.2 & acis - s / none & 1999 dec 01 + & ( ii ) & 3040 & 69.0 & acis - s / none & 2001 dec 29 + & ( iii ) & 3043 & 67.1 & acis - s / none & 2002 apr 18 + the antennae & ( iv ) & 3042 & 67.3 & acis - s / none & 2002 may 31 + & ( v ) & 3044 & 36.5 & acis - s / none & 2002 jul 10 + & ( vi ) & 3718 & 34.7 & acis - s / none & 2002 jul 13 + & ( vii ) & 3041 & 72.9 & acis - s / none & 2002 nov 22 + ngc 1399 and the antennae are cases , respectively , of an elliptical galaxy and a starforming galaxy that host a large number of ulxs . @xcite compared _ chandra _ and _ hubble space telescope _
( _ hst _ ) images and found 70% of low mass x - ray binaries ( lmxbs ) in ngc 1399 were associated with globular clusters ( gcs ) ; the gc associated sources are brighter than field sources on average .
the x - ray source population in ngc 1399 may be unusual because a recent survey of lmxbs in six elliptical galaxies shows that , in general , there is no significant difference in luminosity between gc associated and field lmxbs @xcite .
several ulxs in ngc 1399 are detected but their natures are still unclear ; they could be imbhs or multiple lmxbs @xcite .
x - ray sources in the antennae have been well studied along with a series of observations from _ chandra _ @xcite .
comparison with optical images shows that the antennae x - ray sources are predominately associated with young star clusters @xcite .
the antennae ulxs have spectra consistent with ulxs in nearby galaxies and are not supernova remnants because they are variable and lack radio emission @xcite .
@xcite argued that imbhs could account for a few of the ulxs in the antennae but not most of them , which were more likely to be supercritical accreting stellar mass x - ray binaries due to their displacement from young star clusters @xcite .
the ulx x-37 in the antennae has been identified as a background quasar @xcite .
the most interesting ulxs are potential imbhs .
ngc 1399 and the antennae are two extremely different host environments for ulxs .
the comprehensive connection between ulxs and young stellar objects makes ulxs in ngc 1399 mysterious .
a comparison of ulxs between these two system may shed light on their natures .
we analyzed _ chandra _ archival data for both galaxies ( [ sec : obs ] ) in order to measure their long - term and short - term time variability ( [ sec : var ] ) and spectral variation ( [ sec : spec ] ) .
we discuss the results and the implications in
[ sec : diss ] .
we adopt the distance to ngc 1399 as 20.5 mpc @xcite and the distance to the antennae as 13.8 mpc @xcite .
this new distance to the antennae is obtained by measurement of the red giant branch tip and appears accurate to within 2 mpc , while the older distance was estimated from the recession velocity .
the new distance is more likely to be accurate .
we note that this revised distance to the antennae has reduced the antennae source luminosities by a factor of 4 relative to previous x - ray studies .
chandra has made seven observations of ngc 1399 during 1999 - 2003 and seven observations of the antennae during 1999 - 2002 .
a study of some of the antennae observations has previously been made @xcite .
the observations are listed in table [ tab : obs ] , in which we assigned letters to index observations of ngc 1399 and roman numerals to observations of the antennae .
only data from the advanced ccd imaging spectrometer ( acis ) were used in this paper .
observations ( a ) , ( c ) , ( g ) were made with the acis - i and ( b ) , ( d ) , ( e ) , ( f ) were from the acis - s but with the high energy transmission grating spectrometer ( hetgs ) applied on ( d ) , ( e ) and ( f ) .
observations from ( i ) to ( vii ) were all made with the acis - s detector .
level 2 events files were reduced from level 1 data using ciao 3.3 with caldb 3.2.0 .
we created images and exposure maps for each observation and searched for point - like sources with the wavdetect command .
exposure maps were produced by assuming a source spectrum of power - law form with a photon index of 1.5 @xcite and a galactic absorption along the line of sight of @xmath9 @xmath10 for ngc 1399 and @xmath11 @xmath10 for the antennae @xcite .
a power - law spectrum and a disk blackbody spectrum have similar luminosities in the 0.310 kev band , but will be quite different in the energy range up to 100 kev . since our data are only effective below 10 kev , an approximation with a single power - law spectrum is adequate . in ngc 1399 , we selected only sources that were located at least 20 arcsec off the nucleus and within the d25 circle ( 6.9 arcmin in diameter from the rc3 catalog ) . in the antennae
, we selected only sources in the active starforming region , which is bounded by a 1.4 arcmin radius circle . since the absolute astrometry accuracy of chandra is 0.6 arcsec , which is worse than the angular resolution , a correction for the aspect is necessary between observations before the same region could be applied in different observations for source selection . for ngc 1399 , we selected all detected sources ( @xmath12 ) in the region of interest as defined above in different observations . then we took observation ( b ) as a reference and searched in other observations for matched sources that were located within 1 arcsec of the selected source in ( b ) . a pointing offset
was calculated by averaging the offsets of all matched sources for each observation .
a similar correction was applied to observations of the antennae with observation ( i ) used as the reference . to examine source variability between observations ,
luminosity - luminosity ( l - l ) diagrams are presented for both galaxies .
first , we selected two observations and generated source lists using wavdetect .
then we matched sources and found their luminosities in both observations .
for an unmatched source in the list , we placed the same elliptical region in the other observation and calculated the net counts by subtracting the background level , estimated with wavdetect , from the total counts encircled by the ellipse .
then we derived the luminosity from the net counts normalized with the exposure map and the galaxy distance .
+ + there are three long observations for ngc 1399 : ( b ) , ( e ) , and ( g ) .
the l - l diagram between observation ( b ) and ( e ) is presented on the top panel of fig .
[ fig : lln ] and that between ( b ) and ( g ) is on the bottom panel of fig .
[ fig : lln ] . a lower threshold of @xmath13 erg s@xmath1 is placed to exclude most neutron stars as well as faint sources .
if the source is undetected with a significance of @xmath14 , a @xmath14 upper limit is shown .
[ fig : lla ] shows l - l diagrams for the antennae sources between observation ( i ) and all others .
we emphasize that we calculate luminosity from counting rate assuming a fixed spectrum .
our quoted luminosities are the counting rate multiplied by a roughly constant factor ( the factor is not exactly constant because different instrument modes were used for ngc 1399 and the detector response changed slowly for the antennae ) .
because we are investigating the variability of individual sources , our comparisons of luminosities is very similar to a comparison of counting rates .
if the assumed spectra form for a given source is incorrect , then the luminosities will be incorrect , but all luminosities for any individual source would be shifted by the same factor ( if the spectral shape of an individual source varies between observations , then that would be evidence for variability ) .
we choose to present luminosities instead of count rates , because the former are easier to compare with measurements of other sources .
we note that an estimate of the true luminosity depends on the spectral form assumed and the energy band used .
the bolometric luminosity of the sources is likely larger than our quoted values , since emission is likely present both above and below the 0.3 - 10 kev band . to determine the magnitude of the uncertainty on the luminosity in our chosen energy band , we fit each source that has enough photons with an absorbed power - law and measured its intrinsic luminosity ( the photon index and column density
are allowed to vary ; see details in section [ sec : spec ] ) .
then we fit the spectrum and obtained another estimate of the luminosity by fixing the absorption to the galactic value and the photon index to 1.5 .
the ratio between these two luminosities varies between [email protected] for bright x - ray sources in ngc 1399 and 1@xmath62 in the antennae .
therefore , a bilateral systematic error of @xmath15 is derived .
we also checked with xspec that the systematic error of @xmath15 could cover spectral changes from @xmath16 for a single power - law spectrum , @xmath17 kev for a single disk blackbody spectrum , or a reasonable combination of them ; the column density was fixed at the galactic value for ngc 1399 but was varied from the galactic value to 10@xmath18 @xmath19 for the antennae .
that means this systematic error is valid for black hole binaries at most spectral states that have been observed in our galaxy .
the systematic error is shown as two dashed lines in the l - l diagram .
sources lying between the two dashed lines are considered constant within the accuracy of our measurement .
lightcurves for the ulxs , which we define as having a luminosity in 0.3 - 10 kev band above @xmath20 ergs s@xmath1 in at least one observation , and for the six most variable sources are shown in fig .
[ fig : lcn ] for ngc 1399 and in fig .
[ fig : lca ] for the antennae .
the variability is defined as the maximum ratio in luminosities , which could be a number with a standard error if the lowest luminosity is detected or a 90% lower bound if the lowest luminosity is an upper limit .
the luminosities in the lightcurves are derived slightly differently than those in the l - l diagram .
first we generate a complete source list , which is based on detected sources in the reference observation ( ( b ) or ( i ) ) , as well as sources only detected in other observations .
we place the same circular region for a given source in all observations and calculate the luminosity .
this difference in technique causes a mild difference with the luminosities in the l - l diagrams but is consistent within the errors .
there are 8 ulxs found in ngc 1399 , psx-1 , psx-2 , psx-4 , psx-7 , psx-3 , psx-21 , psx-25 and psx-6 ( names are from * ? ? ?
* ) , sorted by the maximum luminosity . among them , psx-1 , psx-2 , psx-25 and psx-6 are found to be associated with gcs .
the gc identifications are from the _ hst _ image in @xcite and the source list of @xcite .
there are 6 ulxs detected in the antennae , x-16 , x-11 , x-44 , x-42 , x-37 and x-29 ( names are from * ? ? ?
* ) , sorted in the same way .
we note that the antennae x-37 has been identified as a background quasar @xcite and x-29 is the nucleus of ngc 4039 @xcite .
we excluded these two ulxs as well as the nucleus of ngc 4038 ( x-25 ) from all analysis in our paper .
for all the x - ray sources in ngc 1399 and the antennae , we calculated their variability defined as described above .
we defined the variability function for a population as the number of sources with a maximum variability larger than a certain value .
the variability functions for all bright sources ( @xmath2 ergs s@xmath1 in 0.3 - 10 kev ) are shown in the top panel of fig .
[ fig : vfun ] .
the fraction of bright sources which are variable ( with a confidence level of 90% above the systematic error of @xmath15 ) is 10/63 for ngc 1399 and 9/15 for the antennae .
the fraction of ulxs which are variable is 1/8 for ngc 1399 and 4/4 for the antennae . for ngc 1399
, the only variable ulx is psx-21 .
the color of the optical counterpart ( @xmath21 , washington @xmath22 and kron @xmath23 , @xcite ) for this source is inconsistent with that of a globular cluster , but the color may be affected by a defect in the @xmath23-band image ( t. richtler private communication ) . of the other 9 variable sources , only two have optical counterparts within @xmath24 .
therefore , the majority of the variable sources are not associated with globular clusters .
this is in sharp contrast to the fact that 70% of all sources in ngc 1399 are associated with globular clusters @xcite .
the detection threshold in all observations is well below our ulx luminosity threshold .
however , the detection thresholds in the shortest observations of ngc 1399 is close to our luminosity cutoff of @xmath13 ergs s@xmath1 for bright sources .
we examined two long observations spaced by similar periods , i.e. , ( b ) and ( g ) for ngc 1399 and ( i ) and ( vii ) for the antennae , to check if the difference in variability between ngc 1399 and the antennae still persists when only observations with detection thresholds well below our luminosity cutoff are used .
we found that 7 of the 63 bright sources in ngc 1399 and 4 of the 15 bright sources in the antennae are variable .
this is consistent with the conclusions that bright x - ray sources in ngc 1399 are less variable than those in the antennae .
+ to compare with the variability function of x - ray binaries with known characteristics , we selected 14 galactic black hole lmxbs32 , gro j1655@xmath2540 , grs 1009@xmath2545 , gs 1124@xmath25684 , gs 2000@xmath3250 , gs 2023@xmath3338 , gx 339@xmath254 , v4641 sgr , x0620@xmath25003 , x1543@xmath25475 , x1705@xmath25250 , xte j1118@xmath3480 , xte j1550@xmath25564 , xte j1859@xmath3226 ] and 25 neutron star lmxbs238 , gx 13@xmath31 , gx 17@xmath32 , gx 3@xmath31 , gx 340@xmath30 , gx 349@xmath32 , gx 354@xmath250 , gx 5@xmath251 , gx 9@xmath31 , gx 9@xmath39 , lmc x-2 , sco x-1 , ser x-1 , x0614@xmath3091 , x1254@xmath25690 , x1543@xmath25624 , x1608@xmath25522 , x1705@xmath25440 , x1735@xmath25444 , x1812@xmath25121 , x1820@xmath25303 , x1822@xmath25000 , x2127@xmath3119 ] and calculated their variability functions using data from the all sky monitor ( asm ) onboard the _ rossi x - ray timing explorer _ ( _ rxte _ ) .
for each source , we got one variability value by sampling the asm one day average lightcurve spaced in the same way as those seven _
_ observations of ngc 1399 .
then we shifted the first date to 10 days later and got another value until there were 100 values .
the variability of the source was derived from an average of these 100 values and then the variability functions were obtained , respectively , for sources of black hole and neutron star lmxbs .
we sampled the asm data spaced in the same way as the antennae observations and the results looked similar .
we also examined the variability of two individual galactic sources : grs 1915@xmath3105 and 4u 1820@xmath2530 .
the galactic black hole binary grs 1915@xmath3105 has a luminosity in the ulx range and is unusual in that it has been a persistent x - ray emitter for 14 years @xcite .
it is also the most massive stellar mass black hole that has been found so far .
the bright neutron star lmxb 4u 1820@xmath2530 ( @xmath26x1820@xmath25303 ) residing in the galactic gc ngc 6624 has a similar environment with sources in ngc 1399 and is an ultracompact binary , though its luminosity is low relative to the sources considered here .
the variability function for each of these sources was obtained from 100 asm flux measurements .
the normalized variability functions of selected black hole and neutron star lmxbs , grs 1915@xmath3105 , 4u 1820@xmath2530 , ngc 1399 and the antennae are shown together in the bottom panel of fig .
[ fig : vfun ] .
+ to estimate the typical outburst duration of sources in ngc 1399 , we used the two deepest observations of ngc 1399 .
these are ( b ) and ( g ) and are separated by @xmath27 yrs .
if a source was detected in both observations , we assume the source lasted in its outburst state for @xmath28 .
we found 60 bright sources and 8 ulxs had at least one detection in either ( b ) or ( g ) , and among them 50 bright sources and 7 ulxs had two . assuming a fixed outburst duration @xmath29 for x - ray sources in ngc 1399 , we estimated that @xmath30 , where @xmath31 was the fraction of twice detected sources among sources with at least one detection .
we thus derived @xmath32 yrs for bright sources and @xmath33 yrs for ulxs . to search for short - term variability for every x - ray source in each observation
, we used the k - s test to examine if the 0.3 - 10 kev x - ray count rate is constant .
no source in ngc 1399 shows any short - term variability above a significance level of 3@xmath34 . however , four antennae x - ray sources , including two ulxs , are variable with a significance level above 3@xmath34 .
a 3.8@xmath34 probability of non - constant emission is shown for the antennae x-16 in observation ( v ) , and an 8.0@xmath34 level for x-42 in observation ( vii ) . the total number of trials , equal to the number of detected sources multiplied by the number of observations is 105 . taking into account the number of trials ,
the variability from x-16 is significant at the 98.7% level , while the variability from x-42 is highly significant with a chance probability of occurrence of @xmath35 .
we examined the light curves in several bands and found that the variability is most pronounced in the 1 - 7 kev band ; there is no significant variability below 1 kev . the 1 - 7 kev lightcurves for x-16 and x-42 and the power spectrum density ( psd ) for x-42 are shown in fig .
[ fig : psd ] . strong variability at a time scale around 20 ks in x-42 is explicit .
the psd looks like a flat form that breaks at @xmath36 hz , which is similar to the break psds of galactic black holes .
however , the power spectrum is dominated by a single flare with a duration of about 5 ks about halfway through the observation . therefore , it would be imprudent to over interpret the power spectrum and attempt to derive a mass for the compact object .
we note that no significant power is detected in any other observation .
we examined all the x - ray spectra for ulxs in both galaxies with xspec 11.3.1 .
all sources in ngc 1399 and the antennae could be reasonably described by an absorbed power - law , whereas some sources in the antennae may show line features but with low significance ( see @xcite ) .
two sources ( psx-2 and psx-4 ) in ngc 1399 present an alternative favored fitting with a multicolor disk ( mcd ) blackbody ( diskbb in xspec ) model ( see fig . [ fig : diskbb ] ) .
all fittings adopt an absorption with solar abundance ( wabs in xspec ) as a free parameter but setting the lower bound to the galactic value .
fitting parameters of sources psx-2 and psx-4 are shown in table [ tab : psx ] .
the best fitted 0.3 - 10 kev luminosity versus photon index in every observation that contains enough photons are plotted in fig .
[ fig : nlg ] for ngc 1399 ulxs and fig .
[ fig : alg ] for the antennae ulxs .
the best fitted column densities for ulxs in ngc 1399 are consistent within 3@xmath34 with the galactic value , except for psx-2 in observation ( b ) which differs by nearly 6@xmath34 .
the best fitted column densities for ulxs in the antennae are shown on each plot panel in fig .
[ fig : alg ] . all photon indices and luminosities in fig .
[ fig : nlg ] & [ fig : alg ] are obtained by fitting a single power - law model with absorption .
lllll[!h ] psx-2 & diskbb & @xmath37 & @xmath38 & @xmath39 + & powerlaw & @xmath40 & @xmath41 & @xmath42 + psx-4 & diskbb & @xmath43 & @xmath44 & @xmath45 + & powerlaw & @xmath46 & @xmath47 & @xmath48 + + in fig .
[ fig : nlg ] , a weak correlation between the luminosity and the power - law photon index is found for all ulxs except psx-1 in ngc 1399 with a correlation coefficient of 0.66 and a probability of 3.7% for a coincidence from uncorrelated data .
this correlation does not exist in all ulxs in the antennae , but in observations for individual sources x-11 , x-44 and x-42 , the correlation coefficients with chance probabilities in brackets are 0.61(14.6% ) , @xmath250.57(17.5% ) and @xmath250.66(10.6% ) , respectively .
these correlations are weak and more observations are needed to expand the samples .
the most dramatic difference between ngc 1399 and the antennae is that the ulxs as well as the bright x - ray sources in the elliptical galaxy are less variable . in fig .
[ fig : lln ] , only a couple sources show apparent variability in the l - l diagram of ngc 1399 between ( b)-(e ) or ( b)-(g ) observations . in fig .
[ fig : lla ] , the number of sources in the antennae with significant variability is obviously larger than that in ngc 1399 .
this is verified by lightcurves of ulxs and the variability functions of bright sources .
the variability functions of galactic lmxbs show that most black holes are highly variable , with the notable exception of grs 1915@xmath3105 , whereas bright neutron stars are less variable .
the normalized variability function ( bottom panel of fig . [ fig : vfun ] ) presents the bright sources in ngc 1399 as similar to 4u 1820@xmath2530 , a little less variable than the group of neutron star lmxbs and grs 1915@xmath3105 , and far less variable than the group of black hole lmxbs .
ulxs in ngc 1399 are even less variable than other bright sources , with 7 of 8 showing constant flux and the other having little variability .
none of the bright x - ray sources or ulxs in ngc 1399 show short - term variability with a significance level of 3@xmath34 in any individual observation .
the lack of variability should provide a strong constraint on the nature of these sources .
a natural way to produce low variability would be if the bright x - ray sources in ngc 1399 consist of multiple unresolved lmxbs .
however , a large number of sources is required . from the asm data ,
we estimated that neutron star lmxbs are above 10% of their maximum flux 80% of the time , while black holes are above their 10% of their maximum flux only 15% of time ( the estimate for black holes is an upper bound because several of our selected black hole systems have had only one outburst during the lifetime of rxte ) . using the average luminosity of 4u 1820@xmath2530 of @xmath49 , about 8 neutron star binaries would be needed to produce the least luminous of our bright sources and 50 would be needed to produce the least luminous ulx .
large numbers of neutron stars are known to be present in some globular clusters in our galaxy , e.g. there are around 50 in the galactic gc ngc 6266 @xcite , but only very rarely does a cluster contain more than one bright x - ray source .
it is very puzzling to understand how individual clusters in ngc 1399 could contain so many bright x - ray sources .
though the luminosity of black holes in the accreting phase is , generally , several times larger than that of neutron stars , the smaller duty cycle partially offsets the increased luminosity and comparable numbers of black hole binaries are needed to reach the same luminosity .
perhaps , the most promising interpretation of the ngc 1399 bright sources is that they are are individual black holes of a type similar to grs 1915@xmath3105 .
as noted above , grs 1915@xmath3105 has been a persistent x - ray source since its discovery 14 years ago @xcite and has reached luminosities in the ulx range .
the variability function of grs 1915@xmath3105 is similar to that of the ngc 1399 bright sources .
the companion star in grs 1915@xmath3105 is a k or m giant @xcite .
the time scale for the evolution of such a binary system to the mass - transferring phase is roughly 6 gyr @xcite .
allowing for different companion star masses , similar mass transferring binaries could currently be present in either the old ( @xmath50 gyr ) or young ( @xmath51 gyr ) globular clusters in ngc 1399 @xcite .
@xcite has previously suggested that the ulxs in ellipticals may be similar to grs 1915@xmath3105 .
@xcite suggested that the bright x - ray sources in ellipticals are x - ray binaries with roche - lobe filling giant companions with mass transfer driven by nuclear evolution of the companion .
they predicted outburst durations similar to what we find for ngc 1399 .
for the sources associated with globular clusters , another possibility is that the companion stars are initially isolated stars which are captured by the black hole after evolving to the giant phase .
the capture cross - section is enhanced for stars in the giant phase .
however , grs 1915@xmath3105 shows strong variability on short time scales , which we do not observe from the bright x - ray sources in ngc 1399 .
the sensitivity of chandra to short timescale variability is limited due to the low counting rate , but our detection of short timescale variability from sources in the antennae ( with comparable counting rates ) indicates that similar flares would have been detected from the ngc 1399 sources .
the short timescale variability of sources in ngc 1399 should be investigated in more detail with new observations .
@xcite has suggested that the bright x - ray sources in ngc 1399 may be neutron stars accreting from white dwarf companions in ultracompact binaries , similar to 4u 1820@xmath2530 .
the variability of 4u 1820@xmath2530 is similar to the ngc 1399 bright sources .
hence , neutron star ultracompact binaries may account for some or all of the bright sources if mildly super - eddington ( apparent ) luminosities can be achieved .
however , the short lifetimes expected for the high accretion rates required given the very low - mass companions ( @xmath52 ) make it unlikely that the ulxs in ngc 1399 are neutron star ultracompact binaries .
another possible interpretation of the bright sources is that they are neutron - star binaries similar to the neutron - star x - ray binaries known as z sources .
z sources are persistent with variability similar to that seen from the ngc 1399 sources .
again , mildly super - eddington ( apparent ) luminosities would be required to produce the dimmest of the bright sources and such objects are unlikely to explain the ulxs .
interestingly , @xcite suggested that the z sources have giant star companions with mass loss driven by nuclear evolution .
this would make the systems the neutron star analogs of black hole / giant star binaries such as grs 1915@xmath3105 .
the three most luminous x - ray sources in ngc 1399 are interesting .
the brightest one , psx-1 , has a hard spectrum ( @xmath53 ) , little variability ( @xmath54 ) , and a high luminosity around @xmath4 ergs s@xmath1 .
the spectrum indicates the source is not in its high / soft state ( @xmath55 ) but at the hard extreme of the low / hard state ( @xmath56 ) @xcite .
the low variability indicates that the source is probably not relativistically beamed @xcite , because relativistically beamed sources are expected to be highly variable .
this source is associated with a bright gc with @xmath57 , @xmath58 @xcite , which means it is unlikely to be a background agn .
a huge number of sources would be required to explain the emission as a collection of neutron star or black hole x - ray binaries .
if the source is an imbh in the low state , its practically constant luminosity is reasonable if compared to the long - term behavior of m82 x-1 , which varies in luminosity by a factor of 2 - 3 @xcite and shows a hard spectrum with a power - law index @xmath59 @xcite .
ngc 1399 psx-2 and psx-4 display typical behaviors of black holes in the high / soft state .
psx-2 is associated with a bright gc ( @xmath60 , @xmath61 ) but psx-4 is not .
the plausible cool disk ( @xmath62 kev ) emission is consistent with the soft spectral index ( @xmath63 ) .
their low variability ( @xmath64 for psx-2 and @xmath65 for psx-4 ) is suggestive of some galactic black holes in the high state , but no galactic source has been found to remain in the high state for four years @xcite . the cool disk has been a useful probe in finding imbh candidates @xcite .
however , their disk temperatures are somewhat higher than most imbh candidates , and their luminosities ( @xmath66 ergs s@xmath1 ) are also around the lower bound of those candidates @xcite .
these sources do not have high mass companion stars to produce geometrically beamed emission as suggested by @xcite .
therefore , it could be possible that these two sources are 10@xmath6100@xmath7 black holes emitting at the high state if the disk components are true . one possible way to determine the reality of the disk emission is an accurate measurement of the column density , since these two sources , in particular psx-2 , show significantly higher intrinsic absorption if modeled by a single power - law spectrum .
the bright x - ray sources in the antennae are variable .
this suggests that they are black hole binaries and accreting from massive stars to produce unstable disks @xcite .
they are not as variable as black hole lmxbs ( fig .
[ fig : vfun ] ) . considering the variability together with their high luminosity
, we suggest they are hmxbs accreting via roche - lobe overflow . there is no disk emission found in any of the 4 ulxs .
the most luminous source , x-16 , manifests a very hard spectra ( @xmath67 ) and has strong variability with a change of luminosity of @xmath610 ( fig .
[ fig : alg ] ) .
the source could be an imbh with a mass of a few hundred to thousand @xmath7 emitting at the hard state .
the x - ray behavior is also consistent with a relativistically beamed jet pointing at us , but @xcite pointed out that there is no radio emission found from any ulx in the antennae .
it is interesting that none of the ulxs shows the very soft spectrum normally seen in ulxs in nearby galaxies @xcite .
we note that x-37 was reported to have a soft component in its spectrum , but has been subsequently identified as a background agn . x-11 and x-42 are likely to be in the hard state ( @xmath68 ) , x-44 looks like it is between the hard and soft states ( @xmath69 ) . a plausible correlation ( @xmath70 ) is found in x-11 between luminosity and @xmath5 , which shows similar behaviors as stellar mass black holes , but we should be cautious because of the weak correlation coefficient .
the x - ray properties of these three sources are consistent with hmxbs , but whether they are imbhs or @xmath620@xmath7 black holes emitting in near or super eddington luminosities is still unclear .
we thank tom richtler for providing us the gc list in ngc 1399 and giving instructions on the gc identification and an anonymous referee for comments which improved our interpretation of the results .
hf thanks zach prieskorn for helping correct the english in the paper .
we acknowledge support from nasa grants hst - go-10001 and chandra grant go4 - 5086a .
pk acknowledges support from a university of iowa faculty scholar award . | the temporal and spectral properties of ultraluminous x - ray sources ( ulxs , @xmath0 ergs s@xmath1 ) and bright x - ray sources ( @xmath2 ergs s@xmath1 ) are examined and compared in two extremely different host environments : the old elliptical galaxy ngc 1399 and the young , starforming antennae galaxies ( ngc 4038/4039 ) .
ulxs in ngc 1399 show little variability on either long or short time scales . only 1 of 8 ulxs and 10 of 63 bright sources in ngc 1399 are variable at a confidence level of 90% . on long timescales , the ngc 1399 sources are steadier than most galactic black hole x - ray binaries , but similar to grs 1915@xmath3105 .
the outburst duration of the ngc 1399 sources is about 20 yrs , again , similar to that of grs 1915@xmath3105 .
the bright x - ray sources in ngc 1399 may be black hole x - ray binaries with giant star companions similar to grs 1915@xmath3105 .
the brightest ulx ( psx-1 ) in ngc 1399 is coincident with a globular cluster , shows a hard spectrum with a photon index around 1.5 , and has a nearly constant luminosity around @xmath4 erg s@xmath1 .
it may be an intermediate - mass black hole ( imbh ) in a hard spectral state .
in contrast to ngc 1399 , the ulxs in the antennae are all variable and a large fraction of the bright sources ( 9 of 15 ) are also variable .
the variability and luminosity of ulxs in the antennae suggest they are black hole high mass x - ray binaries accreting via roche - lobe overflow .
a flare with a duration of about 5 ks is found from antennae x-42 .
the most luminous ulx , x-16 , with a very hard spectrum ( @xmath5 = [email protected] ) and a luminosity which varies by a factor of 10 , could be an imbh candidate . |
the effect of weak quenched disorder on large - scale properties and phase diagrams of many - body systems attracted considerable attention for decades . despite significant efforts ,
there still remain many open questions .
the prominent example is the @xmath0 model in which an @xmath8-component order parameter ( the magnetization in the spin system notation ) is coupled to a symmetry - breaking random field . for @xmath9 , it is known as the random field ising model ( rfim).@xcite for @xmath10 , one has to distinguish the random field ( rf ) case where the order parameter couples linearly to disorder , and the random anisotropy ( ra ) case , where the coupling to disorder is bilinear .
these models are relevant for a variety of physical systems such as amorphous magnets,@xcite diluted antiferromagnets in a uniform external magnetic field,@xcite liquid crystals in porous media,@xcite critical fluids in aerogels,@xcite nematic elastomers,@xcite and vortex phases of impure superconductors .
@xcite for @xmath9 , the ra model reduces to the random - temperature model , where the randomness couples to the local energy density as , for example , in diluted ferromagnets.@xcite in contrast to the systems with random - temperature - like disorder , the rf and ra models suffer from the so - called dimensional reduction ( dr ) . a straightforward analysis of the feynman diagrams giving the leading singularities yields to all orders that the critical behavior of the rf @xmath0 model in @xmath11 dimension is the same as that of the pure system in @xmath12 dimensions.@xcite consequently
, the lower critical dimension is @xmath13 for ising - like systems and @xmath14 for systems with continuous symmetry .
this can elegantly be demonstrated using supersymmetry.@xcite however , simple imry - ma arguments show that the lower critical dimension of the rfim is @xmath15.@xcite the deviation from dr is also confirmed by the high - temperature expansion.@xcite thus dr breaks down , rendering standard field theoretic methods useless .
another known problem where the perturbation theory is spoiled by dr is elastic manifolds in disordered media .
there , two methods were developed to overcome difficulties related to dr : the gaussian variational approximation ( gva ) in replica space and the functional renormalization group ( frg ) .
the gva is supposed to be exact in the limit @xmath16 .
unfortunately , this approach when applied to the rf problem leads to very complicated equations which do not allow us to compute the critical exponents.@xcite considering the rf @xmath0 model , fisher @xcite showed that expansion around the lower critical dimension @xmath17 generates an infinite number of relevant operators which can be parametrized by a single function .
however , he found that the corresponding one - loop frg equation has no analytic fixed point ( fp ) solution . only recently , using the progress in the elastic manifold problem,@xcite it was realized that the scaling properties of systems exhibiting metastability are encoded in a nonanalytic fp .
feldman @xcite has shown that , indeed , in @xmath18 and for @xmath19 ( a more precise computation@xcite gives @xmath20 ) , there is a nonanalytic fp with a cusp at the origin .
this fp provides the description for the ferromagnetic - paramagnetic phase transition in the rf @xmath0 model and allows one to compute the critical exponents which are different from the dr prediction .
recently , le doussal and wiese @xcite extended the frg analysis to two - loop order .
the extension beyond one - loop order is highly nontrivial due to the nonanalytic character of the renormalized effective action , which leads to anomalous terms in the frg equation.@xcite the two - loop calculations were also independently performed in ref . , and the truncated exact frg was proposed in ref .
. the more accurate analysis of the frg flows for the rf model showed that for @xmath21 there is a crossover to a fp with weaker nonanalyticity resulting in the dr critical exponents.@xcite a similar picture was found for the ra @xmath0 model with the main difference that @xmath22 and @xmath23.@xcite a more peculiar issue concerns the phase diagram of the rf and ra models below @xmath24 .
it is known that for the rf model and models with isotropic distributions of random anisotropies true long - range order is forbidden below @xmath17 ( for anisotropic distributions , long - range order can occur even below @xmath24).@xcite however , quasi - long - range order ( qlro ) with zero order parameter and an infinite correlation length can persist even for @xmath25 . for instance , the gva predicts that the vortex lattice in disordered type - ii superconductors can form the so - called bragg glass exhibiting slow logarithmic growth of displacements.@xcite this system can be mapped onto the rf @xmath26 model , in which the bragg glass corresponds to the qlro phase .
indeed , for @xmath27 and @xmath25 , the frg equations have attractive fps which describe the qlro phases of rf and ra models .
@xcite in order to study the transition between the qlro phase and the disordered phase expected in the limit of strong disorder , one has to go beyond the one - loop approximation .
the truncated exact frg ( ref . ) and the two - loop frg ( ref . ) performed using a double expansion in @xmath28 and @xmath29 give access to a different singly unstable fp which is expected to control the transition .
both methods give qualitatively similar pictures of the frg flows : the critical and attractive fps merge in some dimension @xmath30 that is therefore the lower critical dimension of the qlro - disordered transition .
for the rf @xmath26 model , both methods predict approximately the same lower critical dimension @xmath31 , and thus , suggest that there is no bragg glass phase in @xmath32 .
however , one has to take caution when extrapolating results obtained for small @xmath28 and @xmath29 .
moreover , in contrast to the model of refs . and which belongs to the so - called `` hard - spin '' models , the system studied in ref . corresponds to `` soft spins '' , and thus , is expected to belong to a different universality class . in terms of vortices , the soft spin rf model allows for topological defects which destroy the bragg glass .
most studies of the rf and ra models are restricted to either short - range ( sr ) correlated disorder or uncorrelated pointlike defects .
however , real systems often contain long - range ( lr ) correlated disorder or extended defects in the form of linear dislocations , planar grain boundaries , three - dimensional cavities , etc .
systems with anisotropic orientation of extended defects can be described by a model in which all defects are strongly correlated in @xmath33 dimensions and randomly distributed over the remaining @xmath34 dimensions .
the case @xmath35 is associated with uncorrelated pointlike defects , while extended columnar or planar defects are related to the cases @xmath36 and 2 , respectively .
the critical behavior of the @xmath0 model with random - temperature - like extended defects was studied in refs . using a perturbative rg analysis in conjunction with a double expansion in @xmath37 and @xmath33 . in the case of an isotropic distribution of disorder ,
power - law correlations are the simplest example with the possibility for a scaling behavior with new fps and new critical exponents .
the critical behavior of systems with random - temperature disorder correlated as @xmath38 for large separation @xmath2 was studied in refs . .
the power law can be ascribed to extended defects of internal dimension @xmath4 , distributed with random orientation in @xmath11 dimensional space . in general , one would probably not expect a pure power - law decay of correlations .
however , if the correlations of defects arise from different sources with a broad distribution of characteristic length scales , one can expect that the resulting correlations will , over several decades , be approximated by an effective power law.@xcite power - law correlations with a noninteger value @xmath4 can be found in systems containing defects with fractal dimension @xmath4 .
@xcite for example , the behavior of @xmath39 in aerogels is argued to be described by an xy model with lr correlated defects .
@xcite this is closely related to the behavior of nematic liquid crystals enclosed in a single pore of aerosil gel which was recently studied in ref . , using the approximation in which the pore hull is considered a disconnected fractal .
the frg was used to investigate the statics and dynamics of elastic manifolds in media with lr correlated disorder in ref . .
the critical behavior of the @xmath0 model with lr correlated rf was studied in refs . .
however , the methods used in these works fail to describe properly the case of sr correlated rf .
therefore , there is a necessity to reexamine the critical behavior of the lr rf model using methods which are successful in the sr case . in the present paper ,
we study the lr correlated rf and ra @xmath0 models using the frg to one - loop order .
the paper is organized as follows . in sec .
[ sec - model ] , we introduce the lr rf and ra models , and derive the frg equations . in sec .
[ sec - rf ] , we study the lr rf model . in sec .
[ sec - ra ] , we consider the lr ra model and discuss the application to superfluid @xmath7he in aerogels .
the final section summarizes our results .
the large - scale behavior of the @xmath0 symmetric spin systems at low temperatures can be described by the nonlinear @xmath4 model with the hamiltonian @xmath40 = \int d^d x \left [ \frac{1}{2 } ( \nabla \vec{s}\,)^2 + v(x,\vec{s } ) \right],\end{aligned}\ ] ] where @xmath41 is the @xmath8-component classical spin with a fixed - length constraint @xmath42 .
@xmath43 is the random disorder potential , which can be expanded in spin variables as follows : @xmath44 the corresponding coefficients have simple physical interpretation : @xmath45 is a random field , @xmath46 is a random second - rank anisotropy , and @xmath47 are general @xmath48th tensor anisotropies . as was shown in ref .
, even if the system has only finite number of nonzero @xmath47 , the rg transformations generate an infinite set of high - rank anisotropies preserving the symmetry with respect to rotation @xmath49 if it is present in the bare model .
for instance , starting with only second - rank anisotropy corresponding to the ra model , all even - rank anisotropies will be generated by the rg flow .
we will reserve the notation ra for the systems which have this symmetry and the notation rf for the systems which do not . in the present work ,
we consider the case of gaussian distributed long - range correlated disorder with zero mean and cumulants given by @xmath50,\end{aligned}\ ] ] with @xmath51 .
for the sake of convenience , we fix the constant in fourier space , taking @xmath52 .
to average over disorder , we introduce @xmath53 replicas of the original system and compute the replicated hamiltonian @xmath54,\ \ \ \ \ \label{hn}\end{aligned}\ ] ] where @xmath55 .
the properties of the original disordered system ( [ h0 ] ) and ( [ v0 ] ) can be extracted in the limit @xmath56 . according to the above definition of the rf and ra models , the functions @xmath57 are arbitrary in the case of the rf and even for the ra .
power counting suggests that @xmath58 is the lower critical dimension for both models.@xcite at criticality or in the qlro phase , the correlation functions of the order parameter exhibit scaling behavior .
in contrast to the models with temperature - like disorder in the models under consideration , the connected and disconnected correlation functions may scale with different exponents .
this reflects metastability and the breaking of the dr .
for instance , the connected two - point function behaves as @xmath59 while the disconnected function scales as @xmath60 here , @xmath61 is the fourier component of the order parameter and the angle brackets stand for the thermal averaging .
schwartz and soffer@xcite proved that the exponents of the rf model obey the inequality @xmath62 . since in the ra case
the coupling to disorder is bilinear , the schwartz - soffer inequality can not be applied directly to @xmath63 and @xmath64 . in the ra case , it is convenient to introduce the correlation functions of the form eqs .
( [ cor - fun - con ] ) and ( [ cor - fun - dis ] ) not for @xmath65 , but for the field @xmath66 ^ 2 - 1/n$ ] , and define exponents @xmath67 and @xmath68 which satisfy the schwartz - soffer type inequality : @xmath69 . @xcite vojta and schreiber@xcite generalized this inequality to correlated rf and obtained a more restrictive bound @xmath70 since @xmath71 . for ra models
similar arguments lead to @xmath72 . to derive the one - loop frg equations
, we straightforwardly generalize the methods developed in refs . , and to model ( [ hn ] ) .
we express the order parameter @xmath73 as a combination@xcite @xmath74 of a fast field @xmath75 fluctuating at small scales @xmath76 which is orthogonal to a slow field @xmath77 of unit length , changing at scales @xmath78 . here , @xmath79 is the uv cutoff and @xmath80 .
the field @xmath77 can be considered as the coarse - grained order parameter ( local magnetization ) whose fluctuations at low temperature are small , @xmath81 . integrating out the fast variables @xmath75 , we rescale in such a way that the effective hamiltonian of the slow fields @xmath77 would have the structure of the bare hamiltonian ( [ hn ] ) .
it is convenient to change variable to @xmath82 .
the frg equations to first order in @xmath6 and @xmath4 are given by @xcite @xmath83 ^ 2 - a r_1 '' ( \phi)\nonumber \\ & & - ( n-2 ) \bigg\ { 2a r_1(\phi ) + a r_1'(\phi)\cot\phi \nonumber \\ & & - \frac1{2\sin^2{\phi } } \big [ { r_1'(\phi ) + r_2'(\phi ) } \big]^2 \bigg\ } , \label{frg-1 } \\
\partial_{\ell } r_2(\phi ) & = & - ( \varepsilon-\sigma ) r_2(\phi ) - \big\ { ( n-2)\big[2r_2(\phi ) \nonumber \\ & & + r_2'(\phi ) \cot\phi\ , \big ] + r_2''(\phi)\big\ } a , \ \
\label{frg-2}\end{aligned}\ ] ] where @xmath84 .
we have absorbed the factor of @xmath85 in redefinition of @xmath86 and introduced @xmath87 in terms of the variable @xmath88 , the functions @xmath89 become periodic with period @xmath90 in the rf case and @xmath91 in the ra case .
the flow equation for the temperature to one - loop order reads @xmath92 according to eq .
( [ t ] ) , the temperature is irrelevant for @xmath93 and sufficiently small @xmath94 .
although we expect @xmath95 in the vicinity of a fp , one has to take caution whether the found fp survives in three dimensions.@xcite the scaling behavior of the system is controlled by a zero - temperature fp of eqs .
( [ frg-1 ] ) and ( [ frg-2 ] ) @xmath96 $ ] , such that @xmath97 .
an attractive fp describes a phase , while a singly ( unidirectionally ) unstable fp describes the critical behavior .
the critical exponents are determined by the frg flow in the vicinity of the fp and to one - loop order are given by @xmath98 it is convenient to introduce the following reduced variables : @xmath99 to check the stability of the fp @xmath100 $ ] , we linearize the flow equations around this fp : @xmath101 and obtain @xmath102 \big[y_1''(\phi ) + y_2''(\phi ) \big]\nonumber \\ & & - a y_1''(\phi ) - a_0 r_1^{*\prime\prime}(\phi ) -(n-2 ) \big\ { 2a_0 r_1^*(\phi ) \nonumber \\ & & + a_0 r_1^{*\prime}(\phi)\cot\phi + 2a y_1(\phi ) + a y_1'(\phi)\cot\phi \nonumber \\ & & - \big [ r_1^{*\prime}(\phi ) + r_2^{*\prime}(\phi ) \big ] \nonumber \\ & & \times \big [ y_1'(\phi ) + y_2'(\phi ) \big]/\sin^2{\phi } \big\ } , \ \ \ \ \ \ \ \ \ \label{frgl-1 } \\ \lambda
y_2(\phi ) & = & - y_2(\phi ) - a_0 \big\ { ( n-2)\big[2r_2^*(\phi ) + r_2^{*\prime}(\phi ) \cot\phi \big ] \nonumber \\ & & + r_2^{*\prime\prime}(\phi ) \big\ } - a \big\ { ( n-2)\big[2y_2(\phi ) + y_2'(\phi ) \cot\phi \big ] \nonumber \\ & & + y_2''(\phi ) \big\ } , \label{frgl-2}\end{aligned}\ ] ] where we have introduced the eigenvalues @xmath103 ( measured in units of @xmath104 ) and defined @xmath105 .
the fp is attractive if all @xmath106 fulfill the inequality @xmath107 .
a singly unstable fp has only one eigenvalue @xmath108 such that @xmath109 , which determines the third independent exponent @xmath110.\ ] ] this exponent characterizes the divergence of the correlation length in the vicinity of transition .
cdddddd & & & & & + @xmath111 & 1.271 & @xmath1121.0000 & @xmath1130.5811 & @xmath1131.271 & @xmath1132.429 + & 2 & @xmath1120.4657 & @xmath1130.0891 & @xmath1131.218 & @xmath1132.000 + & 3 & @xmath1120.3320 & @xmath1130.0041 & @xmath1131.198 & @xmath1131.618 + & 4 & @xmath1120.2668 & @xmath1120.0207 & @xmath1131.192 & @xmath1131.414 + @xmath114 & 2 & @xmath1120.1743 & @xmath1130.0192 & @xmath1131.167 & @xmath1131.366 + & 3 & @xmath1120.1132 & @xmath1120.0073 & @xmath1131.160 & @xmath1131.225 + & 4 & @xmath1120.0819 & @xmath1120.0138 & @xmath1131.144 & @xmath1131.158 + @xmath115 & 2 & @xmath1120.0941 & @xmath1130.0080 & @xmath1131.145 & @xmath1131.215 + & 3 & @xmath1120.0549 & @xmath1120.0058 & @xmath1131.127 & @xmath1131.135 +
we now focus on the phase diagram and critical behavior of the lr rf @xmath0 model .
et al . _
@xcite studied the critical behavior of the lr rf @xmath0 model using a @xmath116 expansion around the upper critical dimension @xmath117 .
they found that , to lowest order in @xmath118 , the critical properties are that of a pure system in @xmath119 dimensions ; however , this generalized dr was found to fail at higher orders .
chang and abrahams@xcite applied to the lr rf @xmath0 model a low - temperature version of the rg . expanding around the lower critical dimension @xmath58 , they obtained the rg recursion relations for the three parameters @xmath120 , @xmath121 , and @xmath122 , which in our notation correspond to @xmath123 .
as we have shown in the previous section , the truncated rg neglects an infinite number of relevant operators .
they found the nontrivial zero - temperature fp , which in our notation @xmath124 reads @xmath125 the correlation length exponent @xmath126 is determined by eq .
( [ nu - def ] ) , with the relevant eigenvalue @xmath108 given by @xmath127}{\hat{\varepsilon}\,^2(n-3)^2}}. \ \ \ \ \ \ \label{nu - dr}\end{aligned}\ ] ] note that the expression for the relevant eigenvalue reported in ref . is incorrect and gives values which are several times larger than that computed using eq .
( [ nu - dr ] ) .
although the truncated rg scheme when applied to the model with only sr correlated rf ( @xmath128 ) results in the dr fp @xmath129 and exponent @xmath130 , the exponent ( [ nu - dr ] ) differs from the generalized dr prediction @xmath131 . nevertheless , one may doubt about the applicability of the truncated rg to the lr rf @xmath0 model even if the dr is broken . .
the borderline between the sr and lr fps is computed numerically using eq .
( [ criterion - rf ] ) ; the borderline between the sr tt and lr tt regions is given by eq .
( [ tt0 ] ) and that between the lr and lr tt regions by eq .
( [ tt ] ) .
inset : schematic phase diagram on the @xmath132 plane for a particular value of @xmath133 .
the solid line is @xmath134 ; the dashed and dotted lines are given by eqs .
( [ criterion - rf ] ) and ( [ tt ] ) , respectively.,width=307 ] we now reexamine the long - distance behavior of the lr rf @xmath0 model by means of the full one - loop frg derived in sec .
[ sec - model ] .
equation ( [ frg-2 ] ) is linear in the function @xmath135 and can be solved analytically .
the fp solution fulfilling the rf boundary condition , i.e. , @xmath90 periodicity , is given by @xmath136 which is an analytic function .
note that the analyticity of the lr part of the disorder correlator was also revealed in the elastic manifold problem.@xcite this also gives us the fp value of @xmath137 , @xmath138 which , following eqs .
( [ exp - eta ] ) , completely fixes the values of the critical exponents @xmath63 and @xmath64 , @xmath139 exponents ( [ rf - exp ] ) satisfy the generalized schwartz - soffer inequality at equality .
this is at variance with the sr rf models , where the schwartz - soffer inequality was found to be strict,@xcite but in agreement with the results for the lr rf spherical model with long - range interactions.@xcite one may conjecture that the generalized schwartz - soffer inequality is satisfied as equality for any @xmath8 including the lr rf ising model,@xcite but there is no argument that this persists in higher orders of loop expansion .
let us first discuss the critical behavior of the lr rf model above the lower critical dimension , @xmath140 .
note that @xmath141 exactly coincides with @xmath142 given by eqs .
( [ delta-1 - 2 ] ) .
therefore the truncated rg would give the same values of @xmath63 and @xmath64 as the full frg , at least to first order in @xmath6 and @xmath4 , if these exponents would be computed in ref . .
to obtain the function @xmath143 and the amplitude @xmath144 , we integrate eq .
( [ frg-1 ] ) numerically . since the coefficients of eq .
( [ frg-1 ] ) are singular at @xmath145 and @xmath91 , to compute the solution in the vicinity of these points , we use expansions of @xmath143 in powers of @xmath146 and @xmath147 , respectively . the expansion around @xmath148 is completely determined by the value of @xmath149 , @xmath150 } + \frac{r_1^{*\prime\prime}(0 ) \phi^2}{2 } \nonumber \\ & \pm & \frac{\sqrt{r_1^{*\prime\prime}(0)(\hat{\varepsilon}-1)(n-3)^2 + 1 } } { 3(n-3)\sqrt{n+2 } } |\phi|^3 + o(\phi^4 ) , \ \ \ \ \label{series1}\end{aligned}\ ] ] while to get the explicit expansion around @xmath91 we need know @xmath149 and @xmath151 : @xmath152(n-3)+1\big\}^2 + 2 r_1^{*\prime\prime}(\pi)(n-3)}{2(n-3 ) [ ( n-3)\hat{\varepsilon } - 2(n-2 ) ] } \nonumber \\ & & + \frac{r_1^{*\prime\prime}(\pi ) ( \pi-\phi)^2}{2 } + o[(\pi-\phi)^4 ] .
\label{series2}\end{aligned}\ ] ] from eq .
( [ series1 ] ) , we see that the sr part of the disorder correlator is nonanalytic at small @xmath88 , so that we have to distinguish the left and right derivatives . in
what follows , we adopt @xmath153 .
we use numerical integration to continue the solutions given by expansions ( [ series1 ] ) and ( [ series2 ] ) inside the interval @xmath154 $ ] and match them by adjusting the shooting parameters @xmath149 and @xmath151 . only the series with `` @xmath155 '' in eq .
( [ series1 ] ) can be matched with the solution computed using expansion ( [ series2 ] ) . following the third term of eq .
( [ series1 ] ) , the fp solution exists only if @xmath156 .
we found that this inequality is always strict , however , the difference @xmath157 becomes smaller in the limit of large @xmath8 .
thus the truncated rg fails to give the correct values of @xmath149 and @xmath144 although it gives the correct sum .
consequently , the relevant eigenvalue @xmath108 and the exponent @xmath126 obtained from the truncated rg are also expected to be incorrect .
the computed values of @xmath149 and @xmath151 are shown in table [ tab - rf ] .
let us check the stability of fps .
first , we examine the stability of the sr fp @xmath158 $ ] found numerically in ref . .
linearized about the fp , eq .
( [ frgl-2 ] ) can be solved analytically giving @xmath159 and @xmath160 .
the sr fp is stable against the introduction of lr disorder if @xmath161 .
taking into account that @xmath162 and @xmath163 , we can rewrite the criterion of the sr fp stability in the following form @xmath164 which was derived using general rg scaling arguments in ref . .
the regions of stability of the lr and sr fps are shown in fig .
[ fig1-rf ] .
tissier and tarjus ( tt ) argued that for @xmath165 , the cuspy sr fp becomes more than once unstable , but a different critical ( singly unstable ) sr tt fp arises with the function @xmath166 being only @xmath167 times differentiable at the origin.@xcite although this weak nonanalyticity should reflect some metastability in the system,@xcite it leads to the dr critical exponents .
according to eq .
( [ criterion - rf ] ) , the sr tt fp is stable with respect to the lr correlated disorder for @xmath165 and @xmath168 ( @xmath169 ) .
, width=307 ] we now check the stability of the lr rf fp @xmath170 $ ] . substituting the lr fp in eqs .
( [ frgl-1 ] ) and ( [ frgl-2 ] ) , we obtain @xmath171 and @xmath172 to compute the eigenfunction @xmath173 and eigenvalue @xmath103 , we solve eq .
( [ frgl-1 ] ) numerically using shooting and imposing the @xmath90-periodic boundary condition .
the obtained values of @xmath108 are shown in table [ tab - rf ] .
as one can see from the table , the relevant eigenvalues @xmath108 computed using the full frg , and consequently , the exponent @xmath126 differ significantly from that obtained using the truncated rg .
the critical exponents are expected to be continuous functions of @xmath6 , @xmath174 , and @xmath8 . in the region controlled by the sr fp
the correlation length exponent is @xmath175 , independent on @xmath8.@xcite the first line in table [ tab - rf ] corresponds to a point on the borderline separating the regions of stability ( see fig . [ fig1-rf ] ) , and thus , on the borderline we indeed have @xmath176 .
we checked that the second eigenvalue @xmath177 computed at the lr fp vanishes exactly on the borderline , indicating that the lr fp becomes twice unstable in the region dominated by the sr fp .
we also looked for an analog of the tt phenomena in the presence of lr correlated disorder and found that for @xmath178 there is a crossover to a new singly unstable lr fp of a tt type with the function @xmath166 being only @xmath167 times differentiable at the origin and @xmath179 .
the values of @xmath180 and @xmath181 at the lr tt fp are simply given by eq .
( [ delta-1 - 2 ] ) .
thus , the lr tt fp leads to the critical exponents ( [ rf - exp ] ) and the relevant eigenvalue ( [ nu - dr ] ) obtained from the truncated rg . finally , we discuss the lr rf @xmath0 model below the lower critical dimension , @xmath182 . for @xmath183 ,
the frg equations have two attractive fps : the sr and lr , which describe the sr qlro and lr qlro phases , respectively .
the phase diagram computed from the stability analysis of different fps is depicted in fig .
[ fig2-rf ] .
the exponents characterizing the power - law behavior of the connected and disconnected correlation functions in the lr qlro phase are given by eq .
( [ rf - exp ] ) . the physically interesting case @xmath184 describing the bragg glass phase in the presence of the lr correlated disorder
was considered by one of the authors in ref . .
in the bragg glass , the displacements @xmath185 of a periodic structure ( e.g. , vortex lattice ) grow logarithmically as @xmath186 . note that in ref .
the period was fixed to @xmath187 while in the present work it is @xmath90 , so that the relation between the universal amplitude @xmath188 and the exponent @xmath63 is given by @xmath189 . for @xmath184 , the crossover from sr qlro with the critical exponents @xmath190 and @xmath191 to lr qlro with the exponents @xmath192 and @xmath193 happens for @xmath194 ( see fig . [ fig2-rf ] ) .
in this section , we study the lr ra case which corresponds to @xmath91-periodic functions @xmath195 . equation ( [ frg-2 ] ) has a family of ( at most ) @xmath91-periodic solutions which can be expressed as polynomials in @xmath196 of power @xmath197 with @xmath198 for instance , for @xmath199 , we have @xmath200[n\cos^2\phi-1],\label{r2-ra }
\\ a^*_{\mathrm{lrra}}&=&\frac14 , \label{a - ra}\end{aligned}\ ] ] and for @xmath201 , we have @xmath202 and @xmath203 . \nonumber\end{aligned}\ ] ] the stability analysis shows that due to the inequality @xmath204 for @xmath205 and @xmath206 , all fps with @xmath207 are unstable .
it can be easily seen for @xmath184 when the functions @xmath208 ( @xmath209 ) become @xmath210-periodic , and thus , are unstable with respect to a @xmath91-periodic perturbation .
henceforth we consider only the lr fp determined by eqs .
( [ r2-ra ] ) and ( [ a - ra ] ) , which give the following values of the critical exponents @xmath211 exponents ( [ exp - ra-2 ] ) satisfy the generalized schwartz - soffer inequality @xmath72 also at equality .
analogously to the lr rf model , in order to obtain the function @xmath143 , we solve eq .
( [ frg-1 ] ) numerically using shooting and imposing the @xmath91-periodic boundary condition . since the coefficients of eq .
( [ frg-1 ] ) are singular at @xmath145 , we use an expansion of @xmath143 in powers of @xmath146 which reads @xmath212}{16(n-2 + 2 \hat{\varepsilon } ) } + \frac{r_1^{*\prime\prime}(0 ) \phi^2}{2 } \nonumber \\ & & \pm \sqrt{1+\frac{16r_1^{*\prime\prime}(0)(\hat{\varepsilon}-1)}{n+2 } } \frac { |\phi|^3}{12 } + o(\phi^4 ) . \
\ \ \ \label{series - ra}\end{aligned}\ ] ] as one can see from eq .
( [ series - ra ] ) , the sr disorder correlator @xmath143 has a cusp at the origin .
only the solution with `` + '' in eq .
( [ series - ra ] ) fulfills all conditions .
we now check the stability of the sr ra fp with respect to the lr correlated disorder .
analogously to the rf case , linearized around the fp , eq .
( [ frgl-2 ] ) allows for an analytical solution giving @xmath213 and @xmath214 .
the sr fp is stable if @xmath161 , which can be rewritten as @xmath215 above the lower critical dimension , @xmath140 , inequality ( [ ra - ineq ] ) holds for all @xmath216 so that the sr fp is stable with respect to the weak lr correlated disorder .
although the lr correlated disorder does not change the critical behavior for @xmath217 , it can modify the critical behavior which should exist for @xmath218 but which is not accessible in the one - loop approximation.@xcite ( @xmath169).,width=297 ] below @xmath24 , the lr ra model can not develop a true long - range order , but there can exist different types of qlro .
the sr qlro phase is controlled by the sr ra fp , with @xmath219 and @xmath220 computed in ref . .
it was found to be stable in the subspace of the sr correlated disorder for @xmath221 .
we find that it is also stable with respect to the lr correlated disorder for @xmath4 fulfilling inequality ( [ ra - ineq ] ) .
the new lr qlro phase is controlled by an attractive lr ra fp with @xmath91-periodic functions @xmath222 and @xmath223 , which we computed numerically by integrating eq .
( [ frg-1 ] ) with the initial condition given by eq .
( [ series - ra ] ) and using @xmath149 as a shooting parameter .
the exponents describing the correlation functions in the lr qlro phase are given by eqs .
( [ exp - ra-1 ] ) and ( [ exp - ra-2 ] ) .
note that below @xmath24 we have @xmath224 . to check the stability of the lr ra fp , we substitute it in eq .
( [ frgl-2 ] ) and obtain @xmath213 and @xmath225 the eigenfunction @xmath173 and eigenvalue @xmath103 are computed numerically using shooting .
this allows us to determine the stability regions for different qlro phases , which are shown in fig .
[ fig3-ra ] . as one can see from fig .
[ fig3-ra ] in contrast to sr qlro , lr qlro can exist even for @xmath226 and @xmath217 .
the above results may be relevant for the behavior of @xmath7he in aerogels .
it has recently been observed in nmr experiments that the phase a of @xmath7he confined in aerogel exhibits two different types of magnetic behavior called @xmath227 and @xmath228 states.@xcite depending on the cooling , one can obtain either a nearly pure @xmath227 state or a mixed @xmath229 state , which gives two overlapping lines @xmath227 and @xmath228 in the transverse nmr spectrum .
although the pure @xmath228 state has not been observed , there is an evidence that the @xmath229 state is inhomogeneous and consists of regions with two different magnetic orders @xmath227 and @xmath228.@xcite the order parameter of the @xmath7he - a can be parametrized by the complex vector @xmath230 and the real unit vector @xmath231 , which characterize the orbital and magnetic anisotropy , respectively.@xcite only the orbital part of the order parameter given by the real unit vector @xmath232 $ ] interacts with the aerogel matrix , which can be treated as quenched random anisotropy disorder .
the spin part @xmath231 is not coupled directly to the disorder , but there is a weak spin - orbit ( dipole ) interaction between @xmath233 and @xmath231 which generates the nmr frequency shift . in ref . , the existence of the two states was interpreted in terms of different `` random textures '' of the field @xmath233 in the larkin - imry - ma state ( qlro phase in our notation ) .
the measured dependencies of the transverse nmr signal on the tipping angle can be explained if one assumes that in the @xmath227 state the vectors @xmath233 and @xmath231 are almost uncorrelated , while in the @xmath228 state they are partially locked .
this is possible if in the @xmath227 state the characteristic length of the texture , i.e. , the larkin length , is much smaller than the characteristic length of the dipole interaction , @xmath234 , while in the @xmath228 state they are of the same order.@xcite however , the nature of different random textures exhibiting different characteristic length scales is not clear .
alternatively , one can try to interpret the @xmath228 state as a network of topological defects pinned by the aerogel.@xcite we argue that these two different states may be the sr and lr qlros found above in the lr ra model .
indeed , aerogel is a porous medium formed by tangled silicon strands which exhibit a fractal mass distribution .
nmr and small - angle x - ray scattering ( saxs ) experiments give the mass - to - distance relation @xmath235 , with the fractal dimensions @xmath236 varying in the range of @xmath237
@xmath238 depending on the fabrication process .
this scaling holds up to the fractal correlation length , which can exceed 100 nm.@xcite thus , the effective ra disorder is expected to be long - range correlated with @xmath239 up to the scale of the fractal correlation length or even more.@xcite as a result , the sr qlro phase is unstable to formation of islands with lr qlro ( see fig . [ fig3-ra ] ) . to compute the larkin length , one has to solve the flow equations ( [ frg-1 ] ) and ( [ frg-2 ] ) starting from a particular bare disorder correlator and looking for the scale at which the cusp is developing .
@xcite we can estimate the larkin length using flory - type arguments.@xcite assuming that @xmath240 is an elastic constant and @xmath86 is a strength of disorder , we obtain @xmath241 and @xmath242 for the sr and lr disorders , respectively . for aerogel , we have @xmath243 and @xmath244 so that the larkin lengths in both phases may differ significantly .
this can explain the experimentally observed coexistence of regions with different spin states .
further investigations , however , are clearly needed .
in this work , we investigated the long distance properties of the @xmath0 model with random fields and random anisotropies correlated as @xmath38 for large separation @xmath2 .
we derived the functional renormalization group equations to one - loop order , which allow us to describe the scaling behavior of the models below and above the lower critical dimension @xmath58 . using a double @xmath3 and @xmath4 expansion
, we obtained the phase diagrams and computed the critical exponents to first order in @xmath6 and @xmath4 . for the lr rf model
, we found that the truncated rg developed in ref .
to study the critical behavior above the lower critical dimension is able to give the correct one - loop values of exponents @xmath63 and @xmath64 , but not the phase diagram and the critical exponent @xmath126 except for the region controlled by the weakly nonanalytic lr tt fp .
thus , although the truncated rg overcomes the dimensional reduction , it fails to reproduce all properties which can be obtained using the functional renormalization group .
we found a new lr qlro phase existing in the lr rf model below the lower critical dimension for @xmath183 and determined the regions of its stability in the @xmath245 parameter space .
we obtained that the weak lr correlated disorder does not change the critical behavior of the ra model above @xmath24 for @xmath216 , but can create a new lr qlro phase below @xmath24 .
the existence of two qlro phases in lr ra systems may explain the two different states of @xmath7he - a in aerogel observed recently in nmr experiments.@xcite however , many questions are still open .
in particular , the paramagnetic - ferromagnetic transition should exist also for @xmath218 , though it was not found in the one - loop approximation .
even for the sr correlated disorder , it is still unclear whether it remains perturbative.@xcite it would also be interesting to find a connection with the replica symmetry breaking picture.@xcite we would like to thank pierre le doussal and kay wiese for many stimulating discussions and critical reading of the manuscript .
we are also grateful to g.e .
volovik for drawing our attention to recent results on superfluid helium in aerogels .
thanks lpt ens for hospitality during this work .
acknowledges support from the european commission under contract no .
mif1-ct-2005 - 021897 .
clark , t. bellini , r.m .
malzbender , b.n .
thomas , a.g .
rappaport , c.d .
muzny , d.w .
schaefer , and l. hrubesh , phys .
lett . * 71 * , 3505 ( 1993 ) ; t. bellini , n.a .
clark , and d.w .
schaefer , _ ibid . _ * 74 * , 2740 ( 1995 ) . | we study the long - distance behavior of the @xmath0 model in the presence of random fields and random anisotropies correlated as @xmath1 for large separation @xmath2 using the functional renormalization group .
we compute the fixed points and analyze their regions of stability within a double @xmath3 and @xmath4 expansion .
we find that the long - range disorder correlator remains analytic but generates short - range disorder whose correlator develops the usual cusp .
this allows us to obtain the phase diagrams in @xmath5 parameter space and compute the critical exponents to first order in @xmath6 and @xmath4 .
we show that the standard renormalization group methods with a finite number of couplings used in previous studies of systems with long - range correlated random fields fail to capture all critical properties .
we argue that our results may be relevant to the behavior of @xmath7he - a in aerogel . |
davis famously constructed , for every @xmath0 , a closed aspherical @xmath1-manifold whose universal cover is not homeomorphic to @xmath2 @xcite .
we prove : [ thm : univ cover not rn ] if the universal cover of an open smooth manifold @xmath3 is diffeomorphic to @xmath2 with @xmath0 , then the tangential homotopy type of @xmath3 contains a continuum of open smooth @xmath1-manifolds whose universal covers are not homeomorphic to @xmath2 .
our proof also works in pl and top categories . recall that a homotopy equivalence @xmath4 of @xmath1-manifolds is _ tangential _ if @xmath5 and @xmath6 are stably isomorphic , where @xmath7 denotes the tangent bundle if @xmath8 is a smooth manifold and the tangent microbundle in the pl or top cases . by (
* theorem 2.3 ) this is equivalent to requiring that @xmath9 is homotopic to a cat - homeomorphism for some @xmath10 ( where cat equals diff , pl or top ) .
\(1 ) a result of curtis - kwun @xcite that for a boundary connected sum @xmath13 of a countable family of compact @xmath1-manifolds , the homeomorphism type of @xmath14 determines the isomorphism class of @xmath15 .
\(2 ) a recent result of calcut - king - siebenmann @xcite that any countable collection of cat properly embedded @xmath16 s in @xmath2 is cat unknotted , which generalizes classical results of cantrell and stallings .
\(3 ) the existence of infinitely many smooth compact contractible @xmath1-manifolds whose boundary homology spheres have freely indecomposable fundamental groups .
( such examples can be found in @xcite , and more examples are now known , see e.g. casson - harer @xcite for aspherical homology @xmath17-spheres that bound smooth contractible @xmath18-manifolds , while kervaire @xcite showed that the fundamental group of any homology @xmath17-sphere appears as the fundamental group of the boundary of a smooth contractible @xmath1-manifolds for any @xmath11 ) .
since @xmath3 is open , it contains a cat properly embedded ray whose cat regular neighborhood is an embedded closed halfspace , see e.g. ( * ? ? ?
* section 3 ) .
hence @xmath3 is cat isomorphic to the interior of a noncompact manifold @xmath19 whose boundary is an open disk . by the strong version of the cantrell - stalling hyperplane unknotting theorem proved in ( * ? ? ?
* corollary 9.3 ) , the universal cover of @xmath19 can be compactified to @xmath20 , the @xmath1-disk , in which the preimage of @xmath21 becomes the union of a countable collection of round open disks with pairwise disjoint closures .
let @xmath22 be an infinite sequence of compact contractible @xmath1-manifolds , such that @xmath23 are pairwise non - isomorphic and freely indecomposable .
given a subset @xmath24 , let @xmath25 be a boundary connected sums of @xmath26 s with indices in @xmath27 .
( for our purposes the choices involved in defining boundary connected sums will always be irrelevant ) .
fix a closed @xmath28-disk @xmath29 , and let @xmath30 be a boundary connected sum of @xmath19 and @xmath25 obtained by identifying @xmath31 with a closed disk in @xmath21 .
a deformation retraction @xmath32 extends to a deformation retraction of @xmath33 , so @xmath34 is tangentially homotopy equivalent to @xmath3 . if @xmath35 denotes a boundary connected sum of countably many copies of @xmath25 , then the interior of the universal cover of @xmath30 is homeomorphic to the interior of a boundary connected sum of @xmath20 and @xmath35 , which is homeomorphic to @xmath36 . by
* theorem 4.1 ) if @xmath35 , @xmath37 have homeomorphic interiors , then @xmath38 , @xmath39 have isomorphic fundamental groups .
now @xmath40 is a free product in which each factor @xmath41 , @xmath42 appears countably many times .
each @xmath41 is freely indecomposable , so @xmath43 by grushko s theorem .
thus the universal covers of @xmath44 lie in a continuum of homeomorphism types .
the proof of ( * ? ? ?
* theorem 4.1 ) is quite technical , which may be due to the fact that the tools of siebenmann s thesis were not yet available at the time , so we summarize it in a modern language : given @xmath45 it is easy to construct a cofinal family @xmath46 of neighborhoods of infinity in @xmath35 such that in the corresponding inverse sequence of fundamental groups , the group @xmath47 is the free product of @xmath48 , and the map @xmath49 is a retraction onto the first @xmath50 factors , and in particular , is surjective .
hence the inverse sequence of groups is mittag - leffler , and therefore its pro - equivalence class depends only on the homeomorphism type of @xmath35 .
now if @xmath35 , @xmath37 have homeomorphic interiors , then a simple diagram chase in the commutative diagram from the definition of pro - equivalence shows that each free factor of @xmath51 occurs as a free factor of @xmath52 , so @xmath43 .
theorem [ thm : univ cover not rn ] should hold for @xmath53 , but our proof fails .
one could try substituting the boundary connected sum of @xmath19 and @xmath35 by the end sum of @xmath3 with a suitable whitehead manifold , but the multiple hyperplane unknotting is no longer true , due to existence of an exotic @xmath54\times \mathbb r^2 $ ] @xcite .
this makes analyzing the fundamental group of infinity more delicate . for the same reason we do not attempt to prove theorem [ thm : univ cover not rn ] for @xmath3
whose universal cover is not @xmath2 . by the cartan - hadamard theorem
any ( complete riemannian ) @xmath1-manifold of nonpositive curvature is covered by @xmath2 .
so given an open @xmath1-manifold of nonpositive curvature with @xmath0 , theorem [ thm : univ cover not rn ] yields a continuum of @xmath1-manifolds in the same tangential homotopy type that admit no metric of nonpositive curvature . in @xcite the author studied obstructions to nonpositive curvature on open manifolds that go beyond the cartan - hadamard theorem . | we show that any open aspherical manifold of dimension @xmath0 is tangentially homotopy equivalent to an @xmath1-manifold whose universal cover is not homeomorphic to @xmath2 . |
let @xmath1 be a graph , let @xmath2 be an assignment of lists of colors to vertices of @xmath1 .
we say that a coloring @xmath3 is an @xmath4-coloring , or _
proper coloring from lists @xmath4 _ if @xmath5 for @xmath6 , and @xmath7 if @xmath8 .
when such an @xmath4-coloring occurs , we say that @xmath1 is _
@xmath4-colorable_. for extensive literature on list - colorings of planar graphs we refer the reader to @xcite .
thomassen @xcite proved that if @xmath1 is a planar graph and @xmath9 for each vertex @xmath6 , then @xmath1 is @xmath4-colorable .
let @xmath10 be a subset of vertices in a graph @xmath1 .
we say that a _ precoloring of p _ is _ extendable to a @xmath0-list coloring _ of @xmath1 if for every list assignment @xmath2 , such that @xmath11 for @xmath12 and @xmath9 for @xmath13 , @xmath1 is @xmath4-colorable .
albertson @xcite posed the following question .
let @xmath1 be a plane graph .
is there a @xmath14 such that whenever @xmath15 is such that the distance between every pair of vertices of @xmath10 is at least @xmath16 , then every precoloring of @xmath10 extends to a @xmath0-list coloring of @xmath1 ? here , the distance _
dist@xmath17 _ between a pair of vertices @xmath18 and @xmath19 is the number of edges in a shortest path joining them .
tuza and voigt showed in @xcite , see also @xcite , that the condition of a large distance between precolored vertices is essential by finding a planar graph @xmath1 with a set of precolored vertices at pairwise distance at least @xmath20 such that the precoloring is not extendable to a @xmath0-list coloring of @xmath1 .
so , the distance @xmath16 in the above question should be at least @xmath0 .
does this question have a positive answer if @xmath21 ?
the original theorem of thomassen @xcite implies that if there are two adjacent precolored vertices assigned distinct colors , then the precoloring is extendable to a @xmath0-list coloring of @xmath1 .
bhme et al .
@xcite described when the precoloring of vertices on a short face with at most six vertices can be extended to a @xmath0-list coloring of a planar graph . in this manuscript
, we introduce a technique using shortest paths in planar graphs which allows us to answer albertson s question for a wide class of planar graphs .
we prove that a proper precoloring of a pair of vertices can always be extended to a @xmath0-list coloring of a planar graph provided they are not separated by @xmath22- or @xmath20-cycles .
we also provide results about extensions of precolorings of vertices on one face .
finally , we answer albertson s question in the case where there are no @xmath22- or @xmath20-cycles separating precolored vertices and there is a special tree containing all of the precolored vertices . to state our main results in all their generality , we need to define several notions .
we say a set of vertices @xmath23 separates a set of vertices @xmath10 in a connected graph @xmath1 if there are at least two vertices of @xmath10 contained in distinct connected components of @xmath24 .
if @xmath23 is an @xmath25-vertex set separating @xmath10 in @xmath1 and spanning @xmath26 , we say @xmath1 contains a @xmath10-separating @xmath26 .
if @xmath23 separates @xmath27 we say @xmath23 separates @xmath1 , or @xmath23 is a separating set in @xmath1 . for a set of vertices
@xmath10 in a graph @xmath1 , _ dist@xmath28 dist@xmath29 _ is the smallest distance in @xmath1 between two vertices of @xmath10 . for a path @xmath30 , with endpoints @xmath31 and @xmath32 , we say a vertex @xmath33 is central if the distances in @xmath30 from @xmath33 to @xmath31 and from @xmath33 to @xmath32 differ by at most @xmath34 .
note there are at most two central vertices in @xmath30 . for graph theoretic terminology
not defined here , we refer the reader to @xcite .
let @xmath1 be a planar graph , @xmath10 a subset of vertices of @xmath1 .
fix a positive integer @xmath16 .
let @xmath35 be a tree with @xmath36 .
let the set of special vertices be the union of @xmath10 and the set of vertices of degree either @xmath34 or at least @xmath22 in @xmath35 . a path in @xmath35 with special vertices as endpoints and containing no other special vertices
is called a branch of @xmath35 .
we say a tree @xmath35 is * @xmath37-steiner * if + ( 1 ) every branch has length at least @xmath38 , + ( 2 ) every branch is a shortest ( in @xmath1 ) path between its endpoints , + ( 3 ) if @xmath39 is a center of a branch of @xmath35 , then a shortest ( in @xmath1 ) path between @xmath39 and every vertex in another branch has length at least @xmath16 , and + ( 4 ) no two vertices of @xmath35 from distinct branches have a common neighbor outside of @xmath35 nor are they adjacent . for example , when @xmath40 is a set of two vertices at distance @xmath41 from each other , a shortest @xmath42-path is a @xmath43-steiner tree with a single branch .
we say that a set @xmath23 of four vertices of degree at most @xmath0 in a graph @xmath1 forms the configuration @xmath44 , if @xmath45 $ ] is isomorphic to @xmath46 .
see figure [ fig : d - reducible ] for an illustration of @xmath47 .
we say that a set @xmath23 of seven vertices of degree at most @xmath48 in a graph @xmath1 forms the configuration @xmath49 , if @xmath45 $ ] induces a @xmath48-wheel , formed from a central vertex @xmath33 adjacent to a @xmath48-cycle @xmath50 such that @xmath51 and @xmath52 have degree at most @xmath0 in @xmath1 .
see figure [ fig : w - reducible ] for an illustration of @xmath53 . for a graph @xmath1 and a set of precolored vertices @xmath10 ,
let @xmath54 , a * reduction of * @xmath1 , be a graph obtained by one of the following operations : ( 1 ) for a separating @xmath55 or @xmath56 that does not separate @xmath10 , remove from @xmath1 the vertices and edges in the region that is bounded by the separating @xmath55 or @xmath56 and that does not contain any vertices of @xmath10 , ( 2 ) for a configuration @xmath44 such that @xmath57 , remove @xmath23 from @xmath1 , or ( 3 ) for a configuration @xmath49 such that @xmath57 , remove @xmath23 from @xmath1 .
if none of these operations can be carried out , let @xmath58 .
consider a sequence of graphs @xmath59 such that @xmath60 for @xmath61 and @xmath62 .
call such a graph @xmath63 a * reduced * graph of @xmath1 .
a reduced graph does not have a separating @xmath55 or @xmath56 that does not separate @xmath10 and it contains no configurations @xmath64 or @xmath65 with @xmath57 .
we shall show that if a reduced graph of @xmath1 has a coloring extension of @xmath10 , then so does @xmath1 .
we now state the main results of this manuscript .
[ steiner ] let @xmath1 be a plane graph , let @xmath10 be a set of vertices such that there is no @xmath10-separating @xmath55 or @xmath56 in @xmath1 .
if there is a reduced graph of @xmath1 that has a @xmath66-steiner tree , then every precoloring of @xmath10 is extendable to a proper @xmath0-list coloring of @xmath1 .
[ two ] let @xmath1 be a plane graph and @xmath67 .
if @xmath1 has no @xmath68-separating @xmath55 or @xmath56 , then every proper precoloring of @xmath69 is extendable to a proper @xmath0-list coloring of @xmath1 .
[ one - face ] let @xmath1 be a plane graph and @xmath70 a set of vertices of a facial cycle of @xmath1 .
let @xmath71 , where the vertices of @xmath10 are labeled cyclically around @xmath70 . then every proper precoloring of @xmath10 is extendable to a @xmath0-list coloring of @xmath1 if one of the following conditions holds : 1 . @xmath72 $ ] consists of disjoint vertices and edges with pairwise distance at least @xmath22 , 2 .
@xmath73 and none of the following occur : 1 .
there is a vertex @xmath74 adjacent to at least five vertices of @xmath10 such that @xmath75 consists of the colors assigned to those five vertices .
2 . @xmath76 and there is an edge @xmath77 and a color @xmath78 such that , for @xmath79 , the vertex @xmath80 is adjacent to @xmath81 , where addition of indices is modulo @xmath82 , and @xmath83 consists precisely of the colors assigned to those four vertices and @xmath78 .
3 . @xmath84 and there is a triangle @xmath85 and colors @xmath86 , such that , for @xmath87 , the vertex @xmath80 is adjacent to @xmath88 , where addition of indices is modulo @xmath82 , and @xmath83 consists of the colors assigned to those three vertices and @xmath86 .
[ two - faces ] let @xmath10 be a set of vertices in a plane graph @xmath1 , @xmath89 , such that there are two faces @xmath90 where the vertices of @xmath10 lie on the boundaries of @xmath91 and @xmath92 .
assume @xmath1 contains no @xmath10-separating @xmath55 or separating @xmath56 .
then every precoloring of @xmath10 is extendable to a proper @xmath0-list coloring of @xmath1 .
the rest of the paper is organized as follows . in section [ preliminaries ]
we state known results mentioned above in detail and prove some technical lemmas .
we prove all of the theorems in section [ proofs ] .
finally , we state open problems and comments in section [ conclusions ] .
[ t ] let @xmath1 be a plane graph , @xmath93 the set of vertices of a face of @xmath1 , @xmath94 an assignment of lists of colors to vertices of @xmath1 such that @xmath95 for all @xmath96 , @xmath97 with @xmath98 for some adjacent vertices @xmath99 , and @xmath100 for @xmath101
. then @xmath1 is @xmath4-colorable .
there is a planar graph @xmath1 and a set @xmath10 of vertices with @xmath102 and an assignment of lists of size @xmath22 to vertices of @xmath10 and lists of size @xmath0 to the remaining vertices such that @xmath1 is not colorable from these lists .
[ bms ] let @xmath103 be a plane graph with facial cycle @xmath70 of length @xmath104 , where the vertices of @xmath70 are labeled cyclically @xmath105 .
if @xmath11 for @xmath106 , @xmath9 for @xmath107 , and @xmath108 $ ] is @xmath4-colorable , then @xmath1 is @xmath4-colorable unless one of the following occurs : 1 .
there is a vertex @xmath109 adjacent to five vertices in @xmath70 and @xmath75 consists exactly of the colors assigned to those five vertices .
2 . @xmath84 and there is an edge @xmath77 , @xmath110 and a color @xmath78 such that , for @xmath79 , the vertex @xmath80 is adjacent to @xmath81 , where addition of indices is modulo @xmath82 , and @xmath83 consists precisely of the colors assigned to those four vertices and @xmath78 .
3 . @xmath84 and there is a triangle @xmath85 , @xmath111 and colors @xmath86 , such that , for @xmath87 , the vertex @xmath80 is adjacent to @xmath88 , where addition of indices is modulo @xmath82 , and @xmath83 consists of the colors assigned to those three vertices and @xmath86
. see figure [ fig : onefaceforb ] for illustrations of the forbidden configurations described in conditions 2 and 3 of theorem [ bms ] .
note how these compare to the forbidden configurations of theorem [ one - face ] . for a vertex @xmath6
, we denote @xmath112 the neighborhood of @xmath32 in @xmath1 . for a vertex @xmath32 and a set of vertices @xmath23 , we write @xmath113 if @xmath32 is adjacent to all vertices in @xmath23 .
let @xmath114 be a subgraph of @xmath1 and @xmath115 be a vertex coloring of @xmath114 .
for @xmath116 , let @xmath117 be the set of colors used on neighbors of @xmath32 in @xmath114 .
let @xmath118 be the size of the neighborhood of @xmath32 in @xmath114 .
for a vertex set @xmath23 in @xmath1 , let @xmath119)$ ] be the set of neighbors of vertices from @xmath23 not in @xmath23 . for an induced subgraph @xmath114 of @xmath1 and @xmath120 , let @xmath121 .
when the subgraph @xmath114 is clear , we use @xmath122 .
we say @xmath114 is colored _ * nicely * _ by a coloring @xmath115 with respect to lists @xmath4 if @xmath115 is an @xmath4-coloring of @xmath114 and for every vertex @xmath123 , @xmath124 we also say @xmath115 is a _ nice _ coloring of @xmath114 in this case .
a vertex from @xmath125 adjacent to at least three vertices in @xmath114 is called a _ three - neighbor _ , or simply @xmath22-neighbor , of @xmath114 .
we denote the set of @xmath22-neighbors of @xmath114 by @xmath126 .
let @xmath127 $ ] be the subgraph of @xmath1 induced by vertices of @xmath114 and its @xmath22-neighbors . for a path @xmath128 , and two vertices @xmath129 of @xmath30 we write @xmath130 to denote the subpath @xmath131 of @xmath30 .
the following proposition is almost identical to thomassen s theorem 5.3 of @xcite , with the added condition that @xmath114 contains all precolored vertices .
the proof is included for completeness .
[ nicelyh ] let @xmath1 be a planar graph and @xmath10 a set of vertices .
let @xmath4 be an assignment of lists of colors such that @xmath11 for @xmath12 and @xmath9 for @xmath132 .
if there is an induced connected subgraph @xmath114 of @xmath1 containing all vertices from @xmath10 such that it can be nicely colored with respect to @xmath4 , then @xmath1 is @xmath4-colorable .
note if @xmath133 for each @xmath134 then every proper coloring of @xmath114 is a nice coloring .
consider a nice coloring @xmath115 of @xmath114 .
then @xmath135 for all @xmath123 and @xmath136 for all @xmath137 .
therefore , by thomassen
s theorem , @xmath138 is @xmath139-colorable .
together with the coloring @xmath115 of @xmath114 , this gives a proper @xmath4-coloring of @xmath1 as @xmath140 for all @xmath137 .
[ properties - shortest - path ] let @xmath30 be a shortest @xmath42-path in a planar graph @xmath1 , where @xmath128 with @xmath141 , @xmath142 . then the following properties hold : + ( 1 ) for all @xmath143 , @xmath144 , + ( 2 ) for every @xmath145 , @xmath146 in @xmath1 unless @xmath147 , for @xmath148 , + ( 3 ) if @xmath149 for some @xmath143 , then @xmath150 , for @xmath151 , + ( 4 ) if there is no separating @xmath55 or @xmath56 in @xmath1 , then for each @xmath25 with @xmath151 there is at most one vertex @xmath143 such that @xmath150
. items ( 1)-(3 ) hold because @xmath30 is a shortest @xmath42-path . to see the validity of item ( 4 ) ,
assume there are two vertices adjacent to @xmath152 .
then it is easy to verify that there is either a separating @xmath55 or a separating @xmath56 in @xmath1 .
note that lemma [ properties - shortest - path ] implies that if @xmath30 is a shortest path between two vertices of a planar graph @xmath1 , then every block of @xmath153 with at least three vertices of @xmath30 has vertex set @xmath154 , for some @xmath155 , where @xmath156 are consecutive vertices of @xmath30 and @xmath157 , for @xmath158 .
observe that because @xmath30 is a shortest path and there are no separating @xmath55s or @xmath56s in @xmath1 , the vertices of @xmath159 form an independent set .
we call a block of @xmath153 with @xmath25 vertices of @xmath30 an * i - block * , @xmath160
. see figure [ fig : blocks ] for examples of blocks in @xmath153 .
note also that the block - cut - vertex tree of @xmath153 is a path . , where the bold line indicates @xmath30 .
] note that if @xmath153 has a cut - edge , that edge is in @xmath30 , and if @xmath153 has a cut - vertex , that vertex is in @xmath30 . we shall need a notion of a nontrivial block which will allow us to focus on subpaths of @xmath30 and not worry about the boundary conditions . for a shortest @xmath161-path @xmath162
, we say an edge @xmath163 is a * nontrivial cut - edge * of @xmath164 if @xmath163 is a cut - edge not incident to either @xmath165 or @xmath166 ; we say @xmath167 is a * nontrivial block * of @xmath164 if @xmath167 is a block that does not contain @xmath165 or @xmath166 .
we say a block @xmath167 is a * remote nontrivial block * of @xmath164 if @xmath168 where @xmath169 and @xmath170 are distinct nontrivial blocks of @xmath164 .
let @xmath171 and let @xmath172 . if @xmath163 is a nontrivial cut - edge in @xmath164 , then it is easy to see that @xmath163 is a cut - edge in @xmath153 ; if @xmath167 is a nontrivial block of @xmath164 , then @xmath167 is a block of @xmath153 .
[ nice1 ] let @xmath30 be a shortest @xmath42-path in a planar graph @xmath1 , where @xmath128 with @xmath141 and @xmath142 .
let @xmath1 have no separating @xmath55 and no separating @xmath56 .
let @xmath173 be an assignment of lists of colors to the vertices of @xmath1 where @xmath174 with @xmath175 and @xmath176 for @xmath177 .
then @xmath30 can be nicely colored with respect to @xmath4 .
assume @xmath178 , where @xmath179 , has been colored nicely by @xmath115 and @xmath180 , @xmath181 . if there is no @xmath143 such that @xmath182 , then color @xmath183 arbitrarily from its list so that @xmath184 . if there is a @xmath143 such that @xmath182 , choose a color for @xmath183 more carefully .
if @xmath34 or @xmath185 is not in @xmath186 , then choose @xmath187 from @xmath188 .
otherwise , @xmath189 , for some colors @xmath190 . if @xmath191 , let @xmath192 . if @xmath193 , then there is @xmath194 .
let @xmath195 . in each case
, we have constructed a nice coloring of @xmath196 . since @xmath197 for @xmath198
, the above argument may be applied along @xmath30 up through @xmath32 so that @xmath30 is nicely colored .
[ nice2 ] let @xmath30 be a shortest @xmath42-path in a planar graph @xmath1 , where @xmath199 with @xmath141 and @xmath142 .
let @xmath1 have no separating @xmath55 and no separating @xmath56 .
let @xmath173 be an assignment of lists of colors to the vertices of @xmath1 where @xmath200 and @xmath176 for @xmath201 .
then @xmath30 can be nicely colored with respect to @xmath4 .
the proof is by induction on @xmath202 .
+ if @xmath203 , the statement follows trivially .
if @xmath204 , we can assume that there is a vertex @xmath33 , where @xmath205 , otherwise color @xmath30 properly from @xmath4 . if @xmath206 for some @xmath207 , let @xmath208 and color @xmath209 arbitrarily from @xmath210 . if @xmath211 , then @xmath212 and there is a color @xmath213
assume without loss of generality that @xmath214 .
then let @xmath215 , and color @xmath216 arbitrarily from their lists so the path @xmath217 is properly colored .
as a result @xmath218
. now assume the result holds for shortest paths on fewer than @xmath219 vertices .
let @xmath220 .
color @xmath221 nicely with a coloring @xmath115 .
if there is no vertex outside of @xmath30 adjacent to @xmath222 and @xmath32 , then choose @xmath223 from @xmath224 .
this gives a nice coloring of @xmath30 .
so assume there is a vertex @xmath143 such that @xmath225 .
if @xmath226 or @xmath227 , then let @xmath228 . if @xmath229 and @xmath230 , then @xmath231 contains a color not in @xmath186 .
assign this color to @xmath32 to obtain a nice coloring of @xmath30 .
so we can assume @xmath232 or @xmath233 .
if @xmath234 , let @xmath235 providing a nice coloring of @xmath30 .
thus we can assume @xmath236 , @xmath237 for all @xmath238 , and @xmath239 .
apply induction to @xmath240 in the graph @xmath241 induced in @xmath1 by this path and its neighbors , with a new list @xmath242 assigned to @xmath243 and all other old lists .
there is a nice coloring @xmath244 of @xmath240 in @xmath241 .
note that it is a nice coloring of @xmath240 in @xmath1 .
we either have @xmath245 , for some @xmath238 , or @xmath246 . if @xmath247 , color @xmath248 first so that if there is @xmath249 , then @xmath250 . then let @xmath251 . if @xmath252 , without loss of generality say @xmath253 , then let @xmath254 and color @xmath248 so that @xmath250 . [ nice3 ]
let @xmath30 be a shortest @xmath42-path in a planar graph @xmath1 , where @xmath128 with @xmath141 and @xmath142 .
let @xmath1 have no separating @xmath55 and no separating @xmath56 .
let @xmath173 be an assignment of lists of colors to the vertices of @xmath1 with @xmath97 and @xmath176 for @xmath201 .
assume @xmath153 has at least two cut - edges .
then @xmath30 can be nicely colored with respect to @xmath4 .
let @xmath255 and @xmath256 be two cut - edges of @xmath153 , where @xmath257 . using lemma [ nice1 ] color @xmath258 and @xmath259 nicely with a coloring @xmath115 .
if @xmath260 , we are done by giving @xmath261 a color different from @xmath262 and @xmath263 .
otherwise , delete @xmath262 from @xmath264 , delete @xmath263 from @xmath265 and color @xmath266 nicely from the updated lists using lemma [ nice2 ] .
since @xmath258 , @xmath266 , and @xmath259 do not have pairwise common neighbors in @xmath267 , this gives a nicely colored @xmath30 .
the next lemma is a key lemma in this paper , stating that either a given shortest path between two precolored vertices could be nicely colored , or another subgraph that is close to that path could be nicely colored . for a path @xmath162 , a center @xmath39 of @xmath162 , and an even positive integer @xmath268
, we call the two vertices of @xmath162 at distance ( in @xmath162 ) @xmath269 from @xmath39 the * @xmath16-tag vertices * with respect to @xmath39 , or simply tag vertices , of @xmath162 .
[ nice4 ] let @xmath30 be a shortest @xmath42-path in a planar graph @xmath1 with a center @xmath39 and @xmath270-tag vertices @xmath271 , @xmath272 with respect to @xmath39 , where @xmath273 with @xmath274 and @xmath275 .
assume @xmath1 contains no separating @xmath55 or @xmath56 and no configuration @xmath64 or @xmath65 with @xmath276 .
let @xmath173 be an assignment of lists of colors to the vertices of @xmath1 such that @xmath97 and @xmath176 for @xmath277 .
then there is a connected graph @xmath278 such that every vertex of @xmath279 is at distance at most @xmath280 from @xmath39 , and @xmath114 can be nicely colored from @xmath4 .
recall that every vertex in @xmath281 is adjacent to at most three vertices in @xmath30 , and if a vertex from @xmath159 is adjacent to vertices in @xmath30 , these vertices in @xmath30 must be consecutive .
_ observation 1 .
_ if @xmath153 has a @xmath282-block @xmath167 , for a @xmath283 , then there is a shortest @xmath42-path @xmath284 such that @xmath285 has a nontrivial cut - edge .
let @xmath167 contain @xmath286 and vertices @xmath287 not in @xmath288 , where @xmath289 , for @xmath290 .
consider the shortest @xmath42-path @xmath291 then it is a routine check to see that @xmath292 is a nontrivial cut - edge in @xmath285 , as shown in figure [ fig : observation1 ] .
is a nontrivial cut - edge in @xmath285 . ] _ observation 2 .
_ we can assume at least one of the following holds : + ( 1 ) for every shortest @xmath293-path @xmath284 , each nontrivial block of @xmath285 is either a @xmath22- , @xmath20- , or @xmath0-block , + ( 2 ) for every shortest @xmath294-path @xmath284 , each nontrivial block of @xmath285 is either a @xmath22- , @xmath20- , or @xmath0-block .
if there is a shortest @xmath293-path @xmath162 such that @xmath164 has a nontrivial cut - edge and there is a shortest @xmath295-path @xmath296 such that @xmath297 has a nontrivial cut - edge , then lemma [ nice3 ] implies @xmath298 can be nicely colored .
assume , without loss of generality that for every shortest @xmath293-path @xmath284 , @xmath285 has no nontrivial cut - edges .
then observation 1 implies there is no @xmath282-block of @xmath285 with @xmath283 .
assume that part ( 1 ) of observation 2 holds .
let @xmath299 , @xmath300 .
let @xmath35 be a shortest @xmath301-path with the largest number of @xmath22-neighbors .
_ observation 3 .
_ if , for some shortest @xmath161-path @xmath35 with maximum number of @xmath22-neighbors , @xmath302 has a nontrivial @xmath22-block , then there is a graph @xmath303 satisfying the conditions of the lemma .
if such a block @xmath167 were to exist , say with consecutive vertices @xmath304 of @xmath35 and @xmath305 , @xmath306 , then there is no vertex @xmath307 , such that @xmath308 is adjacent to @xmath33 and two other vertices of @xmath35 , otherwise there is a shortest @xmath161-path with more @xmath22-neighbors than @xmath35 .
let @xmath309 be the graph induced by vertices of @xmath310 and @xmath33 .
nicely color @xmath311 and nicely color @xmath312 , then properly color @xmath33 and @xmath313 from remaining available colors in their lists .
since there is no @xmath22-neighbor of @xmath309 adjacent to @xmath313 and there is no such @xmath22-neighbor adjacent to @xmath33 , this coloring is a nice coloring of @xmath303 .
thus , we can assume that all nontrivial blocks of @xmath302 are @xmath20- or @xmath0-blocks .
since @xmath314 , there are remote nontrivial blocks in @xmath302 .
_ observation 4 .
_ if @xmath302 has a remote @xmath0-block for some shortest @xmath161-path @xmath35 with maximum number of @xmath22-neighbors , then there is a graph @xmath303 satisfying the conditions of the lemma .
assume there is such a block @xmath167 with consecutive vertices @xmath315 of @xmath35 and vertices @xmath316 , @xmath317,@xmath318 not in @xmath35 such that @xmath319 , for @xmath320 . from observation 3
, we can assume that every nontrivial block of @xmath302 is either a @xmath20- or a @xmath0-block .
note first that there is no vertex @xmath33 adjacent to @xmath316 and two vertices of @xmath35 , and there is no vertex @xmath33 adjacent to @xmath318 and two vertices of @xmath35 . indeed , assume otherwise that there is a vertex @xmath33 adjacent to @xmath316 and two vertices of @xmath35 . then @xmath321 .
since all nontrivial blocks of @xmath35 have at least four vertices of @xmath35 , and there are nontrivial blocks @xmath169 and @xmath170 of @xmath302 such that @xmath168 , we see there is a vertex @xmath322 adjacent to @xmath323 and there is a vertex @xmath324 adjacent to @xmath325 , as shown in figure [ fig : obs4case2 ] .
then @xmath326 is a shortest @xmath161-path @xmath296 with a block in @xmath297 having at least six vertices of @xmath296 , as can be seen in figure [ fig : obs4case2 ] .
this is a contradiction to observation 2 .
similarly , it is impossible to have a vertex @xmath33 adjacent to @xmath318 and two vertices of @xmath35 .
assume now that there is no vertex @xmath33 adjacent to @xmath316 and @xmath318 and a vertex of @xmath35 .
let @xmath309 be a graph induced by vertices of @xmath310 and @xmath327 , as shown in figure [ fig : obs4case1 ] .
note that while @xmath317 is shown in the figure , it is not a vertex in the graph @xmath303 .
, as described in a case of observation 4 . ]
to color @xmath328 nicely , first color @xmath329 and @xmath330 nicely , then color @xmath331 properly so @xmath317 has at least three colors remaining in its list after the removal of colors used on adjacent vertices , and finally color @xmath316 and @xmath318 using available colors . _ fact .
_ we can assume for every shortest @xmath161-path @xmath332 , where @xmath333 , @xmath334 , with maximum number of @xmath22-neighbors and for every remote nontrivial @xmath0-block @xmath167 of @xmath35 with vertices @xmath335 of @xmath35 and @xmath336 , for @xmath337 , there is a vertex @xmath338 .
consider the shortest @xmath161-path @xmath339 .
there must be a vertex @xmath340 , otherwise @xmath341 is a cut - edge in @xmath342 .
there must also be a vertex @xmath343 , otherwise @xmath344 is a cut - edge in @xmath342 .
next , consider the shortest @xmath161-path @xmath345 .
there must be a vertex @xmath346 , otherwise @xmath347 is a cut - edge in @xmath348 .
there must also be a vertex @xmath349 , otherwise @xmath350 is a cut - edge in @xmath348 .
finally , consider the shortest @xmath161-paths @xmath351 by the fact above , there must be vertices @xmath352 and @xmath353 such that @xmath354 and @xmath355 .
thus , @xmath45 $ ] where @xmath356 corresponds to the configuration @xmath65 in @xmath1 , as seen in figure [ fig : obs4case3 ] , where the bold vertices represent @xmath23 . as it arises locally around @xmath35 . ]
this completes the proof of observation 4 . to summarize , we know that for any shortest @xmath161-path @xmath35 , every nontrivial block of @xmath302 is a @xmath22- , @xmath20- , or @xmath0-block .
moreover , if @xmath35 has the largest number of @xmath22-neighbors among all such shortest @xmath301-paths , then every remote nontrivial block of @xmath302 is a @xmath20-block . to conclude the proof of lemma [ nice4 ] ,
let @xmath35 be a shortest @xmath161-path with the largest number of @xmath22-neighbors among all such shortest @xmath161-paths .
consider a remote nontrivial block of @xmath302 with consecutive vertices @xmath357 of @xmath35 and vertices @xmath358 not in @xmath35 such that @xmath359 for @xmath360 .
there is no @xmath33 adjacent to @xmath316 and two vertices of @xmath35 , and there is no vertex @xmath33 adjacent to @xmath317 and two vertices of @xmath35 . let @xmath309 be the graph induced by vertices of @xmath310 and @xmath358 .
to color @xmath328 nicely , first color @xmath329 and @xmath361 nicely , then color @xmath362 $ ] properly .
there is , without loss of generality , a vertex @xmath33 adjacent to @xmath316 and two vertices of @xmath35 .
if @xmath363 for some vertex @xmath308 , then consider the path @xmath364 . then in @xmath164 there is a @xmath282-block with @xmath365 , a contradiction as shown on the left in figure [ fig : lemma5case2 ] .
so assume @xmath366 .
consider a path @xmath367 .
observe that the edge @xmath368 is a nontrivial cut - edge in @xmath297 unless there is a vertex @xmath308 adjacent to @xmath369 and another vertex of @xmath296 .
this third vertex is either @xmath370 or @xmath33 .
it could not be @xmath370 as shown before .
such that @xmath371 . ]
thus , @xmath372 .
see the right hand side of figure [ fig : lemma5case2 ] .
similarly , by considering the path @xmath373 , we see there is a vertex @xmath374 .
finally , by considering the path @xmath375 , we have a vertex @xmath376 .
but now the graph @xmath45 $ ] , where @xmath377 , gives the configuration @xmath64 in @xmath1 , as seen in figure [ fig : lemma5case2 ] where the bold vertices represent @xmath23 .
we see now , that a graph @xmath303 in all the cases above was constructed by taking the union of @xmath378 , @xmath379 , and a graph @xmath279 induced by a shortest @xmath161-path @xmath35 ( of length @xmath380 ) and , perhaps some vertices at distance @xmath34 from @xmath35 .
thus , any vertex of @xmath279 is at distance at most @xmath280 to @xmath381 .
note if the two precolored vertices are adjacent , then the coloring is extendable by thomassen s theorem . in general , we use induction on @xmath382 where the base case is precolored @xmath31 and @xmath32 connected by an edge .
assume @xmath1 is connected , otherwise the result follows trivially by induction . _ claim .
_ @xmath1 has no separating @xmath55 or @xmath56 .
+ let @xmath383 be a vertex set of such a separating cycle . by the assumption of the theorem , @xmath383 does not separate @xmath69 .
let @xmath384 and @xmath385 be the vertex sets of disconnected plane graphs obtained by removing @xmath386 $ ] from @xmath1 , such that @xmath387 . by induction ,
color @xmath388 $ ] from @xmath4 .
this gives a proper coloring @xmath115 of @xmath383 .
now , in @xmath389 $ ] , there is a face with vertex set @xmath383 having color lists of size @xmath34 and all other vertices have color lists of size @xmath0 .
thus , by theorem [ bms ] , @xmath389 $ ] is colorable from the corresponding lists .
let @xmath390 be a shortest @xmath42-path in @xmath1 , with @xmath274 and @xmath275 , for @xmath391 .
by lemma [ nice1 ] there is a nice coloring @xmath115 of @xmath392 . by lemma [ properties - shortest - path](4 )
there is at most one vertex adjacent to @xmath393 and at most one vertex adjacent to @xmath394 , if @xmath395 .
let @xmath396 .
if there is no vertex @xmath18 , with @xmath397 , and no vertex @xmath18 , with @xmath398 , then @xmath115 is a nice coloring of @xmath30 .
assume that there is a vertex @xmath19 , with @xmath399 , and there is no vertex @xmath18 , with @xmath397 , or , the other way around , there is no vertex @xmath18 , with @xmath398 and there is a vertex @xmath19 , with @xmath400 .
then @xmath115 is a proper coloring of @xmath30 such that @xmath401 for every @xmath402 , and @xmath403 . deleting @xmath30 and
the corresponding colors from the lists of their neighbors in @xmath404 produces a list assignment where all vertices in a face containing @xmath405 have lists of size at least @xmath22 ( except for @xmath19 ) , and all other vertices have lists of size @xmath0 . using thomassen s theorem ,
@xmath404 can be colored from these lists .
together with the coloring @xmath115 of @xmath30 , it gives a proper @xmath4-coloring of @xmath1 .
finally , assume there is a vertex @xmath18 , with @xmath398 , and there is a vertex @xmath33 , with @xmath406 .
note that there is at most one additional vertex adjacent to @xmath248 and @xmath407 , call it @xmath352 if it exists .
delete @xmath30 from @xmath1 and add two new adjacent vertices @xmath408 and @xmath409 in the resulting face , also add edges @xmath410 , where @xmath411 and @xmath412 , where @xmath413 .
choose two new colors @xmath78 and @xmath414 not used in any of the lists assigned to vertices of @xmath1 .
let @xmath415 , @xmath416 , @xmath417 , @xmath418 , @xmath419 , @xmath420 , and @xmath421 .
for every other vertex of this modified graph , let @xmath422 be equal to @xmath139 .
see figure [ fig : thm2 ] for an illustration of this process . and
@xmath409 in @xmath404 . ] observe that @xmath422 satisfies the conditions of thomassen s theorem , so there is a proper @xmath422-coloring of this graph .
thus , there is a proper @xmath422-coloring of @xmath404 , where no vertex uses colors @xmath78 or @xmath414 .
this is a proper @xmath139-coloring of @xmath404 .
together with the coloring @xmath115 of @xmath30 , it gives a proper @xmath4-coloring of @xmath1 .
let @xmath35 be a @xmath66-steiner tree in @xmath241 , a reduced graph of @xmath1 satisfying the conditions of the theorem .
let @xmath4 be an assignment of lists of colors to vertices of @xmath1 such that @xmath423 for @xmath12 and @xmath9 for @xmath424 .
we first color @xmath241 , then extend it to a proper @xmath4-coloring of @xmath1 .
to color @xmath241 , first color special vertices of @xmath35 which are not in @xmath10 arbitrarily from their lists .
let @xmath425 be the set of branches in @xmath35 and let @xmath426 with endpoints @xmath427 .
let @xmath428 be the graph obtained by applying lemma [ nice4 ] to @xmath30 and @xmath429 be a nice coloring of @xmath430 from the corresponding lists ( see figure [ fig : h ] ) .
finally , let @xmath115 be a coloring of @xmath431 , such that @xmath432 if @xmath433 .
_ claim 1 . _
the coloring @xmath115 is a nice coloring of @xmath114 .
let @xmath434 be two vertices of @xmath114 that do not belong to the same @xmath430 .
we shall prove that @xmath18 and @xmath435 do not have common neighbors outside of @xmath114 and they are not adjacent .
let @xmath436 , @xmath437 , @xmath438 , @xmath439 . if @xmath440 , then @xmath18 and @xmath435 do not have a common neighbor outside of @xmath35 and they are not adjacent by part @xmath441 of the definition of a @xmath37-steiner tree . if @xmath442 , then @xmath443 , thus @xmath444 , where @xmath445 is a center of @xmath284 , as follows from lemma [ nice4 ] . from part @xmath446 of the definition of a @xmath37-steiner tree ,
we have that @xmath447 .
thus @xmath448 when @xmath449 . finally if @xmath450 , then @xmath451 and @xmath443 .
thus @xmath452 , where @xmath453 are centers of @xmath30 and @xmath284 , respectively . moreover @xmath454 .
thus @xmath455 if @xmath456 . it follows that @xmath115 is a proper coloring of @xmath114 . to show that @xmath115 is nice ,
consider a vertex @xmath32 adjacent to @xmath114 .
we see that @xmath32 is adjacent to non - special vertices of @xmath430 for at most one branch @xmath30 of @xmath35 .
since @xmath115 is a nice coloring of @xmath430 , it follows that @xmath457 . to conclude the proof of claim 1 ,
recall that @xmath114 is a connected graph containing all vertices of @xmath10 .
proposition [ nicelyh ] implies that @xmath241 is colorable from @xmath4 . to show that @xmath1 is colorable , it is sufficient to observe the following . obtained in the proof of theorem [ steiner ] . ] _ claim 2 .
_ let @xmath93 be a graph , @xmath10 be a set of vertices , and @xmath4 be an assignment of lists of size @xmath0 to vertices of @xmath458 and lists of size @xmath34 to vertices of @xmath10 .
let @xmath459 be a reduction of @xmath93 .
if @xmath460 has a proper coloring from lists @xmath4 then @xmath93 has a proper coloring from lists @xmath4 .
let @xmath115 be a proper coloring of @xmath460 from lists @xmath4 .
if @xmath460 was obtained from @xmath93 by removing the vertices in a region separated by @xmath55 or @xmath56 , these vertices can be colored properly from @xmath4 using theorem [ bms ] .
if @xmath460 was obtained from @xmath93 by removing the set @xmath23 of @xmath20 vertices , @xmath461 of configuration @xmath47 , we see that @xmath462 , @xmath463 for the two vertices @xmath464 of degree two in @xmath465 $ ] and @xmath466 , @xmath467 for the two vertices @xmath468 of degree three in @xmath465 $ ] . in the subgraph
@xmath465 $ ] each vertex has list size equal to its degree under list assignment @xmath139 .
an @xmath139-coloring of @xmath465 $ ] can be found directly or by the results of @xcite .
thus @xmath93 has a proper coloring from lists @xmath4 .
if @xmath460 was obtained from @xmath93 by removing the set @xmath23 of @xmath469 vertices @xmath470 of configuration @xmath53 , then we see that @xmath471 , @xmath472 , and @xmath473 .
let @xmath474 , so color @xmath33 with @xmath78 and remove @xmath78 from @xmath475 .
what remains to be colored is a @xmath48-cycle with vertices having lists of size at least @xmath185 , which is colorable by the classification of all @xmath185-choosable graphs by erds et al .
since @xmath465 $ ] is properly colorable from lists @xmath139 , @xmath93 is properly colorable from lists @xmath4 .
this proves claim 2 .
since @xmath241 was obtained from @xmath1 via a sequence of reductions , the theorem follows .
\(1 ) let @xmath4 be an assignment of lists of colors to vertices of @xmath1 such that @xmath176 for all @xmath476 and @xmath477 for all @xmath478 .
if @xmath10 is a set of vertices and edges with pairwise distance at least @xmath22 , then for all @xmath479 , @xmath18 is adjacent to at most two vertices of @xmath10 .
thus , for every proper coloring @xmath115 of @xmath72 $ ] from the corresponding lists @xmath4 and for all @xmath476 , we have @xmath480 .
moreover , @xmath481 belongs to the frontier of a face in @xmath482 .
thus , by proposition [ nicelyh ] , @xmath1 is colorable from lists @xmath4 .
\(2 ) without loss of generality , assume @xmath70 is on the unbounded face of @xmath1 .
let @xmath483 be a set of at most six precolored vertices on the boundary of @xmath70 .
fix an assignment @xmath4 of lists of colors to the vertices of @xmath1 with @xmath9 for all @xmath132 and @xmath477 for all @xmath478 .
we shall show that @xmath1 is @xmath4-colorable provided the three forbidden configurations are not present .
we shall create a new graph @xmath241 on the vertex set of @xmath1 with new lists @xmath422 .
let @xmath484 be distinct colors not present in @xmath231 for any @xmath6 .
let @xmath422 be a new list assignment with @xmath485 for @xmath486 and @xmath487 for each @xmath132 , where @xmath488 is the set of colors used in lists @xmath4 of vertices in @xmath489 and @xmath490 is an arbitrary subset of the set of colors used in lists @xmath422 of vertices of @xmath491 , such that @xmath492 .
in creating @xmath422 we simply replaced the colors originally assigned to @xmath10 with new distinct colors , and replaced the old colors in the lists of vertices in the neighborhood of vertices of @xmath10 .
let a new plane graph @xmath241 be obtained from @xmath1 by removing the edges @xmath493 for @xmath494 that correspond to non - consecutive vertices of @xmath70 , and adding all edges @xmath493 for @xmath494 in the unbounded face of @xmath1 .
the resulting graph has a new unbounded face with vertex set @xmath10 , and , perhaps , some new edges . by theorem 7 , @xmath241 is @xmath422-colorable by a coloring @xmath115 provided the three forbidden configurations are not present .
moreover , for any @xmath424 , we have @xmath495 , so @xmath496 and @xmath497 if @xmath498 . to create a proper @xmath4-coloring of @xmath1 ,
replace the color @xmath499 with an element of @xmath500 for @xmath494 .
[ proof of theorem [ two - faces ] ] delete @xmath10 and the corresponding colors from the lists of adjacent vertices .
there are at most two faces , @xmath501 or @xmath501 and @xmath502 , in the graph @xmath482 such that the vertices adjacent to @xmath10 in @xmath1 belong to the boundaries of these two faces .
these vertices have lists of size at least @xmath20 , and all other vertices in @xmath482 have lists of size at least @xmath0 .
call the resulting lists @xmath422 . add a vertex @xmath503 to the face @xmath504 and make it adjacent to all vertices on @xmath504 , @xmath505 , or @xmath506 .
let @xmath78 be a color not used in any of the lists @xmath231 , @xmath507 .
let @xmath508 , @xmath509 , if @xmath510 and @xmath511 .
for all other vertices , let @xmath512 . applying theorem [ two ] to the resulting graph with lists @xmath513 allows for this graph to be properly colored from these lists .
we note here that it is not hard to see that this new graph does not contain any @xmath514-separating @xmath55s or @xmath56s because such a separating @xmath55 or @xmath56 would have to be made up of vertices and edges from the original graph and would have separated some of the precolored vertices of @xmath1 , a contradiction .
this coloring gives a proper coloring of @xmath482 from lists @xmath422 , and thus it gives a proper coloring of @xmath1 from lists @xmath4 .
we proved the question of albertson has a positive answer if there are no short cycles separating precolored vertices and there is a nice tree containing precolored vertices .
we note here that by the definition of a @xmath37-steiner tree , theorem [ steiner ] can be applied to plane graphs with precolored vertices that are not far apart .
for example , let @xmath1 be a 100-cycle with vertices @xmath515 and @xmath516 . then @xmath1 contains a @xmath517-steiner tree obtained from deleting @xmath518 and incident edges .
the centers of the branches are far apart , but @xmath519 .
we believe that in a planar triangulation either such a tree could always be found , or there are small reducible configurations such as shown in figure [ fig : reducible ] .
the reducible configurations @xmath47 and @xmath53 are just two in a family of many reducible configurations of those types . modifying the definition of a reduced graph to include the removal of every reducible @xmath46 and every reducible @xmath48-wheel leads us to the following question .
is it the case that every reduced planar triangulation with a set @xmath10 of precolored vertices with @xmath520 contains a @xmath66-steiner tree ?
if the above question has a positive answer , then by theorem [ steiner ] , the precoloring of @xmath10 extends to a @xmath0-list coloring of @xmath1 .
we did not strive to improve the constants here . with more careful calculations ,
one could easily obtain smaller constants .
the condition of no separating short cycles seems to be essential .
reducing the sizes of the lists , increasing the sizes of lists on so - called `` precolored '' vertices , or eliminating the distance condition in this problem is not possible even for a small number of precolored vertices , see figure [ fig : distsize ] .
figure [ fig : tentacles ] shows we can not reduce the sizes of the lists , even if the vertices are on the same face .
this graph belongs to a family of graphs where the length of each path along the unbounded face from the outer triangles to the inner triangle must be divisible by three .
it is not hard to see that if the vertices of the inner triangle are assigned colors @xmath521 , respectively , then one of the vertices with lists of size @xmath185 can not be colored .
however , we conjecture that a precoloring of two far - apart vertices is always extendable to a @xmath0-list coloring of a planar graph .
the authors wish to thank the anonymous referee for many helpful suggestions , especially an observation that greatly simplified and improved the proof of theorem [ one - face ] ( 2 ) .
99 m. albertson , _ you ca nt paint yourself into a corner _ ,
j. combin .
theory ser .
b 73 ( 1998 ) 189194 .
t. bhme , b. mohar , m. stiebitz , _
dirac s map - color theorem for choosability _ ,
j. graph theory 32 ( 1999 ) , no .
4 , 327339 . p. erds , a. l. rubin , h. taylor , _ choosability in graphs _ , proceedings of the west coast conference on combinatorics , graph theory and computing ( humboldt state univ . ,
arcata , calif . , 1979 ) , pp .
125157 , congress .
numer . , xxvi , utilitas math . ,
winnipeg , man . , 1980 .
a. v. kostochka , m. stiebitz , b. wirth , _ the colour theorems of brooks and gallai extended , _ discrete math . 162
( 1996 ) , no . 1 - 3 , 299303 . m. mirzakhani , _ a small non-4-choosable planar graph _ , bull
17 ( 1996 ) 1518 . c. thomassen , _ color - critical graphs on a fixed surface _ , j. combin .
theory ser .
b 70 ( 1997 ) , no . 1 , 67 - 100 . c. thomassen , _ every planar graph is 5-choosable _ , j. combin .
theory ser .
b 62 ( 1994 ) 180181 . c. thomassen ,
_ exponentially many 5-list - colorings of planar graphs _ ,
j. combin .
theory ser .
b 97 ( 2007 ) , no .
4 , 571583 . zs .
tuza , m. voigt , _ a note on planar 5-list - colouring : non - extendability at distance 4 _ , discrete math . 251
( 2002 ) 169172 . v. g. vizing ,
_ coloring the vertices of a graph in prescribed colors , _ methody diskret .
v teorii kodov i shem 29 ( 1976 ) , 310 ( in russian ) .
m. voigt , _ list colourings of planar graphs , _ discrete math . 120 ( 1993 ) , no . 1 - 3 , 215219 . m. voigt , b. wirth , _ on @xmath22-colorable non-@xmath20-choosable planar graphs ,
_ j. graph theory 24 ( 1997 ) , no .
3 , 233235 .
d. b. west , _ introduction to graph theory _ ( second edition ) , prentice hall ( 2001 ) . | a celebrated result of thomassen states that not only can every planar graph be colored properly with five colors , but no matter how arbitrary palettes of five colors are assigned to vertices , one can choose a color from the corresponding palette for each vertex so that the resulting coloring is proper .
this result is referred to as @xmath0-choosability of planar graphs .
albertson asked whether thomassen s theorem can be extended by precoloring some vertices which are at a large enough distance apart in a graph . here , among others ,
we answer the question in the case when the graph does not contain short cycles separating precolored vertices and when there is a wide " steiner tree containing all the precolored vertices . |
consider a heat conductor @xmath0 having ( positive ) constant initial temperature while its boundary is constantly kept at zero temperature .
this physical situation can be described by the following initial - boundary value problem for heat equation : @xmath1 here @xmath0 the _ heat conductor _ is a bounded domain in the euclidean space @xmath2 @xmath3 with lipschitz boundary and @xmath4 denotes the normalized temperature of the conductor at a point @xmath5 and time @xmath6 a _ hot spot _
@xmath7 is a point at which the temperature @xmath8 attains its maximum at each given time @xmath9 that is such that @xmath10 if @xmath0 is convex ( in this case @xmath11 is said a _ convex body _ ) , it is well - known by a result of brascamp and lieb @xcite that @xmath12 is concave in @xmath13 for every @xmath14 and this , together with the analyticity of @xmath8 in @xmath15 implies that for every @xmath14 there is a unique point @xmath16 at which the gradient @xmath17 of @xmath8 vanishes ( see also @xcite ) . the aim of this paper is to give quantitative information on the hot spot s location in a convex body
. a description of the evolution with time of the hot spot can be found in @xcite ; we summarize it here for the reader s convenience .
a classical result of varadhan s @xcite tells us where @xmath7 is located for small times : since @xmath18 ( here @xmath19 is the distance of @xmath13 from @xmath20 ) , we have that @xmath21 where @xmath22 and @xmath23 is the _ inradius _ of @xmath24 . in particular
, we have that @xmath25 for large times instead , we know that @xmath7 must be close to the maximum point @xmath26 of the first dirichlet eigenfunction @xmath27 of @xmath28 indeed , denoting with @xmath29 the eigenvalue corresponding to @xmath27 , we have that @xmath30 converges to @xmath27 locally in @xmath31 as @xmath32 goes to @xmath33 ; therefore ( see @xcite ) @xmath34 while it is relatively easy to locate the set @xmath35 by geometrical means , does not give much information : locating either @xmath7 or @xmath26 has more or less the same difficulty . in this paper , we shall develop geometrical means to estimate the location of @xmath7 ( or @xmath26 ) , based on two kinds of arguments .
the former is somehow reminiscent of the proof of the maximum principle of alexandrov , bakelmann and pucci and of some ideas contained in @xcite , concerning properties of solutions of the monge - amp ` ere equation .
the estimates obtained in this way are applicable to any open bounded set , not necessarily convex .
let @xmath0 be a bounded open set and denote by @xmath36 the closure of its convex hull ; we shall prove the following inequality ( see theorem [ th : nonconvessi ] ) : @xmath37^n}\,.\ ] ] here , @xmath38 is the diameter of @xmath0 and @xmath39 is a constant , depending only on the dimension @xmath40 , for which we will give the precise expression ; observe that the quantity @xmath41 is scale invariant .
when @xmath24 is convex , more explicit bounds can be derived ; for instance , the following one relates the distance of @xmath26 from @xmath20 to the inradius and the diameter : @xmath42 where again @xmath43 is a constant depending only on @xmath40 ( see theorem [ th : conv ] for its expression ) .
we point out that the so called _ santal point _ of @xmath24 always satisfies , hence this can also be used to locate such a point ( see section [ sec : polar ] and remark [ santalo ] ) .
the latter argument relies instead on the following idea from @xcite .
let @xmath44 be the unit sphere in @xmath45 for @xmath46 and @xmath47 define the hyperplane @xmath48 and the two half - spaces @xmath49 ( here the symbol @xmath50 denotes the usual scalar product in @xmath51 ) .
suppose @xmath52 has non - empty intersection with the interior of the conductor @xmath0 and set @xmath53 then _ if the reflection @xmath54 of @xmath55 with respect to the hyperplane @xmath52 lies in @xmath56 then @xmath52 can not contain any critical point of @xmath57 _ this is a simple consequence of _ alexandrov s reflection principle _ based on hopf s boundary point lemma ( see section [ sec : alexandrov ] for details ) .
based on this remark , for a convex body @xmath36 we can define a ( convex ) set @xmath58 the _ heart _ of @xmath36 such that @xmath59 for every @xmath14 ( in fact , we will prove that @xmath60 can not contain the hot spot @xmath7 for any @xmath14 ) .
the heart @xmath58 of @xmath36 is easily obtained as the set @xmath61 as we shall see , the two methods have their advantages and drawbacks , but they are , in a sense , complementary . on the one hand , while inequalities and are quite rough in the case in which @xmath0 has some simmetry ( e.g. they do not allow to precisely locate @xmath26 even when @xmath24 is a ball ) , by the second argument , the problem of locating @xmath7 is quite trivial ; on the other hand , while in some cases ( e.g. when @xmath0 has no simmetries or @xmath20 contains some flat parts , as example [ exa : joint ] explains ) , we can not exclude that the heart of @xmath36 extends up to the boundary @xmath62 of @xmath63 estimates and turn out to be useful to quantitatively bound @xmath7 away from @xmath64 thus , we believe that a joint use of both of them provides a very useful method to locate @xmath7 or @xmath26 . studies on the problem of locating @xmath26 can also be found in @xcite : there , by arguments different from ours and for the two - dimensional case , the location of @xmath26 is estimated within a distance comparable to the inradius , uniformly for arbitrarily large diameter . in section
[ sec : folding ] we shall relate @xmath58 to a function @xmath65 of the direction @xmath66 the _ maximal folding function _ and we will construct ways to characterize it .
we will also connect @xmath67 to the fourier transform of the characteristic function of @xmath36 : this should have some interest from a numerical point of view .
finally , in section [ sec : algoritmo ] , we will present an algorithm to compute @xmath67 when @xmath36 is a polyhedron : based on this algorithm , we shall present some numerical computations .
in this section , if not otherwise specified , @xmath24 is a bounded open set and we denote by @xmath36 the closure of the convex hull of @xmath24 .
notice that @xmath36 is a convex body , that is a compact convex set , with non empty interior . in
what follows , @xmath68 denotes the @xmath40-dimensional lebesgue measure of a set @xmath69 and @xmath70 the @xmath71-dimensional hausdorff measure of its boundary ; also , @xmath72 will be the volume of the unit ball in @xmath73 we recall here some notations from @xcite .
the _ gauge function @xmath74 of @xmath36 centered at a point @xmath75 _ is the function defined by @xmath76 observe that we have @xmath77 for every @xmath14 ; in particular , if @xmath78 then @xmath79 is 1-homogeneous .
we set @xmath80 so that @xmath81 is the convex function whose graph is the cone projecting @xmath62 from the point @xmath82 .
it is also useful to recall the definition of the _ support function _
@xmath83 of @xmath63 that is @xmath84 as it is easily seen , @xmath83 is a @xmath85-homogeneous convex function ; viceversa , to any convex @xmath85-homogeneous function @xmath86 it corresponds exactly one convex body whose support function is @xmath87 ( refer to @xcite , for instance ) .
the _ polar set of @xmath36 with respect to @xmath88 _ is the convex set @xmath89 coinciding with the unit ball of the `` norm '' . ]
@xmath90 centered at @xmath88 , that is @xmath91 if @xmath88 is in the interior of @xmath36 , then @xmath92 is compact .
observe that this can be equivalently defined as @xmath93 we also recall that for every convex body @xmath36 the function @xmath94 $ ] defined by @xmath95 attains a positive minimum at some point @xmath96 , which is called the _ santal point _ of @xmath36 ( see @xcite ) .
when referring to the polar of @xmath36 with respect to its santal point , we simply write @xmath97 , instead of @xmath98 : we will see that the method developed in the next subsection for estimating the hot spot , applies to the santal point as well .
it is not difficult to see that @xmath92 coincides ( up to a translation ) with the _ subdifferential _ @xmath99 of @xmath81 at the point @xmath88 , i.e. @xmath100 where for every @xmath101 @xmath102 finally , we will need the following monotonicity property of the subdifferential of a function : the proof can be found in @xcite .
[ th : sub ] let @xmath103 and @xmath104 be continuous convex functions on @xmath36 such that @xmath105 on @xmath64 define : @xmath106 if @xmath107 in @xmath36 , then @xmath108 the following result holds for a general domain @xmath0 and is the cornerstone of our estimates .
[ thmestimx0 ] ( i ) let @xmath8 be the solution of the initial - boundary value problem and , for every fixed time @xmath109 , let @xmath110 be a hot spot at time @xmath32 , that is a point where the value @xmath111 is attained .
then @xmath112 is the _ contact set _ at time @xmath32 , i.e. the subset of @xmath0 where @xmath113 coincides with its convex envelope .
\(ii ) let @xmath114 and @xmath27 be respectively the first dirichlet eigenvalue and eigenfunction of @xmath115 in @xmath116 let @xmath26 be a maximum point of @xmath27 in @xmath0 and set @xmath117 then @xmath118 is the contact set of @xmath27 , i.e. the subset of @xmath0 where @xmath119 coincides with its convex envelope .
\(i ) for @xmath120 let @xmath121 denote the convex envelope of @xmath122 then , @xmath123 is a continuous convex function in @xmath36 , such that @xmath124 on @xmath125 the function @xmath126 defined by @xmath127 for @xmath128 is such that @xmath129 and hence @xmath130 by lemma [ th : sub ] . by the rescaling properties of the subdifferential and ,
we know that @xmath131 thus , @xmath132 on the other hand , by sard s lemma and the formula for change of variables ( see for instance ( * ? ? ?
* section 1.4.2 ) ) , we obtain @xmath133 with @xmath134 .
observe that the contact set is not empty , thanks to the fact that @xmath135 and moreover we have @xmath136 .
now , by the arithmetic - geometric mean inequality , we have in @xmath137 that @xmath138 which yields @xmath139 therefore , we finally obtain that @xmath140^{-n}\int_{\cc(t ) } |{\delta}u(x , t)|^n\ dx\,.\ ] ] and we conclude the proof by simply using the equation @xmath141
\(ii ) the proof runs similarly to case ( i ) .
estimates and are generally difficult to handle . the following weaker forms of may be more useful .
[ cor : estimxinfty ] under the same assumptions as in theorem [ thmestimx0 ] , we have : latexmath:[\[\label{coroll } and latexmath:[\[\label{coroll2 } we recall that @xmath97 denotes the polar set of @xmath36 with respect to the santal point .
[ rem : coroll ] as is well - known , @xmath143 can be estimated from below by @xmath144 where @xmath145 is a positive constant ( see @xcite ) .
thus , becomes @xmath146^n.\ ] ] [ rem : gnn ] in @xcite the following problem is posed : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ [ ... ] suppose @xmath147 is a solution of @xmath148 in a bounded domain @xmath24 in @xmath149 , say @xmath150 .
is there some @xmath151 depending only on @xmath24 ( i.e. , independent of @xmath152 and @xmath8 ) such that @xmath8 has no stationary point in an @xmath153-neighbourhood of @xmath20 ?
this is true for @xmath154 in case @xmath155 for @xmath156 , but for @xmath157 the problem is open . [ ... ] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ here , we point out that by the same arguments used for theorem [ thmestimx0 ] , we can easily prove the following estimate : @xmath158 is the lipschitz constant for @xmath152 and @xmath101 is a point where @xmath8 achieves its maximum @xmath159 when @xmath160 we obtain the inequality @xmath161 from below in a way similar to that of theorem [ th : nonconvessi ] below .
an interesting instance of occurs when @xmath162 in this case @xmath8 is the _ torsional creep _ of an infinite bar with cross - section @xmath163 we thus obtain : @xmath164 this inequality can also be viewed as an estimate for the maximum @xmath165 in the spirit of the alexandrov - bakelman - pucci principle . using the definition of the polar set , it is easy to see that @xmath166 goes to @xmath33 as the point @xmath13 approaches the boundary .
the following lemma gives a quantitative version of this fact and helps us to provide explicit estimates of the position of @xmath167 [ lmestimk * ] let @xmath88 be any point belonging to the interior of @xmath36 and define @xmath168 .
then @xmath169 and @xmath170 obviously @xmath36 is contained in the ball @xmath171 centered at @xmath88 with radius @xmath172 and in the halfspace @xmath173 supporting @xmath36 at any point @xmath174 such that @xmath175 set @xmath176 then @xmath177 whence @xmath178 now notice that @xmath179 is the convex envelope of the union of the ball @xmath180 and the point @xmath181 ; its volume is explicitly computed : @xmath182 with @xmath183.$ ] thus , is readily obtained by observing that the function of @xmath184 into the braces is increasing and hence bounded below by @xmath185 we are now ready to prove the first quantitative estimate on the location of @xmath26 : this will result from a combination of the previous lemma and .
[ th : nonconvessi ] under the same assumptions of theorem [ thmestimx0 ] , we have that @xmath186 in particular , the following estimate holds true : @xmath187 applying lemma [ lmestimk * ] with @xmath188 and corollary [ cor : estimxinfty ] gives : @xmath189 thus , easily follows by observing that @xmath190 . finally , using the isodiametric inequality @xmath191^n\!\!,\ ] ] in conjunction with ,
we show the validity of . estimates and involve the first eigenvalue @xmath114 , which in general is not easy to compute explicitly ; when @xmath24 is convex , we can estimate @xmath192 from above by means of basic geometric quantities , thus providing an easily computable lower bound on @xmath193 .
this is the content of the following theorem , which represents the main contribution of this section .
[ th : conv ] if @xmath24 is convex , then @xmath194,\ ] ] where @xmath195 is the inradius of @xmath0 , @xmath196 denotes the first dirichlet eigenvalue of @xmath197 in the unit ball and @xmath198 is the isoperimetric ratio of @xmath0 . in particular
, the following bound from below on @xmath193 holds true @xmath199.\ ] ] the proof of readily follows by combining to the following upper bound on @xmath114 @xmath200 proved in ( * ? ? ?
* theorem 2 ) . using and the two inequalities @xmath201^{n-1},\ ] ]
we end up with . observe that using and the inequality @xmath202 which follows from the monotonicity and scaling properties of @xmath203 , we can infer @xmath204,\ ] ] thus providing a lower bound which is strictly greater than , as long as the ratio @xmath205 is strictly smaller than @xmath206 and @xmath207 .
inequality is in fact a corollary of a sharper inequality holding for starshaped sets .
we can then give a refinement of which holds in this larger class : to this end , we borrow some notations from @xcite .
a set @xmath0 is said to be _ strictly starshaped _ with respect to a point @xmath208 if it is starshaped with respect to @xmath101 and if its support function centered at @xmath101 , i.e. @xmath209 is uniformly positive , that is @xmath210 .
let @xmath24 be a strictly starshaped set with locally lipschitz boundary , as in @xcite we define @xmath211 where @xmath212 denotes surface measure on @xmath213 according to this notation , ( * ? ? ?
* theorem 3 ) states : @xmath214 arguing as in theorem [ th : conv ] and using in place of gives the following estimate .
[ th : ambiguo ] let @xmath24 be a strictly starhaped set with locally lipschitz boundary and denote by @xmath36 the closure of the convex hull of @xmath24 .
then @xmath215 we remark that is sharper and more general than , and it is at the same time more explicit than , in the sense that , differently from @xmath216 the number @xmath217 can be computed directly from the support function ( which exactly determines a convex set ) . [ santalo ] it is worth noticing that the santal point @xmath218 of @xmath36 always satisfies ( as well as for every @xmath14 ) , then it satisfies all the estimates we proved for @xmath26 in this section .
in particular , theorem [ th : conv ] ( or theorem [ th : ambiguo ] ) can be used as well to estimate the location of the santal point of a convex set .
in this section , for the reader s convenience , we recall some relevant facts about _ aleksandrov s symmetry principle , _ which has been extensively used in many situations and with various generalizations ( see @xcite for a good reference ) . for @xmath219 let @xmath220 @xmath221 and @xmath222 be the sets defined in and . also , define a linear transformation @xmath223 by the matrix : @xmath224 where @xmath225 is the kronecker symbol and the @xmath226 are the components of @xmath227 .
then the application @xmath228 defined by @xmath229 represents the reflection with respect to @xmath230 as already mentioned , if @xmath0 is a subset of @xmath231 we set @xmath232 [ th : nohotspot ] let @xmath0 be a bounded domain in @xmath233 with lipschitz continuous boundary @xmath20 and suppose the hyperplane @xmath52 defined by has non - empty intersection with @xmath116 assume that @xmath234 if @xmath0 is not symmetric with respect to @xmath52 , then @xmath52 does not contain any ( spatial ) critical point of the solution @xmath8 of . for @xmath235 and @xmath14
the function @xmath236 is well - defined and is such that @xmath237 hence @xmath238 in @xmath239 by the strong maximum principle for parabolic operators ( see @xcite ) .
since @xmath240 on @xmath241 we obtain that @xmath242 on it ( @xmath66 is in fact the interior normal unit vector ) , by hopf s boundary lemma for parabolic operators .
we conclude by noticing that @xmath243 on @xmath244 with the same arguments and a little more work , one can extend this result to more general situations , involving nonlinearities both for elliptic and parabolic operators . as an example , here we present the following result .
[ th : nocritical ] let @xmath0 and @xmath245 satisfy the same assumptions as those of proposition [ th : nohotspot ] ; in particular suppose that @xmath234 let @xmath246 be a solution of class @xmath247 of the system : @xmath248 where @xmath152 is a locally lipschitz continuous function .
if @xmath0 is not symmetric with respect to @xmath52 , then @xmath52 does not contain any critical point of @xmath57 the proof runs similarly to that of proposition [ th : nohotspot ] ; the relevant changes follow .
the function @xmath249 defined for @xmath250 satisfies the conditions : @xmath251 where the function @xmath252 defined by @xmath253 is bounded by the lipschitz constant of @xmath152 in the interval @xmath254.$ ] hence @xmath255 in @xmath256 by the arguments used in @xcite .
let @xmath257 then @xmath258 and the strong maximum principle can be applied to obtain that @xmath238 in @xmath259 the conclusion then follows as before by hopf s boundary lemma .
an immediate consequence of this theorem is the following result .
[ coro : nohotspot ] let @xmath0 and @xmath245 satisfy the same assumptions as those of proposition [ th : nohotspot ] .
let @xmath103 be the first ( positive ) eigenfunction of @xmath197 with homogeneous dirichlet boundary conditions .
if @xmath0 is not symmetric with respect to @xmath52 , then @xmath52 does not contain any critical point of @xmath260
in what follows , we shall assume that @xmath261 is a convex body , that is a compact convex set with non - empty interior .
occasionally , we will suppose that @xmath261 is of class @xmath262 i.e. a set whose boundary @xmath62 is an @xmath71-dimensional submanifold of @xmath51 of class @xmath263 we are interested in determining the function given by @xmath264 which will be called the _
maximal folding function _ of @xmath265 @xmath67 defines in turn a subset of @xmath36 the _ heart _ of @xmath36 as @xmath266 of course , @xmath58 is a closed convex subset of @xmath267 observe that @xmath67 can be bounded below and above by means of the support functions of @xmath58 and @xmath268 @xmath269 the following results motivate our interest on @xmath58 and @xmath270 [ prop : heart ] let @xmath36 be a convex body . * the hot spot @xmath7 of @xmath63 the point @xmath26 and any limit point of @xmath7
as @xmath271 always belong to @xmath272 moreover , @xmath7 and @xmath26 must fall in the interior of @xmath273 whenever this is non - empty . *
the center of mass of @xmath63 @xmath274 always belongs to the heart @xmath58 of @xmath267 * if @xmath36 is strictly convex , the incenter @xmath275 of @xmath36 belongs to @xmath276 * let @xmath277 if there exist @xmath278 @xmath279 independent directions @xmath280 such that @xmath281 @xmath282 then @xmath283 in particular , if @xmath284 then @xmath58 reduces to @xmath285 and the hot spot of @xmath36 is stationary .
* let @xmath286 then @xmath287 items ( i ) and ( iv ) follow by observing that , for @xmath288 the set @xmath289 is contained in @xmath36 and is symmetric with respect to @xmath230 hence , @xmath290\ dy\end{aligned}\ ] ] and the last term is non - negative , vanishing if and only if @xmath36 is @xmath66-symmetric . items ( ii ) and ( iii ) are easy consequences of proposition [ th : nohotspot ] and corollary [ coro : nohotspot ] .
for a fixed @xmath291 let us define @xmath292 which is non negative , thanks to ( ii ) . then @xmath293 and @xmath294 for every @xmath46 such that @xmath295 .
hence @xmath296 thus taking the maximum as @xmath297 varies on @xmath44 we obtain .
informations on convex heat conductors with a stationary hot spot can be found in @xcite .
formula deserves some comments : observe that for every fixed @xmath298 , the minimum problem inside the braces amounts to finding a direction @xmath66 close to @xmath297 , so to maximize @xmath299 , and such that at the same time we can fold @xmath36 as much as possible , so to minimize the difference @xmath300 .
we conclude this subsection by an example that shows how the simultaneous application of proposition [ prop : heart ] and the results of section 2 substantially benefits the problem of locating @xmath167 [ exa : joint ] let us consider a spherical cap @xmath301 with @xmath302 .
thanks to the simmetry of @xmath303 , it is easily seen that its heart is given by @xmath304 which is a vertical segment touching the boundary @xmath305 at the point @xmath306 . in particular , by this method we can not exclude that the hotspot @xmath7 ( or the point @xmath26 ) is on the boundary . however , we can now use the results of section [ sec : polar ] , to further sharpen this estimate on the location of @xmath26 : indeed , applying theorem [ th : conv ] , we get @xmath307,\ ] ] where we used that @xmath308 and @xmath309 . the following theorem whose proof can be found in ( * ?
* theorem 5.7 ) guarantees that , for a regular set ( not necessarily convex ) , the maximal folding function is never trivial .
[ th : fraenkel ] let @xmath0 be a bounded open ( not necessarily convex ) subset of @xmath2 with @xmath310 boundary @xmath20 , and denote by @xmath36 the convex hull of @xmath311 for every @xmath312 , there exists @xmath313 such that , for every @xmath314 in the interval @xmath315 we have : 1 .
@xmath316 ; 2 .
@xmath317 , for every @xmath318 .
unfortunately , the previous result is just qualitative and does not give any quantitative information about the maximal folding function .
moreover , notice that the @xmath310 assumption on @xmath20 can not be dropped , even in the case of a convex domain : think of the spherical cap in example [ exa : joint ] , for which we have @xmath319 .
.3 cm in order to compute @xmath320 we need some more definitions .
we set @xmath321 and for every @xmath322 we define the segment @xmath323 then , we denote by @xmath324 the projection operator on @xmath325 , that is the application defined by @xmath326 and , for @xmath327 in the set @xmath328 the _ shadow of @xmath36 in the direction @xmath66 _ we define : @xmath329 we say that a convex body @xmath36 is _ @xmath66-strictly convex _ if @xmath62 does not contain any segment parallel to @xmath330 if @xmath36 is @xmath66-strictly convex , then for every @xmath331 ( equivalently @xmath332 ) such that the normal @xmath333 to @xmath62 at @xmath13 is orthogonal to @xmath66 , the set @xmath334 degenerates to the singleton @xmath335 we point out that @xmath336 is a convex function on @xmath337 , while @xmath338 is concave ; moreover , if we set @xmath339 we have that @xmath340 and , as soon as @xmath36 is @xmath66-strictly convex , @xmath341 where obviously @xmath342 denotes the graph of the relevant functions .
[ th : brasco ] let @xmath261 be a convex body . for @xmath46
consider the function @xmath343 given by @xmath344 then @xmath345 observe that @xmath346 let @xmath347 since @xmath348 then , for every point @xmath349 with @xmath350 and @xmath351 we have that @xmath352 in particular , for @xmath353 we obtain that @xmath354 and hence @xmath355 thus , @xmath356 if @xmath357 maximizes @xmath358 by taking @xmath359 we see that @xmath360 for every @xmath350 and @xmath361 therefore , @xmath362 and hence @xmath363 if we now remember that , for a convex domain @xmath36 , the quantity @xmath364 is the _ width of @xmath36 in the direction @xmath66 _ , we immediately get a nice consequence of the previous theorem .
[ th : corewidth ] let @xmath261 be a convex body
. then we have the following estimate for the width of @xmath58 in the direction @xmath66 : @xmath365 we first observe that @xmath366 , so that @xmath367 and yields @xmath368 then , from the definition of width , using ( [ supp_core ] ) , ( [ lambda ] ) and ( [ -lambda ] ) , we get @xmath369 thus concluding the proof . in general , inequality is strict .
for example , in @xmath370 consider the ellipse given by @xmath371 with @xmath372 .
the function @xmath67 can be easily computed in this case : for every @xmath373 we get @xmath374 the set @xmath375 is the image of a _ quadrifolium _ ( a _ rhodonea _ with @xmath376 petals ) by the mapping @xmath377 .
thus , for example , by choosing the direction @xmath378 the right - hand side of equals @xmath379 while clearly the left - hand side is zero since @xmath380 , due to the simmetries of @xmath36 .
this example also highlights the interest of the quantity @xmath381 , which can be seen as a measure of the lack of symmetry of @xmath36 in the direction of @xmath66 .
the function @xmath382 in can be explicitly computed by the use of the fourier transform : this is the content of the next result .
[ th : fourier ] let @xmath36 be a convex body and for @xmath46 , let @xmath382 be the function defined in .
@xmath383 where @xmath384 denotes the fourier transform of the characteristic function of @xmath36 and @xmath385 differentiation in the direction @xmath330 for @xmath386 and @xmath387 we write @xmath388 and @xmath389 with @xmath390 , @xmath391 and @xmath392 .
by fubini s theorem we compute @xmath393 for @xmath394 we then obtain : @xmath395\ , e^{-iy\cdot\eta}dy.\end{aligned}\ ] ] therefore , by the inversion formula for the fourier transform , we have : @xmath396 by , we also obtain that @xmath397 @xmath398 and hence @xmath399 formula follows from and at once .
if @xmath36 is a polygon , @xmath384 can be explicitly computed in terms of the vertices of @xmath36 .
let @xmath400 be a ( convex ) polygon with vertices @xmath401 we assume that @xmath402 are ordered counterclockwise and we set @xmath403 rewriting @xmath384 as a boundary integral ( see @xcite ) by means of the divergence theorem , we have that @xmath404 where @xmath405 is the exterior normal to the @xmath406-th side of @xmath267 also , @xmath407 is easily computed from the previous expression : @xmath408 ^ 2}\\ & + { \displaystyle}\frac{i}{|\eta|^2}\,{\sum\limits}_{j=1}^n |p_{j+1}-p_j|(\nu_j\cdot\eta)\ , \frac{(p_{j+1}\cdot{\omega})\,e^{-ip_{j+1}\cdot\eta}-(p_j\cdot{\omega})\,e^{-ip_j\cdot\eta}}{(p_{j+1}-p_j)\cdot\eta}. \end{split}\ ] ] we conclude this section by presenting some necessary conditions for the optimality of @xmath409 to this aim , we first state and prove an easy technical result for the subdifferential of a function .
[ th : tech ] let @xmath410 be a convex open set .
let @xmath411 and @xmath412 be a convex and , respectively , a concave function from @xmath0 to @xmath413 . if @xmath414 attains its maximum at a point @xmath415 , then
@xmath416 it is clear that both @xmath417 and @xmath418 are non - empty .
since @xmath419 is a maximum point , we get @xmath420 and hence @xmath421 for every @xmath422 and @xmath423 if @xmath424 then we have @xmath425 as a consequence of the definitions of @xmath336 and @xmath338 , we have that @xmath426 implies that @xmath419 belongs to the boundary of @xmath427 we are now in a position to state a necessary optimality condition .
[ th : necessary ] let @xmath261 be a convex body and @xmath312 .
suppose that @xmath382 attains its maximum at a point @xmath428 that is @xmath429 set @xmath430 and for every @xmath431 denote by @xmath432 its reflection with respect to the optimal hyperplane , that is @xmath433 then : 1 .
if @xmath434 , we have @xmath435 where @xmath436 denotes the _ normal cone _ of @xmath36 at a point @xmath437 ; 2 . if @xmath438 , there holds @xmath439 where @xmath440 we can suppose for simplicity that @xmath441 we first suppose that @xmath419 is an interior point .
since by its very definition @xmath382 is the sum of a convex function and a concave one , by lemma [ th : tech ] @xmath442 let us now set @xmath443 the reflection of @xmath101 in the optimal hyperplane is @xmath444 and @xmath445 for @xmath446 then , since @xmath447 , implies .
if @xmath419 is on the boundary of @xmath448 then @xmath449 or @xmath450 may be empty : this is clearly the case if @xmath451 or @xmath452 have some vertical parts .
observe that actually we have the following possibilities : 1 .
@xmath453 ; 2 .
@xmath454 .
if ( 1 ) holds , then at every @xmath455 , the convex body @xmath36 has only supporting hyperplanes parallel to @xmath456 : these are invariant with respect to the action of @xmath457 , so that their reflections are supporting hyperplanes for @xmath36 at @xmath432 and formula or easily follows . if ( 2 ) holds , we have @xmath458 and let us call @xmath459 .
then @xmath460 and we have @xmath461 , so that @xmath462 by observing that @xmath463 , follows . under the same notations of theorem [ th : necessary ] , if @xmath62 admits a ( unique ) unit normal @xmath464 at the point @xmath101 , then it admits a unit normal at the point @xmath432 too and @xmath465 in particular , if @xmath458 we have @xmath466 .
it is sufficient to observe that in this case @xmath467 and hence is a consequence of or . using theorem [ th : necessary ]
, we obtain an interesting upper bound on the maximal folding function for a strictly convex domain , in terms of its support function .
if @xmath36 is strictly convex , then for every @xmath468 @xmath469 first observe that thanks to the @xmath470homogeneity of the support function , the maximization problem in can be equivalently settled in @xmath44 .
the strict convexity of @xmath36 implies that @xmath471 ( see @xcite ) .
moreover , @xmath472 for every @xmath473 , with @xmath437 .
thus , with the same notations as in theorem [ th : necessary ] , @xmath474 for every @xmath475 and using the condition , we also get @xmath476 ; on the other hand , @xmath477 and @xmath478 , which implies the following @xmath479 hence we can conclude by simply applying theorem [ th : brasco ] .
if @xmath36 is symmetric with respect to a hyperplane orthogonal to @xmath227 , then equality holds in and both quantities equal @xmath480 .
otherwise , in general inequality is strict as figure @xmath85 informs us .
@xmath481 @xmath482 the intersection of the two straight lines corresponds to @xmath483 the dark dot corresponds to @xmath484 notice that , following an argument similar to that of the proof of proposition [ paolo ] , we can in fact give a precise characterization of the maximal folding function @xmath67 in terms of the support function @xmath83 .
precisely the following holds @xmath485 where @xmath486 if @xmath36 is not strictly convex ( and then @xmath83 is not @xmath310 ) the above formula still remains valid , up to suitably interpreting the gradient of @xmath83 as the subdifferential @xmath487 .
if @xmath36 is a convex polyhedron , then the conclusions of theorem [ th : brasco ] can be improved : roughly speaking , we can discretize the optimization problem ( [ lambda ] ) , by only visiting the projections of the vertices of @xmath36 on @xmath488 we begin with the following general result . in other words ,
if @xmath489 touches " @xmath490 from the interior at @xmath13 and @xmath13 is contained in the interior of some segment on the boundary of @xmath489 , then the boundary of @xmath490 must contain all the segment at the same . indeed , let @xmath495 be a support hyperplane to @xmath490 at @xmath13 and denote by @xmath496 the half - space delimited by @xmath495 and containing @xmath490 ; then @xmath495 is also a support hyperplane to @xmath489 at @xmath13 and @xmath497 .
thus , @xmath498 while @xmath499 this implies @xmath500 since @xmath13 is not an endpoint of @xmath278 , and hence @xmath501 . now , let @xmath36 be a convex polyhedron .
if @xmath88 is not a vertex of @xmath63 then @xmath88 belongs to the relative interior of some @xmath519-dimensional facet of @xmath520 with @xmath521 and hence it belongs to the relative interior of a segment @xmath278 with ( at least ) one end at some vertex @xmath522 of @xmath64 and @xmath58 in two examples .
in the first picture observe that , by means of , we also know that @xmath26 is at a positive ( and computable ) distance from the boundary of @xmath267,title="fig : " ] and @xmath58 in two examples . in the first picture observe that , by means of , we also know that @xmath26 is at a positive ( and computable ) distance from the boundary of @xmath267,title="fig : " ] in case ( ii ) , the @xmath519-dimensional facet @xmath524 of @xmath62 containing @xmath518 must be orthogonal to the hyperplane @xmath525 however , the same argument used for case ( i ) can easily be worked out in @xmath526 the first author was partially supported by the european research council under fp7 , advanced grant n. 226234 `` analytic techniques for geometric and functional inequalities '' , while the second and third authors were partially supported by the prin - miur grant `` aspetti geometrici delle equazioni alle derivate parziali e questioni connesse '' .
h. l. brascamp , e. h. lieb , _ on extensions of brunn - minkowski and pr`ekopa - leindler theorems , including inequalities for log - concave functions , and with an application to the diffusion equation _ , j. funct .
( 1976 ) , 366389 .
m. chamberland , d. siegel , _ convex domains with stationary hot spots _
20 ( 1997 ) 11631169 .
l. e. fraenkel , introduction to maximum principles and symmetry in elliptic problems , cambridge university press 2000 .
p. freitas , d. krejcirk , _ a sharp upper bound for the first dirichlet eigenvalue and the growth of the isoperimetric constant of convex domains _ , proc .
american math .
( 2008 ) , 29973006 .
b. gidas , w. m. ni , l. nirenberg , _ symmetry and related properties via the maximum principle _ ,
68 ( 1979 ) , no .
3 , 209243 .
d. grieser , d. jerison , _ the size of the first eigenfunction on a convex planar domain _ , j. amer .
soc . 11 ( 1998 ) , 4172 .
r. gulliver , n.b .
willms , _ a conjectured heat flow problem , _ in solutions , siam review 37 , ( 1995 ) 100104 . c. e. gutierrez , the monge - ampre equation , progress in nonlinear differential equations and their applications , 44 .
birkhuser boston , inc . ,
boston , ma , 2001 . c. s. herz , _ fourier transforms related to convex sets , _ ann .
75 ( 1962 ) 8192 .
b. kawohl , _ a conjectured heat flow problem , _ in solutions , siam review 37 ( 1995 ) 104105 .
m. s. klamkin , _ a conjectured heat flow problem , _ in problems , siam review 36 ( 1994 ) 107
. n. korevaar , _ convex solutions to nonlinear elliptic and parabolic boundary value problems , _ indiana univ .
j. 32 ( 1983 ) , 603614 .
r. magnanini , s. sakaguchi , _ on heat conductors with a stationary hot spot _ ,
ann . mat .
pura appl . 183
( 2004 ) , 123 .
r. magnanini , s. sakaguchi , _ polygonal heat conductors with a stationary hot spot _
, j. anal . math . 105
( 2008 ) , 118 .
m. h. protter , h. f. weinberger , maximum principles in differential equations , prentice - hall , englewood cliffs , n. j. , 1967 .
s. sakaguchi , _ behavior of spatial critical points and zeros of solutions of diffusion equations , _ selected papers on differential equations and analysis , 1531 , amer . math .
ser . 2 , 215 , amer .
soc . , providence , ri , 2005 .
j. serrin , _ a symmetry problem in potential theory , _ arch .
rational mech .
43 ( 1971 ) , 304318 .
r. schneider , convex bodies : the brunn - minkowski theory , cambridge university press 1993 .
g. talenti , _ some estimates of solutions to monge - ampre type equations in dimension two _ , ann . scuola norm .
( 1981 ) , 183230 . s. r. s. varadhan , _ on the behaviour of the fundamental solution of the heat equation with variable coefficients , _ comm .
pure appl .
20 , ( 1967 ) , 431455 . | we investigate the location of the ( unique ) hot spot in a convex heat conductor with unitary initial temperature and with boundary grounded at zero temperature .
we present two methods to locate the hot spot : the former is based on ideas related to the alexandrov - bakelmann - pucci maximum principle and monge - amp ` ere equations ; the latter relies on alexandrov s reflection principle .
we then show how such a problem can be simplified in case the conductor is a polyhedron .
finally , we present some numerical computations . |
in this section , we introduce the related works , motivation and the problems that we are facing . the concept of fuzzy sets was introduced by zadeh in 1965 @xcite . since
then fuzzy sets and fuzzy logic have been applied in many real applications to handle uncertainty .
the traditional fuzzy set uses one real number @xmath0 $ ] to represent the grade of membership of fuzzy set @xmath1 defined on universe @xmath2 .
sometimes @xmath3 itself is uncertain and hard to be defined by a crisp value .
so the concept of interval valued fuzzy sets was proposed @xcite to capture the uncertainty of grade of membership .
interval valued fuzzy set uses an interval value @xmath4 $ ] with @xmath5 to represent the grade of membership of fuzzy set @xmath1 . in some applications such as expert system , belief system and information fusion
, we should consider not only the truth - membership supported by the evidence but also the false - membership againsted by the evidence . that is beyond the scope of fuzzy sets and interval valued fuzzy sets . in 1986 , atanassov introduced the intuitionistic fuzzy sets @xcite which is a generalization of fuzzy sets and provably equivalent to interval valued fuzzy sets .
the intuitionistic fuzzy sets consider both truth - membership and false - membership .
later on , intuitionistic fuzzy sets were extended to the interval valued intuitionistic fuzzy sets @xcite . the interval valued intuitionistic fuzzy set uses a pair of intervals @xmath6 , \mbox { } 0 \leq t^- \leq t^+ \leq 1 $ ] and @xmath7 , \mbox { } 0 \leq f^- \leq f^+ \leq 1 $ ] with @xmath8 to describe the degree of true belief and false belief .
because of the restriction that @xmath8 , intuitionistic fuzzy sets and interval valued intuitionistic fuzzy sets can only handle incomplete information not the indeterminate information and inconsistent information which exists commonly in belief systems .
for example , when we ask the opinion of an expert about certain statement , he or she may say that the possibility that the statement is true is between @xmath9 and @xmath10 and the statement is false is between @xmath11 and @xmath12 and the degree that he or she is not sure is between @xmath13 and @xmath14 . here is another example ,
suppose there are 10 votees during a voting process . in time @xmath15 , three vote yes " , two vote no " and five are undecided , using neutrosophic notation , it can be expressed as @xmath16 . in time @xmath17 , three vote yes " , two vote no " , two give up and three are undecided , it then can be expressed as @xmath18 .
that is beyond the scope of the intuitionistic fuzzy set .
so , the notion of neutrosophic set is more general and overcomes the aforementioned issues . in neutrosophic set , indeterminacy
is quantified explicitly and + truth - membership , indeterminacy - membership and false - membership are independent .
this assumption is very important in information fusion when we try to combine the data from different sensors .
neutrosophy was introduced by florentin smarandache in 1980 .
it is a branch of philosophy which studies the origin , nature and scope of neutralities , as well as their interactions with different ideational spectra " @xcite .
neutrosophic set is a powerful general formal framework which generalizes the concept of the classic set , fuzzy set @xcite , interval valued fuzzy set @xcite , intuitionistic fuzzy set @xcite , interval valued intuitionistic fuzzy set @xcite , paraconsistent set @xcite , dialetheist set @xcite , paradoxist set @xcite , tautological set @xcite .
a neutrosophic set @xmath1 defined on universe @xmath19 .
@xmath20 with @xmath21 and @xmath22 being the real standard or non - standard subsets of @xmath230 ^ - , 1^+[$ ] .
@xmath24 is the degree of true membership function in the set @xmath1 , @xmath25 is the degree of indeterminate membership function in the set @xmath1 and @xmath22 is the degree of false membership function in the set @xmath1 .
the neutrosophic set generalizes the above mentioned sets from philosophical point of view . from scientific or engineering point of view , the neutrosophic set and set - theoretic operators need to be specified . otherwise , it will be difficult to apply in the real applications . in this paper
, we define the set - theoretic operators on an instance of neutrosophic set called interval neutrosophic set ( ins ) .
we call it as interval " because it is subclass of neutrosophic set , that is we only consider the subunitary interval of @xmath26 $ ] .
an interval neutrosophic set @xmath1 defined on universe @xmath2 , @xmath27 with @xmath24 , @xmath25 and @xmath22 being the subinterval of @xmath26 $ ] .
interval neutrosophic set can represent uncertainty , imprecise , incomplete and inconsistent information which exist in real world .
the interval neutrosophic set generalizes the following sets : 1 .
the _ classical set _ , @xmath28 , @xmath29 or @xmath30 , @xmath31 or @xmath30 and @xmath32 .
the _ fuzzy set _ , @xmath28 , @xmath33 $ ] , @xmath34 $ ] and @xmath32 .
3 . the _ interval valued fuzzy set _
, @xmath28 , @xmath35 $ ] , @xmath36 and @xmath37 .
the _ intuitionistic fuzzy set _ , @xmath28 , @xmath33 $ ] , @xmath38 $ ] and @xmath39 . 5 .
the _ interval valued intuitionistic fuzzy set _ , @xmath28 , @xmath40 $ ] and @xmath39 .
the _ paraconsistent set _
, @xmath28 , @xmath33 $ ] , @xmath38 $ ] and @xmath41 . 7 .
the _ interval valued paraconsistent set _
, @xmath28 , @xmath42 $ ] and @xmath43 .
the relationship among interval neutrosophic set and other sets is illustrated in fig [ fig0 ] . #
1#2#3#4#5 @font ( 4821,4434)(0,-10 ) ( 2475,3825)(2475,4200 ) ( 2505.000,4080.000)(2475.000,4200.000)(2445.000,4080.000)(2505.000,4080.000 ) ( 1050,2925)(2400,3525 ) ( 2302.527,3448.849)(2400.000,3525.000)(2278.158,3503.678)(2302.527,3448.849 ) ( 3975,2925)(2550,3525 ) ( 2672.238,3506.082)(2550.000,3525.000)(2648.955,3450.784)(2672.238,3506.082 ) ( 825,1950)(825,2475 ) ( 855.000,2355.000)(825.000,2475.000)(795.000,2355.000)(855.000,2355.000 ) ( 3975,1950)(3975,2475 ) ( 4005.000,2355.000)(3975.000,2475.000)(3945.000,2355.000)(4005.000,2355.000 ) ( 825,900)(825,1350 ) ( 855.000,1230.000)(825.000,1350.000)(795.000,1230.000)(855.000,1230.000 ) ( 825,225)(825,600 ) ( 855.000,480.000)(825.000,600.000)(795.000,480.000)(855.000,480.000 ) ( 1875,4275)(0,0)[lb ] ( 1500,3600)(0,0)[lb ] ( 675,2550)(0,0)[lb ] ( 0,2775)(0,0)[lb ] ( 3525,2775)(0,0)[lb ] ( 3375,2550)(0,0)[lb ] ( 3375,1800)(0,0)[lb ] ( 0,1425)(0,0)[lb ] ( 450,675)(0,0)[lb ] ( 375,0)(0,0)[lb ] ( 0,1725)(0,0)[lb ] note that @xmath44 in fig .
[ fig0 ] such as @xmath45 means that b is a generalization of a. we define the set - theoretic operators on interval neutrosophic set ( ins ) .
various properties of ins are proved , which are connected to the operations and relations over ins .
the rest of paper is organized as follows .
section [ section2 ] gives a brief overview of neutrosophic set .
section [ section3 ] gives the definition of interval neutrosophic set and set - theoretic operations .
section [ section4 ] gives some properties of set - theoretic operations .
section [ section5 ] gives the definition of convexity of interval neutrosophic sets and prove some properties of convexity .
section [ section6 ] concludes the paper . to maintain a smooth flow throughout the paper
, we present the proofs to all theorems in appendix .
this section gives a brief overview of concepts of neutrosophic set defined in @xcite .
here , we use different notations to express the same meaning .
let @xmath46 and @xmath47 be two real standard or non - standard subsets , then @xmath48 , @xmath49 .
@xmath50 , @xmath51 .
@xmath52 .
let @xmath2 be a space of points ( objects ) , with a generic element in @xmath2 denoted by @xmath53 .
+ a neutrosophic set @xmath1 in @xmath2 is characterized by a _ truth - membership function _
@xmath54 , a _ indeterminacy - membership function _ @xmath55 and a _ false - membership _ function @xmath56 . @xmath57 and @xmath58 are real standard or non - standard subsets of @xmath230 ^ - , 1^+[$ ] . that is @xmath590 ^ - , 1^+ [ , \\ i_a : x & \rightarrow & ] 0 ^ - , 1^+ [ , \\
f_a : x & \rightarrow & ] 0 ^ - , 1^+[.\end{aligned}\ ] ] there is no restriction on the sum of @xmath60 , @xmath61 and @xmath58 , so @xmath62 .
the _ complement _ of a neutrosophic set @xmath1 is denoted by @xmath63 and is defined by @xmath64 for all @xmath53 in @xmath2 .
a neutrosophic set @xmath1 is _ contained _ in the other neutrosophic set @xmath65 , @xmath66 , if and only if @xmath67 for all @xmath53 in @xmath2 .
the _ union _ of two neutrosophic sets @xmath1 and @xmath65 is a neutrosophic set @xmath68 , written as @xmath69 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath70 for all @xmath53 in @xmath2 .
the _ intersection _ of two neutrosophic sets @xmath1 and @xmath65 is a neutrosophic set @xmath68 , written as @xmath71 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath72 for all @xmath53 in @xmath2 .
the _ difference _ of two neutrosophic sets @xmath1 and @xmath65 is a neutrosophic set @xmath68 , writeen as @xmath73 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath74 for all @xmath53 in @xmath2 .
let @xmath1 be the neutrosophic set defined on universe @xmath75 and @xmath65 be the neutrosophic set defined on universe @xmath76 . if @xmath77 and @xmath78 , then the _ cartesian product _ of two neutrosophic sets @xmath1 and @xmath65 is defined by @xmath79
in this section , we present the notion of _ interval neutrosophic set _ ( ins ) .
interval neutrosophic set ( ins ) is an instance of neutrosophic set which can be used in real scientific and engineering applications .
let @xmath2 be a space of points ( objects ) , with a generic element in @xmath2 denoted by @xmath53 .
+ an interval neutrosophic set ( ins ) @xmath1 in @xmath2 is characterized by truth - membership function @xmath54 , indeterminacy - membership function @xmath55 and false - membership function @xmath56 . for each point @xmath53 in @xmath2 , @xmath80 $ ] .
an interval neutrosophic set ( ins ) in @xmath81 is illustrated in fig .
[ fig1 ] . # 1#2#3#4#5 @font ( 5538,3609)(0,-10 ) ( 525,225)(525,3450 ) ( 555.000,3330.000)(525.000,3450.000)(495.000,3330.000)(555.000,3330.000 ) ( 525,225)(5400,225 ) ( 5280.000,195.000)(5400.000,225.000)(5280.000,255.000)(5280.000,195.000 ) ( 537,852)(538,853)(539,855 ) ( 542,858)(547,864)(554,872 ) ( 564,883)(576,896)(590,912 ) ( 607,931)(627,952)(649,976 ) ( 674,1001)(701,1028)(730,1057 ) ( 760,1086)(793,1117)(828,1147 ) ( 864,1178)(901,1209)(940,1239 ) ( 981,1269)(1023,1297)(1068,1325 ) ( 1114,1352)(1162,1378)(1213,1402 ) ( 1267,1425)(1323,1446)(1383,1466 ) ( 1447,1483)(1513,1498)(1584,1510 ) ( 1657,1519)(1734,1525)(1812,1527 ) ( 1883,1525)(1955,1521)(2025,1513 ) ( 2094,1503)(2161,1491)(2226,1476 ) ( 2289,1460)(2350,1442)(2409,1423 ) ( 2466,1402)(2522,1380)(2576,1357 ) ( 2629,1333)(2680,1309)(2730,1283 ) ( 2780,1257)(2828,1230)(2875,1203 ) ( 2920,1175)(2965,1147)(3008,1120 ) ( 3050,1092)(3090,1066)(3129,1039 ) ( 3165,1014)(3199,991)(3230,968 ) ( 3259,948)(3285,929)(3307,912 ) ( 3327,898)(3343,885)(3357,875 ) ( 3367,867)(3375,861)(3381,857 ) ( 3384,854)(3386,853)(3387,852 ) ( 525,1575)(526,1576)(527,1577 ) ( 530,1580)(535,1584)(541,1590 ) ( 550,1598)(561,1608)(575,1620 ) ( 591,1634)(610,1650)(631,1667 ) ( 654,1686)(680,1706)(707,1727 ) ( 737,1749)(768,1771)(802,1793 ) ( 836,1815)(873,1837)(911,1859 ) ( 951,1880)(993,1900)(1037,1919 ) ( 1083,1937)(1131,1955)(1182,1970 ) ( 1236,1985)(1294,1998)(1355,2009 ) ( 1420,2018)(1488,2025)(1561,2029 ) ( 1638,2031)(1718,2030)(1800,2025 ) ( 1872,2018)(1944,2010)(2015,1999 ) ( 2085,1986)(2154,1972)(2221,1956 ) ( 2286,1939)(2350,1921)(2411,1902 ) ( 2472,1882)(2530,1861)(2587,1840 ) ( 2643,1817)(2698,1795)(2751,1771 ) ( 2804,1747)(2856,1723)(2906,1699 ) ( 2956,1674)(3004,1649)(3052,1624 ) ( 3098,1600)(3142,1575)(3185,1551 ) ( 3227,1528)(3266,1505)(3303,1484 ) ( 3337,1464)(3369,1445)(3398,1428 ) ( 3423,1412)(3446,1399)(3466,1387 ) ( 3482,1377)(3495,1369)(3506,1362 ) ( 3514,1357)(3519,1354)(3522,1352 ) ( 3524,1351)(3525,1350 ) ( 525,2250)(526,2249)(527,2248 ) ( 530,2245)(535,2241)(541,2235 ) ( 550,2227)(561,2217)(575,2205 ) ( 591,2191)(610,2175)(632,2158 ) ( 655,2139)(681,2119)(710,2098 ) ( 740,2076)(772,2054)(806,2032 ) ( 842,2010)(880,1988)(920,1966 ) ( 961,1945)(1005,1925)(1051,1906 ) ( 1099,1888)(1151,1870)(1205,1855 ) ( 1263,1840)(1324,1827)(1390,1816 ) ( 1460,1807)(1535,1800)(1614,1796 ) ( 1697,1794)(1784,1795)(1875,1800 ) ( 1950,1806)(2026,1815)(2102,1825 ) ( 2176,1837)(2250,1850)(2321,1865 ) ( 2391,1881)(2460,1897)(2527,1915 ) ( 2592,1934)(2656,1953)(2718,1974 ) ( 2779,1994)(2839,2016)(2898,2038 ) ( 2956,2060)(3012,2083)(3068,2106 ) ( 3123,2130)(3177,2153)(3230,2177 ) ( 3281,2201)(3332,2224)(3381,2248 ) ( 3428,2271)(3473,2293)(3517,2315 ) ( 3558,2335)(3596,2355)(3632,2373 ) ( 3665,2390)(3695,2406)(3721,2420 ) ( 3745,2432)(3765,2443)(3781,2451 ) ( 3795,2459)(3806,2464)(3813,2469 ) ( 3819,2472)(3822,2474)(3824,2475)(3825,2475 ) ( 525,2700)(526,2700)(528,2699 ) ( 532,2697)(538,2695)(546,2692 ) ( 558,2687)(572,2681)(590,2674 ) ( 611,2666)(636,2657)(665,2646 ) ( 696,2634)(731,2621)(769,2607 ) ( 809,2592)(852,2577)(897,2560 ) ( 944,2544)(993,2527)(1043,2510 ) ( 1095,2493)(1148,2476)(1202,2459 ) ( 1257,2442)(1313,2426)(1370,2410 ) ( 1429,2394)(1488,2379)(1549,2364 ) ( 1611,2350)(1675,2336)(1740,2323 ) ( 1807,2311)(1876,2299)(1946,2289 ) ( 2019,2279)(2093,2270)(2169,2263 ) ( 2245,2257)(2323,2253)(2400,2250 ) ( 2489,2249)(2574,2251)(2656,2254 ) ( 2734,2260)(2808,2267)(2877,2276 ) ( 2942,2286)(3003,2297)(3060,2310 ) ( 3115,2323)(3166,2337)(3215,2352 ) ( 3261,2368)(3305,2385)(3347,2401 ) ( 3387,2419)(3425,2436)(3462,2454 ) ( 3496,2472)(3529,2489)(3560,2506 ) ( 3589,2523)(3616,2538)(3640,2553 ) ( 3662,2567)(3682,2579)(3698,2590 ) ( 3712,2599)(3724,2607)(3733,2613 ) ( 3740,2618)(3745,2621)(3748,2623 ) ( 3749,2624)(3750,2625 ) ( 525,1200)(525,1199)(525,1197 ) ( 525,1194)(525,1190)(524,1185 ) ( 524,1179)(525,1171)(525,1162 ) ( 526,1151)(527,1140)(529,1127 ) ( 531,1113)(534,1099)(539,1084 ) ( 544,1068)(550,1052)(558,1036 ) ( 567,1020)(578,1004)(591,988 ) ( 605,972)(623,957)(642,942 ) ( 665,928)(690,914)(719,901 ) ( 752,888)(789,877)(831,866 ) ( 877,856)(929,847)(987,840 ) ( 1050,834)(1119,829)(1195,826 ) ( 1275,825)(1342,826)(1411,827 ) ( 1481,830)(1553,834)(1624,839 ) ( 1696,845)(1768,852)(1839,859 ) ( 1910,867)(1980,876)(2050,885 ) ( 2120,894)(2189,904)(2258,915 ) ( 2326,926)(2394,937)(2461,949 ) ( 2528,961)(2595,973)(2661,985 ) ( 2727,998)(2792,1010)(2857,1023 ) ( 2920,1036)(2983,1049)(3044,1061 ) ( 3104,1074)(3162,1086)(3217,1098 ) ( 3271,1110)(3322,1121)(3370,1131 ) ( 3415,1141)(3456,1150)(3494,1159 ) ( 3528,1167)(3558,1173)(3585,1179 ) ( 3607,1185)(3626,1189)(3641,1192 ) ( 3653,1195)(3662,1197)(3668,1198 ) ( 3672,1199)(3674,1200)(3675,1200 ) ( 525,750)(526,748)(527,746 ) ( 529,743)(532,739)(535,733 ) ( 540,726)(547,717)(554,707 ) ( 563,696)(573,684)(585,670 ) ( 599,656)(614,642)(631,627 ) ( 649,612)(670,597)(693,582 ) ( 718,567)(745,553)(776,539 ) ( 809,526)(846,513)(887,501 ) ( 932,491)(982,481)(1038,472 ) ( 1099,464)(1166,458)(1240,453 ) ( 1321,450)(1408,449)(1500,450 ) ( 1571,452)(1644,456)(1717,460 ) ( 1791,466)(1865,472)(1938,479 ) ( 2011,487)(2082,495)(2153,504 ) ( 2223,514)(2292,524)(2361,534 ) ( 2428,545)(2495,556)(2562,568 ) ( 2627,580)(2692,592)(2757,604 ) ( 2820,617)(2883,629)(2946,642 ) ( 3007,655)(3068,668)(3127,681 ) ( 3185,694)(3242,706)(3296,718 ) ( 3349,730)(3398,742)(3446,753 ) ( 3490,763)(3531,773)(3569,781 ) ( 3603,789)(3633,797)(3659,803 ) ( 3682,808)(3701,813)(3716,817 ) ( 3728,820)(3737,822)(3743,823 ) ( 3747,824)(3749,825)(3750,825 ) ( 375,75)(0,0)[lb ] ( 5400,0)(0,0)[lb ] ( 0,3225)(0,0)[lb ] ( 0,3450)(0,0)[lb ] ( 0,3000)(0,0)[lb ] ( 375,2775)(0,0)[lb ] ( 1425,1575)(0,0)[lb ] ( 1425,2100)(0,0)[lb ] ( 2925,2625)(0,0)[lb ] ( 3450,2100)(0,0)[lb ] ( 1425,900)(0,0)[lb ] ( 1425,525)(0,0)[lb ] when @xmath2 is continuous , an ins @xmath1 can be written as @xmath82 when @xmath2 is discrete , an ins @xmath1 can be written as @xmath83 consider parameters such as capability , trusthworthiness and price of semantic web services .
these parameters are commonly used to define quality of service of semantic web services . in this section
, we will use the evaluation of quality of service of semantic web services @xcite as running example to illustrate every set - theoretic operation on interval neutrosophic set .
[ example1 ] assume that @xmath84 $ ] .
@xmath85 is capability , @xmath86 is trustworthiness and @xmath87 is price .
the values of @xmath88 and @xmath87 are in @xmath89 $ ] .
they are obtained from the questionaire of some domain experts , their option could be degree of good , degree of indeterminacy and degree of poor .
@xmath1 is an interval neutrosophic set of @xmath2 defined by + @xmath90,[0.3,0.5],[0.3,0.5 ] \rangle / x_1 + \langle [ 0.5,0.7],[0,0.2],[0.2,0.3 ] \rangle / x_2 + \\
\langle [ 0.6,0.8],[0.2,0.3],[0.2,0.3 ] \rangle / x_3 $ ] .
+ @xmath65 is an interval neutrosophic set of @xmath2 defined by + @xmath91,[0.1,0.3],[0.1,0.3 ] \rangle / x_1 + \langle [ 0.2,0.3],[0.2,0.4],[0.5,0.8 ] \rangle / x_2 + \\ \langle [ 0.4,0.6],[0,0.1],[0.3,0.4 ] \rangle / x_3 $ ] . + an interval neutrosophic set @xmath1 is _ empty _ if and only if its @xmath92 , @xmath93 and @xmath94 , for all @xmath53 in @xmath2 .
we now present the set - theoretic operators on interval neutrosophic set .
the _ complement _ of an interval set @xmath1 is denoted by @xmath63 and is defined by @xmath95 for all @xmath53 in @xmath2 .
[ example2 ] let @xmath1 be the interval neutrosophic set defined in example [ example1 ] .
then , @xmath96,[0.5,0.7],[0.2,0.4 ] \rangle / x_1 + \langle [ 0.2,0.3],[0.8,1.0 ] , [ 0.5,0.7 ] \rangle / x_2 + \\ \langle [ 0.2,0.3],[0.7,0.8],[0.6,0.8 ] \rangle / x_3 $ ] .
+ an interval neutrosophic set @xmath1 is _ contained _ in the other interval neutrosophic set @xmath65 , @xmath66 , if and only if @xmath97 for all @xmath53 in @xmath2 .
note that by the definition of containment , @xmath2 is partial order not linear order . for example , let @xmath1 and @xmath65 be the interval neutrosophic sets defined in example [ example1 ] .
then , @xmath1 is not contained in @xmath65 and @xmath65 is not contained in @xmath1 .
two interval neutrosophic sets @xmath1 and @xmath65 are _ equal _ , written as @xmath98 , if and only if @xmath66 and @xmath99 the _ union _ of two interval neutrosophic sets @xmath1 and @xmath65 is an interval neutrosophic set @xmath68 , written as @xmath69 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath100 for all @xmath53 in @xmath2 .
[ example3 ] let @xmath1 and @xmath65 be the interval neutrosophic sets defined in example [ example1 ] .
then , @xmath101,[0.1,0.3],[0.1,0.3 ] \rangle / x_1 + \\
\langle [ 0.5,0.7 ] , [ 0,0.2 ] , [ 0.2,0.3 ] \rangle / x_2 + \langle [ 0.6,0.8],[0,0.1],[0.2,0.3 ] \rangle / x_3 $ ] .
+ the intuition behind the union operator is that if one of elements in @xmath1 and @xmath65 is true then it is true in @xmath102 , only both are indeterminate and false in @xmath1 and @xmath65 then it is indeterminate and false in @xmath102 . the other operators should be understood similarly .
@xmath102 is the smallest interval neutrosophic set containing both @xmath1 and @xmath65 .
the _ intersection _ of two interval neutrosophic sets @xmath1 and @xmath65 is an interval neutrosophic set @xmath68 , written as @xmath71 , whose truth - membership , indeterminacy - membership functions and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath103 for all @xmath53 in @xmath2 .
[ example4 ] let @xmath1 and @xmath65 be the interval neutrosophic sets defined in example [ example1 ] .
then , @xmath104,[0.3,0.5],[0.3,0.5 ] \rangle / x_1 + \\
\langle [ 0.2,0.3 ] , [ 0.2,0.4 ] , [ 0.5,0.8 ] \rangle / x_2 + \langle [ 0.4,0.6],[0.2,0.3],[0.3,0.4 ] \rangle / x_3 $ ] .
+ @xmath105 is the largest interval neutrosophic set contained in both @xmath1 and @xmath65 .
the _ difference _ of two interval neutrosophic sets @xmath1 and @xmath65 is an interval neutrosophic set @xmath68 , writeen as @xmath73 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath106 for all @xmath53 in @xmath2 .
[ example5 ] let @xmath1 and @xmath65 be the interval neutrosophic sets defined in example [ example1 ] .
then , @xmath107,[0.7,0.9],[0.5,0.7 ] \rangle / x_1 + \\
\langle [ 0.5,0.7 ] , [ 0.6,0.8 ] , [ 0.2,0.3 ] \rangle / x_2 + \langle [ 0.3,0.4],[0.9,1.0],[0.4,0.6 ] \rangle / x_3 $ ] .
+ @xmath108 the _ addition _ of two interval neutrosophic sets @xmath1 and @xmath65 is an interval neutrosophic set @xmath68 , writeen as @xmath109 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath110 for all @xmath53 in @xmath2 .
[ example6 ] let @xmath1 and @xmath65 be the interval neutrosophic sets defined in example [ example1 ] .
then , @xmath111,[0.4,0.8],[0.4,0.8 ] \rangle / x_1 + \\ \langle [ 0.7,1.0 ] , [ 0.2,0.6 ] , [ 0.7,1.0 ] \rangle / x_2 + \langle [ 1.0,1.0],[0.2,0.4],[0.5,0.7 ] \rangle / x_3 $ ] .
+ the _ cartesian product _ of two interval neutrosophic sets @xmath1 defined on universe @xmath112 and @xmath65 defined on universe @xmath113 is an interval neutrosophic set @xmath68 , writeen as @xmath114 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 and @xmath65 by @xmath115 for all @xmath53 in @xmath112 , @xmath116 in @xmath113 .
[ example7 ] let @xmath1 and @xmath65 be the interval neutrosophic sets defined in example [ example1 ] .
then , @xmath117,[0.03,0.15],[0.03,0.15 ] \rangle / x_1 + \\ \langle [ 0.6,0.79 ] , [ 0,0.08 ] , [ 0.1,0.24 ] \rangle / x_2 + \langle [ 0.76,0.92],[0,0.03],[0.03,0.12 ] \rangle / x_3 $ ] .
+ the _ scalar multiplication _ of interval neutrosophic set @xmath1 is an interval neutrosophic set @xmath65 , writeen as @xmath118 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 by @xmath119 for all @xmath53 in @xmath2 , @xmath120 .
the _ scalar division _ of interval neutrosophic set @xmath1 is an interval neutrosophic set @xmath65 , writeen as @xmath118 , whose truth - membership , indeterminacy - membership and false - membership functions are related to those of @xmath1 by @xmath121 for all @xmath53 in @xmath2 , @xmath120 .
now we will define two operators : truth - favorite ( @xmath122 ) and false - favorite ( @xmath123 ) to remove the indeterminacy in the interval neutrosophic sets and transform it into interval valued intuitionistic fuzzy sets or interval valued paraconsistent sets .
these two operators are unique on interval neutrosophic sets .
the _ truth - favorite _ of interval neutrosophic set @xmath1 is an interval neutrosophic set @xmath65 , writeen as @xmath124 , whose truth - membership and false - membership functions are related to those of @xmath1 by @xmath125 for all @xmath53 in @xmath2 .
[ example8 ] let @xmath1 be the interval neutrosophic set defined in example [ example1 ] .
then , @xmath126,[0,0],[0.3,0.5 ] \rangle / x_1 + \langle [ 0.5,0.9 ] , [ 0,0 ] , [ 0.2,0.3 ] \rangle / x_2 + \\ \langle [ 0.8,1.0],[0,0],[0.2,0.3 ] \rangle / x_3 $ ] .
+ the purpose of truth - favorite operator is to evaluate the maximum of degree of truth - membership of interval neutrosophic set .
the _ false - favorite _ of interval neutrosophic set @xmath1 is an interval neutrosophic set @xmath65 , writeen as @xmath127 , whose truth - membership and false - membership functions are related to those of @xmath1 by @xmath128 for all @xmath53 in @xmath2 .
[ example9 ] let @xmath1 be the interval neutrosophic set defined in example [ example1 ] .
then , @xmath129,[0,0],[0.6,1.0 ] \rangle / x_1 + \langle [ 0.5,0.7 ] , [ 0,0 ] , [ 0.2,0.5 ] \rangle / x_2 + \\ \langle [ 0.6,0.8],[0,0],[0.4,0.6 ] \rangle / x_3 $ ] .
+ the purpose of false - favorite operator is to evaluate the maximum of degree of false - membership of interval neutrosophic set .
for every two interval neutrosophic sets @xmath1 and @xmath65 : + 1 .
@xmath130 2 .
@xmath131 3 .
@xmath132 4 .
in this section , we will give some properties of set - theoretic operators defined on interval neutrosophic sets as in section [ section3 ] . the proof of these properties is left for the readers .
@xmath134 , @xmath135 , @xmath136 , @xmath137 @xmath138 , + @xmath139 , + @xmath140 , + @xmath141 .
@xmath142 , @xmath143 .
@xmath144 , @xmath145 , @xmath146 , @xmath147 .
@xmath148 , @xmath149 , where @xmath150 , @xmath151 and @xmath152 , @xmath153 .
@xmath154 , @xmath155 .
@xmath156 , @xmath157 , where @xmath158 , @xmath151 and @xmath159 , @xmath153 .
@xmath160 , @xmath161 @xmath162 , @xmath163 .
@xmath164 here , we notice that by the definitions of complement , union and intersection of interval neutrosophic set , interval neutrosophic set satisfies the most properties of class set , fuzzy set and intuitionistic fuzzy set .
same as fuzzy set and intuitionistic fuzzy set , it does not satisfy the principle of middle exclude .
we assume that @xmath2 is a real euclidean space @xmath165 for correctness . an interval neutrosophic set @xmath1 is convex if and only if @xmath166 for all @xmath85 and @xmath86 in @xmath2 and all @xmath167 in @xmath26 $ ] . fig .
[ fig1 ] is an illustration of convex interval neutrosophic set .
if @xmath1 and @xmath65 are convex , so is their intersection .
an interval neutrosophic set @xmath1 is + strongly convex if for any two distinct points @xmath85 and @xmath86 , and any @xmath167 in the + open interval @xmath168 , @xmath169 for all @xmath85 and @xmath86 in @xmath2 and all @xmath167 in @xmath26 $ ] .
if @xmath1 and @xmath65 are strongly convex , so is their instersection .
in this paper , we have presented an instance of neutrosophic set called interval neutrosophic set ( ins ) .
the interval neutrosophic set is a generalization of classic set , fuzzy set , interval valued fuzzy set , intuitionistic fuzzy sets , interval valued intuitionistic fuzzy set , interval type-2 fuzzy set @xcite and paraconsistent set .
the notions of inclusion , union , intersection , complement , relation , and composition have been defined on interval neutrosophic set .
various properties of set - theoretic operators have been proved . in the future
, we will create the logic inference system based on interval neutrosophic set and apply the theory to solve practical applications in areas such as such as expert system , data mining , question - answering system , bioinformatics and database , etc .
[ theorem0 ] @xmath102 is the smallest interval neutrosophic set containing both @xmath1 and @xmath65 .
let @xmath69 .
@xmath170 , @xmath171 , @xmath172 . @xmath173 , @xmath174 , @xmath175 .
@xmath176 , @xmath177 , @xmath178 , + @xmath179 , @xmath180 , @xmath181 , @xmath182 , @xmath183 , @xmath184 .
+ @xmath185 , @xmath186 , @xmath187 .
that means @xmath68 contains both @xmath1 and @xmath65 .
+ furthermore , if @xmath188 is any extended vague set containing both @xmath1 and @xmath65 , then @xmath189 , @xmath190 , so @xmath191 .
@xmath192 , @xmath193 , so @xmath194 .
@xmath195 , @xmath196 , so @xmath197 .
@xmath198 , @xmath199 , so @xmath200 .
@xmath201 , @xmath202 , so @xmath203 .
@xmath204 , @xmath205 , so @xmath206 .
that implies @xmath207 .
@xmath105 is the largest interval neutrosophic set contained in both @xmath1 and @xmath65 .
the proof is analogous to the proof of theorem [ theorem0 ] .
@xmath108 @xmath208 .
for every two interval neutrosophic sets @xmath1 and @xmath65 : + 1 .
@xmath130 2 .
@xmath131 3 .
@xmath132 4 .
@xmath133 we now prove the first identity .
let @xmath69 .
+ @xmath209 , + @xmath210 , + @xmath211 , + @xmath212 , + @xmath213 , + @xmath214 .
+ @xmath215 , + @xmath216 , + @xmath217 , + @xmath218 , + @xmath219 .
+ @xmath220 , + @xmath221 , + @xmath222 , + @xmath223 , + @xmath224 .
+ @xmath225 , + @xmath226 , + @xmath227 , + @xmath228 , + @xmath229 .
+ @xmath230 , + @xmath231 , + @xmath232 , + @xmath233 , + @xmath234 .
+ because , + @xmath235 , + @xmath236 , + @xmath237 , + @xmath238 , + @xmath239 , + @xmath240 , + so , @xmath130 .
the other identities can be proved in a similar manner .
[ theorem4 ] if @xmath1 and @xmath65 are convex , so is their intersection .
let @xmath71 , then + @xmath241 , @xmath242 , @xmath243 , @xmath244 , @xmath245 , @xmath246 , since @xmath1 and @xmath65 are convex : @xmath247 , @xmath248 , @xmath249 , @xmath250 , @xmath251 , @xmath252 , + @xmath253 , @xmath254 , @xmath255 , @xmath256 , + @xmath257 , + @xmath258 , + hence , + @xmath259 , @xmath260 , + @xmath261 , + @xmath262 , + @xmath263 , @xmath264 .
if @xmath1 and @xmath65 are strongly convex , so is their instersection .
the proof is analogus to the proof of theorem [ theorem4 ] .
the authors would like to thank dr .
florentin smarandache for his valuable suggestions .
atanassov , k. ( 1986 ) .
intuitionistic fuzzy sets " .
_ fuzzy sets and systems _ * 20 * 8796 .
atanassov , k. , gargov , g. ( 1989 ) .
interval valued intuitionistic fuzzy sets " .
_ fuzzy sets and systems _ * 31 * 343349 .
liang , q. , mendel , j. ( 2000 ) .
interval type-2 fuzzy logic systems : theory and design " . _ ieee transactions on fuzzy systems _ * 8 * 535550 .
turksen , i. ( 1986 ) .
interval valued fuzzy sets based on normal forms " .
_ fuzzy sets and systems _ * 20 * 191210 . | neutrosophic set is a part of neutrosophy which studies the origin , nature , and scope of neutralities , as well as their interactions with different ideational spectra .
neutrosophic set is a powerful general formal framework that has been recently proposed .
however , neutrosophic set needs to be specified from a technical point of view . to this effect
, we define the set - theoretic operators on an instance of neutrosophic set , we call it interval neutrosophic set ( ins ) .
we prove various properties of ins , which are connected to the operations and relations over ins . finally , we introduce and prove the convexity of interval neutrosophic sets .
* interval neutrosophic sets * + _ department of computer science + georgia state university + atlanta , georgia 30302 , usa + email : [email protected] , \{cscpnmx , yzhang , raj}@cs.gsu.edu _ _ key words : _ neutrosophic set , interval neutrosophic set , set - theorectic operator , convexity |
the genetic information carried by the mrna and then translated into proteins is encoded into nucleotide triplets called _ codons_. four alternate nucleotidic bases ( a , u , c , g ) compose the mrna , so that there are @xmath4 possible codons that have to code for only @xmath5 naturally occurring amino acids .
the genetic code is therefore redundant : while a few amino acids correspond to a single codon , most amino acids can be encoded by different codons .
different codons coding for the same amino acid are known as synonymous codons , and in a wide variety of organisms synonymous codons are used with different frequencies a phenomenon known as _
codon bias_. with the advent of whole - genome sequencing of numerous species , genome - wide patterns of codon bias are emerging in the different organisms .
various factors such as expression level , gc content , recombination rates , rna stability , codon position , gene length , environmental stress and population size , can influence codon usage bias within and among species @xcite . while the biological meaning and origin of codon bias is not yet fully understood , there is a large consensus that the degeneration of the genetic code might provide an additional degree of freedom to modulate accuracy and efficiency of translation @xcite . indeed , population genetic studies @xcite have shown that synonymous sites are under weak selection , and that codon bias is maintained by a balance between mutation - selection ( random variability in genetic sequences followed by fixation of the optimal codons ) and genetic drift ( allowing for the occurrence of non - optimal codons ) . in fact
, highly expressed genes feature an extreme bias by using a small subset of codons , optimized by translational selection @xcite . on the other hand ,
the persistence of non - optimal codons in less - expressed sequences causes long breaks during protein synthesis ; this could be the result of genetic drift and have a key role in the protein folding process @xcite .
in addition , codon usage appears to be structured along the genome , with neighboring genes having similar codon compositions @xcite , and codon bias seems positively correlated to gene length ( as a result of selection for accuracy in the costly production of long proteins ) @xcite . in the last years
there has been a wide effort in developing effective ways to measure codon bias @xcite .
the most widely used indices include the _ codon adaptation index _
( @xmath2 ) @xcite , the _ trna adaptation index _ ( @xmath1 ) @xcite , and the _ effective number of codons _ ( @xmath3 ) @xcite , each of them having specific advantages and drawbacks .
for instance , @xmath2 and @xmath1 correlate well with gene expression levels , however such correlation is a natural consequence of their definition : they are tuned on a reference set of highly expressed genes .
@xmath3 is instead basically a measure of the entropy of the codon usage distribution , and thus shows a lower correlation with expression levels . in this work
we propose a novel codon bias index named _ competition adaptation index _
( @xmath0 ) , which does not rely on information about gene expression levels , but instead has a self - consistent biological meaning based on trna availability and competition between cognate and near - cognate trnas . in other words
, @xmath0 is a parameter - free index that does not require a set of reference genes for its calibration , a fact that constitutes its main advantage with respect to @xmath2 and @xmath1 .
moreover , @xmath0 is designed to extract genetic signals that could be directly correlated to experimental measures for translation speeds , an emerging and challenging issue still to be explored . in order to show the advantage of the novel codon bias index , we perform a genome - wide comparison of @xmath2 , @xmath1 , @xmath3 and @xmath0 for _ escherichia coli _ ( e.coli ) .
our analysis reveals that the information on gene conservation across species and gene essentiality is better captured by codon bias metrics that build on trna availability ( @xmath1 and @xmath0 ) .
we also study codon bias in relation to the connectivity patterns of the protein - protein interaction network ( pin ) @xcite of e.coli .
we thus show that translational selection systematically favors proteins with the highest number of interactions and belonging to the most densely connected community of the network , at least when the bias is measured by @xmath0 and , to a smaller extent , by @xmath1 .
additionally , we address the issue of how much a similarity in the codon usage bias of a set of genes is reflected on the interactions among the corresponding proteins .
a principal component analysis for the variability of codon bias indices indeed reveals that closeness of a set of genes in the space of the two principal components likely results in the corresponding proteins to interact in comparison with an appropriate null model .
overall , our study reveals that @xmath0 captures more information than the other indices about the connection between codon bias and the topology of the interactome . besides , we recall that @xmath0 does not require calibration on gene expression levels and has a consistent biological meaning based on the competition between cognate and near - cognate trnas .
these observations stress the potential of the new index to both measure and explain codon usage bias .
in this work we investigate the genome of e.coli k-12 substr .
mg1655 , whose 4005 coding mrna sequences have been collected from genbank @xcite .
the gene copy numbers coding for each trna ( tgcn ) were derived from the genomic trna database @xcite . in order to have an index for gene `` conservation '' , we use the normalized evolutionary retention index ( eri ) @xcite : for each gene in e.colis genome , its eri measures how much that gene is shared among other 32 bacterial species ( having at least an ortholog of the given gene ) .
a low eri value thus denotes that a gene is specific to e.coli , whereas , high eri is characteristic of highly shared ( and therefore conserved ) genes .
concerning gene `` essentiality '' , we use the classification of gerdes et al .
@xcite for the e.coli genome into 606 essential and 2940 non - essential genes , based on experimental measures of gene resistance against transposon insertion . codon usage bias can be assessed , for each gene in a given genome , by various indices that can be classified into broad groups based on : ( i ) codon frequencies ; ( ii ) reference gene sets ; ( iii ) deviation from a postulated distribution ; ( iv ) information theory ; ( v ) interactions among trnas ( see @xcite for an overview ) .
we focus here on the most widely used indices : @xmath1 @xcite , that belongs to groups ( ii ) and ( v ) by requiring calibration on a set of highly expressed genes ; @xmath2 @xcite , a group ( i ) and ( ii ) index built on local statistics of codon usage and on a reference list of optimally expressed genes ; @xmath3 @xcite , a group ( i ) index based on the number of different codons used in a coding sequence .
the novel codon bias index we propose in this work , @xmath0 , is instead based on the competition of cognate and near - cognate trnas to bind to the a - site on the ribosome during translation , and is thus a group ( v ) index that does not need tuning on a reference set of highly expressed genes .
while the formal definition for @xmath0 and the rationale behind are given below , we refer to the supporting information for the operative definition of @xmath2 , @xmath1 and @xmath3 . * competition adaptation index * ( @xmath0 ) .
it is generally accepted that translation speed depends on the efficiency of the codon / anticodon pairing in the a site of the ribosome @xcite .
hence , for a given codon , the rate of amino acid synthesis is essentially influenced by two dominant processes : the number of collisions of the corresponding trna with the ribosomal a site ( which strongly depends on trna concentration in the cell ) and the specificity of the codon - anticodon pairing .
such a pairing process satisfies the watson - crick base - pairing rules ( g - c and a - u , and vice versa ) for the first two bases , whereas , the rule on the third ( or _ wobble _ ) base is more relaxed and non - standard pairing is allowed in some cases ( wobble complementarity)@xcite .
hence , there are cases in which several trnas pair with the same codon ( provided that these are identical in the first two bases ) and are called isoacceptor or cognate trnas .
codon - anticodon interactions are thus characterized by competition between cognate trna ( with perfect or wobble complementarity between mrna codon and trna anticodon ) , near - cognate trna ( with a mismatch in only one of the first two bases ) and non - cognate trna ( with at least two mismatches ) .
discrimination between correct and wrong trna according to base pairing features very high fidelity ( error rate @xmath6 ) .
rejection of the wrong trna can occur in two distinct phases @xcite : initial selection of the ternary complex ef - tu - gtp - aa - trna and subsequent proofreading of aa - trna after gtp hydrolyzation .
the first interaction is fast and does not depend on the choice of codons , in order to allow the ribosome to quickly screen for the available trnas .
the second step is instead sensitive to base complementarity , featuring the first selection between cognate and near - cognate trnas : non - cognate are excluded almost immediately with @xmath7 , and then a more strict and efficient proofreading takes place , excluding near - cognate with @xmath8 .
this means that near - cognate trna ( unlike non - cognate trna ) can enter into the interaction process between the ternary complex and the site a of the ribosome , and ( when not accepted , in very few cases ) can be rejected at the stage of initial recognition or during proofreading . in any event , this process results in a time delay of translation , because near - cognate rejection brings the ribosome back to the initial state of waiting for the correct trna .
the rationale behind the definition of @xmath0 is precisely that of building an index which is based both on trna availability and on competition between cognate and near - cognate trnas that could modulate the speed of translation of mrnas into proteins .
note that , since in vivo experimental determinations of trna concentrations are available only for few organisms , we will implement @xmath0 using the number of trna gene copies ( tgcn ) which , at least in simple organisms , has a high and positive correlation with trna abundance @xcite ( a similar approach is adopted in the definition of @xmath1 @xcite ) . for each codon
@xmath9 we define its absolute adaptiveness value ( @xmath10 ) as : @xmath11.\ ] ] here @xmath12 is the number of isoacceptor trna sequences ( anticodons ) that recognize codon @xmath9 ( _ i.e. _ , containing either the anti - codon @xmath9 or all its cognates that are read by @xmath9 ) and @xmath13 is the gene copy number of the @xmath14-th of such trnas , whereas @xmath15 is the number of trna sequences that are near - cognate of @xmath9 and @xmath16 is the gene copy number of the @xmath14-th of such trnas ( see also fig . [ fig.cognate ] of supporting figures ) .
the amount in square brackets represents a penalty introduced by the competition with near - cognate trnas , assuming unit or zero values in the cases of smaller and higher competition , respectively .
this term thus assumes the role of selective constraint on the efficiency of the codon - anticodon coupling .
importantly , and at odds with @xmath1 , these terms do not result from optimization on expression levels , but have a biological justification based on cognate / near - cognate competition . note that , in the computation of @xmath10 for a given codon , we count as isoacceptor trnas those with perfect or wobble base pairing that also carry the same amino acid of @xmath9 s anticodon .
computation of @xmath0 continues by defining for each codon @xmath9 its relative adaptiveness value @xmath17 , where @xmath18 is the maximum value between all the @xmath10 of codons .
@xmath0 of gene @xmath19 is finally defined as the harmonic mean of the relative adaptiveness of its codons : @xmath20 the choice of the harmonic mean ( rather than geometric as for @xmath2 and @xmath1 ) is consistent with the association of @xmath0 with the rate of protein synthesis .
indeed , the translation speed of codon @xmath9 can be defined as the reciprocal of the concentration of the corresponding trna isoacceptors @xcite .
therefore , if codon @xmath9 is read at a speed proportional to @xmath21 , then the average translation speed of a gene is given by the harmonic mean of the @xmath22 associated to its codons .
@xmath0 takes values between 0 and 1 , where values close to 0 ( 1 ) indicate highest ( lowest ) competition , and therefore a low ( high ) translation rate . in this study we use protein interaction data collected in string ( known and predicted protein - protein interactions ) @xcite . in such database , each predicted interaction is assigned with a confidence level or probability @xmath23 , evaluated by comparing predictions obtained by the different techniques @xcite with a set of reference associations , namely the functional groups of kegg ( kyoto encyclopedia of genes and genomes ) @xcite . in this way ,
interactions with high @xmath23 are likely to be true positives , whereas , a low @xmath23 likely corresponds to a false positive .
since the percentage of false positives can be very high @xcite , we select a stringent cut - off @xmath24 that allows a fair balance between coverage and interaction reliability ( see the probability distribution @xmath25 in the left panel of fig .
[ fig.p_kw ] of supporting figures ) .
we thus build the protein - protein interaction network ( pin ) of e.coli by placing a link between each pair of proteins ( nodes ) @xmath26 provided that @xmath27 . the resulting number of connections or _
degree _ for a given protein @xmath9 is denoted as @xmath28 . to detect communities of pin we resort to molecular complex detection ( mcode ) @xcite . in a nutshell ,
mcode iteratively groups together neighboring nodes with similar values of the core - clustering coefficient , which for each node is defined as the density of the highest @xmath29-core of its immediate neighborhood times @xmath29 @xcite .
mcode detects the densest regions of the network and assigns to each found community a score that is its internal link density times the number of nodes belonging to it @xcite .
we also characterize each found community @xmath30 with the mean value @xmath31 and standard deviation @xmath32 of codon bias values within the community , and use them to compute a @xmath33-score as @xmath34 ( where @xmath35 and @xmath36 are , respectively , the mean value and standard deviation of codon bias values computed on the whole network ) . in this way
, a value of @xmath37 ( @xmath38 ) indicates that community @xmath30 features significantly higher ( lower ) codon bias than the population mean . finally note that
each node of pin is photogenically classified according to the clusters of orthologous groups @xcite ( cogs ) of proteins @xcite .
cogs are generated by comparing predicted and known protein sequences in all completely sequenced genomes to infer sets of orthologs .
each cog consists of a group of proteins found to be orthologous across at least three lineages and likely corresponds to an ancient conserved domain @xcite .
principal component analysis ( pca ) @xcite is a multivariate statistical method to transform a set of observations of possibly correlated variables into a set of linearly uncorrelated variables ( called principal components ) spanning a space of lower dimensionality .
the transformation is defined so that the first principal component accounts for the largest possible variance of the data , and each succeeding component in turn has the highest variance possible under the constraint that it is orthogonal to ( _ i.e. _ , uncorrelated with ) the preceding components .
we use this technique on the space of codon bias indices , so that each gene of e.coli is represented as a 4-dimensional vector with coordinates ( @xmath0 , @xmath2 , @xmath1 , @xmath3 ) .
such coordinates are separately normalized to zero mean and unit variance over the whole genome .
we then obtain the associated covariance matrix between the four dimensions of codon bias and diagonalize it .
the eigenvectors of the covariance matrix , ordered according to the magnitude of the corresponding eigenvalues , are the principal components of the original data . in order to assess how significant are the codon usage patterns observed for the pin , we need to compare the e.coli interactome with a suitable null model for it , _
i.e. _ , an appropriate randomization of the network . here
we follow the most common approach in statistical mechanics of networks of using the _ configuration model _ ( cm ) @xcite .
the basic idea is to build the null model as an ensemble @xmath39 of graphs with maximum entropy , except that the ensemble average of the node degrees are constrained to the values observed for the real network : @xmath40 @xmath41 .
this leads to a probability distribution over @xmath39 which is defined via a set of lagrange multipliers @xmath42 ( one for each node ) associated to the constraints @xcite .
once all @xmath42 are found , the cm reduces to having a link between nodes @xmath9 and @xmath14 with probability @xmath43 , independently on all other links .
then , the null hypothesis is that any given network property @xmath44 varies in the range @xmath45 $ ] , where both average and standard deviation of @xmath44 over the ensemble can be obtained either analytically or numerically ( by drawing sample networks from @xmath39 ) @xcite .
the number of standard deviations by which the empirical and expected values of @xmath44 differ is given by the @xmath33-score @xmath46=(\chi-\langle \chi \rangle_{\omega})/\sigma_{\omega}[\chi]$ ] : large positive ( negative ) values of @xmath46 $ ] indicate that @xmath47 is substantially larger ( smaller ) than expected , whereas , small values signal no significant deviation from the null model .
as the starting point of our analysis , we first check how the different codon bias indices correlate over e.colis genome .
[ corr ] shows that , interestingly , @xmath0 is strongly ( and positively ) correlated with @xmath1 , whereas it does not show any significant correlation with @xmath2 nor with @xmath3 .
this result can be easily explained as @xmath0 and @xmath1 elaborate on the same genetic information , that is the abundance of trnas , whereas @xmath2 and @xmath3 are based on codon usage statistics ( see the supporting information ) .
is strongly and positively correlated with @xmath1 ( @xmath48 ) , but not with both @xmath2 nor @xmath3 ( @xmath49).,scaledwidth=75.0% ] we move further and analyze the correlation between the various codon bias indices and the evolutionary retention index ( eri ) @xcite for e.coli genes ( we recall that a gene with a low eri value is peculiar to e.coli , whereas a gene with high eri is shared among different species ) . fig .
[ correri ] reports the average values and standard deviations of the codon bias indices for every group of genes having the same eri value .
interestingly , the evolutionary codon adaptation measured by @xmath0 and @xmath1 tends to increase for genes that are less specific to e.coli .
[ correri ] also suggests that it is possible to make a threefold separation of genes by looking at the rate of variation of @xmath1 and @xmath0 with eri .
we thus identify group a ( eri @xmath50 0.2 : 1597 low eri genes that are specific to e.coli ) , group b ( 0.2 @xmath50 eri @xmath50 0.9 : 1804 intermediate eri genes ) and group c ( eri @xmath51 0.9 : 231 high - eri genes that are highly conserved and shared among several bacterial species ) . in each group ,
the correlation between codon bias and eri is maximized ( see the corresponding correlation coefficients in the figure ) . @xmath2 and @xmath3 are instead less structured with respect to eri , as shown by the very small correlation coefficients ( and by the impossibility to identify gene groups ) . denoting the corresponding correlation coefficients . in the left panels ,
the fits are performed separately for the three groups of genes a ( eri @xmath50 0.2 ) , b ( 0.2 @xmath50 eri @xmath50 0.9 ) and c ( eri @xmath51 0.9 ) .
both @xmath0 and @xmath1 monotonously increase with eri , whereas @xmath2 and @xmath3 show a low correlation with eri.,scaledwidth=75.0% ] we now study the patterns of codon usage bias in essential and non - essential genes , according to the classification scheme of gerdes et al .
@xcite ( see materials and methods ) . as a preliminary result , fig .
[ esseri ] reports the percentage of essential genes in each set of genes sharing the same eri .
we see that the three groups a , b , c of genes identified as in the previous paragraph feature different percentages of essential genes : approximately , 10% for group a , 15% for group b and above 30% for group c. essentiality and eri thus seems to capture similar genetic features .
[ e - ne ] shows instead that @xmath0 and @xmath1 are more sensitive than @xmath2 and @xmath3 in distinguishing essential from non - essential genes .
overall , figs .
[ correri ] and [ e - ne ] provide a clear indication that codon bias , as measured by @xmath1 and @xmath0 , is more pronounced for genes that are highly conserved ( _ i.e. _ , with high eri ) and essential , on the other hand @xmath2 and @xmath3 are less sensitive to these quantities .
-eri and @xmath1-eri ) .
the groups have different incidences of essential genes : 10% for group a ( eri @xmath50 0.2 ) , 15% for group b ( 0.2 @xmath50 eri @xmath50 0.9 ) and more than 30% for group c ( eri @xmath51 0.9).,scaledwidth=37.5% ] and @xmath1 can effectively separate essential from non - essential genes .
in fact , in the left panels the average codon bias values for essential and non - essential genes have a relative variation of about 5% , whereas , in the right panels such values are almost coincident and the errors overlap.,scaledwidth=75.0% ] we now perform a kind of gene ontology to check how the three gene groups a , b , c are projected over the clusters of orthologous genes ( cogs ) and their functional annotations @xcite . to this end
, for each group we evaluate the bayesian probability that its genes belong to a given cog : @xmath52 , where @xmath53 is estimated as the fraction of the genome belonging to the group , @xmath54 as the fraction of the genome belonging to the cog and @xmath55 is the fraction of genes in the cog that belong to a particular group . fig .
[ histo ] shows the histogram of @xmath56 over the 17 cogs , for the three groups a , b , c defined above . assuming an arbitrary discriminating threshold of 10%
, we observe that each group is prevalently projected over a limited set of cogs ( reported in the legend box of fig .
[ histo ] ) .
group a genes ( those with low eri values ) mostly insist over cogs k and g
( transcription , carbohydrate metabolism ) ; group b ( genes with intermediate eri ) is enriched in cogs g and e ( again , carbohydrate metabolism , amino acid metabolism and transport ) ; finally , group c ( genes with the highest eri ) is dominated by the functional annotations associated with cogs j and l ( translation , ribosome structure and biogenesis , replication , recombination and repair ) .
indeed , group c , composed of the highly adapted , essential , and conserved genes of e.coli , is the set of genes that code for ribosomal proteins . over the cogs for the three gene groups a , b , c*.
each group is characterized by one or a few predominant cogs , indicated within parenthesis in the legend ( assuming a threshold of @xmath57 and excluding generic cogs r and s , for which function prediction is too general or missing).,scaledwidth=80.0% ] we now turn our attention to the network of interacting proteins in e.coli .
we start by studying codon bias in relation with the connectivity patterns of the network .
first , note that the degree distribution of proteins is scale - free ( see the right panel of fig .
[ fig.p_kw ] of supporting figures ) , meaning that the network features a large number of poorly connected proteins and a relatively small number of highly connected hubs .
[ grado ] notably shows that these hub proteins are systematically characterized by higher values of codon bias of the corresponding genes when this is measured by @xmath1 and @xmath0 .
@xmath2 and @xmath3 are instead clearly less sensitive to protein connectivity . of the corresponding proteins in the pin of e.coli*. solid lines are linear fits . @xmath0 and @xmath1 of a gene definitely increase with the connectivity of the corresponding protein in the pin , whereas the other two indices are less sensitive to this parameter.,scaledwidth=75.0% ] we move further and consider codon bias in relation with the community structure of the pin .
we recall that a community is a group of proteins that are more densely connected within each other than with the rest of the network .
table [ tab.comm ] shows the features of the communities that are assigned by mcode a score higher than 10 , together with their cog composition , average degree and , for eri and the various codon bias indices , the internal average value @xmath31 and the @xmath33-scores ( comparing the distribution of bias inside the community with that of the whole network ) .
we see that such topologically determined communities , ordered by score , are evolutionarily and functionally characterized by a dominant cog , shared by the majority of the proteins in the community .
this suggests that the identified communities can be associated with specific metabolic functions : they correspond to functional modules , essential for the life - cycle of the organism .
.*features of top - scoring communities*. number of nodes ( @xmath58 ) , community score ( @xmath58 times the internal density ) , predominant cog label and percentage ; then , for eri and the various codon bias indices , mean values @xmath31 internal to the community and @xmath33 scores ( between square brackets ) .
values of @xmath59 are reported in bold . [ cols="^,^,^,^,^,^,^,^,^,^,^",options="header " , ] [ tab.comm ] let us focus on the first community , that includes only 60 proteins ( 4.5% of the whole network ) but as much as 32.6% of the total number of links in the network , and that basically overlaps with the main core of the pin ( _ i.e. _ , the @xmath29-core with the highest possible degree ) .
notably , proteins belonging to this community have on average a codon bias index ( as measured by @xmath1 and , even more , by @xmath0 ) that is significantly higher than the average of the rest of the network ( the @xmath33-score is bigger than 1 ) . as noticed above , this core
is essentially composed of ribosomal proteins , that are usually highly expressed , have the highest codon usage bias , and are broadly conserved and essential across different taxa @xcite .
finally we perform pca over the space of the four codon bias indices ( @xmath0 , @xmath2 , @xmath1 , @xmath3 ) measured for each e.coli gene .
the two first principal components ( pc1 and pc2 ) turn out to represent for as much as 85% of the total variance of codon bias over the genome ( left plot of fig .
[ pca ] ) .
projection of the first two principal components on the individual codon bias indices ( loadings ) shows that none of the four indices predominantly contributes to the data variability ( right plot of fig .
[ pca ] ) .
thus , the placement of a gene in the pc1-pc2 plane depends on a weighted contribution of all the indices .
interestingly , the genes encoding for the proteins of the eight top mcode communities are well localized and separated in this reduced space ( fig . [ pca_comm ] ) . in particular , the first community ( _ i.e. _
, the core of ribosomal proteins characterized by high values of both @xmath0 and @xmath1 ) is located in the upper left part of the graph , isolated from the others .
this represents an important evidence : proteins that belong to the densest connected cores of the interactome are well - localized in the space of the two principal components .
is anticorrelated with the other codon bias indices , pc1 results from a weighted and coherent contribution of all the indices , whereas , for pc2 the contribution of @xmath0 and @xmath1 is opposite to that of @xmath2 and @xmath3.,scaledwidth=50.0% ] in other words , if a set of proteins are physically and functionally connected in a module , then their corresponding genes should share common codon bias features .
conversely , we can obtain an estimate for the conditional probability @xmath60 of a functional interaction between proteins , provided that their relative genes fall within a distance @xmath61 in the plane of the two principal components pc1 and pc2 .
reasonably , we compare @xmath60 estimated on the real interactome with @xmath62 estimated on the configuration model ( cm ) which , we recall , is a degree - conserving randomization ( re - wiring ) of the network .
[ p_link ] shows the @xmath33-score for @xmath63 ) as a function of @xmath61 , and reveals a peculiar behavior : for small distances ( @xmath64 ) the probability of finding a connection between two proteins is much higher than what could have been expected from a ( degree - conserving ) random link placement .
conversely , for medium distances ( @xmath65 ) , the linking probability is lower than that of the cm , whereas , the real network and the cm become compatible for large distances , where , however , connections are rather few .
this analysis shows that sets of genes sharing similar codon usage patterns encode for proteins that are much more likely to interact than in situations where chance alone is responsible for the structure of the interactome .
-score for @xmath66 for each pair of genes and their respectively encoded proteins .
* @xmath61 is the euclidean distance between pairs of genes in the space of the first two pca components , and @xmath60 is the conditional probability of having a link in the pin between two proteins given that their encoding genes are localized within a distances @xmath61 in the pc1-pc2 plane .
the @xmath33-score is obtained as @xmath67=[pr(\mbox{link}|d)-\langle pr(\mbox{link}|d ) \rangle_{\omega}]/\sigma_{\omega}[pr(\mbox{link}|d)]$ ] .
the gray dashed lines mark the significance interval of @xmath68.,scaledwidth=50.0% ]
in this work we have introduced @xmath0 , a novel codon bias index that is inspired by @xmath1 , though conceptually distinct .
in fact , @xmath0 does not make reference to lists of highly expressed genes , and is thus unsupervised and based on intrinsic information about co - evolution of genes that code for proteins and trnas .
conceptually , the definition of @xmath0 is based on a model that postulates a competition between cognate and near - cognate trnas for the same codon , exposed on the ribosome at each step of protein synthesis .
competitive mechanisms in the machinery of ribosomal translation of genes into proteins have been repeatedly suggested and studied in the literature @xcite and deserve further attention in order to understand their role for translation efficiency . in particular ,
@xmath0 was designed in order to provide information about the speed of protein synthesis , being based on proofreading delay mechanisms .
our genome - wide analysis of codon bias in e.coli using @xmath0 as well as other commonly used indices revealed that codon usage metrics resting on counting trna genes ( @xmath0 and @xmath1 ) are strongly and positively correlated among themselves in spite of their conceptually different definition .
it would then be quite interesting to check in the future whether this correlation is specific to e.coli or it is universally observed in the genomes of bacterial species that are either ecologically and evolutionarily close or , by contrast , very far from e.coli .
we also found that both @xmath0 and @xmath1 correlate with eri , the degree of conservation for a gene among similar species , and gene essentiality , whereas , @xmath2 and @xmath3 are less sensitive to these quantities . @xmath0 and
@xmath1 values also allow to distinguishing three groups of genes , that are differently characterized by both codon choice adaptation , eri and degree of essentiality , and that also feature specific predominant cog signatures .
in particular the third group ( c ) , composed of the few genes that are highly conserved and with the strongest codon bias adaptation , consists for 30% of essential genes with predominant cogs j and l that refer to translation , ribosome structure and biogenesis , replication , recombination and repair .
these represent house - keeping and control functions that must be continuously executed by the cell , meaning that the genes responsible for them have to be expressed most of the time during the cell cycle .
these observations strongly support the idea that an increasing selection of codons and , in parallel , a correlated modulation of trna availability co - evolved along the evolutionary history of a species .
finally , we have addressed a theme as relevant as the connection between codon usage bias and protein functional or physical interactions .
our main result indicates that , in the course of the evolution of a genome , the functional structuring of the complex of interactions between proteins has interfered with the peculiar codon - coding formulation of the corresponding genes .
in particular we have shown here , for the first time to our knowledge , that communities of highly connected proteins in the interactome of e.coli correspond to encoding genes that share the same degree of evolutionary adaptation , as expressed by codon bias indices that synthetically represent genetic information encoded in the trnas sector of the genome .
indeed , @xmath0 , that is based on a simple representation of trna competition , seems to detect the codon bias signal behind communities more consistently than the other indices here considered .
conversely , we have provided evidence that if two genes have similar codon usage patterns then the corresponding proteins have a significant probability of being functionally connected and interacting .
this result points out that codon bias should be a relevant parameter in the fundamental problem of predicting unknown protein - protein interactions from genomic information .
this study is a first exploratory step towards a more complete investigation on how communities within protein - protein interaction networks rest on a consistent but still to be decoded codon bias signal .
indeed , the connection of the topology of a network with an underneath semantics is far from trivial , as recently pointed out in the specialized literature @xcite .
biological pins and codon bias offer an interesting case study worth to be investigated in a wider perspective @xcite .
here we define and explain the rationale behind the various codon bias indices that have been proposed in the literature . * codon adaptation index * ( @xmath2 ) @xcite .
the pattern of codon usage is ruled by two simultaneous effects @xcite : translational selection towards optimal codons for each amino acid , and genetic drift that allows the persistence of non - optimal codons .
it is natural to assume that selection is stronger for codons of highly expressed genes , which thus feature a more pronounced bias in the use of codons .
the principle behind @xmath2 is exactly that codon usage in highly expressed genes can reveal the optimal ( _ i.e. _ , most efficient for translation ) codons for each amino acid .
hence , @xmath2 builds on a reference set of highly expressed genes to assess , for each codon @xmath9 , the relative synonymous codon usages ( @xmath69 ) and the relative codon adaptiveness ( @xmath21 ) : @xmath70 in the @xmath69 , @xmath71 is the number of occurrences of codon @xmath9 in the genome , and the sum in the denominator runs over the @xmath72 synonyms of @xmath9 ; @xmath73s thus measure codon usage bias within a family of synonymous codons .
@xmath21 is then defined as the usage frequency of codon @xmath9 compared to that of the optimal codon for the same amino acid encoded by @xmath9_i.e .
_ , the one which is mostly used in a reference set of highly expressed genes .
the @xmath2 for a given gene @xmath19 is calculated as the geometric mean of the usage frequencies of codons in that gene , normalized to the maximum @xmath2 value possible for a gene with the same composition of amino acid : @xmath74 where the product runs over the @xmath75 codons belonging to that gene ( except the stop codon ) .
the critical aspect in the definition of @xmath2 is that it requires to define _ a priori _ reference set of highly expressed genes that is specific for a given organism .
@xmath2 is then not always transferable ; yet , since it is tuned on highly expressed genes , it is generally well correlated with gene expression levels in genomes for which reference gene sets are available . in this work ,
@xmath2 for e.coli genes was computed using the dambe 5.0 package @xcite * trna adaptation index * ( @xmath1 ) @xcite .
the speed of protein synthesis is bound to the waiting time for the correct trna to enter the ribosomal a site @xcite , and thus depends on trna concentrations @xcite ( and , indirectly , on gene copy numbers ) .
the consequent adaptation of codon usage to trna availability @xcite is at the basis of @xmath1 , which follows the same mathematical model of @xmath2defining for each codon @xmath9 its absolute ( @xmath10 ) and relative ( @xmath21 ) adaptiveness value : @xmath76 where @xmath12 is the number of isoacceptor trnas that recognize codon @xmath9 ( _ i.e. _ , trnas that carry the same aminoacid that is encoded by @xmath9 ) , tgcn@xmath77 is the gene copy number of the @xmath14-th trna that recognizes the @xmath9-th codon , @xmath78 is a selective constraint on the efficiency of the codon - anticodon coupling , @xmath18 is the maximum @xmath10 value and @xmath79 is the geometric mean of all @xmath21 with @xmath80 .
the @xmath1 of gene @xmath19 is eventually defined as the geometric mean of the relative adaptiveness values of its codons , thus estimating the amount of adaptation of gene @xmath19 to its genomic trna pool : @xmath81 the critical issue for @xmath1 is the selection of a meaningful set of @xmath78 values , _
i.e. _ , weights that represent wobble interactions between codons and trnas . assuming that trna usage is maximal for highly expressed genes , these values are chosen in order to optimize the correlation of @xmath1 values with expression levels
exactly as @xmath2 . besides , while the efficiencies of the different codon - trna interactions are expected to vary among different organisms , @xmath78 values are based on the gene expression in _ saccharomyces cerevisiae _
@xcite thus lacking universality @xcite . in this work we have evaluated @xmath1 values of e.coli genes using the codonr package @xcite .
* effective number of codons * ( @xmath3 ) @xcite .
@xmath3 is a measure that quantifies the departure of a gene from the random usage of synonymous codons .
given a sequence of interest , the computation of @xmath3 @xcite starts from the quantity defined for each family @xmath82 of synonymous codons : @xmath83 where @xmath84 is the number of codons in @xmath82 ( each appearing @xmath85 times in the sequence ) and @xmath86 .
@xmath3 then weights these quantities in order to measure amount of entropy in the codon usage of the sequence : @xmath87 where @xmath88 is the number of families with one codon only and @xmath89 is the number of families with degeneration @xmath90 ( families with degeneration 6 are divided into two families of degeneration 2 and 4 , as they often are subject to different selective forces ) .
note that @xmath3 reaches is maximal value ( 61 ) when all codons are used equally and its minimal value ( 23 ) when only one codon is used per amino acid ( extreme bias ) . differently from both @xmath2 and @xmath1 ,
@xmath3 is a more immediate measure of codon usage that does not require any _ a priori _ information nor makes any biological hypothesis ( which constitute its weakness and , and the same time , its strength ) . yet
, since the effect of selection is a reduction of entropy for codon usage in a sequence , @xmath3 provides a reliable measure for this effect . in this paper
we have obtained @xmath3 values through dambe 5.0 @xcite .
, with the vertical line indicating the cut - off we use to separate true from false positives .
right plot : * distribution of degrees * @xmath91 when @xmath24 , with the insets showing the same distribution for the original network ( @xmath92).,scaledwidth=100.0% ] hershberg r , petrov da .
selection on codon bias .
ann rev genet .
shabalina sa , spiridonov sa , kashina a. sounds of silence : synonimous nucleotides as a key to biological regulation and complexity .
nucleic acids res
. 2013;41(4):20732094 .
tuller t. challenges and obstacles related to solving the codon bias riddles .
biochem soc trans . 2014;42(1):155159 .
bennetzen jl , hall bd .
codon selection in yeast .
j biol chem .
gouy m , gautier c. codon usage in bacteria : correlation with gene expressivity .
nucleic acids res
. 1982;10(22):70557074 .
jansen r , bussemaker hj , gerstein m. revisiting the codon adaptation index from a whole - genome perspective : analyzing the relationship between gene expression and codon occurrence in yeast using a variety of models .
nucleic acids res . 2003;31(8):22422251 .
li gw , burkhardt d , gross c , weissmann js . quantifying absolute protein synthesis rates reveals principles underlying allocation of cellular resources .
cell 2014;157(3):624635 .
pop c , rouskin s , ingolia nt , han l , phizicky em , weissman js .
causal signals between codon bias , mrna structure , and the efficiency of translation and elongation .
mol syst biol .
daubin v , perrire g. g+c3 structuring along the genome : a common feature in prokaryotes .
mol biol evol .
eyre - walker a. synonymous codon bias is related to gene length in escherichia coli : selection for translational accuracy ?
mol biol evol .
alexander roth ma , cannarozzi gm .
measuring codon usage bias . in : cannarozzi gm ,
schneider a , editors .
codon evolution mechanisms and models .
oxford university press ; 2012 .
p. 189217 .
sharp pm , li wh . the codon adaptation index a measure of directional synonymous codon usage bias , and its potential applications .
nucleic acids res
. 1987;15(3):12811295 .
dos reis m , savva r , wernisch l. solving the riddle of codon usage preferences : a test for translational selection .
nucleic acids res
. 2004;32(17):50365044 .
wright f. the `` effective number of codons '' used in a gene . gene .
szklarczyk d , franceschini a , kuhn m , simonovic m , roth a , minguez p , et al .
the string database in 2011 : functional interaction networks of proteins , globally integrated and scored .
nucleic acids res .
2011;39(database issue):d561d568 .
benson da , karsch - mizrachi i , clark k , lipman dj , ostell j , sayers ew .
nucleic acids res .
2012;40(database issue):d48d53 .
chan pp , lowe tm .
gtrnadb : a database of transfer rna genes detected in genomic sequence .
nucleic acids res .
2009;37(database issue):d93d97 .
gerdes sy , scholle md , campbell jw , balsi g , ravasz e , daugherty md , et al .
experimental determination and system level analysis of essential genes in escherichia coli mg1655 .
. 2003;185(19):56735684 .
rodnina mv , wintermeyer w. fidelity of aminoacyl - trna selection on the ribosome : kinetic and structural mechanisms .
annu rev biochem .
anticodon pairing : the wobble hypothesis .
j mol biol .
gromadski kb , rodnina mv .
kinetic determinants of high - fidelity trna discrimination on the ribosome . mol cell .
ikemura t. correlation between the abundance of escherichia coli transfer rnas and the occurrence of the respective codons in its protein genes : a proposal for a synonymous codon choice that is optimal for the e. coli translational system .
j mol biol .
dong h , nilsson l , kurland cg .
co - variation of trna abundance and codon usage in escherichia coli at different growth rates .
j mol biol .
percudani r , pavesi a , ottonello s. transfer rna gene redundancy and translational selection in saccharomyces cerevisiae .
j mol biol .
kanaya s , yamada y , kudo y , ikemura t. studies of codon usage and trna genes of 18 unicellular organisms and quantification of bacillus subtilis trnas : gene expression level and species - specific diversity of codon usage based on multivariate analysis . gene .
zhang g , ignatova z. generic algorithm to predict the speed of translational elongation : implications for protein biogenesis .
chien ct , bartel pl , sternglanz r , fields s. the two - hybrid system : a method to identify and clone genes for proteins that interact with a protein of interest .
proc natl acad sci usa .
phizicky em , fields s. protein - protein interactions : methods for detection and analysis .
microbiol rev .
puig o , caspary f , rigaut g , rutz b , bouveret e , bragado - nilsson e , et al .
the tandem affinity purification ( tap ) method : a general procedure of protein complex purification .
kanehisa m , goto s. kegg : kyoto encyclopedia of genes and genomes .
nucleic acids res
. 2000;28(1):2730 .
huang h , jedynak bm , bader js . where have all the interactions gone ? estimating the coverage of two - hybrid protein interaction maps .
plos comput biol .
bader gd , hogue cwv . an automated method for finding molecular complexes in large protein interaction networks .
bmc bioinformatics .
newman mej .
modularity and community structure in networks .
proc natl acad sci usa .
tatusov rl , natale da , garkavtsev iv , tatusova ta , shankavaram ut , rao bs , et al .
the cog database : new developments in phylogenetic classification of proteins from complete genomes .
nucleic acids res
. 2001;29(1):228 .
jolliffe it . principal component analysis .
series : springer series in statistics .
springer ; 2002 .
487 : p. 28 .
park j , newman mej .
statistical mechanics of networks .
phys rev e 2004;70(6 ) : 066117 .
squartini t , garlaschelli d. analytical maximum - likelihood method to detect patterns in real networks .
new j phys .
butland g , peregrn - alvarez jm , li j , yang w , yang x , canadien v , et al .
interaction network containing conserved and essential protein complexes in escherichia coli .
fluitt a , pienaar e , viljoen h. ribosome kinetics and aa - trna competition determine rate and fidelity of peptide synthesis .
comput biol chem .
2007;31(5 - 6):335346 .
hric d , darst rk , fortunato s. community detection in networks : structural communities versus ground truth .
phys rev e. 2014;90(6):062805 .
boccaletti s , bianconi g , criado r , del genio ci , gmez - gardees j , romance m , et al .
the structure and dynamics of multilayer networks .
. 2014;544(1):1122 .
bulmer m. the selection - mutation - drift theory of synonymous codon usage .
xia x. dambe5 : a comprehensive software package for data analysis in molecular biology and evolution .
mol biol evol .
varenne s , buc j , lloubes r , lazdunski c. translation is a non - uniform process .
effect of trna availability on the rate of elongation of nascent polypeptide chains .
j mol biol .
srensen ma , kurland cg , pedersen s. codon usage determines translation rate in escherichia coli .
j mol biol .
ikemura t. codon usage and trna content in unicellular and multicellular organisms .
mol biol evol .
sabi r , tuller t. modelling the efficiency of codon - trna interactions based on codon usage bias .
sun x , yang q , xia x. an improved implementation of effective number of codons ( nc ) .
mol biol evol .
the density of a graph @xmath93 with @xmath58 nodes and @xmath94 links is the ratio between @xmath94 and the maximum number of possible links , namely @xmath95 , whereas , a @xmath29-core is a graph @xmath93 of minimal degree @xmath29 , meaning that each node belonging to @xmath93 has degree greater or equal than @xmath29 .
note that communities found by mcode can overlap , and some nodes can be excluded from all communities : there is no strict partition of the network , which would have been at least questionable since the same protein can be involved in different metabolic processes and thus belong to more than one community .
for these reasons and also because it considers both density and connectivity , mcode performs much better for our purposes than other community detection methods like modularity maximization @xcite that are based only on intra / inter - community densities and create a strict partition of the network .
here is the functional classification of cogs : l , replication , recombination and repair ; j , translation , including ribosome structure and biogenesis ; k , transcription ; d , cell division and chromosome partitioning ; o , molecular chaperons and related functions ; m , cell wall structure , biogenesis and outer membrane ; n , secretion , motility and chemotaxis ; p , inorganic ion transport and metabolism ; t , signal transduction ; c , energy production and conversion ;
g , carbohydrate metabolism and transport ; e , amino acid metabolism and transport ; f , nucleotide metabolism and transport ; h , coenzyme metabolism ; i , lipid metabolism ; r , general function prediction only ; s , no function prediction . | synonymous codons , _ i.e. _ , dna nucleotide triplets coding for the same amino acid , are used differently across the variety of living organisms .
the biological meaning of this phenomenon , known as _ codon usage bias _ , is still controversial . in order to shed light on this point
, we propose a new codon bias index , @xmath0 , that is based on the competition between cognate and near - cognate trnas during translation , without being tuned to the usage bias of highly expressed genes .
we perform a genome - wide evaluation of codon bias for _ e.coli_ , comparing @xmath0 with other widely used indices : @xmath1 , @xmath2 , and @xmath3 .
we show that @xmath0 and @xmath1 capture similar information by being positively correlated with gene conservation , measured by eri , and essentiality , whereas , @xmath2 and @xmath3 appear to be less sensitive to evolutionary - functional parameters .
notably , the rate of variation of @xmath1 and @xmath0 with eri allows to obtain sets of genes that consistently belong to specific clusters of orthologous genes ( cogs ) .
we also investigate the correlation of codon bias at the genomic level with the network features of protein - protein interactions in _
e.coli_. we find that the most densely connected communities of the network share a similar level of codon bias ( as measured by @xmath0 and @xmath1 ) .
conversely , a small difference in codon bias between two genes is , statistically , a prerequisite for the corresponding proteins to interact .
importantly , among all codon bias indices , @xmath0 turns out to have the most coherent distribution over the communities of the interactome , pointing to the significance of competition among cognate and near - cognate trnas for explaining codon usage adaptation . |
the present paper is motivated by a question of w. khnel @xcite : is it true that every @xmath1-connected graph @xmath2 contains a subdivision @xmath3 of the complete graph @xmath4 on four vertices such that @xmath5 is @xmath1-connected ?
an affirmative answer of a slightly stronger result taking care of possible chords in that subdivision , as provided in the present paper , can be employed to construct an embedding of an arbitrary @xmath1-connected graph into @xmath15 whose intersection with any open half space is connected @xcite .
problems on the existence of nonseparating or , more generally , high - connectivity - keeping subgraphs form a part of graph connectivity theory . of particular interest
are questions on the existence of high - connectivity keeping paths or cycles through prescribed elements or of prescribed parity .
for example , it has been conjectured by lovsz that for every @xmath16 there exists a smallest @xmath17 such that for any two vertices @xmath18 of an @xmath17-connected graph @xmath2 there exists an induced @xmath18-path @xmath19 such that @xmath20 is @xmath16-connected ( cited according to @xcite ) . presently , we know that @xmath21 and @xmath22 , but the existence of @xmath17 is open for all @xmath23 .
another example along these lines is thomassen s and toft s theorem that every @xmath24-connected graph contains a nonseparating induced cycle @xcite , and thomassen s theorem that every @xmath25-connected graph @xmath2 contains an induced cycle @xmath26 such that @xmath27 is @xmath16-connected ( where the bound @xmath28 is sharp for all @xmath29 ) @xcite .
there are variations of these results where the objects might have chords , i. e. are not necessarily induced , and also variations and solutions where the edges are deleted instead of the vertices ( see , for example , ( * ? ? ? * conjecture 4 ) or @xcite ) . by considering paths and cycles as subdivisions of @xmath30 and @xmath6 , respectively
, one might ask if there are similar results for more complex substructures ( like for subdivisions or induced subdivisions of @xmath4 ) . the general , qualitative answer to this
is `` yes '' : [ t1 ] suppose that @xmath31 is a class of graphs such that there is an @xmath32 such that every @xmath32-connected graph contains a member from @xmath31 as a subgraph .
then , for every @xmath16 , there exists an @xmath33 such that every @xmath33-connected graph @xmath2 contains a member @xmath3 of @xmath31 such that @xmath5 ( alternatively : @xmath34 ) is @xmath16-connected .
this follows immediately from theorem 1 in @xcite that the vertex set of every @xmath35-connected graph @xmath2 admits a partition into two sets @xmath36 such that @xmath37,g[b]$ ] are @xmath16-connected and every vertex in @xmath38 has at least @xmath16 neighbors in @xmath39 ; any copy of @xmath40 in @xmath37 $ ] will do ( cf .
@xcite ) . the possibly easiest question beyond paths and cycles is to ask for the smallest @xmath16 such that every @xmath16-connected graph contains a nonseparating subdivision of @xmath4 .
brute force application of theorem [ t1 ] leads to the fact that @xmath16 exists and is at most 55.296 ( quite large on the other , hand this yields a subdivision @xmath3 of @xmath4 such that @xmath5 is even @xmath24-connected ) . a better bound
can be obtained by the following recent generalization of thomassen s theorem on high - connectivity keeping cycles by fujita and kawarabayashi @xcite ; a _ theta _ is a graph isomorphic to a subdivision of @xmath41 , where @xmath41 is obtained from @xmath4 by deleting an edge .
@xcite [ t2 ] for @xmath29 , every @xmath25-connected graph @xmath2 contains either @xmath41 as a subgraph , or it contains an induced theta @xmath42 such that @xmath43 is @xmath16-connected . from this
it follows that every @xmath44-connected graph @xmath2 has a nonseparating subdivision of @xmath4 : if @xmath2 contains @xmath4 as a subgraph then we are done , otherwise we first find an induced theta @xmath42 ( that might be a @xmath41 ) such that @xmath43 is @xmath24-connected by theorem [ t2 ] we then take neighbors @xmath18 in @xmath45 of two inner vertices of distinct subdivision paths . by what we have mentioned above ( `` @xmath21 '' ) ,
there is an @xmath18-path @xmath19 in the @xmath24-connected graph @xmath43 such that @xmath46 is connected , so that @xmath47 $ ] is a nonseparating subgraph of @xmath2 , which contains the desired subdivision of @xmath4 as a spanning subgraph .
similarly , we might use our knowledge on @xmath48 to prove that every @xmath49-connected graph @xmath2 has a subdivision @xmath3 of @xmath4 such that @xmath5 is @xmath50-connected . here
we prove first that every @xmath1-connected graph has a nonseparating subdivision of @xmath4 , where , in addition , we can avoid an arbitrary prescribed vertex ( see corollary [ c1 ] below ) .
this is sharp in the sense that there are infinitely many @xmath24-connected graphs where every @xmath4-subdivision separates ; take @xmath51 for @xmath52 .
the result relies on a well - known structural characterization of the class of @xmath1-connected graphs stating that every @xmath1-connected graph can be reduced to a smaller one by contracting a single edge and simplifying , unless it is isomorphic to the square of a cycle or the line graph of a cubic graph ( theorem [ t3 ] below , see @xcite and @xcite ) .
the main technicalities to overcome here are * to construct appropriate nonseparating subdivisions in @xmath2 from appropriate subdivisions in the graph @xmath53 obtained from @xmath2 by contracting a single edge ( for the induction step ) , and * to find a nonseparating induced theta in certain cubic simple graphs ( to let the induction start ) .
it turned out , however , that these two steps do _ not _ rely on the @xmath1-connectedness of the graphs in question , and that ( a ) depends only on degree constraints to both @xmath2 and @xmath53 .
so , apart from the application of fontet s and martinov s theorem on @xmath1-connected graphs , our proof `` works '' for connected graphs of minimum degree at least @xmath1 . to really let it work in this case , we have designed a _ structure theorem for the simple connected graphs of minimum degree at least @xmath1 _ : we can reduce such a graph to a smaller such graph by deleting or contracting a single and simplifying , unless it is the square of a cycle or the edge disjoint union of certain bricks according to a pattern described by a connected hypergraph of rank at most @xmath24 as in theorem [ t5 ] .
whereas in the @xmath1-connected case above , these bricks are just triangles , we need a number of isomorphism types of bricks for this , as depicted in figure [ f1 ] below .
an alternative version of this in terms of decomposition into edge - disjoint subgraphs can be found in the abstract .
we begin with a lemma on cubic graphs , which is the starting point of the inductive proofs of theorem [ t4 ] and theorem [ t6 ] .
[ l1 ] let @xmath0 be a vertex of a simple connected cubic graph @xmath2 nonisomorphic to @xmath4 .
then @xmath2 has a nonseparating induced subgraph isomorphic to a subdivision of @xmath41 avoiding @xmath0 .
* we are looking for an induced nonseparating theta @xmath42 in @xmath2 avoiding @xmath0 .
we proceed by induction on @xmath54 .
if @xmath55 then @xmath2 must be a prism @xmath56 or @xmath57 .
the one - vertex - deleted subgraphs of these are subdivisions of @xmath41 , so that the induction starts .
suppose that @xmath58 .
suppose first that @xmath0 is contained in a subgraph @xmath59 isomorphic to @xmath41 .
since @xmath2 is connected and @xmath60 , @xmath59 is an induced subgraph of @xmath2 , hence @xmath61 consists of either one or two vertices .
reduction if @xmath0 is contained in some @xmath41.,width=196 ]
if @xmath62 then @xmath63 has exactly one neighbor @xmath64 not in @xmath65 , and @xmath66 is a bridge . let @xmath67 be the components of @xmath68 , where @xmath69 , and @xmath70 and @xmath71 . let @xmath3 be the cubic simple graph obtained from @xmath72 by adding a single edge @xmath73 connecting @xmath74 and adding a single edge @xmath75 connecting @xmath76 ( see upper half of figure [ f2 ] ) .
since @xmath69 is connected , @xmath3 is connected , and we may apply induction to @xmath77 for @xmath78 to obtain a nonseparating induced theta @xmath42 avoiding @xmath79 in @xmath3 . as
@xmath42 does not contain @xmath79 , it does not contain @xmath73 and hence neither @xmath75 , ( as @xmath80 form a cut in @xmath3 but @xmath42 is @xmath50-connected ) , implying that @xmath42 is and induced subgraph of either @xmath81 or @xmath82 . since @xmath81 contains only @xmath24 vertices
, @xmath42 is an induced subgraph of @xmath82 .
it is easy to see that @xmath42 is a nonseparating induced theta in @xmath2 avoiding @xmath65 and hence avoiding @xmath0 .
if @xmath83 for distinct _ nonadjacent _ vertices @xmath18 , then let @xmath3 be the simple connected cubic graph @xmath3 obtained from @xmath84 by adding a single edge from @xmath63 to @xmath64 .
if @xmath3 is nonisomorphic to @xmath4 then we find a nonseparating induced theta in @xmath3 avoiding @xmath63 by induction , and it will be the desired nonseparating induced theta in @xmath2 avoiding @xmath0 . if @xmath3 is isomorphic to @xmath4 then we take @xmath85 .
if @xmath83 for distinct _ adjacent _ vertices @xmath18 , then let @xmath86 and @xmath87 , where @xmath88 are in @xmath65 and @xmath89 are not . it might be the case that @xmath90 , but the graph @xmath3 obtained from @xmath72 by adding a single edge connecting @xmath74 and adding a single edge connecting @xmath76 is simple , cubic , and connected ( see lower half of figure [ f2 ] ) . by applying induction to @xmath77 we find a nonseparating induced theta in @xmath3 avoiding @xmath79 , which is at the same time a nonseparating induced theta in @xmath2 avoiding @xmath65 , as above .
hence we may assume that @xmath0 is not contained in a subgraph isomorphic to @xmath41 .
if @xmath0 is on a triangle @xmath91 then @xmath92 consists of three independent edges , so that the graph @xmath3 obtained from @xmath2 by contracting @xmath91 to a single vertex @xmath93 is simple , connected , and cubic . since @xmath94
we find a nonseparating induced theta in @xmath3 avoiding @xmath93 by induction , which is at the same time a nonseparating induced theta in @xmath2 avoiding @xmath95 .
hence we may assume that @xmath0 is not on a triangle .
let @xmath96 and @xmath97 .
observe that @xmath98 are distinct as @xmath0 is not on a triangle .
if @xmath89 are not adjacent then the graph @xmath3 obtained from @xmath99 by adding a single edge connecting @xmath100 and adding a single edge connecting @xmath89 is simple ( and cubic , but not necessarily connected ) .
if the component @xmath26 of @xmath3 containing @xmath100 is isomorphic to @xmath4 then @xmath101 induces a nonseparating theta in @xmath2 avoiding @xmath0 .
otherwise , we apply induction to @xmath102 to find a nonseparating induced theta in @xmath26 avoiding @xmath64 , which will serve for @xmath2 , too
. hence we may assume that @xmath103 induce a triangle @xmath91 in @xmath2 .
if @xmath92 consists of three independent edges then we find a nonseparating induced theta @xmath42 avoiding @xmath0 in the graph @xmath3 obtained from @xmath2 by contracting @xmath91 to a single vertex @xmath93 .
if @xmath42 does not contain @xmath93 then it obviously serves for @xmath2 , otherwise @xmath104 $ ] will do it . in the remaining case , there exists a common neighbor @xmath105 of @xmath89 distinct from and nonadjacent to @xmath63 , so that @xmath106 $ ] is isomorphic to @xmath41 . if @xmath59 is nonseparating then we are done , otherwise @xmath107 for some @xmath79 distinct from and nonadjacent to @xmath0 . the graph @xmath3 obtained from @xmath84 by adding a single edge connecting @xmath108 is simple , cubic , connected , and nonisomorphic to @xmath4 ( because @xmath109 is a bridge there )
hence the statement follows again by induction applied to @xmath110 .
@xmath111 observe that , in lemma [ l1 ] we can not achieve that one of the subdivision paths of @xmath42 has length @xmath112 , because there are ( cubic ) graphs without cycles with a unique chord as an induced subgraph .
the petersen graph and the heawood graph are basic examples , a full characterization can be derived from the main result in @xcite .
in this section , we prove lemma [ l2 ] below , which will be employed as an inductive device to obtain nonseparating induced subdivisions in some graph @xmath2 from those in a minor of @xmath2 obtained by contracting a single edge and simplifying . there is a version in terms of _ deleting _ edges , which comes with a slightly less technical proof , but as we do not need it for our main result on @xmath1-connected graphs ( theorem [ t4 ] ) , we postpone it to a later section . the attribute _ principal _ refers to objects of a subdivision @xmath3 of some graph @xmath19 which correspond to those of @xmath19 .
so the _ principal vertices _ of @xmath3 are those of degree at least @xmath24 in @xmath3 if @xmath113 , principal vertices are _ principally adjacent _ in @xmath3 if they are adjacent in @xmath19 , and so on .
moreover , we use path notation to describe certain basic subgraphs of the subdivision @xmath3 as follows : provided that @xmath19 is simple ( as it will be always the case ) , two vertices @xmath114 are on at most one common subdivision path .
if there is such a path , say @xmath115 , then we denote the unique @xmath18-subpath of @xmath115 by @xmath116 $ ] . in this case , let us define , moreover , @xmath117-b$ ] , @xmath118:=h[a , b]-a$ ] , and @xmath119 .
we call the isomorphism class of @xmath19 the _ type _ of our subdivision @xmath3 . hence two subdivisions @xmath120 of two simple graphs @xmath121 , respectively , of minimum degree at least @xmath24 have the same type if and
only if @xmath122 . recall that the _ wheel _
@xmath123 is obtained from a cycle @xmath26 of length @xmath32 by adding a new vertex @xmath0 and a new edge connecting @xmath0 to @xmath79 for each @xmath124 .
the vertices in @xmath101 are called the _ rim vertices _ of the wheel , @xmath0 is called its _ center_. [ l2 ] suppose that @xmath53 is a graph of minimum degree at least @xmath1 obtained from a simple graph @xmath2 of minimum degree at least @xmath1 by contracting an edge @xmath109 to a single vertex @xmath93 and then simplifying , and that @xmath125 is a nonseparating induced subgraph in @xmath53 isomorphic to a subdivision of a wheel or of a prism @xmath56 or of @xmath57 .
then there exists a nonseparating induced subgraph @xmath3 in @xmath2 isomorphic to a subdivision of a wheel or of a prism @xmath56 or of @xmath57 such that @xmath126 if @xmath127 and @xmath128 otherwise . * proof . * if @xmath127 then obviously @xmath129 is as desired .
so suppose that @xmath130 .
we then consider different cases according to the degree of @xmath93 in @xmath125 and the type of @xmath125 ( subdivision of a prism / @xmath57 / @xmath4 / larger wheel ) , and also on the different types of adjacency in @xmath2 of the neighbors of @xmath93 in @xmath53 : all edges not incident with @xmath93 in @xmath53 are single edges in @xmath2 , but whenever there is an edge @xmath131 in @xmath53 then the subgraph induced by @xmath132 in @xmath2 is formed by either the edge @xmath109 plus one or two edges connecting @xmath133 to @xmath0 or @xmath79 or both ; accordingly , we say that @xmath131 is an _
@xmath0- , @xmath79- , or @xmath109-edge _ and that @xmath133 is an _
@xmath0- , @xmath79- , or @xmath109-vertex _ , respectively .
the strategy is as follows : in each case , we try to identify an appropriate @xmath3 as an induced subgraph of @xmath134 $ ] , where @xmath135 as @xmath136 is connected by assumption , it suffices to find at least one edge connecting every component of @xmath137 $ ] to @xmath138 in order to prove that @xmath3 is nonseparating .
this is where the degree conditions to @xmath2 and @xmath53 are used ; in most cases , we will leave the easy argument to the reader . * case 1 .
* @xmath93 has at most three neighbors in @xmath125 and among them there is no @xmath0-vertex or no @xmath79-vertex or no @xmath109-vertex .
this case does not depend on the actual type of @xmath125 , @xmath3 will be of the same type .
let @xmath139 denote the neighbors of @xmath93 in @xmath125 .
if all of @xmath139 are @xmath109-vertices then there is a @xmath140 having a neighbor in @xmath138 since @xmath141 , and we take @xmath142 $ ] . hence we may assume that at least one vertex of @xmath139 is not an @xmath109-vertex .
if there is no @xmath79-vertex in @xmath139 then one vertex of @xmath139 is an @xmath0-vertex , so that @xmath79 has a neighbor in @xmath138 , and we may take @xmath143 $ ] .
similarly if there is no @xmath79-vertex in @xmath139 then we may take @xmath144 $ ] . in the remaining subcase ,
all of @xmath139 are @xmath0- or @xmath79-vertices , and both types do occur ; but then we may take @xmath145 $ ] . @xmath146
* @xmath93 has at most three neighbors .
how to find @xmath3 in @xmath134 $ ] in the proof of lemma 2 .
the picture shows the entire subgraph induced by @xmath147 , where the fat lines resemble single edges and all other lines are subdivision paths ; most of these paths connect @xmath108 or the principal vertices of @xmath125 , displayed in white .
the solid vertices resemble neighbors of @xmath93 in @xmath125 ; some of them might coincide with principal vertices of @xmath125 distinct from @xmath93 , i. e. the thin line which actually connects them to a white vertex might represent a path of length @xmath148.,width=377 ] we may assume that @xmath93 has exactly three neighbors @xmath149 in @xmath125 and that @xmath149 is an @xmath0- , @xmath79- , @xmath109-vertex , respectively , for otherwise we are in case 1 .
we first look at the case that @xmath125 is a subdivision of @xmath4 , with principal vertices @xmath150 . by symmetry
, we may assume that @xmath149 are on @xmath151 $ ] , @xmath152 $ ] , @xmath153 $ ] , respectively .
the situation is illustrated in figure [ f3 ] , top left .
if @xmath154 then @xmath145 $ ] is an induced subdivision of a @xmath1-wheel , otherwise @xmath145 $ ] is a subdivision of the prism .
next look at the case that @xmath125 is a subdivision of @xmath57 , where @xmath155 and @xmath156 form the two classes of pairwise principally nonadjacent principal vertices ( `` principal color classes '' ) . by symmetry
, we may assume that @xmath149 are on @xmath151 $ ] , @xmath152 $ ] , @xmath153 $ ] , respectively ( see figure [ f3 ] , top right ) .
but then @xmath157 $ ] is a nonseparating subdivision of @xmath4 in @xmath2 .
if @xmath125 is a subdivision of the prism then the situation is slightly less symmetric : let @xmath156 be the principal neighbors of @xmath93 in @xmath125 , where we may assume that @xmath89 are principally adjacent , and let @xmath158 and @xmath159 denote the unique principal neighbor not in @xmath160 of @xmath161 and @xmath162 , respectively .
the two options for @xmath163 are displayed in the middle row of figure [ f3 ] : if @xmath163 is on @xmath153 $ ] then we may assume by symmetry that @xmath18 are on @xmath151 $ ] , @xmath152 $ ] , respectively ; in this subcase , @xmath164 $ ] is a nonseparating induced subdivision of @xmath4 in @xmath2 ( observe that @xmath158 has a neighbor in @xmath138 ) . if , otherwise , @xmath163 is not on @xmath153 $ ] then we may assume by symmetry that @xmath163 is on @xmath151 $ ] , and the same conclusion holds ( no matter if @xmath63 is on @xmath152 $ ] and @xmath64 on @xmath153 $ ] or vice versa ) .
finally , suppose that @xmath125 is a subdivision of a wheel with center @xmath161 , and rim vertices @xmath165 ( in this order ) , where @xmath166 and @xmath167 .
one possible case is illustrated in figure [ f3 ] , bottom left : if @xmath163 is not on @xmath151 $ ] then , by symmetry , we may assume that @xmath163 is on @xmath168 $ ] , so that @xmath169 ) ) ] $ ] is a nonseparating induced subdivision of @xmath4 in @xmath2 .
the same conclusion holds if @xmath163 is on @xmath151 $ ] but not equal to @xmath161 .
if @xmath170 then @xmath145 $ ] is a subdivision of a wheel with @xmath171 rim vertices .
so suppose that we are not in case 1 or case 2 .
then @xmath125 is a wheel with center @xmath93 and rim vertices @xmath172 , where @xmath166 ( in this order on the rim ) .
let us denote by @xmath173 the neighbors of @xmath93 in @xmath125 , where @xmath174 is on @xmath175 $ ] for @xmath176 and is either an @xmath0- , @xmath79- , or @xmath109-vertex .
for @xmath177 and @xmath178 , let @xmath179 ) ) \ ] ] ( indices modulo @xmath32 ) be the set of vertices in @xmath2 corresponding to the principal vertices @xmath180 of @xmath125 and to the vertices of all subdivision paths in between .
observe that @xmath181 $ ] is a nonseparating subgraph of @xmath2 , because @xmath182 has a neighbor in @xmath138 and is connected to all vertices in @xmath183 $ ] by a path .
* at least two @xmath174 are @xmath109-vertices .
we then may assume by symmetry that @xmath184 and @xmath185 for some @xmath186 are @xmath109-vertices , and that for all @xmath187 with @xmath188 , @xmath174 is either an @xmath0- or a @xmath79-vertex .
if @xmath189 then @xmath190 $ ] is a subdivision of @xmath4 .
so we may assume that @xmath191 . if there is a @xmath140 such that all @xmath174 with @xmath188 are @xmath192-vertices then @xmath193 $ ] is a subdivision of a wheel with @xmath194 rim vertices and center @xmath192 .
this is illustrated in figure [ f3 ] , bottom right , with @xmath195 .
hence we may assume that there are both @xmath0- and @xmath79-vertices among @xmath196 . by symmetry
, we may assume that there is an @xmath197 with @xmath198 such that @xmath199 is an @xmath0-vertex and @xmath200 are @xmath79-vertices .
but then @xmath201 $ ] is a subdivision of a wheel with @xmath202 rim vertices and center @xmath79 . @xmath146 * case 4 .
* there is exactly one @xmath109-vertex .
we then may assume by symmetry that @xmath184 is the unique @xmath109-vertex among the @xmath174 and that there are at least two @xmath0-vertices among the @xmath174 .
if there is no @xmath79-vertex then @xmath143 $ ] is a nonseparating subdivision of a wheel with @xmath32 rim vertices and center @xmath0 in @xmath2 , and if there is exactly one @xmath79-vertex @xmath174 then @xmath203 $ ] is a nonseparating subdivision of a wheel with @xmath204 rim vertices and center @xmath0 , because in these cases @xmath205}(y)<4 $ ] .
hence there are at least two @xmath79-vertices , too . by symmetry
, we may assume that there exists an @xmath197 with @xmath206 such that @xmath199 is an @xmath0-vertex and @xmath200 are @xmath79-vertices , so that , again , @xmath201 $ ] is a subdivision of a wheel with @xmath202 rim vertices and center @xmath79 .
@xmath146 hence we are left with the case that all @xmath174 are @xmath0- or @xmath79-vertices .
suppose first that , for some @xmath207 , there exist consecutive @xmath192-vertices @xmath208 .
then there exist @xmath209 and @xmath210 such that @xmath208 , @xmath211 , @xmath212 are @xmath192-vertices and either none of @xmath213 is a @xmath192-vertex or @xmath214 , where all indices are taken modulo @xmath32 .
( that is , we take a `` maximal nontrivial circular subsequence '' of consecutive @xmath192-vertices . )
we choose @xmath215 such that , under these constraints , @xmath16 is minimum . without loss of generality
we may assume that @xmath216 and @xmath217 . if @xmath214 then all @xmath174 are @xmath0-vertices , and @xmath218 $ ] is a nonseparating subdivision of a wheel with @xmath32 rim vertices and center @xmath0 in @xmath2 .
if @xmath219 then @xmath134 $ ] is a nonseparating subdivision of a wheel with @xmath32 rim vertices and center @xmath0 in @xmath2 .
if @xmath220 then @xmath221 are consecutive @xmath79-vertices and none of @xmath222 is a @xmath79-vertex ; hence @xmath223 by minimality of @xmath16 , and @xmath134 $ ] is a subdivision of the prism @xmath56 in @xmath2 . finally , if @xmath224 then @xmath225 $ ] is a subdivision of a wheel with rim vertices @xmath226 and center @xmath0 .
hence we may assume that there are neither consecutive @xmath0-vertices @xmath208 nor consecutive @xmath79-vertices .
if @xmath227 then @xmath134 $ ] is a subdivision of @xmath57 , and otherwise @xmath228 $ ] is a subdivision of @xmath4 , with principal vertices @xmath229 .
@xmath111 observe that our proof almost produces the statement obtained from lemma [ l2 ] without the option of an induced @xmath57 : if @xmath125 itself is not a subdivision of @xmath57 then we might be forced to take @xmath3 as a subdivision of @xmath57 only in the very last paragraph of the proof , where we are in the special case that @xmath125 is a subdivision of a wheel of length @xmath1 , where the principal central vertex @xmath93 has exactly two @xmath0-neighbors in @xmath125 , which are principally nonadjacent , and exactly two @xmath79-neighbors ( which are then principally nonadjacent , too ) .
in this section we prove our main result for @xmath1-connected graphs ( theorem [ t4 ] ) . for the proof ,
we combine the previous technical lemmas ( which do _ not _ rely on @xmath1-connectedness ) with a fundamental result on @xmath1-connected graphs by fontet @xcite and martinov @xcite ( cf .
@xcite ) .
@xcite @xcite [ t3 ] every @xmath1-connected graph @xmath2 nonisomorphic to the square of a cycle and nonisomorphic to a line graph of a cubic graph has an edge @xmath73 such that the graph obtained from @xmath2 by contracting @xmath73 to a single vertex ( and then simplifying ) is @xmath1-connected .
it is clear that not every cubic graph @xmath230 yields a @xmath1-connected line graph @xmath231 ; in fact , @xmath231 is @xmath1-connected if and only if @xmath230 is simple , triangle free , @xmath24-edge - connected , and the cuts consisting of @xmath24 edges are precisely the edge - neighborhoods of the single vertices .
[ t4 ] let @xmath0 be a vertex of a simple @xmath1-connected graph @xmath2
. then @xmath2 contains an induced subgraph @xmath3 isomorphic to a subdivision of a wheel or of a prism or of @xmath57 such that @xmath5 is connected and contains @xmath0 .
* induction on @xmath232 . a nonseparating induced subdivision of @xmath233 in @xmath234 , and a nonseparating induced subdivision of @xmath235 in @xmath236.,width=415 ]
suppose first that @xmath2 is the square of a cycle @xmath237 of length @xmath52 . if @xmath32 is even then @xmath238 induce a nonseparating subdivision of a wheel of lenght @xmath1 , and if @xmath32 is odd then @xmath239 induce a nonseparating subdivision of @xmath4 .
( figure [ f4 ] illustrates this for @xmath240 and @xmath241 . ) in both cases , we get a nonseparating induced subgraph avoiding at least one vertex , and since the squares of cycles are vertex transitive , we can achieve that they avoid _ any _ prescribed vertex .
suppose now that @xmath2 is the line graph of a cubic graph @xmath230 .
@xmath230 is simple , and if @xmath242 then @xmath2 is the square of a @xmath243-cycle , which we have already considered .
hence we may apply lemma [ l1 ] to @xmath230 and any vertex @xmath93 of @xmath230 such that the given vertex @xmath0 is on the triangle formed by @xmath244 in @xmath231 : this yields an induced nonseparating theta @xmath42 in @xmath230 avoiding @xmath93 .
if the two degree-@xmath24-vertices of @xmath42 are connected by an edge then @xmath245 induces a nonseparating subdivision of a @xmath1-wheel in @xmath246 , and in the other case @xmath245 induces a nonseparating subdivision of a prism ; in both cases we avoid @xmath0 . in the remaining case
, there exists an edge @xmath73 such that the graph @xmath53 obtained from @xmath2 by contracting @xmath73 to a single vertex @xmath93 and simplifying is @xmath1-connected
. let @xmath247 if @xmath248 and let @xmath249 otherwise . by induction , we find a subdivision in @xmath53 as in the statement , avoiding @xmath250 , so that the statement follows for @xmath78 by lemma [ l2 ] .
@xmath111 it is obvious that we can not replace the list of graphs ( wheels / prism / @xmath57 ) in theorem [ t4 ] by a proper sublist : for @xmath3 from the list , consider the graph @xmath2 obtained from @xmath3 by adding a new vertex @xmath0 and adding an edge from @xmath0 to each @xmath251 ; then the only induced subdivision of a graph from the list in @xmath2 which avoids @xmath0 is @xmath3 itself . the situation might change if we do not insist on prescribing a vertex ; so let us try to find @xmath1-connected graphs with just one type of nonseparating induced subdivision from the list .
suppose that @xmath230 is a cubic graph such that @xmath231 is @xmath1-connected .
it is easy to see that @xmath231 can not contain induced subdivisions of @xmath123 with @xmath252 or of @xmath57 .
if @xmath230 has no cycles with a unique chord , as it is the case for the petersen graph and the heawood graph , then @xmath231 does not contain nonseparating induced subdivisions of @xmath233 ; this shows that we really need the prism in our list .
now suppose that @xmath230 is planar ( i. e. it is the geometric dual of an arbitrary spherical triangulation ) , fix an embedding of @xmath230 in the plane and suppose , to the contrary , that @xmath231 has a nonseparating induced subdivision @xmath3 of the prism .
it is easy to see that this corresponds to an induced theta @xmath42 in @xmath230 without a subdivision path of length @xmath112 , i. e. @xmath253 . for each subdivision path @xmath19 of @xmath42 , choose an edge incident with an inner vertex of @xmath19 but not in @xmath254 .
then two among these three edges are embedded in distinct regions of @xmath255 , so that there is no path in @xmath256 connecting them .
it follows that @xmath257 is disconnected , contradiction .
this shows that we need @xmath233 in our list .
next look at @xmath258 , @xmath259 , with color classes @xmath36 , where the vertices in @xmath38 have degree @xmath1 and those in @xmath39 have degree @xmath159 .
suppose that there exists an induced subgraph @xmath3 isomorphic to a subdivision of a graph in our list . since @xmath3 has a vertex of degree @xmath24 , we know that @xmath260 or @xmath261 as @xmath3 is an induced subgraph of @xmath2 . if @xmath3 had nonprincipal vertices then , analogously , @xmath262 or @xmath263 , so that @xmath264 , which is not possible
. hence every vertex of @xmath3 is a principal vertex , which implies @xmath265 .
hence we can not omit @xmath57 in our list .
a similar argument shows that @xmath266 with @xmath267 does not admit a nonseparating induced subgraph isomorphic to a subdivision of _ any _ graph in the list and shows that the connectivity bound in theorem [ t4 ] is sharp .
the icosahedron shows that we can not remove @xmath268 from our list ( this needs a little more work ) .
i do not know if we can omit the wheels @xmath269 with @xmath270 : [ p1 ] is it true that every @xmath1-connected graph contains a nonseparating induced subgraph isomorphic to a subdivision of @xmath4 , @xmath233 , @xmath268 , @xmath56 , or @xmath57 ? however , as all graphs listed in theorem [ t4 ] have a spanning subdivision of @xmath4 , we get the following corollary : [ c1 ] let @xmath0 be a vertex of a @xmath1-connected graph @xmath2 .
then @xmath2 contains a subdivision @xmath3 of @xmath4 as a subgraph such that @xmath5 is connected and contains @xmath0 .
in particular , every @xmath1-connected graph admits a nonseparating subdivision of @xmath4 , that is , we get an affirmative answer to khnel s question mentioned in the introduction .
we will develop a structure theorem for the class @xmath271 of simple connected graphs of minimum degree at least @xmath1 . throughout this section ,
let us call an edge @xmath73 of a graph @xmath2 in @xmath271 _ essential _ if the graph @xmath272 obtained from @xmath2 by deleting @xmath73 is not in @xmath271 , and let us call @xmath73 _ critical _ if the graph @xmath273 obtained from @xmath2 by contracting @xmath73 and simplifying is not in @xmath271 .
it is obvious that @xmath73 is essential if and only if at least one of its endvertices has degree @xmath1 or @xmath73 is a bridge , and that @xmath73 is critical if and only if the endvertices of @xmath73 have a common neighbor of degree @xmath1 or @xmath274 consists of three common neighbors of the endvertices of @xmath73 .
we are interested in the _ minimal critical _ graphs in @xmath271 , i. e. graphs where every edge is both essential and critical .
these graphs are always bridgeless , because bridges are not critical . the nine bricks . vertices of attachment are displayed solid.,width=377 ] for the description , we use a number of building blocks , as depicted in figure [ f1 ] .
suppose that @xmath3 is a graph whose automorphism group acts transitively on vertices , edges , sets of two independent edges whose endvertices have a common neighbor .
the complete graph @xmath7 on five vertices and the octahedron @xmath8 are such graphs .
we define @xmath275 and @xmath276 as the graphs obtained from @xmath3 deleting a single edge or deleting three edges forming a triangle , respectively , and @xmath277 as the graph obtained from @xmath2 by taking @xmath278 and two independent edges @xmath80 with @xmath279 , deleting @xmath0 from @xmath2 , adding two new vertices @xmath88 , and adding two new edges from @xmath79 to @xmath280 and two new edges from @xmath192 to @xmath281 .
a _ brick _ is a graph isomorphic to one of @xmath7 , @xmath8 , @xmath9 , @xmath10 , @xmath11 , @xmath12 , @xmath13 , @xmath14 , or @xmath6 .
each brick comes together with its _ vertices of attachment _ : for @xmath7 and @xmath8 , this is an arbitrary single vertex , for the other seven bricks these are its vertices of degree less than @xmath1 .
the remaining vertices of the respective brick are its _ inner vertices _ , and the edges connecting two inner vertices are called its _ inner edges_. observe that every brick @xmath39 has @xmath112 , @xmath50 , or @xmath24 vertices of attachment , and that they are pairwise nonadjacent unless @xmath39 is _ triangular _ , i. e. isomorphic to @xmath6 .
it turns out that the minimal critical graphs from @xmath271 are either squares of cycles of length at least @xmath282 , or they are the edge disjoint union of bricks , following certain rules .
this is made precise in the following theorem .
[ t5 ] a graph @xmath2 is a minimal critical graph in @xmath271 if and only if it is the square of a cycle of length at least @xmath282 or arises from a nonempty connected multihypergraph @xmath3 of minimum degree at least @xmath50 and @xmath283 for all hyperedges @xmath73 by replacing each hyperedge @xmath73 by a brick @xmath284 ( see figure [ f1 ] ) such that the vertices of attachment of @xmath284 are those in @xmath280 and at the same time the only objects of @xmath284 contained in more than one brick , and * @xmath284 is triangular only if every vertex @xmath0 of @xmath73 is incident with precisely one hyperedge @xmath285 , where @xmath286 is none of @xmath287 , and for distinct @xmath288 from @xmath280 we have @xmath289 only if not both of @xmath290 are triangular and @xmath291 only if @xmath292 is @xmath13 or @xmath14 .
* proof . * squares of cycles of length at least @xmath282 are minimal critical graphs in @xmath271 ; the graphs @xmath2 obtained from a hypergraph @xmath3 as in the statement are in @xmath271 , and , moreover , every edge of a non - triangular brick @xmath284 except for inner edges of @xmath293 in such a graph is incident with a degree-@xmath1-vertex of @xmath2 in @xmath294 , and its endvertices have a common neighbor of degree @xmath1 in @xmath294 ; condition ( tb ) ensures this for the edges of triangular blocks , whereas inner edges of @xmath293 are both essential and critical by definition
. it thus suffices to prove that _ every _ minimal critical graph in @xmath271 has this shape .
so suppose that @xmath2 is a minimal critical graph in @xmath271 .
the proof runs as follows : we construct a hypergraph @xmath3 on a subset of @xmath54 by subsequently defining ( _ collecting _ ) hyperedges . as a book - keeping device ,
each of these hyperedges @xmath73 receives a label @xmath284 , which is the actual subgraph of @xmath2 it represents : @xmath284 will allways be a brick , and its vertices of attachment will be those of @xmath280 and at the same time the only objects of @xmath284 that may occur in labels of previously collected hyperedges .
( in most cases , we will leave the easy checking to the reader . ) ( tb ) will be satisfied because we will ensure that the the vertices of triangular bricks have degree @xmath1 in @xmath2 and its edges are contained in only one triangle in @xmath2 , and that vertices of attachment of any brick are nonadjacent in @xmath2 unless the brick is isomorphic to one of @xmath295 .
if we manage to cover all edges of @xmath2 by these @xmath284 then the statement follows .
* step i. * let us first look at the set @xmath296 of edges which are contained in no subgraph of @xmath2 isomorphic to @xmath41 ( induced or non - induced ) .
observe that if @xmath73 is an edge in @xmath296 then it forms a triangle in @xmath2 together with two further edges @xmath297 , because @xmath73 is critical . since @xmath73 is not on @xmath41 , each of @xmath298 is on exactly one triangle in @xmath2 , and
hence @xmath297 are in @xmath296 , too .
hence the subgraph @xmath299 is the edge disjoint union of its triangles .
if @xmath0 is any vertex from @xmath300 , then there is a triangle @xmath91 in @xmath301 containing @xmath0 , and @xmath0 is the only common neighbor of the two remaining vertices in @xmath302 by construction , so that @xmath0 has degree @xmath1 in @xmath2 .
therefore , each vertex has either degree @xmath1 or degree @xmath50 in @xmath301 , and the degree-@xmath1-vertices of @xmath301 are not incident with any edge outside @xmath296 . for each triangle @xmath91 of @xmath301
, we collect a @xmath24-hyperedge labelled @xmath91 whose vertices are those of @xmath95 .
* step ii .
* now we look at the remaining edges , not contained in @xmath296 .
as long as there is an edge @xmath109 on _ more _ than two triangles not in a label of a previously collected hyperedge , we process it as follows .
since one of @xmath108 has degree @xmath1 , say , @xmath0 , it follows that @xmath108 have exactly three common neighbors @xmath149 .
if @xmath79 has degree exceeding @xmath1 in @xmath2 then @xmath149 have degree @xmath1 as @xmath303 are essential .
moreover , since @xmath304 are critical , each of @xmath149 must have a neighbor among @xmath149 .
without loss of symmetry , we may assume that @xmath100 both are adjacent to @xmath63 .
but then @xmath100 are adjacent , for if the neighbor @xmath93 of @xmath64 distinct from @xmath305 was not equal to @xmath163 then the edge @xmath306 would not be critical .
hence , in this case , we see that @xmath307 \cong k_5 $ ] forms an endblock of @xmath2 , with cutvertex @xmath79 , and we collect a @xmath112-hyperedge labelled @xmath39 whose vertex is @xmath79 .
let us now assume that both @xmath108 have degree @xmath1 , so that @xmath308 .
suppose that two of them , say @xmath18 , were adjacent .
if one of @xmath18 had degree exceeding @xmath1 , say @xmath64 , then @xmath63 has degree @xmath1 and the neighbor @xmath309 of @xmath63 distinct from @xmath310 must be @xmath163 as @xmath311 is critical ; if @xmath100 are nonadjacent then @xmath312 \cong
k_5 ^ -$ ] , and we collect a @xmath50-hyperedge labelled @xmath39 with vertices @xmath100 . otherwise , if @xmath100 are adjacent , then we conclude , as before , that @xmath313=b \cong k_5 $ ] is an endblock of @xmath2 with cutvertex @xmath64 and collect a @xmath112-hyperedge labelled @xmath39 whose vertex is @xmath64 .
so suppose that @xmath18 are adjacent and both have degree @xmath1 .
let @xmath93 be the neighbor of @xmath64 distinct from @xmath305 . since @xmath306 is critical , @xmath93 is adjacent to one of @xmath314 , so that @xmath93 is equal to the neighbor @xmath309 of @xmath63 distinct from @xmath310 , or @xmath93 is equal to @xmath163 . if @xmath93 is equal to @xmath163 then it follows , by symmetry , that @xmath309 is equal to @xmath163 ; in that case , @xmath307 $ ] is an endblock of @xmath2 , and if @xmath163 has degree @xmath1 then @xmath2 is isomorphic to @xmath7 , which is the square of a cycle of length @xmath282 , whereas otherwise @xmath163 was a cutvertex of @xmath315 $ ] and we collect a @xmath112-hyperedge labelled @xmath316 with vertex @xmath163 . in the remaining case , @xmath317 , @xmath318 , and
we collect a @xmath50-hyperedge with vertices @xmath319 .
let us label this hyperedge by the subgraph formed by the edges incident with @xmath320 .
this is isomorphic to @xmath13 , and is , up to an edge connecting @xmath321 , even equal to @xmath322 $ ] but observe that such an edge necessarily belongs to the label of a @xmath24-hyperedge collected in step i. hence we end in the subcase that @xmath149 are pairwise nonadjacent .
since @xmath109 is critical , at least one of @xmath149 must have degree @xmath1 in @xmath2 , and we collect a @xmath24-hyperedge with vertices @xmath149 and label @xmath315 \cong k_5^\triangledown$ ] .
* step iii . * after having processed all edges of @xmath2 satisfying the preconditions in step i or step ii , we are now in a situation that every edge not occuring in a label of a previously collected hyperedge is on at most two triangles and in a subgraph @xmath41 . in particular , the endvertices of each of these edges have a common neighbor of degree @xmath1 .
as long as there is such an edge @xmath323 , we process it as follows .
let @xmath324 be a subgraph of @xmath2 isomorphic to @xmath41 and containing @xmath73 .
observe that ( * ) every triangle not contained in the label of any previously collected hyperedge is edge - disjoint from the label of every previously collected hyperedge , as ( * * ) these are induced subgraphs of either @xmath2 or @xmath325 ( the reader might want to confirm ( * * ) a posteriori for the label of the hyperedge collected _ next _ ) . from this
it follows that @xmath324 is edge - disjoint from the label of every previously collected hyperedge . hence we may assume , without loss of generality , that @xmath108 have degree @xmath24 in @xmath324 and that the two vertices @xmath18 in @xmath326 are the only common neighbors of @xmath108 in @xmath324 and hence in @xmath2 , and that @xmath0 has degree @xmath1 in @xmath2 .
observe that @xmath18 can not be adjacent for otherwise one of them , say @xmath63 , had degree @xmath1 ( as @xmath66 is essential ) and the neighbor @xmath309 of @xmath63 distinct from @xmath327 must be adjacent to some @xmath328 as @xmath311 is critical ; but then @xmath329 is on at least three triangles and thus contained in the label of some hyperedge collected in step ii , contradiction .
let @xmath163 denote the neighbor of @xmath0 distinct from @xmath330 .
if @xmath79 had degree exceeding @xmath1 then @xmath18 have degree @xmath1 because @xmath331 are essential , and @xmath163 has degree @xmath1 and is adjacent to @xmath18 because @xmath332 are critical .
since @xmath108 have only two common neighbors , the neighbor @xmath105 of @xmath163 distinct from @xmath333 is also distinct from @xmath79 , and it is adjacent to one of @xmath18 because @xmath334 is critical . by symmetry , we may assume that @xmath105 is adjacent to @xmath63 .
it follows that the neighbor @xmath93 of @xmath64 distinct from @xmath335 is equal to @xmath105 , as @xmath336 is critical .
if @xmath105 is adjacent to @xmath79 then @xmath105 has degree @xmath1 since @xmath337 is essential ; in this case , we observe that by ( * ) all edges of @xmath338 \cong k_{2,2,2}$ ] are disjoint from the label of any previously collected hyperedge ; @xmath39 is an endblock of @xmath2 with cutvertex @xmath79 , and we collect a @xmath112-hyperedge with vertex @xmath79 and label @xmath39 . in the other case
, it follows again that @xmath339 is edge - disjoint from the label of any previously collected hyperedge , and we collect a @xmath50-hyperedge with vertices @xmath340 and label @xmath39 .
hence we may assume that both @xmath108 have degree @xmath1 .
let @xmath163 be as above , and let now @xmath105 be the neighbor of @xmath79 distinct from @xmath341 .
observe that @xmath342 .
as @xmath343 and @xmath337 are critical , one of @xmath18 is a common neighbor of @xmath344 of degree @xmath1 , and one of @xmath18 is a common neighbor of @xmath345 of degree @xmath1 .
let us assume first that @xmath346 .
that is , @xmath314 induces a triangle with neighborhood @xmath347 in @xmath2 .
if this neighborhood is independent in @xmath2 then we collect a @xmath24-hyperedge with vertices @xmath348 and label @xmath349 \cong k_{2,2,2}^\triangledown$ ] as usual .
if @xmath348 form a triangle then at most one of @xmath348 has degree exceeding @xmath1 .
if none has then @xmath2 is isomorphic to @xmath350 , which is the square of a @xmath243-cycle , and if @xmath351 has degree exceeding @xmath1 then we collect a @xmath112-hyperedge with vertex @xmath192 and label @xmath350 as usual .
if @xmath348 induce a path in @xmath2 then we may assume , without loss of generality , that @xmath352 are not adjacent .
if @xmath64 had degree exceeding @xmath1 then @xmath163 has degree @xmath1 as @xmath353 is essential , so that there is a unique neighbor @xmath309 of @xmath163 distinct from @xmath341 but @xmath354 can not be critical , contradiction .
hence @xmath64 has degree @xmath1 , and we collect a @xmath50-hyperedge with vertices @xmath352 and label @xmath339 . in the remaining case
, we may assume that @xmath334 is the unique edge connecting @xmath355 ( by symmetry ) .
one of @xmath352 must have degree @xmath1 , say , @xmath163 .
let @xmath309 be the unique neighbor of @xmath163 distinct from @xmath356 ; since @xmath357 is critical , we obtain that @xmath105 has degree @xmath1 and is adjacent to @xmath309 .
we then collect a @xmath50-hyperedge with vertices @xmath358 and the label @xmath359 formed by the edges incident with @xmath360 . if @xmath358 are nonadjacent then this is an induced subgraph of @xmath2 , otherwise observe that the edge @xmath361 belongs to @xmath296 and is therefore contained in the label of some @xmath24-hyperedge collected in step i. in the remaining case we are left with the situation that neither @xmath63 nor @xmath64 ( by symmetry ) , are common neighbors of @xmath352 .
we will show that @xmath2 is the square of a cycle . without loss of generality
we may assume that @xmath163 is adjacent to @xmath63 but not to @xmath64 and @xmath105 is adjacent to @xmath64 but not to @xmath63 .
we relabel the vertices @xmath362 by @xmath363 , respectively , and observe that @xmath364 $ ] induces the square of a path @xmath365 , where @xmath366 have degree @xmath1 in @xmath2 and @xmath367 is the only common neighbor of @xmath368 in @xmath2 . by ( * * ) , the edges of this square do not occur in the label of any previously collected hyperedge .
let @xmath369 , @xmath166 , be a maximal subpath of @xmath2 such that its vertex set induces its square @xmath370 , @xmath371 have degree @xmath1 in @xmath2 and @xmath367 is the only common neighbor of @xmath368 in @xmath2 , and the edges of @xmath370 do not occur in the label of any previously collected hyperedge .
in particular , every edge incident with one of @xmath372 is in @xmath370 .
since @xmath373 has degree @xmath24 in @xmath370 , there exists a uniqe vertex in @xmath374 in @xmath375 .
since @xmath376 is critical , @xmath377 has degree @xmath1 and is adjacent to @xmath374 , too ( where for @xmath227 we used the fact that @xmath367 is the only common neighbor of @xmath368 in @xmath2 ) . hence the subgraph formed by @xmath370 and the two edges @xmath378 and @xmath379 is the square of the path @xmath380 , but not an _ induced _ subgraph , by maximality of @xmath19 . hence @xmath374 is adjacent to one of @xmath381 ( or to both ) .
let @xmath382 be the unique vertex in @xmath383 .
since @xmath384 is critical it follows that @xmath374 has degree @xmath1 and is adjacent to @xmath382 .
now let @xmath385 be the uniqe neighbor of @xmath374 distinct from @xmath386 .
if @xmath387 then @xmath388 and @xmath374 must have a common neighbor of degree @xmath1 among @xmath389 .
it can not be @xmath390 or @xmath367 because @xmath374 is not adjacent to these , and it can not be @xmath391 because @xmath367 is the unique common neighbor of of @xmath368 in @xmath2 . therefore , @xmath392 . then @xmath390 and @xmath374 must have a common neighbor of degree @xmath1 , and that can be only @xmath382 . now
if @xmath393 were nonadjacent then the unique neighbor @xmath309 of @xmath388 distinct from @xmath389 would be distinct from all @xmath394 , implying that @xmath395 can not have a common neighbor of degree @xmath1 , contradiction .
hence @xmath2 is isomorphic to the square of the cycle @xmath396 .
in this section we prove the statement of lemma [ l2 ] where @xmath53 is obtained from @xmath2 by _ deleting _ an edge instead of _ contracting _ it .
[ l3 ] suppose that @xmath53 is a graph of minimum degree at least @xmath1 obtained from a simple graph @xmath2 ( of minimum degree at least @xmath1 ) by deleting an edge @xmath109 , and that @xmath125 is a nonseparating induced subgraph in @xmath53 isomorphic to a subdivision of a wheel or of a prism or of @xmath57 .
then there exists a nonseparating induced subgraph @xmath3 in @xmath2 isomorphic to a subdivision of a wheel or of a prism or of @xmath57 such that @xmath126 .
* proof . *
if @xmath109 does not connect elements from @xmath397 then we take @xmath129 .
if @xmath109 is a chord of a subdivision path @xmath398 $ ] then we obtain an induced subdivision @xmath3 of the same type as @xmath125 from @xmath125 by adding @xmath109 and deleting all inner vertices of the @xmath108-subpath in @xmath398 $ ] ; as these have degree @xmath50 in @xmath125 , they have a neighbor in @xmath399 , so that @xmath5 remains nonseparating .
so let us suppose that @xmath109 connects elements from distinct subdivision paths of @xmath125 .
* @xmath125 is a subdivision of @xmath57 .
then consider the two principal color classes @xmath400 and @xmath401 .
if @xmath109 connects two principally nonadjacent principal vertices , say , @xmath402 , then @xmath403 $ ] is a nonseparating induced subdivision of @xmath4 in @xmath2 , and if otherwise one of @xmath108 is an internal vertex of a subdivision path , say , @xmath404 , then @xmath405 is .
@xmath146 * case 2 .
* @xmath125 is a subdivision of a prism @xmath56 .
let @xmath149 and @xmath156 form the principal triangles and suppose that @xmath149 is principally adjacent to @xmath156 , respectively . if @xmath109 connects two principally nonadjacent principal vertices , say , @xmath406 , then @xmath407 $ ] is a nonseparating induced subdivision of @xmath4 .
if one of @xmath108 is a subdivision vertex of the two principal triangles , say , @xmath408 then @xmath409 will do it . in the remaining case , we may assume by symmetry that @xmath410)$ ] and @xmath411 , so that @xmath412 $ ] is a nonseparating induced subdivision of @xmath4 .
@xmath146 * case 3 .
* @xmath125 is a subdivision of a wheel .
let @xmath63 be the principal center @xmath63 and @xmath413 be the principal rim vertices , in this order .
if both @xmath108 are on the rim cycle of the subdivision then we may assume , by symmetry , that @xmath414 and @xmath415 for some @xmath191 , @xmath416 ( indices modulo @xmath32 ) .
if @xmath417 then @xmath418 ) \cup \bigcup_{j=1}^{k-1 } v(h'[b_j , b_{j+1 } ] ) \cup v(h'[x , b_1 ] ) \cup v(h'[b_k , y ] ) ] $ ] is an induced nonseparating subdivision of a wheel with rim vertices @xmath419 and center @xmath63 , otherwise @xmath420 ) \cup v(h'[a , b_1 ] ) \cup v(h'[a , b_2 ] ) \cup v(h'[b_0,b_1 ] ) \cup v(h'[b_1,b_2 ] ) \cup v(h'[b_2,y ] ) ] $ ] is either a nonseparating subdivision of @xmath4 with principal vertices @xmath421 ( if @xmath79 is not adjacent to @xmath422 in @xmath125 ) , or @xmath423 is a nonseparating subdivision of the prism @xmath56 with principal triangles @xmath424 and @xmath425 ( if @xmath79 is adjacent to @xmath422 in @xmath125 ) .
now if @xmath426 then the only situation we have not considered so far , by means of relabelling , is that @xmath108 are internal vertices of disjoint subdivision paths ; in this case it is easy to see that @xmath427 $ ] is a subdivision of @xmath57 .
hence it suffices to consider the case that @xmath166 and @xmath428 ( by symmetry ) .
if @xmath429 then @xmath427 $ ] is a subdivision of a wheel with center @xmath63 and rim vertices @xmath430 . otherwise , @xmath431 $ ] is a nonseparating subdivision of a wheel with center @xmath63 and rim vertices @xmath432 .
let us now combine lemma [ l1 ] , lemma [ l2 ] , lemma [ l3 ] , and theorem [ t5 ] to generalize theorem [ t4 ] to simple connected graphs of minimum degree at least @xmath1 .
[ t6 ] let @xmath0 be a vertex of a simple connected graph @xmath2 of minimum degree at least @xmath1 .
then @xmath2 contains an induced subgraph @xmath3 isomorphic to a subdivision of a wheel or of the prism or of @xmath57 such that @xmath5 is connected and contains @xmath0 .
* recall that @xmath271 denotes the class of simple connected graphs of minimum degree at least @xmath1 .
we do induction on @xmath232 .
suppose that we have proved the statement for all graphs with less vertices than @xmath2 .
if there is a non - essential edge or a non - critical edge then we proceed by induction using lemma [ l3 ] or lemma [ l2 ] , respectively .
if @xmath2 is the square of a cycle of length at least @xmath282 then the statement is true by theorem [ t4 ] ( or just by the respective easy argument in the proof of theorem [ t4 ] ) .
otherwise , there exists a hypergraph and edge disjoint bricks @xmath284 as in theorem [ t5 ] from which we can obtain @xmath2 as described there .
if there is a @xmath24-hyperedge @xmath73 with a nontriangular brick @xmath284 then we look at the graph @xmath53 obtained from @xmath2 by replacing that brick by a triangular one , i. e. we take @xmath433 and add three new edges connecting the vertices from @xmath280 pairwise . as @xmath280 is independent in @xmath2 , we will not generate multiple edges , so that @xmath53 is in class @xmath271 .
let @xmath434 if @xmath435 , otherwise we take any @xmath436 adjacent to @xmath0 .
by induction , we find an induced subgraph @xmath125 in @xmath53 avoiding @xmath250 and isomorphic to a subdivision of a wheel , the prism , or @xmath57 .
if this subdivision contains at most one of the new edges then there is at least one vertex in @xmath437 .
we may replace the new edge in the subdivision ( if present ) by a path of length @xmath50 in @xmath284 connecting its endvertices and avoiding @xmath0 ( by choice of @xmath250 ) , as to obtain a nonseparating induced subdivision in @xmath2 , of the same type as @xmath125 .
if , otherwise , this subdivision contains more than one new edge then it must contain all three new edges , and all three vertices from @xmath280 must be principal vertices in @xmath125 .
( in particular , @xmath125 is a subdivision of a wheel or a prism . )
at least two of these have degree @xmath24 in @xmath125 , and we find a nonseparating induced cycle @xmath438 in @xmath125 avoiding one of them and using exactly one of the new edges .
this cycle is an induced nonseparating cycle in @xmath53 , and hence @xmath439 $ ] is a nonseparating subdivision of a wheel with three or four rim vertices ( depending on whether @xmath293 or @xmath440 ) , avoiding @xmath0 .
hence we may assume that @xmath441 for every @xmath24-hyperedge @xmath73 in our representation . if there is a separating @xmath50-hyperedge @xmath73 ( where _ separating _ refers to the hypergraph ) , then its vertices are independent in @xmath2 , and we look at the graph @xmath53 obtained from @xmath2 by contracting @xmath284 to a single vertex @xmath93 .
the resulting graph is in class @xmath271 .
we set @xmath247 if @xmath442 and @xmath249 otherwise , and apply induction to @xmath53 as to obtain an induced subgraph @xmath125 of @xmath53 as in the statement .
since @xmath125 has no separating vertex , it avoids all but one component of @xmath443 , implying that it corresponds to a nonseparating induced subdivision in @xmath2 of the same type , avoiding @xmath0 .
hence we may assume that there is no separating @xmath50-hyperedge
. if there is a vertex @xmath93 of degree exceeding two in our hypergraph then it can not be incident with @xmath24-hyperedges ( cf .
( ii ) in theorem [ t5 ] ) .
suppose that such a @xmath93 is a cutvertex of @xmath2 .
then there exists a component @xmath26 of @xmath444 which does not contain @xmath0 .
if that component is equal to @xmath445 for some @xmath112-hyperedge then it is a wheel with @xmath24 or @xmath1 rim vertices , avoiding @xmath93 , and this will serve for @xmath3 in the statement . if not then @xmath26 corresponds to a component of our hypergraph minus @xmath93 , and there are at least two @xmath50-hyperedges connecting @xmath93 to @xmath101 , because there is no separating @xmath50-hyperedge . but then @xmath446 $ ] is connected and has minimum degree at least @xmath1 , so that we find a nonseparating induced subdivision @xmath3 as desired avoiding @xmath93 in @xmath53 , and this will be nonseparating in @xmath2 , too . hence all vertices in our hypergraph have degree @xmath50 .
we have already seen that for every @xmath112-hyperedge @xmath73 , @xmath447 is a nonseparating wheel with @xmath24 or @xmath1 rim vertices .
hence in this case we find @xmath3 provided that @xmath435 . for a @xmath50-hyperedge with @xmath435 , let us consider @xmath448 , and let @xmath309 be the vertex in @xmath449 .
then @xmath93 is not a cutvertex of @xmath2 ( for otherwise @xmath73 was a separating @xmath50-hyperedge , contradiction ) .
it follows that @xmath450 does not separate @xmath2 , as @xmath451 is an endblock of @xmath444 .
but then either @xmath452 $ ] is a nonseparating wheel in @xmath2 with @xmath24 or @xmath1 rim vertices ( if @xmath284 is @xmath9 or @xmath10 ) , or @xmath453 $ ] is such a wheel ( if @xmath284 is @xmath13 or @xmath14 ; recall that @xmath73 is a nonseparating @xmath50-hyperedge , so that @xmath454 does not separate @xmath2 ) . as @xmath0 is an internal vertex of at most one brick
, we thus find a nonseparating subgraph as desired if there is _ more _ than one hyperedge with at most two vertices .
hence we may assume that all but at most one hyperedges are @xmath24-hyperedges . if _
all _ hyperedges are @xmath24-hyperedges then @xmath2 is the line graph of a simple connected cubic graph and we find the desired subgraph @xmath3 by using lemma [ l1 ] just as in the proof of theorem [ t4 ] .
we reduce the remaining cases to this situation .
so suppose that there is a @xmath112- or @xmath50-hyperedge @xmath73 .
then we may assume that @xmath455 by what we have done earlier .
if @xmath73 is actually a @xmath112-hyperedge then we take two disjoint copies @xmath456 of @xmath433 and identify the two vertices @xmath457 in either copy corresponding to the vertex in @xmath280 ; the new graph @xmath53 is a connected graph and can be represented as the linegraph of a connected cubic graph , hence we find a nonseparating induced subgraph @xmath125 avoiding the vertex @xmath250 corresponding to @xmath457 as desired ; as @xmath250 separates @xmath53 , @xmath125 is contained in either @xmath458 or @xmath459 and hence will correspond to a nonseparating subgraph in @xmath2 avoiding @xmath294 .
if @xmath73 is a @xmath50-hyperedge with @xmath460 and @xmath461 are _ adjacent _ in @xmath2 then @xmath461 belong to the same triangular brick @xmath462 ; the vertex @xmath463 is then a cutvertex in @xmath2 , and @xmath464 $ ] is an endblock of @xmath2 . in that case , we proceed similarly as in the preceeding paragraph : we take disjoint copies of @xmath465 and merge them at the vertices corresponding to @xmath133 , apply lemma [ l1 ] as to find a nonseparating @xmath125 avoiding @xmath133 in one of the copies and hence a nonseparating @xmath3 in @xmath2 avoiding @xmath466 .
hence we are left with the case that @xmath73 is a @xmath50-hyperedge with @xmath460 and _ nonadjacent _ @xmath461 .
then @xmath461 belong to distinct triangular bricks @xmath462 , @xmath467 , respectively .
if @xmath468 then we contract @xmath284 to a single vertex @xmath93 as to obtain @xmath53 ; by induction , there is a nonseparating induced subgraph @xmath125 in @xmath53 avoiding @xmath93 as desired , and hence there is a nonseparating @xmath3 in @xmath2 avoiding @xmath294 .
otherwise , we may write @xmath469 and @xmath470 , where @xmath471 are distinct , and consider the graph @xmath53 obtained from @xmath2 by deleting @xmath472 and adding four new vertices @xmath473 forming a @xmath4 plus the four edges @xmath474 . applying induction to @xmath53 and @xmath63 , we find an appropriate nonseparating induced subdivision @xmath125 in @xmath53 .
if it does not contain both of @xmath471 then it can not contain any of @xmath473 , and it serves as a nonseparating induced subdivision avoiding @xmath294 in @xmath2 .
otherwise , if @xmath397 contains both @xmath471 , then it must contain all vertices from @xmath475 , as it is nonseparating in @xmath53 and avoids @xmath63 .
if @xmath125 avoids all of @xmath473 then it will serves as a nonseparating induced subdvision avoiding @xmath294 in @xmath2 . otherwise , it is easy to see that @xmath476 is an induced path in @xmath2 , so that @xmath477 $ ] is a nonseparating subdivision in @xmath2 avoiding @xmath294 ( of the same type as @xmath125 ) .
@xmath111 as a corollary , we get , again , the following : [ c2 ] for every vertex @xmath0 in a connected graph @xmath2 of minimum degree at least @xmath1 , there exists a subgraph @xmath3 isomorphic to a subdivision of @xmath4 such that @xmath5 is connected and contains @xmath0 .
in contrast to the situation in theorem [ t4 ] on @xmath1-connected graphs , where we do not know if the list of graphs we expect to find subdivided as induced nonseparating subgraphs is minimal ( see problem [ p1 ] ) , for every @xmath166 there exist connected graphs of minimum degree at least @xmath1 such that if @xmath3 is a nonseparating induced subgraph isomorphic to a subdivision of a wheel or the prism or @xmath57 then @xmath478 : just let @xmath2 be the graph by taking the union of @xmath479 disjoint cycles @xmath480 of length @xmath32 , adding two new vertices @xmath108 and making them adjacent to the vertices of the cycles and to each other . it is obvious that an induced subdivision @xmath3 as above is contained in @xmath481 $ ] for some @xmath482 , and that it contains at most one @xmath140 if it is nonseparating ; this implies @xmath483 \cong w_\ell$ ] .
the list of graphs in theorem [ t4 ] and theorem [ t6 ] have been characterized earlier as the @xmath24-connected graphs in which every subgraph isomorphic to a subdivision of @xmath4 is spanning @xcite .
i do not know if this result can be employed here to give a simpler proof of , say , corollary [ c1 ] or just of the fact that every @xmath1-connected graph has a nonseparating subdivision of @xmath4 .
there is a number of potential applications of theorem [ t5 ] .
in fact , one might try to generalize statements where theorem [ t3 ] is employed as an induction tool . in a forthcoming paper with a. s. pedersen
we will use it for a coloring problem related to lovsz s conjecture on double - critical graphs .
* sketch of proof .
* if there is a vertex @xmath0 such that @xmath484 is outerplanar then @xmath2 has a spanning subdivision of @xmath4 and the statement is true .
otherwise , there exists a separating induced cycle @xmath26 in @xmath2 .
we fix an embedding of @xmath2 and take @xmath26 such that the bounded region @xmath485 of @xmath26 is minimal , i. e. , there is no other nonseparating induced cycle whose bounded region is properly contained in @xmath38 .
there is a vertex @xmath93 embedded in @xmath38 , and there are three @xmath486-paths having pairwise only @xmath0 in common ; they trisect @xmath38 and form , together with @xmath26 , a @xmath4-subdivision @xmath3 in @xmath2 . by choice of @xmath38
, @xmath3 contains _ all _ vertices embedded in @xmath38 , hence @xmath5 is the component formed by the vertices embedded in the unbounded region of @xmath38 .
@xmath111 \4 .
let @xmath487 .
the analogous statements of theorem [ t5 ] for the class of connected graphs of minimum degree at least @xmath16 are much easier to prove . for @xmath223
, we get that every simple connected graph @xmath2 of minimum degree at least @xmath50 can be reduced to a smaller such graph by deleting a single edge or contracting a single edge and simplifying unless @xmath2 is a _ friendship graph _ , i.e. obtained from taking @xmath488 disjoint copies of @xmath6 , chosing a single vertex in either of them , and identifying all chosen vertices .
( see ( * ? ? ?
* theorem 6 ) for the _ friendship theorem_. ) for @xmath489 , let us consider the two _ @xmath24-bricks _ : @xmath4 with only one vertex of attachment , and @xmath41 whose vertices of attachment are its two degree @xmath50-vertices .
we get ( without proof ) : [ t8 ] a simple graph @xmath2 of minimum degree at least @xmath24 can not be reduced to a smaller such graph by deleting a single edge or contracting a single edge and simplifying if and only if it is @xmath4 or arises arises from a nonempty connected multigraph @xmath3 of minimum degree at least @xmath50 and @xmath490 for all hyperedges @xmath73 by replacing each edge ( might be a loop ) @xmath73 by a @xmath24-brick @xmath284 such that the vertices of attachment of @xmath284 are those in @xmath280 and at the same time the only objects of @xmath284 contained in more than one brick . as a consequence we get , for example , an analogue to corollary [ c2 ] , that for every vertex @xmath0 of a simple graph @xmath2 of minimum degree @xmath24 there exists a nonseparating induced cycle avoiding @xmath0 .
nonseparating cycles have been studied in @xcite , and we get alternative proofs for some results in there , for example for corollary 2 in @xcite . \5 .
graphs which do not contain a subdivision of @xmath4 as an induced subgraph at all ( no matter whether nonseparating or not ) have been investigated in @xcite .
it turns out that these graphs are @xmath163-colorable for some constant @xmath163 , and even @xmath24-colorable if , in addition , they do not contain a wheel as an ( induced ) subgraph .
we get a weaker version from theorem 6 : if @xmath2 contains no induced subgraph isomorphic to a subdivision of a wheel or of the prism or of @xmath57 then it has a vertex of degree at most @xmath24 ; as this holds for every induced subgraph , we can ( greedily ) @xmath1-color such a graph .
it remains open if this remains true if we exclude only _ nonseparating _ subdivisions as above .
99 and m. kriesell , `` what is on his mind ? '' , discr .
( 2010 ) , 25732583 .
, a. rnyi , and v. t. ss , `` on a problem of graph theory '' , studia sci . math . hungar . 1 ( 1966 )
, 215235 . ,
`` connectivit des graphes automorphismes des cartes : proprits et algorithmes '' , thse detat , universit p. et m. curie , paris ( 1979 ) .
, `` graphes @xmath1-essentiels '' , c. r. acad .
paris ( 1978 ) , t. 287 , 289290 . and
k. kawarabayashi , `` non - separating even cycles in highly connected graphs '' , to appear in combinatorica , preliminary version from the authors .
and w. khnel , `` convex and two - piece - property drawings of graphs '' , preliminary version from the authors ( 2011 ) .
, o. lee , b. reed , and p. wollan , `` a weaker version of lovsz path removal conjecture '' , j. combin .
theory ser .
b 98 ( 2008 ) , no . 5 , 972979 . , `` vertex suppression in @xmath24-connected graphs '' , j. graph theory 57 ( 2008 ) , 4154 . and
d. osthus , `` partitions of graphs with high minimum degree or connectivity '' , j. combin .
theory ser .
b 88 ( 2003 ) , no .
1 , 2943 . , personal communication . , f. maffray , and n. trotignon , `` on graphs with no induced subdivision of @xmath4 '' , preprint from the authors ( 2010 ) . , `` on @xmath16-critically @xmath491-connected graphs '' , progress in graph theory , waterloo 1982 , academic press ( 1984 ) , 389398 . , `` a recursive characterization of the 4-connected graphs '' , discrete math .
84 ( 1990 ) , 105108 . , `` uncontractable 4-connected graphs '' , j. graph theory 6 ( 1982 ) , 343344 . ,
`` graph decomposition with applications to subdivisions and path systems modulo @xmath16 '' , j. graph theory 7 ( 1983 ) , 261271 .
, `` nonseparating cycles in @xmath16-connected graphs '' , j. graph theory 5 ( 1981 ) , no .
4 , 351354 . and
b. toft , `` nonseparating induced cycles in graphs '' , j. combin .
theory ser .
b 31 ( 1981 ) , no .
2 , 199224 . and
k. vukovi , `` a structure theorem for graphs with no cycle with a unique chord and its consequences '' , j. graph theory 63 ( 2010 ) , no .
1 , 3167 . | we first prove that for every vertex @xmath0 of a @xmath1-connected graph @xmath2 there exists a subgraph @xmath3 in @xmath2 isomorphic to a subdivision of the complete graph @xmath4 on four vertices such that @xmath5 is connected and contains @xmath0 .
this implies an affirmative answer to a question of w. khnel whether every @xmath1-connected graph @xmath2 contains a subdivision @xmath3 of @xmath4 as a subgraph such that @xmath5 is connected . the motor for our induction is a result of fontet and martinov stating that every @xmath1-connected graph can be reduced to a smaller one by contracting a single edge , unless the graph is the square of a cycle or the line graph of a cubic graph .
it turns out that this is the only ingredience of the proof where @xmath1-connectedness is used .
we then generalize our result to connected graphs of minimum degree at least @xmath1 , by developing the respective motor , _ a structure theorem for the class of simple connected graphs of minimum degree at least @xmath1 _ : a simple connected graph @xmath2 of minimum degree @xmath1 can not be reduced to a smaller such graph by deleting a single edge or contracting a single edge and simplifying if and only if it is the square of a cycle or the edge disjoint union of copies of certain _ bricks _ as follows : each brick is isomorphic to @xmath6 , @xmath7 , @xmath8 , @xmath9 , @xmath10 , or one the four graphs @xmath11 , @xmath12 , @xmath13 , @xmath14 obtained from @xmath7 and @xmath8 by deleting the edges of a triangle , or replacing a vertex @xmath0 by two new vertices and adding four edges to the endpoints of two disjoint edges of its former neighborhood , respectively .
bricks isomorphic to @xmath7 or @xmath8 share exactly one vertex with the other bricks of the decomposition , vertices of degree @xmath1 in any other brick are not contained in any further brick of the decomposition , and the vertices of a brick isomorphic to @xmath6 must have degree @xmath1 in @xmath2 and have pairwise no common neighbors outside that brick . * ams classification : * 05c75 , 05c83 .
* keywords : * subdivision , minimum degree , complete graph , decomposition , structure theorem . |
all order resummed predictions have long been a common method to extend the predictive power of theoretical qcd estimates for several observables compared to that provided by fixed order perturbative computations alone .
this applies in particular to observables that are measured close to their exclusive limits , such as event shape variables in @xmath0 annihilation @xcite where the highest statistics are typically found close to the two jet limit .
more generally , if a physical variable such as a typical event shape , defined for an arbitrary hard process , is constrained , by observation / measurement , to be near its born value , it turns out to be sensitive to soft emissions .
in fact if the deviation from the born value of an observable ( which we can take to be zero for the following discussion ) is denoted by @xmath1 , near the born limit the cross - section for values up to @xmath1 ( @xmath2 ) is dominated by soft and/or collinear logarithms . in particular
the @xmath3 order perturbative estimate typically goes as : @xmath4 where the dots denote terms less singular in the @xmath5 limit and @xmath6 is a scale relevant to the underlying hard process .
hence the predictive power of fixed order perturbation theory is spoilt in the small @xmath1 region , since the smallness of the expansion parameter @xmath7 is compensated by the presence of large logarithms with upto two powers of @xmath8 for each power of @xmath9 . by
now there are solid techniques in place to handle such logarithmic behaviour to all orders , which result in improved predictions valid over a much larger range of variable values than the fixed order computations alone , which become meaningless at very small @xmath1 .
these resummation techniques rely mainly on factorisation methods ( broadly speaking ) , which in turn follow from qcd coherence properties and the observables dependence on final state emissions being factorisable , usually in some appropriate transform space @xcite . due to the latter property
not all variables turn out to belong to the class of resummable observables
an example is jet rates defined with the jade jet algorithm @xcite .
while resummed predictions are often needed to improve perturbative accuracy , they also serve as a pathway to approach the non - perturbative domain one of the most challenging and certainly the least understood aspect of qcd .
specifically , to access large non - perturbative effects one has to enter the low @xmath10 domain , where @xmath10 is the transverse momentum of a typical gluon emission , leading to a measured value of the observable .
hence in the small @xmath11 region , aside from large logarithms one also has power corrections @xmath12 where the @xmath13 are non - perturbative coefficients . while for @xmath14 the problem becomes highly non - perturbative , nevertheless for hard enough processes a wide range of values exists , @xmath15 , where both perturbative resummations and non - perturbative power corrections are important effects . exploring this region quantitatively
is important as it provides valuable information on the onset of confinement effects . in short resummations
are a useful probe of qcd dynamics , which comes into its own in the infrared region , where parton multiplication is copious and hadronisation effects are important . in what follows
we shall confine the discussion to observables that do admit phase space factorisation and hence are resummable .
we shall begin by considering the anatomy of a typical resummed answer and indicate what is the state - of the art with regard to computation of the various pieces we will mention .
we shall then discuss some significant recent developments , which contributed to the improvement of resummed predictions and consequently reflect an improved understanding of qcd dynamics : in particular the discovery of non global logarithms and the advent of generalised approaches to resummation .
we shall also present some recent comparisons of such improved predictions with data .
consider an observable which has large logarithms in its perturbation expansion as below : @xmath16 one can naturally define leading logarithms ( ll ) as being the double logs @xmath17 , next - to leading logs ( nll ) as those terms with @xmath18 etc and in some cases this nomenclature is indeed employed @xcite .
note however that in the regime where @xmath19 , any truncation at @xmath20 order , in this terminology , is no longer useful since one always has neglected terms @xmath21 with @xmath22 which are larger than one .
it should be pointed out that the state - of the art resummation enables us to go to nnll order in this notation .
however since variables that admit phase space factorisation typically exponentiate , it is possible to have a different ( and in fact more commonly used ) classification of logarithmic terms .
exponentiation means one can write @xcite @xmath23 the first piece in brackets is simply a well behaved perturbative expansion ( with no dependence on @xmath1 ) in @xmath9 .
this piece treats for instance the mismatch between the full real and virtual emission and their soft limits , included in the form factor @xmath24 .
all the singular dependence on @xmath1 is contained in the form factor @xmath24 .
it takes the form @xmath25}.\ ] ] with this exponentiation it is now possible to define @xmath26 as a ll function ( since it is accompanied by an extra logarithm @xmath27 ) , @xmath28 as an nll function ( which is purely single logarithmic @xmath29 ) , @xmath30 as nnll due to the extra factor of @xmath9 relative to the single logarithmic @xmath28 piece . in this notation
the state of the art is nll resummation in that resummed predictions aim to compute up to the @xmath28 piece .
this prediction is then matched to fixed order calculations .
the ` remainder ' piece @xmath31 vanishes in the @xmath5 limit .
having clarified the basic notation and terminology we now turn our attention to some more recent developments .
we begin the next section by providing a short discussion on non global observables .
until very recently the resummed predictions available in the literature were all made employing an independent emission formalism . in other words the matrix element for multiple gluon production was approximated by a factorised product of single soft emission contributions : @xmath32 with @xmath10 the transverse momentum and @xmath33 the rapidity wrt the emitting jet axis . the independent emission pattern ( in one form or another ) was used in conjunction with phase space factorisation to yield the form factor @xmath24 .
independent emission essentially means one can assume each final state gluon to be emitted directly by the hard initiating parton as in an abelian theory .
subsequent branching of these gluons can in fact be neglected for several observables , apart from the contribution of such decays to the running of the coupling
. however the independent emission approximation ( [ eq : indem ] ) was found to be insufficient at single logarithmic level ( ie at the level of the function @xmath28 ) for several observables @xcite . in some of these cases , such as energy flow away from jets @xcite ,
single logarithms were in fact a leading effect due to absence of collinear enhancements , which meant that the function @xmath26 was absent .
in general the observables for which independent emission breaks down at single - log level have one common feature responsible for this breakdown they are sensitive to emissions only in a limited rapidity region .
for this reason such observables and corresponding logarithms are refered to as _ non global _ , while the complementary set of observables is refered to as global .
the origin of the problem is simple to understand and briefly explained below .
consider a measurement of say energy flow away from hard jets in a solid angle @xmath34 as depicted above .
then by restricting , for example , the transverse energy in @xmath34 such that @xmath35 with @xmath6 a hard scale , one is vetoing real radiation above scale @xmath36 into @xmath34 .
assuming an independent emission ansatz one needs to just veto direct emission from the hard jet lines as in diagram ( a ) above
. plotted against @xmath37 for different definitions of @xmath34 .
] this yields a single logarithmic form factor @xmath38.\ ] ] the above result is incorrect at single log level .
this can be seen by expanding it to @xmath39 and comparing with the logarithmic dependence of an exact fixed order estimate at @xmath39 .
in fact an additional source of single logarithms is found starting from @xmath40 which removes the discrepancy with the single logarithms in the fixed order estimate . at order @xmath40
this is just emission of a soft gluon with energy @xmath41 by a relatively harder gluon with energy @xmath42 with @xmath43 outside @xmath34 ( see diagram ( b ) ) .
this and analogous higher order contributions are missed by the independent emission approach which would be correct if both the gluon outside @xmath34 and the softest gluon inside @xmath34 were measured ( vetoed ) .
the effect of the softest gluon would cancel against virtual corrections at single logarithmic level .
however since we are only measuring _ inside _ @xmath34 the softest gluon emission is significant even in the presence of the harder emitter outside @xmath34 and hence the gluon branching contribution appears . the generalization of the above effect to @xmath3 order is the coherent emission of a single softest gluon into @xmath34 by an ensemble of @xmath44 gluons outside @xmath34 ( diagram ( c ) above .
the @xmath44 gluon emitters are themselves ordered in energy , @xmath45 .
the effect of such multiple wide - angle ( non collinear - enhanced ) soft emission is precisely a single logarithm , @xmath46 . resumming these terms needs a change of approach from the independent emission approximation . in ref .
@xcite , the resummation of single logarithms is carried out numerically by using the dipole evolution picture which captures the essential single logarithms in the large @xmath47 limit . the resulting function is plotted in fig .
[ fig : ng ] . following this a non - linear evolution equation corresponding to the dipole evolution
was derived in ref .
@xcite and its numerical solution yielded identical results to those obtained in ref .
@xcite .
subsequent to the discovery of non global logs , there was some effort made at understanding how to define observables in a way such that non global effects may be minimised , which would reduce any uncertainty to do with missing non global @xmath48 suppressed effects which can not yet be resummed .
the reduction or removal of the non global component is also useful for facilitating the phenomenology of multi jet event shapes and energy flows .
this is because non global logs have thus far only been explicitly computed in the two jet case , although a similar structure is expected in the extension to multi - jet events and their calculation will also be along the same lines , employing dipole ( non - linear ) evolution . in the above regard appleby and seymour considered the effect of jet clustering algorithms on non global logarithms .
they found that if they defined rapidity gaps in terms of minijet energy flows ( with the minijets being defined by a clustering alogorithm ) , rather than a sum over hadronic energies in the gap , the non global component was significantly reduced .
this is essentially because the clustering algorithm has the effect of pulling soft hadrons ( partons ) out of the gap and clustering them with harder emissions outside the gap .
hence the non global contribution is not triggered except in specific geometrical configurations which survive the clustering , which reduces its numerical significance ( see @xcite for details ) .
berger , kucs and sterman introduced event - shape / energy flow correlations @xcite aimed at controlling non global effects in energy flow .
this meant simultaneously restricting the energy @xmath36 in @xmath34 alongside limiting the value of an event shape variable @xmath1 , defined in terms of hadron momenta outside @xmath34 .
doing so amounts to restricting transverse momenta of soft gluons both outside the gap region , by controlling the event shape @xmath49 , and inside the gap @xmath50 with @xmath51 .
non global or secondary logarithms now appear as @xmath52 contributions and hence varying the values of @xmath1 and @xmath36 in tandem , allows one to control the significance the of the non - global contribution .
in the course of these studies the same authors also introduced new kinds of event shapes with an adjustable parameter @xmath53 that allows one to control the approach to the two jet limit @xcite .
these variables have a parametric dependence on the transverse momenta @xmath10 and rapidity @xmath33 with respect to the thrust axis , of the form @xmath54 , where the sum extends over all final state partons .
the resummation to nll accutacy , for these observables can be found in ref .
@xcite . a special case , @xmath55 , of the above class of variables is the much studied thrust variable .
dokshitzer and marchesini @xcite further extended the study of event shape / energy flow correlations by showing that the non global part of the answer , which involves logarithms of @xmath56 , factorises from the usual global resummation of the event shape @xmath1 ( this time defined as a sum over all final state hadron contributions in and outside the gap ) . although some effort has been devoted to understanding and computing non global logarithms , the only inclusion of this effect in comparisons to data thus far has been in the case of dis event shape variables . for details
the reader is refered to ref .
. a recent comparison of a matched resummed prediction ( including non global effects and power corrections ) to h1 data is shown in fig . 2 , for the thrust defined wrt the thrust axis in the breit current hemisphere of dis .
an interesting and fairly recent development is that of nll resummed predictions for three jet event shapes in @xmath0 , dis ( two final state + one incoming jet ) and hadron - hadron collisions ( two incoming jets and a final state ( radiated ) jet+vector boson ) . for an example
see @xcite .
values from 15 81 gev . ] for all these processes , at born level one finds that the three hard jets lie in a plane defined say by the thrust and thrust major axes .
small deviations from the born level can be accesed by studying small values of the out of plane momentum @xmath57 .
compared to two jet event shapes such as thrust , the three jet variables offer some new challenges . from the point of view of the resummed prediction , one as to also take account of coherent , soft interjet radiation while in the global two jet case only radiation emitted collinearly ( effectively incoherently ) by each jet contributes to nll accuracy .
the interjet radiation gives a geometry dependence at nll level , characteristic of three jet variables .
such variables would be very interesting to study experimentally especially from the viewpoint of hadronisation corrections .
this is a particularly important intermediate step along the way to understanding power behaved hadronisation effects in multijet configurations , such as in dijet event shapes in hadroproduction ( @xmath58 processes ) .
of late there has been significant effort and progress made in developing resummation formulae that are as general as currently possible and which reduce the need for treating different observables and different processeses on a laborious case - by case basis .
the idea is to write down master formulae which have general applicability and contain process and observable dependence in a few parameters which need to be computed on a case specific basis @xcite .
for example processes involving an arbitrary number of hard partons were treated in ref .
@xcite .
a generalised approach that aims at facilitating the computablity of resummed results , using numerical methods , and in fact automating the entire process of resummation was developed in @xcite .
the idea roughly , is to test the observable s dependence on soft and collinear emissions and determine numerically the parametric dependence on transverse momenta , rapidity and azimuth wrt the nearest leg .
once this dependence is obtained , the results are automatically inserted into a master formula valid for an arbitrary case ( there however being some conditions such as globalness , that the observable has to satisfy for the master formula to be valid ) .
final results are numerically obtained by means of a computer program .
this is a particularly powerful method since almost no analytical effort is needed to generate the final results and the entire procedure is automated .
its advantage is particularly manifest in cases where analytical calculations are too cumbersome , requiring the inversion of multiple mellin and/or fourier transforms or the observable depends in a complicated way on multiple soft emissions .
a good example of this latter point is the durham jet finding algorithm for which the analytic computation of the full nll function @xmath28 was intractable due to effects to do with recombination of soft partons .
the numerical resummation methods were on the other hand able to determine this function for the first time @xcite .
another example of the power of the automated resummation approach was provided by the computation , for the first time , to nll accuracy of an event shape in hadron
hadron collisions ( with a dijet final state ) the transverse thrust distribution @xcite .
this is defined as follows : @xmath59 where as usual the sum extends over all final state particles and @xmath60 is the transverse momentum wrt the beam direction while @xmath61 is a unit vector found by maximising the sum in the numerator .
the results for the different dijet production channels is illustrated in fig .
2 . in conclusion
it is worthwhile to note that there has been significant progress in the development of resummed computations in the last few years .
the consequences of these developments are naturally of great value to experimental studies and future qcd phenomenology .
additionally they point towards unravelling of previously unexplored qcd dynamics and therefore to important progress on the theoretical front .
gev .. fig . taken from ref . @xcite . ]
i would like to thank the authors of ref .
@xcite for the use of their figure for fig . 3
. 9 s. catani , l. trentadue , g. turnock and b.r .
webber nucl .
phys b407 ( 1993 ) 3 .
n.brown and w.j .
stirling , phys .
b 252 ( 1990 ) 657 .
s. catani , yu.l .
dokshitzer and b.r .
webber , phys .
b322 ( 1994 ) 263 . m. dasgupta and g.p . salam , phys .
b512 ( 2001 ) 323 .
m.dasgupta and g.p .
salam , jhep 0203 ( 2002 ) 017 .
a. banfi , g. marchesini and g.e .
smye , jhep 0208 ( 2002 ) 006 .
r.b . appleby and m.h .
seymour , jhep 0212 ( 2002 ) 063 .
berger , t. kucs and g. sterman , phys
d68 ( 2003 ) 014012 .
dokshitzer and g. marchesini , jhep 0303 ( 2003 ) 040 . m. dasgupta and g.p . salam , jhep 0208,(2002 ) 032 .
a.banfi , yu.l .
dokshitzer , g. marchesini and g. zanderighi , phys .
b508 ( 2001 ) .
n. kidonakis , g. oderda and g. sterman , nucl .
phys b531 ( 1998 ) 365 ; e. laenen , g. oderda and g. sterman , phys .
lett . b438 ( 1998 ) 173 .
r. bonciani , s. catani , m.l . mangano and p. nason , hep - ph/0307035 ( 2003 ) .
a. banfi , g.p .
salam and g. zanderighi , preprint lpthe-03 - 12 , hep - ph/0304148 .
a. banfi , g.p .
salam and g. zanderighi , jhep 0201 , ( 2002 ) , 18 . | we provide a brief review of the current status of resummed predictions for qcd final state observables such as event shapes and jet rates in a variety of different hard processes .
particular emphasis is given to more recent developments such as the study of non global observables , development of generalized resummation formulae and resummations for new types of event shape variables . |
device - to - device ( d2d ) communications @xcite have emerged as a promising paradigm for 3gpp lte - advanced ( lte - a ) networks , which provide mobile wireless connectivity , reconfigurable architectures , as well as various wireless applications ( e.g. , network gaming , social content sharing and vehicular networking ) for better user experience . with d2d communications , nearby devices in a cellular network
can communicate with each other directly bypassing the base stations .
conventional d2d communications commonly refer to direct information exchanges among devices in human - to - human and machine - to - machine communications , without the involvement of wireless operators .
however , conventional d2d technologies can not provide efficient interference management , security control and quality - of - service guarantee @xcite .
recently , there is a trend towards operator - controlled d2d communications to facilitate profit making for operators as well as better user experience for devices @xcite .
this paper considers a multi - hop lte - a network consists of devices deployed by multiple operators . in this network , an efficient approach to improve the end - to - end throughput
is to enable cooperative sharing of idle devices among the operators for multi - path routing @xcite .
the cooperation can increase throughput for the devices because a cooperative relay may substantially lead to improved network capacity .
accordingly , a larger amount of user traffic demand can be supported , which will lead to higher aggregated revenue for operators . in this cooperation
, each operator needs to decide on which operators to cooperate with to maximize profit and , given the cooperation behavior of operators .
then , the devices from cooperative operators need to make decision on which devices to cooperate with to maximize throughput .
we call the formation of this interrelated operator cooperation and device cooperation as a hierarchical cooperation problem , which is the main focus of this paper .
the hierarchical cooperation gives rise to two major concerns .
firstly , what is the stable coalitional structure desirable for all operators so that none of operators is willing to leave the coalition ? secondly , what is the stable network structure for cooperative devices to perform multi - path routing ?
this paper addresses these two concerns by formulating a layered game framework to model the lte - a network with operators and devices being the players in the upper and lower layer , respectively .
previous work has also considered game - theoretic framework with hierarchies / layers , e.g. , @xcite in the cognitive radio networks , @xcite in two - tier femtocell networks and @xcite in wrns .
however , most of the works considered the competition relationship between different layers , which belongs to the stackelberg game concept @xcite . in the proposed layered game framework , different layers interact to improve the benefit of each other cooperatively .
we adopt the concepts of an extended recursive core and nash network as the solutions for the proposed games in the upper layer and lower layer , respective . to the best of our knowledge ,
this is the first work to introduce the application of extended recursive core in wireless communications .
we consider an lte - a network consisting of a number of devices belonging to multiple operators .
we denote the set of operators as @xmath0 , and the set of devices of operator @xmath1 as @xmath2 .
the operators are willing to form overlapping coalitions to maximize their individual profits .
an overlapping coalitional structure for a number of operators can be defined in a cover function as the set @xmath3 which is a collection of non - empty subsets of @xmath4 such that @xmath5 and the sub - coalitions could overlap with each other .
@xmath6 is the total number of coalitions in collection @xmath7 .
let @xmath8 denote the set of coalitions that operator @xmath9 belongs to . for multiple access at every hop
, we consider an ofdma - based transmission . in an operator coalition @xmath10 , each relay device @xmath11 not only needs to support internal flow transmission demand , but also can serve as a relay for other devices from cooperative operators @xmath12 . due to the limited transmission power of each device , multi - hop relaying is adopted to route flow sessions from source devices to destination devices .
since single path routing is too restrictive for satisfying traffic demand , we assume each flow session can be split for multi - path routing if necessary .
figure [ model ] illustrates an example for the studied system model and the corresponding notations . in this example
, the lte - a network is composed of @xmath13 devices deployed by @xmath14 different operators , i.e. , @xmath15 , @xmath16 , @xmath17 and @xmath18 . from fig .
[ model ] , there is a multi - path flow sourced from @xmath19 to @xmath20 and a single - path flow sourced from @xmath21 to @xmath22 . from the chosen links , the final coalitional structure of the operators is @xmath23 .
let @xmath24 denote the set of flow sessions from operator @xmath9 .
@xmath25 denote the set of nodes of flow @xmath26 .
the source and destination device of flow @xmath26 is represented as @xmath27 and @xmath28 , respectively .
we denote the link between two devices @xmath29 and @xmath30 as @xmath31
. the channel gain on a link @xmath31 can be obtained from @xcite : @xmath32 where @xmath33 is the antenna related constant , @xmath34 is the path loss exponent , and @xmath35 is the distance between devices @xmath29 and @xmath30 .
@xmath36 denotes the data rate on link @xmath37 attributed to a flow session @xmath38 .
since , for d2d communications , a flow session from a source device may traverse multiple relay devices to reach its destination device , we consider the following two cases about a device .
\1 ) if a device @xmath29 is the source or destination of flow session @xmath26 , i.e. , @xmath39 or @xmath40 , then @xmath41 where @xmath42 is the aggregated rate of flow session @xmath26 , and @xmath43 denotes the set of devices having direct link with @xmath29 .
\2 ) if a device @xmath29 is an intermediate relay device of flow session @xmath26 , i.e. , @xmath44 and @xmath45 , then @xmath46 let @xmath47 denote the capacity of link @xmath48 .
the aggregated data rate on each link @xmath31 can not exceed the link s capacity . thus we have the following constraint , @xmath49 let @xmath50 denote the maximal rate a flow session @xmath26 that is available on link @xmath37 , with the constraints in ( [ source ] ) , ( [ intermediate ] ) and ( [ rate ] ) , and @xmath51 the maximal aggregated rate of a flow session @xmath26 .
we have @xmath52 .
the aggregated rate of flow session @xmath53 is constrained by @xmath54 where @xmath55 is the rate demand of flow session @xmath26 .
we formulate a game - theoretic framework , referred to as the layered coalitional game ( lcg ) , to model the decision - making process of hierarchical cooperation between the operators and devices .
both operators and devices are assumed to be self - interested and rational .
the operators aim to maximize their individual utility , while the devices attempt to maximize their end - to - end throughput with the help of relay devices from cooperative operators . in this lcg ,
the operators need to decide on the coalitional structure and the devices need to make decisions to form a relay network structure , both in distributed ways , with the aim to improve their utilities and payoffs respectively . as both operators and devices only have limited information at their own layer , information exchange between both layers is required .
the operators need to collect the payoff information from devices , to make the decision of coalitional structure formation .
the decision of operators will then be provided to the devices for the purpose of network structure formation . in this regard , there could be multiple interactions between the operator layer and device layer . recognizing the behavior of operators and devices , we propose to use the overlapping coalition formation game ( cfg ) to model the behavior of operators and the coalitional graphical game ( cgg ) to characterize the interaction among devices , which will be introduced in section iii - b and section iii - c , respectively .
since operators and devices have different objectives and concerns during the cooperation , our next step is to define the objective functions to capture the incentives for operators and devices . for a given operator coalition @xmath56
, we define the payoff of a device @xmath57 from operator @xmath58 , which performs flow transmission , as the end - to - end throughput of the flow from device @xmath29 to device @xmath59 , which is expressed as follows , @xmath60 where @xmath61 .
a relay device @xmath30 aims to help a device @xmath29 on transmission to improve the throughput by most .
therefore , to evaluate how much the relay device @xmath30 can help to improve the throughput of the device @xmath29 , we define the payoff of the relay device @xmath30 as the difference between the throughput of the device @xmath29 with the help of device @xmath30 and that without the help of device @xmath30 , which can be expressed as follows : [ rdevice ] v(\{j } ) & = v^(j)(\{i})-v^(/j)(\{i } ) + & = _ _ h _ h _
l^(h ) ^h ^i = n(l^(h ) ) _ j _ i \{j } f_(i , j)(l^(h ) ) + & - _ _ h _ h _
l^(h ) ^h ^i = n(l^(h))_j^ _ i f_(i , j)(l^(h ) ) @xmath62 where @xmath63 represents the payoff of device
@xmath29 with the help of device @xmath30 and @xmath64 represents the payoff of device @xmath29 without the help of device @xmath59 . in the proposed lcg , operators are allowed to form overlapping coalitions to share their idle devices as relays with the aim to improve the aggregated throughput . through multi - path routing , each operator aims to improve the aggregated throughput as much as possible .
thus , in return , higher revenue for providing the flow to meet the customer demand will be rewarded for each operator .
we define the individual utility of the operator as the profit , i.e. , revenue minus the cost of devices in transmitting and forwarding a packet ( e.g. , due to energy consumption ) .
let @xmath65 denote the rewarded revenue per unit throughput achieved per unit time , and @xmath66 denote the operation cost per device @xmath29 per unit time . in this case , we assume @xmath67 .
then , given a partition @xmath68 , for an operator @xmath9 without cooperation , i.e. , @xmath69 , the utility function @xmath70 can be calculated as ( [ operator ] ) , @xmath71 the first term and the second term on the right side of ( [ operator ] ) represents the aggregated revenue for operator @xmath9 and the total device costs , respectively . while cooperation can lead to profit improvement for operators , it may also incur inherent coordination costs , such as packet overhead .
let @xmath72 denote the coalition cost incurs to operator @xmath9 for being coaliton @xmath56 .
the objective for operator cooperation through cooperative sharing of devices , is to maximize their aggregated profit , i.e. , revenue is subtracted by device operation cost and coalition cost . given the partition @xmath7
, we define the utility of an operator coalition @xmath73 as the profit of the coalition as follows , @xmath74 an overlapping cfg is formulated among operators whose interests are to satisfy its internal flows with as less cost as possible . due to interference ,
the utility of any operator is affected by not only the behavior of others in the same coalition , but also that of operators from other coalitions .
thus , the considered operator coalitional game is in a partition form since the aggregated utility of a coalition @xmath73 depends on the coalitional structure @xmath7 of all the operators @xmath4 in the network .
we introduce the framework of an overlapping cfg in partition form with non - transferable utility to model the cooperation among operators .
an overlapping cfg in partition form with non - transferable utility ( ntu ) is defined by a pair @xmath75 where @xmath4 is the set of players , and @xmath76 is a value function that maps every partition @xmath7 and every coalition @xmath77 to a real number that represents the total utility ( profit ) that players in coalition @xmath56 can obtain .
the _ strategy _ of an operator is to form the coalitions to improve its individual utility defined by ( [ operator ] ) .
note that different from non - overlapping cfg where players have to cooperate with all others in the same coalition and each player only stays in one coalition , in an overlapping cfg , each player is able to join multiple different coalitions .
the solution of the overlapping cfg is the stable overlapping coalitional structure for operators , under which no one will deviate .
to this end , we adopt the concept of _ extended recursive core _
@xcite , referred to as @xmath78 , as the solution .
@xmath78 is an extended solution of coalition formation game , which allows coalitions to overlap , accounting for externalities across coalitions . in the proposed game ,
the externalities are represented by the inter - coalition interference between devices .
deviation is a key notion for the definition of the @xmath78 . as a consequence of deviation
, a new partition will be formed .
therefore , the deviation is equivalent to the formation of a new partition . in the proposed operator coalitional game ,
let partition @xmath7 move to @xmath79 by deviation .
* complete deviation : if there exists @xmath80 and @xmath81 such that for all @xmath82 , for all coalition @xmath83 such that @xmath84 , then the player set @xmath85 performs complete deviation .
the players @xmath82 are called complete deviators . *
partial deviation : if there exists @xmath86 containing only overlapping players such that for all @xmath87 , for all @xmath88 , then the player set @xmath89 performs partial deviation .
the players @xmath87 are called partial deviators . in an overlapping cfg in a partition form ,
if a coalition of players performs a complete deviation or partial deviation , this may affect the payoffs of the remaining players .
for the remaining players , we then define the _ residual game _ as following : let @xmath90 be an overlapping cfg in partition form . if a subset of players @xmath56 has already organized themselves into a certain partition @xmath91 .
a residual game @xmath92 is an overlapping cfg in a partition form defined on a set of players @xmath93 .
the players in @xmath94 are called residuals .
the residual game is an overlapping cfg in partition form on its own . in the residual game ,
the players react to the deviation only on the set of remaining players including partial deviators which can play further deviation . given two payoff vectors @xmath95 , if @xmath96 , @xmath97 , and @xmath98 , @xmath99 , we write @xmath100 .
let @xmath101 denote an _ outcome _ of the game , where @xmath102 is a utility vector resulting from a partition @xmath7 .
let @xmath103 denote the @xmath78 of game @xmath90 , and @xmath104 denote the set of all the possible partitions of @xmath4 .
@xmath78 can be found inductively in four main steps @xcite \1 ) _ trivial game _ : given a coalitional game @xmath90 , the @xmath78 of a coalitional game with @xmath105 is composed of the only outcome with the trivial partition : @xmath106 .
\2 ) _ inductive assumption _ : given the @xmath78 @xmath107 for each game with @xmath108 players , the _ assumption _ about the residual game @xmath109 is defined as follows : @xmath110 \3 ) _ dominance _ : an outcome @xmath101 is dominated via a coalition @xmath56 if for at least one @xmath111 there exists an outcome @xmath112 such that @xmath113 .
\4 ) _ @xmath78 generation _ : the @xmath78 of a game of @xmath114 players is the set of undominated outcomes .
the concept of dominance expresses that , given a current partition @xmath7 and the respective payoff vector * x * , an undominated coalition @xmath56 represents a deviation from @xmath7 in such a way that the reached outcome @xmath115 is more rewarding for the players in coalition @xmath56 , compared to @xmath102 .
thus , @xmath78 can be seen as a set of partitions , under which the players cooperate in self - organized overlapping coalitions that provides them with the highest payoff . during the coalition formation process of the operator coalitional game
, in order to reach an outcome lies in @xmath78 , we let each operator iteratively joins and leaves the coalitions to ensures a maximum payoff ( i.e. , an undominated outcome ) . to prevent loop ,
we introduce a variable couple _ history _
@xmath116 for each operator to record all the coalitions that it has ever joined and the corresponding utility .
if the new coalition for operator @xmath9 to join has been recorded and the utility that it is about to get is the same as the _ history _ , then operator @xmath9 will maintain the current coalition set @xmath8 even if its utility will better off .
once operator @xmath9 changes its coalition set , the new coalition set is included in _
we propose the coalition formation algorithm for operators in * algorithm 1 * to reach the stable coalitional structures in @xmath78 .
* initial state * + in the starting network , the operators are partitioned by @xmath117 with non - cooperative operators . * coalition formation process * + * phase 1 * _ network discovery _
+ devices from the same operators perform the _ dynamic virtual link formation _
algorithm specified in * algorithm 2*. based on the information feedback from device layer , each operator @xmath9 calculates the corresponding utility in non - cooperative case . * phase 2 * _ coalition formation _ the operators play their strategies sequentially in a random order . *
repeat * 1 ) each operators @xmath1 sequentially engages in pair- wise negotiations with another operator @xmath118 to identify potential cooperator . during this process ,
the devices from the operator pair perform _ dynamic _ _ virtual link algorithm_. based on the information feedback , each operator calculates its potential utility .
2 ) based on the potential utility information and _ history _ , the pair of operators decides to form a new coalition @xmath119 if it ensures the utility improvement .
3 ) the operators already that have cooperation with any operator in @xmath56 update their utility .
based on the updated utility and _ history _ , they perform deviation with the operator(s ) in @xmath56 if it leads to utility improvement .
* until * any further coalition formation does not result in utility improvement of at least one operator , i.e. , conver- gence to a stable partition in the @xmath78 .
* phase 3 * _ cooperative sharing _
the operators share their relay devices with cooperative operators for multi - path routing according to the final network graph @xmath120 .
the convergence of the operator coalitional game is guaranteed due to the fact that @xmath121 the total number of possible partitions with overlapping coalitions is finite , @xmath122 the transition from a partition to another leads to the increase of individual utility , and @xmath123 the game contains mechanism to prevent the operators to re - visit a previously formed coalitional structure . as a result , each cooperation buildup and breakup will lead to a new partition . as the number of the partition can be visited
is finite , the game is guaranteed to reach a final partition , under which the utility of each operator can be no longer increased .
furthermore the last partition lies in the @xmath78 because @xmath121 during the coalition formation process only the partitions that bring improvement of individual utility for each operator are formed , and @xmath122 in the final convergent partition , there are no dominated coalitions which the operators are better off by deviating from . in the cgg , the source devices need to play _ transmission strategy _ , while the relay devices need to play not only transmission strategy but also _ relay strategy_. the transmission strategy for a source device @xmath29 is to send a link establishment proposal to a relay @xmath124 which can help to improve its payoff defined in ( [ device ] ) .
the relay strategy for a relay device @xmath30 is to accept or reject a link proposal from a transmitting device @xmath125 which increases its payoff defined in ( [ rdevice ] ) .
once the relay device accepts the link establishment proposal , it will need to play transmission strategy like a source device .
we consider that an device @xmath29 can have multiple incoming and outgoing flows simultaneously , constrained by ( 2 ) , ( 3 ) and ( 4 ) , and the maximum transmit power is limited to @xmath126 .
we denote @xmath127 the strategy space which consists of all the strategies of device @xmath29 .
when device @xmath29 plays strategy @xmath127 , while all other devices maintain their current strategies denoted by a vector @xmath128 , the resulting network graph is denoted by @xmath129 .
all the devices are considered to be myopic in the sense that each device responds to improve the payoff given the current strategies of the other devices .
that is , each devices plays myopically without the foresight of the future evolution of the network . based on this
, we define the concept of _ best response _ for devices as follows : a strategy @xmath130 is a best response for a device @xmath131 , if @xmath132 , @xmath133 . based on the concept of best response
, we introduce the myopic dynamics algorithm for the proposed cgg shown in * algorithm 2*. we define an iteration as a round of myopic plays during which each device chooses to play its current best response @xmath134 sequentially in a random order with the aim to maximize its payoff given the current strategies of others .
the dynamic virtual link formation process may consist of one or more iterations , as the best strategy of each device may change over time .
all the devices play best strategies based on their flow demands .
each source device has a certain self - generated flow demand , while the flow demand of a relay device is equal to its incoming throughput . to meet the flow demand ,
the device on transmission may propose a link establishment proposal to multiple relay devices .
the iteration stops when either all the flow demands of devices are satisfied or none of the devices can unilaterally change its strategy to improve its payoff . in other words ,
when the algorithm converges , it results in a network in which none of the devices can unilaterally improve its throughput .
this is referred to as a nash network , which gives the stability concept of the final network structure @xmath120 . specifically , the nash network is defined as follows .
[ nash ] a network graph @xmath135 with @xmath136 denoting the set of all notes , i.e. devices , and @xmath137 denoting the set of all edges , i.e , links between pairs of devices , is the nash network in its strategy space @xmath138 , @xmath139 , if no device @xmath131 can improve its payoff by a unilateral change in its strategy @xmath140 . in the nash network , all the links are chosen based on the best responses of devices and are thus in the nash equilibrium . in a network with finite number of devices ,
the final network structure @xmath120 resulting from the proposed cgg is a nash network . *
initial state * + in the starting network , each source device transmits directly to its destination device .
* network structure formation process * + * repeat * * phase 1 * _ dynamic virtual link replacement _ the devices play their strategies sequentially in a random order .
* repeat * 1 ) during each iteration @xmath141 , every device on transmis- sion performs pairwise negotiation with other idle devices from cooperative operators and calculates its potential payoff improvement under cooperation .
2 ) after negotiation , each device @xmath29 plays its best response @xmath142 , based on its flow rate demand . *
until * none of the devices can further improve its payoff by a unilateral change in its strategy
. * phase 2 * _ feedback _ each device @xmath29 sends the information about the link @xmath143 back to its operator for a coalition formation decision . * until * acceptance of the convergent network structure by all the operators which perform * algorithm 1 * , i.e. , * algorithm 1 * converges . *
phase 3 * _ multi - hop routing _ all the source and relay devices perform multi - path routing according to the final network structure @xmath120 .
for simulation , we consider an lte - a network with a tdd - ofdma scheme , locating within a @xmath144 area . the bandwidth available in this network
is @xmath145 .
the maximum transmit power of each device is @xmath146 .
we calculate the capacity of each link according to shannon capacity .
the noise level is @xmath147 . for the wireless propagation , as in @xcite
, we set the path loss exponent @xmath148 and antenna related constant @xmath149 .
four operators are considered in this lte - a networks . for each operator , there is one internal flow session . for each flow session ,
the source device and destination device are randomly selected .
the number of devices from each operator is varied from @xmath14 to @xmath13 .
the results presented in this section are averaged over @xmath150 times of run with random location of devices .
we first examine the convergence speed of the cgg at the device layer . to this end , we set the coalition cost to be @xmath151 . in this case , the grand coalition is always one of the stable coalitional structures .
hence , the simulation with this setting is equivalent to performing the cgg given the coalitional structure of grand coalition at the operator layer .
each flow is assigned with a random rate requirement within @xmath152 kb / s$ ] . in fig .
[ average_iteration ] , we show the average and the maximum number of iterations required till convergence of the algorithm from the initial network structure ( i.e. , direct transmission ) , as a function of the number of devices
. as expected , with the increase of the number of devices , more interactions among devices is required for the cgg to converge . observing from fig .
[ average_iteration ] , we find that when the total number of devices varies from @xmath153 to @xmath154 , the average and maximum numbers of iterations vary from @xmath155 and @xmath156 to @xmath157 and @xmath158 , respectively .
thus , on average , the convergence speed of the algorithm is satisfactory .
we then examine the proposed lcg in a lte - a network with uniform distribution of devices .
the revenue obtained in a time unit for successfully transmitting a unit throughput ( i.e. , @xmath159 ) is @xmath160 .
the operation cost @xmath66 is considered to be the power consumption cost .
for each device to transmit or forward flow in a time unit , the cost is @xmath161 _ per _ watt .
we assume that when the operators make cooperation with another , a fixed cost @xmath162 is incurred to both of them .
thus , the total coalition cost afforded by operator @xmath9 for being in coalition @xmath56 is @xmath163 .
the total coalition cost for operator @xmath9 can be calculated by @xmath164 .
we set @xmath165 . in fig .
[ average_utility ] , we evaluate the efficiency of the proposed lcg by presenting the average aggregated utility of all operators achieved as a function of the number of total devices . for comparison , we also study an lcg variant ( labeled as lcg - variant " ) which substitutes the proposed cfg at the operator layer with a coalition formation game only enabling non - overlapping coalitions .
we use the _ recursive core _ as the stability concept for the partition of operators and adopt a solution similar to the merge - only algorithm proposed in @xcite to find a stable partition .
moreover , we use the result of non - cooperative case ( labeled as non - cooperative " ) , in which the operators work independently with each individual performing cgg among its own devices , as the lower bound performance
. we can observe from the figure that cooperation brings significant performance gains over non - cooperative case .
the proposed lcg employing overlapping cfg outperforms the lcg variant with non - overlapping cfg .
this is because the overlapping cfg allows more freedom in coalitional structure formation for potential utility improvement .
in addition , we can observe that the performance gap between lcg and the lcg variant increases with the number of devices in the network .
the hierarchical cooperation in lte - a networks is a promising solution for high speed data transmission and wide - area coverage . in this paper , we have presented a game - theoretic framework to model the hierarchical cooperation problem .
specifically , we have proposed a layered coalitional game ( lcg ) to model the cooperation behavior among the operators and devices in the different layers of lte - a networks .
the concept of extended recursive core has been advocated as the solution of stable coalitional structures .
we proposed an overlapping coalition formation game for operators to find a stable coalitional structure lies in the extended recursive core that benefits all the cooperative operators . while , a coalitional graphical game has been introduced for devices to form the stable network structures for multi - path routing .
numerical results have shown that the proposed lcg yields notable gains relative to both the non - cooperative case and a lcg variant . the future work will characterize the performance gap between the proposed lcg and the optimal solutions obtained by centralized approaches .
this work was supported in part by singapore moe tier 1 ( rg18/13 and rg33/12 ) , k. doppler , m. rinne , c. wijting , c. ribeiro , and k. hugl , device - to - device communications as an underlay to lte - advanced networks , " _ ieee communication magazine _ , vol .
42 - 49 , dec . 2009 .
y. xiao , g. bi , d. niyato and l. a. dasilva , a hierarchical game theoretic framework for cognitive radio networks , " _ ieee journal on selected areas in communications - cognitive radio series _ ,
2053 - 2069 , november 2012 .
x. kang , r. zhang , and m. motani , price - based resource allocation for spectrum - sharing femtocell networks : a stackelberg game approach , " _ ieee journal on selected areas in communications _ ,
538 - 549 , 2012 .
f. meshkati , h. v. poor , s. c. schwartz , and n. b. mandayam , an energy - efficient approach to power control and receiver design in wireless data networks " , _ ieee transaction on communication _ , vol .
1885 - 1894 , november 2005 .
s. guruacharya , d. niyato , m. bennis , and d. i. kim ,
dynamic coalition formation for network mimo in small cell networks , _ ieee transaction on wireless communication _ , vol .
5360 - 5372 , oct . | device - to - device ( d2d ) communications , which allow direct communication among mobile devices , have been proposed as an enabler of local services in 3gpp lte - advanced ( lte - a ) cellular networks .
this work investigates a hierarchical lte - a network framework consisting of multiple d2d operators at the upper layer and a group of devices at the lower layer .
we propose a cooperative model that allows the operators to improve their utility in terms of revenue by sharing their devices , and the devices to improve their payoff in terms of end - to - end throughput by collaboratively performing multi - path routing . to help understanding the interaction among operators and devices , we present a game - theoretic framework to model the cooperation behavior , and further
, we propose a layered coalitional game ( lcg ) to address the decision making problems among them .
specifically , the cooperation of operators is modeled as an overlapping coalition formation game ( cfg ) in a partition form , in which operators should form a stable coalitional structure . moreover ,
the cooperation of devices is modeled as a coalitional graphical game ( cgg ) , in which devices establish links among each other to form a stable network structure for multi - path routing .
we adopt the extended recursive core , and nash network , as the stability concept for the proposed cfg and cgg , respectively .
numerical results demonstrate that the proposed lcg yields notable gains compared to both the non - cooperative case and a lcg variant and achieves good convergence speed . _
keywords- _ d2d communications , lte - advanced network , layered coalitional game , coalitional structure formation , multi - path routing , extended recursive core , coalitional graphical game , nash network . |
in many parts of high energy physics like early universe physics and heavy ion collisions one wants to follow quantum field systems in time .
typical phenomena include reheating after inflation : after a phase of parametric resonance most energy is in a small part of the spectrum , a non - thermal initial condition . in order to study these systems and to describe phase transitions and thermal fluctuations between classical minima ,
non - perturbative approximation schemes are needed .
examples of high temperature approximations , that have been used successfully are the classical approximation@xcite and the hartree or gaussian approximation@xcite .
the latter assumes a gaussian density matrix , such that all information is contained in the 1- and 2-point functions , higher point functions can be factorised using wick s theorem .
instead of the commonly used hartree approximation , which has a spatially constant mean field we used an improved version , allowing for _ inhomogeneous _ configurations .
we studied this approximation in a scalar @xmath1 theory in @xmath2 dimensions : @xmath3 in the gaussian approximation we can expand the operator field in mode functions : @xmath4 where the _ time - independent _ @xmath5 and @xmath6 satisfy the usual commutation relations .
we can choose them such that : @xmath7 for a free thermal system the @xmath8 have a bose - einstein form and the @xmath9 are plane waves , with suitable normalisation .
the heisenberg equations of motions in the gaussian approximation in eq .
( [ eq : eom_phi ] ) while omitting eq .
( [ eq : eom_mode ] ) .
the large @xmath10 approximation leads to @xmath11 in eq .
( [ eq : eom_phi ] ) and eq .
( [ eq : eom_mode ] ) .
we chose not to use large @xmath10 to avoid problems with would - be goldstone bosons in @xmath2 dimensions . ]
become : @xmath12 } & \phi(x , t ) & = 0 , \label{eq : eom_phi } \\
\mbox { [ } \partial_t^2 - \triangle + \mu^2 + 3 \lambda \phi^2(x , t ) + 3 \lambda \sigma(x , t ) \mbox { ] } & f_\alpha(x , t ) & = 0 , \label{eq : eom_mode}\end{aligned}\ ] ] @xmath13 we mention that there exists an effective hamiltonian which leads to the above equations , suggesting that in the end the system will equilibrate classically according to this @xmath14 .
but we hope that this has a timescale much larger than that of all interesting processes .
[ [ initial - conditions ] ] initial conditions + + + + + + + + + + + + + + + + + + we need to specify the initial fields and their conjugate momenta , which together with the @xmath15 specifies an initial gaussian density matrix . by averaging over different runs with different initial conditions we can build non - gaussian density matrices , allowing for much more general initial conditions .
only the time evolution is evaluated with gaussian density matrices .
we will take the mean field in its zero - temperature minimum , with all energy in a few of its momentum modes , like after parametric resonance : @xmath16 where the @xmath17 s are random phases and @xmath18 is of the order @xmath19
. we will average over several runs , typically 10 - 20 .
for simplicity we will take the quantum modes equal to the zero - temperature vacuum form : @xmath20 with @xmath21 , @xmath19 chosen as the zero - temperature mass and @xmath22 the size of the system . to put the system on a computer we discretise it on a space - time lattice . in the limit of ( spatial )
lattice spacing @xmath23 to zero , the mode sum @xmath24 occurring in the equations of motion is logarithmically diverging , but this can be fixed by a simple mass renormalisation : @xmath25 . [ [ thermalisation ] ] thermalisation + + + + + + + + + + + + + + if @xmath26 is small and the system is not too far from equilibrium the interaction of the mode functions with the typically _ inhomogeneous _ mean field can be viewed as scattering of particles via the mean field . because of this scattering they will hopefully thermalise according to a thermal i.e. bose - einstein distribution . to check this ,
a definition of particle number is needed . for a free thermal system
the 2-point functions read : @xmath27 we want to define an @xmath28 and @xmath29 in a similar way as eq .
[ eq:2pnt ] .
this can be done by coarsening the measured 2-point functions over the initial conditions , a time interval , and space .
in fig . [ fig : phi ] we plotted for a typical run ( @xmath30 lattice points , size @xmath31 , @xmath32 ) the spatial averages of the mean field and the energy densities as a function of time . we see that the energy , initially completely in the mean field , is going to the modes with a timescale of the order 100 in mass units .
the total energy is conserved up to a few percent . the energy staying in the mean field around 20% is much more than expected from classical equipartition : around @xmath33 for @xmath30 mode functions .
[ fig : distrib ] shows a plot of the calculated particle numbers obtained after averaging over whole space , approximately 5 oscillation periods and 10 configurations . on the left
the situation is plotted at shorter times , on the right at larger times .
we see that starting at low momenta , a bose - einstein distribution is emerging
. only the zero mode is showing deviations
. runs at twice this energy density show very similar results , but with faster thermalisation . at large times deviations from the straight line
be s start to show up , which we interpret as signs of classical equipartition . for strong coupling or after a long times we see the emergence of an offset , which might be a chemical potential following from approximate particle conservation .
we finally mention checks with different kinds of initial conditions ( not presented here ) , which also show thermalisation according to a bose - einstein distribution .
using our hartree ensemble approximation with inhomogeneous mean fields , we find thermalisation according to a bose - einstein distribution , starting in the low momentum modes
. the time - scales heavily depend on the size of the coupling and the energy density .
signs of classical equilibration occur only at timescales much larger than that for the onset of the bose - einstein thermalisation , so even if they will prevent the be thermalisation to reach arbitrary high values of @xmath34 , they do not influence processes with an intermediate timescale .
grigoriev , d. , and rubakov , v. , _ nucl .
* b299 * , 67 - 78 ( 1988 ) aarts , g. , bonini , g.f . , and wetterich , c. , ` hep - ph/0007357 ` mihaila , b. , athan , t. , cooper , f. , dawson , j. , and habib , s. , ` hep - ph/0003105 ` | using an improved version of the hartree approximation , allowing for ensembles of inhomogeneous configurations , we show in a @xmath0 theory , that initially the system thermalises with a bose - einstein distribution .
for later times and larger couplings we see deviations . |
low luminosity gamma - ray bursts ( _ ll_grbs ) , or sub - energetic grbs , are a subsample of grbs that emit @xmath0 @xmath1 erg / s of gamma - rays or hard x - rays .
this is several orders of magnitude below the typical luminosity of long grbs .
additionally , these bursts are characterized by a single pulsed smooth light curve and by a very soft spectrum . even though _ ll_grbs are more numerous than typical long grbs @xcite , because of their low luminosity only a few of these _ ll_grbs have been discovered to date .
these include grb 980425 @xcite , grb 031203 @xcite , grb 060218 @xcite and grb 100316d @xcite .
_ ll_gbrs also have a radio afterglow , which indicates a comparable energy in mildly relativistic ejecta @xcite .
this energy is only a small fraction of the total explosion energy in the supernovae ( @xmath2 @xmath3 erg ) that follow _ ll_grbs .
several authors have suggested that _
ll_grbs are produced by shock breakouts @xcite .
in particular , @xcite have shown that _
ll_grbs jets are not powerful enough to punch a hole in a stellar envelope and as such they can not be produced via the common collapsar model for long grbs . at the same time , by studying the emission from a relativistic shock breakout , @xcite have provided strong evidence that the `` prompt '' emission in _
ll_grbs is produced by this mechanism .
the breakout of a shock that travels outward through a stellar envelope produces the first observable light from a stellar explosion .
as the shock crosses the stellar envelope , and as the envelope s density decreases outwards , the shock accelerates and it breaks out at a larger velocity than it had in the stellar interior , carrying , however , only a small fraction of the explosion energy . before the breakout , the shock is radiation - dominated . at the time of the breakout , all photons in the transition layer escape and produce a short and bright flash .
newtonian shock breakout has been explored before by many authors @xcite .
for relatively slow shocks , the radiation in the shock breakout layer is in thermal equilibrium and it produces a signal that peaks in the uv . for faster radiation - mediated shocks ,
the radiation is not in thermal equilibrium @xcite , and it produces a signal that peaks in x - rays @xcite .
the signal from a relativistic shock breakout has only been recently studied by @xcite , and it has been shown to produce a flash of soft gamma - rays .
in particular , they found a `` relativistic breakout closure relation '' between the breakout energy , temperature and duration .
remarkably , the observed 4 _ ll_grbs satisfy this relationship , strengthening their interpretation in the context of the relativistic shock breakout model . following the relativistic shock breakout the envelope expands and accelerates in a way such that outer parts of the envelope are faster and less energetic , and inner parts are slower and more energetic .
this ejecta interacts with the circumstellar medium and drives an external shock , which accelerates particles and generates synchrotron emission . as slower and more energetic material catches up with the decelerating ejecta it re - energizes the blast wave . in this paper
, we estimate the expected synchrotron afterglow signal from the forward shock that arises from the interaction of these ejecta with the circum - burst material .
we then compare the expected signal with the available multiwavelength _
ll_grb afterglow data .
this can serve as an additional test of the relativistic shock breakout model for the `` prompt '' signal in _ ll_grbs .
the expected afterglow from a relativistic shock breakout is similar to the afterglow that arises from the long - lasting blast wave in regular grbs @xcite .
the main difference arises from the fact that within the shock breakout the blast wave is continuously re - energized by slower ejecta .
the relativistic shock breakout model provides the distribution of the kinetic energy of the ejecta as a function of its velocity ( or its lorentz factor ) . using this distribution and the `` refreshed shock model '' considered before in the context of grbs ( e.g. , @xcite ) , we estimate the resulting dynamics of the blast wave and its emission .
we note that the relativistic shock breakout scenario and the interaction of the ejecta with the external medium has been considered before by @xcite , and recently considered in the context of the breakout from neutron star mergers @xcite .
however , in this work we consider it in the context of _
ll_grbs , for which we have evidence suggesting that this mechanism takes place . moreover , for _
ll_grbs we have available multiwavelength data that allow us to put this theory to the test .
we also consider not only the relativistic phase , but also the non - relativistic phase , which is important in some _
ll_grbs , where the shock breakout is mildly relativistic .
the paper is organized as follows . in section [ dynamics ] , we consider the dynamics of the relativistic shock breakout scenario .
we calculate , in section [ afterglow_signal ] , the light curves and spectra expected as the ejected material following the breakout interacts with the external medium , both in the relativistic and the non - relativistic phases . in section [ application ] , we calculate the expected afterglow signal from _ ll_grbs using the general formalism presented in section [ afterglow_signal ] and compare it with the multiwavelength afterglow observations of _ ll_grbs . we discuss each _ ll_grb in section [ application ] .
we examine the constraints on the x - ray emitting region in section [ xray_constraints_section ] .
we discuss the implications of these results and our conclusions in section [ final_section ] .
we begin summarizing the dynamics of the shock propagation within a stellar envelope before and after the breakout . before the breakout , a relativistic shock accelerates as @xmath4 where @xmath5 . here
, @xmath6 is the pre - shock density and @xmath7 is the lorentz factor ( the subscript `` i '' stands for `` initial '' ) . we treat the whole expanding envelope as a series of successive shells . following the breakout ,
these shells expand and accelerate to a final lorentz factor given by @xmath8 the density near the stellar edge is given by @xmath9 , where @xmath10 , with @xmath11 being the stellar radius and @xmath12 the distance from the star center , and @xmath13 the polytropic index .
typically , for the stars in which shocks will be relativistic , @xmath14 .
the mass of each of the shells at the edge is approximately given by @xmath15 .
the energy of each shell after it has reached @xmath16 is @xmath17 , where @xmath18 is the speed of light . using eq .
( [ gamma_f ] ) , we can express @xmath19 as a function of @xmath16 and find @xmath20 as a function of @xmath16 as follows @xmath21 where we have used @xmath22 in the last expression .
this expression is consistent with , e.g. , @xcite and @xcite .
as each one of the shells interacts with the circumstellar medium , they will drive a forward and a reverse shock .
particles will be accelerated in these shocks and radiation will be produced . in the scenario
considered above , shells will continuously inject energy to the blast wave , since slower shells contain more energy than faster ones and they catch up with the decelerating ejecta ( e.g. , @xcite ) .
the energy injection , in this scenario is typically characterized as @xmath23 .
the results of the last section show that , for a relativistic breakout from a stellar envelope @xmath24 where we have used @xmath22 in the last expression . at the time of breakout ,
most of the energy is emitted by a shell with pre - explosion optical depth of order unity .
we denote this shell as the _ breakout shell _ , and define its final lorentz factor as @xmath25 and its final energy as @xmath26 .
subsequent shells will be more energetic but slower than the breakout shell .
the final lorentz factor of the breakout shell is crucial in determining the total bolometric energy radiated from the breakout and the duration / temperature of the breakout emission .
it allows for the derivation of a `` relativistic breakout closure relation '' @xcite .
the optical depth of the breakout shell , @xmath27 , with @xmath28 @xmath29 g@xmath30 , implies that the mass of the breakout shell is @xcite : @xmath31 where @xmath32 is the stellar radius in units of @xmath33 . using @xmath34 we find @xmath35
therefore , we can normalize the proportionality relation eq .
( [ e ] ) using @xmath26 and @xmath25 as @xmath36 in summary , by identifying the emission from the breakout and using the relativistic breakout relation , we can determine @xmath25 and the breakout stellar radius .
with these two quantities , we can determine @xmath26 using eq .
( [ energy_norm ] ) and determine how much energy is injected to the blast wave using eq .
( [ e_vs_gamma ] ) .
the breakout shell , with energy @xmath26 and lorentz factor @xmath25 , will decelerate in a time ( see , e.g. , @xcite ) : @xmath37 as it interacts with an external density given by @xmath38 ( not to be confused with the polytropic index ) , where @xmath39 is the distance from the center of the explosion . for a constant density medium ( ism )
@xmath40 , and for a wind medium @xmath41 , where @xmath42 is the wind parameter g cm@xmath30 , where @xmath44 is the wind mass - loss rate and @xmath45 is its velocity .
the reference value for @xmath46 was scaled using @xmath47 m@xmath48 yr@xmath30 and @xmath49 km s@xmath30 . ] in units of @xmath50 g cm@xmath30 .
the slower / more energetic shells will catch up with the faster / less energetic ones when the latter have decelerated to a comparable lorentz factor . at this point , the lorentz factor of the shocked material is @xcite : @xmath51 and the blast wave energy increases as @xmath52 where time @xmath53 is the observer s time since breakout .
thus , due to the energy injection of slower / more energetic shells , the blast wave lf decrease with time slower compared with the case when no energy injection proceeds ( @xmath54 ) . for the case considered here , @xmath55 , the lf decreases as @xmath56 ( @xmath57 ) for ism ( wind ) medium .
meanwhile , the blast wave energy increases as , see eq .
( [ e_vs_time ] ) , @xmath58 ( @xmath59 ) for the ism ( wind ) case . the _ relativistic phase _ of the afterglow , for the typical parameters considered here , will last until a time , @xmath60 , when @xmath61 in eq .
( [ gamma_f_vs_time ] ) : @xmath62 afterward , the energy injection of non - relativistic shells will be important .
we note that acceleration continues beyond the breakout shell , up to the outermost shell , as long as the temperature in the shock exceeds @xmath63 kev , which is required for the creation of a significant amount of pairs ( see @xcite for details ) .
that is , although the breakout shell carries most of the energy during the breakout event , it is preceded by shells with lower energy as prescribed by eq .
( [ e_vs_gamma ] ) but that move faster .
a conservative lower limit of the _ final _ lorentz factor of the outermost shell is given by @xcite @xmath64 this faster material will decelerate earlier than @xmath65 , since the deceleration time depends much stronger on the lorentz factor than on the energy ( see eq .
[ t_dec0 ] ) .
these estimates are limited to @xmath66 , which corresponds to @xmath67 ( see @xcite ) .
after a time @xmath60 the blast wave velocity becomes non - relativistic .
slower shells continue to inject energy to the blast wave , similarly as in section [ dynamics ] , but the shock velocity instead of the lorentz factor
increases as @xmath68 , see eq .
( [ acceleration_lf ] ) , where @xmath69 @xcite . following a similar procedure as in section [ dynamics ]
, it can be shown that the energy increases as a steep function of the velocity @xmath70 @xcite @xmath71 which for @xmath22 , as considered above , yields @xmath72 , where @xmath73 in the non - relativistic phase satisfies @xmath74 . to take into account energy injection to the external shock
we follow the same procedure as in @xcite .
the total energy in the blast wave is @xmath75 where @xmath76 is the mass of the collected external medium up to radius @xmath39
. the velocity decreases and the energy increases with time as : @xmath77 and @xmath78 for the case considered here , @xmath79 , the velocity decreases slowly as @xmath80 ( @xmath81 ) , while the blast wave energy increases as @xmath82 ( @xmath83 ) for the ism ( wind ) case .
we note that the energy injection in the non - relativistic phase will proceed until a characteristic velocity of the order of @xmath84 is reached , where @xmath20 and @xmath85 are the total explosion energy and the total ejected mass , respectively .
we turn now to estimate the spectra and light curves expected from these refreshed shocks . initially , until the deceleration of the fastest `` shell '' , see eq .
( [ gamma_max ] ) , this shell coasts at a constant velocity and the afterglow is dominated by its emission .
this phase is called the `` coasting phase '' . synchrotron emission from this phase has been obtained before ( see , e.g. , @xcite and recently , ) . here , we present the light curves for @xmath86 ( other cases can be found easily ) . for the ism [ wind ]
case , the light curves are @xmath87 \nu^2 & \mbox{$\nu < \nu_m$ } \\ t^2 [ t^{5/2 } ] \nu^{5/2 } & \mbox{$\nu_m < \nu < \nu_a$ }
\\ t^3 [ t^{-\frac{(p-1)}{2 } } ] \nu^{-\frac{(p-1)}{2 } } & \mbox{$\nu_a < \nu < \nu_c$ } \\ t^2 [ t^{-\frac{(p-2)}{2 } } ] \nu^{-\frac{p}{2 } } & \mbox{$\nu_c < \nu$. } \end{cases}\ ] ] since the velocity during this phase is constant in time , these light curves are valid both for a relativistic and a non - relativistic shell . the light curves and spectra presented in the next two subsections are valid _ after _ the deceleration of the fastest `` shell '' , when the blast wave energy starts to be refreshed by slower moving ejecta .
using the formalism presented in @xcite one can calculate the spectra and light curves of the forward shock synchrotron emission when energy injection occurs as described above ( that is , for a given set of parameters : @xmath26 , @xmath25 and @xmath88 ) .
@xcite do provide a rough estimate of the synchrotron emission from the reverse shock , however , in order to calculate this emission more accurately , one needs to consider the exact numerical calculation of the strength of the reverse shock presented in @xcite for the relativistic case and in @xcite for the non - relativistic ( newtonian ) case .
the density behind the reverse shock is larger than that behind the forward shock , however , the reverse shock will peak at a much lower frequency than that of the forward shock .
for this reason , we ignore the emission from the reverse shock and focus on the forward shock emission . in section
[ application ] , we will show that the forward shock model fits the radio data of _ ll_grbs , and thus , the accompanying reverse shock synchrotron emission would be at even lower frequencies . for a given set of afterglow parameters : density ( @xmath38 ) , power - law index of energy distribution of electrons ( @xmath89 ) and microphysical parameters ( @xmath90 and @xmath91 )
, we can use @xcite to predict the expected afterglow emission at any wavelength and time .
we follow a more precise procedure , which is to calculate the energy of the blast wave at any given time and , with it , calculate the forward shock emission using @xcite , since this last work presents more precise numerical coefficients in the forward shock synchrotron emission calculation .
the predicted spectrum will be given by the synchrotron spectrum of a power - law distribution of electrons .
it is characterized by three frequencies : the self - absorption frequency , @xmath92 , the minimum frequency , @xmath93 , and the cooling frequency , @xmath94 . to sketch the calculation
, we provide the scalings of the characteristic frequencies as a function of time and energy for a general density stratification ( for compactness ) , although we focus on the interstellar medium ( ism ) and wind cases .
the energy at a given time , as a function of @xmath88 and @xmath95 , is given by eq .
( [ e_vs_time ] ) . for the case of @xmath86
the peak flux is @xmath96 at @xmath93 .
the specific flux at each of the synchrotron segments is @xmath97 , @xmath98 , @xmath99 , @xmath100 .
we have ( see , e.g. , @xcite ) @xmath101 for the case of @xmath102 , the peak flux is at @xmath92 ( see below ) and given by @xmath103 .
the specific flux at each of the synchrotron segments is @xmath97 , @xmath104 , @xmath99 , @xmath100 .
both @xmath93 and @xmath94 behave as above .
however , @xmath92 is given by @xmath105 light curves at a fixed frequency are obtained by constructing the synchrotron spectrum with the given characteristic frequencies and letting them evolve with time @xcite .
for simplicity , we do not consider the smoothing between different power - law segments in the synchrotron spectrum ( see , e.g. , @xcite ) .
we now provide the `` closure relations '' , which is a relation between the temporal decay index , @xmath106 , and the spectral index , @xmath107 ( not to be confused with the velocity of the ejecta ) , where the specific flux at frequency @xmath108 is defined as @xmath109 .
we focus on the light curves above the peak frequency ( the other regimes can be found easily with the equations above ) , and find @xmath110 these expressions are consistent with @xcite .
for example , for @xmath111 and @xmath55 , the ism ( wind ) case yields decaying light curves , with the light curve for @xmath112 being steeper ( shallower ) than that of @xmath113 .
the relativistic and non - relativistic solutions should be joined at @xmath60 . at @xmath114 ,
we use the non - relativistic expressions for the energy injection discussed above and the expressions for the synchrotron characteristic frequencies presented in @xcite .
we note that this is an idealized case .
there will be a smooth transition both in the dynamics of the blast wave @xcite and in the exact shape of the synchrotron spectrum @xcite as the blast wave transitions from the relativistic to the non - relativistic phases .
in general , this will make @xmath60 larger than reported in eq .
( [ t_nr ] ) . similarly , as done for the relativistic phase , to sketch the calculation we provide the scalings of the characteristic frequencies as a function of time and energy for a general density stratification ( for compactness ) , although we focus on the ism and wind cases .
the energy at a given time , as a function of @xmath73 and @xmath95 , is given by eq .
( [ e_non_rel ] ) . for the case of
@xmath86 we have ( see , e.g. , @xcite ) @xmath115 as for the relativistic phase , for the case of @xmath116 , both @xmath93 and @xmath94 remain the same .
however , @xmath92 is given by @xmath117 the `` closure relations '' for the non - relativistic phase are given by @xmath118 for example , for @xmath111 and @xmath79 , the ism ( wind ) case yields increasing ( decaying ) light curves , with the light curve for @xmath112 being shallower than that of @xmath113 .
based on the observed properties of the prompt emission of _ ll_grbs , @xcite suggested that the shock breakout in these events is ( mildly ) relativistic . in this section
we compare the resulting afterglows , calculated with the formalism presented above and based on the estimated breakout properties of @xcite , with the afterglow observations of these _ ll_grbs . using the prompt properties of _ ll_grbs , @xcite determine the breakout shell final lorentz factor , @xmath25 , and the breakout radius @xmath39 ( see table [ table1 ] ) . with these quantities , we determine @xmath26 , the initial energy injected to the blast wave , using eq .
( [ energy_norm ] ) .
subsequent shells will inject more energy as prescribed in eq .
( [ e_vs_time ] ) until @xmath60 , when the energy injection will proceed in the non - relativistic phase as prescribed in eq .
( [ e_non_rel ] ) . given that these _ ll_grbs are accompanied by supernovae with estimated kinetic energies between @xmath2 and @xmath3 erg ,
an estimate of this energy can serve as a good quantity of the total energy in the explosion and the point at which energy injection should stop .
we note however that this energy is not reached until a time which is much longer than the time of the last available observations in the radio and x - ray bands . as in @xcite
, we divide the 4 _ ll_grbs in two pairs with different properties .
grbs 980425 and 031203 both have observed prompt peak energies in excess of @xmath63 kev @xcite . in these two grbs , significant amount of pairs are created in the shock , and ejecta are accelerated to lorentz factor larger than @xmath25 , up to @xmath119 , see eq .
( [ gamma_max ] ) @xcite .
grbs 060218 and 100316d both have observed prompt peak energies marginally consistent with @xmath63 kev @xcite . in these two grbs , significant amount of pairs
are only marginally created , and ejecta with lorentz factor larger than @xmath25 is most likely absent . in that case
the afterglow emission is produced in the coasting phase , as described in section [ coasting ] , since the deceleration of the breakout shell takes place only after very long time . moreover , since the late time radio light curves of these two grbs decrease with time , then the ejecta must interact with a wind medium ( otherwise all light curves would increase with time ) . .using the values of the breakout radius ,
@xmath39 , and the breakout shell final lorentz factor , @xmath25 , reported in @xcite we find the total energy of the breakout shell , @xmath26 , using eq .
( [ energy_norm ] ) .
the total observed isotropic energy in gamma - rays / hard x - rays , @xmath120 , is also presented for comparison . [ cols="^,^,^,^,^",options="header " , ] to calculate the expected light curves , we use the values of @xmath26 and @xmath25 from table [ table1 ] and use @xmath55 and @xmath79 as found above . , that is , a radiative stellar envelope . for a convective envelope @xmath121 , and @xmath88 and @xmath73
are determined by eq .
( [ s ] ) and following eq .
( [ s ] ) . in this case
the temporal decay of the light curves above the peak are only modified by @xmath122 according to eqs .
( [ closure1 ] ) and ( [ closure2 ] ) . ]
we fix the electron energy power - law index to @xmath111 ( we will comment on this point below for each burst ) .
we also fix the fraction of shocked energy in electrons to @xmath123 , since @xmath90 seems to be narrowly clustered around this value in afterglow modeling studies ( see figure 1 in @xcite ) . for cases
in which a satisfactory fit is not reached with @xmath124 , we consider other values .
we calculate both light curves , for an ism ( @xmath40 ) with density @xmath13 , and for a wind medium ( @xmath41 ) with wind coefficient @xmath42 .
we scan the parameter space in density , @xmath125 @xmath126 and @xmath127 for the ism and wind case , and in @xmath128 .
we look for a set of parameters that best fits the radio observations within a factor of @xmath129 .
we later check if the fit to the radio light curves also gives a good fit to the x - ray observations . in the following subsections
, we will discuss each one of the four _ ll_grbs described in table [ table1 ] in the context of the model described above .
we do not discuss them in chronological order , but in an order that reflects our overall understanding of each burst within this model .
grb 031203 was monitored extensively in the radio bands @xcite , and it was also followed in the x - ray band starting 6 hours after the burst @xcite .
we calculate the predicted radio fluxes in the ism scenario and present them in fig .
we find that the model can fit both the radio and _ also _ the x - ray observations quite well .
we find that the x - ray band before 6 days is above @xmath93 but below @xmath94 .
it decreases as @xmath130 @xcite .
the closure relations in eq .
( [ closure1 ] ) yield a consistent temporal decay with the chosen value of @xmath111 . within this model
, the reason why the x - ray light curve steepens at @xmath131 days is due to the crossing of the synchrotron cooling frequency through the x - ray observing band ( the synchrotron cooling frequency decreases with time in this case ) .
the observed x - ray spectrum at @xmath132 days is @xmath133 and @xmath134 at 6 hours and 3 days , respectively @xcite , roughly consistent with the predicted spectrum @xmath135 .
the x - ray spectrum after @xmath136 days should steepen by @xmath137 , which can be tested with the _ chandra _ data at @xmath63 days @xcite .
the blast wave energy , which changes with time as energy is injected , is @xmath138 erg at 10 days in this case ; the time dependence of the blast wave energy can be found in eq .
( [ e_vs_time ] ) . the non - relativistic phase for this burst will commence beyond 300 days , which is a factor of few larger than the time of the last observations and all the light curves would slightly rise .
we note that solutions with a slightly smaller density and larger @xmath91 also fit the data well and the non - relativistic phase might be delayed for an even longer time .
the wind case is not able to fit the radio observations , even if we allow @xmath90 to vary from the fiducial adopted value of @xmath139 .
also , the radio light curves would be steeper than @xmath140 for @xmath141 , according to eq .
( [ closure1 ] ) , which is much steeper than observed ( the radio light curves are quite flat at late times ) .
we therefore do not show a figure for this scenario .
interestingly , ir - optical observations starting 9 hr after the burst show a bright ir emission with a very steep soft spectrum , @xmath142 @xcite . this emission fades within a day , before the main component of the supernova ( sn ) light rises a week later .
the very soft spectrum of early ir emission indicates that it has , most likely , a different origin than that of both the radio and the x - rays .
it is still consistent with the radio and x - ray being both generated by synchrotron emission , as our model predicts an ir flux that is fainter by a factor of @xmath143 than the one observed .
we do not discuss here the origin of this unique ir emission , as it is most likely related to the cooling envelope phase of the sn . in that case
the bright yet fast fading signal may indicate on a progenitor composed of a compact core and an extended low - mass envelope @xcite .
@xcite have modeled the radio emission of this burst in the context of the external forward shock emission ( without energy injection ) .
they also prefer an ism with similar density as the one we adopt here .
however , energy injection via the blast wave enables us explain also the x - ray emission , which @xcite had found was a factor of @xmath144 smaller than observed .
grb 980425 was monitored extensively in the radio bands @xcite , and its x - ray data showed an unusually flat behavior for more than 100 days @xcite .
we use the scenario described above for the parameters of this _
ll_grb to model the radio light curves .
the initial blast wave energy of this _
ll_grb is the smallest of all _ ll_grbs , therefore , we need to consider the non - relativistic phase since @xmath60 falls within the available radio observations for @xmath145 @xmath126 ( @xmath146 ) in the ism ( wind ) case , see eq .
( [ t_nr ] ) .
we consider first the ism case .
the light curves above the minimum frequency in the non - relativistic phase increase with time , see eq .
( [ closure2 ] ) , which is inconsistent with the time - decreasing trend of the radio data at late times .
therefore , we consider the wind medium , for which the solution lies entirely in the non - relativistic phase . using a value of @xmath139
as discussed above , we find that the wind medium is not able to fit the radio observations .
we lower @xmath90 to a smaller value ( @xmath147 ) and find that the predicted light curves give an overall good fit to the radio observations ( however , the radio observations at @xmath148 days and the slight increase in flux at @xmath149 days are not well fit ) .
we note that the radio light curves beyond @xmath149 days are above @xmath92 and @xmath93 , but still below @xmath94 .
observations indicate that they decrease roughly as @xmath150 @xcite . using the closure relations in eq .
( [ closure2 ] ) , we find that a value of @xmath151 fits the decay of the radio observations , and we present the light curves in fig .
[ fig2 ] for this case .
the predicted x - ray emission ( which lies above @xmath94 ) is smaller than the observed value by at least a factor of a few tens .
the blast wave energy at 10 days in this case is @xmath152 erg ; the time dependence of the blast wave energy can be found in eq .
( [ e_non_rel ] ) .
interestingly , in the wind case and using the same parameters as above , but changing the value of @xmath89 to @xmath153 , we find an x - ray flux consistent with the observations , both in normalization and temporal decay .
the x - ray light curve until @xmath154 days decays as @xmath155 @xcite , but including the flux at @xmath156 days , the overall light curve decays as @xmath157 .
this is consistent with the predicted temporal decay @xmath158 , according to eq .
( [ closure2 ] ) .
the observed x - ray spectrum for @xmath159 days is @xmath160 @xcite , consistent with the predicted value of @xmath161
. the predicted radio light curve with @xmath162 would be too shallow ( @xmath163 ) , but would qualitatively fit the radio observations within a factor of @xmath129 until @xmath164 days , and the discrepancy would increase up to a factor of @xmath144 at @xmath165 days .
these results suggest that both the radio and the x - rays can be generated by synchrotron emission from a single population of electrons in the forward shock , if the electron spectrum is concave .
namely , if the electron spectrum is steep ( soft ) at low energies and shallow ( hard ) at high energies .
such spectrum is predicted by some _ fermi _ acceleration models in newtonian shocks , due to the nonlinear feedback of the accelerated particles on the shock structure ( e.g. , * ? ? ?
@xcite have modeled the radio data of grb 980425 also in the wind case and they present several solutions : one with @xmath166 and @xmath167 , another with @xmath168 , @xmath169 and @xmath170 , and they also find acceptable models with @xmath147 .
our solution for this grb has @xmath171 , however , the two models are inherently different .
@xcite allow for the energy of the blast wave to increase _ instantaneously _ by a factor of @xmath172 to explain the increase by a factor of @xmath173 in the radio light curves at @xmath149 days .
however , within out model energy is _ continuously _ injected to the blast wave and we ignore this slight increase in the radio light curves at @xmath149 days .
the lorentz factor of the breakout shell for this burst is close to unity and creation of pairs in the shock is marginal .
therefore , it is likely that no shells are accelerated to velocities faster than the breakout shell , in which case the forward shock is coasting at a constant velocity for a long time until the breakout shell decelerates . in this case
only a wind medium can fit the observed decaying radio afterglow and this is the model we consider here .
the blast wave moves with a constant lorentz factor of @xmath174 until it decelerates at @xmath154 days . during the coasting
phase the blast wave energy increases linearly with time for the wind case , and reaches @xmath175 erg at the deceleration time .
we note that solutions with a smaller density also fit the data well ( and also underestimate the x - ray fluxes ) and the deceleration phase might be delayed for an even longer time .
radio and x - ray afterglow observations of grb 060218 have been reported in @xcite .
the radio light curve at 8.46 ghz decays as @xmath176 @xcite .
we find that this radio band is between @xmath92 and @xmath94 ( @xmath93 is below @xmath92 ) .
( [ f_nu_coasting ] ) , we find that a value of @xmath177 fits the decay of the radio observations , and this is the reason we present the light curves for this value of @xmath89 in fig .
we find an overall good fit to the radio observations ( however , some upper limits at 1.43 , 4.86 and 8.46 ghz are consistent only to within a factor of @xmath129 ) .
we also find @xmath178 ghz at 5 days , and that @xmath93 is about ten times smaller than @xmath92 , consistent with the spectrum at 5 days presented in @xcite .
with that model the predicted x - ray emission is below the observed value by at least a factor of @xmath143 . as in the case of grb 980425 , we lower the value of @xmath89 to obtain a rough estimate if a concave electron spectrum may fit the radio and x - ray data simultaneously .
this increases the predicted x - ray flux , although , it makes the light curve shallower , see eq . ( [ f_nu_coasting ] ) , which is inconsistent with the observed decay @xmath179 @xcite .
nevertheless , the discrepancy between the predicted and observed x - ray flux shrinks with a smaller value of @xmath89 .
even if such a modification would be capable of explaining the x - ray flux , the predicted x - ray spectrum is @xmath180 .
this stands in contrast with the observed x - ray spectrum @xmath181 @xcite .
such a steep spectrum can not be explained within this model , unless the electron spectrum has a cut - off at energies of the x - ray emitting electrons .
this solution is contrived , and we conclude that the x - ray and the radio are most likely of different origin . if faster shells than the breakout shell are present , then the blast wave would start decelerating at much earlier times and it can be relativistic for long enough for the decreasing radio emission to be consistent also with a constant density medium .
however , in that case @xmath182 ghz @xmath183 at 5 days , which leads to a flux at 1.43 ghz that is at least a factor of @xmath184 brighter than the upper limit at that time .
this is marginal even if there is strong scintillation , but can not be completely ruled out . the x - ray emission in that case is still underpredicted by the model if a single electron power - law is assumed and the difficulty to reproduce the x - ray spectrum persists .
the last burst in our sample is grb 100316d , whose x - ray data are presented in @xcite , and additional x - ray data and the radio data are presented in @xcite .
since this burst is almost identical to grb 060218 , and it is explained with the same parameters within the relativistic shock breakout model , we use the same afterglow model , namely synchrotron emission from a coasting blast wave in a wind medium .
the radio data of this burst are sparse , available only in two frequencies and possibly affected by scintillation .
therefore , it is not very constraining and many models can fit the radio data . here
we present one such model in fig .
[ fig4 ] , which fits the radio data reasonably well .
most importantly , @xmath92 is not well constrained , e.g. , @xcite suggest that @xmath185 ghz at @xmath149 days , whereas in the model presented here this value of @xmath92 is obtained at @xmath136 days . as a result
the radius and thus the velocity of the radio emitting region can significantly vary between various acceptable models .
the model underpredicts x - ray flux compared with the observed value by a factor of more than @xmath186 . as for grb 060218 , a smaller value of @xmath89 ( concave electron spectrum ) increases the predicted x - ray flux , but the x - ray light curve is too shallow and it is inconsistent with the observed decay @xmath187 @xcite . at the same time the observed x - ray spectrum , @xmath188 @xcite , is similar to that of grb 060218 and it indicates that also here the x - ray and the radio are most likely generated by different processes .
the blast wave moves with a constant velocity and its energy increases linearly with time until it reaches @xmath175 erg at the deceleration time of @xmath189 d. we note that solutions with a smaller density also fit the data well ( and also underestimate the x - ray fluxes ) and the deceleration phase might be delayed for an even longer time . if faster shells than the breakout shell would be present , then an ism or wind medium would also marginally fit the radio data , since the data are not very constraining .
@xcite have modeled the radio emission of grb 060218 , and @xcite the one of grb 100316d in the context of the external forward shock emission .
the external densities they find are different than ours .
however , here the afterglow is produced in the coasting phase , whereas they considered a decelerating blast wave .
the x - ray afterglows in grb 060218 and grb 100316d are most likely not generated by the same process that produces their radio emission .
nevertheless , several strong constraints on the x - ray source can be derived from simple considerations .
these constraints are regardless of whether _ ll_grbs are shock breakouts or not , or even of the exact models of both the radio and the x - ray emission . in both bursts
the total energy emitted in x - rays by @xmath190 s is @xmath191 erg ( assuming isotropy ; see below ) .
thus , the first general constraint is that for any reasonable radiative efficiency , the source energy at that time is @xmath192 erg , and most likely significantly higher .
the x - ray energy output is roughly constant for every logarithmic time interval , implying that the energy constraint at @xmath193 s is lower by about an order of magnitude .
a second constraint follows from the fact that the gamma - ray emission , which led to the bursts detection , is not highly collimated , namely opening angle @xmath194 and inverse beaming factor @xmath195 ( e.g. , * ? ? ? * ) .
given that the x - ray emission is seen in both bursts , it is most likely not highly collimated as well .
this implies that the source can not be a highly relativistic narrow jet .
in fact , the radio emission implies that the x - ray source can not be highly relativistic even if it is not narrowly collimated .
the reason is that without being collimated it can not run ahead of the radio source , otherwise it will sweep the circum - burst medium , thereby preventing the interaction that leads to the radio emission .
now , the radio source , at least in grb 060218 where @xmath92 is measured , is mildly relativistic .
@xcite measures @xmath196 ghz and @xmath197 mjy at 5 days for grb 060218 .
using @xcite we find that if we require that the energy in the radio emitting region is @xmath198 and we allow for a ( large ) uncertainty in the measurement of @xmath92 by a factor of @xmath173 , then these observations imply a lorentz factor of the radio emitting region @xmath199 at 5 days . ] .
therefore , the x - ray source is at most mildly relativistic , implying that the x - ray emission radius , @xmath200 , is limited to @xmath201 . on the other , hand the x - ray emitting radius can not be too low .
the line profiles of sn 2006aj ( the sn that accompanied grb 060218 ) do not show signs of anisotropy , and there is no polarization ( @xmath199 per cent at the 3-@xmath202 level ) either in the lines nor in the continuum , implying that the bulk of the sn ejecta has no strong deviation from spherical symmetry @xcite . this ejecta is about a solar mass moving at a velocity @xmath203c @xcite that lies between the observer and the center of the explosion .
no radiation that is emitted behind this mass can escape to the observer on a time - scale that is shorter than about a week ( since the diffusion time is too long ) .
therefore , the x - ray source must be at all time _ ahead of _ the bulk of the sn mass .
thus , together with the limit that the radio emitting region is ahead of the x - ray source , we obtain @xmath204 .
given that the afterglow is observed over two decades in time , the radius of the region where x - rays are generated must increase in time together with the sn ejecta .
this constraint implies that the x - ray source is most likely within the explosion ejecta , between the bulk of the sn mass ( the slowest part of the ejecta ) and the leading radio source ( the fastest part of the ejecta ) .
this disfavors emission models that are powered by internal dissipation in an unrelated component , such as internal shocks or magnetic reconnection in an outflow that is unrelated to the one producing the sn and the radio emission .
the constraints derived above are non - trivial and can not be easily met by most types of x - ray sources . specifically , they are most difficult to reconcile with any source that gains its energy continuously from a central engine , as has often been suggested ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the main problem in such sources is that the energy which is generated in the center must penetrate through several solar masses of sn ejecta , while the source is not relativistic and it is not narrowly collimated ( @xmath194 ) .
furthermore , this should take place without the energy being dissipated below the photosphere and without leaving any major imprint on the asymmetry inferred from the sn line profiles .
we note that a simple `` inward '' extrapolation in energy ( from the energy of the fast moving material to the sn energy ) described at the end of section [ nr phase ] does not work .
namely , the energy - velocity relation is too shallow compared to that seen in regular sne and to that expected in a simple spherical blast wave that propagates in a typical star ( e.g. , @xcite ) .
this is true even when one accounts for the shallow relativistic slope described here and in previous works ( e.g. , @xcite ) .
this probably implies that the explosion is not fully spherical ( e.g. , @xcite ) , or that the stellar structure is different than in typical sne , or a combination of the two .
a likely explanation for this point is discussed in a future work ( nakar , in preparation ) .
regardless of the reason for the shallow energy profile , the observations indicate that the asphericity , if exists , is not high ( inverse beaming factor @xmath195 ) and that the spherical approximation for the gamma - ray emission and the following blast wave is reasonable .
note that if the breakout deviates from spherical symmetry this may result in breakout velocities that are slower in some parts of the stellar surface @xcite .
but , unless the breakout asphericity is high , the breakout from most of the stellar surface will not be strongly affected ; neither the resulting gamma - rays nor the following blast wave .
we have calculated the synchrotron emission from the forward shock that arises during the interaction between the stellar ejecta and the circum - burst medium following a relativistic shock breakout .
the hydrodynamics of the shock breakout ejecta is described following @xcite , where we assumed that the progenitor has a standard radiative stellar envelope , characterized by a polytropic index @xmath22 .
for the subsequent blast wave evolution we assume that the circum - burst density profile is either constant ( ism ) or @xmath205 ( wind ) .
the main difference between this model and the standard grb afterglow model is that here energy is continuously injected into the slowly decelerating blast wave by slower ejecta whose profile is dictated by the sn shock that unbinds the stellar envelope .
once the behavior of the blast wave is determined the emission follows the standard grb synchrotron afterglow model @xcite .
we have applied this model to the four _ ll_grbs with good data , grbs 980425 , 031203 , 060218 and 100316d .
the prompt emission provides us the initial conditions ( the energy and lorentz factor profiles ) of the ejecta @xcite .
we first fit , for each burst , the free parameters of the emission model to the observed radio afterglow and then check if the same model parameters can also explain the observed x - ray afterglow .
we find that the model can successfully explain the radio emission of all four _ ll_grbs . in some _
ll_grbs , the x - ray emission can also be generated by the same population of electrons that generate the radio emission , while in others it can not . for grb 031203 , we find that the simplest model can successfully fit _ both _ the observed radio and x - ray afterglows light curves and spectra with a single set of parameters , if the external medium is an ism .
this burst is the only _
ll_grb for which we expect the blast wave producing the afterglow to be relativistic , similarly to regular grbs , and indeed the parameters we find for the external density and microphysical parameters are similar to those found in regular grbs .
this is the first time that _ both _ the prompt and the afterglow emission can be explained for a single _ ll_grb : the prompt emission is produced by the relativistic shock breakout signal and the radio and x - ray afterglow are explained by its accompanying external shock
. for grb 980425 , a model with a wind medium fits the radio data .
the density of the wind is such that the blast wave is already newtonian at the time of the first observation . using the simplest model , the same parameters that fit the radio data result in an x - ray spectrum that is too soft and an x - ray flux that is too low .
however , the model can simultaneously explain the radio and x - ray observations ( both the light curve and the spectrum ) with a minor modification that relaxes the assumption of a single power - law electron distribution . a concave electron distribution with a logarithmic derivative of @xmath206 for the radio emitting electrons and @xmath207 for the x - ray emitting electrons can fit both observations .
interestingly such a spectrum is predicted by some _ fermi _ acceleration models in newtonian shocks , due to nonlinear feedback of the accelerated particles on the shock structure ( e.g. , * ? ? ?
* ) . for grb 060218 and 100316d ,
the relativistic shock breakout model predicts a mildly relativistic breakout that is consistent with the absence of ejecta faster than the breakout shell . in that case
the breakout shell decelerates after a long time , and the observed afterglow is produced via synchrotron emission in the coasting phase .
the observed radio afterglow , which decays at late times , is consistent only with a wind medium ( in the ism case all light curves rise during the coasting phase ) .
this model is able to fit the radio data .
however , with the same parameters , it underestimates the x - ray flux .
moreover , the afterglow x - ray spectra in these two bursts are very steep . in fact
, it is very peculiar compared to other grb afterglows .
yet , the light curve in both cases ( a very bright simple power - law , roughly @xmath208 ) resembles that of regular grbs and is uncommon in ordinary sne .
the steep spectra implies , first , that it is highly unlikely that the x - rays and the radio are generated by the same population of electrons , as was already found earlier @xcite .
secondly , the combination of the x - ray spectra and light curves in these two bursts , suggests that the x - ray source is uncommon to either regular grbs or regular sne .
we stress that while we have found models that are consistent with the radio emission of all four _ ll_grbs , these are by no means the only consistent models .
first , within the synchrotron forward shock model there is enough freedom to allow for a range of parameters that fit the observations .
secondly , we did not attempt to calculate here the synchrotron emission from the reverse shock or the contribution of inverse compton emission from either the forward or the reverse shocks .
it is most likely that there are solutions where the observed radio emission is dominated by one of these components ( e.g. , reverse shock synchrotron ) .
the main puzzle that remains is the source of the x - ray afterglow in _
ll_grbs 060218 and 100316d .
we draw here some general constraints on the source of this emission , finding that it is not narrowly collimated and that it is not ultra - relativistic .
the source itself is most likely moving at velocity @xmath209c and a lorentz factor @xmath199 , and it is located at all time somewhere within the ejecta , between the bulk of the sn mass and the leading radio source .
previous authors invoked the existence of a continuous central engine to explain the x - ray afterglow @xcite , without discussing how such a source produces this emission .
the constraints we derive here show that long lasting central engine activity is unlikely to be the correct solution .
the major difficulty for this model is the need for the central engine energy , that is not narrowly collimated , to penetrate through the opaque sn ejecta without leaving a trace of anisotropy in the spectral line profiles @xcite .
additionally , the combination of the x - ray light curve and spectrum seen in these bursts is unique and does not resemble any other cases where we have a good reason to suspect that they are a result of central engine activity . in our view , given that the x - ray source is probably moving together with the ejecta , the most natural scenario is that the x - ray source is within the interaction layer that is also responsible for the radio emission .
since the x - ray and the radio do not arise from the same process , it is possible that either the long lived reverse shock ( within a refreshed shocks scenario ) or inverse compton emission can explain these unique x - ray signatures .
rbd dedicates this work to the memory of francisco pla boada .
rbd thanks useful discussions with paz beniamini , raffaella margutti , lara nava and rongfeng shen .
we thank christopher matzner for useful comments on the manuscript .
this work was supported by an erc advanced grant ( grb ) , by an erc starting grant ( grb / sn ) , by the i - core program of the pbc and the isf ( grant 1829/12 ) , isa grant and isf grants .
this work made use of data supplied by the uk swift science data centre at the university of leicester . | the prompt emission of low luminosity gamma - ray bursts ( _ ll_grbs ) indicates that these events originate from a relativistic shock breakout . in this case
, we can estimate , based on the properties of the prompt emission , the energy distribution of the ejecta .
we develop a general formalism to estimate the afterglow produced by synchrotron emission from the forward shock resulting from the interaction of this ejecta with the circum - burst matter .
we assess whether this emission can produce the observed radio and x - ray afterglows of the available sample of 4 _ ll_grbs .
all 4 radio afterglows can be explained within this model , providing further support for shock breakouts being the origin of _ ll_grbs .
we find that in one of the _ ll_grbs ( grb 031203 ) , the predicted x - ray emission , using the same parameters that fit the radio , can explain the observed one . in another one ( grb 980425 ) , the observed x - rays can be explained if we allow for a slight modification of the simplest model . for the last two cases ( grbs 060218 and 100316d ) , we find that , as is the case for previous attempts to model these afterglows , the simplest model that fits the radio emission underpredicts the observed x - ray afterglows . using general arguments ,
we show that the most natural location of the x - ray source is , like the radio source , within the ejecta - external medium interaction layer but that emission is due to a different population of electrons or to a different emission process .
radiation mechanisms : non - thermal - methods : analytical - gamma - ray burst : general |
we consider the ground state and the elementary excitations of an array of spin - peierls chains coupled by elastic and magnetic interactions .
it is expected that the effect of the magnetic interchain coupling will be to reduce the dimerization amplitude and that of the elastic coupling will be to confine the spin one - half solitons corresponding to each isolated chain .
we show that this is the case when these interactions are not frustrated . on the other hand , in the frustrated case
we show that the amplitude of dimerization in the ground state is independent of the strength of the interchain magnetic interaction in a broad range of values of this parameter .
we also show that free solitons could be the elementary excitations when only nearest neighbour interactions are considered .
the case of an elastic interchain coupling is analyzed on a general energetic consideration . to study the effect of the magnetic interchain interaction the problem
is simplified to a two - leg ladder which is solved using density matrix renormalization group ( dmrg ) calculations .
we show that the deconfinement mechanism is effective even with a significantly strong antiferromagnetic interchain coupling .
elementary excitations of antiferromagnetic systems are spin - one magnons which are bosonic in character and correspond to flipping locally the spin of one electron .
they give rise to a definite peak in the dynamical structure factor which is observed in neutron scattering measurements . in a one - dimensional antiferromagnet ,
this picture breaks down and the concept of fractionalized excitation has been proposed in the last years to explain the excitation spectra .
the excitation spectrum of a 1d antiferromagnetic heisenberg model is a paradigmatic example of a fractional quantum state .
the elementary excitations are spin-@xmath0 spinons which are topological excitations identified as quantum domain walls .
the spectral signal corresponding to the excitation of two spinons shows a highly dispersive continuum without a definite one - particle peak in the dynamic susceptibility .
measurements of the two - spinon continuum have been achieved in the quasi - one - dimensional antiferromagnetic system kcuf@xmath1 @xcite . since the discovery of the cuprate superconductors an intense activity has been carried out to see if a fractional quantum state is possible in a two - dimensional antiferromagnet .
an experimental realization seems to be found in cs@xmath2cucl@xmath3 @xcite .
some analogies appear in the excitation spectra when the magnetism is coupled to phonons in the so - called spin - peierls systems .
this coupling leads to a magneto - structural transition towards a dimerized low - temperature phase , opening a gap in the magnetic spectra .
even though the excitation spectrum of an isolated spin - phonon chain is not exactly known , an early semiclassical analysis of a bosonized field theory @xcite gives some insight to this problem .
the elementary excitation is a soliton , a topological defect which separates two different phases of the dimerized order .
it is a mixed magnetic and structural excitation carrying a spin-@xmath0 .
therefore , the dynamical response of a pure spin - phonon system should be dominated by a two - soliton continuum above the spin gap .
numerical calculations in small chains confirm this scenario @xcite .
the role of the interchain coupling should be taken into account for application to a real compound .
it was particularly analyzed in connection with neutron and optical spectra measurements in the inorganic spin - peierls system cugeo@xmath4 @xcite .
it has been concluded that solitons are always confined by the interchain coupling .
it means that no trace of solitons is present in the excitation spectrum .
different routes have been followed to analyze the effect of the interchain coupling . in one of them
a linear confinement potential between the solitons has been considered as a result of a mean field treatment of the interchain coupling @xcite .
a ladder of bound soliton - antisoliton states appears before a magnon continuum . in other
approach @xcite , a mixed lattice and magnetic excitation called _ domain _ was found in a semiclassical approximation of the bosonized theory of the two - dimensional spin - phonon system .
the domain is a triplet excitation .
excitations inside the domain give rise to a series of states before the appearance of a two - domain continuum .
experiments in cugeo@xmath4 @xcite have shown the existence of a peak separated from a continuum , in consistency with the previous pictures .
the peak could be associated with a magnon excitation and the continuum with a two - magnon continuum .
we note that by integrating out the phonon coordinates the adiabatic - antiadiabatic crossover has been studied in the one - dimensional model @xcite .
moreover , the dynamic correlation function has been studied in the antiadiabatic limit for the problem of coupled chains @xcite .
the effective one - chain problem has been solved by using exact results from the sine - gordon theory corresponding to the continuum limit of the effective magnetic problem .
the interchain interactions have been taken into account by a rpa approach .
the dynamic susceptibility has also a peak corresponding to a spin-1 excitation ( which we call a magnon ) and a continuum .
now the question arises about if free solitons could be observed in a real system .
that is to say , if it is possible to observe a two - soliton continuum above the gap in a spin - peierls material .
the complete answer to this question is a formidable problem because it implies the calculation of the dynamic spin - spin correlation function for a coupled 2d or 3d spin - phonon problem . as a first step we use a semiclassical approach in the present work .
our purpose is to show that a frustrated nearest neighbour interchain coupling in a quasi - one - dimensional spin - peierls system does not confine the in - chain solitons .
our work is also motivated by the recently identified materials tiocl and tiobr as spin - peierls compounds @xcite .
these materials deviate from a canonical spin - peierls behaviour due to the existence of an intermediate incommensurate phase ( between the uniform and dimer phases ) .
it has been shown that a frustrated interchain coupling in the bi - layer structure could be the origin of that incommensurate phase @xcite .
therefore we propose as a minimal model for this material a set of antiferromagnetic spin - peierls chains coupled by a frustrated magneto - elastic interaction .
the result of the present paper could be taken as a starting point to interpret the magnetic excitation spectra in the low - temperature phases of these materials .
aiming to discuss the previous posed problem , in this work we compare two different models for the interchain coupling . in both models
an array of one - dimensional heisenberg chains coupled to the lattice deformation is described by the following hamiltonian : @xmath5\ , \textbf{s}_{i}^{j } \cdot \textbf{s}_{i+1}^{j } \label{hchains}\end{aligned}\ ] ] where we denote by @xmath6 the site of the @xmath7th chain , and the in subscript indicates that this part of the hamiltonian takes into account only the in - chain interactions .
@xmath8 are spin-@xmath0 operators with exchange constant @xmath9 along the x - axis of a non - deformed underlying lattice . in our simplified model
we took the scalar ionic coordinates @xmath10 to be the relevant for the dimerization process and @xmath11 are their conjugate momenta .
@xmath12 is the elastic coupling along the chain .
furthermore @xmath13 measures the deformation effect on the magnetic exchange constant .
the two cases we compare are differentiated by the interchain coupling configuration . in the first one , the magnetic sites lay on a lattice with square geometry .
the interchain coupling reads : @xmath14 in the other case they reside on an underlying triangular lattice . equation ( [ hintersquare ] ) changes to : @xmath15 \nonumber\\ & + & j_\mathrm{inter}\sum_{i , j } ( \textbf{s}_{i}^{j } \cdot\textbf{s}_{i}^{j+1 } + \textbf{s}_{i}^{j } \cdot\textbf{s}_{i+1}^{j-1 } ) \label{hintertriang}\end{aligned}\ ] ] where @xmath16 and @xmath17 are the elastic and magnetic interchain coupling constants .
we have considered only nearest neighbour interactions in both cases .
we finally get the complete hamiltonian as follows : @xmath18 where we must replace the second term of the sum with equation ( [ hintersquare ] ) or ( [ hintertriang ] ) depending on the case .
let us begin with a qualitative energetic interpretation in order to describe the ground state and the elementary excitations of these systems .
we treat the problem in the adiabatic approximation , i.e. we consider the ionic coordinates as classical static variables and neglect the kinetic term in ( [ hchains ] ) .
it is known that the one - dimensional spin - phonon model given by ( [ hchains ] ) dimerizes as the result of a competition between the elastic and magnetic interactions . in figure [ dimerlatsq ]
we show the array of dimerized chains ( horizontally positioned ) in the square geometry .
schematic representation of the dimerized phase in the square lattice ( gray background ) for ( a ) the in - phase and ( b ) an out - of - phase configuration .
arrows signal the displacement of the atoms .
the thin dashed lines correspond to unperturbed interchain springs and the blue continuous ones forming zigzag paths are dilated springs.,width=302 ] each spin belongs to a singlet and we therefore assume that @xmath17 is not active ( at least for a small enough value of @xmath17 , see the analysis for ladders below ) . therefore , we are not going to consider its effects in the present analysis .
in this figure we represent two possible configurations of dimerized chains . in the upper graphic ,
each chain is in phase with its neighboring chains , i.e. the @xmath6th ion of chain @xmath7 moves in the same direction of the @xmath6th ion in chains @xmath19 and @xmath20 . in this case
we can observe that this in - phase configuration does not cost additional interchain elastic energy because every interchain spring is unperturbed .
the lower part of the figure shows a possible out - of - phase configuration where the @xmath6th ion of chain @xmath7 moves in opposite direction to that of the @xmath6th ion of chains @xmath19 and @xmath20 .
here , there is an additional energy cost due to the expansion of every interchain spring . any other possible out - of - phase configuration will also have perturbed interchain springs so the in - phase configuration is the only one that minimizes the total energy .
now let us see what happens in the triangular geometry .
again , in figure [ dimerlattr ] we show two possible arrangements , the in - phase one in the upper part of the figure and an out - of - phase configuration ( where dimers in next - nearest neighbour chains are mutually out of phase ) in the lower part . schematic representation of the dimerized phase in the triangular lattice ( gray background ) for ( a ) the in - phase arrangement and ( b ) an out - of - phase configuration .
the thin dashed lines correspond to unperturbed interchain springs , the thick blue continuous ( green dashed ) ones forming zigzag paths are dilated ( compressed ) springs.,width=302 ] in this figure we can observe that there is alternately a compressed spring , an unperturbed one , a dilated one and then another unperturbed spring . here
we have the first difference with respect to the square geometry : meanwhile in the square geometry there is no additional elastic energy cost in the dimerization process , in the triangular one there are always deformed interchain springs .
the second difference arises when comparing the two panels of figure [ dimerlattr ] where we observe that in both cases we have the same deformation pattern of interchain springs , i.e. due to the frustrated configuration of the model the lower energy state of the system is degenerate with a degeneracy that depends on all the possible combinations of dimerization patterns .
note that if there were next - nearest neighbour interactions in the model , the system would lock next - nearest neighbour chains in phase but there is still no restriction with respect to nearest neighbour chains .
now let us include solitons as point defects dephasing the dimerization order in the in - phase arrangements of figures [ dimerlatsq ] and [ dimerlattr ] . we include two of such defects in the same chain , and
therefore the situation could be represented as shown in figure [ 2solitint ] .
schematic representation of the dimerized phase including two solitons ( signalled by an s ) in the central chain for ( a ) square and ( b ) triangular lattices .
the thick red lines show the perturbation induced by the excitations on the interchain springs .
note that in ( a ) every intersoliton spring is affected and in ( b ) the excitations produce a local perturbation with respect to the dimerized configuration.,width=302 ] note the essential difference in both cases , for the square lattice the intermediate zone dimerizes in antiphase to the background .
the energy cost of this configuration is proportional to the distance between the solitons , so they are always confined with a linear confining potential . in the triangular geometry
this situation does not take place because the interchain energy does not depend on the relative dimerization phase of two nearest neighbour chains and the only additional energy respect to the dimerized state is the local cost to create the solitons .
this means that the solitons are not interacting excitations in the triangular geometry .
the previous analysis could be checked quantitatively by the following construction . as both the lattice dimerization and soliton formation
could be taken into account including a smooth perturbation of the dimerization order , we can approximate @xmath21 .
the elastic interchain energies of ( [ hintersquare ] ) and ( [ hintertriang ] ) could be expanded in a gradient expansion of the lattice constant @xmath22 .
@xmath23 \label{heltriang}\end{aligned}\ ] ] note that in the triangular lattice we should retain one higher order in @xmath22 ( the last term of ( [ heltriang ] ) ) to have a non - null interchain coupling .
if it were not for such term the only effect of the interchain coupling would be to change the in - chain elastic coupling @xmath24 by @xmath25
. therefore the main effect of the interchain coupling in this case is to weaken the dimerization , reducing the magnetic gap and increasing the width of the solitons . as in the previous analysis
, a soliton could be included as point defect , thus neglecting its width .
we consider that all but the 0-th chain is dimerized as the background @xmath26 for @xmath27 .
for the 0-th we propose : @xmath28 being @xmath29 the heaviside step function , and where we have put two solitons at positions @xmath30 and @xmath31 ( @xmath32 ) . by replacing in expressions ( [ helsq ] )
we obtain the interacting energy @xmath33 which increases linearly with the distance between the solitons . moreover using ( [ heltriang ] ) the interacting energy vanishes showing the independence of the solitons . in the previous descriptions
we have neglected the interchain magnetic interaction .
moreover , we have used a point approximation for the solitons . in order to undertake a more precise study which also includes the effect of @xmath17
, we have to consider the two - chain problem , i.e. we extract a two - leg ladder from the model given in equation ( [ hchains ] ) , ( [ hintersquare ] ) and ( [ hintertriang ] ) . for this purpose , we have performed a numerical analysis for these models , using the dmrg finite algorithm which nowadays is accepted as the best numerical tool in low dimensional correlated systems @xcite .
the ladders were mapped into one - dimensional chains , where interchain interactions become next - nearest neighbour couplings . in order to avoid fictitious edge effects
, we have performed the calculations with periodic boundary conditions .
we have chosen 100-site ladders ( 50 sites each leg ) to plot our results .
we used the soliton width as a parameter to show that the thermodynamic limit was reached .
furthermore , this size is long enough to avoid an unwanted overlapping of possibly deconfined solitons due to their finite width .
it is worth noting that when the number of sites in each leg is even , as in our case , both solitons will appear in one of the two chains . on the other hand , with an odd number of sites in each chain
, one would obtain one soliton per chain but the conclusions would remain the same .
we assert the reliability of the calculations keeping 200 states with a truncation error of order @xmath34 .
lattice coordinates were taken into account by solving iteratively the adiabatic equations , i.e. we take @xmath10 as parameters , calculate the magnetic energy by dmrg , and minimize the total energy ( magnetic + elastic ) until convergence .
details of the self - consistent procedure are given elsewhere @xcite .
from now on we set @xmath35 as the energy scale and @xmath36 as the displacement one .
we define the dimensionless spin - phonon coupling @xmath37 .
the numerical results were obtained fixing @xmath38 and changing @xmath16 and @xmath17 .
we start by the ground state corresponding to @xmath39 .
the iterative method converge to the situation shown in figure [ ui_sz0 ] .
both the square ( sl ) and the triangular ladders ( tl ) dimerize , and in the case where there is no magnetic interchain coupling , the amplitude of dimerization is smaller in the tl as previously proved . lattice distortion in the dimerized state corresponding to the @xmath39 subspace for : ( a ) square and ( b ) triangular ladders . as the distortions in both chains are the same we only show results for the upper one .
we have fixed @xmath40.,width=302 ] note that a more essential difference arises if we pay attention to the variation of the dimerization amplitude @xmath41 with respect to @xmath17 .
the variation of @xmath41 with @xmath42 is shown in figure [ dimervsjinter ] for both the sl and the tl .
dimerization amplitude as a function of @xmath42 for the sl and the tl .
again we use the same elastic parameter indicated in the caption of figure [ ui_sz0].,width=302 ] the dimerized state resist a stronger @xmath17 for the tl than the sl
. moreover , the dimerization amplitude is independent of @xmath17 for the tl in a wide range of this parameter .
this is an important fact and implies that in the dimerized state the system remains one - dimensional even with a strong interchain coupling .
a possible interpretation for this result is that each spin is paired in a singlet and in consequence the frustrated interaction can not compete with this state until it is strong enough to break the singlet and generate an effective uniform 1d heisenberg model along the zig - zag path .
now let us analyze the excited state . in figure [ usqsz_1 ] and [ szsqsz_1 ]
we show the lattice deformation and the local magnetization for the square ladder .
lattice distortion for the upper and lower chains of the sl in the @xmath43 subspace .
the green curve is a fitting to the analytic result given by equations ( [ udomain1 ] ) and ( [ udomain2]).,width=302 ] local magnetization for the upper and lower chains of the square ladder in the subspace @xmath43.,width=302 ] there are two defects of the dimerization pattern .
distortions far away from that defects correspond to the dimerization pattern with the same value as in figure [ ui_sz0 ] ( @xmath44 ) . the defect can not be considered as two independent solitons because they always fall at the same distance independently of the initial distortion pattern used in the method .
the deformation of the intermediate region between the defects does not correspond to the dimerization of the bulk and correspondingly @xmath45 does not vanish in this intermediate zone . therefore we fail to describe this pattern as a two - soliton form that would be given by @xmath46 where @xmath47 is the soliton width .
it is better described as the domain excitation found in @xcite for the square lattice .
it was obtained by bosonization techniques and the self - consistent harmonic approximation ( scha ) . for comparison ,
we redo the calculation of @xcite for the two - chain spin - phonon system coupled by an elastic interaction . to solve the scha equations we fix the bosonic field representing the spin variables in one of the chains ( say 1 ) to its value in the @xmath39 dimerized state @xcite .
the displacement fields @xmath48 and @xmath49 are then given by the following expressions : @xmath50\label{udomain1}\\ u_2(x)=\frac{u_0}{\frac{\epsilon}{2}+1}\left[\frac{\epsilon}{4}+(1+\frac{\epsilon}{4 } ) t(x - x_\mathrm{m})\right ] \label{udomain2}\end{aligned}\ ] ] where @xmath51 , @xmath52 sets the position of the domain in the chain 2 and the function @xmath53 is given by : @xmath54 { \mathop{\mathrm{sech}}}\left[\frac{(x+x_0)}{\xi}\right]\end{aligned}\ ] ] with : @xmath55,\hspace{0.5cm}b=\frac{1}{1+\frac{4}{\epsilon } } \label{x0}\end{aligned}\ ] ] being @xmath47 a characteristic width of the wall of the domain .
@xmath56 measures the radius of the domain configuration .
note that for each interchain coupling @xmath57 , @xmath58 is fixed by equation ( [ x0 ] ) , i.e. the wall of the domain accommodates at a given equilibrium distance . only for negligible values of @xmath57
the domain dissociate in two separate domain walls as given in equation [ 2solfit ] . for any finite @xmath57 ,
the walls are as close as they can in order to reduce the interchain energy .
an accurate fitting could be found between the dmrg results and the analytic form given in equation ( [ udomain2 ] ) .
lines in figure [ usqsz_1 ] were obtained from this equation with @xmath59 and @xmath60 .
@xmath61 has not been taken as a fitted parameter but obtained from equation ( [ x0 ] ) .
the slight reduction of the dimerization in the lower chain is well predicted by the analytic expression in equation ( @xmath62 ) . with the same values for the parameters a very good agreement
is found as shown in the lower panel of figure [ usqsz_1 ] .
the overall coincidence prove that the spin-1 excitation we found numerically could be identified with the domain predicted in analytic calculations based on bosonization .
on the other hand , we also study the effect of a magnetic interchain coupling ( @xmath63 ) .
the only perturbation is a small magnetic polarization of the lower chain debilitating the singlets in this chain as observed in figure [ szsqsz_1 ] .
the displacements @xmath64 do not change with the presence of @xmath17 , and the distortions are shown in figure [ usqsz_1 ] .
a quite different situation occurs in the triangular ladder . in this case
the two defects appear separate enough to give the possibility that the intermediate zone dimerizes as the bulk .
the situation is shown in figure [ uzzsz_1 ] .
lattice distortion for the upper and lower chains of the tl in the subspace @xmath65.,width=302 ] the two defects of the dimerization order are now far apart .
the deformation pattern can now be fitted by a two - soliton function as given in equation ( [ 2solfit ] ) .
for the set of parameters we used , we have found a soliton width of @xmath66 .
different runs give patterns like those of figure [ uzzsz_1 ] but the distance between the solitons @xmath67 depends on the initial distortion pattern . as our method minimizes the total energy
this is an indication that solitons are independent excitations .
let us analyze the effect of the magnetic interchain coupling . as a soliton
could be thought as a free spin in a background of singlets it is expected that a weak @xmath17 should not have much influence in the soliton formation and their interactions .
this is in fact the case for @xmath63 as shown in figure [ szzzsz_1 ] .
local magnetization for the upper and lower chains of the tl in the subspace @xmath65.,width=302 ] it is interesting to remark that even when the interchain coupling equals the in - chain one the basic physics does not change .
the distortion pattern is described by a two - soliton function and the local magnetization of the upper chain increase but the perturbation remains localized .
therefore we predict that independent solitons could be the elementary excitations even in this case . until now we have found the lattice deformation by minimizing the total energy of the system .
now we take the solitons as elementary entities and analyze the question of their interaction energy .
once the distortions have been fitted by the function ( [ 2solfit ] ) we modify the distance between the solitons changing the values of @xmath67 .
we calculate the total energy as a function of the distance @xmath68 .
results are shown in figure [ eintezz ] .
dependence of the interaction energy with the distance between two solitons , for the triangular ladder.,width=302 ] we have subtracted the energy of the dimerized state in order to obtain the creation energy of the solitons plus their interaction energy .
the energy increases at small distances .
this effect could be interpreted as a repulsion between the solitons when their distance is smaller than twice their width . for larger separations
the energy does not depend on the distance .
this fact shows that solitons are indeed independent when they are far apart .
the inclusion of a @xmath69 does not modify this picture as shown in the figure [ eintezz ] .
the role of the interchain coupling frustration in stabilizing deconfined spinons was recently discussed in purely magnetic coupled chains by nersesyan and tsvelik @xcite . to make contact with this study we should consider our spin - phonon model beyond the adiabatic approximation .
actually the partition function of the model given by equations ( [ hchains ] ) and ( [ hintertriang ] ) could be formulated as a path integral . moreover ,
the lattice coordinates @xmath64 could be integrated out producing a retarded interaction between the spin variables . in the antiadiabatic limit
this interaction becomes instantaneous .
the effective magnetic model will contain , in addition to the original interchain magnetic interactions , four - spin exchange interactions as the one included in previous models leading to deconfined spinons @xcite .
it would be very interesting to study the spin - phonon model by the techniques used in @xcite to analyze the appearance of deconfined spinons when the phonon dynamics is included . on a general ground
we could assert that the deconfined scenario discussed in the present paper is connected with a local symmetry transformation that acts on each individual chain changing one of the two possible dimerized states into the other one .
this emergent @xmath70 gauge symmetry is at the heart of the deconfinement mechanism found in the magnetic model in @xcite .
therefore , we expect the appearance of a deconfined quantum critical point in this model as found previously in the simplest spin system
. note the differences with the model of spin - phonon chains coupled by non - frustrated interactions .
this model could be mapped into an effective ising model of coupled chains @xcite .
the mapping takes into account the underlying _ global _ @xmath70 symmetry of the dimer order parameter . the symmetry is spontaneously broken in the low temperature phase and the solitons are confined . in the frustrated case the dimerization phases between different chains are uncorrelated and the mapping of ref .
@xcite leads to decoupled ising chains which is a manifestation of the deconfinement of solitons discussed in the present paper .
let us finish with the following remark .
as previously discussed , in order to fix the dimerization order in the direction perpendicular to the magnetic chains , we should add a next - nearest neighbour elastic interaction .
analogously to the argument used for the square geometry , this effect will provide a confinement mechanism for the solitons . even though this coupling might be necessary to modelize the compounds tiocl and tiobr
, the results of the present paper could be the starting point to describe the spectrum at not too low temperatures , i.e. above the point where the elastic next - nearest neighbour energy starts to be washed out by the thermal energy .
in summary , we have shown an essential difference in the ground state and the elementary excitations of coupled spin - peierls chains depending on the nature of the interchain coupling , i.e. with and without frustration . in the frustrated case
we have found that the amplitude of dimerization in the ground state is independent of the strength of the magnetic interchain coupling in a broad range of values of the parameter @xmath17 .
moreover , the analysis of the elementary excitations showed that solitons of the individual chains survive the inclusion of a nearest - neighbour interchain coupling .
as our approach relay in an static approximation of the phonon field we can not make precise predictions on the dynamical response of the system . on a general ground
we can speculate that a two - soliton continuum should appear in the frustrated system .
this is a different behaviour than that measured in a non - frustrated spin - peierls system .
we thank d. cabra for helpful discussions .
this work was partially supported by pip conicet ( grant 5036 ) . |
|
reaction cross sections @xmath3 of unstable nuclei provide a powerful tool to study the structure of unstable nuclei such as density distribution and deformation @xcite . for instance , a largely extended structure , referred to as `` halo '' , of unstable nuclei such as @xmath4li @xcite , @xmath4be @xcite , and @xmath5c @xcite has been found with such measurements .
the halo structure is one of the characteristic features of weakly bound nuclei , and has attracted lots of attention ( see ref .
@xcite for a recent discovery of halo structure in @xmath6ne nucleus ) .
experimentally large odd - even staggering ( oes ) phenomena have been revealed in reaction cross sections of unstable nuclei close to the neutron drip line , _
e.g. , _ in the isotopes @xmath7c @xcite , @xmath8c @xcite , @xmath9ne@xcite , @xmath10ne @xcite , and @xmath11 mg @xcite . in ref .
@xcite , we have argued that the pairing correlations play an essential role in these oes .
that is , the oes in reaction cross sections is intimately related to the so called pairing anti - halo effect discussed in ref .
@xcite . on the other hand ,
there have been contradictory arguments whether the pairing gap disappears @xcite or persists@xcite when a nucleus reaches at the neutron drip line , _ i.e. _ , the single - particle energy of the last occupied orbit approaches the zero energy .
if the pairing gap disappeared , the oes effect might be either quenched or disappeared completely , unless the deformation parameter is significantly different among the neighboring nuclei @xcite . in this paper , we first discuss the pairing correlations close to the zero energy by the hartree - fock bogoliubov ( hfb ) method .
we carry out hfb calculations for the neutron 3s@xmath12 orbit in @xmath13cr changing the separation energy in a mean field potential , and examine different definitions for an effective pairing gap parameter .
this problem is also related with the superfluidity of neutron gases in the outer crust of neutron stars @xcite .
the second motivation of this work , in addition to giving the details of the analysis in ref .
@xcite , is to propose a formula to measure the odd - even staggering in the reaction cross sections .
notice that the oes of the isotope shift of stable nuclei has been discussed mostly to clarify the deformation changes in odd - mass and even - mass nuclei @xcite .
the present issue of oes in the reaction cross sections has a similar aspect to the previous study in ref .
@xcite in one sense , but different in another aspect since it aims at studying the existence of the pairing correlation in nuclei close to the neutron drip line .
the paper is organized as follows . in sec .
ii , we discuss the pairing correlation in neutron - rich nuclei using the hfb method . in sec .
iii , we apply a glauber theory in order to calculate reaction cross sections .
we introduce the oes parameter for reaction cross sections , and discuss it in relation with the pairing correlations in weakly bound nuclei .
we then summarize the paper in sec .
in the coordinate space representation , the hfb equations read @xcite ( cc - & ( ) + ( ) & -+ ) ( c u_i ( ) + v_i ( ) ) = e_i ( c u_i ( ) + v_i ( ) ) , where @xmath14 is the mean - field hamiltonian , @xmath15 being the nucleon mass . @xmath16 and @xmath17 are the mean - field and the pairing potentials , respectively , and @xmath18 is a quasi - particle energy . here
, we have assumed that the nucleon - nucleon interaction is a zero range force so that these potentials are local .
the upper component of the pair wave function @xmath19 is a non - localized wave function if the quasi - particle energy @xmath18 is larger than the fermi energy @xmath20 , while the lower component @xmath21 is always localized .
the pair potential @xmath17 in general has a larger surface diffuseness than the mean field potential @xmath16 , and goes beyond it due to the non - localized property of the upper component of the wave function @xmath19 , that is , due to the coupling to the continuum spectra @xcite . in the mean field approximation without the pairing correlations ( _ i.e. _ , @xmath22 ) , the halo structure originates from an occupation of a weakly - bound @xmath0 or @xmath1 orbit by the valence nucleons near the threshold @xcite .
the asymptotic behavior of a single particle wave function for @xmath23 wave reads @xmath24 where @xmath25 is defined as @xmath26 with the hartree - fock ( hf ) energy @xmath27 . the mean square radius of this wave function is then evaluated as @xmath28 which will diverge in the limit of vanishing separation energy , @xmath29 .
it has been shown that this divergence occurs not only for @xmath23 wave but also for @xmath30 wave , although the dependence on @xmath31 is now @xmath32 for @xmath1 @xcite .
in contrast , in the presence of the pairing correlations ( _ i.e. , _
@xmath33 ) , the lower component of the hfb wave function , which is relevant to the density distribution , behaves as @xcite @xmath34 where @xmath35 is given by , @xmath36 using the quasi - particle energy @xmath18 . with the canonical basis @xmath37 in the hfb theory ,
the quasi - particle energy may be approximately given by @xcite @xmath38 where @xmath39 and @xmath40 .
in the zero binding limit , @xmath41 and @xmath42 , the asymptotic behavior of the wave function @xmath43 is therefore determined by the gap parameter as , @xmath44 .
\label{hfb}\ ] ] the radius of the hfb wave function will then be given in the limit of small separation energy @xmath45 as @xmath46 if the gap parameter @xmath47 stays finite in the zero energy limit of @xmath48 , the extremely large extension of a halo wave function in the hf field will be reduced substantially by the pairing correlations and the root - mean - square ( rms ) radius will not diverge .
this is referred to as the anti - halo effect due to the pairing correlations @xcite .
it was shown in ref .
@xcite that this is the main reason for the observed oes in the reactions cross sections of several drip line nuclei . in order to study the behavior of the pairing gap in weakly bound nuclei ,
we carry out hfb calculations for the neutrons in @xmath13cr nucleus . to this end
, we use a spherical woods - saxon ( ws ) potential , @xmath49 with @xmath50 } , \label{fermifunction}\ ] ] for the mean - field potential @xmath51 . following ref .
@xcite , we take @xmath52 mev , @xmath53 fm , @xmath54 mev@xmath55@xmath56 , and @xmath57=0.67 fm . for the hfb calculations , we use a density - dependent contact pairing interaction of surface type , with which the pairing potential is given by @xmath58 here , @xmath59 and @xmath60 are the total particle density and the neutron pairing density , respectively , given by @xmath61 we again follow ref . @xcite and take @xmath62=0.16 @xmath63 and @xmath64 mev@xmath55@xmath65 with the energy cut off of 50 mev above the fermi energy . in order to construct the proton density , we use the same mean - field potential as in eq .
( [ vmf ] ) , but with @xmath66 mev .
we also add the coulomb potential for a uniform charge with a radius of @xmath67 .
we discretize the continuum spectra with the box boundary condition .
we take the box size of @xmath68=60 fm , and include the angular momentum up to @xmath69 .
notice that we determine the pairing potential self - consistently in this model according to eq .
( [ pairpot ] ) , although the woods - saxon potential is fixed for the mean - field part .
the fermi energy is also determined self - consistently according to the condition for the average particle number conservation , a = ( r ) d = _
i|v_i()|^2 d. these self - consistencies are particularly important to increase the pairing gap for an extremely loosely bound orbit @xcite .
cr nucleus as a function of the single particle energy @xmath70 for the 3s@xmath12 orbit in a woods - saxon potential .
the top panel shows the mean square radius of the 3s@xmath12 wave function obtained with ( the solid line ) and without ( the dashed line ) the pairing correlation .
the middle panel shows the root - means - square ( rms ) radii for @xmath71cr ( the dotted line ) , @xmath72cr ( the dashed line ) , and @xmath13cr ( the solid line ) .
these are obtained with the hartree - fock ( @xmath71cr and @xmath72cr ) and the hartree - fock - bogoliubov ( @xmath13cr ) methods .
the bottom panel shows the effective pairing gap for the 3s@xmath12 state .
the solid line is the pairing gap defined with the canonical basis , that is , the expectation value of the pair potential with respect to the canonical basis wave function for the 3s@xmath12 state , while the dot - dashed line shows the lowest quasi - particle energy for the s@xmath12 states . ] the top panel of fig .
1 shows the mean square radius of the 3@xmath73 state . in order to study the dependence on the binding energy , we vary the depth of the ws potential @xmath74 for neutron @xmath23 wave states while we keep the original value for the other angular momentum states .
we also arbitrary change the single - particle energy for the 2d@xmath75 state from @xmath76 mev to @xmath77 mev so that the 3s@xmath12 state lies below the 2d@xmath75 state .
the dashed line is obtained with the single - particle wave function for the mean - field hamiltonian @xmath78 , while the solid line is obtained with the wave function for the canonical basis in the hfb calculations .
the radius of the single - particle wave function for the @xmath23-wave state increases rapidly as the single - particle energy @xmath79 approaches zero , and eventually diverges in the limit of @xmath80 .
in contrast , the hfb wave function shows only a moderate increase of the radius even in the limit of @xmath80 .
the middle panel of fig .
1 shows the root - mean - square ( rms ) radius for the whole nucleus , by taking into account the contribution of the other orbits as well . for comparison
, we also show the rms radii for @xmath71cr and @xmath72cr obtained with the same mean - field potential but without including the pairing correlation .
notice that the rms radius of @xmath13cr is larger than that of @xmath72cr in the range of @xmath81 mev .
this is due to the coupling of single - particle wave functions to a larger model space , including continuum , induced by the pairing correlations . on the other hand , in the limit of @xmath80
, the rms radius of @xmath72cr shows a divergent feature while that of @xmath13cr is almost constant . for the 3s@xmath12 state .
the solid line shows the pairing gap evaluated with the canonical basis , @xmath47 for @xmath82s@xmath12 , while the dot - dashed line is the lowest quasi - particle energy @xmath18 .
notice that the right hand side of eq .
( [ ehfb ] ) is simply a diagonal component of the hfb hamiltonian in the canonical basis , while the left hand side is obtained by diagonalizing the hfb matrix@xcite . due to the off - diagonal components , eq .
( [ ehfb ] ) holds only approximately , and thus @xmath18 may be smaller than @xmath83 in actual calculations .
one can see in the figure that the effective pairing gaps persist even in the limit of @xmath80 , leading to the reduction of the radius of hfb wave function as is shown in the top panel of fig
. 1 . is determined self - consistently in this model .
the dot - dashed lines in the top and the middle panels denote the results of the self - consistent calculations shown in fig . 1 . ] in the simplified hfb model of ref .
@xcite , it was claimed that the effective paring gap is diminished or quenched substantially for low @xmath84 orbits with @xmath0 and @xmath1 . in this model , the radial dependence of the pairing field @xmath85
is fixed either as a fermi - type function ( volume - type ) or a derivative of the fermi function ( surface - type ) with the same surface diffuseness parameter as in the mean - field potential .
furthermore , the fermi energy is set equal to the single - particle energy , @xmath79 , for the mean - field hamiltonian @xmath78 , that is , @xmath86 .
the effective pairing gap is defined in ref .
@xcite to be identical to the corresponding quasi - particle energy .
if the energy in the canonical basis were the same as the single - particle energy , eq .
( [ ehfb ] ) indeed yields e_i=_i^(can ) .
[ delta - eff ] care must be needed , however , since in general @xmath87 deviates from the single - particle energy , @xmath88 in the hfb .
moreover , when the effective gap is plotted as a function of @xmath79 as has been done in ref .
@xcite , setting @xmath86 leads to a violation of particle number in this model , whose effect may be large in the limit of @xmath89 . and the neutron number calculated with the simplified hfb model .
these are plotted as a function of the single particle energy @xmath90 for the 3s@xmath12 orbit in a woods - saxon potential .
the solid lines are obtained by adjusting the fermi energy so that the neutron number is @xmath91=52 , while the dashed lines are obtained by setting @xmath92 . ] in order to investigate the consistency of the simplified model of ref .
@xcite , we repeat the same calculations shown in fig . 1 by assuming that the pair potential @xmath85 is proportional to @xmath93 , where @xmath94 is given by eq . ( [ fermifunction ] ) .
we use the proportional constant of @xmath95 mev , that leads to the same value for the average pairing gap , @xmath96 as that in the self - consistent calculations shown in fig . 1 for @xmath97 mev .
we keep this value in varying the depth of the woods - saxon potential , @xmath98 .
we first keep the particle number to be a constant ( @xmath91=52 ) and determine the fermi energy self - consistently within this simplified model .
figure 2 show the results of such calculations . for comparison , the top and the middle panels also show by the dot - dashed lines the results of the self - consistent calculations , which have already been shown in fig .
it is remarkable that this model yields a similar rms radius to that of the self - consistent calculation .
the effective pairing gaps show somewhat different behavior from those in the self - consistent calculation , especially for the pairing gap defined with the canonical basis ( see the bottom panel ) .
however , it should be emphasized that the pairing gaps stay finite in this calculation in the limit of vanishing single - particle energy , as in the self - consistent calculation shown in fig .
this implies that the self - consistency for the pair potential is not important , as far as the rms radius is concerned .
state , @xmath99 . ]
we next carry out the calculation by setting the fermi energy to be the same as the single - particle energy for the 3s@xmath12 state , @xmath100 . in this calculation , the number of particle changes as we vary the depth of the woods - saxon potential .
the fermi energy @xmath101 and the neutron number are shown in the upper and the lower panels of fig .
3 , respectively , by the dashed lines . for comparison , the figure also shows the results of the previous calculation shown in fig .
2 , that is , those obtained by adjusting the fermi energy so that the neutron number is a constant ( see the solid lines ) .
the variation of the particle number is large , that is , from 47 to 54 in the range of single - particle energy shown in fig .
the radii and the effective pairing gaps are shown in fig .
4 . as one can clearly see , this non - self - consistent calculation yields considerably different results from the self - consistent calculations .
firstly , the reduction of the mean square radius of the single - particle orbit is somewhat underestimated , although the effect is still large ( see the top panel ) .
secondly , the rms radius obtained with this model is completely inconsistent with the result of the self - consistent calculation as shown in the middle panel .
lastly , the effective pairing gaps drops off in the limit of @xmath102 ( see the bottom panel ) . particularly , the lowest quasi - particle energy is substantially dismissed , that is a similar behavior as that shown in ref .
evidently , the claim of ref .
@xcite that the pairing gap disappears in the zero energy limit is an artifact of setting @xmath86 .
if the fermi energy is determined self - consistently for a given particle number , the effective pairing gap persists even if the pairing potential is pre - fixed as shown in fig .
let us now investigate how the pairing correlation affects the reaction cross sections of weakly - bound nuclei . to this end , we use the glauber theory @xcite . in the optical limit approximation of the glauber theory ,
the reaction cross section @xmath3 can be calculated as @xcite , @xmath103 with @xmath104 here , @xmath105 and @xmath106 are the projectile and the target densities , respectively , and @xmath107 is the impact parameter . @xmath108 and @xmath109 are the transverse components of @xmath110 and @xmath111 , respectively , that is , @xmath112 and @xmath113 , where @xmath114 is the unit vector parallel to @xmath107 .
@xmath115 is the profile function for the @xmath116 scattering , for which we assume to take a form of @xcite , @xmath117 with @xmath118 being the total @xmath116 cross section .
it has been known that the optical limit approximation overestimates reaction cross sections for weakly - bound nuclei @xcite . in order to cure this problem , abu - ibrahim and suzuki
have proposed modifying the phase shift function @xmath119 in eq .
( [ pshift ] ) to @xcite @xmath120 \nonumber \\ & & -\frac{1}{2}\int d{\mbox{\boldmath $ r$}}'\rho_t({\mbox{\boldmath $ r$ } } ' ) \nonumber \\ & & \times\left[1-\exp\left(-\int d{\mbox{\boldmath $ r$}}\rho_p({\mbox{\boldmath $ r$ } } ) \gamma_{nn}({\mbox{\boldmath $ s$}}'-{\mbox{\boldmath $ s$}}+{\mbox{\boldmath $ b$}})\right)\right ] .
\label{chi}\end{aligned}\ ] ] with this prescription , the effects of multiple scattering between a projectile nucleon and the target nucleus , and that between a target nucleon and the projectile nucleus , are included to some extent @xcite .
cr + @xmath121c reactions at @xmath122=240 mev / nucleon , obtained with the modified optical limit approximation of the glauber theory , eq .
( [ chi ] ) .
the density distributions for the @xmath123cr are constructed from the hfb calculations shown in fig . 1 .
( the lower panel ) the staggering parameter defined by eq .
( [ eq - gam ] ) . ] the upper panel of fig .
5 shows the reaction cross sections for @xmath123cr + @xmath121c reactions at @xmath122=240 mev / nucleon , obtained with the phase shift function given by eq .
( [ chi ] ) . for the density of the cr isotopes
, we use the results of the hfb calculations shown in fig .
1 . for the density of the target nucleus @xmath121c
, we use the same density distribution as that given in ref . @xcite . in the actual calculation , we treat the proton - neutron and the proton - proton / neutron - neutron scattering separately and use the parameters given in table i in ref .
@xcite for the profile function @xmath115 . in order to evaluate the phase shift function
, we use the two - dimensional fourier transform technique @xcite .
we give its explicit form in the appendix .
the reaction cross sections shown in fig .
5 show a similar behavior as in the rms radii shown in the middle panel of fig .
1 , as is expected .
that is , the reaction cross sections for @xmath13cr and @xmath72cr are inverted at a small biding energy , due to the pairing effect shown in fig .
1 . this leads to a large odd - even staggering in reaction cross sections for weakly bound nuclei@xcite .
defined by eq . as a function of the neutron separation energy s@xmath124 for the odd - mass nuclei .
the solid and the dotted lines correspond to the reaction cross sections for @xmath125o+@xmath121c and @xmath123cr+@xmath121c at @xmath122=240 mev / nucleon , respectively .
the dashed and the dot - dashed lines show the results for @xmath10ne+@xmath121c at @xmath122=240 mev / nucleon , in which the valence neutron in @xmath6ne occupies the 2p@xmath126 or the 1f@xmath127 orbits , respectively . ] in order to quantify the oes of reaction cross sections , we introduce the staggering parameter defined by _ 3=(-)^a , [ eq - gam ] where @xmath128 is the reaction cross section of a nucleus with mass number @xmath129 .
we can define the same quantity also for rms radii .
notice that this staggering parameter is similar to the one often used for the oes of binding energy , that is , the pairing gap @xcite .
the lower panel of fig .
5 shows the staggering parameter @xmath130 for the @xmath123cr nuclei as a function of the single - particle energy , @xmath90 .
one can clearly see that the staggering parameter @xmath130 increases rapidly for small separation energies , and goes up to a large value reaching @xmath131 80 mb . in order to find out a general trend of the staggering parameter , fig .
6 shows the value of @xmath130 for various orbits with 2s@xmath12 , 3s@xmath12 , 2p@xmath126 and 1f@xmath127 .
the values for the 2s@xmath12 and 2p@xmath126 orbits correspond to the reaction cross sections for the @xmath125o and @xmath10ne nuclei , respectively , calculated in ref .
the value for the 1f@xmath127 orbits corresponds to the reaction cross sections for @xmath10ne nuclei , obtained with the diffuseness parameter of the mean - field woods - saxon potential of @xmath57=0.65 fm .
one can clearly see that @xmath130 for the low @xmath84 orbits with @xmath0 and @xmath1 show a rapid increase at small separation energies , the @xmath0 orbits increasing more rapidly than the @xmath1 orbit .
in contrast , the high @xmath84 orbit with @xmath2 does not show any anomaly in the limit of @xmath102 .
these features are quite similar to the growth of a halo structure only in the low @xmath84 orbits due to a zero or small centrifugal barrier . of reaction cross sections defined by eq . for the ne isotopes with the @xmath121c target at @xmath122=240 mev / nucleon .
this is plotted as a function of the neutron separation energy s@xmath124 of the odd-@xmath129 nuclei .
the experimental data for the reaction cross sections are taken from ref .
@xcite , while the empirical separation energies are taken from refs .
the dashed line is the calculated staggering parameter for the @xmath10ne isotopes , assuming that the valence neutron in of @xmath6ne occupies the 2p@xmath132 orbit . ]
the experimental staggering parameters @xmath130 are plotted in fig . 7 for ne isotopes as a function of the neutron separation energy for the odd - mass nuclei .
we use the experimental reaction cross sections given in ref .
@xcite while we evaluate the separation energies with the empirical binding energies listed in ref .
for the neutron separation energy for the @xmath6ne nucleus , we use the value in ref .
the experimental uncertainties of the staggering parameter are obtained as @xmath133 where @xmath134 is the experimental uncertainty for the reaction cross section of a nucleus with mass number @xmath129 .
the figure also shows by the dashed line the calculated staggering parameter for the @xmath10ne nuclei with the 2p@xmath126 orbit , that has been shown also in fig .
one sees that the experimental staggering parameter agrees with the calculated value for @xmath10ne nuclei when one assumes that the valence neutron in @xmath6ne occupies the 2p@xmath126 orbit .
furthermore , although the structure of lighter odd - a ne isotopes is not known well , it is interesting to see that the empirical staggering parameters closely follow the calculated values for the 2p@xmath126 orbit .
this may indicate that the low-@xmath84 single - particle orbits are appreciably mixed in these ne isotopes due to the deformation effects@xcite .
we have studied the odd - even staggering ( oes ) of the reaction cross sections by using the hartree - fock - bogoliubov model . to this end , we have introduced the staggering parameter @xmath130 defined with a 3 point difference formula in order to clarify the relation between the magnitude of oes and the neutron separation energy .
we have shown that the oes parameter increases largely for low-@xmath84 orbits with @xmath0 and @xmath1 at small separation energies .
the experimental staggering parameter for the ne isotopes show a similar increase . on the other hand
, we have found that the staggering parameter stays almost a constant value , @xmath135 mb for higher @xmath84 orbits with _
the increase of @xmath130 is induced by the finite pairing correlations in the zero separation energy limit . in this respect , we have shown that the effective pairing gap for the 3s@xmath12 orbit in the @xmath13cr nucleus persists even in the limit of vanishing separation energy . this remains the same even if the pair potential is prefixed , as long as the chemical potential is adjusted to keep the particle number to be the same .
we have shown that such simplified hfb model well reproduces the results of the self - consistent hfb model for the root - mean - square radius .
the staggering parameter proposed in this paper provides a good measure for the oes of reaction cross sections .
further systematic experimental studies would be helpful in order to clarify the pairing correlations in weakly - bound nuclei and in the limit of zero neutron separation energy .
we would like to thank m. takechi for fruitful discussions on the experimental data .
we thank also p. schuck and c.a .
bertulani for useful discussions .
this work was supported by the japanese ministry of education , culture , sports , science and technology by grant - in - aid for scientific research under the program numbers ( c ) 22540262 and 20540277 .
in this paper , we evaluate the phase shift functions given by eqs .
( [ pshift ] ) and ( [ chi ] ) using the fourier transform technique @xcite .
first we notice that eq .
( [ pshift ] ) can be expressed as @xmath136 with @xmath137 the two - dimensional fourier transform of @xmath138 in eq .
( [ pshift ] ) then reads @xmath139 where @xmath140 , and @xmath141 are the two - dimensional fourier transform of @xmath142 , and @xmath143 , respectively . for the profile function given by eq .
( [ gamma_nn ] ) , its fourier transform reads , @xmath144 the fourier transform of the density distribution is evaluated as @xmath145 where @xmath146 is the three - dimensional fourier transform of the density at @xmath147 . for a spherical density , @xmath59
, @xmath146 depends only on @xmath148 , that is , @xmath149 where @xmath150 is the spherical bessel function of zero - th order . taking the inverse fourier transform of eq .
( [ 2dft ] ) , the phase shift function is calculated as @xmath151 where @xmath152 is the bessel function of zero - th order .
a similar technique has been used to evaluate a double folding potential in heavy - ion reactions @xcite .
one can apply the same method to evaluate the phase shift function given by eq .
( [ chi ] ) .
first notice that @xmath153 depends only on @xmath154 .
this leads to @xmath155 where @xmath156 is calculated as @xmath157 | we investigate the odd - even staggering ( oes ) in reaction cross sections of weakly bound nuclei with a glauber theory , taking into account the pairing correlation with the hartree - fock - bogoliubov ( hfb ) method .
we first discuss the pairing gap in extremely weakly bound nuclei and show that the pairing gap persists in the limit of zero separation energy limit even for single - particle orbits with the orbital angular momenta @xmath0 and @xmath1 .
we then introduce the oes parameter defined as the second derivative of reaction cross sections with respect to the mass number , and clarify the relation between the magnitude of oes and the neutron separation energy .
we find that the oes parameter increases considerably in the zero separation energy limit for @xmath0 and @xmath1 single - particle states , while no increase is found for higher angular momentum orbits with _
e.g. , _
@xmath2 .
we point out that the increase of oes parameter is also seen in the experimental reaction cross sections for ne isotopes , which is well accounted for by our calculations . |
the properties of intercombination transitions in alkaline earth atoms and atoms with similar electronic structure has become an object of a growing number of experimental and theoretical studies .
it is mostly caused by a variety of new applications in the physics of ultra cold atoms : from laser cooling and trapping @xcite to optical frequency standards @xcite .
great progress in this area has been achieved for ytterbium ( yb ) , which has 7 stable isotopes with atomic weights 168 , 170 , 171 , 172 , 173 , 174 , and 176 .
quantum degenerate gases have been obtained for the bosonic isotopes @xmath2 @xcite , @xmath3 @xcite , and @xmath4 @xcite , and the fermionic isotopes @xmath5 and @xmath6@xcite .
photoassociation spectroscopy has been carried out for bosons @xcite as well as fermions @xcite .
a two - color photoassociation experiment has allowed a precise determination of the ground state s - wave scattering lengths for all combinations of yb isotopes @xcite .
finally , it was demonstrated that the significant change of the scattering properties for colliding ground state atoms can be achieved with optical feshbach resonances in these systems @xcite .
this work is devoted to the theoretical study of the photoassociation spectra near the intercombination transition @xmath0@xmath1 of yb for bosonic isotopes .
we take advantage of the experimental photoassociation spectra for @xmath7 and @xmath8 from tojo _
et al_. @xcite to precisely determine the binding energies of excited state molecular energy levels .
in addition , we report new measurements of the binding energies of excited @xmath9 and @xmath10 .
analysis of the photoassociation spectra also requires knowledge of the scattering properties in the ground electronic state of two colliding atoms .
the necessary information and experimental data for all ground state isotopic combinations can be found in ref .
@xcite .
yb is an excellent example of a system for which the scattering properties can be easily tuned by the change of the isotopic combination , thus changing the reduced mass of the colliding pair .
such mass tuning of the scattering properties can also be very useful in other similar species with several isotopes .
development in the laser trapping and cooling of atoms such as cd @xcite and hg @xcite will hopefully allow photoassociation investigations of these systems in the future .
cadmium and hg , like yb , are good candidates for mass tuning of the scattering length because of their numerous isotopes .
in addition , hg is seen as a very promising candidate for future optical frequency standards .
the clock frequency shift induced by black body radiation is especially small in hg @xcite , compared with other group ii elements @xcite .
the two - body loss rate coefficient @xmath11 in the photoassociation process for a thermal cloud of ultracold atoms at temperature @xmath12 needs to be evaluated as an average over all possible momenta of two colliding atoms .
this loss rate is directly dependent on the light intensity @xmath13 leading to the photoassociation as well as on the detuning of the light from the atomic resonance @xmath14 and detuning @xmath15 corresponding to the optical resonance coupling the scattering ground state `` @xmath16 '' with the excited `` @xmath17 '' bound state @xcite .
the averaged loss rate can be written as @xcite : @xmath18 where @xmath19^{2}+[\gamma_{e}(i,\varepsilon_{r})/2]^{2}}\ ] ] is the loss rate @xcite corresponding to particular momentum vectors of the relative motion of the two colliding atoms @xmath20 as well as the motion of their center of mass @xmath21 .
contributions from all possible transitions between excited bound and ground scattering states are included in this expression .
they are taken in the sum with weights dependent on the total angular momentum @xmath22 of the two - atom system . in eq .
( [ equ2 ] ) the magnitude of the wave vector corresponding to the relative motion is @xmath23 , the kinetic energy of relative motion is @xmath24 , and @xmath25 is the reduced mass of the colliding atoms . the doppler shift is described by @xmath26 where the magnitude of the laser light wave vector is @xmath27 and the mass of the molecule created in the photoassociation process is @xmath28 .
the shift of the photoassociation resonance @xmath29 caused by the photon recoil is also included here . finally , the light induced shift of a given photoassociation resonance can be expressed as a sum @xmath30 of contributions @xmath31 of all possible optical transitions between the excited bound state `` @xmath17 '' and ground scattering states `` @xmath16 '' . similarly , the total width of the resonance @xmath32 , where @xmath33 is the light induced width between the `` @xmath17 '' and `` @xmath16 '' states and @xmath34 describes all possible other processes leading to loss of the excited state . if radiative decay is the dominant loss process , then @xmath34 can be taken as the natural width of the excited molecular state .
the light induced width @xcite @xmath35 is linearly dependent on the light intensity @xmath13 through the operator @xmath36 describing optical coupling between particular excited and ground states .
for the case investigated here more details about this operator can be found in refs .
this width also depends on the kinetic energy of the relative motion of the two colliding atoms @xmath37 .
this is because the energy normalized ground scattering state @xmath38 strongly depends on this energy .
finally the magnitude of the light - induced width is dependent on the unit normalized excited bound state @xmath39 .
the light induced shift @xcite can be calculated from the fano theory @xcite @xmath40 where @xmath41 is a principal part integral over all collision energies , and the sum occurring in this expression is taken over all unity normalized bound state @xmath42 of the ground electronic potential .
like in the case of the light induced width , the light induced shift is linearly dependent on the laser intensity @xmath13 and of course on @xmath37
. the photoassociation process can be viewed as a case of the optical feshbach resonance @xcite .
the optical coupling of the ground scattering state to the excited bound state changes both the amplitude and the phase of the scattering wave function . the amplitude can be detected through loss of atoms from the trap .
the phase change can also be detected @xcite .
this approach leads to conclusion that the scattering length describing ultra cold collisions of the two ground state atoms can be modified by light due to the coupling with an excited bound state .
this effect was demonstrated with a bose - einstein condensate ( bec ) of @xmath43rb @xcite .
the expression for a scattering length modified by light can be written in the following form @xcite : @xmath44 } { [ \delta - \delta_{e}-e_{e}(i,0)-e_{\rm rec , mol}]^{2}+[\gamma_{e}(i,0)/2]^{2 } } \,,\ ] ] where @xmath45 is the background scattering length of the system in absence of the light .
the optical length @xmath46 @xcite is @xmath47 because of the @xmath48-wave threshold law , this quantity approaches an energy - independent constant as @xmath49 that varies linearly with light intensity @xmath13 .
the optical length characterizes the strength of an optical feshbach resonance , namely the ability of light to change the scattering length of the ground state atoms ; see also ref .
@xcite for a review of feshbach resonances , including optical ones .
the scattering length can be changed on the order of its background magnitude @xmath50 while minimizing losses if the optical length is very large compared with @xmath50 so that the optical control can be achieved at large detuning .
the particular properties of the photoassociation resonances are strongly dependent on the properties of the atomic interaction of the colliding atoms . in general
we describe the colliding system using the hamiltonian operator @xmath51 . in this expression @xmath12
is the kinetic energy operator for relative radial motion , @xmath52 is the atomic hamiltonian operator representing internal atomic degrees of freedom , @xmath53 is the interaction operator described by nonrelativistic molecular born - oppenheimer potentials , and @xmath54 is the rotational energy operator .
reference @xcite gives more details for the case investigated here .
let us first focus on the interaction in the excited state .
in this paper we limit our discussion to photoassociation near the intercombination transition @xmath0@xmath1 .
therefore we will discuss the atomic interaction properties only near the dissociation limit @xmath55 . in this case
it is convenient to use the @xmath56 basis . here
@xmath57 is the total electron angular momentum , @xmath58 is the rotational angular momentum , and @xmath59 is the total angular momentum .
the projections of @xmath60 on a space - fixed @xmath61 axis is @xmath62 .
finally @xmath63 is the total parity .
we add an index @xmath17 to all quantities introduced here to indicate that they correspond to the excited electronic state .
it should be noted that @xmath64 as well as @xmath65 are good quantum numbers of the hamiltonian @xmath66 . in our case with @xmath67 and @xmath68 we can solve the schrdinger equation using only two channels for each @xmath64 : one with @xmath69 and one with @xmath70 . in this basis
the interaction operator @xmath53 is not diagonal .
its matrix elements can be expressed in terms of potentials @xmath71 and @xmath72 which correspond to states with @xmath73 and @xmath74 , respectively , where @xmath75 is the projection of the total electron angular momentum @xmath76 along the interatomic axis .
the other components of the hamiltonian operator , @xmath12 , @xmath52 , and @xmath54 , are diagonal in our basis .
the matrix elements of the sum @xmath77 can be written in a compact form @xcite : @xmath78 \\
\frac{\sqrt{j_{e}(j_{e}+1)}}{2j_{e}+1}[v_{1}(r)-v_{0}(r ) ] & \frac{j_{e}+1}{2j_{e}+1}v_{0}(r)+\frac{j_{e}}{2j_{e}+1}v_{1}(r)+b(r)(j_{e}+2)(j_{e}+1 ) \\
\end{array } \right ) \end{array } \begin{array}{c } \vspace{3 mm } \\
\vspace{3 mm } l_{e}=j_{e}-1 \vspace{3mm}\\ l_{e}=j_{e}+1 \end{array}\ ] ] where @xmath79 .
henceforth we will omit explicit indication of the @xmath80-dependence of @xmath81 , @xmath82 and @xmath83 .
analytic diagonalization of this matrix gives the following eigenvalues : @xmath84 \right\ } \
\label{equ9 } { \cal v}_{1 } & = & \frac{1}{2}\left\ { v_{0}+v_{1}+(v_{1}-v_{0})\sqrt{1-\frac{4b}{v_{1}-v_{0}}+\frac{4b^{2}(2j_{e}+1)^{2}}{(v_{1}-v_{0})^{2}}}+2b[j_{e}(j_{e}+1)+1 ] \right\ } \,.\end{aligned}\ ] ] these eigenvalues @xmath85 and @xmath86 can be used as effective potentials for single channel calculations of excited bound states .
such an approach allows one to approximate the influence of coriolis coupling on the molecular structure in simple single channel calculations .
the quantum number @xmath75 becomes a good one in the range of @xmath80 where @xmath87 , i e. , when the anisotropy of the interaction between two atoms is much bigger than the rotation energy .
in such a case the eigenvalues and corresponding eigenvectors of the matrix in eq .
( [ equ7 ] ) can be written in the following form : @xmath88 \longrightarrow \left ( \begin{array}{c } -\sqrt{j_{e}/(2j_{e}+1)}\vspace{3mm}\\ \sqrt{(j_{e}+1)/(2j_{e}+1 ) } \end{array } \right ) \,,\\ \label{equ11 } { \cal v}_{1 } & = & v_{1}+bj_{e}(j_{e}+1 ) \hspace{8 mm } \longrightarrow \left ( \begin{array}{c } \sqrt{(j_{e}+1)/(2j_{e}+1)}\vspace{3mm}\\ \sqrt{j_{e}/(2j_{e}+1 ) } \end{array } \right ) \,.\end{aligned}\ ] ] in many cases these potentials provide a successful approximation for single channel calculations of bound states of real systems .
such an approach is particularly good for yb , because the potential curves at long range are dominated by the resonant dipole interaction . in the case of yb
@xmath89 up to very large @xmath80 .
this is a quite different case from sr @xcite , where the @xmath90 value is more than 20 times smaller than for yb . in the case of sr
the competition between the van der waals and resonance interactions is important . in the range of @xmath80 where the interaction is dominated by the resonant dipole term the ungerade potentials for @xmath73 and @xmath74 are attractive and repulsive , respectively .
therefore only the potential @xmath85 will support a series of bound states having @xmath91 symmetry near the dissociation limit .
the @xmath86 potential becomes attractive at small @xmath80 due to chemical bonding and may support a bound state near the dissociation limit . in the rest of the paper
we will describe excited bound states of @xmath91 symmetry using single channel solutions of the schrdinger equation .
this simple approach to the description of atomic interactions allows us to write the appropriate expressions for the light induced width and shift : @xmath92 and @xmath93 in eqs .
( [ equ12 ] ) and ( [ equ13 ] ) @xmath94 is the natural decay width of the atomic transition and @xmath95 the wavelength of the atomic transition .
the dimensionless rotational line strength factor for a transition from the ground scattering state to the excited @xmath91 state in a bosonic isotope has the simple approximate form obtained by machholm _ et al_. @xcite : @xmath96 for even @xmath22 .
this expression can be obtained using the approach described in ref .
@xcite . to extract the rotational line strength factor @xmath97
one can start from eq .
( [ equ3 ] ) and use eq .
( b3 ) from ref .
@xcite . in our case , this allows us to show that the laser radiation couples the ground scattering channel having @xmath98 only with the excited state channel having the same @xmath99 .
therefore the light induced width , eq .
( [ equ3 ] ) , is proportional to the fraction of the total wave function of the excited bound state in the channel with @xmath100 .
this fraction can be found from eqs .
( [ equ10 ] ) and ( [ equ11 ] ) .
moreover , the light induce width is proportional to the square of the proper clebsch - gordan coefficient @xmath101 , where @xmath102 for the dipole transition and @xmath103 describes the polarization of light .
this way one can introduce the rotational line strength factor @xmath104 which is dependent on the light polarization @xmath105 and the projection @xmath106 of the total angular momentum @xmath22 in the ground electronic state , see ref .
@xcite . the rotational line strength factor given by eq .
( [ equ14 ] ) can be obtained as an average @xmath107 over all possible orientations of the total angular momentum @xmath22 ; compare refs .
this quantity is independent of the light polarization @xmath105 . to derive eq .
( [ equ14 ] ) one can use the following relation @xmath108 which is fulfilled for @xmath109 , @xmath102 and @xmath110 .
equation ( [ equ14 ] ) can be also expressed in terms of a wigner 3-@xmath111 symbol , see ref .
@xcite : @xmath112 this work is limited only to the case of weak interaction with light .
therefore we can ignore the dependence of the light induced width and shift on the projection @xmath106 and use the projection independent eq .
( [ equ14 ] ) .
the total angular momentum @xmath22 in the ground scattering state is the same as the rotational angular momentum @xmath113 .
this is a consequence of the fact that the atomic interaction in the ground electronic state has @xmath114 symmetry and the total electronic angular momentum @xmath115 . in homonuclear pairs of bosonic atoms
only even @xmath22 are allowed .
finally , the franck - condon factors @xcite : @xmath116 and @xmath117 can be expressed in terms of the regular @xmath118 and irregular @xmath119 solutions of the the schrdinger equation for the ground scattering state and the wave function @xmath120 for the excited bound state .
the expressions for the light induced width and shift can be further simplified by using the reflection approximation @xcite .
the franck - condon factors can then be written in the following form : @xmath121 and @xmath122 where @xmath123 the fraction @xmath124 is the mean vibrational spacing , @xmath125 and @xmath126 are the effective potentials in the excited and ground electronic states . here
@xmath127 is the condon point .
when the reflection approximation is applicable , the condon point is approximately the classical outer turning point for the excited bound state .
the reflection approximation is very good for a number of yb bound states close to the @xmath55 dissociation limit .
this is in contrast with other species such as ca or sr , for which the reflection approximation needs to be modified @xcite . since eqs .
( [ equ12 ] ) and ( [ equ21 ] ) show that the light induced width is proportional to @xmath128 , photoassociation spectroscopy provides an excellent tool to investigate the properties of the ground state scattering wave function @xcite .
this will be explored below for photoassociation of yb near the intercombination line .
experimental information is crucial for determining the parameters that describe the long range interaction near the @xmath129s@xmath130@xmath131p@xmath132 dissociation limit .
we used the experimental data obtained by tojo _
et al_. @xcite to find the parameters needed to describe @xmath133 for the @xmath91 state . as described in ref .
@xcite , yb atoms were decelerated by a zeeman - slowing laser for the @xmath129s@xmath130@xmath129p@xmath132 transition , and were collected in a magneto - optical trap ( mot ) with a laser for the @xmath129s@xmath130@xmath131p@xmath132 transition .
after the compression of the mot , the atoms were transferred into a crossed optical trap with laser beams at 532 nm .
the atoms were evaporatively cooled by decreasing the trap potential depth .
typically , mot duration time is 10 s and evaporative cooling time is 6 s. a laser beam for photoassociation was applied to the trapped atoms , and the number of remaining atoms was measured through an absorption image with the @xmath129s@xmath130@xmath129p@xmath132 transition after the release from the optical trap .
the applied photoassociation light intensity varied between 6.5 @xmath25w/@xmath134 to 90 mw/@xmath134 , depending on which excited level was being probed .
we obtained the atom - loss spectra by scanning the photoassociation laser frequency .
these spectra allowed the determination of the bound states energies of a series of @xmath135 and @xmath136 @xmath91 levels for each of the two isotopic species .
table [ tab1 ] lists the measured energies along with their error bars .
the data for the @xmath137yb@xmath138 @xmath135 states were taken at about 4 @xmath139k , and the other data in table [ tab1 ] were taken at temperatures in the range from 5 to 27 @xmath139k .
the shift with temperature and light intensity of the photoassociation features was taken into account in the data analysis .
these shifts are mostly responsible for the magnitude of the error bars listed in table [ tab1 ] .
[ cols= " > , > , > , > , > , > , > , > , > , > , > , > " , ] reference @xcite fits ground state binding energy data for two isotopes using a similar form for the ground state potential without the dipolar term , @xmath140 - \frac{c^{(g)}_{8}}{r^8 } \,;\ ] ] this model with mass scaling predicted the observed binding energies for 4 other isotopic molecules with an error on the order of @xmath141 mhz or less .
here we have refitted the ground state potential by simultaneously fitting the data for all isotopes in ref . @xcite .
the fit has a slightly better @xmath142 than the previous one , but does not represent a significant improvement . for the sake of completeness , we give the model parameters for the global fit with enough significant digits to reproduce the calculated values in table [ tab4 ] : @xmath143 , @xmath144 and @xmath145 .
the @xmath146 and @xmath147 obtained here as well as in ref .
@xcite agree very nicely with those calculated by zhang and dalgarno @xcite .
the reported values in ref .
@xcite are @xmath148 and @xmath149 , respectively , with an estimated uncertainty of 10% .
it should be emphasized that the shape of our potential can significantly differ form the real one .
nevertheless our model potential should correctly represent the number of vibronic bound states in the ground electronic state of yb@xmath150 molecule . therefore even if our model is significantly different from an _ ab initio _ potential like that reported by buchachenko
_ et al_. @xcite it should give about the same number of vibronic bound states . it should be noted that similar analysis to that for yb @xcite was carried out recently for the scattering properties in the ground electronic state of various isotopes of sr by martinez _
et al _ , using a realistic potential @xcite .
the scattering properties from this work are in very good agreement with results obtained using a potential derived from fourier transform spectroscopy by stein _
et al_. @xcite .
as discussed in section [ singchan ] , the strength and shape of a photoassociation line is to a large extent determined by the ground state scattering wave function at the condon point for the transition . since the condon points for photoassociation to
near - threshold excited states tend to be at quite large internuclear separation @xmath80 , the scattering wave function needs to be known only at relatively large @xmath80 .
consequently , in order to describe photoassociation in the ultra low scattering energy regime , the detailed knowledge of the atomic interaction at short range can be compressed to a very few parameters .
the most important quantity describing scattering during an ultracold collision is the scattering length @xmath151 .
if the long range potential has the van der waals form @xmath152 , the scattering length is very well approximated by a simple analytical relation given by gribakin and flambaum @xcite @xmath153\ ] ] see also refs . @xcite . here
the mean scattering length @xmath154 is a characteristic length associated with the van der waals potential , where @xmath155 is the gamma - function .
this length also defines a characteristic energy for the van der waals potential , @xmath156 .
the phase @xmath157 is defined by @xmath158 where @xmath159 is the inner classical turning point of @xmath160 at zero energy .
the number of bound states @xmath161 in the potential is @xcite @xmath162 + 1,\ ] ] where the brackets @xmath163 $ ] mean the integer part .
the scattering length varies periodically with phase @xmath157 , having a singularity when @xmath164 .
this variation can be observed experimentally for different isotopes of the same species .
if we assume that the interaction potential is the same for all isotopes so that mass scaling applies , @xmath165 can be varied by changing the reduced mass @xmath25 .
the relation between scattering length , the energies of near threshold bound states and the reduced masses of the colliding atoms was carefully studied by kitagawa _
et al_. @xcite for ground state interactions of various yb isotopes .
it is useful to define a reduced mass difference @xmath166 needed to change @xmath165 by @xmath167 , that is , to change the number of bound states in the potential by one .
it is approximately @xmath168 since @xmath169 for @xmath137yb , we see that a mass difference of 5 atomic mass units is sufficient to change the number of yb@xmath138 bound states by one .
alternatively , varying @xmath25 continuously by 5 atomic mass units will cause the scattering length to vary across a singularity over its full range from @xmath170 to @xmath171 , as found by ref .
there are actually 7 stable isotopes of yb , and thus 28 different discrete physical values of @xmath172 that are available in the laboratory using different isotopic combinations . and @xmath173 , the respective ground state and excited state potentials .
the upper curve shows the outer turning points of each calculated @xmath174 , @xmath135 bound level for @xmath137yb .
the dotted lines indicate the experimentally observed range of levels .
the arrow indicates the mean scattering length @xmath175 .
the lower panel shows the centrifugal barriers for the long range ground state potential for @xmath176 0 ( @xmath48-wave ) , 2 ( @xmath177-wave ) , 4 ( @xmath16-wave ) , and 6 ( @xmath178-wave ) .
an energy @xmath179 1 mk is equivalent to @xmath180 21 mhz . ]
the long - range ground state scattering wave function at very low collision energy depends on three basic parameters , namely the separation between colliding atoms @xmath80 , the relative kinetic energy of collision @xmath37 , and the quantum defect associated with the phase @xmath157 or , equivalently , the scattering length @xmath151 .
the reduced mass @xmath25 is a parameter that allows control of the quantum defect .
thus , the long - range scattering wave function for the van der waals system has an universal character that can be expressed by a system - independent function of the dimensionless variables @xmath181 , @xmath182 and @xmath183 .
all of the needed information about the ground state scattering and bound state wave functions can be calculated from a knowledge of the scaling parameters @xmath181 , @xmath182 and @xmath184 and the value of reduced mass @xmath185 for which the scattering length is singular in a given system .
the analytical solutions of the schrdinger equation are known for several class of potentials , thus allowing the derivations of compact expressions for the scattering length @xcite .
van der waals systems such as those discussed here can be nicely described and very well understood using analytical theory .
gao @xcite used the framework of quantum defect theory for a van der waals system to work out a number of practical formulas and results .
for example ,
when the @xmath48-wave scattering length is singular , there is a bound state at @xmath186 not only for the @xmath48-wave but also for @xmath187 . on the other hand
if the scattering length is equal to @xmath188 there will be zero energy bound states for @xmath189 .
the practical consequence for real potentials is that when the @xmath48-wave scattering length is near such special values , these other threshold bound states show up as shape resonances for other partial waves .
a shape resonance is a quasibound state with enhanced amplitude trapped behind a centrifugal barrier that can lead to enhanced photoassociation .
while the wave functions we show are calculated numerically by solving the schrdinger equation for our potential , the analytic results are very helpful for interpreting their features , and for evaluating approximations like the reflection formula .
as we already emphasized before , a photoassociation experiment can be used to map the square of the scattering wave function in the ground electronic state at the condon points corresponding to the different excited bound levels .
figure [ fig1 ] shows on the lower panel the effective potential , @xmath190 , in the ground electronic state , where @xmath191 is the rotational quantum number for the partial wave @xmath192 .
[ fig1 ] shows the centrifugal barrier for a few partial waves with low @xmath191 .
the upper panel of this figure shows the excited state @xmath174 potential together with bound states and their outer turning points , which are essentially the same as their condon points .
this panel shows how the @xmath193 value that is probed changes with the binding energy of the excited bound level . probing experimental levels with binding energies @xmath194 between 2000 mhz and 3 mhz
allow probing the scattering wave function between around 80 a@xmath195 and 800 a@xmath195 . in order to illustrate some properties of the scattering wave function we will begin by setting the collision energy to a constant value corresponding to @xmath196 @xmath25k .
as we will show in the next section , shape resonances could be clearly observed at this energy , which is below the top of the @xmath16-wave centrifugal barrier .
figure [ fig2 ] shows the dependence of the square of the wave function magnitude on @xmath25 and @xmath80 .
changing @xmath25 corresponds to the change of the quantum defect of the colliding system , which changes @xmath151 and the bound state spectrum .
calculations have been done for a few lowest partial waves @xmath48 , @xmath177 , @xmath16 , and @xmath178 . in fig .
[ fig2 ] the brighter areas show regions of higher amplitude and the dark lines show the nodes of the wave function .
the effect of the exclusion of the wave function from short range by the centrifugal barrier is especially evident for the @xmath16-wave and the @xmath178-wave . for this yb system ,
the singularities in the @xmath48-wave scattering length occur for @xmath197 167.3 , 172.1 , and 176.9 @xmath198 @xcite .
near these values , an @xmath48-wave node moves in to smaller distances on the order of @xmath199 as @xmath172 increases and a new bound state appears in the potential .
there is also a @xmath16-wave shape resonance with a dramatic enhancement of short - range amplitude near these @xmath172 values where @xmath151 is singular , as expected from the analytic van der waals quantum defect theory .
similarly , there are @xmath177- and @xmath178-wave shape resonances near the values of @xmath200 169.7 and 174.5 @xmath198 where @xmath201 for the yb system . near these shape resonances
, a node moves to shorter distances inside the centrifugal barrier as @xmath172 increases and there is an extra bound state in the potential .
if the collision energy @xmath37 is lowered , the wave functions show similar patterns , except that the short - range wave function is much more attenuated due to lower penetration through the centrifugal barrier . clearly
, with 28 different physical values for @xmath172 available , we can expect significant isotopic variation in the photoassociation spectra , depending on the range of condon points sampled and the temperature of the sample .
spectral lines with condon points outside of the barrier will have less isotopic sensitivity , whereas lines with condon points in or inside the region of the barrier will show much more sensitivity to the isotopic combination .
in this section we will show how the scattering properties of the different yb isotopes affect their photoassociation spectra . as an example we performed calculations of the light induced trap - loss coefficient at various gas temperatures in range from @xmath202 , to @xmath203 .
results for @xmath204 are particularly interesting , since the line shapes are relatively sharp and ground scattering resonances can have significant influence on the spectrum . to emphasize the qualitative differences between spectra of the different isotopes
we first focus our attention on relatively deep bound states having binding energy around 2000 mhz . as can be seen in fig .
[ fig1 ] this corresponds to the condon point placed near or inside the centrifugal barriers .
bound states of the @xmath205yb@xmath150 molecule near the @xmath135 level at @xmath206 mhz ; other rotational levels with the same vibrational quantum number are evident .
the top panel shows contributions to the light induced width @xmath207 coming from the optical coupling of the ground scattering states having @xmath176 0 ( @xmath48-wave ) , 2 ( @xmath177-wave ) , 4 ( @xmath16-wave ) , and 6 ( @xmath178-wave ) with excited bound states having @xmath208 , as a function of the collision energy @xmath37 , calculated for the light intensity @xmath209 .
the @xmath48-wave scattering length is quite large in magnitude , @xmath210 @xcite , so that this system has a @xmath16-wave shape resonance near @xmath211 @xmath25k that leads to resonantly enhanced @xmath212 and 5 features in the photoassociation spectrum .
the lower three panels show contributions to the fully thermally averaged loss rate @xmath213 coming from the optical coupling of the ground scattering states having @xmath176 0 ( @xmath48-wave ) , 2 ( @xmath177-wave ) , 4 ( @xmath16-wave ) , and 6 ( @xmath178-wave ) with excited bound states having @xmath208 and @xmath214 , as a function of the laser detuning from the atomic resonance , calculated for various temperatures @xmath215 , @xmath216 , and @xmath217 and the light intensity @xmath218 . ] we start our discussion from @xmath205yb which has a large negative scattering length near an @xmath48-wave singularity @xcite .
consequently , we expect a @xmath16-wave shape resonance at low scattering energy .
the upper panel of fig .
[ fig3 ] shows the energy dependence of the light induced width for the @xmath2191 , 3 , 5 and 7 levels with the same vibrational quantum number as the @xmath220 level .
the light induced widths were calculated for scattering states with total angular momentum @xmath221 .
the widths for the same scattering partial wave @xmath191 but with @xmath222 will be very similar in energy variation .
the @xmath16-wave resonance can be clearly seen in upper panel of fig .
[ fig3 ] . to see how the photoassociation spectrum changes with the temperature of the gas sample we have calculated the thermally averaged light induced trap - loss coefficients for temperatures @xmath22310 @xmath25k , 100 @xmath25k and 1 mk ; see fig .
these correspond to @xmath224 of 210 khz , 2.1 mhz , and 21 mhz respectively . at temperature 10
@xmath25k the spectrum is narrow , with a width determined mainly by the spontaneous emission rate , and is dominated by @xmath48-wave scattering , for which only the @xmath135 state is visible .
the calculated spectrum at temperature @xmath204 has three lines .
one is the @xmath135 line due to @xmath48-wave scattering .
this line has a normal thermal line shape @xcite .
two other strong and sharp lines corresponding to @xmath136 and @xmath225 bound states are due to the @xmath16-wave resonance shown in upper panel of fig .
both lines have subthermal width ; see ref .
the shape of these two lines is determined mostly by the shape of the @xmath16-wave resonance .
finally , at a temperature of 1 mk one can clearly see the very broad thermally broadened lines coming from @xmath48 and @xmath177-wave partial waves .
the @xmath177-wave feature is much broader than the @xmath48-wave one .
this is caused by the fact that for collision energy around 1 mk the light induced width for the @xmath48-wave decreases with an increase of the collision energy , while for the @xmath177-wave the light induced width increases .
this affects the shape of the line and leads to some shifting and widening of the line coming from the @xmath177-wave compared to that coming from the @xmath48-wave .
in addition , the @xmath16-wave shape resonance gives rise to a sharp subthermal feature . bound state of the @xmath137yb@xmath150 molecule near the @xmath135 level at @xmath226 mhz .
the panels are the same as described in the caption of fig [ fig3 ] .
the @xmath48-wave scattering length is @xmath227 , which is about @xmath228 , and there is a broad @xmath177-wave shape resonance near @xmath229 @xmath25k near the top of the @xmath177-wave centrifugal barrier @xcite . ]
the second example is for @xmath137yb , which has a measured scattering length relatively close to @xmath188 @xcite . in such case
one can expect that both a @xmath177-wave and an @xmath178-wave resonance can occur near threshold .
as with fig .
[ fig3 ] for @xmath205yb fig .
[ fig4 ] shows the light induced width for the @xmath137yb bound - states with the total angular momentum @xmath2191 , 3 , 5 and 7 , with the same vibronic quantum number as the level at @xmath230 mhz .
the figure clearly shows the @xmath177 and @xmath178-wave resonances .
as for @xmath205yb , the spectrum at a temperature of 10 @xmath25k is dominated by the @xmath48-wave component . however , unlike @xmath205yb , the @xmath177-wave shape resonance leads to a weak @xmath136 feature even at such low temperature .
the spectra at @xmath231 @xmath25k and 1 mk clearly show the influence of the @xmath177-wave resonance .
this resonance had a crucial role in the correct interpretation of photoassociation spectra near the resonance transition @xmath232 and in the determination of the scattering length for this isotope @xcite .
when the spectrum is dominated by the features due to the ground state @xmath177-wave , the intensity ratio of the @xmath136 and @xmath135 features should be 3/2 from eq .
( [ equ14 ] ) .
such @xmath177-wave doublets in @xmath137yb were observed by tojo _
et al _ @xcite , who found that the transition to the bound state with @xmath136 can be stronger than the transition to @xmath135 at relatively low collisions energies corresponding to a temperature of about @xmath233 , what can be explained by our model .
the spectrum at temperature 1 mk is also dominated by the @xmath177-wave component .
however one can notice sharp structure on the top of a weak and broad feature due to the @xmath225 bound - state , mostly due to the ground state @xmath16-wave .
the sharp structure is a consequence of the @xmath178-wave resonance .
the @xmath178-wave component has a subthermal width connected to the @xmath178-wave resonance seen in upper panel of fig .
one can also notice a weak line corresponding to @xmath234 bound state supported by this @xmath178-wave resonance .
bound state of the @xmath235yb@xmath150 molecule near the @xmath135 level at @xmath236 mhz .
the panels are the same as described in the caption of fig [ fig3 ] .
the @xmath48-wave scattering length is @xmath237 , not near any special value , but there is a broad @xmath16-wave shape resonance at relatively high energy near @xmath238 mk near the top of the @xmath16-wave centrifugal barrier @xcite . ]
figure [ fig5 ] shows our calculations for @xmath235yb .
this isotope has a relatively small negative scattering length @xcite .
the figure shows the light induced width for @xmath235yb bound - states with the total angular momentum @xmath2191 , 3 , 5 and 7 , with the same vibronic quantum number as the level at @xmath239 . although the scattering length is not near any special value for van der waals quantum defect theory , a wide @xmath16-wave resonance is clearly seen relatively high in energy near 1 mk . the spectrum at 10 @xmath25k is typical with only an @xmath48-wave line that can be observed . upon increasing the temperature to 100 @xmath25k
the spectrum is still dominated by the @xmath48-wave component , although noticeable @xmath177-wave components begin to appear .
moreover very weak @xmath16-wave components occur in this spectrum .
the picture changes dramatically when the temperature increases to 1 mk , where the @xmath16-wave components dominate the spectrum .
bound state of the @xmath235yb@xmath150 molecule near the @xmath135 level at @xmath240 mhz .
the panels are the same as described in the caption of fig [ fig3 ] . in this case
the binding energy is much smaller than in figs .
[ fig3 ] , [ fig4 ] , and [ fig5 ] , and the condon points are outside the centrifugal barriers except when the collision energy is very low ; see fig .
[ fig1 ] . ]
finally we would like to show how the spectrum is affected when the condon point for the transition is moved outside the region of the centrifugal barrier .
this is done by choosing a level with much smaller binding energy in our previous examples .
figure [ fig6 ] shows the light induced width for the @xmath205yb bound - states with the total angular momentum @xmath2191 , 3 , 5 and 7 , with the same vibronic quantum number as the level at @xmath241 mhz .
the upper panel of fig .
[ fig6 ] shows that there are no shape resonances , since there are no amplitude enhancements from being behind the centrifugal barrier .
the lower panels of fig .
[ fig6 ] show the calculated spectra at 10 @xmath25k , 100 @xmath25k , and 1 mk .
the higher temperature 100 @xmath25k spectrum is dominated by broad @xmath177-wave components , on top of which a sharp subthermal @xmath48-wave component is clearly seen .
this is because there is a node in the @xmath48-wave scattering wave function at the condon point at relatively low scattering energy , unlike for other partial waves .
this sharp structure is a nice example of the subthermal line shapes discussed by machholm _
et al_. @xcite in the context of alkaline earth photoassociation .
increasing the temperature to 1 mk leads to a quasi - continuum as several partial waves contribute , as seen from fig .
[ fig6 ] . the examples in this section show that the photoassociation spectra of various isotopes of the same species can differ qualitatively .
this variety is directly connected with the quantum defect in the ground electronic state which is dependent on reduced mass and scattering length of the colliding species .
this sensitivity of the rotational structure of photoassociation lines can help in the determination or verification of the scattering length , as it was recently done in case of calcium by vogt _
et al_. @xcite . of the @xmath48-wave photoassociation features as a function of the detuning @xmath14 of the resonance from the atomic transition for the several bosonic isotopes of yb .
the optical length determines the strength of an optical feshbach resonance .
the optical length varies linearly with laser intensity @xmath13 and these values have been normalized to @xmath242 for @xmath243 .
] a realistic description of near threshold interaction in the ground and excited electronic states of the yb@xmath150 molecule is possible thanks to the experimental data @xcite and the calculations we have shown here .
our models of the interaction potentials can be used to predict the magnitude of the optical lengths in eq .
( [ equ6 ] ) for the set of optical feshbach resonances corresponding to the photoassociation lines of the different yb bosonic isotopes ( fermionic isotopes are more complicated because of hyperfine structure in the excited state @xcite ) . figure [ fig7 ] shows our calculated optical lengths for the homonuclear bosonic isotopes of yb .
the isotope @xmath205yb offers the strongest optical feshbach resonance because of its large negative scattering length , which enhances the amplitude of the @xmath48-wave ground state wave function in the region of the condon points .
bound levels near the excited state threshold have optical lengths at 1 w @xmath244 on the order of @xmath245 a@xmath195 , similar to a value measured for a weakly bound excited level of @xmath246sr @xcite .
this suggests that these resonances may be of practical use for changing the scattering length for useful time scales while reducing spontaneous emission loss processes by using large detuning .
et al . _
@xcite have experimentally demonstrated the possibility of some degree of optical feshbach control in ultracold yb gases .
our predicted optical lengths should be helpful in choosing good resonances for controlling ultracold yb collisions by light .
one could naively think that this is a simple system so that not much difference should be observed between the spectra of three bosonic isotopes like @xmath205yb , @xmath137yb , and @xmath235yb .
however , our calculations show that for temperatures about 100 @xmath25k the spectra for various isotopes are qualitatively different .
this difference is manifested by the fact that different rotational lines are apparent in different isotopes .
moreover the shapes of photoassociation lines are also affected .
some lines have subthermal widths depending on the isotope .
these differences can be well understood as due to the isotopic variation in the properties of the scattering wave function in the ground electronic state .
there is a clear connection between shape resonances in ground scattering states and the appearance of strong lines coming from higher partial waves at relatively low temperature of the order 100 @xmath25k .
however , this only happens for excited bound states that have sufficiently short - range condon points inside the location of the centrifugal barrier . on the other hand , near - threshold bound states with long - range condon points outside the centrifugal barrier
do not show resonance enhancement .
for example , we showed that the presence of @xmath136 lines for @xmath205yb at small detunings does not indicate that a @xmath177-wave resonance occurs .
thus , a good experimental method to determine whether resonances are present or not would be to measure photoassociation lines for more deeply bound excited states that have turning points near to or less that the ground state centrifugal barrier .
the existence of resonances correlates with the approximate value of the @xmath48-wave scattering length .
the intercombination photoassociation spectra of yb are qualitatively different from spectra for group ii atoms such as ca and sr . for the case of ca theory shows that the excited state potential in the turning point range for levels near the dissociation limit is dominated by the van der waals potential , since the resonant dipolar interaction is so small @xcite .
the reflection approximation is not applicable in such a case @xcite .
strontium @xcite is a system which is half way between ca and yb .
only the last three sr bound states closest to the dissociation limit can be treated as dominated by the resonant dipole interaction , while more deeply bound levels are determined mostly by the van der waals part of the interaction .
photoassociation can be observed in ca or sr for two series of bound states of @xmath91 and @xmath247 symmetry , respectively .
although the resonant dipole interaction for the @xmath247 state is repulsive , its weakness allows it to be overcome by the attractive van der waals interaction .
in contrast the yb interaction in the excited electronic state with dissociation limit @xmath55 is dominated by the resonance interaction , so that only one potential curve of @xmath91 symmetry is attractive at long range and supports a series of detectable bound states .
the reflection approximation is well applicable in the case of @xmath91 levels of yb .
since the @xmath247 potential becomes attractive at much shorter range of the interatomic separation , the density of bound states near threshold is much less , and only one bound state with @xmath247 symmetry is expected near within a few ghz of threshold .
the heavier group iib elements such as cd and hg will be similar to yb .
for both species the interaction in the excited state is dominated by the resonance interaction , and therefore one can expect that the reflection approximation will be applicable . moreover both species have numerous stable isotopes .
consequently , mass tuning of the scattering length should be applicable for both of these species , and the results obtained in this work can be treated as universal and generic for cd and hg .
this work was partially supported by grant - in - aid for scientific research of jsps ( 18204035 ) and gcoe `` the next generation of physics , spun from universality and emergence '' from mext of japan .
psj was supported in part by the office of naval research .
the research is part of the program of the national laboratory famo in toru , poland and partially supported by the polish mnisw ( project
n n202 1489 33 ) . | the properties of bosonic ytterbium photoassociation spectra near the intercombination transition @xmath0@xmath1 are studied theoretically at ultra low temperatures .
we demonstrate how the shapes and intensities of rotational components of optical feshbach resonances are affected by mass tuning of the scattering properties of the two colliding ground state atoms .
particular attention is given to the relationship between the magnitude of the scattering length and the occurrence of shape resonances in higher partial waves of the van der waals system .
we develop a mass scaled model of the excited state potential that represents the experimental data for different isotopes .
the shape of the rotational photoassociation spectrum for various bosonic yb isotopes can be qualitatively different . |
spin effects in high energy fragmentation processes have attracted much attention recently [ see , e.g. , refs . [ 1 - 12 ] and the references given therein ] .
one of the important issues is the study of spin transfer from the fragmenting quark to the produced hadrons .
it can be studied by measuring the polarizations of the hadrons produced in the fragmentation of a polarized quark .
hyperon polarization has been frequently used in this connection@xcite . on the other hand ,
it has been pointed out@xcite that the vector meson polarization can also be used for this purpose since the polarization of a vector meson can also be studied easily in experiments .
furthermore , compared with hyperon , the production rate of a vector meson is usually higher and the origin is simpler due to the much smaller contribution from the decay of heavier hadrons .
this implies less theoretical uncertainties in calculations and better statistics in experiments .
therefore , the study of polarization of the vector mesons may provide us important information to understand the role of spin in hadronic reactions .
the polarization of a vector meson is described by the spin density matrix @xmath2 or its element @xmath3 , where @xmath4 and @xmath5 label the spin component along the quantization axis , usually the @xmath6-axis of the frame .
the diagonal elements @xmath7 , @xmath8 and @xmath9 for the unit - trace matrix are the relative intensities of meson spin component @xmath4 to take the values @xmath10 , @xmath11 , and @xmath12 respectively . in experiment ,
@xmath8 can be measured from the angular distribution of its decay products and a deviation of @xmath8 from @xmath13 indicates spin alignment . in the helicity basis , i.e. , in the case that the @xmath6-axis is chosen as the moving direction of the vector meson , the matrix is usually called helicity density matrix and @xmath8 represents the probability of the vector meson to be in the helicity zero state .
measurements have been carried out in different reactions@xcite in particular in @xmath0 annihilation at lep recently @xcite .
the data@xcite show that @xmath8 is significantly larger than @xmath13 for vector mesons with large @xmath6 and increases with increasing @xmath6 ( here @xmath14 , where @xmath15 is the momentum of the vector meson , @xmath16 is the total @xmath0 center of mass energy ) .
this means that there exist a significant spin alignment for vector mesons in helicity frame in @xmath0 annihilation at the @xmath17 pole , and the effect is more significant for the large momentum fraction region .
the simple statistic model@xcite based on spin counting gives the result of @xmath18 .
a qcd - inspired model is given in ref .
[ [ augustin80 ] ] , where a fast quark combines a soft anti - quark to form a vector meson and the soft anti - quark preferentially has the same helicity as the fast one .
this model leads to @xmath19 .
another model@xcite describes the production of vector meson through the channel @xmath20 , with the vector meson coupling to the quark like a vector current . in this case , the helicity - conserving vector current gives rise to @xmath21 .
none of these predictions is consistent with the data@xcite . in a recent paper@xcite ,
we calculated the helicity density matrix of vector mesons inclusively produced in @xmath0 annihilation at @xmath17 pole by taking the spin of a vector meson which contains a fragmenting quark as the sum of the spin of the polarized fragmenting quark and that of the anti - quark created in the fragmentation process . by comparing the obtained results with the data@xcite
, we showed that the experimental results for @xmath8 imply a significant polarization for the anti - quark which is created in the fragmentation process and combines with the fragmenting quark to form the vector meson .
it should be polarized in the opposite direction as that of the fragmenting quark and the polarization can approximately be written as , @xmath22 where @xmath23 is a constant ; @xmath24 is the polarization of the anti - quark in the moving direction of the fragmenting quark and @xmath25 is the longitudinal polarization of the fragmenting quark of flavor @xmath26 .
the results of @xmath8 obtained from above relation are in good agreement with the data for different mesons .
since the calculations are rather straight - forward , the relation given by eq . ( [ eq1 ] ) can be considered as a direct implication of the data@xcite in @xmath0 annihilation .
it implies that there exists a spin interaction between the fragmenting quark and the anti - quark during the hadronization process
. it should be interesting to extend the calculations to other processes and make further tests to see whether the relation is universal , in the sense that it is true for quark fragmentation in all different reactions .
hence we made similar calculations for polarized lepton - nucleon deep - inelastic scattering ( dis ) process and polarized @xmath1 collisions .
the results for dis have been given in ref.[[xl02 ] ] .
in this paper , we present the calculations and the results for high @xmath27 vector mesons in polarized @xmath1 collisions , where factorization theorem can be used so that fragmentation effects and other effects such as those from the structure functions can be studied separately .
compared with those in the lepton - induced reactions such as @xmath0 annihilations and dis , the study of high @xmath27 vector mesons in @xmath1 collisions has the following properties .
( 1 ) the involved hard scatterings are strong interaction processes rather than the electroweak processes in the lepton - induced reactions , so the corresponding cross section should in general be larger . furthermore , the luminosity of the incoming proton beams can in general be made higher than that for leptons .
hence , the statistics in experiment can be better .
( 2 ) not only longitudinally but also transversely polarized quarks can be produced , so that we can study the properties in both cases .
( 3 ) there are many different hard subprocesses which contribute to the high @xmath27 hadron production .
this makes the study more interesting and , at the same time , more complicated than those in lepton - induced reactions .
for example , the gluon fragmentation is also involved here , which is unclear even for the unpolarized case .
it is very important to know whether its contribution is significant in different kinematic regions .
we will first make an estimation of it in next section using a monte carlo event generator and then present the calculation method of spin alignment of vector mesons in @xmath1 collisions . in section iii , we calculate the spin alignment of vector mesons with high @xmath27 in the helicity frame in @xmath1 collisions with one longitudinally polarized proton beam and present the results . in section iv , we extend the study to the the spin alignment of high @xmath27 vector mesons for the case that one proton beam is transversely polarized .
finally , a short summary is given in section v.
in this section , we first summarize the calculation method of spin alignment of vector mesons that was used in @xmath0 annihilation , and then extend it to the inclusive high @xmath27 vector meson production in longitudinally polarized @xmath1 collisions .
the calculation method for vector mesons produced in @xmath0 annihilation has been presented in ref.[[xll01 ] ] .
the main points are summarized as follows . for @xmath0 annihilation at a given energy ,
we need to consider vector mesons produced in the fragmentation of a polarized initial quark @xmath28 of flavor @xmath26 and those of polarized initial anti - quark @xmath29 .
the polarization of @xmath28 or @xmath29 is a given constant which can be calculated using the standard model for electroweak interaction . to calculate the spin density matrices of these vector mesons , we divide them into the following two groups and consider them separately : ( a ) those which contain the fragmenting quark @xmath28 ; ( b ) those which do nt contain the fragmenting quark .
the spin density matrix @xmath30 for the vector meson @xmath31 is given by : @xmath32 where @xmath33 and @xmath34 are respectively the average number and spin density matrix of the vector mesons from group ( a ) ; @xmath35 and @xmath36 are those from ( b ) ; @xmath37= @xmath38 + @xmath35 is the total number of vector mesons and @xmath6 is the momentum fraction of the initial quark s momentum carried off by the vector meson .
the average numbers @xmath33 and @xmath35 are independent of the spin properties and can be calculated using an event generator based on a fragmentation model which gives a good description of the unpolarized data .
we used generator @xmath39@xcite in our calculations . in terms of the usually used cross section and fragmentation functions , eq.([eq2 ] )
has the following form , @xmath40 where @xmath41 and @xmath42 are the fragmentation functions of the initial quark @xmath43 ( anti - quark ) into the vector mesons of group ( a ) and ( b ) with momentum fraction @xmath6 respectively ; @xmath44 is the usual fragmentation function ; @xmath45 is the cross section of the process @xmath46 .
we see that , @xmath47 is just the cross section of the production of the meson @xmath31 which contains the initial quark via the fragmentation of @xmath26-flavor quark in @xmath0 annihilation , i.e. , @xmath48 .
it is proportional to the average number @xmath33 in eq .
( [ eq2 ] ) .
the denominator is the total cross section of inclusive vector meson production , which corresponds to @xmath37 of eq .
( [ eq2 ] ) .
there are many different possibilities for the production of of the vector mesons in group ( b ) , we take them as unpolarized , i.e. , @xmath36=1/3 . for those vector mesons from group ( a ) , i.e. , those which contain @xmath28 and an anti - quark @xmath49 created in the fragmentation , the spin is taken as the sum of the spins of @xmath43 and @xmath49 .
the polarization of @xmath43 is taken as the same as that before the fragmentation .
then the spin density matrix @xmath34 can be calculated from the direct product of the spin density matrix @xmath50 for @xmath28 and @xmath51 for @xmath52 .
transforming the direct product , @xmath53=@xmath50@xmath54@xmath55 , to the coupled basis @xmath56 ( where @xmath57=@xmath58+@xmath59 ) , we can obtain the spin density matrix of the vector meson @xmath34 . in the helicity frame of @xmath43 ,
i.e. , @xmath6-axis is taken as the moving direction of @xmath43 , the density matrix of the vector meson of type ( a ) is given by , @xmath60 where @xmath25 is the longitudinal polarizations of @xmath28 and @xmath61 is the polarization vector of the anti - quark @xmath49 . hence , the 00-component of the density matrix takes the following simple form , @xmath62 in which there is only one unknown variable @xmath24 , the polarization of @xmath49 along the moving or polarization direction of @xmath43 . using eqs .
( [ eq2 ] ) and ( [ eq4 ] ) , we can determine @xmath24 in different cases by fitting the data@xcite in the @xmath0 annihilation at @xmath17 pole for the production of different vector mesons .
as mentioned in the introduction , we found that@xcite , the data for different vector mesons can reasonably be fitted if we take the @xmath24 satisfying the relation given by eq.(1 ) .
now , we insert eq.(1 ) into eq .
( [ eq4 ] ) and obtain that@xcite , @xmath63 finally , from eqs .
( [ eq2 ] ) and ( [ eq5 ] ) , we have @xmath64 now we come to the high @xmath27 vector meson production in polarized @xmath1 collisions .
such vector mesons come mainly from the fragmentation of outgoing partons in the hard scattering subprocesses
. the outgoing parton can be a gluon or a quark , and the cross section of gluon - involved subprocess is even larger than others .
however , the gluon distribution and fragmentation functions are both poorly known yet , especially for the polarized case .
fortunately , the momentum fractions carried by the gluons in a proton are usually very small , so the gluon contributions to very high @xmath27 hard scattering subprocesses are suppressed .
in addition , it is known that the gluon fragmentation function is softer than the quark fragmentation function .
consequently , for the final hadron production with high @xmath27 , the contribution from gluon fragmentation is much smaller than that from quark fragmentation@xcite . to see it numerically
, we calculate these contributions using @xmath39 . in fig .
[ meta132 ] , we show the results for inclusive @xmath65 production with @xmath27@xmath6613 gev at @xmath67=500 gev in @xmath1 collisions .
we see that , at such high @xmath27 cutoff , the contribution of gluon fragmentation is indeed much smaller than that of quark fragmentation , especially for the large @xmath68 region ( @xmath68 is the pseudo - rapidity of the vector meson ) .
thus , as we did in studying hyperon polarization in high @xmath27 jets@xcite , we neglect the contribution of polarized gluon fragmentation to the spin alignment of vector mesons , and consider only the contribution of polarized quark s fragmentation . for quark fragmentation
, we can use the same method of calculating spin alignment as that used in @xmath0 annihilation . to do it
, we need to know the polarization of the outgoing quark in the hard subprocess .
such quark is also longitudinally polarized when one proton beam is longitudinally polarized and its polarization can be calculated in pqcd .
the obtained polarization can be written as , @xmath69 where @xmath70 is the longitudinal polarization transfer factor in the scattering from the incoming parton to the outgoing parton ; @xmath70 is only a function of @xmath71 which is defined as @xmath72 for the subprocess @xmath73 , where @xmath74 , @xmath75 , @xmath76 and @xmath77 are the four - momenta of the partons @xmath78 , @xmath79 , @xmath80 and @xmath81 respectively .
@xmath70 has been calculated using pqcd and the leading order results can be found in many publications , for example , in ref.[[xll02 ] ] .
we take the parton @xmath78 as the parton from the polarized proton beam , and the polarization of the incoming parton @xmath78 is given by @xmath82 , where @xmath83 and @xmath84 are the helicity and unpolarized distribution functions at momentum fraction @xmath85 and scale @xmath86 . therefore , the longitudinal polarization @xmath87 is the function of variables @xmath85 , @xmath86 and @xmath71 .
since we are now discussing @xmath1 collisions with one polarized beam , it is clear that the outgoing quarks with positive momenta in @xmath6 direction have a much larger probability of being polarized than those with negative momenta when the positive axis of @xmath6 is taken as the direction of the incoming polarized proton .
similar to that in @xmath0 annihilation , to calculate the spin density matrix of the high @xmath27 vector mesons in @xmath1 collisions , we also divide them into two groups ( a ) and ( b ) according to whether they contain an outgoing quark in the hard scattering and consider them separately .
the difference is that , now the fragmenting quark is an outgoing quark in the hard scattering subprocesses and its polarization is not a constant .
it depends on the origin of this quark and is given by eq.([polq7 ] ) . obviously , it is a function of @xmath85 , @xmath86 and @xmath71 .
hence , @xmath8 for vector mesons of group ( a ) should be also a function of @xmath85 , @xmath86 and @xmath71 and it is given by , @xmath88 to obtain @xmath8 for vector mesons at a given pseudo - rapidity @xmath68 , we need to replace the products in eq.([eq3 ] ) with the corresponding convolutions , i.e. , @xmath89\otimes \sigma_{pp\to q_fx } } { \sum_f d_{v / f}(z_c)\otimes\sigma_{pp\to q_fx } } \label{eqpp01}\ ] ] to write it more precisely , we have , @xmath90 } { \int d^2{p_{t}}\sum_{abcd } \int dx_a dx_b f_a(x_a,\mu^2)f_b(x_b,\mu^2 ) \frac { d \hat { \sigma } } { d\eta}(ab \to cd ) d_{v / c}(z_c,\mu^2 ) } } , \label{rhopp}\ ] ] where @xmath27 is the transverse momentum of the vector meson ; @xmath91 and @xmath92 are the unpolarized distribution functions of partons @xmath78 and @xmath79 in proton at the scale @xmath93 , @xmath85 and @xmath94 are the corresponding momentum fractions carried by @xmath78 and @xmath79 ; @xmath95 is the usual fragmentation function of parton @xmath80 into vector meson @xmath31 , @xmath96 and @xmath97 are those of parton @xmath80 into vector mesons of group ( a ) and ( b ) respectively , @xmath98 is the momentum fraction of parton @xmath80 carried by the produced @xmath31 ; @xmath99 is the cross section at the parton level .
the cross section of the hard subprocess can be calculated using perturbative qcd .
the summation in eq .
( [ rhopp ] ) runs over all possible subprocesses .
the fragmentation functions are independent of the spin properties and can be calculated using an available monte carlo event generator . in practice
, we generate a @xmath1 collision event and then search for the produced vector meson @xmath31 .
we then calculate its contribution to @xmath100 by tracing back its origin : if the vector meson @xmath31 belongs to group ( b ) , then its contribution to @xmath100 is given by @xmath101 .
if it belongs to group ( a ) , we then need to further trace back the origin of the fragmenting quark to calculate its polarization @xmath102 using eq.([polq7 ] ) .
we insert the @xmath102 into eq.([eq8 ] ) to get the contribution @xmath103 of such vector meson to @xmath100 . after running the program for sufficiently large number of events
, we obtain the final @xmath100 . in terms of a mathematic formula
, the calculation in this procedure is expressed as , @xmath104/ { n(\eta ) } , \label{eq11}\ ] ] where @xmath105 and @xmath106 are the numbers of vector mesons of group ( a ) and ( b ) as a function of @xmath68 , and @xmath107 is the total number of vector mesons . the results for @xmath100 in longitudinally polarized @xmath1 collisions are presented in next section .
using the method described in last section , we calculate the @xmath8 s for different vector mesons in @xmath1 collisions with one proton beam longitudinally polarized .
we now summarize the obtained results in the following . using the generator @xmath39@xcite
, we first calculate the different contributions to vector meson production , i.e. , those of group ( a ) and ( b ) . as an example
, we show the results for @xmath108 at @xmath67=500 gev and @xmath27@xmath6613 gev in fig .
[ korg ] . from these results
, we can see clearly that , the decay contribution is indeed very small .
the contribution of those containing an outgoing quark in the hard scattering subprocess is very high , even larger than @xmath109 for large @xmath68 region .
the results of @xmath8 s for different vector mesons are shown in fig .
[ vfig13 ] .
we see that , the spin alignment for @xmath108 , @xmath110 and @xmath65 are significantly high .
the obtained @xmath8 s for these mesons increase from 1/3 to about 0.45 with increasing @xmath68 .
there is nearly no spin alignment in the negative @xmath68 region since the scattered quark moving in the negative @xmath6 direction is almost unpolarized .
the spin alignment for @xmath111 is smaller in both positive and negative @xmath68 region , this is because @xmath81 quark fragmentation dominate @xmath111 production at high @xmath27 and the polarization of @xmath81 quark @xmath112 is smaller than that of @xmath113 quark @xmath114 .
we can also make calculations for other vector mesons , such as @xmath115 , @xmath116 , and @xmath117 .
however , since these mesons do not contain @xmath113 or @xmath81 valence quark , which dominate the high @xmath27 hadron production in @xmath1 collisions , their production rates at high @xmath27 are small .
this makes the statistics much worse than those for @xmath108 , @xmath111 , @xmath110 and @xmath65 . for the same reason ,
the @xmath118 quark distribution is much more important .
hence , the spin alignment for these mesons are much more sensitive to the helicity distribution of the strange quark in nucleon , which is less precisely determined yet .
it is thus not a good choice to study the spin transfer in fragmentation process by measuring the spin alignment of these mesons .
but , it is possible to use them to study the helicity distribution of the strange sea in nucleon if they can be measured with good accuracy .
when one of the proton beams is transversely polarized , the incoming quark @xmath78 for the elementary hard scattering process can also be transversely polarized in the same direction .
its polarization @xmath119 is determined by the transversity distribution function@xcite @xmath120 and the unpolarized distribution function @xmath84 , i.e. , @xmath121 .
there is no gluon transversity distribution at leading twist@xcite , thus the incoming gluon in the hard scattering is not transversely polarized , which is different from the longitudinally polarized case .
the transverse polarization can also be transferred from the incoming quark to the outgoing quark in the hard scattering .
the polarization direction of the outgoing quark is also transverse to the moving direction of the quark but is in general different from that of the incoming quark .
both the magnitude and the direction of the polarization of the outgoing quark can be calculated@xcite using pqcd for the hard elementary process .
the results are summarized in the following .
we recall that , for a quark @xmath78 with transverse polarization @xmath119 , the spin density matrix @xmath122 in the helicity basis is given by , @xmath123 where @xmath117 is the angle between the polarization direction and the @xmath124 axis .
the @xmath125-plane is perpendicular to the moving direction of the quark , which is taken as the @xmath6-direction ; and the @xmath124-axis is taken as the normal direction of the scattering plane .
the spin density matrix of the outgoing quark in the helicity basis can be obtained from @xmath126 given in eq .
( [ eqrho1 ] ) and the helicity amplitudes of the scattering .
the result is given by@xcite , @xmath127 where @xmath128 is a real function of @xmath71 defined in section 2b for the scattering and is called the polarization transfer factor . @xmath128 for different subprocesses have been calculated using pqcd and can be found , for example , in ref.[[collins94 ] ] . from eq .
( [ eqrho2 ] ) , we see that the outgoing quark is also transversely polarized .
the magnitude of the the polarization is given by @xmath129 , i.e , @xmath130 we see also that , the angle between the polarization direction and the @xmath124-axis remains the same before and after the scattering .
more precisely , the polarization vector of the outgoing quark is in the plane transverse to the moving direction of the quark and the angle between the polarization direction and the normal of the scattering plane is the same as that for the incoming quark . in another word , the direction of transverse polarization of the incoming and that of the outgoing quark are related to each other by a rotation around the normal of the scattering plane , which changes the moving direction of the quark from the incoming to the outgoing direction .
fig . 2 of ref.[[collins94 ] ] ) .
we note that the direction of the polarization of the outgoing quark depends not only on the scattering angle but also on the azimuthal direction .
in contrast , the magnitude of the polarization depends only on the scattering angle .
the dependence is given by the corresponding spin transfer function @xmath128 , which is only a function of @xmath71 defined above .
having seen that the outgoing quarks are also transversely polarized if the incoming protons are transversely polarized , we now discuss the spin alignment of the vector mesons produced in the fragmentation of such quarks .
we recall that , the relation given in eq . ( [ eq1 ] ) is obtained for the fragmentation of the longitudinally polarized quarks by fitting the @xmath0 annihilation data@xcite .
it shows that the anti - quark which combines the fragmenting quark to form the vector meson is polarized in the opposite direction as that of the fragmenting quark . to test whether the relation is in general true for the fragmentation of a longitudinally polarized quark in any high energy reactions
, we extended the calculations of the spin alignment of vector mesons to other lepton - induced reactions in [ [ xl02 ] ] and longitudinally polarized @xmath1 collisions in last sections .
now we further assume that the relation is also true for the fragmentation of a quark polarized in any direction , i.e , we extend the relation to , @xmath131 we use this relation to calculate the spin alignment of the vector mesons in transversely polarized @xmath1 collisions so that this extension can be tested in future experiments . the calculations are carried out in a way similar to that in the longitudinally polarized case .
we now present the calculations and the results in the following two different frames .
\(1 ) in the helicity frame : in the helicity frame of the outgoing quark of the hard scattering , its spin density matrix is off - diagonal and is given by eq .
( [ eqrho2 ] ) . from the relation given by eq .
( [ barqt ] ) , the spin density matrix of the anti - quark which combines with the fragmenting quark to form the vector meson is , @xmath132 we build the direct product of @xmath133 given by eq.([eqrho2 ] ) with @xmath134 given by eq.([eqrho3 ] ) , transform it to the coupled basis , and obtain the spin density matrix of the vector meson of group ( a ) as , @xmath135 where the superscript
@xmath136 " denotes the helicity frame .
hence , the @xmath137 for such vector mesons in the helicity basis is given by , @xmath138 we can see that , as can be expected , if we look only at the @xmath139-component of the spin density matrix , the result is independent of the azimuthal direction of the transverse polarization , i.e. independent of the angle @xmath117 .
it depends only on the magnitude of the transverse polarization @xmath140 , thus on the variables @xmath141 and @xmath71 .
we see also that @xmath142 is smaller than @xmath13 , which is different from the results in the longitudinally polarized case .
this shows that , if the fragmenting quark is transversely polarized , the produced vector meson should have a larger probability to be in the helicity @xmath143 states .
as long as there exists the contribution of the vector mesons of group ( a ) , the final results for @xmath144 in this frame should also be smaller than @xmath13 .
we know from fig .
2 that the contribution of the group ( a ) increase with increasing @xmath68 , so we expect a decreasing @xmath144 as @xmath68 increases for vector mesons such as @xmath108 and @xmath145 .
it is significantly smaller than @xmath13 for large @xmath68 .
hence , the measurements of @xmath144 of the vector mesons in transversely polarized @xmath1 collisions should give a good check of the extension of the relation of eq .
( 1 ) to the transversely polarized case .
\(2 ) in the transversity frame : to compare with the results in the helicity frame for the longitudinally polarized case , we also calculate the spin alignment of the vector mesons in the transversely polarized case in the `` transversity frame '' . in this frame ,
similar to the helicity frame for longitudinally polarized case , we take the ( transverse ) polarization direction of the fragmenting quark as the quantization axis . the spin density matrix of the outgoing quark and that of the anti - quark are both diagonal in this frame , and they are given by @xmath146 @xmath147 where the superscript
@xmath148 " denotes the transversity frame .
we see that they have completely the same form as those for the corresponding quarks and anti - quarks in longitudinally polarized case in the helicity frame .
the corresponding spin density matrix @xmath149 , which can be obtained from the direct product @xmath150 , is also diagonal , and the result for the @xmath139-component @xmath151 is given by , @xmath152 we note that , since @xmath149 is diagonal in the transversity frame , there is a simple relation between @xmath142 and @xmath151 .
the relation can easily be obtained by using @xmath153 and @xmath154 .
( here , @xmath155 denotes the helicity and @xmath156 denote the spin component along the quantization axis in the transversity frame . ) it is given by , @xmath157 we see that , @xmath158 can also be obtained from this relation and the result for @xmath159 given by eq .
( [ eq14 ] ) .
we also see that , the obtained @xmath158 given by eq.([eq15 ] ) is in general larger than @xmath13 .
this means that , the probability for the produced vector mesons to be in the spin states which are transverse to the moving direction are relatively smaller .
we see that , although the values of @xmath158 and @xmath159 are different , they represent the same physical meaning .
we should note that , to measure @xmath160 in experiments , we need to determine the quantization axis in this frame . since quantization axis is chosen as the polarization direction of the quark before the fragmentation , it can be different for different vector mesons .
the polarization direction is determined by eq.([eqrho2 ] ) . in practice ,
we first find out the normal direction of the scattering plane and then determine the polarization direction according to the rules described below eq .
( [ eqpqt ] ) .
( c.f . , also fig.2 of ref.[[collins94 ] ] ) .
the moving direction of the fragmenting quark can be replaced by the thrust axis of the quark jet .
inserting eqs .
( [ eq14 ] ) and ( [ eq15 ] ) into eq.([eq11 ] ) , we can calculate @xmath161 for different vector mesons in the above two frames in @xmath1 collisions with one beam transversely polarized . for these calculations ,
we first need to calculate @xmath140 using eq.([eqpqt ] ) .
so far , the transversity distribution @xmath162 is not known yet . in order to get a feeling of how large the spin alignment can be
, we make an estimation using the simple form for @xmath163 in the light - cone su(6 ) quark - spectator model@xcite as input . in this model , the @xmath163
is given by , @xmath164{\widetilde { m}}_s(x ) -\frac{1}{6}d_v(x){\widetilde { m}}_v(x);\end{aligned}\ ] ] @xmath165 where @xmath166 and @xmath167 are related to the melsh - wigner rotation , which can be calculated according to the procedure given in [ [ bqma98 ] ] .
the obtained results of @xmath161 for different vector mesons at @xmath168 gev and @xmath169 gev in the helicity frame are shown in fig .
[ vfigt13h ] and those in the transversity frame are shown in fig .
[ vfigt13 ] . for comparison
, we also show the results obtained by taking the upper limit of @xmath162 in soffer s inequality@xcite , @xmath170.\ ] ] from these figures , we see that , in the helicity frame , the @xmath8 s of vector mesons are smaller than @xmath13 in the large @xmath68 region , but the effect is not very significant . in the transversity frame , however , the @xmath8 s of @xmath108 , @xmath110 and @xmath65 are larger than @xmath13 and increase to about 0.5 with increasing @xmath68 , which are similar to the results of longitudinally polarized case .
however , the results of @xmath111 s @xmath8 in both frames are much near the unpolarized case of 1/3 .
this is because , the @xmath81 quark fragmentation dominates the @xmath111 s production at high @xmath27 , and according to the soffer s inequality , @xmath171 should be small , since @xmath172 is negative for valence quark .
as we mentioned above , the gluon is not transversely polarized , so it is much safer to assume that there is no contribution of gluon fragmentation to the spin alignment of vector mesons in transversely polarized @xmath1 collisions .
therefore , the high @xmath27 cut which we invoked in longitudinally polarized case to ensure a small contribution of gluon fragmentation is unnecessary now . here
, only a relative high @xmath27 cut is required to ensure the validity of pqcd .
we therefore make also the calculations for @xmath173 gev at @xmath168 gev and the results are given in fig .
[ vfigt8h ] and fig .
[ vfigt8 ] .
the results are much similar to those obtained in the case of higher @xmath27 cut @xmath169 gev .
in summary , we calculate the spin alignment of vector mesons with high @xmath27 in high energy @xmath1 collisions with one longitudinally polarized beam by taking the spin of vector meson as the sum of the spin of the scattered quark and that of the secondary anti - quark produced in the fragmentation process . the results for different vector mesons at rhic energy are presented .
these results show that there exist quite significant spin alignment for @xmath108 , @xmath110 and @xmath65 .
we also study the spin alignment of vector mesons in the case that one of the proton beams is transversely polarized . using the same method as that in the longitudinally polarized case
, we obtain the results in the helicity and the transversity frame .
the results show that , in the large @xmath68 region , the @xmath8 s in the helicity frame are smaller than @xmath13 , while the @xmath8 s in the transversity frame are larger than @xmath13 . the magnitudes of the effect depend on the transversity distribution of the quark in proton .
they are expected to be more significant in the large @xmath68 region .
measurements of them can provide useful information for spin effects in strong interaction and the spin structure of nucleon . 0.3 cm we thank li shi - yuan , si zong - guo , xie qu - bing and other members in the theoretical particle physics group of shandong university for helpful discussions .
this work was supported in part by the national science foundation of china ( nsfc ) under the approval number 10175037 and the education ministry of china under huo ying - dong foundation .
99 j. collins , nucl .
b*396 * , 161 ( 1993 ) .
k. chen , g.r .
goldstein , r.l .
jaffe , and x. ji , nucl .
b*445 * , 380 ( 1995 ) .
jaffe , phys .
rev . d*54 * , r6581 ( 1996 ) .
a.d . adamov and g.r .
goldstein , phys .
rev . d*56 * , 7381 ( 1997 ) ; d*64 * , 014021 ( 2001 ) .
[ ada97 ] c. boros and liang zuo - tang , phys .
rev . d*57 * , 4491 ( 1998 ) .
a. kotzinian , a. bravar and d. von harrach , eur .
j. c*2 * , 329 ( 1998 ) .
m. anselmino , m. boglione , u. dalesio , and f. murgia , eur .
j. c*21 * , 501(2001 ) ; m. anselmino , d. boer , u. dalesio , and f. murgia , phys . rev . d*63 * , 054029 ( 2001 ) ; m. anselmino , m. bertini , f. murgia , p. quintairos , eur .
j. * c11 * , 529 ( 1999 ) .
[ abm00 ] b.q .
ma , i. schmidt and j.j .
yang , phys .
d*61 * , 034017 ( 2000 ) ; b.q . ma ,
i. schmidt , j. soffer and j.j .
yang , phys .
rev . d*62 * , 114009 ( 2000 ) ; phys .
rev . d*64 * , 014017 ( 2001 ) , erratum - ibid , d*64 * , 099901 ( 2001 ) ; nucl .
a*703 * , 346 ( 2002 ) .
liu chun - xiu and liang zuo - tang , phys .
rev . d*62 * , 094001 ( 2000 ) ; liu chun - xiu , xu qing - hua and liang zuo - tang , phys .
rev . d*64 * , 073004 ( 2001 ) ; liang zuo - tang and liu chun - xiu , phys .
rev . d*66 * , 075302 ( 2002 ) .
[ lxl01 ] xu qing - hua , liu chun - xiu and liang zuo - tang , phys .
d 65 , 114008 ( 2002 ) .
[ xll02 ] xu qing - hua , liu chun - xiu and liang zuo - tang , phys . rev .
d 63 , 111301(r ) ( 2001 ) .
[ xll01 ] xu qing - hua , and liang zuo - tang , phys . rev .
d 66 , 017301 ( 2002 ) .
[ xl02 ] i. cohen _ et al_. , phys
d 25 , 634 , ( 1982 ) . [ epr82 ] bebc wa59 collab . , w. witttek
_ et al_. , phys .
* b187 * , 179 ( 1987 ) ; bbcn collab .
et al_. , z. phys . * c66 * , 583 ( 1995 ) .
excharm collab .
, a. n. aleev _
et al_. , phys .
b485 * , 334 ( 2000 ) .
delphi collab .
, p. abreu _ et al_. , phys . lett . *
b406 * , 271 ( 1997 ) ; opal collab . , k. ackerstaff
_ et al_. , phys . lett . *
b412 * , 210 ( 1997 ) ; opal collab . , k. ackerstaff _
et al_. , z. phys . * c74 * , 437 ( 1997 ) ; opal collab . , g. abbiendi
_ et al_. , eur .
j. * c16 * , 61 ( 2000 ) .
[ kress01 ] i.i .
y. bigi , nuovo cimento * a41 * , 581 ( 1977 ) . j.e .
augustin and f.m .
renard , nucl . phys . *
b162 * , 341 ( 1980 ) .
[ augustin80 ] j.f .
donoghue , phys . rev . * d19 * , 2806 ( 1979 ) .
[ donoghue ] t. sjstrand _ et al_. , computer physics commun .
135 , 238 ( 2001 ) .
we note that the result of eq.(5 ) is obtained in the helicity frame of the fragmenting quark .
but , in experiment , the spin alignment of produced vector meson is usually measured in the helicity frame of the vector meson . as we mentioned in ref.[[xll01 ] ] , when the energy of the fragmenting quark is high , such as at lep and rhic , the moving direction of the produced meson is very close to that of the fragmenting quark .
the difference of the final results between these two frames is small . j.p .
ralston , d.e .
soper , nucl .
b*152 * , 109 ( 1979 ) ; x. artru , m. mekhfi , z. phys . *
c45 * , 669 ( 1990 ) ; r.l .
jaffe , x.d .
ji , phys .
67 , 552 ( 1991 ) ; nucl .
b*375 * , 527 ( 1992 ) ; v. barone , a. drago , p.g .
ratcliffe , phys .
359 , 1 ( 2002 ) .
collins , s.f .
heppelmann , and g.a .
ladinsky , nucl .
b*420 * , 565 ( 1994 ) .
[ collins94 ] b.q .
ma , i. schmidt , and j. j. yang , phys . rev .
* d 63 * , 037501 ( 2001 ) .
[ bqma ] b.q .
ma , i. schmidt , and j. soffer , phys . lett .
* b 441 * , 461 ( 1998 ) .
[ bqma98 ] j. soffer , phys .
74 , 1292 ( 1995 ) .
[ soffer ] m. glck , e. reya , m. stratmann , w. vogelsang , phys . rev . d*63 * , 094005 ( 2001 ) .
[ grsv2000 ] m. glck , e. reya , and a. vogt , eur .
j. * c 5 * , 461 ( 1998 ) .
fig.1 : ( a ) contributions of quark and gluon fragmentation to @xmath174 production , and ( b ) the ratio of the quark fragmentation s contribution to the total rate , in @xmath175@xmath176@xmath177 at @xmath67=500 gev and @xmath27@xmath6613 gev . 0.3 cm fig.2 : different contributions to @xmath108 production in @xmath175@xmath176@xmath178 at @xmath67=500 gev and @xmath27@xmath6613 gev as functions of @xmath68 . here
containing hard u " and similar denote the contributions of those containing the outgoing quark from the hard scattering subprocess . 0.3 cm fig.3 : spin alignment of different vector mesons with @xmath27@xmath6613 gev in @xmath175@xmath176@xmath179 at @xmath67=500 gev in helicity frame when one proton beam is longitudinally polarized .
the standard lo set of grsv2000@xcite and lo set of grv98@xcite are used for helicity and unpolarized distribution functions respectively .
the scale @xmath93 is taken as the transverse momentum of the hard scattering subprocess .
0.3 cm fig.4 : spin alignment of different vector mesons with @xmath27@xmath6613 gev in @xmath180@xmath176@xmath179 at @xmath67=500 gev in the helicity frame when one of the proton beams is transversely polarized .
the solid lines denote the results obtained using the light - cone quark - spectator model for @xmath163 ; the dotted lines correspond to the results obtained using the upper limit in soffer s inequality .
0.3 cm fig.5 : spin alignment of different vector mesons with @xmath27@xmath6613 gev in @xmath175@xmath176@xmath179 at @xmath67=500 gev in the transversity frame when one of the proton beams is transversely polarized .
the solid lines denote the results obtained using the light - cone quark - spectator model for @xmath163 ; the dotted lines correspond to the results obtained using the upper limit in soffer s inequality . | using the same method as that used in @xmath0 annihilation and lepton - nucleon deep - inelastic scattering , we calculate the spin alignment of vector mesons with high transverse momenta in high energy @xmath1 collisions with one longitudinally polarized beam .
we present the results obtained at rhic energies .
we also study the spin alignment of such vector mesons in the case of a transversely polarized proton beam and present its relation to the transversity distribution in the nucleon .
numerical results obtained using simple models for the transversity distributions are also given .
these results show that there exist significant spin alignment in both cases , and they can be measured in experiments such as those at rhic . |
in this paper a _ hyperbolic surface _ is a two dimensional complete riemannian manifold @xmath0 with sectional curvature equal to @xmath15 .
such a surface is isomorphic to the quotient @xmath16 , of the poincar upper halfplane @xmath17 by a _ fuchsian group _ @xmath18 , i.e. a discrete torsion - free subgroup of psl(2,@xmath19 ) .
the _ laplace operator _ on @xmath17 is the differential operator which associates to a @xmath20-function @xmath21 the function @xmath22 since the action of psl(2 , @xmath23 ) on @xmath24 leaves @xmath25 invariant , @xmath25 induces a differential operator on @xmath26 which extends to a self - adjoint operator @xmath2 densely defined on @xmath27 it is a general fact that the laplace operator is a non - positive operator whose spectrum is contained in the smallest interval @xmath28 \subset { \mathbb r^{- } } \cup \{0\}$ ] with @xmath29 .
[ d1 ] let @xmath30 be a real number and @xmath31 be a nonzero function on @xmath0 .
the pair @xmath32 is called an eigenpair of @xmath0 if @xmath33 on @xmath0 where @xmath34 and @xmath21 are respectively called an _ eigenvalue _ and an _ eigenfunction _ ( sometimes a _ @xmath34-eigenfunction _ ) .
when @xmath35 , we add the adjective _ small _ i.e. @xmath32 , @xmath34 and @xmath21 are respectively called a _ small eigenpair _ , a _ small eigenvalue _ and a _ small eigenfunction .
_ we begin with a noncompact , finite area hyperbolic surface @xmath0 of type @xmath1 i.e. @xmath9 .
the laplace spectrum of such a surface is composed of two parts : _ the discrete part _ and _ the continuous part _ @xcite .
the continuous part covers the interval @xmath36 and is spanned by _
eisenstein series _ with multiplicity @xmath37 .
eisenstein series are not eigenfunctions although they satisfy @xmath38 because they are not in @xmath39 .
for this reason , they are called _ generalized eigenfunctions_. the discrete spectrum consists of eigenvalues .
they are distinguished into two parts : _ the residual spectrum _ and _ the cuspidal spectrum_. an eigenpair @xmath32 is called _ residual _ if @xmath21 is a linear combination of residues of meromorphic continuations of eisenstein series .
such @xmath34 and @xmath21 are respectively called a _ residual eigenvalue _ and a _ residual eigenfuction_. the residual spectrum is a finite set contained in @xmath40 . on the other hand ,
an eigenpair @xmath32 is called _ cuspidal _ if @xmath21 tends to zero at each cusp . in this case @xmath34 and @xmath21 are respectively called a _ cuspidal eigenvalue _ and a _ cuspidal eigenfuction_. these eigenvalues with multiplicity are arranged by increasing order and we denote @xmath41 the @xmath37-th cuspidal eigenvalue of @xmath0 . for an arbitrary fuchsian group @xmath18 , it is not known whether the cardinality of the set of cuspidal eigenvalues of @xmath42 is infinite . however , a famous theorem of a. selberg says that it is the case when @xmath18 is arithmetic .
any cuspidal eigenpair @xmath32 with @xmath43 is called a _ small cuspidal eigenpair _ and in that case , @xmath34 and @xmath21
are respectively called a _ small cuspidal eigenvalue _ and a _ small cuspidal eigenfunction_. in @xcite , jean - pierre otal and eulalio rosas proved that the total number of small eigenvalues of any hyperbolic surface of type @xmath1 is at most @xmath44 . in the same paper
they formulate the following : * conjecture . *
_ let @xmath0 be a noncompact , finite area hyperbolic surface of type @xmath1
. then @xmath45 . _
this conjecture is motivated by the following two results * ( huxley @xcite , otal @xcite ) * let @xmath0 be a finite area hyperbolic surface of genus @xmath46 or @xmath47
. then @xmath0 does not carry any small cuspidal eigenpair . *
( otal @xcite ) * let @xmath0 be a finite area hyperbolic surface of type @xmath1 .
then the multiplicity of a small cuspidal eigenvalue of @xmath0 is at most @xmath48 .
the set @xmath49 carries a topology for which two surfaces @xmath50 and @xmath51 are close when the groups @xmath18 and @xmath52 can be conjugated inside psl@xmath53 so that they have generators which are close . with this topology @xmath49
is not compact .
however it can be compactified by adjoining @xmath54 s for each @xmath55 with @xmath56 . in this compactification
a sequence @xmath57 converges to @xmath58 if and only if for any given @xmath59 the @xmath60-thick part @xmath61 converges to @xmath62 in the gromov - hausdorff topology .
recall that the @xmath60-thick part of a surface @xmath0 is the subset of those points of @xmath0 where the _ injectivity radius _ is at least @xmath60 .
recall also that the injectivity radius of a point @xmath63 is the radius of the largest geodesic disc that can be embedded in @xmath0 with center @xmath64 . for any @xmath65 and @xmath66
we define the sets @xmath67 it is clear that @xmath68 for @xmath69 . with this notation
the conjecture can be formulated by saying that @xmath70 in this paper , we study the sets @xmath71 .
the methods developed here are not sufficient to prove the conjecture but we show that the sets @xmath72 and @xmath73 ( @xmath74 ) contains neighborhoods of certain strata in the compactification of @xmath75 .
@xmath76 for any integer @xmath77 , @xmath78 is an open subset of @xmath79 . + * ( ii ) * @xmath80 contains a neighborhood of @xmath12 in @xmath6 .
+ * ( iii ) * @xmath13 contains a neighborhood of @xmath14 in @xmath6 .
observe that it is theoretically possible for a residual eigenfunction to converge to a cuspidal eigenfunction .
therefore indicating that @xmath81 may not be continuous .
also , the result @xcite suggest that @xmath81 may not be continuous at those @xmath9 where it takes value strictly more than @xmath82 .
therefore , the first assertion is not completely trivial .
the paper is organized as follows . in 1 we recall some preliminaries for convergence of hyperbolic surfaces in @xmath6 . in 2 and 3 we study convergence properties of eigenpairs on converging hyperbolic surfaces .
similar study has already been carried out by scott wolpert @xcite , lizhen ji @xcite and christopher judge @xcite .
we shall first make precise the notions of convergence in @xmath83 and the notion of convergence of a sequence of functions on a converging sequence of surfaces .
let @xmath5 be a sequence of surfaces in @xmath3 converging to a surface @xmath84 in the compactification @xmath85 .
another way of understanding this convergence is as follows : let @xmath86 and let @xmath87 ( @xmath88 is the margulis constant ; see thick / thin decomposition for details ) be a fixed constant .
let @xmath89 .
up to a conjugation of @xmath90 in psl@xmath53 , one may assume that @xmath91 is mapped to @xmath92 under the projection @xmath93 .
then up to extracting a subsequence we may suppose that @xmath90 converges to some funchsian group @xmath94 .
we say that the pair @xmath95 converges to @xmath96 where @xmath97 is the image of @xmath91 under the projection @xmath98 .
let @xmath84 be the hyperbolic surface of finite area whose connected components are the @xmath99 s for different choices of base point @xmath100 in different connected components of @xmath101 .
the surface @xmath102 does not depend , up to isometry , on the choice of the base point @xmath92 in a fixed connected component of @xmath101 ( i.e. if @xmath103 be a point in the same connected component of @xmath101 as @xmath92 then the corresponding limiting surfaces are isometric ) .
one can check that @xmath104 in @xmath85 .
+ fix an @xmath59 and choose a base point @xmath105 for each @xmath106 .
assume that the pair @xmath95 converges to @xmath96 where , for each @xmath107 , the point @xmath91 maps to @xmath92 under the projection @xmath108 .
for a @xmath109 function @xmath21 on @xmath110 denote by @xmath111 the lift of @xmath21 under the projection @xmath93 .
let @xmath112 be a sequence of functions in @xmath113 .
one says that @xmath112 _ converges to _ a continuous function @xmath114 if @xmath115 converges , uniformly over compact subsets of @xmath24 , to @xmath116 for each choice of base points @xmath105 and for each @xmath117 . with the above understanding of convergence of functions we shall prove the following theorem which has close resemblance with ( * ? ? ?
* theorem 1.2 ) and ( * ? ? ? * theorem 4.2 ) . however , our result does not follow from these
. we would like to mention that a similar limiting theorem might not be true ( see ( * ? ? ? *
p-71 ) if one considers @xmath118 instead of @xmath119 ( see theorem 2 ) . in the following , for a function @xmath31
, we shall denote the @xmath120 norm of @xmath21 by @xmath121 .
also , for @xmath122 and @xmath123 we denote the @xmath39-norm of the restriction of @xmath21 to @xmath124 by @xmath125 .
a function @xmath122 will be called _ normalized _ if @xmath126 .
an eigenpair @xmath127 will be called normalized if @xmath128 is normalized .
let @xmath129 in @xmath85 .
let @xmath130 be a normalized small cuspidal eigenpair of @xmath110 .
assume that @xmath131 converges to @xmath132 .
then one of the following holds : + @xmath133 there exist strictly positive constants @xmath134 such that @xmath135 .
then , up to extracting a subsequence , @xmath136 converges to a @xmath137-eigenfunction @xmath138 of @xmath84 .
+ @xmath139 for each @xmath140 the sequence @xmath141 then @xmath142 and @xmath143 .
moreover , there exist constants @xmath144 such that , up to extracting a subsequence , @xmath145 converges to a linear combination of eisenstein series and ( possibly ) a cuspidal @xmath132-eigenfunction of @xmath84 .
for @xmath146 , by eisenstein series we understand a linear combination of the following two : + * ( i ) * the classical ( meromorphic continuation ) eisenstein series @xmath147 corresponding to the cusps ( @xmath7 is the index for cusps ) on the surface , + * ( ii ) * the derivatives @xmath148 of @xmath149 at @xmath150 .
+ the first fourier coefficient of such functions in any cusp have the form @xmath151 .
each moderate growth @xmath82-eigenfunction is a linear combination of eisenstein series , in the above sense , and ( possibly ) a cuspidal eigenfunction .
theorem 2 will be applied to prove all three statements of theorem 1 .
the first one is a direct application ; in 4 we prove : the second statement of theorem 1 is also an easy application of theorem 2 and the buser construction @xcite : we explain it now since the proof is short .
we argue by contradiction and assume that there is a sequence @xmath5 in @xmath75 such that @xmath110 converges to @xmath152 and @xmath153 .
then @xmath84 has exactly @xmath154 components of which exactly @xmath37 are thrice punctured spheres .
observe that each component of @xmath84 contains an _ old cusp _
i.e. cusps of @xmath84 which are limits of cusps of @xmath110 ( see proof of theorem 2 ) .
the construction used in the proof of ( * ? ? ?
* theorem 8.1.3 ) implies that , for @xmath106 large , @xmath110 has at least @xmath37 eigenvalues that converge to zero as @xmath106 tends to infinity . let us suppose by contradiction that one of the corresponding eigenfunctions @xmath155 is cuspidal .
then by theorem 2 , @xmath155 converges uniformly over compacta to a function @xmath128 and @xmath128 is an eigenfunction for the eigenvalue @xmath46 .
so @xmath128 is constant in each component of @xmath84 . on those components of @xmath156 that contains an old cusp @xmath128 is necessarily zero because @xmath155 being cuspidal the average of @xmath155 over any horocycle is zero . on the the other component (
the one that does not contain an old cusp ) @xmath128 is zero because the mean of @xmath128 over @xmath84 is equal to the mean of @xmath155 over @xmath110 which is zero ( follows from theorem [ h ] ) .
therefore , @xmath128 is the zero function which is a contradiction by theorem 2 .
hence , for large @xmath106 each eigenfunction corresponding to any of the first @xmath37 eigenvalues of @xmath110 is necessarily residual . now if @xmath153 then each @xmath110 has at least @xmath157 small eigenvalues .
this is a contradiction to ( * ? ? ?
* theorem 2 ) .
therefore we have proved that @xmath158 contains a neighborhood of @xmath12 in @xmath6 . in the last section
we prove the last statement of theorem 1 .
we consider @xmath159 as a subset of @xmath160 and show the following [ neighbrhd ] there exists a neighborhood @xmath161 of @xmath159 in @xmath6 such that for each @xmath162 :
@xmath163 i.e. @xmath164 now we briefly sketch a proof of this proposition .
we argue by contradiction and consider a sequence @xmath5 in @xmath75 that converges to @xmath84 in @xmath165 such that @xmath166 .
then , for @xmath167 and for each @xmath106 , we choose a small cuspidal eigenpair @xmath168 of @xmath110 such that @xmath169 is an orthonormal family in @xmath170 @xmath171 is the @xmath7-th eigenvalue of @xmath110
. for @xmath167 let @xmath172 converges to @xmath173 as @xmath174 . by theorem 2
there are two possible types of behavior that the sequence @xmath175 can exhibit .
either , for each @xmath167 the sequence @xmath175 converges to a @xmath173-eigenfunction @xmath176 on @xmath84 , or for some @xmath7 the sequence @xmath168 satisfies condition @xmath139 in theorem 2 .
however , in our case we have the following lemma : _ for each @xmath7 , @xmath167 , up to extracting a subsequence , the sequence @xmath175 converges to a @xmath177-eigenfunction @xmath178 of @xmath84 .
the limit functions @xmath176 and @xmath179 are orthogonal for @xmath180 i.e. @xmath84 has at least @xmath181 small eigenvalues . moreover
none of the @xmath176 is residual . _
then we count the number of small eigenvalues of @xmath84 using @xcite to conclude that at least one of @xmath176 is nonzero on the component of @xmath84 of type @xmath182 .
this leads to a contradiction by huxley @xcite or ( * ? ? ?
* proposition 2 ) .
the author would like to express his sincere gratitude to his advisor jean - pierre otal for his patience , encouragement and insight .
the author was supported during this research by the indo - french cefipra - ifcpar grant .
in this section we shall recall some preliminary concepts that are important for our purpose .
metric convergence of a sequence @xmath57 to @xmath183 is one of the prime aspects of our study .
we start by explaining the thick / thin decomposition of a hyperbolic surface which is convenient to understand the metric convergence .
let @xmath184 recall that for any @xmath140 , _ the @xmath60-thin part of @xmath0 _ , @xmath185 , is the set of points of @xmath0 with injectivity radius @xmath186 .
the complement of @xmath185 , _ the @xmath60-thick part of @xmath0 _ , denoted by @xmath187 , is the set of points where the injectivity radius of @xmath0 is @xmath188 .
let @xmath189 be a simple closed geodesic on @xmath0 .
it can be viewed as the quotient of a geodesic in @xmath24 by a hyperbolic isometry @xmath190 fixing the geodesic .
we may conjugate @xmath190 such that the geodesic is the imaginary axis and the isometry is @xmath191 , @xmath192 being the length of the geodesic .
we define the _ hyperbolic cylinder @xmath193 with core geodesic @xmath189 _ as the quotient @xmath194 .
recall that the _ fermi coordinates _ on @xmath195 assign to each point @xmath196 the pair @xmath197 where @xmath198 is the signed distance of @xmath64 from @xmath189 and @xmath199 is the projection of @xmath64 on @xmath189 @xcite .
these coordinates give a diffeomorphism of this hyperbolic cylinder to @xmath200 . in terms of these
coordinates the hyperbolic metric is given by : @xmath201 for @xmath202 we define the _ collar _ @xmath203 around @xmath189 by @xmath204 then @xmath205 is diffeomorphic to an annulus whose each boundary component has length @xmath206 .
the _ collar theorem _ of linda keen @xcite says that @xmath207 embeds in @xmath0 ( more precisely , @xmath208 embeds in @xmath0 where @xmath209 and @xmath210 ) .
@xmath0 has @xmath37 ends called _
punctures_. _ cusps _ are particular neighborhood of the punctures .
denote by @xmath211 the parabolic isometry @xmath212 . for a choice of @xmath213 , a cusp @xmath214 is the half - infinite cylinder @xmath215 .
the boundary curve @xmath216 is a _ horocycle _ of length @xmath217 that we identify with @xmath218 .
one can parametrize @xmath214 using the _ horocycle coordinates _
@xcite with respect to its boundary horocycle @xmath216 .
the horocycle coordinates assigns to a point @xmath219 the pair @xmath220 where @xmath198 is the distance from @xmath64 to the horocycle and @xmath199 the projection of @xmath64 on the horocycle . in terms of
these coordinates the hyperbolic metric takes the form : @xmath221 recall that the cusp @xmath222 ( in fact @xmath223 ) around each puncture embeds in @xmath0 and that those cusps corresponding to distinct punctures have disjoint interiors ( ref .
* chapter 4 ) ) .
we call them _
standard cusp_s .
observe that the area and boundary length of a standard cusp is equal to @xmath47 .
for @xmath224 denote the disjoint union @xmath225 by @xmath226 where @xmath227 ranges over distinct cusps in @xmath0 .
by margulis lemma there exists a constant @xmath228 , the margulis constant , such that for all @xmath229 , @xmath230 is a disjoint union of embedded collars , one for each geodesic of length less than @xmath231 , and of embedded cusps , one for each puncture .
the collar around a geodesic of length @xmath232 is called a _
margulis tube_.
we describe how a collar around a geodesic of length @xmath233 converges as @xmath234 tends to zero to a pair of cusps .
first shift the origin of the fermi coordinates of @xmath235 to the right boundary of @xmath235 by making the change of variable @xmath236 . in the shifted fermi
coordinates the metric on @xmath235 is equal to @xmath237 for @xmath198 in a compact region we have the limiting @xmath238 now the hyperbolic metric on @xmath223 is equal to @xmath239 with respect to the boundary horocycle @xmath240 of @xmath223 . choose a base point @xmath241 on the right half of @xmath242 .
then by above , as @xmath243 , the pair @xmath244 converges , up to extracting a subsequence , to @xmath245 where @xmath246 .
since one can choose the base point on the left half of @xmath235 also , the metric limit of @xmath235 is a pair of @xmath223 .
our goal is to study the behavior of sequences of small cuspidal eigenpairs @xmath247 of @xmath248 when @xmath249 converges to @xmath183 and finally to prove theorem 1 . for this
we need to understand how the _ mass _ ( @xmath120 norm ) of a small eigenfunction is distributed over the surface , and in particular how it is distributed with respect to the _ thin / thick _ decomposition .
let @xmath9 .
recall that for any @xmath229 the @xmath60-thin part , @xmath185 , of @xmath0 consists of cusps and margulis tubes .
we separately study the mass distribution of a small cuspidal eigenfunction over these two different types of domains .
for @xmath250 consider the annulus @xmath251 contained in a cusp @xmath222 and bounded by two horocycles of length @xmath252 and @xmath253 .
we begin our study with the following lemma .
[ cusp ] for any @xmath254 there exists @xmath255 such that for any small cuspidal eigenpair @xmath32 of @xmath222 one has @xmath256 if @xmath257 for some @xmath258 then there exists a constant @xmath259 depending on @xmath260 and @xmath261 such that for any small eigenpair @xmath32 one has @xmath262 furthermore , @xmath263 as @xmath264
. * proof . *
we begin with the first part .
since @xmath21 is cuspidal inside @xmath222 it can be expressed as @xmath265 where @xmath266 and @xmath267 is the _ whittaker function _ ( see ( * ? ? ?
* proposition 1.5 ) ) .
the meaning of ( [ expres ] ) is that the right hand series converges to @xmath21 in @xmath268 and that the convergence is uniform over compact subsets .
recall also that for @xmath269 the whittaker functions is defined by @xmath270 where @xmath271 is the _ mcdonald s function _ and that for any @xmath60 ( see @xcite ) @xmath272 whenever the integral makes sense . from the expression
it is clear that the functions @xmath273 form an orthogonal family over @xmath274 ( independent of the choices of @xmath275 and @xmath260 ) .
hence ( 1 ) will follow from the following claim .
[ cusp3 ] let @xmath276 $ ] .
then for any @xmath277 there exists @xmath278 such that for all @xmath279 @xmath280 furthermore , @xmath281 as @xmath264 .
* from the expression of @xmath267 we have @xmath282 to prove the claim we may suppose that @xmath283 .
our next objective is to obtain bounds for the functions @xmath284 for @xmath276 $ ] .
we start from the above integral representation of @xmath285 .
we write @xmath286 where @xmath287 and @xmath288 now we treat @xmath289 and @xmath290 separately . :
we have @xmath291 @xmath292 since @xmath293 , we have : @xmath294 therefore @xmath295 to obtain a lower bound , we write @xmath296 @xmath297 since for all @xmath298 $ ] one has @xmath299 hence @xmath300 combining the above two inequalities @xmath301 : @xmath302 @xmath303 now for any @xmath304 , @xmath305 where @xmath306 .
@xmath308 @xmath309 thus combining the estimates for @xmath289 and @xmath290 we obtain @xmath310 let @xmath311 observe that for @xmath312 and @xmath313 @xmath314 so , for @xmath313 large enough @xmath315 going back to the expression of @xmath267 , for @xmath276 $ ] , we find : @xmath316 @xmath317 @xmath318 @xmath319 i.e. @xmath320 also , @xmath321
@xmath322 @xmath323 i.e. @xmath324 in the last inequality , we used the following estimate from ( * ? ? ?
* section 3.2 ) : @xmath325 with an absolute constant for the @xmath326-term for @xmath327 . comparing and
we get , for any @xmath269 @xmath328 where @xmath329 from the expression it is clear that @xmath330 is bounded independent of @xmath331 ( once @xmath260 is large enough ) and @xmath276 $ ] .
so we obtain the claim by choosing some @xmath332 sufficiently large ( once and for all ) such that the @xmath326-terms in the expression of @xmath333 are small enough .
it is also clear from the expression that when @xmath264 , @xmath281 .
this proves the claim [ cusp3 ] and hence the first part of lemma [ cusp ] .
now we prove the second part .
let @xmath257 for some @xmath258 and let @xmath32 be a residual eigenpair .
the fourier expansion of @xmath21 inside @xmath222 has the form @xmath334 where @xmath335 , @xmath336 ( see @xcite ) and @xmath337 .
since @xmath338 and @xmath339 are orthogonal and since the first part can be applied to @xmath339 , one needs only to prove the lemma for the term @xmath338 .
so we calculate : @xmath340 therefore , for @xmath277 , @xmath341 the lemma is satisfied by @xmath342 such that @xmath343 from the expression it is clear that @xmath342 depends only on two quantities : @xmath260 and @xmath344 . since @xmath345 , @xmath346 when @xmath347 .
this proves the second part .
now we study the distribution of the mass of a small eigenfunction over margulis tubes .
let @xmath189 be a simple closed geodesic of length @xmath348 . recall that @xmath349 denotes the collar around @xmath189 bounded by two equidistant curves of length @xmath275 .
any @xmath350 can be written as a fourier series in the @xmath199-coordinate : @xmath351 the functions @xmath352 and @xmath353 are defined on @xmath354 $ ] and are called the _ @xmath355-th fourier coefficients of @xmath21 _ ( in @xmath207 ) .
when @xmath21 is a @xmath34-eigenfunction , @xmath356 and @xmath357 are solutions of the differential equation @xmath358 we set @xmath359_0}={a_0}(r)$ ] and @xmath359_1}= f - { [ f]_0}$ ] . the following lemma concerns the distribution of masses of @xmath359_0}$ ] and @xmath359_1}$ ] inside @xmath207 .
[ t ] for any @xmath360 there exist constants @xmath361 , depending only on @xmath60 , such that for any small eigenpair @xmath32 of @xmath207 the following inequalities hold : @xmath362_1 } \rvert}_{\mathcal{c}^\epsilon } } < { t_1}(\epsilon ) { { \lvert{[f]_1}\rvert}_{{\mathcal{c}^1 } \setminus { { \mathcal{c}^\epsilon}}}}\ ] ] and @xmath363_0 } \rvert}_{{\mathcal{c}^{\epsilon_0 } } \setminus { \mathcal{c}^\epsilon } } } < { t_2}(\epsilon ) { { \lvert{[f]_0}\rvert}_{{\mathcal{c}^1 } \setminus { { \mathcal{c}^{\epsilon_0}}}}}.\ ] ] therefore , for any @xmath360 and any small eigenpair @xmath32 of @xmath207 one has @xmath364
if @xmath257 for some @xmath365 then there exists a constant @xmath366 , depending only on @xmath261 and @xmath60 , such that @xmath367_0 } \rvert}_{\mathcal{c}^\epsilon } } < { t_0}(\epsilon , \eta ) { { \lvert{[f]_0}\rvert}_{{\mathcal{c}^1 } \setminus { { \mathcal{c}^\epsilon}}}}.\ ] ] furthermore , @xmath368 as @xmath369 . before starting the proof of the above lemma we make a few observations about the solutions of .
the change of variable @xmath370 transforms into @xmath371 let @xmath372 ( resp .
@xmath373 ) be the solution of satisfying the conditions : @xmath374 and @xmath375 ( resp . @xmath376 and @xmath377 ) .
since is invariant under @xmath378 one has : @xmath379 and @xmath380 for all @xmath381 .
therefore there exists @xmath382 such that @xmath383 and @xmath384 on @xmath385 $ ] .
now we prove the following claim .
[ g ] let @xmath386 .
let @xmath387 \rightarrow \mathbb{r}$ ] be a @xmath20-function which satisfies the inequality : @xmath388 for some @xmath389 . if @xmath390 then @xmath391 is a monotone increasing function of @xmath198 in @xmath392 $ ] .
* observe that @xmath393 consider the function @xmath394 defined on @xmath395 $ ] by @xmath396 since @xmath339 is a @xmath20 function @xmath394 is continuous on @xmath395 $ ] .
observe that the claim follows if @xmath397 in @xmath392 $ ] . now for any @xmath398 $ ] @xmath399 therefore for @xmath400 , @xmath401 .
hence the claim.@xmath402 * proof of lemma [ t]*. we need to estimate , for @xmath403 , the quantities : @xmath404_1 } } \rvert}_{{\mathcal{c}^w } \setminus { \mathcal{c}^t}}^2 } = { l_\gamma } { \int_{-{l_w}}^{-{l_t } } } \bigg ( { \sum_{j= 1}^\infty}{\alpha_j^2 } + { \beta_j^2 } \bigg ) dr + { l_\gamma } { \int_{{l_t}}^{{l_w } } } \bigg ( { \sum_{j= 1}^\infty}{\alpha_j^2 } + { \beta_j^2 } \bigg ) dr\ ] ] and @xmath404_0 } } \rvert}_{{\mathcal{c}^w } \setminus { \mathcal{c}^t}}^2 } = { l_\gamma } { \int_{-{l_w}}^{-{l_t } } } { \alpha_0 ^ 2 } dr + { l_\gamma } { \int_{{l_t}}^{{l_w } } } { \alpha_0 ^ 2 } dr\ ] ] where @xmath405 , @xmath406 , @xmath407 and @xmath408 . since @xmath372 is odd and @xmath373 is even , for any symmetric subset @xmath409 $ ] , @xmath410 and @xmath411 are orthogonal in @xmath412 .
now @xmath413 and @xmath414 are linear combinations of @xmath410 and @xmath411 for @xmath415 and @xmath416 is a linear combination of @xmath417 and @xmath418 .
therefore , since @xmath372 and @xmath373 are orthogonal , it is enough to prove the lemma with @xmath410 and @xmath411 instead of @xmath359_1}$ ] and with @xmath417 and @xmath418 instead of @xmath359_0}$ ] .
we detail the computations for @xmath410 .
the computations for @xmath411 are similar .
let us choose @xmath60 such that @xmath419 .
the lemma reduces to find @xmath420 , depending on @xmath60 , and @xmath421 , depending on @xmath422 ( @xmath423 ) , such that @xmath424 and @xmath425 let @xmath426 and set @xmath427 and set for @xmath415 , @xmath428 .
notice that @xmath429 on @xmath430 . hence by @xmath431 satisfies the inequality : @xmath432 hence by claim [ g ] @xmath433 , for @xmath434 , is strictly increasing on @xmath435 .
the same is true for @xmath436 ( even when @xmath437 ) .
we begin with the proof of the second part of the lemma .
so we assume @xmath438 . for @xmath439
consider the integral : @xmath440 since @xmath441 is strictly increasing we have @xmath442 now choosing @xmath443 and @xmath444 the last inequality in gives @xmath445 next choosing @xmath446 and @xmath447 the first inequality in gives @xmath448 therefore @xmath449 where @xmath450 we see that @xmath451 depends only on @xmath452 and @xmath453 . now
therefore , for @xmath60 and @xmath455 fixed , and @xmath453 small @xmath456 and the constant @xmath457 is independent of @xmath453 as soon as @xmath453 is small compared to @xmath60 .
thus we can choose @xmath458 independent of @xmath453 satisfying .
this proves ( [ case3 ] ) for @xmath410 , @xmath415 , exactly the same computations for @xmath459 work with @xmath460 replaced by @xmath461 .
hence in this case our constant , @xmath462 depends only on @xmath60 .
this proves .
now we prove .
since @xmath463 \rightarrow { \mathbb{r}^{+}}$ ] is strictly increasing we have : @xmath464 combining the two inequalities we obtain @xmath465 where @xmath466 the constant @xmath467 is independent of @xmath453 as soon as @xmath453 is small compared to @xmath60 .
thus we can choose @xmath468 independent of @xmath453 satisfying .
this proves .
let @xmath0 be a finite area hyperbolic surface with @xmath37 punctures .
denote by @xmath469 the standard cusp around the @xmath7-th puncture . recall that @xmath469 s have disjoint interiors and that each of them is isometric to the half - infinite annulus @xmath222 ( see [ cuspd ] ) .
applying lemma [ cusp ] in each @xmath469 separately we obtain the following corollary which will be useful in our analysis .
[ crl ] for any @xmath470 there exists @xmath471 , depending only on @xmath60 , such that for any small cuspidal eigenpair @xmath32 of @xmath0 one has @xmath472 } } \setminus { s_c^{(0 , { \epsilon}]}}}}.\ ] ] if @xmath473 for some @xmath438 then for any @xmath470 there exists @xmath474 , depending only on @xmath60 and @xmath261 , such that for any @xmath34-eigenfunction @xmath21 of @xmath0 one has @xmath475 } } \setminus { s_c^{(0 , { \epsilon}]}}}}.\ ] ] furthermore , @xmath476 and @xmath477 tends to zero as @xmath369 . using this corollary and we deduce the following [ crl2 ] for any @xmath478 there exists a constant @xmath479 , depending only on @xmath60 ,
such that for any small cuspidal eigenfuction @xmath21 of @xmath0 one has @xmath480 now we give a new proof of the following theorem of d. hejhal @xcite .
[ h ] consider a sequence @xmath481 converging to @xmath482 .
let @xmath130 be a normalized small eigenpair of @xmath110 such that @xmath483 if @xmath484 then , up to extracting a subsequence , @xmath485 converges to a normalized @xmath132-eigenfunction @xmath138 of @xmath84 .
d. hejhal s proof uses _ convergence of green s functions _ of @xmath110 to that of @xmath84 .
our approach is more elementary and uses the above estimates on the mass distribution of eigenfunctions over thin part of surfaces .
* proof of theorem [ h ] . *
first we prove that , up to extracting a subsequence , @xmath485 converges to a @xmath132-eigenfunction @xmath138 of @xmath84 . by theorem 2 ( which will be proven in 3 )
it is enough to prove that there exist @xmath486 such that @xmath487 up to extracting a subsequence .
we argue by contradiction .
suppose that for any @xmath488 the sequence @xmath489 as @xmath174 .
let @xmath365 , such that @xmath490 for all @xmath491 .
by lemma [ t ] we have @xmath492 therefore from and we have @xmath493 hence if @xmath494 as @xmath174 then @xmath495 as @xmath174 .
this is a contradiction to the fact that each @xmath155 is normalized i.e. @xmath496 .
next we prove that @xmath497 by uniform convergence over compacta , in each cusp and in each pinching collar , the fourier coefficients of @xmath155 will converge to the corresponding fourier coefficients of @xmath138 .
therefore , by , and , @xmath155 s are uniformly integrable : for any @xmath498 there exist @xmath59 such that for all large values of @xmath106 @xmath499 hence @xmath497 this finishes the proof .
let @xmath5 be a sequence in @xmath3 which converges in @xmath85 to @xmath102 .
let @xmath500 be such that @xmath501 and @xmath502 .
recall that the convergence @xmath503 means that for any fixed positive constant @xmath504 ( @xmath88 is the margulis constant ) and a choice of base point @xmath505 , after conjugating @xmath90 so that the projection @xmath506 maps @xmath7 to @xmath507 , @xmath508 converges to a component @xmath509 of @xmath102 .
we begin by fixing some @xmath117 and @xmath510 . in the following
we assume that @xmath511 , @xmath507 , @xmath90 , @xmath512 and @xmath94 satisfy the previous statement . to simplify notations
we shall assume that only one closed geodesic @xmath513 gets pinched as @xmath514 .
in particular the limit surface @xmath84 ( which may be disconnected ) has two new cusps . denote the standard cusps of @xmath110 by @xmath515 @xmath516 and the limits of these in @xmath517 by @xmath518, ... ,@xmath519 and denote by @xmath520 the _ new cusps _ which arise due to the pinching of @xmath189 .
the cusps @xmath521 for @xmath522 will be called _
old cusps_. recall that we have a sequence of small cuspidal eigenpairs @xmath130 of @xmath501 such that the @xmath39-norm of @xmath155 is 1 and @xmath523 . in what follows @xmath524 will denote the area measure on @xmath110 for @xmath525 and @xmath526 will denote the area measure on @xmath24 . the lift of @xmath527 to @xmath24 under the projection @xmath528 , defined as above ,
will be denoted by @xmath111 . by green s formula one
has : @xmath529 let @xmath530 be compact .
one can cover @xmath330 by finitely many geodesic balls of radius @xmath531 .
if @xmath531 is sufficiently small then each of these balls maps injectively to @xmath110 since @xmath532 . therefore , since @xmath533 @xmath534 is bounded depending only on @xmath330 . from the mean value formula ( *
* corollary 1.3 ) there exists a constant @xmath535 such that for @xmath536 close to @xmath137 , @xmath537 for each @xmath538 where @xmath539 denotes the closed neighborhood of radius @xmath198 of @xmath330 in @xmath24 .
next we use the @xmath540-schauder estimates ( * ? ?
* theorem 4 , sect .
ii.5.5 ) to obtain a uniform bound for @xmath541 on @xmath542 .
this makes @xmath543 an equicontinuous family .
so , by arzela - ascoli theorem , up to extracting a subsequence , @xmath544 converges to a continuous function @xmath545 on @xmath330 . by a diagonalization argument
one may suppose that the sequence works for all compact subsets of @xmath24 .
therefore , up to extracting a subsequence , @xmath546 uniformly over compacta . by this uniform convergence
it is clear that @xmath545 is a _ weak solution _ of the laplace equation : @xmath547 .
therefore , by elliptic regularity , @xmath545 indeed a smooth and satisfies @xmath548 also by the convergence @xmath545 induces a function @xmath138 on @xmath84 that satisfies @xmath549 however , @xmath138 may not be an eigenfunction since it could be the zero function . in order to discuss this point
, we shall consider two cases according to whether the @xmath120-norm @xmath550 of the restriction of @xmath485 to @xmath551 is bounded below by a positive constant or not .
we may assume that @xmath553 .
then by the uniform convergence of @xmath546 over compacta , @xmath554 therefore @xmath138 is not the zero function and its @xmath39 norm is less than @xmath47 .
therefore it is a @xmath132-eigenfunction
. then we will prove the following statements : @xmath556 @xmath142 , @xmath557 @xmath558 and @xmath559 @xmath552 constants @xmath560 such that , up to extracting a subsequence , @xmath561 converges uniformly to a function which is a linear combination of eisenstein series and ( possibly ) a @xmath82-cuspidal eigenfunction .
suppose by contradiction that @xmath562 .
then all the cusps of @xmath84 are old cusps .
let @xmath563 denote the _ systole _ of @xmath84 .
then , for @xmath564 and for @xmath106 large enough , we have @xmath565 .
therefore , applying corollary [ crl ] , the assumption @xmath555 implies that @xmath495 .
this is a contradiction since each @xmath155 is normalized .
thus @xmath566 . follows from theorem [ h ] .
fix some @xmath60 , @xmath478 .
choose constants @xmath567 such that @xmath568 therefore the sequence @xmath569 must diverge to @xmath570 . using mean value formula (
* corollary 1.3 ) , @xmath571-schauder estimates @xcite and elliptic regularity , as earlier , and corollary [ crl2 ] we obtain that , up to extracting a subsequence , @xmath572 converges , uniformly over compacta , to a @xmath109 function @xmath545 that satisfies @xmath573 moreover , @xmath545 induces a function @xmath138 on @xmath84 that satisfies @xmath574 using the uniform convergence over compacta we have @xmath575 therefore @xmath138 is not the zero function . from lemma [ cusp ] and lemma [ t ]
we deduce that @xmath576 satisfies _ moderate growth _ condition @xcite in each cusp .
it is known that for any @xmath577 the space of _ moderate growth @xmath34-eigenfunctions _ of @xmath84 is spanned by eisenstein series and ( possibly ) @xmath34-cuspidal eigenfunctions ( see 3 in @xcite ) . in particular
, @xmath138 is a linear combination of eisenstein series and ( possibly ) a cuspidal eigenfunction .
this finishes the proof of @xmath578
we begin by proving which says that @xmath579 is open in @xmath75 .
empty set is open by convention .
therefore , we argue by contradiction and assume that there exists a @xmath580 such that every neighborhood of @xmath0 contains points from @xmath581 .
in other words , there exists a sequence @xmath582 that converges to @xmath0 and , for all @xmath106 , @xmath583 . for @xmath584 ,
let us denote by @xmath585 a normalized @xmath586-cuspidal eigenfunction such that @xmath587 is an orthonormal family in @xmath588 .
since we are considering small eigenvalues , up to extracting a subsequence , the sequence @xmath589 converges . for simplicity
we assume that , for @xmath584 , the sequence @xmath589 converges and denote by @xmath177 its limit .
observe that , for @xmath584 , @xmath590 now , since @xmath591 by theorem 2 , up to extracting a subsequence , @xmath175 converge to @xmath177-eigenfunction @xmath178 of @xmath0 .
moreover , by the result about uniform integrability inside cusps in corollary [ crl ] : @xmath592 .
hence @xmath593 is an orthonormal family in @xmath594 so that the @xmath77-th cuspidal eigenvalue @xmath595 of @xmath0 is below @xmath82 .
this is a contradiction because by our assumption @xmath596 as @xmath580.@xmath402 we argue by contradiction and assume that there is a sequence @xmath4 converging to @xmath597 such that @xmath598 . for @xmath167 and for each @xmath106
we choose small cuspidal eigenpairs @xmath168 of @xmath110 such that @xmath169 is an orthonormal family in @xmath170 @xmath171 is the @xmath7-th eigenvalue of @xmath110 . _ for each @xmath7 , @xmath167 , up to extracting a subsequence , the sequence @xmath175 converges to a @xmath177-eigenfunction @xmath178 of @xmath84 .
the limit functions @xmath176 and @xmath179 are orthogonal for @xmath180 i.e. @xmath84 has at least @xmath181 small eigenvalues .
moreover none of the @xmath176 is residual . _ by uniform convergence of @xmath585 to @xmath178 , we have @xmath600 . to prove the first two statements of the lemma it is enough to prove that , for @xmath167 , @xmath601 because this will imply that @xmath602 is not the zero function and that @xmath175 is uniformly integrable over the thick parts : for any @xmath603 there exists @xmath60 such that for all @xmath106 one has , @xmath604 to prove that , for each @xmath167 , @xmath601 we argue by contradiction and assume that for some @xmath167 , @xmath605 to simplify the notation , denote the sequence @xmath168 by @xmath130 and the limit @xmath606 by @xmath607 . by corollary [ crl ] the functions @xmath155 are uniformly integrable over the union of cusps of @xmath110 : for any @xmath213 there exists @xmath59 such that for all @xmath106 one has : @xmath608 since @xmath609 there is only one closed geodesic , @xmath610 , whose length @xmath611 tends to zero . for any @xmath612 and for @xmath106 large enough such that @xmath613 denote by @xmath614 the collar around @xmath513 bounded by two equidistant curves of length @xmath234
in view of the uniform integrability inside cusps , there exists @xmath615 such that for any @xmath229 there exists @xmath616 such that for @xmath617 we have : @xmath618 now we distinguish again two cases depending on whether @xmath619 or @xmath558 . if @xmath484 then we have a contradiction since @xmath620 by theorem [ h ] .
hence we may suppose that @xmath558 .
so , by theorem 2 either @xmath138 is the zero function or , for instance by ( * ? ? ?
* theorem 3.2 ) , @xmath138 is cuspidal . now recall that by lemma [ t ] we have uniform integrability of @xmath621_1 $ ] : for any @xmath217 there exists @xmath60 such that for all @xmath106 : @xmath622_1 } \rvert}_{\mathcal{c}^\epsilon_m } } < t.\ ] ] hence by , there exists @xmath623 such that for any @xmath624 the exists @xmath625 such that for @xmath626 one has : @xmath627_0 } \rvert}_{\mathcal{c}^\epsilon_m } } > \frac{\delta}{4}\ ] ] in particular , if @xmath628_0}|$ ] then , since area of @xmath629 is less than @xmath47 , we have for any @xmath624 and @xmath626 : @xmath630 now we prove that @xmath621_1 $ ] is uniformly small inside @xmath629 .
more precisely , [ pcurve ] let @xmath60 be such that @xmath631 .
there exists a constant @xmath632 , independent of @xmath60 , and @xmath633 such that for @xmath634 and @xmath635 : @xmath636_1}|(z ) < k\frac{\epsilon^\frac{1}{2}}{1 - \epsilon}.\ ] ] * proof . *
consider the expansion of @xmath485 inside @xmath637 with respect to the fermi coordinates ( see [ cylinder ] ) : @xmath638 here , for each @xmath639 , @xmath640 are the @xmath355-th fourier coefficients of @xmath155 inside @xmath641 and are defined for all @xmath642 .
recall that , for any @xmath643 $ ] we denote by @xmath644 the number @xmath645 . recall also that since @xmath485 is a @xmath131-eigenfunction , @xmath646 and @xmath647 satisfy with @xmath648 and @xmath649 .
therefore , for @xmath415 , one can express : @xmath650 where @xmath651 and @xmath652 are the two linearly independent solutions of with @xmath653 and @xmath654 .
recall that @xmath655 and @xmath656 satisfy : @xmath657 since , for @xmath658 , @xmath659 by claim [ g ] , for each @xmath415 , there exists strictly increasing functions @xmath660 \rightarrow { \mathbb{r}_{>0}}$ ] and @xmath661 \rightarrow { \mathbb{r}_{>0}}$ ] such that @xmath662 we denote by @xmath663 and @xmath664 the two new cusps of @xmath84 that appear as the limit of @xmath665 as @xmath174 .
now , let us assume:@xmath666 then , by the uniform convergence of @xmath485 to @xmath576 over compacta , we have a @xmath65 such that for @xmath667 and @xmath668 : @xmath669 by for any @xmath415 : @xmath670 provide : @xmath671 . recall that @xmath672 is odd and @xmath673 is even .
so by and : @xmath674 therefore , by and : @xmath675 therefore , for any @xmath676 : @xmath677 the last term of the inequality is @xmath678 since @xmath679 and @xmath680 are strictly increasing functions ( by ) .
similarly , @xmath681 hence latexmath:[\[\label{infsum } since , for @xmath415 , the function @xmath683 is strictly increasing , for any @xmath658 : @xmath684 now fix an @xmath60 such that @xmath685 . observe that @xmath686 .
so , for @xmath106 large such that @xmath611 is small compared to @xmath60 : @xmath687 where the constant @xmath688 can be chosen independently of @xmath60 as soon as @xmath106 is larger than some number @xmath633 .
therefore , by and , for @xmath634 and @xmath689 latexmath:[\[\label{uniformb } now fix @xmath691 ( see ) such that @xmath692 and choose @xmath693 .
then by lemma [ pcurve ] and : for each @xmath635 @xmath694_1}|(z).\ ] ] so the parallel curve @xmath695 with distance @xmath696 ( @xmath697 ) from @xmath513 such that @xmath698_0}|({r_0})$ ] has the property that @xmath155 has constant sign on it . in other words ,
the nodal set @xmath699 does not intersect this curve .
this is a contradiction to the next lemma .
[ curve ] let @xmath0 be a noncompact , finite area hyperbolic surface of type @xmath1 .
let @xmath189 be a simple closed geodesic that separates @xmath0 into two connected components @xmath700 and @xmath701 such that @xmath700 is topologically a sphere with @xmath702 punctures and @xmath701 is topologically a genus @xmath339 surface with one puncture .
let @xmath21 be a small cuspidal eigenfunction of @xmath0 .
then the zero set @xmath703 of @xmath21 intersects every curve homotopic to @xmath189 . *
* recall that @xmath703 is a locally finite graph @xcite .
let us assume that @xmath703 does not intersect some curve @xmath704 homotopic to @xmath189 .
we have @xmath705 and all the punctures of @xmath0 are contained in @xmath706 .
consider the components of @xmath707 . recall that since @xmath21 is cuspidal @xmath708 contains all the punctures of @xmath0 and
therefore these components give rise to a cell decomposition of a once punctured sphere .
the euler characteristic of the component @xmath709 containing @xmath704 as a puncture is either negative or zero ( since @xmath189 and each component of @xmath708 are essential ; see @xcite ) .
each component of @xmath707 other than @xmath709 ( at least one such exists since @xmath339 changes sign in @xmath700 ) is a _ nodal domain _ of @xmath21 and hence has negative euler characteristic @xcite .
also @xmath703 being a graph has non - positive euler characteristic .
let @xmath710 ( resp .
@xmath711 ) be the union of the nodal domains contained in @xmath706 which are different from @xmath709 and where @xmath21 is positive ( resp .
negative ) .
denote by @xmath712 the euler characteristic of the topological space @xmath713 .
since the euler characteristic of a once punctured sphere is @xmath47 , by the euler - poincar formula one has : @xmath714 this is a contradiction because the right hand side of the equality is strictly negative.@xmath402 now we prove that @xmath138 is not a residual eigenfunction .
it is clear from the uniform convergence that @xmath576 is cuspidal at the old cusps .
if @xmath576 is a residual eigenfunction then the only possibility is that @xmath138 is not cupsidal at one of the two new cusps .
let us assume that @xmath576 is residual in @xmath715 .
then , for sufficiently large @xmath217 , @xmath576 has constant sign in @xmath716 .
therefore , by the uniform convergence @xmath717 it follows that , for all @xmath106 large , @xmath485 has constant sign on a component of @xmath718 . since this component is homotopic to @xmath513 this leads to a contradiction to lemma [ curve ] as well .
this finishes the proof of lemma 2.@xmath402 let us denote the two components of @xmath84 by @xmath719 and @xmath720 such that @xmath721 and @xmath722 .
lemma 2 says that @xmath84 must have at least @xmath181 many small cuspidal eigenvalues . by (
* thorme 0.2 ) the number of non - zero small eigenvalues of @xmath719 is at most @xmath723 . in particular ,
the number of small cuspidal eigenvalues of @xmath719 is at most @xmath723 .
thus for some @xmath7 , @xmath167 , @xmath176 is not the zero function when restricted to @xmath724 i.e. @xmath176 is a cuspidal eigenfunction of @xmath724 .
this is a contradiction because @xmath724 does not have any small cuspidal eigenfunction by @xcite or @xcite .
@xmath402 the arguments in the proof of proposition are applicable to more general settings .
in particular , let @xmath5 be a sequence in @xmath75 that converges to @xmath725 let @xmath130 be a normalized small eigenpair of @xmath110 .
let @xmath726 as @xmath106 tends to infinity .
the arguments show the following : if @xmath727 then there exists a curve @xmath695 , homotopic to a geodesic of length tending to zero , on which , up to extracting a subsequence , @xmath155 has constant sign . | let @xmath0 be a noncompact , finite area hyperbolic surface of type @xmath1 .
let @xmath2 denote the laplace operator on @xmath0 .
as @xmath0 varies over the _ moduli space _ @xmath3 of finite area hyperbolic surfaces of type @xmath1 , we study , adapting methods of lizhen ji @xcite and scott wolpert @xcite , the behavior of _ small cuspidal eigenpairs _ of @xmath2 . in theorem 2 we describe limiting behavior of these eigenpairs on surfaces @xmath4 when @xmath5 converges to a point in @xmath6
. then we consider the @xmath7-th _ cuspidal eigenvalue _ , @xmath8 , of @xmath9 .
since _
non - cuspidal _ eigenfunctions ( _ residual eigenfunctions _ or _ generalized eigenfunctions _ ) may converge to cuspidal eigenfunctions , it is not known if @xmath8 is a continuous function . however , applying theorem 2 we prove that , for all @xmath10 , the sets @xmath11 are open and contain a neighborhood of @xmath12 in @xmath6 . moreover , using topological properties of nodal sets of _ small eigenfunctions _ from @xcite , we show that @xmath13 contains a neighborhood of @xmath14 in @xmath6 .
these results provide evidence in support of a conjecture of otal - rosas @xcite . |
in 1935 , wigner and huntington @xcite predicted that the molecules in solid molecular hydrogen at very high pressure would dissociate to form a metallic atomic solid .
solid hydrogen is expected to transform into a metallic liquid ground state at high pressure,@xcite may exhibit high - t@xmath3 superconductivity and superfluidity,@xcite and plays an important role in astrophysics.@xcite although it is not yet possible to reach the static pressure of more than 400 gpa normally thought necessary to dissociate the hydrogen molecules , recent experimental results obtained using diamond anvil cell techniques have been interpreted as indicating metallization and the transition of molecular hydrogen to an atomic liquid , known as phase iv , below 300 gpa.@xcite other experimental groups disagree , observing no evidence of the optical conductivity expected of a metal over the entire range of temperatures to the highest pressures explored , and no signature of an atomic liquid.@xcite the interpretation of experimental results is complicated by the fact that , unlike phase i ( @xmath4 gpa ) , which is a molecular solid of quantum rotors on a hexagonal close packed lattice , the structures of phase ii ( known as the broken - symmetry phase@xcite ) and phase iii ( @xmath5 gpa ) are unknown.@xcite in addition , it remains unclear whether or not phase iii is metallic.@xcite metallization is believed to occur either via the dissociation of hydrogen molecules and a structural transformation to an atomic metallic phase or via band overlap ( band - gap closure ) within the molecular phase .
there is more evidence in favor of the latter assumption .
quantum monte carlo ( qmc ) calculations of metallic hydrogen @xcite and our knowledge of the experimental equation of state@xcite suggest that the hydrogen molecules remain undissociated at pressures up to 620 gpa .
density functional theory ( dft ) calculations using the ab initio random structure searching approach@xcite predict dissociation near 500 gpa,@xcite whilst dft calculations using evolutionary techniques@xcite predict that the formation of stable non - molecular metallic hydrogen requires pressures much higher than 600 gpa.@xcite these values are so large that it seems likely that metallization takes place by band - gap closure in the molecular phase well before dissociation occurs .
experimental vibrational and scattering results are not yet sufficient to allow an unambiguous determination of the structure of phase iii , leaving theorists free to speculate.@xcite however , the range of structures proposed is so wide and the energy differences between them so small ( typically a few mev / proton ) that it is unclear which , if any , corresponds to the global free energy minimum .
finding the most stable arrangements of atoms in solids and molecules is and will remain a very difficult task , but significant progress has been made over the past few years .
recent advances based on simulated annealing , metadynamics , random sampling , evolutionary algorithms , basin hopping , minima hopping , and data mining , are discussed in ref . .
the ab initio random structure searching method@xcite has also been used to investigate the zero - temperature phase diagram of molecular solid hydrogen.@xcite searches excluding the effects of proton zero - point energy ( zpe ) found a sequence of molecular crystal structures within the pressure range over which phase iii is observed experimentally .
the c2/c structure was stable up to 270 gpa ; cmca-12 was stable from 270 gpa to 385 gpa ; and cmca was stable above 385 gpa .
the consequences of including proton zpe ( see section iv ) were investigated only after the structures had been found . ref . provided many new ideas about the phase diagram of solid molecular hydrogen , but the energies of the crystal structures considered were obtained using a simple semi - local approximation to the unknown exchange - correlation ( xc ) functional of density functional theory ( dft ) , which may or may not be accurate .
moreover , since dft substantially underestimates the fundamental band gap for essentially all materials , it also underestimates the pressure at which the band gap closes.@xcite finally , it is worth noting that crystal structures obtained using random searching do not always agree with structures obtained using other techniques or experiment.@xcite this is not surprising , as the probability that a random search will find the ground state decreases exponentially with increasing system size.@xcite dft is one of very few quantum mechanical methods capable of calculating the energies and enthalpies of all phases of interest over all relevant pressure regimes and has played an important role in studies of high - pressure hydrogen .
however , serious doubts about the accuracy of the results persist .
how do the inter - atomic interaction energy , the bond - stretch energy , the phase diagram , the metallization mechanism , and the phonon spectrum depend on the approximation used for the xc functional ?
how accurate should we expect dft calculations of measured quantities such as infra - red ( ir ) and raman spectra to be ?
answering these questions is necessary to assess the reliability of the many existing dft simulations of high - pressure solid hydrogen.@xcite on the minus side , because the enthalpy differences between rival structures are so small ( a few mev / proton ) , it is very difficult to be sure that a specific structure represents the global minimum .
the failures of dft to describe water and other systems containing hydrogen at low pressure and temperature have already been documented.@xcite perhaps surprisingly , the errors of dft are already troublesome even for a single h@xmath6o molecule : it was shown recently @xcite that some common xc functionals provide rather poor estimates of the o
h bond - stretch energy . because of the limited experimental data , theoretical and computational studies have an unusually important role to play in the study of solid hydrogen , but only if they are correct .
our purpose in this paper is to investigate the successes and failures of several widely used xc functionals in the study of high - pressure solid hydrogen .
we emphasize that the failures we identify do not signal a breakdown of dft itself ; they demonstrate only the limited accuracy of certain xc functionals .
we believe that identifying and investigating deficiencies such as these provides useful information to developers of new xc functionals and may bring us closer to one of the main goals of dft , which is to design functionals capable of providing accurate total energies for as wide a variety of systems as possible .
starting at the static level of theory , we calculate the zero - temperature phase diagram using three different functionals : the perdew - zunger ( pz ) parameterization of the local density approximation@xcite ( lda ) , the perdew - burke - ernzerhof ( pbe ) version of the generalized gradient approximation@xcite ( gga ) , and the semi - local becke - lee - yang - parr ( blyp ) functional.@xcite we focus on four specific structures with space groups p@xmath0/m , c2/c , cmca-12 , and cmca . according to previous static dft calculations using the pbe functional,@xcite these are stable in the pressure ranges @xmath7 105 gpa , 105270 gpa , 270385 gpa , and 385490 gpa , respectively . nuclear quantum effects are generally neglected in dft calculations because of computational cost . however , there is strong evidence to show that this neglect has a significant impact on the results for solid hydrogen , even at finite temperature.@xcite we investigate quantum nuclear effects at zero temperature by calculating the proton zpe as a function of pressure within the quasi - harmonic approximation .
our calculations of the phonon spectrum also enable us to investigate the effect of the choice of xc functional on the frequencies of the raman and ir active modes .
we demonstrate that adding the proton zpe to the calculated enthalpy produces `` dynamic '' phase diagrams in better agreement with experiment than the static phase diagrams obtained excluding zpe .
the paper is organized as follows .
section ii describes the details of our dft calculations .
section iii investigates the effect of the choice of xc functional on the calculated phase diagram and metallization pressure of solid molecular hydrogen at the static level of theory , and shows how the choice of functional affects the pressures at which the phase transitions i @xmath8 ii @xmath8 iii are predicted to take place . in section iv , we go beyond the static level by investigating the effects of zpe .
section vi concludes .
since the energy differences between high - pressure solid molecular structures are very small , the calculations must be done with the highest possible numerical precision .
our dft calculations were carried out within the pseudopotential and plane - wave approach using the quantum espresso suite of programs.@xcite all calculations used norm - conserving pseudopotentials and a basis set of plane waves with a cutoff of 200 ry .
geometry and cell optimizations employed a dense @xmath9 @xmath10-point mesh .
the bfgs quasi - newton algorithm was used for cell and geometry optimization , with convergence thresholds on the total energy and forces of 0.01 mry and 0.1 mry / bohr , respectively , to guarantee convergence of the total energy to better than 1 mev / proton and the pressure to better than 0.1 gpa / proton . to include the effects of zpe and investigate the phonon spectrum ,
vibrational frequencies were calculated using density - functional perturbation theory as implemented in quantum espresso.@xcite the zpe per proton at a specific cell volume @xmath11 was estimated within the harmonic approximation : @xmath12 , where @xmath13 .
here @xmath14 and @xmath15 are the numbers of vibrational modes in the unit cell and phonon wave vectors @xmath16 , respectively , and the summation over @xmath16 includes all @xmath10-points on a @xmath17 grid in the brillouin zone .
mcmahon @xcite demonstrated that a @xmath18-point grid of this density is sufficient to converge zpe differences between structures to within a few percent .
the enthalpy at pressure @xmath19 , including zpe effects , was obtained by minimizing @xmath20 with respect to @xmath11 at constant @xmath19 , where @xmath21 is the static ground - state energy at volume @xmath11 .
this section reports the ground - state phase diagram of solid molecular hydrogen in the pressure range corresponding to phases i , ii and iii ( @xmath1 gpa ) .
three different xc functionals were used to calculate the enthalpies of the cmca , c2/c , cmca-12 , and p@xmath0/m structures found by recent pbe calculations@xcite to be stable ( have the lowest enthalpy ) at different pressures within this range .
these structures are illustrated in figs .
[ p63m300 ] , [ c2c300 ] , [ cmca12300 ] , and [ cmca300 ] .
figure [ enthalpies ] shows the static lattice enthalpy as a function of pressure calculated using the lda , pbe and blyp functionals . according to our lda calculations , the p@xmath0/m , c2c , cmca-12 , and cmca phases are stable in the pressure ranges @xmath7 70 gpa , 70165 gpa , 165260 gpa , and @xmath22 260 gpa , respectively .
our pbe phase diagram is in agreement with previous work,@xcite indicating that the p@xmath0/m , c2/c , cmca-12 , and cmca phases are stable in the pressure ranges @xmath7 110 gpa , 110245 gpa , 245370 gpa , and @xmath22 370 gpa , respectively .
according to experiment , the phase transition from phase i to phase ii happens around 110 gpa and is characterized by a change in the low - frequency region of the raman and ir spectra.@xcite it is interesting that the structural transition from p@xmath0/m to c2/c is also observed at 110 gpa when the pbe functional is used in the static approximation
. the static pbe phase diagram does not provide any information about the transition from phase ii to iii , which is observed experimentally at around 150 gpa and is accompanied by a large low - temperature discontinuity in the raman scattering and a strong rise in the ir molecular vibrons.@xcite the static blyp phase diagram , by contrast , identifies the transition from phase ii to iii with the structural change from p@xmath0/m to c2/c . as fig .
[ enthalpies](c ) illustrates , blyp calculations predict that the p@xmath0/m , c2c , cmca-12 , and cmca phases are stable in the pressure ranges of @xmath7 160 gpa , 160370 gpa , 370430 gpa , and @xmath22 430 gpa , respectively .
( color online ) a layer of the hexagonal p@xmath0/m structure at 300 gpa .
the layers are stacked in an abab fashion .
the primitive unit cell contains 16 atoms , which form hydrogen molecules of two types : 75% of the molecules lie flat within the plane and 25% lie perpendicular to the plane .
, title="fig:",scaledwidth=45.0% ] + ( color online ) a layer of the monoclinic c2/c structure at 300 gpa .
the layers are arranged in an abcda fashion and the primitive unit cell contains 12 atoms .
, title="fig:",scaledwidth=45.0% ] + ( color online ) a layer of the monoclinic cmca-12 structure at 300 gpa .
the layers are arranged in an aba fashion and the primitive unit cell contains 12 atoms .
, title="fig:",scaledwidth=45.0% ] + ( color online ) a layer of the cmca structure at 300 gpa .
the layers are arranged in an aba fashion and the primitive unit cell contains 8 atoms . , title="fig:",scaledwidth=40.0% ] + ( color online ) .
enthalpy per proton as a function of pressure calculated using three different xc functionals : ( a ) lda , ( b ) pbe , ( c ) blyp .
the static lattice enthalpies of three different insulating crystal structures are reported relative to the enthalpy of the metallic cmca structure .
the cmca-12 structure , which is insulating at low pressure , transforms into the metallic cmca structure at 260 , 370 or 430 gpa according to the lda , pbe and blyp approximations , respectively.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
enthalpy per proton as a function of pressure calculated using three different xc functionals : ( a ) lda , ( b ) pbe , ( c ) blyp .
the static lattice enthalpies of three different insulating crystal structures are reported relative to the enthalpy of the metallic cmca structure . the cmca-12 structure , which is insulating at low pressure , transforms into the metallic cmca structure at 260 , 370 or 430 gpa according to the lda , pbe and blyp approximations , respectively.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
enthalpy per proton as a function of pressure calculated using three different xc functionals : ( a ) lda , ( b ) pbe , ( c ) blyp .
the static lattice enthalpies of three different insulating crystal structures are reported relative to the enthalpy of the metallic cmca structure .
the cmca-12 structure , which is insulating at low pressure , transforms into the metallic cmca structure at 260 , 370 or 430 gpa according to the lda , pbe and blyp approximations , respectively.,title="fig:",scaledwidth=45.0% ] figure [ xcgaps ] illustrates how the dft band gaps of the cmca-12 , c2/c and p@xmath0/m structures close as the pressure increases .
the results obtained using different xc functionals differ markedly .
the dft band gap does not correspond to the measured quasi - particle band gap and is usually much smaller , so these results are not expected to agree with experiment ; the true pressures at which the band gaps of these structures close are likely to be significantly higher than suggested by fig .
[ xcgaps ] .
exact - exchange ( exx ) dft calculations yield band gaps 12 ev higher than lda gaps.@xcite although this 12 ev difference was obtained by comparing lda and exx calculations , pickard@xcite investigated the effect of adding 12 ev to the pbe band gap of the c2/c phase . increasing the pbe gap by 1 ev or 2 ev and extrapolating to the pressure at which the increased gap vanished gave metallization pressures of 350 and 410 gpa , respectively .
table [ dftgap ] reports the pressures of band - gap closure obtained using the lda , pbe and blyp functionals with and without band - gap corrections of 1 ev , 1.5 ev and 2 ev .
use of the blyp functional yields the highest calculated metallization pressures for all three structures , even higher than the pressures obtained by applying a 2 ev correction to the lda gap .
( color online ) .
single - particle band gaps versus pressure calculated using ( a ) the lda , ( b ) the pbe , and ( c ) the blyp functionals .
the blyp band gap is larger than the lda and pbe gaps and consequently yields larger metallization pressures .
the p@xmath0/m and cmca-12 phases have the highest and lowest metallization pressures , respectively , regardless of the xc functional used.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
single - particle band gaps versus pressure calculated using ( a ) the lda , ( b ) the pbe , and ( c ) the blyp functionals . the blyp band gap is larger than the lda and pbe gaps and consequently yields larger metallization pressures .
the p@xmath0/m and cmca-12 phases have the highest and lowest metallization pressures , respectively , regardless of the xc functional used.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
single - particle band gaps versus pressure calculated using ( a ) the lda , ( b ) the pbe , and ( c ) the blyp functionals .
the blyp band gap is larger than the lda and pbe gaps and consequently yields larger metallization pressures .
the p@xmath0/m and cmca-12 phases have the highest and lowest metallization pressures , respectively , regardless of the xc functional used.,title="fig:",scaledwidth=45.0% ] the enthalpy - pressure relationships plotted in fig .
[ enthalpies ] show that the lda favors the metallic cmca phase relative to the insulating phases , and hence that the structural transition to the cmca phase occurs at a lower pressure within the lda .
furthermore , fig.[xcgaps ] shows that the band gaps of the insulating phases close at lower pressures when the lda functional is used .
the blyp functional , by contrast , makes insulating structures relatively more stable than the metallic cmca structure and predicts that the insulating cmca-12 phase does not transform into the metallic cmca phase until the pressure reaches 430 gpa .
all three xc functionals suggest that the p@xmath0/m phase has the largest band gap and , consequently , the highest metallization pressure of the three insulating structures considered .
the band gap of the cmca-12 phase closes at the lowest pressure .
.[dftgap ] calculated pressure in gpa at which the band gap of solid molecular hydrogen in the cmca-12 , c2/c and p@xmath0/m structures closes .
the metallization pressure is obtained by plotting the band gap as a function of pressure and extrapolating to the point at which the gap vanishes .
band gaps obtained using the lda , pbe and blyp xc functionals are shown .
the effects of adding band - gap corrections of 1.0 , 1.5 and 2.0 ev to the lda results are also indicated . [ cols="^,^,^,^,^,^,^",options="header " , ] all three xc functionals also indicate that the insulating wide - band - gap hexagonal structure p@xmath0/m is stable in the low - pressure regime . indeed , all three predict the same sequence of structural transitions ,
p@xmath0/m @xmath8 c2/c @xmath8 cmca-12 @xmath8 cmca , although at different pressures .
they also agree that the band gap of the cmca-12 structure ( which is insulating at low pressure ) closes before the transition to the cmca structure ( which is metallic at low pressure ) takes place
. however , the quantitative disagreements between results obtained using different xc functionals are so large that it is barely possible to compare the calculated dft phase diagram with experiment . assuming that the i @xmath8 ii and ii @xmath8 iii phase transitions observed experimentally at around 110 and 150 gpa are related to structural transformations , different functionals ascribe qualitatively different structural transitions to each .
the pbe phase diagram predicts a transition from p@xmath0/m to c2/c at around 105 gpa , which could be the i @xmath8 ii transition .
the lda and blyp phase diagrams show no structural transformations near 110 gpa but two different transformations , p@xmath0/m @xmath8 c2/c and c2/c @xmath8
cmca-12 , in the pressure range 150160 gpa , possibly related to the ii @xmath8 iii transition . none of the three xc functionals is able to explain both phase transitions .
the treatment of nuclear quantum effects ( zpe ) within dft is a challenging problem , but is especially important in hydrogen due to the small nuclear mass . indeed
, we show below that including the proton zpe has a large effect on the phase diagram .
dft is also used to study other properties that depend on lattice vibrations , including raman and ir spectra and the electron - phonon interaction.@xcite as in the static case , the question arises as to how these dynamical aspects of the behavior of high - pressure solid hydrogen are affected by the choice of xc functional .
answering this question is more important in the dynamic than the static case , since the experimental techniques most widely used to study high - pressure solid hydrogen are raman and ir spectroscopy , both of which are interpreted using phonon calculations.@xcite figure [ phonon ] shows the phonon densities of states of the cmca-12 and c2/c structures calculated using the lda , pbe and blyp xc functionals .
the principal effect of the choice of functional is on the size of the phonon band gap , which depends on pressure and decreases as the pressure increases . as in the case of the electronic band gap , the blyp and lda functionals give the largest and smallest phonon gap , respectively .
surprisingly , at low frequencies ( @xmath7 2200 @xmath2 ) , the phonon spectra obtained using the three different xc functionals are quite similar .
the xc functional has a much larger effect in the high - frequency regime ( @xmath22 3500 @xmath2 ) . as fig .
[ phonon ] shows , the lda , pbe and blyp functionals all predict two strong modes above the phonon gap , but at significantly different frequencies .
the uncertainties in the frequencies of these modes complicate the comparison of computational and experimental raman and ir spectra . to illustrate this difficulty , we have investigated the sensitivity of the calculated ir spectrum of c2/c , cmca-12 , and p@xmath0/m phases at 175 gpa to the choice of xc functional . as illustrated in fig .
[ irs ] , the computed ir spectrum of the c2/c phase exhibits a single strong vibron peak , although the peak intensity depends on the choice of xc functional .
the peak frequencies calculated using the lda , pbe , and blyp functionals are approximately 3800 , 4200 , and 4600 @xmath2 , respectively . in the case of the cmca-12 phase , all three functionals produce two peaks of similar intensity in the high frequency regime .
the lda peak frequencies are 3500 and 3750 @xmath2 ; the pbe frequencies are 3820 and 4050 @xmath2 ; and the blyp frequencies are 4250 and 4460 @xmath2 .
the simulated ir spectrum of the p@xmath0/m phase shows a single high - frequency peak , which appears at 4250 , 4470 , or 4860 @xmath2 for the lda , pbe , and blyp functionals , respectively .
experimental ir spectra@xcite taken at 158 gpa show a single strong vibron peak with a frequency of 4400 @xmath2 . as the pressure increases , the frequency of this peak decreases , reaching 4300 @xmath2 at 208 gpa , suggesting a frequency close to 4350 @xmath2 at p=175 gpa .
experiment also shows@xcite that the vibron intensity below 150 gpa is approximately three orders of magnitude lower than at higher pressures .
although no xc functional yields the experimental ir frequency accurately , all indicate that the c2/c phase has a very strong ir peak in roughly the right frequency range . as pointed out by pickard and needs,@xcite this observation is consistent with the suggestion that phase iii has the c2/c structure .
regardless of the choice of xc functional , the c2/c vibron peak is much more intense than the vibron peaks for the other structures studied .
( color online ) .
the phonon densities of states of the cmca-12 and c2/c structures calculated using the lda , pbe and blyp functionals at 175 and 250 gpa .
all three functionals yield almost the same low frequency ( @xmath7 2200 @xmath2 ) phonon spectrum .
the frequencies of the vibron modes in the high - frequency regime ( @xmath22 3500 @xmath2 ) depend strongly on the choice of functional.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
the phonon densities of states of the cmca-12 and c2/c structures calculated using the lda , pbe and blyp functionals at 175 and 250 gpa .
all three functionals yield almost the same low frequency ( @xmath7 2200 @xmath2 ) phonon spectrum .
the frequencies of the vibron modes in the high - frequency regime ( @xmath22 3500 @xmath2 ) depend strongly on the choice of functional.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
the phonon densities of states of the cmca-12 and c2/c structures calculated using the lda , pbe and blyp functionals at 175 and 250 gpa .
all three functionals yield almost the same low frequency ( @xmath7 2200 @xmath2 ) phonon spectrum
. the frequencies of the vibron modes in the high - frequency regime ( @xmath22 3500 @xmath2 ) depend strongly on the choice of functional.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
the phonon densities of states of the cmca-12 and c2/c structures calculated using the lda , pbe and blyp functionals at 175 and 250 gpa .
all three functionals yield almost the same low frequency ( @xmath7 2200 @xmath2 ) phonon spectrum
. the frequencies of the vibron modes in the high - frequency regime ( @xmath22 3500 @xmath2 ) depend strongly on the choice of functional.,title="fig:",scaledwidth=45.0% ] ( color online ) .
simulated ir spectra of the c2/c , cmca-12 and p@xmath0/m phases at 175 gpa .
the intensity scale used for the p@xmath0/m phase is double that used for the cmca-12 and p@xmath0/m phases.,title="fig:",scaledwidth=45.0% ] ( color online ) .
simulated ir spectra of the c2/c , cmca-12 and p@xmath0/m phases at 175 gpa .
the intensity scale used for the p@xmath0/m phase is double that used for the cmca-12 and p@xmath0/m phases.,title="fig:",scaledwidth=45.0% ] ( color online ) .
simulated ir spectra of the c2/c , cmca-12 and p@xmath0/m phases at 175 gpa .
the intensity scale used for the p@xmath0/m phase is double that used for the cmca-12 and p@xmath0/m phases.,title="fig:",scaledwidth=45.0% ] the effect of zpe on the phase diagram is normally studied by applying a simple , additive , pressure - dependent correction to the static ground - state total energy .
the zero - point motion is assumed to be harmonic and the phonon frequencies calculated as a function of volume using dft .
this was the procedure adopted here .
the limitations of the harmonic approximation can be overcome using dft - based path - integral molecular dynamics,@xcite but this is very costly and still relies on the accuracy of the xc functional , which is clearly in doubt .
ground - state qmc simulations treating both the electrons and protons as quantum mechanical particles@xcite do not rely on an approximate xc functional but are even more expensive .
furthermore , only a few qmc simulations of this type have ever been attempted and there are good reasons to doubt their accuracy in practice .
figure [ zpes ] summarizes the results of our proton zpe calculations using the lda , pbe and blyp functionals .
the proton zpes of all the structures investigated , cmca-12 , c2/c , p@xmath0/m , and cmca , are shown in the pressure range of phases i , ii , and iii .
it is clear that the zpe increases with increasing pressure regardless of the crystal structure or choice of xc functional , and that the blyp and lda functionals give the largest and smallest zpes in all cases . for the insulating phases , cmca-12 , c2/c , and p@xmath0/m ,
increasing the pressure also increases the differences between the proton zpes obtained using different xc functionals . for the metallic cmca phase ,
these differences are approximately independent of pressure .
the differences between the proton zpes of the candidate crystal structures can be comparable to the differences between their static total energies .
therefore , structure searching algorithms including an approximate zpe should be considered .
the crystal structures investigated here were all mechanically stable and fully relaxed in the supercells used for structure searching , but the existence of small imaginary phonon frequencies indicates that some may be weakly unstable with respect to structural distortions on longer length scales.@xcite the imaginary phonon frequencies are relatively small , however , and their number can be reduced by time - consuming small relaxations of the structures in larger supercells .
( color online ) .
the proton zero - point energies of the cmca-12 , c2/c , p@xmath0/m and cmca crystal structures within the pressure range of phases i , ii , and iii as calculated using the lda , pbe and blyp functionals .
the zpe per proton increases with pressure . at any given pressure
, the difference between zpes calculated using the blyp and lda functionals is more than 10 mev / proton ; this is large in comparison with the few mev / proton enthalpy differences between structures.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
the proton zero - point energies of the cmca-12 , c2/c , p@xmath0/m and cmca crystal structures within the pressure range of phases i , ii , and iii as calculated using the lda , pbe and blyp functionals .
the zpe per proton increases with pressure .
at any given pressure , the difference between zpes calculated using the blyp and lda functionals is more than 10 mev / proton ; this is large in comparison with the few mev / proton enthalpy differences between structures.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
the proton zero - point energies of the cmca-12 , c2/c , p@xmath0/m and cmca crystal structures within the pressure range of phases i , ii , and iii as calculated using the lda , pbe and blyp functionals .
the zpe per proton increases with pressure . at any given pressure ,
the difference between zpes calculated using the blyp and lda functionals is more than 10 mev / proton ; this is large in comparison with the few mev / proton enthalpy differences between structures.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
the proton zero - point energies of the cmca-12 , c2/c , p@xmath0/m and cmca crystal structures within the pressure range of phases i , ii , and iii as calculated using the lda , pbe and blyp functionals .
the zpe per proton increases with pressure . at any given pressure ,
the difference between zpes calculated using the blyp and lda functionals is more than 10 mev / proton ; this is large in comparison with the few mev / proton enthalpy differences between structures.,title="fig:",scaledwidth=45.0% ] the static phase diagrams presented in fig .
[ enthalpies ] assume that the protons are infinitely massive . by adding the proton zpe to the static results we obtain the dynamic phase diagrams reported in fig .
[ enthalpieszpe ] .
the lda dynamic phase diagram shows three phase transformations : p@xmath0/m to c2/c at around 50 gpa , c2/c to cmca-12 at around 170 gpa , and the cmca-12 to cmca metal - insulator transition at around 200 gpa .
the pbe and blyp dynamic phase diagrams place the p@xmath0/m to c2/c transition at 110130 gpa .
this suggests that the i - ii phase transition observed experimentally at 110 gpa corresponds to the p@xmath0/m to c2/c structural transformation , even though the ir data provides suggests that c2/c corresponds to phase iii .
the dynamic pbe and blyp phase diagrams agree that the cmca-12 structure is never stable in the pressure range associated experimentally with phase iii and that the c2/c structure transforms directly , or almost directly , into the metallic cmca structure .
the transition to cmca-12 proposed by pickard and needs@xcite has not been observed experimentally@xcite and a recent metadynamics calculation @xcite also suggests that , at finite temperature , phase iii may transform to cmca without passing through cmca-12 .
if the cmca-12 phase is ever stable , it is only within a very narrow window just below the pressure of the transition to the metallic cmca structure .
( color online ) .
enthalpy per proton included zpe ( dynamic enthalpy ) as a function of pressure calculated using three different xc functionals : ( a ) lda , ( b ) pbe , ( c ) blyp .
the dynamic lattice enthalpies of three different insulating crystal structures are reported relative to the enthalpy of the metallic cmca structure.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
enthalpy per proton included zpe ( dynamic enthalpy ) as a function of pressure calculated using three different xc functionals : ( a ) lda , ( b ) pbe , ( c ) blyp .
the dynamic lattice enthalpies of three different insulating crystal structures are reported relative to the enthalpy of the metallic cmca structure.,title="fig:",scaledwidth=45.0% ] + ( color online ) .
enthalpy per proton included zpe ( dynamic enthalpy ) as a function of pressure calculated using three different xc functionals : ( a ) lda , ( b ) pbe , ( c ) blyp .
the dynamic lattice enthalpies of three different insulating crystal structures are reported relative to the enthalpy of the metallic cmca structure.,title="fig:",scaledwidth=45.0% ] the inclusion of the effects of proton zpe substantially decreases the metallization pressure .
although the pbe and blyp dynamic phase diagrams can both be interpreted as in agreement with experimental results concerning metallization , they do not agree with the _ same _ experimental results .
eremets and troyan claim that @xcite hydrogen transforms into a metal at 260270 gpa : the conductance increases sharply and changes little on increasing the pressure to 300 gpa or cooling to 30 k ; and the sample reflects light well . as is shown in fig .
[ enthalpieszpe](b ) , the pbe dynamic phase diagram supports this claim : c2/c remains the most stable structure throughout phases ii and iii and transforms directly or almost directly into the metallic cmca phase at 240250 gpa . on the other hand , hemley and co - workers@xcite
see no evidence of a metallic state up to 360 gpa , but report electronic properties consistent with semi - metallic behaviour . the blyp dynamic phase diagram , fig .
[ enthalpieszpe](c ) , shows a metal - insulator transition at 350360 gpa , related to a phase transition from the c2/c structure to the cmca structure .
the dynamic phase diagrams also suggest that metallization of high - pressure solid hydrogen happens through a molecular structural transformation , not band - gap closure .
the pressure at which the band gap of the c2/c phase closes , as shown in table [ dftgap ] and also as calculated using the more accurate @xmath23 approximation,@xcite is considerably higher than the pressure at which the transition to the metallic cmca structure takes place .
since metallization is likely to take place via a structural transformation , the failure of dft to produce accurate band gaps is unlikely to be relevant .
last , but not least , we ask why dft is unable to reproduce the phase diagram of high - pressure solid hydrogen .
we believe that there are two main problems .
one is caused by the very small enthalpy differences of a few mev per proton between phases .
exhaustive tests for molecules , where accurate results can be obtained using quantum chemical methods , have shown that no current xc functional is consistently capable of achieving such accuracy.@xcite the second is the many - body self - interaction ( xc - si ) error present in the xc functionals used here .
a systematic investigation@xcite found the xc - si errors of the lda , gga , and blyp total energies of a single h@xmath6 molecule to be 1.264 , -0.126 , and 0.0846 ev , respectively .
these values are more than two orders of magnitude larger than the dft enthalpy differences between the crystal structures of high - pressure solid hydrogen .
the gga functional has a lower x - sie than the lda functional , but the c - sie , which is negative by definition , is larger than the x - sie in this case .
the xc - sie of the blyp functional is relatively small because the lyp functional by construction does not suffer from the c - sie . in previous work@xcite
we showed that the use of hybrid xc functionals , which generally suffer from smaller xc - si errors than pure density functionals,@xcite changes the phase diagram considerably .
this paper reported dft calculations using the lda , pbe and blyp xc functionals of various proposed structures of solid molecular hydrogen at low temperature and high pressure .
static results obtained using all three xc functionals are consistent with the following scenario : at the highest pressures considered here , the cmca-12 structure , which is insulating at low pressure but becomes semi - metallic with increasing pressure , transforms into the cmca structure , which is metallic even at low pressure .
since the cmca-12 and cmca structures have the same space group , differing only in the number of hydrogen molecules per unit cell , this seems reasonable .
more generally , as the pressure is increased , the band gaps of the insulating structures of solid hydrogen become smaller . as the band gaps decrease , we find that structures with wider gaps become progressively less stable relative to structures with smaller gaps , and a series of transitions to structures with smaller gaps takes place .
this process culminates in the transition to the metallic cmca structure .
the above scenario changes significantly when the effects of proton zpe are added to the static phase diagram .
our zpe calculations were based on the harmonic approximation , which is standard but may be problematic .
however , it is clear from our results that the effects of proton zpe are essential and must be taken into account , even if only approximately .
surprisingly , but perhaps fortuitously , the agreement between the pbe and blyp dynamic phase diagrams is much better than the agreement between the corresponding static phase diagrams . in particular , both dynamic phase diagrams suggest that the experimentally observed i - ii phase transition corresponds to a structural transformation from p@xmath0/m to c2/c , that c2/c is the most stable of the structures considered over the whole pressure range of phases ii and iii , and that metallization happens via a molecular structural transformation from c2/c to cmca without passing through cmca-12 .
the main difference between the pbe and blyp dynamic phase diagrams is the pressure of the c2/c to cmca transition , which the pbe functional places at 240250 gpa and the blyp functional places at 350360 gpa . as discussed above ,
both results can be interpreted as being in agreement with recent experimental results , but different and conflicting ones .
the low - frequency ( @xmath7 2200 @xmath2 ) parts of the phonon spectra of all structures considered are surprisingly independent of the choice of xc functional , but this is not true of the high - frequency vibronic parts . in general , it appears that the blyp functional produces stiffer phonons than the pbe functional and considerably stiffer phonons than the lda functional .
this work focused on four specific crystal structures recently proposed on the basis of pbe calculations.@xcite although the phase diagrams obtained using different xc functionals differed substantially in quantitative terms , especially at the static level , the four structures always appeared in the same order as the pressure was increased .
we think it likely , however , that if we had considered a wider range of structures , we would have seen qualitative as well as quantitative differences between the phase diagrams calculated using different xc functionals .
dft is probably not accurate enough to allow reliable comparisons of the stabilities of the insulating , semi - metallic , and metallic phases of solid molecular hydrogen . indeed ,
since simple dft calculations fail to describe van der waals interactions and severely underestimate band gaps , it would be unreasonably optimistic to expect such comparisons to be reliable .
we conclude that using dft results as the basis of our understanding of high - pressure solid hydrogen is likely to lead to serious errors . to obtain reliable results ,
it is almost certainly necessary to go beyond dft and take many - body effects properly into account.@xcite in any case , it is clear that more work needs to be done , both theoretically and experimentally , to understand solid molecular hydrogen at low temperature . | this paper investigates some of the successes and failures of density functional theory in the study of high - pressure solid hydrogen at low temperature .
we calculate the phase diagram , metallization pressure , phonon spectrum , and proton zero - point energy using three popular exchange - correlation functionals : the local density approximation ( lda ) , the perdew - burke - ernzerhof ( pbe ) generalized gradient approximation , and the semi - local becke - lee - yang - parr ( blyp ) functional .
we focus on the solid molecular p@xmath0/m , c2/c , cmca-12 , and cmca structures in the pressure range from @xmath1 gpa over which phases i , ii and iii are observed experimentally . at the static level of theory , in which proton zero - point energy is ignored ,
the lda , pbe and blyp functionals give very different structural transition and metallization pressures , with the blyp phase diagram in better agreement with experiment .
nevertheless , all three functionals provide qualitatively the same information about the band gaps of the four structures and the phase transitions between them . going beyond the static level , we find that the frequencies of the vibron modes observed above 3000 @xmath2 depend strongly on the choice of exchange - correlation functional , although the low - frequency part of the phonon spectrum is little affected .
the largest and smallest values of the proton zero - point energy , obtained using the blyp and lda functionals , respectively , differ by more than 10 mev / proton . including the proton zero - point energy calculated from the phonon spectrum within the harmonic approximation
improves the agreement of the blyp and pbe phase diagrams with experiment . taken as a whole , our results demonstrate the inadequacy of mean - field - like density functional calculations of solid molecular hydrogen in phases i , ii and iii and emphasize the need for more sophisticated methods . |
a packing is defined as a set of nonoverlapping objects arranged in a space of given dimension @xmath3 .
one packing problem in @xmath3-dimensional euclidean space @xmath12 that has not been generally addressed is that of finding the maximally dense ( optimal ) packing(s ) of @xmath0 nonoverlapping @xmath3-dimensional spheres of unit diameter near ( local to ) an additional fixed central sphere such that the greatest radius @xmath13 from any of the surrounding spheres centers to the center of the fixed sphere is minimized .
this problem is called the densest local packing ( dlp ) problem @xcite , and the minimized greatest radius associated with number of spheres @xmath0 is denoted by @xmath1 . in various limits , the dlp problem encompasses both the kissing number and ( infinite ) sphere packing problems @xcite .
the former is a special case of the dlp problem in that the kissing number @xmath14 , or number of identical @xmath3-dimensional nonoverlapping spheres that can simultaneously be in contact with ( kiss ) a central sphere , is equal to the greatest @xmath0 for which @xmath15 , and the latter is equivalent to the dlp problem in the limit that @xmath16 .
the dlp problem for @xmath17 spheres in @xmath7 dates back to a debate between newton and gregory in 1694 .
newton believed that only @xmath18 identical spheres could simultaneously contact a central same - size sphere , while gregory believed the correct number to be @xmath17 . the first rigorous proof that newton was right came in 1953 @xcite , followed by a more concise proof in 1956 @xcite .
however , the question remains : how close can @xmath17 identical spheres come to a central same - size sphere - how good was gregory s guess ? in another paper @xcite , we showed that for any @xmath3 , the dlp optimal packings in @xmath12 with @xmath19 , @xmath20 the golden ratio , include packings where all @xmath0 sphere centers lie on a spherical surface of radius @xmath1 .
the smallest radius spherical surface onto which the centers of @xmath17 spheres of unit diameter can be placed is strongly conjectured to be @xmath21 , with the centers arranged in a structure first documented in @xcite .
it appears that though gregory was incorrect in conjecturing @xmath22 to be @xmath17 , his guess was nt particularly far - off . for each @xmath0 in the dlp problem in @xmath12
, there is a single optimal @xmath1 , though generally for a given @xmath0 there can be multiple distinct packings that achieve this radius . in the predecessor to this paper @xcite , hereafter referred to as paper i , we studied the optimal packings and corresponding @xmath1 for the dlp problem in @xmath5 for @xmath23 to @xmath24 and for @xmath0 corresponding to full shells of the triangular lattice from @xmath25 to @xmath6 .
we also discussed the general concepts and applications associated with dlp optimal packings for arbitrary @xmath0 and @xmath3 . in paper
i , we reported that a majority of the dlp optimal packings in @xmath26 contain rattlers , or spheres ( disks ) that can be displaced in at least one direction without increasing @xmath1 or displacing any other sphere ( disk ) in the packing .
further , we found that many optimal packings contain cavities at their centers in which the central disk , were it not fixed , could move freely ( as a rattler , or object in a packing that can be displaced without displacing any other objects or the packing boundary ) .
the optimal packings in @xmath26 also exhibit a wide range of rotational symmetries , particularly for smaller @xmath0 , and packings of @xmath0 spheres from certain classes such as the curved @xcite and wedge hexagonal packings were found to be optimal at various @xmath0 .
we additionally observed that as @xmath0 grows large , disks in the _ bulk _ ( as opposed to on the surface ) of dlp optimal packings are largely arranged as subsets of the densest infinite packing in @xmath5 , i.e. , with centers on the sites of the triangular lattice , whereas disks farthest from the central sphere ( centers at distance @xmath1 ) tend to form circular rings .
the dlp problem is related to problems of finding arrangements of @xmath8 points @xmath27 that minimize potential energy . defining the dlp potential energy as the negative of the density of @xmath8 spheres ( including the central sphere ) contained completely within the _ encompassing sphere _ , a sphere of radius @xmath28 centered on the central sphere , the dlp problem becomes a minimal - energy problem with the pair potential between points ( sphere centers ) exhibiting features of long - range attraction and infinite short - range repulsion .
comparisons between dlp optimal packings and minimal - energy configurations for @xmath8 points with pair potentials exhibiting similar features of long - range attraction and short - range repulsion indicate that , though minimal energies for certain values of @xmath0 are similar , optimal configurations of points ( sphere centers ) are in general completely different .
this finding could have implications for nucleation theory , as is discussed in more detail in secs .
[ minenergysection ] and [ conclusions ] .
the dlp problem is relevant to the realizability of functions that are candidates to be the pair correlation function of a packing of identical spheres . for a statistically homogeneous and isotropic packing , the pair correlation function
is denoted by @xmath29 ; it is proportional to the probability density of finding a separation @xmath30 between any two sphere centers and normalized such that it takes the value of unity when no spatial correlations between centers are present .
specifically , no function can be the pair correlation function of a point process ( where a packing of spheres of unit diameter is a point processes with a minimum pair separation distance of unity ) unless it meets certain necessary , but generally not sufficient , conditions known as realizability conditions @xcite .
two of these conditions that appear to be particularly strong for the realizability of sphere packings @xcite are the nonnegativity of @xmath29 and its corresponding structure factor @xmath31 , where @xmath32 with number density @xmath33 and @xmath34 the @xmath3-dimensional fourier transform of the total correlation function @xmath35 .
knowledge of the maximal number of sphere centers that may fit within radius @xmath13 from an additional fixed sphere s center , where that maximal number , denoted by @xmath36 , is equal to the greatest @xmath0 in the dlp problem for which @xmath37 , may be employed to construct a third realizability condition , called the @xmath38 condition @xcite , on @xmath29 .
this condition is written @xmath39 where @xmath40 is defined for a statistically homogeneous packing as the expected number of sphere centers within distance @xmath13 from an arbitrary sphere center .
the function @xmath40 can be related to the pair correlation function @xmath29 , where for a packing of spheres with a pair correlation function @xmath41 that is direction - dependent , @xmath29 is the directional average of @xmath42 , by @xmath43 in eq .
( [ zr ] ) , @xmath33 is the constant number density of sphere centers and @xmath44 is the surface area of a sphere of radius @xmath30 in @xmath12 , @xmath45 as was discussed in previous papers @xcite , the @xmath38 realizability condition has been shown to encode information not included in the nonnegativity conditions on pair correlation functions and their corresponding structure factors alone .
the maximal infinite - volume packing fraction @xmath46 of identical nonoverlapping spheres in @xmath12 can be bounded from above by employing knowledge of the optimal @xmath1 in the dlp problem , where a _ packing fraction _ is the fraction of a given space covered by nonoverlapping objects . as was discussed in paper
i , an upper bound is constructed by measuring the packing fraction @xmath47 as the fraction of the volume of the encompassing sphere covered by the @xmath8 spheres of unit diameter , @xmath48 in the first equality in ( [ packingfraction ] ) , @xmath13 is the greatest radius from any of the @xmath0 surrounding spheres centers to the center of the central sphere , and in the second equality , the number density @xmath33 is the fraction of a sphere of radius @xmath47 covered by the @xmath8 spheres of unit diameter and @xmath49 is the standard gamma function . for small numbers of spheres ( @xmath50 ) in low dimensions ( @xmath51 ) ,
an algorithm combining a nonlinear programming method with a stochastic search of configuration space can be employed on a personal computer to find putative solutions to the dlp problem .
the accuracy of the solutions found by such an algorithm is bounded only by a machine s precision , and in general higher accuracy only requires more computation time . using such an algorithm , the details of which
are described in paper i , we find and present putatively - optimal dlp packings and their corresponding @xmath1 in @xmath7 for @xmath52 to @xmath53 ( accuracy of at least @xmath54 sphere diameters ) , and for selected values of @xmath0 from @xmath55 to @xmath2 ( accuracy of at least @xmath56 sphere diameters ) .
images and coordinates for many of the optimal packings that we have found are located on our website @xcite .
of the @xmath57 studied , some are randomly chosen and some correspond to the numbers of contacting spheres of unit diameter near and equal to the number in subsets of face - centered - cubic ( fcc ) and hexagonal - close - packed ( hcp ) packings with a given number of full coordination shells . in @xmath7 as in @xmath5 , rigorous and repeated testing of the algorithm indicates that it is robust in finding dlp optimal packings .
the identification of rotation and reflection symmetry , spatially precise to @xmath54 or better sphere diameters , in numerous dlp presumed - optimal packings in @xmath7 for @xmath58 supports this conclusion , as does our finding that the minimal @xmath13 found for @xmath59 to @xmath60 and for @xmath61 are smaller than the ( previously ) best - known minimal radii @xcite for the less - restrictive problem of finding the minimal radius ( larger ) sphere into which @xmath8 spheres of unit diameter may be packed .
however , the algorithm is dependent upon initial conditions , and as the number @xmath0 of spheres increases , an increasing number of trials have been necessary before we have found a packing and corresponding @xmath1 that with a high degree of confidence we consider optimal . for this reason , due to computing time constraints
, though we strongly conjecture that the vast majority of our packings are optimal for @xmath62 , as many as @xmath63 or more of packings for @xmath64 may not be strictly optimal ( though we hereafter refer to them as optimal ) .
for these packings , we will present evidence indicating that the smallest radius @xmath13 found for each @xmath65 is very near to the optimal @xmath1 , if not equal to it . over the range of @xmath0 studied , dlp
optimal packings in @xmath7 differ substantially in terms of symmetry , contact networks , and spatial positioning from the barlow packings @xcite , which we recall are the maximally dense infinite packings of identical nonoverlapping spheres in @xmath7 . in general
, we find that optimal packings are dominated by the dense arrangement of spheres on their surface , i.e. , those spheres with centers at precisely distance @xmath1 , where the arrangement of the spheres in the bulk ( interior ) is of secondary importance .
similarly in @xmath5 , dlp optimal packings differ substantially from packings of contacting disks arranged on the points of a triangular lattice , though in both dimensions at sufficiently large @xmath0 , the spheres in the bulk of each optimal packing begin to always resemble , respectively , a subset of a corresponding maximally dense infinite packing in @xmath5 or @xmath7 . in @xmath5 , the approximate @xmath0 and corresponding @xmath1 at which this change occurs
may be identified visually by perusing the dlp optimal packings at various @xmath0 . in @xmath7
however , it can be difficult to visually compare distinct packings .
consequently , we here introduce the concept of a _ similarity metric _ , defined as a metric designed to quantify the degree of similarity between one set of points and a reference set .
as will be discussed in sec .
[ scandbarlow ] , using such a metric allows quantitative comparisons of the relative degree of similarity between the radial spatial positions of spheres configured as dlp optimal packings and of spheres configured as subsets of a maximally dense infinite packing .
our key results and findings are summarized in the following list : * a novel realizability condition on candidate pair correlation functions @xmath29 for sphere packings in @xmath7 is constructed from knowledge of the @xmath1 ( sec .
[ realizabilityandbounds ] ) . *
dlp optimal packings for almost every @xmath0 exhibit the phenomenon of surface - maximization , i.e. , the number of spheres _ on the surface _
( with centers at precisely radius @xmath1 ) is either the maximal or nearly the maximal number that can be placed without overlap with centers on a spherical surface of radius @xmath66 ( sec . [ scandbarlow1 ] ) . for two @xmath0
( @xmath10 and @xmath11 ) , the dlp optimal packings are termed `` maracas '' packings , as packings at these @xmath0 have the maximal number of spheres on the surface , while all spheres _ not _ on the surface are rattlers ( sec .
[ maracassec ] ) . * dlp optimal packings for the @xmath0 studied are almost always _ most _ similar to subsets of fcc barlow packings , and for sufficiently large @xmath0 , the bulk ( as opposed to the surface ) of optimal packings appear to always be structured similarly to a subset of a barlow packing ( sec .
[ scandbarlow2 ] ) . *
the set of @xmath0 for which there are dlp optimal packings that include rattlers is unbounded , and the number of rattlers in a packing appears to grow _ at most _ as quickly as the surface area of the packing ( sec .
[ scandbarlow2 ] ) . *
the @xmath9 lennard - jones ( lj ) energy of dlp optimal packings for some @xmath0 are within a few percent of the minimal lj energy for @xmath8 points , but in general the spatial configurations of sphere centers in dlp optimal packings are completely different from the minimal - energy configurations of @xmath8 points ( sec .
[ minenergysection ] ) .
* many dlp optimal packings for @xmath58 exhibit elements of perfect rotational and/or reflection symmetry ( secs .
[ densesymm ] and [ othersymm ] ) . *
imperfect icosahedral symmetry is present in many dlp optimal packings , but perfect icosahedral symmetry is _ never _ present ( sec .
[ icospackings ] ) .
a function that is a candidate to be the pair correlation function @xmath29 of a point process , where a packing of spheres of unit diameter is a point process in which the minimum pair distance is unity , must be nonnegative for all @xmath30 and correspondingly have a structure factor @xmath31 that is nonnegative for all @xmath67 . for a packing of spheres of unit diameter , such a candidate function
must additionally be identically zero on the interval @xmath68 to reflect the nonoverlap condition between spheres .
however , the two nonnegativity conditions and the nonoverlap condition are only necessary , and generally not sufficient , conditions for a function to be the pair correlation function of a sphere packing .
the @xmath38 realizability condition further constrains candidate pair correlation functions , eliminating a range of functions that obey the two aforementioned nonnegativity conditions and are identically zero on @xmath68 , and yet violate the @xmath38 condition @xcite .
the function @xmath36 in @xmath7 can be compared to @xmath69 , where @xmath69 is the largest number of spheres of unit diameter whose centers can be placed within distance @xmath13 from a central sphere center when packings are constrained to the space of barlow packings .
both @xmath36 and @xmath69 increase roughly linearly with @xmath70 , as the volume of a sphere of radius @xmath13 is proportional to @xmath70 .
the function @xmath36 is clearly always greater than or equal to @xmath69 , but we find strict inequality , i.e. , @xmath71 , for @xmath72 , as can be seen in figure [ zmaxvszbar ] , a plot of @xmath36 and @xmath69 vs. @xmath0 for @xmath23 to @xmath53 .
we also compare each @xmath1 that we have found to each minimal radius @xmath73 , where @xmath73 is defined as the smallest @xmath13 for @xmath0 spheres surrounding a fixed central sphere when packings are constrained to the space of barlow packings .
figure [ rminvsrbar ] plots @xmath1 and @xmath73 vs. @xmath0 for @xmath23 to @xmath53 and for the values of @xmath0 for which we used the algorithm to seek optimal packings from @xmath55 to @xmath74 .
it is clear from the figure that @xmath1 rises roughly with @xmath73 , and that @xmath73 is an upper bound for @xmath1 .
it is also true that @xmath73 can not grow too much larger than @xmath1 ; appendix [ lowerbound ] discusses methods of bounding @xmath1 in @xmath7 from below using the barlow packings .
we do not here explicitly construct a lower bound for @xmath1 that , along with the upper bound in fig .
[ rminvsrbar ] , could be used to specify a range of feasible values for each optimal @xmath1 .
however , we can comment on the accuracy of the presumed - optimal @xmath1 found by the algorithm for @xmath64 by comparing the differences between @xmath73 and the putative @xmath1 for @xmath57 and over the range @xmath75 for which we have confidence in the optimality of the packings .
this comparison of @xmath76 over the two ranges yields values for both the smallest and largest differences that are very near to one another , as can be visually verified by close inspection of fig .
[ rminvsrbar ] . in @xmath7 , knowledge of the barlow packings allows us to bound @xmath1 and consequently @xmath36 .
however , in @xmath12 with @xmath77 where the maximal infinite - volume packing fraction @xmath46 is only known with analytical rigor for @xmath78 and @xmath79 @xcite , optimal @xmath1 can be employed to provide a rigorous upper bound on @xmath46 . as was shown in paper i
, @xmath46 in @xmath12 can be bounded from above with knowledge of any @xmath80 , where @xmath81 is the set of all positive integers for which @xmath82 is the greatest integer @xmath0 such that @xmath83 .
for example , in @xmath84 , @xmath85 , where @xmath86 is the greatest @xmath0 for which @xmath87 @xcite . the rigorous upper bound on @xmath46 , proved in paper i , is @xmath88 where the maximal local packing fraction @xmath89 of a packing of @xmath0 nonoverlapping spheres of unit diameter around a same - size fixed central sphere is defined as the ratio of the volumes of the @xmath8 spheres to the volume of a sphere of radius @xmath1 , or @xmath90 though @xmath46 is known rigorously in @xmath7 , it is informative to calculate the bounds derived from the putative @xmath80 over the range of @xmath0 tested . as @xmath0 increases , the bound becomes sharper ( becoming exact as @xmath91 ) , and calculations can give a sense of how quickly the bound approaches the known value of @xmath46 .
figure [ 3dupperboundplot ] plots the upper bound calculated using relations ( [ upperbound ] ) and ( [ maxlocaldensity ] ) for the putative @xmath80 found in @xmath7 versus the proved maximal infinite - volume packing fraction , @xmath92 .
the upper bound in @xmath7 ( inequality ( [ upperbound ] ) , fig .
[ 3dupperboundplot ] ) with maximal local packing fraction @xmath89 calculated from eq .
( [ maxlocaldensity ] ) converges to @xmath46 more slowly , as a function of @xmath1 , than does the upper bound in @xmath26 @xcite .
this is intuitive .
as the spheres in the bulk of each dlp optimal packing in both @xmath5 and @xmath7 at sufficiently large @xmath0 appear to be packed as the infinite densest packing in their respective space @xmath12 ( this observation is discussed in more detail in sec .
[ scandbarlow2 ] ) , it would be logical to predict that @xmath89 in the bound ( [ upperbound ] ) converges to @xmath46 in dimension @xmath3 roughly as the ratio of sphere surface area to volume , @xmath93 . up to a small corrective factor
that is also dependent on dimension , this appears to be the case , as the following simple model demonstrates .
the maximal local packing fraction @xmath89 ( [ maxlocaldensity ] ) includes in its numerator the total volume of spheres of unit diameter with centers at distance @xmath1 , even though the ( larger ) sphere of radius @xmath1 whose volume is employed in the denominator does not enclose roughly half of these ( smaller ) spheres volume . approximating the spherical surface of radius @xmath1 as a plane , which is a good approximation
when the ratio of radii @xmath94 is small , a simple model for the volume of the spheres not enclosed by the sphere of radius @xmath1 can be built . in this model under the aforementioned approximation , it follows that the @xmath3-dimensional spheres of unit diameter in a dlp optimal packing in @xmath4 with centers on the surface at radius @xmath1 should be packed roughly as densely as possible , i.e. , with centers arranged as the centers of @xmath95-dimensional spheres in the maximally dense packing in @xmath96 .
making these two approximations , which become exact in the limit @xmath97 , the fraction of the volume of the spheres of unit diameter _ not _ enclosed by the sphere of radius @xmath1 to the volume of the sphere of radius @xmath1 is @xmath98 times as large in @xmath7 as in @xmath5 . at any @xmath1 for large enough @xmath0 then
, about @xmath99 more volume is `` added back in '' to the numerator of @xmath89 in @xmath7 than in @xmath5 , meaning that the convergence fraction , @xmath100 at given @xmath1 in @xmath7 should be comparable to the same fraction @xmath101 at @xmath102 in @xmath5 .
this is indeed the case , as is shown in table [ convergecompare ] for two ranges of @xmath0 .
.comparison of @xmath101 ( [ phic ] ) in @xmath5 and @xmath84 .
the values of @xmath101 are compared for @xmath1 in @xmath26 near to @xmath103 in @xmath7 . [ cols="^,^,^,^,^",options="header " , ] [ icostable ] the numbers of spheres in the outer shells of the two perfectly icosahedrally symmetric packings are near to the maximal number @xmath104 that can be placed on spherical surfaces of radii @xmath105 and @xmath106 , respectively . as a result ,
dlp optimal packings at @xmath0 near @xmath107 and @xmath108 can be packed similarly to these icosahedrally symmetric packings while roughly adhering to the empirical rule of surface - maximization . however , none of the spheres in any of the perfectly icosahedrally symmetric packing s shells are in contact with any other spheres within that shell , and the spheres in the second icosahedron ( third shell ) are not in contact with any of the spheres in the icosidodecahedron ( second shell ) .
this lack of contact allows the spheres to be translated away from their perfectly icosahedrally symmetric positions in order to obtain a smaller dlp optimal packing radius @xmath1 for @xmath0 near @xmath109 and @xmath110 .
+ the two shells of the @xmath109 dlp optimal packing are depicted in fig .
[ n42 ] , where the shells contain @xmath18 and @xmath111 spheres , respectively , just as do the shells of the dense icosahedrally symmetric packing described in the first two rows of table [ icostable ] .
the variation on the two perfectly icosahedrally symmetric polyhedra that form the @xmath109 dlp optimal packing achieves an improvement of @xmath112 sphere diameters in @xmath13 , with @xmath113 .
the last shell of the @xmath110 dlp optimal is depicted in fig .
the @xmath110 packing includes ; @xmath18 spheres within radial distances @xmath114 and @xmath115 , @xmath111 within radial distances @xmath116 and @xmath117 , @xmath18 ( of which five are rattlers ) within radial distances @xmath118 and @xmath119 ( icosahedrally symmetric to @xmath120 sphere diameters ) , and @xmath121 at exactly @xmath122 .
the dlp optimal packing for @xmath110 achieves an improvement of @xmath123 over the perfectly icosahedrally symmetric packing . in a spherical region ,
the most area available to place the centers of nonoverlapping spheres is on the surface .
consequently , is is perhaps expected that dlp optimal packings would contain saturated or nearly saturated surfaces .
however , it is surprising that the salient features of certain dlp optimal packings are entirely determined by the distribution of the spheres with centers at precisely radius @xmath1 .
the dlp optimal packings for @xmath10 and @xmath11 are termed `` maracas '' packings ; they are perfect examples of the phenomenon of surface - maximization , and exhibit some of the lowest values of the greatest @xmath124 compared , respectively , to the packings in the sets @xmath125 and @xmath126 .
the maracas packings each consist of a few unjammed spheres free to rattle within a `` husk '' composed of the maximal number of spheres that can be packed with centers at @xmath1 .
further , @xmath127 and @xmath128 is only @xmath129 sphere diameters larger than @xmath130 .
figure [ maracas ] depicts the first and only shells for the maracas packings .
+ as in @xmath5 , for small enough @xmath0 , there are many dlp optimal packings in @xmath7 where a number of jammed spheres form a cavity around the central sphere such that were the central sphere not fixed , it would be a rattler .
the @xmath131 packing already discussed exhibits this characteristic , containing eight spheres with centers at @xmath132 and six with centers at @xmath133 from the center of the central sphere .
the @xmath134 packing has a similar cavity composed of three layers of four spheres each with centers at respective distances @xmath135 , @xmath136 , and @xmath137 , configured as squares . the @xmath138 optimal packing exhibits a cavity composed of four layers of four spheres each with centers configured as squares and at respective distances @xmath139 , @xmath140 , @xmath141 , and @xmath142 .
the cavities around the central sphere in the @xmath143 and @xmath144 packings are composed of ten layers and nine layers , respectively , of two spheres each .
the cavities formed are not , however , always symmetric . for @xmath145 of
the @xmath146 @xmath0 studied , dlp optimal packings were found containing cavities such that the center of the nearest sphere to the central sphere was at distance @xmath147 ; for only seven of these is any layer composed of more than one sphere .
in general , the cavities range in number of spheres from @xmath148 for @xmath149 to @xmath150 for @xmath151 .
the center of the first sphere forming the cavity ranges from distance @xmath152 for @xmath153 to @xmath154 for @xmath151 .
the farthest of the @xmath150 spheres forming the cavity in the @xmath151 dlp optimal packing has center at distance @xmath155 , indicating that the volume of space available in the cavity to the center of the central sphere is more than three times the sphere s volume .
dlp optimal packings in @xmath7 are widely spatially diverse and differ , particularly on the surface , from subsets of the barlow packings at all @xmath0 .
they sometimes display elements of perfect symmetry and often display elements of imperfect symmetry , such as imperfect icosahedral symmetry for sufficiently small @xmath0 .
they are similar in these respects to dlp optimal packings in @xmath5 , which differ from packings of contacting disks with centers on the vertices of the triangular lattice . however ,
at sufficiently large values of @xmath0 in any @xmath12 , the bulk of dlp optimal packings must begin to closely resemble a subset of one of the densest infinite packings in respective @xmath12 or fail to be a densest local packing . in @xmath7 ,
optimal packings tend to have a minimum number of shells and a last shell that is almost always nearly saturated ( or saturated ) .
these features lead to dlp optimal packings most closely resembling , as measured by a scalar similarity metric ( [ simmetric ] ) , subsets of barlow packings consisting of @xmath8 spheres ( a packing in @xmath156 ) with the same distribution of coordination shells as an fcc packing . knowledge of @xmath1 for certain @xmath157 in @xmath12 makes possible the construction of a rigorous upper bound ( [ upperbound ] ) on the maximal density of an infinite sphere packing in @xmath12 ; this bound becomes more restrictive as @xmath82 grows large , and becomes the equality @xmath158 as @xmath159 .
knowledge of the @xmath1 also makes possible the construction of a function @xmath36 that is an upper bound ( [ zrbound ] ) on the expected number of sphere centers @xmath40 within distance @xmath13 from any given sphere center , which can be related to a packing s pair correlation function @xmath29 in @xmath7 by ( [ zr ] ) .
this upper bound is a realizability condition on a candidate pair correlation function @xmath29 for a packing of spheres , similar to the nonnegativity conditions on @xmath29 and its corresponding structure factor @xmath31 .
the function @xmath36 is also a significantly more restrictive upper bound on candidate cross - correlation functions for a packing of a special central sphere and its surrounding spheres , say , a single spherical solute molecule that attracts same - size spherical solvent molecules .
the critical distinction between a cross - correlation function for a single sphere and a system - wide pair correlation function @xmath29 is that @xmath29 is an expected value or average over all identical spheres in a packing , whereas a cross - correlation function for a single sphere applies locally , just as the function @xmath36 is derived locally .
considering , for example , the spheres forming the cavity wall in dlp optimal packings in @xmath7 with cavities around the central sphere , it is clear that not every sphere in a packing can have the maximal number @xmath36 of sphere centers within distance @xmath13 from its center ( except for @xmath13 such that @xmath160 , where the barlow packings realize this criterion ) .
however , as any single ( solute ) sphere can have the maximal number of ( solvent ) sphere centers within distance @xmath13 from its center , a @xmath40 defined in terms of a cross - correlation function of a single central sphere can be equal to @xmath36 at any @xmath13 .
otherwise expressed , for @xmath161 any realizable cross - correlation function between nonoverlapping spherical solute molecules of unit diameter ( of type @xmath162 ) in the dilute limit and same - size spherical solvent molecules ( of type @xmath163 ) , @xmath164 where the notation @xmath165 indicates the mathematical supremum .
the function @xmath36 is thus a significantly more restrictive upper bound for candidate cross - correlation functions of a single solute sphere amongst same - size solvent spheres than for candidate pair correlation functions of packings of indistinguishable spheres .
the characteristics of dlp optimal packings in each dimension @xmath3 are heavily dependent on the underlying differences in packing spheres densely in @xmath4 .
for example , in @xmath5 there is a unique arrangement of @xmath166 disks in contact with a central disk , whereas in @xmath7 there are a continuum of arrangements of @xmath167 spheres in contact with a central sphere . in @xmath5 , there is a unique densest infinite packing of disks and in @xmath7 , there are an uncountably infinite number of densest infinite packings of spheres .
differences in characteristics across dimension are also driven by the existence of particularly locally dense dimensionally - unique packings . in @xmath5 , these include the wedge hexagonal packings described in paper i and the curved hexagonal packings @xcite , the densest local packings for @xmath168 , @xmath169 . in @xmath7 , they include the two perfectly icosahedrally symmetric packings for @xmath109 and @xmath110 ( table [ icostable ] ) that contain only two and four shells , respectively .
it is curious to note that both the curved hexagonal packings and the perfectly icosahedrally symmetric packings for @xmath109 and @xmath110 are composed of a relatively small number of densely - packed shells of spheres ; however , in the curved hexagonal packings , the spheres in any given shell are in contact with one another whereas in the aforementioned perfectly icosahedrally symmetric packings , they are not .
dimensions four , eight and twenty - four are similar to dimension two in that in each of these dimensions there is a unique ( uniqueness is conjectured for @xmath170 and proved for @xmath78 and @xmath79@xcite ) arrangement of spheres with kissing numbers @xmath171 ( recently proved by musin in @xcite ) , @xmath172 , and @xmath173 .
the densest known packings in dimensions four , eight , and twenty - four are also conjectured to be unique , and each is a lattice packing that is self - dual , i.e. , its reciprocal lattice is itself ( the dual of the triangular lattice is similarly a triangular lattice , though resized and rotated @xmath111 degrees ) .
the self - duality of the @xmath174 ( @xmath78 ) and leech ( @xmath79 ) lattices has been exploited to prove , up to a very small numerical tolerance , that identical nonoverlapping spheres with centers on the sites of these lattices are the densest packings of spheres in their respective dimensions @xcite . in dimension five ,
the densest known packings can be described , similarly to barlow packings , as stackings of layers of the densest packings in @xmath175 .
consequently , we might expect to find dlp optimal packings in @xmath12 , @xmath170 , @xmath148 , and @xmath176 similar to the curved hexagonal packings in @xmath5 , whereas we would not expect to find such optimal packings in @xmath177 .
these and other dimension - dependent dense packing characteristics could have an effect on the probability of freezing in a overcompressed liquid of hard spheres in @xmath12 .
recent work @xcite suggests that the phase transition from a overcompressed hard - sphere liquid in @xmath7 to a crystalline solid with sphere centers near to the sites of the spheres centers in a barlow packing may be described as a two - stage process . in the first stage ,
small clusters of spheres form that are denser than either the liquid or crystalline solid states . in the second stage ,
the dense clusters grow in size and decrease in density while their bulk ( interior ) transforms from the center outward into a crystalline solid state @xcite .
an analogy can be drawn between dlp optimal packings in @xmath7 for smaller @xmath0 and the small clusters , in that both are locally denser than corresponding subsets of barlow packings or crystalline solid states and in that neither are similar ( both angularly and radially ) to small subsets of barlow packings .
a similar analogy applies between dlp optimal packings for sufficiently large @xmath0 and the larger clusters with crystalline solid interiors , in that both are very similar to barlow packings in their bulk and not similar on and near their surfaces . in general , the probability per unit time of freezing in hard - sphere liquids at comparable overcompression decreases with increasing dimension , at least for dimensions two through six @xcite .
however , with the previously described two - stage process in mind , consideration of the dimension - dependent characteristics of the densest local packings could lead to a further increase , beyond what might otherwise be predicted by the general trend , in estimates of the probability per unit time of freezing in @xmath5 as compared to in @xmath7 . for example , unlike in @xmath7 , in @xmath5
there are several values of @xmath178 , specifically , @xmath179 , @xmath111 and @xmath180 , for which one of the densest local packings is a subset of the densest infinite packing , and therefore for which the equality in the upper bound ( [ zrbound ] ) can hold . we might expect to see this occurring in @xmath175 , @xmath181 , or @xmath182 as well , but not in @xmath177 .
the equivalence of densest local packings at certain @xmath0 and subsets of the densest infinite packing in @xmath5 suggests that the first stage of the two - stage crystallization process , in which small clusters of spheres form that are denser than either the liquid or crystalline solid states , may be shortened in duration for hard - disk liquids in @xmath5 relative to hard - sphere liquids in @xmath7 . if this is the case , accounting for the equivalence of densest local packings and subsets of the densest infinite packings should result in an increase in the probability per unit time of freezing in @xmath5 and potentially @xmath175 , @xmath181 , and @xmath182 as compared to in @xmath7 and @xmath177 . for hard - sphere liquids at densities near the freezing point , in @xmath5 relative to in @xmath7 , the more pronounced `` shoulder '' @xcite appearing in pair correlation functions between the first and second nearest - neighbor distances
could be evidence indicating such a shortened first stage .
this example and the similarities between dlp optimal packings and the nuclei described in the two - stage description of crystallization suggest that there may be an explicit connection between freezing in overcompressed hard - sphere liquids and dlp optimal packings .
it would be interesting , in the context of a revised nucleation theory , to explore the relationship between nucleation and dense local clusters of spheres configured with centers near the sites of sphere centers in the densest local packings .
* acknowledgements : * s.t .
thanks the institute for advanced study for its hospitality during his stay there .
this work was supported by the division of mathematical sciences at the national science foundation under award number dms-0804431 and by the mrsec program of the national science foundation under award number dmr-0820341 .
knowledge of the densest infinite packings in @xmath7 , the barlow packings , allows @xmath1 to be bounded both from above and below .
a lower bound can be obtained through the observation that it is not possible to remove a finite number @xmath8 of spheres from an infinite barlow packing and replace them by @xmath8 spheres packed as an optimal dlp packing .
if this could be accomplished at even one value of @xmath0 , then the barlow packings would not be the only densest infinite packings of identical nonoverlapping spheres in @xmath84 , as this operation could be repeated _ ad infinitum _ to yield an infinite - volume packing fraction @xmath46 greater than or equal to @xmath183 .
consider removing all @xmath8 spheres ( including a central sphere ) in a barlow packing with centers within distance @xmath184 of the center of the central sphere , where @xmath185 is chosen to include all spheres in the coordination shell at @xmath186 .
we term the set of all such subsets of @xmath187 spheres chosen from all barlow packings @xmath188 , similar to the @xmath156 defined in sec .
[ scandbarlow ] . attempting to replace ( without overlap ) the removed spheres with a dlp packing of @xmath187 spheres with optimal radius @xmath189
, we see that @xmath190 must be in the range @xmath191 .
if @xmath190 were less than @xmath192 , then the dlp packing would nt be optimal .
if @xmath190 were greater than or equal to @xmath114 , then the barlow packings would not be the only densest packings of identical nonoverlapping spheres in @xmath7 , as for @xmath193 , the dlp packing with optimal radius @xmath189 can be always be placed such that none of its spheres overlaps any of the remaining spheres in the barlow packing .
this range of epsilon results in the lower bound @xmath194 , valid for any @xmath195 chosen as stated above . in practice ,
@xmath190 may be reduced significantly below the value of @xmath114 .
for example , as for certain @xmath196 , @xmath197 varies between the packings in the set @xmath188 , @xmath190 is reduced from unity by the difference between the largest and smallest @xmath186 for these @xmath196 .
more generally , the value of @xmath190 can be significantly reduced by investigating the geometric considerations of placing the @xmath8 spheres of any dlp optimal packing of radius @xmath1 into the void created by removing @xmath0 spheres from a barlow packing . reducing @xmath190 to a minimum possible value @xmath198 results in the lower bound @xmath199 .
any set of @xmath8 spheres in a given @xmath156 can be used as a reference set in the similarity metric ( [ simmetric ] ) . however , for @xmath200 , there are always multiple reference sets of @xmath8 spheres that will produce the same @xmath201 and consequently the same value of @xmath124 for a given comparison set . for example , as the metric ( [ simmetric ] ) is insensitive to the angular position of any sphere , for a furthest coordination shell that is not full , the same @xmath201 is produced regardless of which of the spheres in the furthest shell is included in the packing used as a reference set .
when these and all other degeneracies are taken into account , we find that for @xmath156 including only packings of @xmath202 , and @xmath17 barlow stacking layers , where all @xmath203 packings are constructed from @xmath17 layers , there are @xmath204 and @xmath205 distinct sets @xmath201 . for
the @xmath0 studied where the packings in @xmath156 do not all have the same number of layers , the packings in @xmath156 include @xmath206 , @xmath207 , and sometimes @xmath208 layers . for these @xmath156 ,
the set of distinct @xmath201 is a subset of the set of distinct @xmath201 for any @xmath156 including packings of only @xmath208 layers , and the number of distinct sets @xmath201 is between the number for @xmath156 including only packings of @xmath206 layers and for @xmath156 including only packings of @xmath208 layers .
an intuitive analysis of the determination of the shells @xmath201 suggests that the presence of relatively fewer shells in a reference set can generally increase the value of @xmath124 .
an increase of this sort does occur when shells are of radial width large enough such that the number of sphere centers within each shell approaches the number density of the packing , i.e. , when radial width is on the order of a sphere s diameter .
however , this is not the case for the @xmath201 derived from the sets in each @xmath156 .
indeed , direct analysis of the distribution of spheres within each individual shell @xmath209 confirms that spheres in dlp optimal packings for a given @xmath0 are in general radially configured in a relatively small number of shells .
that is , the radial positions of the sphere centers in dlp optimal packings are clustered around a smaller number of distances from the center of the central sphere , relative to the number of coordination shells in the packings in @xmath156 , and the large fraction of dlp optimal packings that are most similar to fcc - derived packings is not the result of a design - flaw in the similarity metric ( [ simmetric ] ) . | the densest local packings of @xmath0 three - dimensional identical nonoverlapping spheres within a radius @xmath1 of a fixed central sphere of the same size are obtained for selected values of @xmath0 up to @xmath2 . in the predecessor to this paper [ a.b .
hopkins , f.h . stillinger and s. torquato ,
phys .
rev .
e * 81 * 041305 ( 2010 ) ] , we described our method for finding the putative densest packings of @xmath0 spheres in @xmath3-dimensional euclidean space @xmath4 and presented those packings in @xmath5 for values of @xmath0 up to @xmath6 .
we analyze the properties and characteristics of the densest local packings in @xmath7 and employ knowledge of the @xmath1 , using methods applicable in any @xmath3 , to construct both a realizability condition for pair correlation functions of sphere packings and an upper bound on the _ maximal _ density of _ infinite _ sphere packings . in @xmath7 , we find wide variability in the densest local packings , including a multitude of packing symmetries such as perfect tetrahedral and imperfect icosahedral symmetry .
we compare the densest local packings of @xmath0 spheres near a central sphere to minimal - energy configurations of @xmath8 points interacting with short - range repulsive and long - range attractive pair potentials , e.g. , @xmath9 lennard - jones , and find that they are in general completely different , a result that has possible implications for nucleation theory .
we also compare the densest local packings to finite subsets of stacking variants of the densest infinite packings in @xmath7 ( the barlow packings ) and find that the densest local packings are almost always most similar , as measured by a _
similarity metric _ , to the subsets of barlow packings with the smallest number of coordination shells measured about a single central sphere , e.g. , a subset of the fcc barlow packing .
additionally , we observe that the densest local packings are dominated by the dense arrangement of spheres with centers at distance @xmath1 .
in particular , we find two `` maracas '' packings at @xmath10 and @xmath11 , each consisting of a few unjammed spheres free to rattle within a `` husk '' composed of the maximal number of spheres that can be packed with centers at respective @xmath1 . |
in real quantum world hamiltonian of most systems is time - dependent and thus solving the time dependent schrdinger equation , in non - relativistic domain , is a needed one .
a class of such systems is systems with moving boundaries .
problem of a particle in a one - dimensional infinite square - well potential with one wall in uniform motion has been noticed from different aspects @xcite .
the concept of quantum effective force , time - derivative of expectation value of the momentum operator , was introduced in @xcite in the context of the quantum deflection of ultracold particles from mirrors . for the problem of a particle in a 1d expanding infinite
well potential it was shown that this force apart from a minus sign is equal to the expectation value of the gradient of the quantum potential @xcite .
exact solution of the schrdinger equation has been found and examined for a particle in a circular trap with a wall in uniform motion @xcite and for a particle in a hard sphere whose wall is moving with a constant velocity @xcite .
the aim of the present work is to probe some aspects of the time - dependent boundary condition for a particle confined in a circular and spherical box focusing on bohmian interpretation of quantum mechanics that have remained already unnoticed .
although bohmian formalism does not give predictions going beyond those of quantum mechanics whenever the predictions of the later are unambiguous , it should be used due to its interpretational advantages .
the plan of the paper is as follows . in section [ sec : qmforce ]
we consider quantum effective force in arbitrary directions for the problem of a particle confined in time - dependent cylindrical and spherical potential separately . in section [ sec : bm ] we present essentials of bohmian trajectory approach to quantum mechanics ; and give the solution to the guidance equation when the particle is initially in an energy eigenstate .
this section contains numerical calculations .
section [ sec : summary ] gives the concluding remarks .
time - derivative of the expectation value of the momentum operator of a free particle in a box with volume @xmath0 bounded by the moving closed surface @xmath1 along the arbitrary _ fixed _
direction @xmath2 is given by @xmath3~,\end{aligned}\ ] ] where we have used leibnitz integral rule in the second equality ; and @xmath4 shows the velocity of the surface @xmath1 .
now , using the divergence theorem and the schrdinger equation @xmath5 for the free particle we get @xmath6 \nabla_{q_{_0 } } \psi~ , % -\psi^ * ( \hat{n}. { \vec{\nabla } } ) \nabla_q \psi)]\end{aligned}\ ] ] where we have used the vanishing on the surface boundary condition .
@xmath7 represents the normal component of the operator @xmath8 .
consider a particle with mass @xmath9 inside an infinite cylindrical - well potential .
the cylinder has a time - dependent radius @xmath10 where @xmath11 is a constant ; and a height @xmath12 , the bottom and the top surfaces being at @xmath13 and @xmath14 .
the exact solutions of the schrdinger equation ( [ eq : sch ] ) for this problem are given by @xcite @xmath15 \frac{\sqrt{2}}{l(t ) } \frac{1}{| j_{m + 1}(x_{m n } ) im\phi}}{\sqrt{2\pi } } \nonumber \\ & \times & \sqrt{\frac{2}{z } } \sin { \left ( k \pi \frac{z}{z } \right ) } \exp \left [ - i \frac{\hbar k^2 \pi^2}{2 \mu z^2 } t \right ] \nonumber\\ & \equiv & f_{mn}(\rho , t ) \frac{1}{\sqrt{\pi z}}~e^{\pm im\phi } \sin { \left ( k \pi \frac{z}{z } \right ) } \exp \left [ - i \frac{\hbar k^2 \pi^2}{2 \mu z^2 } t \right ] ~,\end{aligned}\ ] ] where , @xmath16 ; @xmath17 ; @xmath18 , @xmath19 , @xmath20 and @xmath21 . the first line are solutions for a particle in a _
circular _ box instead of a cylindrical one . the general solution of eq .
( [ eq : sch ] ) is a superposition of functions ( [ eq : norm_waves_cylind ] ) @xmath22 with time - independent coefficients @xmath23 determined from the relation @xmath24 eq .
( [ eq : qmforce ] ) transforms to @xmath25,\end{aligned}\ ] ] in cylindrical coordinates @xmath26 . from eqs .
( [ eq : norm_waves_cylind ] ) and ( [ eq : geral - sol_circular ] ) one has , @xmath27 for arbitrary radial direction @xmath28 one obtains @xmath29 which from eq .
( [ eq : psi_derivatives_cylinder ] ) becomes zero at both the bottom and the top surfaces @xmath13 and @xmath14 ; and the second term of the right - hand - side is zero at the lateral surface @xmath30 .
thus , we get @xmath31 one can similarly obtain quantum effective force @xmath32 along arbitrary azimuthal direction @xmath33 and @xmath34 \exp \left [ - i \frac{\hbar ( k^2 - { k^{\prime}}^2 ) \pi^2}{2 \mu z^2 }
t \right ] ~.\end{aligned}\ ] ] along direction @xmath35 .
when the initial wavefunction is an energy eigenstate of a smaller box with radius @xmath36 , @xmath37 then @xmath38 where , @xmath39 in such a case the modulus of @xmath40 and its derivatives are independent of azimuthal angle @xmath41 and thus integral over @xmath41 of ( [ eq : force_cylinder_rho ] ) , ( [ eq : force_cylinder_phi ] ) and ( [ eq : force_cylinder_z ] ) is zero .
therefore , quantum effective force is zero in any arbitrary direction for the problem of a particle in a cylindrical trap . for a particle with mass @xmath9 inside an infinite spherical - well potential with radius @xmath10 , exact solutions of the schrdinger equation ( [ eq : sch ] )
read @xcite @xmath42 \sqrt{\frac{2}{l^3(t ) } } \frac{1}{| j_{l + 1}(x_{l n } ) | } j_{l } \left ( x_{l n } \frac{r}{l(t ) }
\right ) y_{l m}(\theta , \phi ) ~ , \nonumber\\ & \equiv & g_{ln}(r , t ) y_{l m}(\theta , \phi)~,\end{aligned}\ ] ] where , @xmath43 ; @xmath44 ; @xmath17 and @xmath45 . the general solution of eq .
( [ eq : sch ] ) is a superposition of functions ( [ eq : norm_waves_spherical ] ) @xmath46 with time - independent coefficients @xmath47 determined from the relation @xmath48 in spherical coordinates @xmath49 eq.([eq : qmforce ] ) transforms to @xmath50 from eqs .
( [ eq : norm_waves_spherical ] ) and ( [ eq : geral - sol_spherical ] ) one obtains @xmath51 for an arbitrary radial direction @xmath52 , one has @xmath53
\frac{\partial \psi}{\partial r } \nonumber\\ & + & [ \sin \theta_0 \cos \theta \cos ( \phi-\phi_0 ) - \cos \theta_0 \sin
\theta ] \frac{1}{r } \frac{\partial \psi}{\partial \theta } -\sin \theta_0 \sin ( \phi-\phi_0 ) \frac{1}{r
\sin \theta } \frac{\partial \psi}{\partial \phi}~,\end{aligned}\ ] ] and thus from ( [ eq : psi_der_sphere ] ) @xmath54 \frac{\partial \psi}{\partial r } \bigg|_{r = l(t)}~.\end{aligned}\ ] ] ultimately , we get @xmath55 ~ \bigg| \frac{\partial \psi}{\partial r } ( r = l(t ) , \omega ) \bigg|^2 ~.\end{aligned}\ ] ] in the same way , one can show @xmath56 ~ \bigg| \frac{\partial \psi}{\partial r } ( r = l(t ) , \omega ) \bigg|^2 ~ , \label{eq : force_sphere_theta}\\ % \frac{d}{dt } \langle p_{\phi_{_0 } } \rangle ( t ) & = & \frac{\hbar^2}{2\mu } l^2(t ) \sqrt { \frac{8\pi}{3 } } ~ { \text{im } } \left ( e^{-i\phi_0 } \int d\omega ~y_{11}(\omega ) ~ \bigg| \frac{\partial \psi}{\partial r } ( r = l(t ) , \omega ) \bigg|^2 \right ) ~ , \label{eq : force_sphere_phi}\end{aligned}\ ] ] where eqs .
( [ eq : force_sphere_theta ] ) and ( [ eq : force_sphere_phi ] ) display respectively quantum effective force in arbitrary directions @xmath57 and @xmath58 . when the initial wavefunction is an energy eigenstate of a smaller box with radius @xmath36 , @xmath59 then @xmath60 where , @xmath61 in such a case @xmath62 is independent of the azimuthal angle @xmath41 , thus the first integral of ( [ eq : force_sphere_r ] ) and ( [ eq : force_sphere_theta ] ) ; and the integral of ( [ eq : force_sphere_phi ] ) would be zero .
@xmath63 is an even function of @xmath64 and thus the second integral of ( [ eq : force_sphere_r ] ) and ( [ eq : force_sphere_theta ] ) is also zero .
therefore , quantum effective force is zero in any arbitrary direction while the force is not zero for the corresponding problem in 1d @xcite .
in causal interpretation of quantum mechanics by introducing polar form @xmath65 in the schrdinger equation ( [ eq : sch ] ) one obtains @xmath66 where @xmath67 is called quantum potential energy . in this theory particle trajectory @xmath68
is determined by simultaneous integration of the time - dependent schrdinger equation and the guidance equation @xmath69 by specifying the initial condition @xmath70 . in this context
one can easily show that the time - derivative of the expectation value of _ actual _ momentum of the particle , @xmath71 , is given by @xcite @xmath72 = - \int_v dv~ r^2 \nabla_{q_{_0 } } q = \langle - \nabla_{q_{_0 } } q \rangle ( t)~.\end{aligned}\ ] ] using the polar form of the wavefunction in the second line of eq .
( [ eq : 1 ] ) and vanishing on the surface boundary condition we obtain the equality @xmath73 .
respecting to the newton s second law @xmath74 in bohmian mechanics , last equality in eq .
( [ eq : bm_qmforce ] ) shows the relation @xmath75 for the actual momentum of the particle .
although , the classical potential energy @xmath76 is zero inside the box , the quantum potential energy is non - zero .
this is the reason of interaction of the boundary with the confined particle . in the following
we will compute trajectories of a bohmian particle which is initially in an energy eigenstate of the box .
so , @xmath77 in eqs . (
[ eq : i_int_circ ] ) and ( [ eq : i_int_sphere ] ) is replaced with @xmath78 . for numerical calculations
new quantities @xmath79 with dimension of velocity , @xmath80 with dimension of time and dimensionless quantity @xmath81 are defined for the problem of a particle in a circular trap . corresponding quantities in the case of a particle in a spherical trap are @xmath82 , @xmath83 and @xmath84 . in our calculations velocity of the moving wall is determined versus the above velocities . using eqs .
( [ eq : guaidance ] ) and ( [ eq : cir_special_sol ] ) one finds , @xmath85 @xmath86 is found by numerical solving of eq .
( [ eq : v_rho_circular ] ) and then by using this result in ( [ eq : v_phi_circular ] ) one obtains @xmath87 trajectories for the case of a particle in circular box are displayed in @xmath88 plane in figure [ fig : traj_circular ] for different rates of contraction and expansion . here , the particle is initially in the energy eigenstate @xmath89 .
it should be noted that trajectories cross each other at different times ( different values of the @xmath41-coordinate ) .
thus , there is no problem regarding the single - valuedness of the wavefunction .
( [ eq : guaidance ] ) and ( [ eq : i_int_sphere ] ) lead to @xmath90 for the velocity of bohmian particle in a spherical trap and initially in an energy eigenstate . from eq .
( [ eq : v_theta_spherical ] ) one sees that during the motion , polar angle @xmath91 will not change , i.e. , motion would be on the surface of a cone with opening angle @xmath92 . by numerical solution of eq .
( [ eq : v_r_spherical ] ) one finds @xmath93 and then eq . ( [ eq : v_phi_spherical ] )
is solved , @xmath94 from this equation it s apparent that for @xmath95 azimuthal angle @xmath41 does not change during the motion which is an obvious result : angular momentum in @xmath35-direction is zero in such a case .
thus , in this case motion is along the straight line @xmath96 .
figure [ fig : traj_spherical_1d ] represents trajectories for particle in a spherical trap . here , particle initially is in the ground state @xmath97 which corresponds to a particle in a 1d square box @xcite .
so far we have studied a single - particle system .
now , we consider a two - body system which are confined in our time - dependent spherical trap .
a novel feature of the causal description of a many - body system is the non - local connection of particles : dependence of instantaneous motion of any one particle to the coordinates of all other particles at the same time . in this case
the particle trajectories @xmath98 are determined by the following system of two simultaneous differential equations @xmath99 if the particles are identical the wavefunction of the system must be symmetrized .
there are different statistics , fermi - dirac ( fd ) for which total wavefunction must be antisymmetric with respect to the exchange of particles in the system and bose - einstein ( be ) for which total wavefunction is symmetric under the exchange of particles . when particles are distinguishable they are independent obeying maxwell - boltzmann ( mb ) statistics .
in bohmian perspective the particles are always distinguishable in all cases @xcite .
the initial wavefunction is taken to be @xmath100 \nonumber \\ & = & \frac { \sqrt{2 } } { a } \frac{1}{r_1 r_2 } \left [ \sin \left ( \frac{\pi r_1}{a } \right ) \sin \left ( \frac{2 \pi r_2}{a } \right ) \pm \sin \left ( \frac{2 \pi r_1}{a } \right ) \sin \left ( \frac{\pi r_2}{a } \right ) \right ] y_{00}(\omega_1 ) y_{00}(\omega_2)~,\end{aligned}\ ] ] which is independent of polar and azimuthal angles and so the motion is one - dimensional .
as particles are classically non - interacting this expression will be preserved by the schrdinger evolution .
in this work the problem of a one - body and a two - body system confined in time - dependent traps were studied for particular initial conditions . for the one - body case we gave analytical relations for quantum effective force .
it was seen that when the particle is initially in an energy eigenstate of a smaller box , irrespective of the size of the smaller box expectation value of the momentum operator does not change with time while for the corresponding 1d system the behavior is different .
a selection of bohmian trajectories computed for a particle initially in an energy eigenstate of the time - dependent circular and spherical boxes for different values of the wall velocity .
for the two - particle system details of motions implied by nonfactorizable wavefunctions are subtle : one can not say that fermions are repelled and bosons are attracted , in comparison with distinguishable particles obeying mb statistics .
the root mean square separation confirms one s expectation : it is less ( more ) in be ( fd ) case than in the mb one . in both one - body and two - body system moving wall makes the system to be non - stationary as a result of which bohmian particles move .
99 s. w. doescher and m. h. rice , _ am .
_ * 37 * ( 1969 ) 1246 ; + d. w. schlitt and c. stutz , _ am .
_ * 38 * ( 1970 ) 70 ; + d. n. pinder , _ am .
j. phys . _ * 58 * ( 1990 ) 54 ; + m. g. da luz and bin kang cheng ,
_ j. phys . a _ * 25 * ( 1992 ) l1043 ; + v. v. dodonov , a. b. klimov and d. e. nikonov , _ j. math
_ * 34 * , ( 1993 ) 3391 ; + o. fojon , m. gadella , l.p .
lara , _ comput .
* 59 * ( 2010 ) 964 s. v. mousavi , _ phys .
_ * 86 * ( 2012 ) 035004 v. v. dodonov and m. a. andreata , _ phys
. lett . a _ * 275 * , ( 2000 ) 173 ; + v. v. dodonov and m. a. andreata , _ laser phys . _
* 12 * , ( 2002 ) 57 yce c. , _ phys
a _ * 327 * ( 2004 ) 107 + s. v. mousavi , _ phys .
a _ , * 377 * ( 2013 ) 1513 s. v. mousavi , _ europhys .
_ , * 99 * ( 2012 ) 30002 d. bohm _ phys .
* 85 * ( 1952 ) 166 ; + d. bohm _ phys . rev _
* 85 * ( 1952 ) 180 p. r. holland 1993 , _ the quantum theory of motion _
( cambridge : cambridge university press ) , pp . 277 - 310 | trajectories of a bohmian particle confined in time - dependent cylindrical and spherical traps are computed for both contracting and expanding boxes .
quantum effective force is considered in arbitrary directions .
it is seen that in contrast to the problem of a particle in an infinite rectangular box with one wall in motion , if particle initially is in an energy eigenstate of a tiny box the force is zero in all directions .
trajectories of a two - body system confined in the spherical trap are also computed for different statistics .
computations show that there are situations for which distance between bosons is greater than the fermions .
however , results of average separation of the particles confirm our expectation about the statistics . |
the central parsec of the milky way ( mw ) is a very special region of our galaxy .
it not only hosts the central supermassive black hole , associated with sgr a * , with well determined mass @xcite , but also contains a significant number of early - type stars , including some in the wolf - rayet stage .
the central cavity has a mini - spiral structure of ionized gas streamers , encircled by a circum - nuclear disk ( cnd ) of molecular gas .
most of the ionized gas emission appears in hydrogen and helium recombination lines , very likely originating from ionization produced by the many young and hot ob stars ; the equivalent to some @xmath5 o9 stars would be required to account for the emission @xcite .
an anomalous structure , called mini - cavity , was identified in the mini - spiral , with a diameter of @xmath6 , about @xmath7 to the southwest of sgr a * @xcite , associated with strong [ fe iii ] emission @xcite . for a recent review of the literature on the galactic centre ,
see @xcite .
x - ray binary transients are overabundant in the central parsec of the mw . @xcite
have found that the overabundance is at least a factor of 20 , per mass , in the central parsec when compared to the abundance at a distance between 1 and 23 parsecs .
it is likely that this overabundance is due to three - body interactions between binary star systems and either black holes or neutron stars that have concentrated in the central parsec through dynamical friction .
this is also supposed to be the case with the formation of similar systems in the cores of globular clusters @xcite .
the known x - ray transient closest to sgr a * , at only @xmath8 south of the centre of the mw , is cxogx j174540.0 - 290031 , discovered in 2004 @xcite .
it was found to be an x - ray binary with an orbital period of 7.9 hr @xcite . from k
band observations , an upper limit for any counterpart was set as @xmath9 , suggesting that it is a low - mass x - ray binary ( lmxb ) .
this x - ray transient also revealed a radio outburst @xcite .
this large radio outburst , as well as its x - ray emission , suggests that this source more likely contains a black hole than a neutron star primary @xcite .
in this paper we report the discovery of a high velocity compact nebular filament extending @xmath10 with a strong velocity gradient that could possibly be associated with a colliding shock between the streamers of the mini - spiral .
this paper analyzes archive data obtained with the spectrograph for integral field observations in the near infrared ( sinfoni ) on the vlt , available in a public access data bank .
the observations were made in k band on april 22 , 2007 .
a fore - optics of @xmath11 pixel@xmath0 was used , providing a field of view ( fov ) of @xmath12 .
a total of four exposures of the galactic nucleus were retrieved from the public data bank to be used in this work .
standard calibration images of linearity lamp , distortion fibre ( used in the data reduction to compute spatial distortions and perform a spatial rectification ) , flat lamp , arc lamp and sky field were obtained from the public data bank .
calibration images of an a0v standard star were also retrieved from the data bank .
the data reduction was made with gasgano software and included the following steps : determination of the trim , sky subtraction , correction of bad pixels , flat - field correction , spatial rectification ( including correction for spatial distortions ) , wavelength calibration and data cube construction . at the end of the process , four data cubes were obtained , with spaxels ( spatial pixels ) of @xmath13 . afterwards , with a script developed by us and using data from the a0v standard star , we applied , in an iraf environment , telluric absorption removal and flux calibration to all the data cubes .
details of the data treatment are given in appendix a.
after the data treatment , we applied pca tomography ( see appendix a ) again .
tomogram 1 , shown in green in figure 1-a , displays the young stellar population .
tomogram 2 , shown in blue in figure 1-a , reveals stars with strong infrared excess , presumably very young stars , like irs 21 @xcite and irs 1 .
the red component in figure 1-a corresponds to an image of stars obtained by subtracting the edges of the co bands from the reconstructed data cube , after suppressing eingenvectors 1 and 2 ( see steiner et al .
2009 for a description of `` feature suppression '' ) .
the stellar component revealed by figure 1-a basically shows the known young as well as the old stellar populations .
irs 16 , the closest irs source to the galactic centre , is resolved in the individual stars , as is irs 33 ( sw of the fov ) .
we performed spectral synthesis on the spectrum of each spaxel of the datacube analyzed here , using the starlight software @xcite , which matches the stellar spectrum of a given object with a model corresponding to a combination of template stellar spectra from a pre - established base . in this work
, we used the base of stellar spectra miles ( medium resolution int library of empirical spectra ) @xcite , containing the spectra of 150 stellar populations with ages between @xmath14 years and @xmath15 years and with metallicities between 0.0001 and 0.05 ( with @xmath16 ) .
the spectral synthesis resulted in a synthetic stellar spectrum for each spaxel of the data cube . before performing spectral synthesis on the data cube , however , we prepared the spectra in two steps : correcting the interstellar extinction due to the galaxy , using @xmath17 and the reddening law of @xcite ; and carrying out a spectral re - sampling with @xmath18 per pixel .
the first step used a script written in idl language , while the second used the task `` dispcor '' , of the `` noao '' package , in iraf environment . in order to analyze the emission line spectrum in more detail , we subtracted the synthetic stellar spectra obtained with the spectral synthesis from the observed ones .
this procedure resulted in a data cube with emission lines only .
we then constructed rgb images of the emission lines br@xmath19 and he i @xmath20 of this emission line data cube .
we can see that the recombination lines br@xmath19 and he i @xmath21 are detected as stellar winds in early - type stars and , also , as ionized interstellar lines basically delineating the northern arm .
a number of other lines are also seen in the southwest of the fov : @xmath22 , @xmath23 , @xmath24 , and @xmath25 are the strongest among them and are identified as [ fe iii ] lines ( lutz et al .
we then constructed rgb images of the emission lines br@xmath19 and [ fe iii ] @xmath26 , shown in figure 1-c and figure 1-d , respectively .
we identified a high velocity compact nebular filament , @xmath27 south of sgr a*. the structure extends over @xmath10 and presents a velocity gradient of @xmath28 km s@xmath0 arcsec@xmath0 for he i , br@xmath19 and [ fe iii ] @xmath29 lines ( figure 1-e , and figures [ fig2 ] and [ fig4 ] ) .
the peak of maximum emission seen in the [ fe iii ] and he i lines is located at @xmath30 and @xmath31 with respect to sgr a*. this position is near the star irs 33n , at @xmath32 ; @xmath33 from sgr a*. as can be seen in figure [ fig3 ] , the velocity at the peak , as estimated from the br@xmath19 , he i @xmath34 and [ fe iii ] @xmath35 is @xmath36 km s@xmath0 .
the filament has a position angle of pa @xmath4 ( see figures [ fig2 ] to [ fig4 ] ) .
the observed line ratios at the emission peak of the filament are ( he i @xmath34)/br@xmath19 = @xmath37 and ( [ feiii ] @xmath29)/br@xmath19 = @xmath38 .
the filament is positioned in one of the mini - spiral regions nearest to the galactic centre , in the field where the eastern and the northern arms of the mini - spiral cross .
it is also at the eastern edge of the mini - cavity .
this field shows significant blue - shifted he i emission line ( paumard et al .
2004 ) and also in the h92@xmath39 and h30@xmath39 lines @xcite .
the filament that we see may be described as a `` jet - like '' structure , being brighter at the centre and fading towards the borders , at least as it is seen in the he i and the [ fe iii ] lines ( see figure [ fig3 ] ) .
this is in the part of the mini - spiral closest to sgr a*. the challenge is to explain the compactness of the structure as well as its ionization ( he i ) and excitation ( [ fe iii ] ) .
the filament overlaps spatially with the star irs33n ( see figure 1-b and figure [ fig4 ] ) .
could this star be the origin of a jet ?
irs 33n is a @xmath40 star , presenting a spectral type of b0.5 - 1 i and has a radial velocity of @xmath41 km s@xmath0 @xcite while the centre of the filament is at @xmath3 km s@xmath0 .
although close in terms of spatial projected distance , the nebular filament is kinematically decoupled from the star irs 33n . therefore this association can be discarded .
km s@xmath0 ) ; 2 ( @xmath42 km s@xmath0 ) ; 3 ( @xmath43 km s@xmath0 ) ; 4 ( @xmath44 km s@xmath0 ) and 5 ( @xmath45 km s@xmath0 ) .
the northern edges of the northern and eastern arm streamers , as well as of the mini - cavity , are shown .
three stars of irs 16 as well as of irs 33 are also shown .
the position of sgr a * is marked as a @xmath46 sign .
the error circle of the lmxb ( green ) and the associated radio transients ( red ) are also shown.[fig4 ] ] the orbits of the mini - spiral have been described as containing a one - armed spiral in a keplerian disk ( lacy et al .
1991 ) that properly describes the [ ne ii ] @xmath4712.8 @xmath48 m .
an alternative description is given by @xcite , in which the whole mini - spiral is essentially described by three streamers in elliptical orbits around sgr a * , on the basis of h92@xmath39 observations .
the northern and eastern arm streamers have high eccentricities ( @xmath49 and @xmath50 , respectively ) and cross near the periastron .
the formal solution provides a projected crossing between the two orbits about @xmath51 to the south of the galactic centre .
this solution has been disputed by @xcite who find that the one - armed spiral pattern may be a better description of the [ ne ii ] nebular emission .
the observations of both h92@xmath39 and the [ ne ii ] lines have good spectral resolution but poorer spatial resolution , of @xmath52 to @xmath6 , than the data reported here .
they could not have identified such a strong velocity gradient in a filament so compact .
perhaps the best explanation for the existence of this filament is that it is associated to a shock between streamers in the mini - spiral . in the one - arm pattern ,
the position of the filament is where the one - armed spiral ends ; that could also be seen as the crossing point between the northern arm and the bar . in the three ellipses model ,
the filament is near the position where the projected orbits of the northern and the eastern arm streamers cross .
it is perhaps not a coincidence that the position angle of the filament ( pa @xmath4 ) is similar to that of the eastern arm and also of the bar .
perhaps the most challenging aspect of this interpretation is how to explain the large velocity gradient of @xmath28 km s@xmath0 arcsec@xmath0 .
it probably has to do with the proximity of the supermassive black hole and its steep gravitational potential gradient . the blue arm of the filament points toward the mini - cavity .
this structure was modeled as a bubble of ionized hot gas generated by a fast ( @xmath53 km s@xmath0 ) wind originated from one or several sources , within a few arcseconds from the mini - cavity , and blown into the partially ionized neutral gas streamers orbiting the centre of the galaxy ( eckart et al .
1992 ; wardle & yusef - zadeh 1992 ; lutz et al . 1993
; melia , coker & yusef - zadeh 1996 ) . in fact , after computing the elliptical keplerian orbits of the streamers
, @xcite suggested that the northern and eastern arms may collide in the `` bar '' region , a few arcesconds south of sgr a*. this is the region with the highest specific kinetic energy of both northern and eastern arms ( see figure 21 in zhao et al .
according to @xcite , the high specific kinetic energy density emphasizes the possibility that the mini - cavity was created by strong shocks resulting from the collisions of the two streamers .
the peak of the emission of the filament is located @xmath54 north of the binary x - ray transient cxogx j174540.0 - 290031 , a lmxb with an orbital period of 7.9 hr @xcite .
this source is associated to one transient and one persistent albeit variable , radio sources @xcite .
the x - ray transient source cxogx j174540.0 - 290031 was detected on july , 5 - 7 , 2004 and is located @xmath8 south of sgr a * , with an uncertainty of about @xmath55 @xcite .
observations made at 43 ghz a few months earlier , on march 28 , 2004 , detected two sources , named ne ( northeast , the strongest one , reaching 93 mjy at 43 ghz ) and sw ( southwest ) with a peak flux of 48 mjy .
the coincidental variability of both radio and x - ray sources strongly suggests that they are associated .
@xcite also report radio observations at 1.3 cm and found that the sw source is a persistent one , being detected before the transient phenomenon at the @xmath56 mjy level .
the ne component , not detected in previous observations , was detected after the transient .
these authors interpret the ne component as a hot spot associated with the head of the jet ejected from the persistent sw radio - variable source .
if so , then the stellar counterpart of the x - ray binary is probably near the star irs 33sw , close to the x - ray error box ( figure [ fig4 ] ) .
the position angle of the radio jet is pa @xmath57 , quite distinct from the position angle of the nebular filament .
as the filament is at @xmath58 from the x - ray source and has a distinct position angle , we may rule out any direct association between the x - ray binary and the origin of the nebular filament .
the sinfoni observations ( april 22 , 2007 ) were made a few years after the radio detection ( march 28 , 2004 ) and also after the x - ray transients detections ( july 5 - 7 , 2004 ) . the nebular filament is characterized by a spectrum that is quite distinct from other gas emitting regions in the fov .
br@xmath19 and he i @xmath34 are the strongest lines . other gas emitting regions in the fov
are probably photoionized by the young stellar population ( such as in irs 16 ) that produces uv photons capable of ionizing h and he ( genzel et al . 2010 ) .
such uv photons , because of their low energy , do not heat the nebulae and some 250 o9 equivalent stars are necessary to account for the emission @xcite .
the vicinity of the x - ray source is the only region in the fov that presents such strong forbidden lines .
but shock waves are also efficient in heating gas and enhancing forbidden line emissions .
the field of irs 33 is the hottest in the mini - spiral , as measured from the h30@xmath39 and h92@xmath39 @xcite , with a temperature of @xmath59 k and an electron density of @xmath60 @xmath61 .
the temperature of 13000 k is enough to excite the [ fe iii ] emission .
it is not enough to ionize the gas to the fe@xmath62 phase .
this requires a temperature of @xmath63 k @xcite .
but the beam size of the radio observations is @xmath64 ; thus a compact structure with @xmath63 k could certainly exist . for collisional ionization of helium
, the temperature must be even higher : @xmath65 k. we conclude that shock heating could both produce and excite the fe@xmath62 ion , but not ionize he .
therefore , a combination of non - shock ionization mechanism plus shock heating seems to be required .
could the x - ray transient be , at least partially , responsible for photoionizing the filament ?
hard x - rays are efficient in photoionizing gas and heating it at the same time @xcite , enhancing the emission from low ionization nebular species .
however , the ionization parameter @xmath66 must be @xmath67 in order to produce high ionization species and @xmath68 in order to produce low ionization species @xcite . in the present case ,
the transient x - ray luminosity was @xmath69 erg s@xmath0 .
however , at the time of the transient , diffuse x - ray emission was observed @xmath70 southwest of the source .
this suggests that the real x - ray luminosity was @xmath71 erg s@xmath0 @xcite .
the discrepant x - ray luminosity seen directly is probably due to the fact that the x - ray binary is seen nearly edge - on .
assuming that the x - ray luminosity is @xmath72 erg s@xmath0 we obtain an ionization parameter of @xmath73 at the filament ; this is too low to produce a significant ionization effect .
we conclude that the x - ray binary has no ionization or heating effect on the filament unless the source radiates soft x - rays near the eddington limit .
this could not have been directly observed due to the strong interstellar absorption and would produce @xmath74 at the filament and could be responsible for the strength of the [ fe iii ] emission and , perhaps , of he i. further studies are required to elucidate the nature of this filament and the role of the lmxb in its neighborhood ; it is important to observe the fov with better spatial resolution .
we have analyzed a data cube centered @xmath64 southeast of srg a * , obtained with the sinfoni / vlt ifu spectrograph and studied the kinematics of the gaseous ionized emission in the fov .
our main findings are : * we identified a compact nebular filament , @xmath27 south of sgr a*. the structure extends over @xmath10 and presents a velocity gradient of @xmath28 km s@xmath0 arcsec@xmath0 .
there is a peak of maximum emission seen in [ fe iii ] and he i lines , located at @xmath75 and @xmath76 with respect to sgr a*. this position is located at @xmath77 east of the star irs 33n .
this emission has the most negative velocity in the fov we have analyzed . * the velocity at the emission peak , as estimated from the br@xmath19 , he i @xmath34 and [ fe iii ] @xmath35 is @xmath36 km s@xmath0 .
the nebular filament has a position angle of pa @xmath4 , similar to that of the eastern arm at that position and to the position angle of the mini - spiral bar .
although close , the nebular filament is , kinematically decoupled from the star irs 33n , which has a velocity of @xmath41 km s@xmath0 .
* the peak position is also located @xmath54 north of the binary x - ray transient cxogx j174540.0 - 290031 , a lmxb with an orbital period of 7.9 hr .
this source is associated with one transient plus one persistent and variable radio sources . *
the [ fe iii ] line emission is strong in the filament and its vicinity .
these lines are probably excited by shock heating .
ionization / heating by soft x - ray photons from the nearby lmxb is possible , although unlikely , because of the high luminosity required . *
although we can not rule out the possibility that we are dealing with a jet - like structure , we propose that the filament may represent a shock from the collision between the northern and the eastern arm streamers around the centre , or perhaps , between the northern arm and the bar .
this research has been supported by the agencies cnpq and fapesp .
the data presented herein were obtained with the sinfoni spectrograph on the vlt / eso available on the eso public data bank .
bower g.c .
, roberts d.a . , yusef - zadeh f. , backer d.c . , cotton w.d .
, goss w.m . , lang c.c . lithwick y. , 2005 , apj , 633 , 218 cardelli j.a . , clayton g.c .
, mathis j.s . , 1989 , apj , 345 , 245 cid fernandes r. , mateus a. , sodr l. , stasinska g. , gomes j.m . , 2005 , mnras , 358 , 363 eckart a. , genzel r. , krabbe a. , hofmann r. , van der werf p.p . , drapatz s. , 1992 , nat , 355 , 526 fender r.p . , kuulkers e. , 2001 , mnras , 324 , 923 ferland g.j .
, netzer h. , 1983 , apj , 264 , 105 genzel r. , eisenhauer f. , gillessen s. , 2010 , rvmp , 82 , 3121 ghez a.m. et al .
, 2008 , apj , 689 , 1044 gillessen s. , eisenhauer f. , trippe s. , alexander t. , genzel r. , martins f. , ott t. , 2009 , apj , 692 , 1075 halpern j.p .
, steiner j.e . , 1983 , apj , 269 , l37 irons w.t . , lacy j.h . ,
richter m.j . , 2012 ,
apj , 755 , 90 lacy j.h . , achtermann j.m . ,
serabyn e. , 1991 , apj , 380 , l71 lucy l.b . , 1974 ,
aj , 79 , 745 lutz d. , krabbe a. , genzel r. , 1993 , apj , 418 , 244 melia f. , coker r.f . , yusef - zadeh f. , 1996 , apj , 460 , l33 morris m. , serabyn e. , 1996 , ara&a , 34 , 645 muno m.p . , pfal e. , baganoff f.k . , brandt w.n . ,
ghez a. , lu j. , morris m.r . , 2005a , apj , 622 , l113 muno m.p .
, lu j.r . , baganoff f.k .
, brandt w.n . ,
garmine g.p . , ghez a.m. , hornstein s.d . ,
morris m.r . , 2005b , apj , 633 , 228 osterbrock d.e . , ferland g.j
, 2006 , astrophysics of gaseous nebulae and active galactic nuclei , 2nd edn . university science books , sausalito , ca paumard t. , maillard j .-
p . , morris m. , 2004 , a&a , 426 , 81 paumard t. , et al . , 2006 , apj , 643 , 1011 porquet d. , grosso n. , blanger g. , goldwurm a. , yusef - zadeh f. , warwick r.s . , predehl p. , 2005
, a&a , 443 , 571 richardson w.h . , 1972 , josa , 62 , 55 roberts d.a . , yusef - zadeh f. , goss w.m . , 1996 ,
apj , 459 , 627 shields j.c . ,
ferland g.j . , 1994 , apj , 430 , 236 snchez - blzquez p. , et al . , 2006 , mnras , 371 , 703 steiner j.e . , menezes r.b . ,
ricci t.v . ,
oliveira a.s . , 2009 , mnras , 395 , 64 tanner a. , ghez a.m. , morris m. , becklin e.e . , cotera a. , ressler m. , werner m. , wizinowich p. , 2002
, apj , 575 , 860 wardle m. , yusef - zadeh f. , 1992 , nat , 357 , 308 yusef - zadeh f. , morris m. , ekers r. , 1989 , in morris m. , ed .
136 , the center of the galaxy .
kluwer , dordrecht , p. 443
yusef - zadeh f. , morris m. , ekers r. , 1990 , nat , 348 , 45 zhao jun - hui .
, goss w.m . , lo k.y . , ekers r.d . , 1991 , nat , 354 , 46 zhao jun - hui .
, morris m.r .
, goss w.m .
, an t. , 2009 , apj , 699 , 186 zhao jun - hui . , blundell r. , moran j.m . , downwes d. , schuster k.f . , marrone d.p . , 2010 , apj , 723 , 1097
after the data reduction , a correction of the differential atmospheric refraction was applied to all data cubes , using an algorithm developed by our group .
it is important to mention , however , that the relative shifts observed in the images of all data cubes are not compatible with what is expected for the differential atmospheric refraction effect .
therefore , we believe that these relative shifts are , at least partially , a consequence of some instrumental effect and not entirely due to the differential atmospheric refraction . in order to combine into one the four data cubes obtained after the correction of the atmospheric differential refraction , we divided the data cubes in two groups with three data cubes each . naming the data cubes 1 , 2 , 3 and 4 ( according to the order of the observations ) ,
the division in two groups was the following : + + group 1 : 1 , 2 and 3 + group 2 : 2 , 3 and 4 + + we then calculated a median of the data cubes of each group , which resulted in two data cubes at the end of the process .
finally , we calculated the average of the two data cubes and obtained the final , combined data cube . by performing a spatial re - sampling to all the images of the data cube
, we obtained spaxels of @xmath78 .
this procedure does not change the spatial resolution of the observation , but sharpens the contours of the structures in the images .
the disadvantage of such spatial re - sampling is that it introduces high spatial frequency components in the images ( usually in the form of small vertical and horizontal stripes ) .
the high spatial frequency components , however , are not a serious problem because , after the spatial re - sampling , a butterworth spatial filtering was applied to all the images of the data cube , to remove high spatial frequency noises , including the high frequency components introduced by the spatial re - sampling .
the steps of this process were as follows : calculation of the fourier transforms ( @xmath79 ) of all the images of the data cube ; multiplication of the fourier transforms @xmath79 by the image corresponding to the butterworth filter ( @xmath80 ) ; calculation of the inverse fourier transforms of all the products @xmath81 ; extraction of the real part of the calculated inverse fourier transforms .
the butterworth filter used in this case corresponds to the product of two identical circular filters , with orders @xmath82 and cutoff frequencies along the horizontal and vertical axis equal to 0.27 ny ( nyquist frequency ) .
we detected the presence of a probable instrumental `` fingerprint '' in the data cube , after the butterworth spatial filtering , which appeared in the data cube as a large horizontal strip in the images and also had a low frequency spectral signature . in order to remove it
, we applied the pca tomography technique @xcite to the data cube .
pca transforms data expressed originally in correlated coordinates into a new system of uncorrelated coordinates ( eigenvectors ) ordered by principal components of decreasing variance .
pca tomography consists in applying pca to data cubes . in this case
, the variables correspond to the spectral pixels of the data cube and the observables correspond to the spaxels of the data cube .
since the eigenvectors are obtained as a function of the wavelength , their shape is similar to spectra and , therefore , we call them eigenspectra . on the other hand , since the observables are spaxels , their projections on the eigenvectors are images , which we call tomograms
. the simultaneous analysis of eigenspectra and tomograms allows one to obtain information that , otherwise , would possibly be harder to detect . using pca tomography
, we were able to identify and remove the instrumental fingerprint of the data cube analyzed here ( steiner et al .
, in preparation ) . finally , a richardson - lucy deconvolution @xcite was applied to all the images of the data cube , using a synthetic gaussian psf .
the use of a gaussian psf was possible in this case because we verified that the original psf of this observation had an approximate gaussian shape .
the final data cube obtained after the full data treatment has a point - spread function ( psf ) with fwhm @xmath83 and a spectral resolution of @xmath84 . | the central parsec of the milky way is a very special region of our galaxy ; it contains the supermassive black hole associated with sgr a * as well as a significant number of early - type stars and a complex structure of streamers of neutral and ionized gas , within two parsecs from the centre , representing a unique laboratory .
we report the identification of a high velocity compact nebular filament 2.2 arcsec south of sgr a*. the structure extends over 1 arcsec and presents a strong velocity gradient of 200 km s@xmath0 arcsec@xmath0 .
the peak of maximum emission , seen in [ fe iii ] and he i lines , is located at @xmath1 arcsec and @xmath2 arcsec with respect to sgr a*. this position is near the star irs 33n .
the velocity at the emission peak is @xmath3 km s@xmath0 .
the filament has a position angle of pa @xmath4 , similar to that of the bar and of the eastern arm at that position .
the peak position is located 0.7 arcsec north of the binary x - ray and radio transient cxogx j174540.0 - 290031 , a low - mass x - ray binary with an orbital period of 7.9 hr . the [ fe iii ] line emission is strong in the filament and its vicinity .
these lines are probably produced by shock heating but we can not exclude some x - ray photoionization from the lmxb . although we can not rule out the idea of a compact nebular jet , we interpret this filament as a possible shock between the northern and the eastern arm or between the northern arm and the mini - spiral `` bar '' .
[ firstpage ] techniques : imaging spectroscopy ism : structure galaxy : centre x - rays : binaries |
topology in qcd plays a relevant rle in understanding several low energy properties of the theory .
one of the quantities that have a direct interest in phenomenology is the topological susceptibility @xmath1 which is defined as the correlator at zero momentum of two topological charge density operators @xmath2 .
in particular the value of @xmath1 in the pure gauge theory is interpreted in terms of the mass of the singlet pseudoscalar meson @xcite .
the lattice is an excellent tool to calculate such dimensionful observables and specifically the lattice determination of @xmath1 in the pure gauge theory is in good agreement with phenomenological expectations @xcite .
the calculation of @xmath1 or any other topology related quantity on the lattice requires a regularization @xmath3 of the topological charge density operator .
the formal nave continuum limit must satisfy @xmath4 , ( @xmath5 is the lattice spacing ) .
however in general @xmath3 does not meet the continuum ward identities @xcite .
consequently the lattice definition of the topological susceptibility @xmath6 ( @xmath7 is the lattice volume and @xmath8 ) need not coincide with the physical continuum expression @xmath1 .
the two quantities are related by @xcite @xmath9 where @xmath0 and @xmath10 are renormalization constants which for evident reasons are called multiplicative and additive respectively . in order to extract @xmath1 from the lattice data of @xmath11
one must know the values of @xmath0 and @xmath10 .
let us outline the origin of the two renormalization constants .
any matrix element which contains @xmath12 insertions of the topological charge operator can be calculated either in the continuum ( with an adequate regularization ) or on the lattice . in general the two calculations match only after the inclusion of a multiplicative renormalization constant @xmath0 @xcite for each insertion of @xmath2 . in a formal writing , @xmath13 where @xmath0 is a finite renormalization constant . in the theory with fermions the topological charge mixes with other operators related to the axial anomaly @xcite .
this mixing induces a correction to the above described multiplicative renormalization and consequently also to eq.([chil ] ) .
such a correction is however rather small @xcite and it is usually neglected consistently with the large statistical errors from a typical numerical simulation . if @xmath14 the above matrix element includes further divergences which are the origin of the additive term @xmath10 .
in fact the expression for @xmath1 contains the product of two topological charges at the same spacetime point .
the operator expansion of this product at short distances contains mixings with operators that share the quantum numbers of @xmath1 .
on the other hand it is known that the correlation function of two topological charge operators at nonzero distance is negative @xcite , @xmath15 for @xmath16 .
this inequality also holds on the lattice for any definition of @xmath3 @xcite . since @xmath1 is a positive quantity , part of the contact terms
must be included in the physical definition of @xmath1 .
the rest of the terms , if any , must be subtracted and they are @xmath10 .
a prescription is necessary to calculate @xmath10 . due to the expression @xmath17 where @xmath18 is the partition function of the gauge theory with a theta term , we know that @xmath1 vanishes within the zero topological charge sector .
we then adopt the following definition for @xmath10 : it is the value of @xmath11 in the sector of zero topological charge , @xmath19 where @xmath20 is the value of the total topological charge of a configuration ( as determined by cooling or other means , see later ) .
a nonperturbative method to calculate @xmath0 and @xmath10 has been developed in refs .
the method will be described in section 2 .
it has been used in several calculations of topological properties in qcd and other theories .
various tests and studies of efficiency have also been worked out in the past . in section 3 we will present the main contribution of the paper : a study of systematic errors that may affect the calculation of the renormalization constants and a comparison of results when they are obtained by using different methods for calibrating the background topological sector .
some conclusive comments are given in section 4 .
in refs . @xcite a technique , called `` heating method '' , to calculate the renormalization constants in eq.([chil ] ) was put forward .
the idea behind the heating method is that the uv fluctuations in @xmath3 , which are the ultimate cause for renormalizations , are effectively decoupled from the background topological signal so that , starting from a classical configuration of fixed topological content , it is possible , by applying a few updating ( heating ) steps at the corresponding value of the lattice bare coupling constant @xmath21 , to thermalize the uv fluctuations without altering the background topological content .
this result is favoured by the fact that topological modes have very large autocorrelation times , as compared to other non topological observables ( this autocorrelation time is particularly long in the case of full qcd @xcite and also in the case of a large number of colours @xcite )
. one can thus create samples of configurations within a given topological sector @xmath20 with the uv fluctuations thermalized . if @xmath22 then the measurement of @xmath23 on such a sample leads to the lattice value of the background topological charge @xmath24 .
as described above @xcite , this result must be renormalized to match the continuum value @xmath20 , @xmath25 where the subscript @xmath26 indicates that the thermalization is achieved within the sector of charge @xmath20 . therefore ,
knowing @xmath20 from the classical configuration and determining @xmath24 from the measurement of @xmath23 on the sample , @xmath0 can be extracted .
if @xmath27 then measuring @xmath28 on the sample leads to the value @xmath29 which is precisely @xmath10 , as indicated above .
@xmath10 can also be calculated on samples with nonvanishing topological background @xmath20 .
following eq.([chil ] ) and eq.([qlzq ] ) the additive constant is extracted in this case by using the relation @xmath30 and leads to the same results @xcite .
a sample of configurations belonging to the sector of total topological charge @xmath20 is obtained in the following way .
one starts from a classical configuration with topological charge @xmath20 .
it can be easily obtained either by using cooled configurations where the energy and the topological charge correspond to the presence of one instanton ( if @xmath31 ) . ] or by setting all gauge links to unity ( if @xmath27 ) .
then a few updating steps are applied and the proper operator ( either @xmath23 or @xmath28 ) is measured at each step . moreover at each step the background topological charge is checked by cooling @xcite in order to verify that the configuration still lies in the sector of charge @xmath20 @xcite . when the result of the measurement stabilizes ( data display a plateau ) , while @xmath20 stays fixed , we consider that the uv fluctuations are thermalized and @xmath32 and @xmath33 yield @xmath0 and @xmath10 respectively .
when the cooling applied to a configuration reveals that its background topological charge is no longer equal to @xmath20 then the configuration is discarded from the sample .
however it might also happen the following event : a possible new instanton or anti
instanton created by the various updating steps might evade the cooling probe because the very cooling procedure could destroy it ( this may happen especially when the instanton spans a few lattice spacings ) . in this case
we would include in the sample a configuration which actually does not belong to the sector of topological charge @xmath20 .
such an event would obviously distort the measurement of any of the above topological operators .
for example , it yields an overestimation of @xmath34 because any added instanton or anti instanton only increments the value of the square .
on the other hand since the theory predilects the sector of vanishing topological charge , the updating steps during the calculation of @xmath0 tend to bring the configuration to that sector either by destroying the original instanton or by creating from scratch an anti instanton . as a consequence
the value of @xmath32 becomes underestimated because on average the expectation value of @xmath23 in the sector @xmath27 is less than in the sector with @xmath31 . in the past
we have always been aware of that potential problem and in this paper we present a study where the background topological sector is calibrated by another method in order to compare the results and detect any difference in the form of a systematic error .
the new method is the counting of zero modes by using ginsparg
wilson based operators @xcite and it will be described in the next section .
following the lines described in the above section , we have calculated the values of the renormalization constants in eq.([chil ] ) for the 1smeared topological charge operator defined as @xcite @xmath35 where all link matrices have been substituted by 1smeared links @xcite ( the smearing parameter was @xmath36 ) .
the corresponding renormalization constants will be called @xmath37 and @xmath38 to denote the level of smearing . in the above expression @xmath39
is the plaquette in the @xmath40 plane with the four corners at @xmath41 , @xmath42 , @xmath43 , @xmath44 ( counter clockwise path ) .
links pointing to negative directions mean @xmath45 .
the generalized completely antisymmetric tensor is defined by @xmath46 and @xmath47 .
the calculation was performed for the yang
mills theory with su(3 ) gauge group and wilson action @xcite .
the lattice size was @xmath48 and the bare coupling @xmath49 .
the practical procedure was the following : starting from a classical configuration with the adequate topological background content ( @xmath27 for @xmath37 or @xmath31 for @xmath38 ) , 80 heat bath ( hb ) steps were applied , each step consisting of three cabibbo
marinari @xcite hits , one for each su(2 ) subgroup and kennedy pendleton algorithm @xcite for refreshing the dynamical variables .
the operator ( @xmath50 for @xmath37 or @xmath23 for @xmath38 ) was measured every 4 steps .
this set of 20 measurements is called `` trajectory '' .
a test of the topological sector was accomplished after each measurement on a separate copy of the configuration .
we accepted only those measurements that were thermalized within the corresponding topological sector .
this condition means that the data must have stabilized to a plateau and that the test must have revealed that the configuration lay within the correct topological sector .
the average over all accepted measurements yielded @xmath37 or @xmath38 .
the test was performed by two different methods : either by a traditional cooling @xcite or by counting fermionic zero modes ( cfzm ) .
this last method consists in calculating the net number of zero modes , @xmath51 by enumerating the level crossings in the spectrum of the wilson
dirac operator @xmath52 as the fermion mass @xmath53 is varied @xcite .
this method was utilized in refs .
@xcite to calculate the topological charge .
the main advantage of the cfzm method against cooling is that the first one does not need to modify the configuration to which it is applied , hence the topological content is never altered during the test . on the contrary
its main disadvantage is that its implementation is heavily time consuming .
let us discuss in more detail the cfzm method .
we can stop counting crossings at any mass @xmath54 inside the allowed interval @xmath55 .
thus , in general the measurement of the topological charge will depend on @xmath56 because there can be level crossings all along the interval of masses where the gap is closed .
this makes the zero mode counting method to look ambiguous .
however in ref .
@xcite it is shown that such a dependence is rather mild for @xmath57 as long as @xmath58 . in the present work we have sought crossings up to three different values for the stopping mass : @xmath59 , @xmath60 and @xmath61 .
in particular the last ( and largest possible ) value looks particularly interesting because instantons representing crossings close to @xmath62 span a size of a few lattice spacings @xcite , i.e. : they are the instantons that most likely could evade the cooling test .
the search for crossings was realized by following the same procedure of ref .
an accelerated conjugate gradient algorithm @xcite was employed to extract the lowest eigenvalues of the wilson
dirac operator . the topological charge after the cooling test was required to be equal to 1 or 0 within a tolerance of @xmath63 .
we usually chose @xmath64 although tests with other values were performed ( see later ) proving that the results are very robust against variability of @xmath63 . -2
cm the results for @xmath37 are shown in fig . [ mtotal ] .
we have calculated 1380 trajectories of 80 hb steps and measured the operator @xmath28 every 4 steps after checking that the background charge was zero . for each measurement , in the figure we show the average over all the results .
actually we show the data after the 16@xmath65 step because measurements after too few hb steps are irrelevant as the configuration is surely not yet thermalized .
a plateau seems to set in after approximately 40 steps which indicates that thermalization has been achieved .
then the value of @xmath37 can be extracted by averaging over the data after the 40@xmath65 step .
the error bar was calculated in the following manner : each trajectory was treated as a single data by averaging all accepted ( i.e. : all thermalized and correctly calibrated ) points in it . since
separate trajectories are prepared by independent monte carlo runs , this procedure guarantees the absence of autocorrelations . then the average and the error are easily obtained from this set of 1380 data .
furthermore , due to the fact that different trajectories can contain a variable number of accepted points , each trajectory must be weighted by a factor proportional to the number of accepted points in it . -2
cm if the number of accepted points in the trajectory @xmath66 is @xmath67 and if @xmath68 is the average over all accepted measurements @xmath53 in this trajectory , @xmath69 then @xmath70 fig .
[ mtotal ] displays the results for all four methods of calibration , cooling with @xmath64 and cfzm with three different values for the stopping mass .
the four white symbols on the left side of the figure are the corresponding results for @xmath37 .
they become smaller as the value of @xmath56 is incremented .
this effect is possibly an indication of a systematic error which however has little effect in the calculation since all results look compatible with each other within errors .
it must be stressed that the statistical errors , represented by the error bars in the figure , amount to about 3% which is rather small . in fig .
[ mq0 ] we show , for each measurement , the number of trajectories for which the calibration test gave an acceptable result , @xmath27 .
the plot is shown for all measurements , thermalized or not , from the 20@xmath65 hb step onward .
the maximum possible number is obviously 1380 and it falls off as the amount of hb steps is increased .
the decrease is steeper for the cfzm with larger stopping mass . in the latest step and for the cfzm with @xmath71 about
30% of all trajectories have varied the topological sector , while for the cooling method the analogous percentage is about 17% . such a difference still allows to obtain results for @xmath37 that are compatible within ( small ) errors , as shown by the white symbols in fig .
[ mtotal ] .
the values of @xmath37 obtained by using cooling with varying @xmath63 are indistinguishable from each other ( white circle in fig .
[ mtotal ] ) for @xmath63 ranging from 0.1 to 0.5 . in fig .
[ mcoolq0 ] three tolerance parameters for cooling are compared indicating that the three tests are almost completely equivalent . again the maximum possible number of trajectories with the right background topological charge is 1380 and it diminishes as the amount of hb steps is increased .
this number became 1136 for @xmath72 at the latest step .
-2 cm -2 cm fig .
[ zhator ] displays the analogous study of fig .
[ mtotal ] for the calculation of @xmath38 .
an instanton with topological charge ( as measured with @xmath23 after cooling ) @xmath73 was employed .
the procedure for the calculation of @xmath38 resembles very much that of the @xmath37 .
we prepared 840 independent trajectories and again the calibration was performed by four different tests , as indicated in fig .
[ zhator ] .
the average per trajectory is @xmath68 , @xmath74 and the result for @xmath38 is @xmath75 a plateau sets in at about the 20@xmath65 hb step .
figures similar to fig .
[ mq0 ] and [ mcoolq0 ] are obtained with analogous conclusions .
again the coincident results ( white symbols in fig . [ zhator ] ) indicate that all calibration methods are equivalent in such a way that any systematic error implicit in our method has negligible consequences within our precision ( statistical errors in fig .
[ zhator ] are about 4% ) . ' '' '' @xmath76 & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 + ' '' '' 6.00 & 0.373(20 ) & 0.383(15 ) & 0.370(20 ) & 0.365(20 ) & 0.390(12 ) + ' '' '' 6.20 & 0.432(6 ) & & & & 0.441(5 ) + ' '' '' 6.50 & 0.503(11 ) & & & & 0.506(5 ) + [ z1 ] ' '' '' @xmath76 & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 + ' '' '' 6.00 & 0.500(12 ) & 0.503(13 ) & 0.495(15 ) & 0.487(20 ) & 0.499(16 ) + ' '' '' 6.20 & 0.560(6 ) & & & & 0.569(4 ) + ' '' '' 6.50 & 0.620(10 ) & & & & 0.631(3 ) + [ z2 ] it is well known that the total topological charge of isolated classical instantons in general is not equal to integer numbers when it is calculated with the operator of eq.([qlattice ] ) .
the difference between the value of the charge and the closest corresponding integer ( @xmath82 in our case ) becomes negligible when the inequalities @xmath83 hold ( @xmath84 being the instanton size and @xmath85 the lattice size ) .
a simple calculation shows that the value of @xmath20 for a discretized instanton in a volume @xmath86 is given by @xmath87 while the discretization error is totally negligible if @xmath88 .
the ratio @xmath89 can be estimated by looking at the action density distribution .
we expect that the infrared effect described in eq.([approxq ] ) cancels out if we divide the charge after heating by the initial value @xmath20 as indicated in eq.([qlzq ] ) .
this fact was carefully checked in @xcite for the 2d @xmath90 nonlinear sigma model .
it was also used in eq.([z997 ] ) although in that case and within our errors , the @xmath73 in the denominator is indistinguishable from 1 . in the present study
we have verified it for our gauge theory : in table [ z1 ] and [ z2 ] the multiplicative constant is calculated at several values of the gauge bare coupling @xmath21 on a @xmath91 volume starting from various initial instantons for the 1 and 2smeared operators ( they are constructed as in eq.([qlattice ] ) after substituting all links with 1 and 2smeared links respectively @xcite ) . a number of trajectories ranging from 200 to 500 were used . notice that within errors the values for @xmath38 and @xmath92 display a dependence on @xmath21 but not on @xmath20 as long as data are divided by the initial value @xmath20 , as described above following eq.([qlzq ] ) .
if instead data were divided by the integer closest to @xmath20 then the results for @xmath93 , @xmath94 would display a fake dependence also on @xmath20 .
this fact is seen in fig .
[ z2smear ] where the values of @xmath92 are displayed as a function of @xmath20 after dividing by @xmath20 ( squares ) or by the closest integer to @xmath20 ( triangles ) : only the squares show constancy with respect to @xmath20 . within our statistical errors
the systematic error is negligible down to @xmath95 .
possibly if the statistics were increased this limit could grow . in any case to calculate @xmath0 it is better to make use of instantons with @xmath20 as close as possible to an integer .
we have studied two possible sources of systematic errors in the nonperturbative determination of the renormalization constants for the evaluation of the topological susceptibility on the lattice , @xmath0 and @xmath10 ( see eq.([chil ] ) ) : * _ i ) _ the cooling test of the background topological charge along the heating process could yield a wrong information since the cooling procedure could delete the unwanted instanton or anti
instanton created by the heating steps .
an independent check of the cooling test has been performed by applying calibration methods based on the counting of fermionic zero modes .
no sizeable systematic effects have been observed within our ( rather small ) statistical errors ( 3% for @xmath10 and 4% for @xmath0 ) . * _ ii ) _ in the calculation of @xmath0 for the operator @xmath23 of eq.([qlattice ] ) the simulation must be started from a configuration with a topological charge @xmath20 different from zero . in the infinite volume limit @xmath20 takes on integer values , however on the lattice , the determination of @xmath20 by using the operator in eq.([qlattice ] ) , leads to results which in general are close to but not strictly equal to integers .
we have argued that this is a potential source of error that can affect the calculation of @xmath0 .
however it can be kept under control if one uses eq.([qlzq ] ) to extract @xmath0 with @xmath20 not rounded to its closest integer . in conclusion
one can safely use the field theoretical method to study topology on the lattice since any possible systematic error of the method is well under control .
moreover the method is much less demanding in computer time than the ginsparg
wilson based method .
the apemille facility in pisa was used for part of the runs .
we thank m.i.u.r . for financial support , project number 2004021808_004 .
x e. witten , nucl .
b156 ( 1979 ) 269 . g. veneziano , nucl . phys .
b159 ( 1979 ) 213 .
b. alls , m. delia , a. di giacomo , phys .
d71 ( 2005 ) 034503 .
b. lucini , m. teper , u. wenger , nucl .
b715 ( 2005 ) 461 . l. del debbio , l. giusti , c. pica , phys . rev .
94 ( 2005 ) 032003 .
r. j. crewther , riv .
nuovo cimento 2 ( 1979 ) 8 . m. campostrini , a. di giacomo , h. panagopoulos , e. vicari , nucl . phys .
b329 ( 1990 ) 683 .
m. campostrini , a. di giacomo , h. panagopoulos , phys .
b212 ( 1988 ) 206 .
b. alls , e. vicari , phys .
b268 ( 1991 ) 241 .
d. espriu , r. tarrach , z. phys .
c16 ( 1982 ) 77 .
b. alls , a. di giacomo , h. panagopoulos , e. vicari , phys .
b350 ( 1995 ) 70 .
k. osterwalder in `` constructive field theory '' , lecture notes in physics , n@xmath96 25 , 1973 , edited by g. velo , a. s. wightman .
e. seiler , i. o. stamatescu , preprint mpi - pae / pth 10/87 ( 1987 ) unpublished .
p. menotti , a. pelissetto , nucl .
b ( proc . suppl . ) 4 ( 1988 ) 644 .
b. alls , m. delia , a. di giacomo , r. kirchner , phys .
d58 ( 1998 ) 114506 .
i. horvth et al . , phys .
lett . b617 ( 2005 ) 49 .
ilgenfritz , k. koller , y. koma , g. schierholz , t. streuer , v. weinberg , hep lat/0512005 .
a. di giacomo , e. vicari , phys .
b275 ( 1992 ) 429 .
b. alls , m. campostrini , a. di giacomo , y. gnd , e. vicari , phys .
d48 ( 1993 ) 2284 .
b. alls , g. boyd , m. delia , a. di giacomo , e. vicari , phys .
b389 ( 1996 ) 107 .
b. alls et al . , phys .
d58 ( 1998 ) 071503
. l. del debbio , h. panagopoulos , p. rossi , e. vicari , phys .
d65 ( 2002 ) 021501 .
m. delia , nucl .
b661 ( 2003 ) 139 .
a. a. belavin , a. m. polyakov , a. s. schwartz , yu .
s. tyupkin , phys .
b59 ( 1975 ) 85 . i. a. fox , m. l. laursen , g. schierholz , j. p. gilchrist , m. gckeler , phys .
b158 ( 1985 ) 332 .
m. teper , phys .
b171 ( 1986 ) 81 ; phys .
b171 ( 1986 ) 86 .
b. alls , l. cosmai , m. delia , a. papa , phys .
d62 ( 2000 ) 094507 . f. farchioni , a. papa , nucl .
b431 ( 1994 ) 686 .
h. ginsparg , k. g. wilson , phys .
d25 ( 1982 ) 2649 .
p. di vecchia , k. fabricius , g. c. rossi , g. veneziano , nucl .
b192 ( 1981 ) 392 ; phys . lett .
b108 ( 1982 ) 323 . c. christou , a. di giacomo , h. panagopoulos , e. vicari , phys .
d53 ( 1996 ) 2619 . k. g. wilson , phys . rev .
d10 ( 1974 ) 2445 .
n. cabibbo , e. marinari , phys .
b119 ( 1982 ) 387 .
a. d. kennedy , b. j. pendleton , phys .
b156 ( 1985 ) 393 .
r. narayanan , h. neuberger , nucl .
b443 ( 1995 ) 305 .
h. neuberger , phys .
d61 ( 2000 ) 085015 .
r. g. edwards , u. m. heller , r. narayanan , nucl . phys . b535
( 1998 ) 403 .
l. del debbio , c. pica , jhep 02 ( 2004 ) 003 .
b. bunk , k. jansen , m. lscher , h. simma , `` conjugate gradient algorithm to compute the low lying eigenvalues of the dirac operator in lattice qcd '' , unpublished report ( 1994 ) ; t. kalkreuter , h. simma , comput .
93 ( 1996 ) 33 .
b. alls , m. beccaria , f. farchioni , phys .
d54 ( 1996 ) 1044 . | a ginsparg wilson based calibration of the topological charge is used to calculate the renormalization constants which appear in the field theoretical determination of the topological susceptibility on the lattice .
a systematic comparison is made with calculations based on cooling .
the two methods agree within present statistical errors ( 3%4% ) .
we also discuss the independence of the multiplicative renormalization constant @xmath0 from the background topological charge used to determine it . |
the present paper is part of the local cosmology from isolated dwarfs ( lcid ) project .
the hubble space telescope ( hst ) has been used to obtain deep photometry of six isolated dwarf galaxies in the local group : ic 1613 , leo a , cetus , tucana , and phoenix .
color - magnitude diagrams ( cmd ) reaching the oldest turn - off points have been derived .
the advanced camera for surveys ( acs , * ? ? ?
* ) has been used to observe the first five , while phoenix was observed with the wide field and planetary camera-2 ( wfpc2 ) .
the main goal of the lcid project is to derive the star formation history ( sfh ) , age - metallicity relation ( amr ) , and stellar population gradients of this sample of galaxies .
our objective is to study their evolution at early epochs and to probe effects of cosmological processes , such as the cosmic uv background subsequent to the onset of star formation in the universe or physical processes such as the gas removal by supernovae ( sne ) feedback .
our sample consists of field dwarfs which were chosen in an effort to study systems as free as possible from environmental effects due to strong interactions with a host , massive galaxy .
the sfh is a powerful tool to derive fundamental properties of these systems and their evolution , but to study the earliest epochs of star formation , deep cmds , reaching the oldest main sequence turn - offs , are required . here
we present our analysis of .
is a dirr / dsph low - luminosity galaxy discovered by @xcite . from its location
, it appears to be a member of the andromeda subgroup of the local group @xcite .
the distance to the galaxy has been estimated by several authors .
@xcite , @xcite , @xcite , and @xcite used the tip of the red giant branch ( trgb ) to estimate distances ranging from 0.6 to 0.96 mpc .
this measurement is inherently challenging because the trgb region of is very sparsely populated .
@xcite also used the trgb but complemented with observations of the horizontal branch ( hb ) , the red clump ( rc ) , and the overall match of the cmd to model cmds to estimate a distance of @xmath11 mpc .
finally , from a study of the rr - lyr stars in lgs-3 , @xcite derived a distance of @xmath12 mpc ( @xmath13 mag )
. we will assume this value in this paper .
the stellar content of has been studied by @xcite , @xcite , @xcite , and @xcite .
@xcite discovered the first cepheid variable candidates in this galaxy .
these works show that contains stars of all ages ; it has been forming stars since at least 12 gyr ago and shows a low recent star formation rate ( sfr ) @xcite .
the sfh shows a dependency with radius @xcite .
no regions have been found in the galaxy @xcite , which is consistent with the low recent star formation activity .
the metallicity of the stars has been estimated from the color of the red giant branch ( rgb ) . in this way
, @xcite obtained a mean metallicity of @xmath14=-2.10\pm 0.22 $ ] , while @xcite derived a metallicity range @xmath15\leq -1.0 $ ] .
@xcite derived @xmath14=-1.5\pm 0.3 $ ] for the stars older than 8 gyr , and @xmath14\sim -1 $ ] for the most recent generation of stars , the hi content of was first studied with single dish spectra by @xcite and , more recently , with interferometry by @xcite . using the results by the latter authors and the adopted distance of 0.65 mpc ,
the results from @xcite correspond to a total hi mass of @xmath16 m@xmath8 , which corresponds to a total atomic gas mass of @xmath17 m@xmath8 when corrected for helium .
they also measured a gas velocity dispersion of 8 km s@xmath9 and negligible rotation at a galactocentric radius of 470 pc .
the corresponding virial mass estimate is @xmath18 m@xmath8 . in this paper
, we present the sfh of obtained from observations with the acs on the hst .
the photometry reaches the oldest main sequence turn - offs of the galaxy , allowing us to obtain an accurate sfh even for the oldest stellar populations .
the structure of the paper is as follows : in [ secred ] the observations and data reduction are discussed , the cmd is presented , and the completeness tests are described . in [ secsfhmethod ]
we describe many of the details of the iac - pop / minniac methodology and its application to the lcid program observations .
the derived sfh of is presented in [ secsfhresult ] and
is compared with those of other lcid galaxies in [ seclcid ] .
the consequences of sne driven winds and cosmic reionization on the sfh of are discussed in [ seccosmo ] .
the main conclusions of the work are summarized in [ seccon ] . finally , in appendix [ secimf ] , the uniqueness , self - consistency , stability , and statistical significance of the solution are discussed .
the acs observations of were obtained between september 12 and 17 , 2005 .
the f475w and f814w bands were selected as the most efficient combination to trace age differences at old ages , since they provide the smallest relative error in age and metallicity in the main - sequence and sub - giant regions .
total integration times were 15072 s in f475w and 13824 s in f814w .
the observations were organized in 6 visits of 2 orbits each , and each orbit was split into one f475w and one f814w exposure . in the first orbit of each visit , exposure times were 1250 and 1147 s respectively while in the second orbit they were 1262 and 1157 s respectively .
dithers of a few pixels between exposures were introduced to minimize the impact of pixel - to - pixel sensitivity variations ( `` hot pixels '' ) in the ccds .
the observed field of is shown in figure [ f1 ] .
the core radius of is 082 and its optical scale length is 078 @xcite , so the 34 @xmath19 34 format of the acs covers a substantial fraction of the stars in .
table [ logobs ] gives a summary of the observations .
we analyzed the images taken directly from the stsci pipeline ( bias , flat - field , and image distortion corrected ) .
two psf - fitting photometry packages , daophot / allframe @xcite and dolphot @xcite , were used independently to obtain the photometry of the resolved stars .
non - stellar objects and stars with discrepant large uncertainties were rejected based on estimations of profile sharpness and goodness of fit .
the final photometric lists contain @xmath20 stars each .
see @xcite for more details about the photometry reduction procedures .
individual photometry catalogs were calibrated using the equations provided by @xcite .
the zero - point differences between the two sets of photometry are @xmath21 for f475w - f814w and @xmath22 for f814w .
these small differences are typical for obtaining hst photometry with different methods @xcite . in [ secsfhmethod ]
we discuss a technique which minimizes the already small impact of these differences in the obtained sfhs .
the cmd of is shown in figure [ f2 ] for the dolphot photometry .
individual stars are plotted in the left panel and density levels are shown in the right panel .
the left axis shows magnitudes in the acs photometric system corrected for extinction .
absolute magnitudes are given on the right axis using the adopted values for the distance modulus ( @xmath23 ) and extinctions ( @xmath24 and @xmath25 mag ) @xcite .
in the left panel , the lines across the bottom show the 0.25 , 0.50 , 0.75 , and 0.90 completeness levels derived from false star tests ( discussed later in this section ) . finally , in order to highlight the main features of the cmd , three isochrones from the basti stellar evolution library @xcite are also shown in the right panel .
the cmd of shown in figure [ f2 ] reaches @xmath26 mag below the oldest main - sequence turn - off ; these observations are @xmath27 mag fainter than the deepest cmd previously obtained for this galaxy @xcite .
a comparison of the observations with the over - plotted isochrones and the presence of an extended hb indicate that a very old , very low metallicity stellar population is present in the galaxy .
the gap produced in the hb by the rr - lyrae variables is visible as well as a narrow rgb .
the latter is populated by stars all older than 1 - 2 gyr and its narrowness is an indication of a low spread in metallicity .
the rc can be observed at the red end of the hb . a ms with stars younger than 200 myr is clearly apparent .
some blue and red core helium - burning stars might be also present above the rc ( @xmath28 ; @xmath29 ) .
the rgb bump can be also observed at @xmath30 and @xmath31 mag .
this evolutionary feature may be used as indicator of a dominant coeval stellar population homogeneous in metallicity @xcite .
signal - to - noise limitations , detector defects , and stellar crowding can all impact the quality of the photometry of resolved stars with the resulting loss of stars , changes in measured stellar colors and magnitudes , and systematic uncertainties . to characterize these observational effects ,
we have used the standard technique of injecting a list of false stars in the observed images and obtaining their photometry in an identical manner as for the real stars .
these observational effects must be simulated in the stars of the global , synthetic cmd ( scmd ) ( hereafter _ synthetic _ stars ) to be used for the analysis of the sfh ( see below ) .
@xcite provide a detailed description of the procedures we adopt for the characterization and simulation of these observational effects .
for the present case of , @xmath32 false stars were used for the dolphot photometry and @xmath33 were used for the daophot photometry .
the completeness factor , @xmath34 , is defined as the rate of recovered to injected stars as a function of magnitude and color .
figure [ f2 ] , right panel , shows @xmath35 and 0.90 for the dolphot photometry .
results for daophot are similar . at the magnitude level of the oldest ms turn - off stars , @xmath36 , allowing for very strong constraints on the oldest epochs of star formation .
we simulate the observational effects in the stars of the scmd with the routine _ obsersin_. it uses the unrecovered false stars and the differences between the injected ( @xmath37 ) and measured magnitudes ( @xmath38 ) of the recovered stars for the simulation @xcite . since the density of stars ( and thus the amount of crowding ) normally varies substantially across the galaxy s field , spatial information has also been taken into account in the completeness test .
to do so , first the spatial density distribution of observed stars in is measured . for that , a grid of @xmath39 pixel size bins is defined , stars are counted in each bin and the result is smoothed and normalized .
this is used as a density probability distribution to assign positions to each synthetic star .
then , for each synthetic star with magnitude @xmath40 , a list of the false stars with @xmath41 ; @xmath42 is created for each filter , where @xmath43 is the spatial distance between synthetic and false star and @xmath44 and @xmath45 free input searching intervals . if the number of false stars selected in this way , in common to both filters , is @xmath46 , the parameters @xmath44 and @xmath45 are increased in turn until 10 or more false stars are found .
then , one of them is randomly selected from this list .
if that star was unrecovered , the synthetic star is eliminated from the scmd .
if the selected star was recovered and @xmath47 and @xmath48 are its injected and recovered magnitudes in a given filter , then @xmath49 will be the magnitude of the synthetic star with observational effects simulated .
the same is done for both filters .
our analysis of the sfh of and the rest of lcid galaxies is done upon two fundamental requirements : i ) the consistency of the method , in order to obtain sfhs that are readily comparable and ii ) the stability and uniqueness of the solution . to accomplish i ) , we have adopted the same criterion on the selection of input parameters to derive the sfh for all galaxies of the lcid project . for ii ) , we have further developed our method and created new tests and control procedures .
details are given in the following and appendix [ secimf ] .
the iac method to solve the sfh is based in three main codes : ( i ) iac - star @xcite is used for the computation of synthetic stellar populations and cmds ; ( ii ) iac - pop @xcite is the core algorithm for the computation of sfh solutions , and ( iii ) minniac ( introduced in * ? ? ?
* but explained in more detail in this paper ) is a suite of routines created to manage the process of sampling the parameter space , creating input data , and averaging solutions . as a summary , figure [ f3 ] shows a data flow diagram for the sfh computation process .
considering that time and metallicity are the most important variables in the problem ( see * ? ? ?
* ) , we define the sfh as an explicit function of both .
we will use two definitions , one in terms of number of formed stars and another in terms of gas mass converted into stars .
formally , @xmath50 is the number of stars formed within the time interval @xmath51 $ ] and within the metallicity interval @xmath52 $ ] , while @xmath53 is the mass converted into stars within the same intervals .
the relation between @xmath54 and @xmath55 is the average stellar mass at birth .
both can be identified with the usual definitions of the sfr , but as a function of time and metallicity .
several criteria can be used to grid the cmd .
the simplest is to use a uniform grid and count stars within each box .
however , not all the regions of the cmd and stellar evolutionary phases contain information of the same quality for the purpose of the sfh derivation .
an alternative is to give higher weights to regions with more certain stellar evolutionary phases ( like the main sequence or the sub - giant branch ) , regions which are more heavily populated , or regions in which the photometry is more accurate .
to do so , we have introduced what we call _ bundles _ ( see * ? ? ?
* and figure [ f4 ] ) , that is , macro - regions on the cmd that can be subdivided into boxes using different appropriate samplings .
the number and sizes of boxes in each bundle determines the weight of that region in the derived sfh .
this is an efficient and flexible approach for two reasons .
first , it allows a finer sampling in those regions of the cmd where the models are less affected by uncertainties in the input physics .
second , the box size can be increased , and the impact on the solution decreased , in those regions where the small number of observed stars could introduce noise due to small number statistics or where we are less confident in the stellar evolution predictions .
iac - pop provides a solution for the input set of data and parameters .
it can also provide several solutions for the same set by changing the initial random number generator seed .
alternatively , it can deliver several solutions by randomly changing the input data @xmath56 ( the number of stars in the original , ocmd , boxes ) according to a poisson statistics .
however all of these solutions are eventually sensitive to several other input functions and parameters . to better analyze their effects
, we can divide them into four groups .
the first group consists of _ sampling parameters _ ; the main ones are the number and boundaries of parent , pcmds , and the criteria to define the bundles and grid arrays for sampling the cmds .
the second group consists of _ external parameters _ ; these are the photometric zero points , distance modulus , and reddening .
the third group consists of _ model functions _ , which are those functions defining the scmd ; the main ones are the initial mass function ( imf ) , @xmath57 , and the function accounting for the frequency and relative mass distribution of binary stars , @xmath58 ( see * ? ?
finally , we have the _ external libraries _ , which consists of stellar evolution and bolometric correction libraries .
limiting and measuring the effects of all these parameters in the sfh solution requires several thousand runs of iac - pop using different input sets .
minniac is used to test for the effects of the four groups are treated separately . in general , the solutions obtained with varying _ sampling parameters _ are averaged ; their rms deviation is also a good estimate of the solution internal error . as for the _ external parameters _ and _ model functions _ , the set with the best @xmath59 is usually adopted .
regarding the _ external libraries _ , the solution with the best @xmath59 may be adopted , but all the solutions are usually conserved ; their differences can provide an indication of the external uncertainties associated to such libraries .
minniac ( from _ minn_esota and _ iac _ ) is a suite of routines developed specifically to manage the process of selection of sampling parameters , creating input data sets , and averaging solutions .
more in detail , regarding the parameter selection and data input , minniac is used for two tasks .
first , to divide the scmd in the corresponding pcmd according to the selected age and metallicity bins .
second , to define the bundles and box grids in the cmds and counting stars in them . with this information
, minniac creates the input files for each iac - pop run .
several selections of both simple populations and bundle grids are used within each sfh solution process .
minniac does this task automatically , requiring only the input of bundles and corresponding grid sizes , age and metallicity bin sizes and the steps by which the starting points of the cmd grid as well as age and metallicity bins should be modified from one iac - pop run to another .
a second module of minniac is used to average single solutions found by iac - pop as required . in this section
we will describe with some details the specific application of the iac method to the case of , detailing the different effects mentioned in
[ secsfhmethod ] : _ sampling parameters _ , _ external parameters _ , _ model functions _ and _ external libraries_. our choices are summarized in table [ varpar ] . for each item , columns 2 , 3 , and 4 give minimum ( or choice 1 ) , maximum ( or choice 2 ) and the step used ( if relevant ) , respectively . column 5 gives the final adopted value ( if relevant ) .
the sfh of has been derived independently for the dolphot and daophot photometry sets following the same procedure as described below .
using iac - star , several scmds were computed , one with each of the imfs and binarity functions listed in table [ varpar ] .
each scmd was built with @xmath60 stars with a constant star formation rate in the range of age 013.5 gyr and a metallicity range 0.00010.005 .
the latter value has been assumed after a preliminary test with isochrones which showed that stars with metallicity larger than 0.005 do not exist in .
using a large number of synthetic stars in the scmd is important in order to minimize the stochastic effects of the monte carlo simulations used in iac - star ; to have a statistically significant number of stars in each pcmd ( @xmath61 stars ) , and to reduce the uncertainties in the stellar composition of simple populations in the pcmds .
the bundles and one of the grid sets used to bin the cmds are shown in figure.[f4 ] . in total , 72 combinations of imf and binarity ( _ model functions _ ) have been tested . for each of them ,
24 solutions have been obtained by varying the cmd binning and the simple stellar population sampling ( _ sampling parameters _ ) .
the average @xmath62 of these 24 solutions is adopted as the solution for the specific imf and binarity .
these averages have been obtained by using a boxcar of width 1 gyr and step 0.1 gyr for @xmath63 and width @xmath64 and step @xmath65 for @xmath66 .
a @xmath67 is also obtained as the average of the 24 single @xmath68 values .
the average solution with the smallest @xmath67 is assumed as the best one .
the imf and the binary fraction corresponding to this solution are assumed here as the ones best representing our photometry data , although we do not intend to reach any strong conclusions concerning their true values . to minimize the effects of _ external parameters _
( distance , reddening and photometry zero - point ) , the sfh is derived for different offsets in color and magnitude applied to the ocmd .
the offset giving the minimum @xmath67 is assumed as the best .
these are obtained when the ocmd are shifted by @xmath69f475w@xmath70f814w@xmath71f814w@xmath72=[0.02,0.11]$ ] and @xmath73 $ ] for the dolphot and daophot photometry sets , respectively .
these tests also provide a way to estimate how distance , reddening and zero - point uncertainties translate into the solution of the sfh . defining @xmath74 as the @xmath67 value of the best solution ,
if @xmath75 is its corresponding rms deviation , all the solutions with @xmath76 should be considered valid .
we can assume that the rms deviation of all the corresponding single solutions is a proxy of the error of the adopted solution , @xmath77 .
the former error , @xmath77 , includes the effects of _ sampling _ and _ external parameters_. however , there is one more source of uncertainty that must be taken into account , namely the effect of statistical sampling in the ocmd . to estimate this ,
the solution is again obtained after varying the numbers of stars in the ocmd grid according to a poisson statistic .
this is done 20 times and the rms deviation of the solutions , @xmath78 , is obtained .
the final adopted error is the sum in quadrature @xmath79 .
finally , solutions have been obtained using both the basti @xcite and the @xcite stellar evolutionary libraries ( _ external libraries _ ) . in both cases the bolometric corrections by @xcite were used .
both produced qualitatively similar results and a comparison between them together with the results obtained with other methods is presented in [ comresult ] . for detailed analysis presented in the following sections , for simplicity , we use only the basti stellar evolutionary libraries .
figure [ f5f6 ] shows plots of the sfh of , @xmath55 , as a function of both @xmath63 and @xmath66 derived from the two different photometry sets . @xmath80 and @xmath81
are also shown in the @xmath82 and @xmath83 planes , respectively .
the projection of @xmath55 on the age - metallicity plane shows the amr , including the metallicity dispersion . in figure [ f7 ]
we show @xmath80 , the amr , and the cumulative mass fraction of , as a function of time with their associated errors .
three vertical dashed lines show the ages of the 10th , 50th , and 95th - percentile of the integral of @xmath80 .
table [ meansolpar ] gives a summary of the integrated and mean quantities for the observed area and estimates for the whole galaxy .
the main features of the sfh of can be observed in figure [ f7 ] .
the bulk of the star formation of occurred at old epochs , with a main peak located @xmath84 gyr ago .
the initial episode of sf lasted until roughly 9 gyr ago .
approximately 85% of the stars were formed in this initial episode . since then
, the star formation rate has been very low and slowly declining until the present .
the sfr at the measured main peak is a factor of 15 larger than the mean sfr between 0 and 9 gyr ago ( because of limited time resolution , this is a lower limit , see [ secconstraints ] ) .
it is worth noting that the age at which the rate of the amr starts to increase more steeply matches up with the age at which the sfr decreases to a low level .
this may indicate a scenario in which the stars were formed at a higher rate at old epochs from fresh , low metallicity gas . when the initial episode of star formation was over , the gas reservoir would have been nearly exhausted . as a consequence ,
subsequent star formation would occur at a much lower rate and from an ism including a high fraction of processed gas , hence significantly increasing the metallicity enrichment rate .
it is interesting to test whether the number of sne produced in during its lifetime are sufficient to produce the chemical enrichment we have derived .
focusing on the main episode , @xmath85 m@xmath8 of gas were converted into stars from 13.5 to 8.5 gyr ago . assuming a minimum progenitor mass for core collapse sne of 6.5 m@xmath8 @xcite we obtain a total of @xmath86 sne produced in the old episode .
accounting for stellar mass loss , and assuming a minimum initial mass of 10 m@xmath8 for sne progenitors , then the total number of sne decreases to @xmath87 .
this can be compared to an estimate of the total number of sne necessary to produce the observed chemical enrichment . assuming the solar abundance distribution from @xcite , that only sne ii are involved , and using the iron mass production rate from @xcite ( see * ? ? ? * for more details ) , we estimate that @xmath2 80 sne ii are enough to produce an increment of metallicity from 0 to 0.0007 , which is the upper value obtained for 8.5 gyr ago .
this estimate was derived assuming the iron mass production of a 13 @xmath88 star ( which may vary by a factor of 23 with different assumptions ) .
although relatively uncertain , our calculation shows that the galaxy needs to retain only roughly 1% of the material released by sne in order to produce the observed metallicity increase .
the same calculation can be made for the chemical evolution from 8.5 gyr ago to the present . in this case , the mass of gas converted into stars is @xmath89 m@xmath8 and the corresponding numbers of sne are @xmath90 and @xmath91 . in this time , the metallicity increases from @xmath92 to a current upper value of @xmath93 , requiring 290 sne .
this is less than 15% of the total , a much larger fraction than calculated for the initial star formation episode .
this implies that a significantly larger fraction of the newly synthesized material was captured at later times at lower sfrs .
note that only sne ii have been considered for this estimates ; if sne ia were also included , even lower fractions of captured newly synthesized material would result .
note also , that much of the latter enrichment could be achieved by sne ia from the initial star formation episode .
the sfh shown in figure [ f7 ] corresponds to the observed area , which covers the center and a major fraction of the galaxy .
the total mass of stars created in this area over the lifetime of the galaxy is @xmath94 m@xmath8 .
assuming an exponential density distribution of scale - length 112 pc ( s. hidalgo et al .
, in prep . )
, we estimate that the total mass of stars ever formed in is @xmath95 m@xmath8 . adding in the gas mass ( * ? ? ?
* see [ secint ] ) , then the total baryonic mass of is @xmath96 m@xmath8 .
these values correspond to the best solution with the adopted imf ( @xmath97 for the interval @xmath98 m@xmath8 ; @xmath99 for the interval @xmath100 m@xmath8 ; see table [ varpar ] ) .
if a salpeter imf with an exponent of 2.35 for the range 0.1 to 100 m@xmath8 is used , then the result would be larger by a factor of @xmath26 .
figure [ f8 ] shows the observed cmd , the best - fit cmd , and the corresponding residuals .
the calculated cmd has been built using iac - star with the solution sfh of as input .
residuals are given in units of poisson uncertainties , obtained as @xmath101 , where @xmath102 and @xmath103 are the number of stars in bin @xmath104 of an uniform grid defined on the observed and calculated cmds , respectively .
observational uncertainties as well as the uncertainties inherent to the sfh computation procedure result in a smoothing of the derived sfh and a decrease of the age resolution ( see appendix [ stab ] ) . having a accurate estimate of this age resolution
is fundamental in order to obtain reliable information of the actual properties of the galaxy .
we have calculated the age resolution as a function of time by recovering the sfh of a set of mock stellar populations corresponding to very narrow bursts occurring at different times .
five mock populations have been computed , each one with a single burst at ages of 1 , 4 , 8 , 11.5 , and 13 gyr .
we used iac - star to create the cmds associated with each mock stellar population . for the mock populations at ages of 11.5 and 13 gyr ,
five simulations have been done with different random number seeds in order to better characterize them .
the assumed metallicities and sfrs correspond to those obtained for .
observational uncertainties were simulated in the cmds by using the results of the dolphot completeness tests .
the sfhs of the mock stellar populations were then obtained using the identical procedure as for the real data . in all cases ,
the recovered sfh is well fitted by a gaussian profile with @xmath105 depending on the burst age .
figure [ f9 ] shows the input bursts and the corresponding gaussian profile fits to the solutions , including their peak age and @xmath105 . from these results we estimate an intrinsic age resolution of @xmath106 gyr at the oldest ages which decreases to @xmath107 gyr at an age of 1 gyr
also , a slight shift is observed between the input and solution ages . for the oldest trials , the shift is toward younger ages ( up to @xmath108 gyr for the oldest trial ) and for the youngest trials it is toward older ages ( about 0.2 gyr ) .
it is particularly important to determine the oldest age at which star formation occurred in and the real duration of the first , main star formation episode .
our sfh solution ( figure [ f7 ] ) is well fitted by a gaussian profile with a peak , @xmath109 gyr and @xmath110 gyr in the age range 9.513.5 gyr ( using the dolphot photometry ) .
the shifts shown in figure [ f9 ] indicate that the actual star formation peak would have occurred at about 11.7 gyr ago .
a comparison of the solution with our estimated resolution indicates that the initial episode of star formation is just barely resolved in our observations .
a quadratic subtraction of the @xmath0 that we have obtained for a sharp burst at this age from the derived @xmath110 provides a first estimate of the actual , intrinsic dispersion of @xmath111 .
given the importance of this measurement , we have pursued a more detailed analysis .
we have extended the former tests to a set of mock stellar populations with sfhs defined as gaussian profile shaped bursts of different @xmath105 and centered at 11.5 gyr .
observational uncertainties were simulated in the associated cmds and the sfhs were recovered following the same process as for the galaxy .
in all cases , the recovered sfhs are well fitted by gaussian profiles in the age range 9.513.5 gyr .
figure [ f10 ] shows a quadratic fit to the standard deviation of the solutions , @xmath112 , as a function of the input ones , @xmath113 .
if the actual sfh were well described as a gaussian profile , its @xmath105 could be inferred by interpolating in this fit , which gives @xmath114 gyr .
we will adopt this value for the forthcoming discussion .
as for the rest of the lcid galaxies , we have obtained the sfh of using the match @xcite and cole @xcite methods in addition to the iac method .
these two methods use @xcite for the stellar evolution library and we have adopted the dolphot photometry to be used with them for this comparison ( for a description of the main features of these methods and the particulars of their application to lcid galaxies , see * ? ? ?
for a proper comparison , we have also obtained the sfh with iac - star / minniac / iac - pop using the same library and photometry .
this also allows us to compare the solutions obtained with the iac method using two different stellar evolution libraries .
figure [ f11 ] shows the cumulative mass fraction of as obtained with the three tested methods .
the agreement between the three methods resulting in the same cumulative mass fraction ( 0.8 ) at 8 gyr is very good .
this strongly supports the conclusion the majority the star formation in occurred during this initial episode of star formation .
also , all three methods find star formation continuing at much lower rates , until the present .
the similar behavior of the three solutions allows us to be confident that our scientific conclusions are independent of the sfh solution method . in particular , the fact that the galaxy has formed most of the stars after the reionization epoch is also confirmed and even reinforced .
figure [ f11 ] also shows effect of using different stellar models by comparing the iac - star / minniac / iac - pop results using two different stellar libraries .
here we see that the sfh derived using the basti stellar evolution library has a slightly faster rise compared to the sfh derived using the @xcite library . however , otherwise , the two are in very good agreement , and the main features of the sfh of are found in both derived models at roughly equivalent levels .
an analysis of the actual resolution provided by the match and cole methods , of the kind of that done for iac - star / minniac / iac - pop in [ secconstraints ] would be necessary before going deeper in this comparison , which is beyond our objectives here . in the following
we will use only the results obtained with iac - star / minniac / iac - pop and the basti stellar evolution library .
in this section , we will compare the sfh of with that of phoenix ( the other dsph / dirr transition galaxy of the lcid sample , * ? ? ? * ) and with those of cetus and tucana ( the two dsphs of the sample , * ? ? ? * ; * ? ? ? * ) .
figure [ f12 ] shows the sfh and the amr of compared to those of phoenix @xcite .
the correspondence between the main properties of both galaxies is remarkable .
both galaxies show a dominant period of initial star formation at early times , with a result that both galaxies had formed 50% of their total stellar mass about 10.511.0 gyr ago , or before a redshift of 2 .
both galaxies had also formed more than 95% of their total stellar mass by 2 gyr ago .
the amrs of phoenix and show similar behaviors from the earliest times , with @xmath115\sim-1.7 $ ] , until reaching 95% of the cumulative mass fraction , with @xmath115\sim-1.2 $ ] .
note that , in both cases , the slope of the amr increases about @xmath116 gyr ago , after the initial episode of star formation .
the sfhs of both transition type galaxies are similar to those of the two dsphs in the sense that an old , main episode exists and that a sharp drop off in the star formation rate occurred @xmath117 gyr ago . after this
, the transition galaxies , phoenix and , were able to form stars at low , steadily decreasing rates while the dsphs completely stopped their star formation .
@xcite showed that the first , main episode of star formation occurred earlier in tucana than in cetus .
the main episode of star formation has an age comparable to that of cetus and is also clearly younger than that of tucana .
there are significant differences between the amrs of the transition and the dsph galaxies even for @xmath118 gyr .
the metal enrichment from the earliest times up to @xmath117 gyr ago is less than 0.1 dex in and phoenix , while it is more than 0.4 dex in the same period in the dsph galaxies @xcite . in the case of the transition galaxies , more than 80% of the stars
were formed with roughly the same metallicity .
this result is supported by the narrow rgb of as described in [ secred ] , indicative of a low metallicity spread for the intermediate - to - old age stars .
the differences between the amrs of the transition and dsph galaxies is intriguing .
galaxy mass is the fundamental parameter which is thought to determine a galaxy s ability to retain gas .
this might naively imply that the dsphs were initially more massive systems than the transition galaxies .
however , today , cetus , tucana , , and phoenix all have comparable luminosities .
tidal stripping is thought to be very important in the evolution of dsph galaxies ( see * ? ? ?
* and references therein ) .
the comparable present day luminosities , combined with the difference in amrs , could point to significantly larger halo masses for the isolated dsphs which have been lost to tides , or , in other words , a much smaller impact of tides on the evolution of the transition galaxies .
dwarf galaxies are the focus of a major cosmological problem affecting the cdm scenario ; the number of dark matter sub - halos around milky way type galaxies predicted by cdm simulations is an order of magnitude larger than observed @xcite .
several explanations have been proposed to solve this problem but most of them use two main processes that can dramatically affect the formation and evolution of dwarf - sized halos : heating from the ultraviolet ( uv ) radiation arising from cosmic reionization and internal supernovae ( sne ) feedback .
both are , in principle , capable of stopping the star formation in a dwarf halo and even to fully remove its gas .
the detailed information about the sfh at early and intermediate ages , such as we have obtained here , may provide fundamental insight into the problem and help to discriminate between the different scenarios proposed so far .
detailed discussions about this complex problem can be found in @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , and @xcite , among others . here we will give a short summary just to place our forthcoming discussion in context .
feedback from sne can suppress star formation or even blow out the gas completely from the smallest systems .
@xcite presented detailed models showing that the mass ejection efficiency is low in galaxies with baryonic mass above @xmath119 m@xmath8 and that only galaxies with baryonic mass below @xmath120 m@xmath8 could have their gas completely blown away almost independently of the sne mechanical luminosity .
it is worth mentioning that , according to the assumptions of @xcite , a baryonic mass of @xmath120 m@xmath8 would correspond to a total mass of @xmath121 m@xmath8 , while a baryonic mass of @xmath119 m@xmath8 would correspond to @xmath122 m@xmath8 .
there is a consensus that uv background heating establishes a characteristic time - dependent minimum mass ( a _ filtering _ mass ) for halos that can accrete gas ( e.g. , * ? ? ?
* ; * ? ? ?
. the uv background will heat the gas in low mass halos , preventing star formation , and eventually photo - evaporate them . as a result , halos not massive enough to have accreted gas before the reionization epoch would fail to do so later , unless they accrete mass faster than the rate at which the filtering mass increases .
the red - shift of the reionization epoch has been estimated from polarization observations of the cmb to be @xmath123 ( wmap 5-year results , * ? ? ?
however , the presence of the gunn - peterson trough in quasars at @xmath124 shows that the universe was not yet fully re - ionized at earlier epochs @xcite .
according to models ( see e.g. , * ? ? ?
* ) , the minimum circular velocity for a dwarf halo to accrete and cool gas in order to produce star formation is @xmath125 kms@xmath9 , which corresponds to a total mass of @xmath126 m@xmath8 .
however , most dwarf galaxies in the local group , including all the milky way satellites and five out of the six of the lcid sample , show circular velocities well below that value
. nevertheless , there are currently several proposed ways to overcome this apparent contradiction . @xcite
has proposed that dsph and the rest of the smallest halos would have been formed during the pre - reionization era .
the processes of photoionization feedback resulted in the suppression of the star formation below an observable level in 90% of them .
this was supported by @xcite , who found that if positive feedback is considered , a fraction of low mass halos ( total mass below @xmath127 m@xmath8 ) would have been able to form stars before the reionization era . the star formation in these galaxies
would have been halted by internal mechanisms , like photo - dissociation of h@xmath128 or sne feedback in advance of the reionization era .
another possibility is that masses of dark matter halos of the dsphs galaxies would be much larger than those measured at the optical limit of the galaxies and larger than the limit required by the reionization scenario @xcite . in these scenarios , dark matter halos of lower masses
would have failed to form stars and remain dark .
alternatively , dark matter halos could have been much larger in the reionization epoch , with total masses above the mass required for star formation , but they lost a large fraction of their mass afterward due to tidal harassment @xcite .
it is also possible that a self - shielding mechanism would be present , protecting the gas in the inner regions or in regions denser than some limit from being heated by the uv background @xcite .
finally , it has been suggested that inhomogeneous reionization can lead to large variations in the masses of halos that survive and produce stars @xcite . in summary , it is useful to note that the first scenario @xcite and also the last ones @xcite , allow star formation in low mass halos ( total mass below @xmath129 m@xmath8 ) , while the second and third are compatible with star formation in more massive halos only ( total mass above @xmath130 m@xmath8 ) .
additionally , the first scenario would imply that star formation would have been produced before reionization and would have been halted afterward in the smallest galaxies .
recently , @xcite have presented a set of high - resolution hydrodynamical simulations of the formation and evolution of isolated dwarf galaxies including the most relevant physical effects , namely metal - dependent cooling , star formation , feedback from type ii and ia sne , and uv background radiation .
their results are very useful for a direct comparison with observations .
they study halos with present day total masses between @xmath131 m@xmath8 and @xmath132 m@xmath8 and reach an interesting set of conclusions that include many of the aforementioned ones .
first , halos that are not massive enough to accrete sufficient gas to form stars before @xmath133 lose their gas subsequently due to uv background heating . in this sense
, reionization sets a lower mass limit of the halo for the star formation .
second , in halos that form stars , feedback is the main process driving the evolution of the galaxy and regulating its star formation .
it alone can blow out all the gas of a dwarf system before @xmath134 , while uv background alone has almost no effect .
however , if feedback has previously made the gas diffuse and reduced its radiative cooling efficiency , then the uv background has a strong effect , producing a sharp cut off in the star formation at the epoch of reionization .
finally , self - shielding is effective only in the inner regions of the halos at the more massive end .
all the aforementioned mechanisms are founded on solid and extensive theoretical modeling .
however to decide to what extent one or another should be preferred , detailed observational evidence is necessary .
one such observable is whether the star formation in the smallest galaxies was halted at reionization . to answer this question ,
a precision of @xmath135 gyr is required in the determination of the sfh at @xmath136 gyr ago @xcite , which is indeed what we have reached in our work . in the following
we will discuss the different possibilities in light of our results for paying attention first to sne feedback effects and second to uv background effects .
the sfh obtained for allows us to make a direct estimate of the importance of feedback in . to do so
, the mechanical luminosity of the sne produced in the main star formation episode and the mass of the galaxy can be compared with the results of @xcite .
they used models to calculate the conditions for blow - away , partial mass loss , and no mass loss of dwarf galaxies as a function of the baryonic mass of the galaxy and the mechanical luminosity , @xmath137 , of the sne produced during a central , instantaneous star formation episode .
the kinetic energy released to the interstellar medium by the sne produced in the old , main star formation episode in , can be computed as follows . for a continuous star formation rate of 1 m@xmath8 yr@xmath9
, the starburst99 models @xcite predict a mechanical luminosity of @xmath138 .
this is the equilibrium value reached after a few tens of million years from the first star formation .
the star formation rate is for a mass range 1 to 100 m@xmath8 and is computed using a classical imf with exponent @xmath139 . for direct comparison , we convert this rate to the same stellar mass interval and imf that we have used and obtain 1.7 m@xmath8 yr@xmath9 . scaling these values to the sfr obtained for
, the total mass converted into stars in the initial episode is @xmath85 m@xmath8 . assuming a gaussian profile of @xmath140 gyr , or a duration ( fwhm ) of 1.4 gyr , results in an average sfr of @xmath141 m@xmath8 yr@xmath9 .
scaling starburst99 models to this average sfr results in a mechanical luminosity of @xmath10 erg s@xmath9 .
alternatively , we can compute the mechanical luminosity from core collapse sne only ( note that starburst99 includes all sources of mechanical luminosity ) . from
[ secsfhmethod ] , the total number of sne formed during the old , main episode was @xmath86 or @xmath87 for minimum core collapse progenitor masses of 6.5 m@xmath8 or 10 m@xmath8 , respectively @xcite .
assuming an energy release per sn of @xmath142 erg @xcite , we obtain a total mechanical luminosity released during the old , main episode of @xmath143 erg s@xmath9 or @xmath144 erg s@xmath9 , respectively .
since these values are compatible with the starburst99 estimate , we will adopt @xmath10 erg s@xmath9 . combining this with a baryonic mass of @xmath96 m@xmath8 ( [ secsfhfeatures ] ) , the model results of @xcite shown in their figure 1 , place in the regime of mass loss , but fairly close to the blow - away regime .
however , equation 1 from @xcite gives a relation between the baryonic and dark matter mass of dwarfs which is an extrapolation to low mass galaxies of the relation derived by @xcite for more massive galaxies ( total masses above @xmath145 m@xmath8 ) .
from this relation , the total mass for is @xmath146 m@xmath8 .
alternately , @xcite give the final ( @xmath134 ) stellar , gaseous , and total masses for their model runs .
the relations between these parameters depend on the model ingredients , but for a baryonic mass like that of , the total mass is in the range from @xmath147 m@xmath8 to @xmath7 m@xmath8 . using this estimate instead of the @xcite extrapolation places well within the regime of mass - loss due to feedback and far from the blow - away regime . a flat einstein - de sitter universe with @xmath148 , @xmath149 ( as derived from the wmap 5-year data ; * ? ? ?
* ) , has an age @xmath150 gyr .
the wmap 5-year data indicate the epoch of reionization occurred at @xmath151 corresponding to a look - back time of @xmath152 gyr .
evidence from quasar spectra indicate the epoch of reionization is over at a redshift of @xmath124 or a look - back time of @xmath6 gyr @xcite .
we will focus on the latter since it is more conservative from the point of view of the forthcoming discussion .
did form most of its stars before or after the epoch of reionization ? using the cumulative mass fraction from figure [ f7 ]
, we find that more than 80% of the stellar mass of has been formed later than @xmath153 gyr ago ( @xmath154 ) .
thus , the observations indicate that the majority of the stars in were formed after the epoch of reionization
. we can use synthetic data to better assess this constraint . in figure
[ f13 ] , we show the results of two synthetic sfhs , one in which all stars are formed between 13.4 and 12.4 gyr ago ( before the epoch of reionization ) and another where all of the stars are formed between 11.9 and 10.9 gyr ago .
figure [ f13 ] shows that for the model where all of the stars formed before the epoch of reionization is clearly inconsistent with the sfh of . on the other hand , the model with all stars forming after the epoch of reionization is a good match to our sfh for , indicating that it is highly unlikely that formed a large fraction of its stars before reionization had concluded .
however , the absolute ages in our comparisons are intrinsically dependent on the stellar evolution models which have intrinsic uncertainties ( e.g. , * ? ? ?
* ; * ? ? ?
modern stellar evolution models have addressed many of the shortcomings in earlier generations of models ( e.g. , see discussion in * ? ? ?
* ) , and the results of the relative ages derived from main sequence stars are in excellent agreement ( better than 0.5 gyr ) as demonstrated by @xcite
. the accuracy of the absolute ages is also increasing .
for example , the average age of the oldest globular clusters obtained from the age analysis by @xcite , using the basti library , is 12.3 gyr with an age dispersion of @xmath155 gyr ( marn - franch , private communication ) . in this regard , the complete inconsistency of our sfh for with the model in which all stars are formed before reionization is persuasive . nonetheless ,
as argued in @xcite , the strongest constraints come from relative ages . because all of the lcid galaxies were observed and analyzed in an identical manner we can make the following argument . since we find an age difference of @xmath156 gyr between the peak of star formation in and that of tucana , and since that difference is larger than the difference between the age of the universe and the epoch of reionization , it is highly unlikely that a significant fraction of the stars in have been formed before the epoch of reionization .
an identical conclusion was found for the cetus dsph @xcite . in summary ,
figure [ f5f6 ] provides strong evidence that most , if not all , of the star formation in occurred after the end of reionization .
although is a low mass galaxy , do theoretical models expect it to form its star before reionization ? before putting the former result in the context of current theoretical model predictions , some estimates of integrated physical quantities of are necessary .
the velocity dispersion of the gas in is 8 kms@xmath9 at a radius of 470 pc @xcite which corresponds to a total mass within that radius of @xmath157 m@xmath8 .
according to @xcite this value should _ not _ be considered indicative of the total mass of the galaxy , but only of the virial mass at the measured radius . following the results of @xcite , the total mass of
would be expected to be in the interval @xmath158 m@xmath8 to @xmath159 m@xmath8 .
equation 16 of @xcite gives a relation between circular velocity and total mass in a density distribution given by an modified isothermal sphere .
using @xmath160 , it can be written as @xmath161 , where @xmath162 is the total mass in units of @xmath163 m@xmath8 and @xmath164 is given in km s@xmath9 .
thus , for a total mass in the range of @xmath147 m@xmath8 to @xmath159 m@xmath8 , the resulting circular velocity for is between 28 km s@xmath9 and 36 km s@xmath9 .
these values are close to or somewhat above the limit of 30 km s@xmath9 below which the uv background is expected to prevent any subsequent star formation . in summary
, it is possible that was massive enough that its star formation was not halted by uv background heating and it was also able to preserve at least part of its gas against feedback from sne .
however , models by @xcite indicate that , for its estimated mass , the _ combination _ of the the uv background together with sn feedback should have halted the star formation at @xmath124 ( or @xmath6 gyr ago ) . since this is evidently not the case , a mechanism such as self - shielding seems necessary to allow the galaxy continue forming stars , at least in its inner regions , as suggested by @xcite . if the total mass of would be of the order of @xmath127 m@xmath8 or lower ( as obtained from the relation used by * ?
* ) , the circular velocity would drop to values of the order of 20 km s@xmath9 . for this total mass estimate , our observation that most of the star formation in has taken place after @xmath124 would be difficult to reconcile with models .
however , one possibility is that a mechanism like mass loss due to tidal harassment ( as proposed by * ? ? ?
* ) would be at play .
on the other hand , the @xcite relation requires a large extrapolation for the mass of , and this may simply imply that the @xcite mass estimates are to be preferred .
we have presented an analysis of the sfh of the transition - type dwarf galaxy , based on deep hst photometry obtained with the acs .
the assumed distance of the galaxy is 652 kpc @xcite .
a summary of the results follows . *
a deep cmd reaching the oldest main - sequence turn - offs with completeness of @xmath165 has been obtained .
exhaustive crowding tests have been performed in order to properly characterize the observational effects .
careful simulations of the observational effects have been conducted , taking into account the variable amount of crowding within the galaxy . *
the sfh for the entire lifetime of the galaxy has been obtained .
tests of self - consistency , uniqueness , and stability of the solution have been performed together with tests exploring the dependency of the solution on the photometric reduction package , stellar evolution library , and the sfh derivation code .
* the solution shows that the sfh of is dominated by an old , main episode with maximum occurring @xmath1 gyr ago and a duration estimated in 1.4 gyr ( fwhm ) .
subsequently , has continued forming stars until the present at a much lower rate . *
the total mass of stars produced is @xmath94 m@xmath8 . of that
, @xmath85 m@xmath8 corresponds to the old , main star formation episode .
the current mass in stars and stellar remnants is @xmath95 m@xmath8 . using a gas mass of @xmath17 m@xmath8 @xcite
the total baryonic mass of the galaxy is @xmath96 m@xmath8 .
* there is little evidence of chemical enrichment during the first main episode of star formation lasting until @xmath117 gyr ago . after that
, the metallicity increased more steeply to a present day value of @xmath3 .
the epoch in which the enrichment rate increased is coincident with the epoch at which the star formation drops to a low value .
this suggests a scenario in which formed stars mainly from fresh gas during the first part of its life .
after @xmath4 gyr ago much of the gas supply was exhausted and stars formed from a larger fraction of recycled gas .
this resulted in both a lower star formation rate and an increase of the chemical enrichment rate .
* the difference between the lack of early chemical enrichment in the isolated transition galaxies compared to the significant early chemical enrichment in the isolated dsph galaxies of comparable luminosities may be indicative that the dsph galaxies experienced significantly more tidal stripping .
* most or all the star formation was produced in after the reionization epoch , assumed to occur @xmath6 gyr ago . *
the mass of and the mechanical luminosity from sne associated with the old , main episode ( @xmath166 erg s@xmath9 ) indicate that mass - loss by galactic winds is important , but that complete blow - away was not likely according to the models of sn feedback by @xcite . in other words , seems to be massive enough to conserve at least a fraction of its gas against sne feedback . *
according to models by @xcite , the total mass of is about @xmath147 m@xmath8 to @xmath7 m@xmath8 , which corresponds to a circular velocity of 28 km s@xmath9 to 36 km s@xmath9 @xcite .
these values are close to or somewhat above the limit of 30 km s@xmath9 below which the uv background radiation would prevent any subsequent star formation . * according to @xcite , the combined effects of feedback and uv background radiation could have stopped the star formation in a galaxy like after reionization , unless some other mechanism like self - shielding is present . indeed , comparison of the sfh of with model results by @xcite indicate that self - shielding should be at play to prevent complete cessation of star formation in at the epoch of reionization .
the results presented in this paper allow us to sketch the following scenario for the early evolution of .
the peak of star formation occurred @xmath1 gyr ago , and the star formation rate dropped dramatically @xmath117 gyr ago or earlier although a low star formation activity is observed until present .
most or all the star formation occurred in the galaxy after the end of reionization ( although some star formation before that time can not be ruled out ) .
feedback from sne would have created outflows , but would have been insufficient to completely blow - away the gas .
the drop in star formation at @xmath117 gyr ago may have been due to the combined results of feedback , uv background heating , and gas exhaustion due to star formation . according to models presented by @xcite , at the epoch of reionization , ( in the absence of self - shielding )
, the combination of heating by cosmic uv - background and sne feedback should have completely stopped the star formation in the galaxy .
the presence of the star formation activity extending to the present leads us to infer that a self - shielding mechanism has been important in the early evolution of .
we are very grateful to dr . c. prez gonzlez and c. muoz - tun for fruitful comments .
the computer network at iac operated under the condor software license has been used .
authors sh , aa , cg and mm are funded by the iac ( grant p3/94 ) and by the science and technology ministry of the kingdom of spain ( grant aya2007 - 3e3507 ) .
sc is funded by the science and technology ministry of the kingdom of spain ( grant aya2007 - 3e3507 ) .
support for this work was provided by nasa through grant go-10515 from the space telescope science institute , which is operated by aura , inc .
, under nasa contract nas5 - 26555 .
this research has made use of nasa s astrophysics data system bibliographic services and the nasa / ipac extragalactic database ( ned ) , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
as we have mentioned in [ secsfhmethod ] , solutions which are unique , stable , self - consistent , and as independent as possible of the choice of parameters and model functions are desirable .
we have done many tests to evaluate these properties in the sfhs derived for lcid galaxies . here
we briefly discuss these aspects of the calculated sfhs .
this condition is fulfilled if combinations of simple populations which are within the error bars of the adopted solution produce cmds which are indistinguishable from the best fit cmd .
it is also required that combinations of simple populations significantly different from the adopted solution produce calculated cmds significantly different from the solution one .
simulations done during the tests for the iac - star / minniac / iac - pop method ( * ? ? ?
* ; * ? ? ?
* and this work ) , consisting in obtaining the sfh of mock stellar population of known sfh , show that the aforementioned conditions are fulfilled in general . by self - consistency
we refer to the capability of the method to reproduce synthetic input .
when a mock population is computed according to a given sfh and analyzed in an identical manner as real data , the resulting sfh should be equal to the input one , within errors and resolution limits .
such tests have been executed in @xcite and repeated here using the best fit sfh given in fig .
[ f5f6 ] to compute the mock population to be analyzed . for simplicity
, we will show here only the tests for dolphot photometry but the result is equivalent for daophot photometry .
see also tests presented in [ secconstraints ] . to minimize the dependency of the test results on the associated statistical fluctuations inherent to the method ( present in the observational uncertainties simulation and in the synthetic cmd modeling ) , we have obtained the sfhs for 10 cmds calculated from the sfh of figure [ f5f6 ] changing only the random number generation seed .
an example is shown in figure [ f8 ] .
figure [ f14 ] shows the average of all of the new sfhs and the amrs obtained .
the input values , i.e. , the sfh and amr of , are also shown for comparison .
the main conclusion of the self - consistency test is that the sfh recovering process introduces no significant bias in the location of the main peak nor in the amr , and that the sfh features are properly recovered , although it introduces some degree of degradation in time resolution .
this has to be taken into account and corrected for when quantities like the maximum intensity of a burst or its duration are relevant ( see [ secconstraints ] ) .
stability is related to uniqueness , but the question we want to address here is what is the maximum excursion of the model functions and external parameters producing a sfh within the error bars of the solution .
we have tested the effects of changing the fraction of binary stars , the imf , and the photometric zero points .
the latter are discussed in [ secsfhresult ] .
the dependency of the solution on the fraction of binary stars and the imf has been analyzed in detail in @xcite and e. skillman et al .
( 2011 , in prep . )
respectively . in short ,
any fraction of binary stars larger than 0.4 gives a sfh within the error bars . in the case of the imf , for the interval of stellar masses
@xmath167 , the range of values producing solutions within the errors is @xmath168 , where @xmath169 is the mass exponent index in @xmath170 .
the solution is independent of @xmath169 for the interval @xmath171 , although the total stellar mass is not . as an indication , the total mass converted into stars in
would be increased by a factor @xmath26 if the imf exponent changes from @xmath172 to @xmath173 for the interval @xmath171 .
the same factor of @xmath26 exists between the corresponding current masses in existing stars and stellar remnants . as a summary , table [ parrange ] shows the range of variation for the binary fraction , imf , and photometric offsets which give solutions within the error bars for the adopted sfh of .
since the number of real stars is limited , the observational sampling could be insufficient and could introduce undesirable biases in the solution . to check this point
we have obtained the sfh of five different random subsets of the observed stars .
each one contains 50% of the total , so that each star may have been included in several subsamples .
figure [ f15 ] shows the sfh of and the average of the sfhs of the five subsamples . for the comparison to be meaningful , the test results have been normalized by a factor .
error bars show the dispersion of the five subsample sfhs .
the agreement is very good and no bias is apparent , which indicates that our solution does not suffer from under - sampling effects .
marn - franch , a. , aparicio , a. , piotto , g. , rosenberg , a. , chaboyer , b. , sarajedini , a. , siegel , m. , anderson , j. , bedin , l.r . ,
dotter , a. , hempel , m. , king , i. , majewski , s. , milone , a.p . ,
paust , n. , reid , i.n .
2009 , , 694 , 1498 , evenly spaced by factors of 2 .
left axis shows magnitudes corrected from extinction .
right axis shows absolute magnitudes .
a distance modulus of @xmath174 and an extinction @xmath175 , @xmath176 have been used .
bottom lines in the right panel show the 0.25 , 0.50 , 0.75 and 0.90 completeness levels .
three isochrones from the basti stellar evolution library have been over plotted .
[ f2],width=529 ] , of obtained from dolphot ( upper panel ) and daophot ( lower panel ) photometry sets .
the sfh as a function of age @xmath80 and metallicity @xmath81 are shown projected on the @xmath177 and @xmath178 planes , respectively . dashed lines give the error intervals .
the age - metallicity relationship is the projection onto the look - back time metallicity plane .
[ f5f6],title="fig:",width=453 ] , of obtained from dolphot ( upper panel ) and daophot ( lower panel ) photometry sets .
the sfh as a function of age @xmath80 and metallicity @xmath81 are shown projected on the @xmath177 and @xmath178 planes , respectively .
dashed lines give the error intervals .
the age - metallicity relationship is the projection onto the look - back time metallicity plane .
[ f5f6],title="fig:",width=453 ] ( upper panel ) , metallicity ( middle panel ) , and cumulative mass fraction ( lower panel ) of obtained from dolphot ( thick , solid lines ) and daophot ( thick , dashed lines ) photometry sets .
thin lines give the uncertainties .
cyan lines show the final adopted solution , which is the mean of dolphot and daophot ones .
vertical dotted - dashed lines indicate the times corresponding to the 10@xmath179 , 50@xmath179 and 95@xmath179 percentiles of @xmath80 , i.e. the times for which the cumulative fraction of mass converted into stars was 0.1 , 0.5 and 0.95 of the current one .
a redshift scale is given in the upper axis , computed assuming @xmath180 , @xmath149 , and a flat universe with @xmath181 .
[ f7],width=529 ] m@xmath8 and their intensities are similar to the sfr of at the corresponding age .
arrows heights are not indicative of burst intensity .
curves show the gaussian profile fits to the solutions .
the gaussian profile areas are normalized to the same value for easier comparison .
the solution peaks ( @xmath182 ) and standard deviations ( @xmath105 ) are given .
[ f9],width=529 ] , of gaussian profile fit solutions obtained for a set of gaussian profile mock input sfhs with standard deviation @xmath113 , centered at about the age of the peak of the main episode of ( 11.5 gyr ) .
observational uncertainties were simulated in the associated mock cmds and the sfhs were recovered following the same procedure as for .
a quadratic fit to the data is shown , together with the @xmath105 value measured in the real data solution ( @xmath183 ) and the corresponding real value inferred using the fit ( @xmath184 ) .
[ f10],width=529 ] , @xmath149 , and a flat universe with @xmath181 .
vertical dotted - dashed lines indicate the times corresponding to the 10@xmath179 , 50@xmath179 and 95@xmath179 percentiles of @xmath80 for each galaxy , i.e. , the times for which the cumulative fraction of mass converted into stars was 0.1 , 0.5 and 0.95 of the current one .
[ f12],width=529 ] ccc 2005 - 09 - 12 16:18:15 & 2512 & f475w + 2005 - 09 - 12 16:42:00 & 2304 & f814w + 2005 - 09 - 14 16:16:02 & 2512 & f475w + 2005 - 09 - 14 16:39:47 & 2304 & f814w + 2005 - 09 - 15 14:38:40 & 2512 & f475w + 2005 - 09 - 15 15:02:25 & 2304 & f814w + 2005 - 09 - 16 13:01:46 & 2512 & f475w + 2005 - 09 - 16 13:25:31 & 2304 & f814w + 2005 - 09 - 17 11:24:47 & 2512 & f475w + 2005 - 09 - 17 11:48:32 & 2304 & f814w + 2005 - 09 - 17 14:36:38 & 2512 & f475w + 2005 - 09 - 17 15:00:23 & 2304 & f814w
+ lcccc stellar evolution library & basti & girardi & & basti + binary fraction & 0.0 & 1.0 & 0.2 & 0.4 + imf ( @xmath185 ) & 0.0 & 4.0 & 0.1 & 1.3 + imf ( @xmath186 ) & 0.0 & 5.0 & 0.1 & 2.3 + @xmath187 & -0.21 & 0.21 & 0.105 & 0.13/0.08 + @xmath188 & -0.15 & 0.15 & 0.075 & 0.11/0.11 + cmd gridding & @xmath189 & @xmath189 & & + age sampling ( @xmath190 ) & 0.2 gyr & 0.5 gyr & & + age sampling ( @xmath191 ) & 1.0 gyr & 1.0 gyr & & + metallicity sampling ( @xmath192 ) & 0.0002 & 0.0002 & & + metallicity sampling ( @xmath193 ) & 0.0005 & 0.0005 & & + | we present an analysis of the star formation history ( sfh ) of the transition - type ( dirr / dsph ) local group galaxy ( pisces ) based on deep photometry obtained with the _ advanced camera for surveys _ onboard the _ hubble space telescope_. our observations reach the oldest main sequence turn - offs at high s / n , allowing a time resolution at the oldest ages of @xmath0 gyr .
our analysis , based on three different sfh codes , shows that the sfh of is dominated by a main episode @xmath1 gyr ago with a duration of @xmath2 1.4 gyr .
subsequently , continued forming stars until the present , although at a much lower rate .
roughly 90% of the stars in were formed in the initial episode of star formation .
extensive tests of self - consistency , uniqueness , and stability of the solution have been performed together with the iac - star / iac - pop / minniac codes and these results are found to be independent of the photometric reduction package , the stellar evolution library , and the sfh recovery method .
there is little evidence of chemical enrichment during the initial episode of star formation , after which the metallicity increased more steeply reaching a present day value of @xmath3 .
this suggests a scenario in which first formed stars mainly from infalling fresh gas , and after about @xmath4 gyr ago , from a larger fraction of recycled gas .
the lack of early chemical enrichment is in contrast to that observed in the isolated dsph galaxies of comparable luminosity , implying that the dsphs were more massive and subjected to more tidal stripping .
we compare the sfh of with expectations from cosmological models .
most or all the star formation was produced in after the reionization epoch , assumed to be completed at @xmath5 or @xmath6 gyr ago .
the total mass of the galaxy is estimated to be between 2 and @xmath7 m@xmath8 , corresponding to circular velocities between 28 km s@xmath9 to 36 km s@xmath9 .
these values are close to but somewhat above the limit of 30 km s@xmath9 below which the uv background is expected to prevent any star formation after reionization .
feedback from sne associated with the initial episode of star formation ( mechanical luminosity from sne @xmath10 erg s@xmath9 ) is probably inadequate to completely blow away the gas .
however , the combined effects of sn feedback and uv background heating might be expected to completely halt star formation at the reionization epoch for the low mass of ; this suggests that self - shielding is important to the early evolution of galaxies in this mass range . |
the analytical treatment of the dynamics of a macroscopic system of @xmath25 particles is not possible , in general .
but there are situations which allow to simplify the dynamical description .
we will restrict ourselves to classical systems .
one of the prominent examples is brownian motion .
a big particle with mass @xmath26 interacts with a solvent . if the solvent particles have a mass @xmath27 much smaller than @xmath26 a time scale separation exists . separating the fast motion of the solvent particles from that of the big one , called brownian particle , a langevin equation results for the motion of the brownian particle .
the influence of the solvent particles occurs through a friction and a fluctuating force .
these two are related to each other by the so - called fluctuation- dissipation theorem of second kind . for a general system with time scale separation ,
the mori - zwanzig projection formalism @xcite allows to derive a langevin equation for the slow variables .
if there is a single slow variable @xmath28 , then one gets @xcite
@xmath29 where @xmath7 is a microscopic frequency , @xmath30 a memory kernel and @xmath31 the fluctuating force . elimination of a macroscopic number of fast degrees of freedom always leads to memory effects .
if @xmath30 relaxes much faster then the average of @xmath32 , one can perform a markov approximation , i.e. @xmath33 . the analytical calculation of @xmath30 is not feasible , in general
. however , there exists a class of models for which the exact elimination of a macroscopic number of degrees of freedom is possible , even if there is no time scale separation .
these models we call kac - zwanzig models @xcite .
they consist of a subsystem coupled to a macroscopic bath of _ harmonic _ oscillators .
the subsystem can be microscopic , i.e. the number of its degrees of freedom is of order one or it can also be macroscopic with , e.g. 10@xmath34 degrees of freedom .
since the equations of motion are linear in the oscillator coordinates , the harmonic degrees of freedom can be microscopic , i.e. the number of its degrees of freedom is of order one or it can also be macroscopic with , e.g. 10@xmath34 degrees of freedom . as a result ,
one obtains for the coordinates of the subsystem a stochastic differential equation of type of eq .. the memory kernel @xmath26 ( which is a matrix in general ) can be expressed by an integral over the frequencies of the harmonic bath .
its spectral properties strongly influence the time dependence of @xmath26 and therefore that of the coordinates of the subsystem , too . in the langevin case where the spectrum is absolutely continuous with infinite support
, the subsystem will relax to equilibrium for all of its initial conditions . in this case
, the harmonic oscillators act as a normal thermostat .
however , if , e.g. that spectrum contains a dense discrete component as it occurs for harmonic systems with quenched disorder , the situation may be different .
then , there exist initial conditions , which will not relax to equilibrium , but may converge to periodic , quasiperiodic or possibly weakly chaotic solutions .
this behavior is closely related to the existence of spontaneous energy localization .
+ spontaneous energy localization in large nonlinear discrete systems is now known as an ubiquitous phenomena related to the existence of discrete breathers ( dbs ) .
when a sufficiently large amount of energy is injected locally in a system which may sustain dbs and which is initially at low temperature , it is found that although some part of this energy spreads over the system , a substantial part of this energy may remain localized as a db @xcite over very long time .
the discovery of this kind of localized modes was so surprising because they existed in systems with discrete translational invariance in space .
dbs also called _ intrinsic localized modes_(ilm ) are spatially localized time periodic solutions of discrete nonlinear hamiltonian systems @xcite .
since many of them are linearly stable and thus can trap energy over long times , they should play an essential role in energy relaxation in complex systems .
the existence of dbs has been rigorously proven by several methods and for a variety of models with optical or acoustic phonons , with or without gaps , in one or any dimension and without or with randomness @xcite .
all these existence proofs require that the db frequency and its harmonics do not belong to the phonon spectrum ( obtained by linearization of the dynamical equations in the vicinity of the ground state ) .
otherwise , the db would be expected to radiate energy in the phonon band and to decay .
in addition , the proofs @xcite only concern the existence of _ out - of - band _ dbs , i.e. their frequencies and those of the higher harmonics are in the band gaps of the phonon spectrum . for certain models
the existence of dbs was even proven before by albanse and frhlich @xcite and by albanse , frhlich and spencer @xcite under the requirement of randomness . in that case
the eigenstates of the linear eigenvalue problem are already localized ( in dimensions larger than two the randomness has to be large enough ) and therefore it is more intuitive that localized time periodic solutions could persist in the presence of nonlinearities , although to prove this is highly nontrivial when the phonon spectrum is discrete , for example in a strongly disordered system with anderson localization , energy radiation of linear phonons by a localized time periodic solution can not occur .
then , the existence of _ intraband _ discrete breathers ( idbs ) with their frequency or some of their harmonics _ inside _ the phonon spectrum , might be a priori possible .
indeed , numerical evidence @xcite suggested that such solutions do exist .
more precisely , it was conjectured that for each localized anderson mode , there is an associated family of idbs with variable frequencies ( and amplitude ) , which converges to this anderson mode at the small amplitude limit .
it was also conjectured that the idbs with respect to their frequency dependence do not form a continuous family of solutions as for ordinary ( out - of - band ) dbs , but are defined only on a fat cantor set of frequencies ( i.e. with nonvanishing lebesgue measure ) .
a forbidden frequency gap is associated with each of the other frequencies of the linear anderson modes .
these frequencies are dense in the linear phonon spectrum but the width of these gaps drops to zero very fast as a function of the spatial distance of the corresponding anderson mode of the center to the idb .
then , the complementary cantor set of frequencies where the idb exists is non void with a strictly positive lebesgue measure .
moreover , the measure density of this cantor set is expected to go to full measure when the amplitude of the idb goes to zero that is when the idb tends to be a linear mode .
the idb should not be confused with so - called embedded solitons in discrete lattice systems ( see @xcite and references therein ) .
these are nonlinear excitations with a _
discrete _ frequency lying within the linear spectrum , whereas idb take frequencies from a set of full measure .
furthermore , in contrast to idb the existence of embedded solitons requires exponentially localized linear modes .
the first proof of existence of such idbs for a certain class of random models was done by albanse and frhlich @xcite and by albanse , frhlich and spencer @xcite .
the models studied by these authors are the _ random _ , discrete schrdinger equation and the wave equation with a cubic nonlinearity .
a perturbational approach was used to prove that all periodic solutions of the linear equation can be continued for small enough @xmath35 , the strength of nonlinearity .
for arbitrary @xmath35 it was proven that there exist at least certain periodic solutions . transforming a _ random _ , nonlinear dynamical system without acoustic phonons into anderson space , i.e. to the space of the localized eigenstates of the linear problem and restricting to nearest neighbor interactions in anderson space , frhlich , spencer and wayne @xcite
have even proven ( under certain conditions ) the existence of localized _ almost periodic _ solutions , i.e. of kam tori . besides the existence of idbs
, it was also conjectured in @xcite the existence of families of multibreather states which exhibit several energy peaks associated with different anderson modes .
we may explain intuitively this conjecture by the fact that the frequency of a nonlinear mode depends on its amplitude .
it is thus possible to tune the amplitude of an arbitrary number of different anderson modes in order they get the same frequency . despite nonlinear modes can not be simply superposed as in the linear case , a time periodic multibreather solution can be constructed numerically .
when , the number of idbs involved in the multibreather state is infinite , these multibreather states are spatially extended and able to transport energy by phase torsion @xcite . as a result , nonlinearity restores the ability of the system to transport energy .
the aim of this paper is to provide an exact proof of these conjectures for a _ special _ class of random models where the underlying physical arguments appear explicitly .
these models , consisting of a microscopic or macroscopic anharmonic subsystem coupled to a bath of harmonic oscillators , are designed on one hand for obtaining simpler mathematical proofs and on the other hand to contain nevertheless the essential features , e.g. for the trajectories @xmath6 of the anharmonic system , which can be found in more general complex and realistic models .
although these results do not provide a rigorous proof ( in the mathematical sense ) for general and fully anharmonic systems , they nevertheless make clearer the fundamental reasons for their existence and thus support the earlier conjecture that they might exist in general .
particulary , we will derive for rather general anharmonic and harmonic subsystems(arbitrary dimension , with or without disorder ) an exact formula which relates the energy dissipation rate @xmath36 of the anharmonic system averaged over a time interval of length @xmath9 to the fourier transform @xmath37 of the dissipation kernels @xmath4 and to @xmath38 , the fourier transform of the velocity - velocity correlation functions of the trajectories @xmath6 . for finite initial energy and
bounded hamiltonian from below it is @xmath39 for @xmath40 .
this will yield the nonoverlapping criterion for the support of @xmath41 and @xmath42 which will have consequences for the possible asymptotic behavior of @xmath6 for @xmath43 .
calculation of the energy dissipation rate @xmath44 for a sinusoidal force with frequency @xmath7 will demonstrate the existence of extreme sensitivity of the dissipated energy by that force on its frequency @xmath7 . compared to ordinary dissipation where @xmath45 ,
superdissipative and subdissipative behavior may occur for which @xmath8 increases , respectively , faster and slower than @xmath9 .
we will also show that there exists a set of full measure such that @xmath8 is bounded for all @xmath7 in that set , i.e. no energy dissipation exists at all .
the possibility of no energy dissipation allows the existence of periodic solutions @xmath46 , also within the phonon band . we will show that such idb exist with frequencies on a fat cantor set . in that case relaxation to equilibrium
is prohibited such that the harmonic bath acts as an _ anomalous _ thermostat .
this behavior is related to the pioneering work of fermi , pasta and ulam , which has given evidence for nonergodic dynamics .
new results confirm that ergodicity in phase space of large and complex dynamical systems is not as smooth as it was believed through langevin bath theory .
when starting from initial states which are not at thermal equilibrium , there are many _ islands _ in phase space which may trap the state of the system for long time and could result in unusual and slow relaxation processes toward the thermal equilibrium .
in order to keep our presentation selfcontained we review and discuss properties of linear spectra which have been proven earlier or are well - admitted in order to define a model and an associated set of hypothesis which could match with the different physical situations . our purpose is to describe the main characteristics which could be expected for large families of nonlinear random models with linear spectrum involving both discrete and continuous components .
continuous spectra occur only in the thermodynamic limit , i.e. for an infinite system . in order to fix the ideas ,
let us consider as an example a random model on a d - dimensional hypercubic lattice , for example a klein - gordon type model with hamiltonian @xmath47 where the momentum @xmath48 is the variable conjugate to the displacement @xmath49 at site @xmath50 .
the local potential @xmath51 is random .
for example it could have the form @xmath52 where @xmath53 is chosen according to some probability law @xmath54 .
@xmath55 denotes the bonds which connect neighboring sites @xmath56 ( counted once ) .
accordingly , the sum in eq . is a double sum over all @xmath50 and @xmath57 which are nearest neighbors .
@xmath58 is the constant of the harmonic coupling between nearest neighbors , which is assumed to be identical for all nearest neighbor bonds .
the dynamical equations are @xmath59 @xmath60 denotes the sites @xmath57 which are nearest neighbors to @xmath50 ,
i.e. the sum in eq . is a single sum over @xmath57 which are nearest neighbors of @xmath50 .
dbs are spatially localized time periodic solutions of eq .. thus , their amplitude must decay to zero at infinity .
consequently , we may approximate this dynamical system by another dynamical one which obeys the same eq .
for the few sites @xmath50 where the amplitudes of the db is not small and the linearized equations @xmath61 for all the other sites @xmath50 .
thus , the system becomes equivalent to a finite anharmonic system coupled linearly by local interactions to a harmonic system ( phonon bath ) .
note , we start with a finite harmonic system and finally take the thermodynamic limit . for a finite harmonic system
the linear spectrum is always discrete .
this kind of approximation strongly simplifies the investigation of dbs . assuming the existence of time periodic solutions ( dbs ) of the _ isolated _ anharmonic system ( i.e. uncoupled with the harmonic phonon bath ) we will discuss under which conditions these solutions could persist when the coupling to the phonon bath is non vanishing , although not too large .
this harmonic phonon bath considered uncoupled to the anharmonic system , is characterized by its frequency spectrum .
when the frequency of the db and its harmonics do not belong to the frequency spectrum of the phonon bath , the proof that the db solutions can be continued to nonzero coupling can be obtained as in @xcite .
this will be performed in the @xmath62 section . when the frequency of the db , one of its harmonics or several belong to the frequency spectrum of the phonon bath , it is essential to consider whether this phonon spectrum is discrete or continuous . figs .
[ fig1 ] and [ fig2 ] , respectively , show possible schemes representing situations where the linear phonons are either spatially extended or localized at the db frequency .
the model we propose is simply to consider the system as harmonic in the regions where the amplitude of the db is expected to be small and to keep it anharmonic in the region where the solution has a large amplitude .
of course , our model shall depend on the location of the idb or multidbs we consider since it determines the selection of the anharmonic subsystem .
this kind of simplification also allows to investigate the relaxation of an arbitrary localized initial excitation .
if the finite initial energy is spread completely over the system the displacements become arbitrary small and the linearization of eq .
is allowed for large enough times . however ,
if the spreading is incomplete , we are allowed to linearize eq . outside the localized region .
when @xmath63 is _ uniformly _ small , the linearized equations which yield the linear phonon spectrum are @xmath64 which has time periodic solutions of the form @xmath65 .
the eigenfrequencies @xmath66 and eigenmodes @xmath67 obey the eigenequation @xmath68 where @xmath69 is the matrix of the second order variation of the potential energy from its ground state .
the eigenfrequencies @xmath70 should not be confused with the frequencies @xmath53 of the small amplitude oscillations in the local potential @xmath51 .
it has been proven @xcite that the linear spectrum of eq .
is discrete and its eigenvectors ( anderson modes ) are square summable and exponentially localized with probability one , when either the spatial dimension is less than three or the coupling is not too large . at larger coupling and
when the dimension of the lattice is at least three , it is believed that there exist mobility edges in the spectrum ( although no rigorous proof is available up to now ) .
there are two mobility edges which determine a central interval where the spectrum is continuous .
the corresponding eigenmodes are extended and non square summable . in two intervals limited by the edge of the spectrum ,
the spectrum is discrete and the corresponding eigenmodes are square summable and exponentially localized .
more generally , we may consider nonlinear models on arbitrary discrete lattices where the small amplitude solution leads to the eigenvalue equation @xmath71 for an arbitrary positive bounded self - adjoint linear operator @xmath72 operating in the hilbert space @xmath73 where @xmath74 represent the sites of an infinite discrete lattice @xmath13 in @xmath75 dimensions ( or any discrete network ) , with the standard hermitian product @xmath76 and @xmath77 .
since we consider physical systems which have to be linearly stable , we must assume the spectrum to be positive or zero and then , the eigenvalues @xmath78 are the square of the eigenmode frequencies
. we may also consider systems with acoustic phonons where the spectrum is gapless and contains the translation mode @xmath79 associated with zero eigenfrequency which is always extended .
standard mathematical theories decompose the spectrum @xmath80 of such a self - adjoint bounded linear operator @xmath72 in three parts : the absolutely continuous spectrum , the discrete spectrum and the singular continuous spectrum @xcite .
the characteristics of the spectrum manifests itself in finite size systems in the limit of large size @xmath25 as explained now .
the eigenvalue equation restricted to a finite subset @xmath81 of @xmath25 indices of @xmath13 corresponds to the diagonalization of a finite self - adjoint matrix @xmath82 which yields @xmath25 orthogonal and normalized eigensolutions @xmath83 with compact support @xmath81 corresponding to real eigenvalues . taking @xmath84 such that the finite support tends to cover the infinite lattice @xmath85 , the nonvanishing accumulation points for the weak topology @xmath86 of the eigensolutions of finite systems @xmath83 are eigensolutions of eq .
for an eigenvalue @xmath87 , the corresponding limit of @xmath88 . the nonvanishing limit solutions
are necessarily square summable with a nonzero norm smaller or equal to @xmath89 and thus can be normalized .
this set of square summable eigensolutions @xmath90 with eigenvalues @xmath91 determines the discrete spectrum of @xmath72 .
since @xmath13 is countable , the discrete set of normalized eigenvectors is also countable ( or finite ) .
the inverse rate of the exponential decay of the eigenmodes determines the so called localization length which is a smooth function of the eigenfrequency and diverges when approaching mobility edges . in systems with acoustic phonons in 3 dimensions and more , it is believed that the spectrum becomes continuous in a nonvanishing positive interval close to zero and then there is only one mobility edge .
extended states ( e.g. plane waves ) are not square summable and their norm diverges proportionally to the system size @xmath25 .
these eigenstates can be nevertheless obtained as nonvanishing accumulations points of @xmath92 .
this set of nonsquare summable solutions and their corresponding eigenvalues of the eigenvalue equation are usually called pseudo eigensolutions and pseudo eigenvalues .
they corresponds to the absolutely continuous spectrum of @xmath72 .
they form uncountable continua .
the limits of the other finite size eigenstates which behave spatially differently for example which develop a fractal structure at large scale or which exhibit algebraic decay at infinity , determine the remaining part of the spectrum which is the singular continuous spectrum of @xmath72 .
although there are examples of operators with purely singular continuous spectrum ( for example with quasiperiodic potentials ) , their contribution to the spectrum of most operators is expected to be marginal , generally , and then it reduces to few isolated specific points of the spectrum ( corresponding to the mobility edges ) .
the spectrum of an operator @xmath72 ( for example random ) is a closed subset @xmath80 of the real axis
. it might be purely absolutely continuous ( for example for operators with spatial periodicity ) or purely discrete ( for example for operators in 1d with randomness or higher dimension with strong disorder ) or it might decompose into two disconnected subsets @xmath93 and @xmath94 where the spectrum is discrete and absolutely continuous , respectively , separated by mobility edges .
this situation is supposed to occur for random operators in 3 dimensions and more at weak enough disorder@xcite .
considering now again , the nonlinear system at the anticontinuous limit @xmath95 , it simply becomes a collection of uncoupled nonlinear oscillators with hamiltonian @xmath96 and thus is trivially integrable .
it is then convenient to represent each anharmonic oscillator with the standard action - angle variables @xmath97 .
then , the local hamiltonian @xmath98 is only a function of its action @xmath99 , and its frequency is @xmath100 .
note , that @xmath101 should not be confused with the frequencies @xmath53 in eq .. time periodic solutions at frequency @xmath7 are obtained by choosing each oscillator @xmath50 either at rest or oscillating at frequency @xmath102 if there exists @xmath99 such that this equality may be fulfilled .
the solution corresponding to a discrete breather at site @xmath50 and at frequency @xmath7 is obtained by choosing all the oscillators @xmath103 at rest , except oscillator @xmath50 which is oscillating at frequency @xmath7 . at @xmath95 , the linear spectrum of eigenfrequencies defined by eq . is the _ closure _ @xmath104 of the set of linear frequencies .
considering a db solution at the anticontinuous limit at site @xmath50 and at a frequency @xmath7 which is nonresonant with the linear spectrum ( that is @xmath105 for any integer @xmath106 ) , and fulfills the condition that @xmath107 at @xmath108 , it was proven by the implicit function theorem @xcite , that this solution can be continued as a spatially localized solution for @xmath109 belonging to a nonvanishing interval . when @xmath110 , the proof for the db existence given in @xcite which requires the invertibility of @xmath111 breaks down . however , numerical investigations suggested the existence of stable idbs with frequency @xmath110 although they are not continuous functions of the model parameters .
otherwise , this numerical work also suggested the existence of intraband multisite breathers called multibreathers .
a direct proof of existence of idb in fully anharmonic models is likely possible ( as an extension of the present work ) but it should be more complex for technical reasons than in the simplified model we propose .
we think , however , that this model captures the essential physical features of general nonlinear random models .
we denote by @xmath112 the set of variables which describe the coupled anharmonic oscillators and by @xmath113 their conjugate variables .
this subsystem is described by an anharmonic hamiltonian @xmath114 without restricting generality the masses are chosen to be equal to one .
the single anharmonic oscillator where @xmath115 is a scalar , is the simplest but a special case .
this anharmonic part is linearly coupled to an infinite harmonic system described by the set of variables @xmath116 and conjugate momenta @xmath117 where the indices @xmath50 are nodes of an arbitrary network .
the harmonic hamiltonian is @xmath118 where @xmath119 is a bounded self - adjoint operator and @xmath24 denotes transposition .
again , the masses can be taken equal to unity without lack of generality .
the total hamiltonian takes the form @xmath120 where the matrix of coupling constants is @xmath121 and @xmath35 is introducted for convenience .
these coupling constants are short ranged that is @xmath122 .
it is more convenient to describe the harmonic system as a collection of noninteracting harmonic oscillators @xmath123 which corresponds to its normal modes .
the normal mode coordinates of the harmonic oscillators of the phonon bath are now the coefficients @xmath124 of the expansion of @xmath63 with respect to the normalized eigenmodes @xmath125 with frequency @xmath126 which fulfills @xmath127 .
they are now coupled to the anharmonic system by the set of coefficients @xmath128 or explicitly @xmath129 . as we assumed above , the set of coupling constants @xmath130 between @xmath115 and @xmath49
are short ranged . in most physical situations
, they decay exponentially , that is there exists two positive constants @xmath131 and @xmath132 such that @xmath133 thus , we may assume the stronger @xmath134 condition @xmath135 without loosing much generality .
then , we obtain a total hamiltonian of the form @xmath136 the reorganization energy @xmath137 is the change in the groundstate energy of the whole system under the constraint of a static displacement @xmath138 of the anharmonic oscillators .
it is obtained by minimizing the last two terms of the energy ( [ phb ] ) with respect to @xmath124 for fixed @xmath115 .
this readily yields @xmath139 where @xmath140 in well - behaved physical models , this sum is always finite since any _ finite _ static displacement @xmath115 should involve a _
finite _ reorganization energy ( and a finite force ) .
this condition requires the assumption that the positive definite series ( ) is always absolutely convergent .
for example , the case of models with acoustic modes where one of the @xmath126 may vanish because there is always a mode at zero frequency ( which corresponds to the free motion of the `` center of mass '' of the whole system ) .
this feature may cause a divergency for coefficients of series ( [ hypf1 ] ) .
but in the limit of low frequency , these ( acoustic ) phonon modes decouple from the local degree of freedom describing the anharmonic part of the system ( @xmath141 ) and thus this reorganization energy ( [ hypf1 ] ) remains finite since @xmath142 is linear or even of higher order in @xmath11 for @xmath143 .
the reorganization energy @xmath137 renormalizes the potential energy of the anharmonic subsystem , i.e. we can introduce an effective energy @xmath144 .
thus , the total hamiltonian ( [ phb ] ) may be written as @xmath145 where @xmath146 is the harmonic displacement minimizing the last two terms of , for fixed @xmath115 .
since we did not specify properties of the harmonic subsystem , which is considered as very general , it is now necessary to introduce some assumptions at this level for ensuring our model to be physically realistic .
the density of the positive eigenfrequencies @xmath126 of the harmonic bath ( called density of states ) is the limit of the positive measure defined as @xmath147 in physical situations , it is generally an absolutely continuous measure , dense on a finite interval ( that is @xmath148 where @xmath149 is a measurable positive function ) .
but this assumption is not necessary and moreover does not characterize whether the phonon spectrum is discrete or continuous .
for example , in spatially periodic harmonic systems the whole phonon spectrum is absolutely continuous with extended eigenstates which are planewaves . in random anderson models in three dimensions and at weak enough disorder ,
the phonon spectrum @xmath80 may split into two disconnected sets @xmath150 which consists of a central interval @xmath94 where the spectrum is supposed to be absolutely continuous ( corresponding to extended eigenstates ) and the union @xmath93 of two side intervals where the spectrum is supposed to be discrete ( and corresponding to square summable eigenstates ) .
these intervals are separated by mobility edges ( isolated points ) . at large disorder in 3d or at lower dimension for any disorder , the whole spectrum is discrete and consists of a single interval @xmath93 . actually , in all these examples , the density of states is an absolutely continuous measure which does not determine a priori whether or not the phonon spectrum is continuous or discrete .
note also that in the case of random potentials , the density of states depends only on the statistical properties of the potential and not on the realizations .
the characteristics of the spectrum ( absolutely continuous , singular continuous or discrete ) appears in our model through the properties of the distribution of coupling constants @xmath151 .
they determine the properties of the kernel which appears after elimination of the harmonic variables and which shall be discussed later .
models for an anharmonic system coupled to a harmonic phonon bath are ubiquitous in physics for describing relaxation processes , for example chemical reactions within kramers theory @xcite . however , in those applications the spectrum of the harmonic bath was always assumed to be absolutely continuous . in that case ,
standard calculations yield after elimination of the phonon bath variables , extra damping terms for the anharmonic system ( plus langevin random forces , if the temperature is not zero ) . when there is a discrete spectrum , the results are totally different , as we shall see now .
the well - known advantage of a coupling with a harmonic phonon bath is that it is straightforward to eliminate the variables of the phonon bath in order to get effective equations for the anharmonic variables @xcite ( we already took advantage of such a situation for proving the existence of ( interband ) discrete breathers in systems with acoustic phonons @xcite ) . making use of eqs . and
, the dynamical equations are @xmath152 where @xmath153 stands for the gradient of @xmath154 with respect to @xmath115 .
the second equation is linear and readily yields @xmath155 where @xmath156 \cos { \omega_{\nu}t } + \frac{\dot{\epsilon}_{\nu}(0)}{\omega_{\nu } } \sin { \omega_{\nu}t } \ ] ] is an arbitrary solution of the homogeneous part of eq .. with respect to the equilibrium position @xmath157 for given initial coordinates @xmath158 . ]
if we assume the phonon bath in thermal equilibrium , @xmath159 is a random variable due to thermal fluctuations and we have at temperature @xmath9 , @xmath160 . substituting @xmath161 from into ,
we obtain @xmath162 where the coefficients @xmath163 of matrix @xmath164 are defined by @xmath165 this series is absolutely convergent and yields a smooth function of @xmath24 because of hypothesis and @xmath166 : @xmath167 @xmath1 is the langevin random force @xmath168 it is convenient to rewrite eq . as @xmath169 where @xmath170 and @xmath171 .
we note @xmath172 which corresponds to the coefficients of the reorganization energy .
eq.([effenp1 ] ) has the form of the langevin equation ( [ eq:1 ] ) , however with the nonlinear term @xmath173 .
the damping kernel @xmath0 is related to the random force @xmath1 by the general langevin relation @xmath174 where @xmath175 denotes the canonical average with respect to @xmath176 and @xmath177 .
definition shows that the damping kernel @xmath178 is the fourier transform of @xmath179 with matrix elements which are measures @xmath180 actually , measure @xmath181 may be written in a more compact way as the result of an integration in the complex plane @xmath182 on a closed contour @xmath183 which encompasses the part of the spectrum of the positive linear phonon operator @xmath119 on the real axis which is smaller than @xmath184 .
here it is @xmath185 .
note that in the case of random potentials , this measure does depend on the realization of the disorder unlike the density of states .
this positive measure @xmath186 can be split into three components which are respectively absolutely continuous , discrete and singular continuous measures .
eq . proves that this decomposition precisely corresponds to the decomposition of the _ spectrum _ @xmath80 of operator @xmath119 which describes the phonon bath , into its absolutely continuous @xmath94 , discrete @xmath93 and singular continuous parts @xmath187 .
thus , the support of each of these measure components on the real @xmath188 axis is independent of @xmath189 and @xmath190
. consequently , the properties of the fourier transform measure of the memory kernel @xmath179 are directly related to the spectral properties of the phonon bath .
we now show how the properties of the kernel @xmath191 and of its fourier transform @xmath179 are related to the properties of the phonon bath .
actually , the fourier transform eq.([posmeas ] ) of the kernel @xmath4 is not a priori a smooth function but is a positive measure @xmath192 which very generally can be decomposed into three parts : the absolutely continuous part @xmath193 where @xmath193 is a positive measurable function of @xmath188 .
the discrete part @xmath194 which is a sum of dirac functions , and the singular continuous part @xmath195 is simply the rest .
these different components are directly related to the corresponding components of the spectrum of the phonon bath .
we will discuss this relationship only qualitatively , without attempting any rigor .
the decomposition of the positive measure @xmath196 into these three parts implies a corresponding decomposition @xmath197.their qualitative time dependence will be discussed below for the absolutely continuous and the discrete part . *
_ absolutely continuous phonon spectrum _
+ for the eigenvalues @xmath198 which belong to the absolutely continuous part of the phonon spectrum , the eigenvectors @xmath125 corresponding to @xmath126 are extended , their components @xmath199 of the normalized state are proportional to @xmath200 where @xmath25 is the number of sites @xmath50 of the harmonic part of our system which is supposed to go to infinity .
since we assumed that the coupling with the phonon bath @xmath201 is local that is is fulfilled , we can write + @xmath202 + where @xmath203 is a finite number independent on @xmath25 for @xmath204 . moreover ,
the extended eigenvector @xmath125 and consequently @xmath203 , depend on a continuous parameter ( for example the wave vector ) .
an important consequence is that the average of @xmath203 for @xmath205 $ ] in the limit of an infinite system and small @xmath206 is a well - defined measurable function of @xmath188 + @xmath207 + thus , the contribution of the absolutely continuous phonon spectrum to the fourier transform eq.([posmeas ] ) of the kernel @xmath192 is nothing but its absolutely continuous part @xmath193 .
if @xmath208 is the density of states of the absolutely continuous phonon spectrum the absolutely continuous measure is given by + @xmath209 + + which implies + @xmath210 + + since @xmath208 is absolutely integrable , @xmath211 decays as @xmath212 or faster , due to the riemann - lebesgue lemma .
this holds e.g. for a spatially periodic system in any dimension and for a three- or higher - dimensional system with disorder which is not too strong .
+ * _ discrete phonon spectrum _
+ for the eigenvalues @xmath213 which belong to the discrete part of the phonon bath , @xmath125 is a localized square summable eigenmode .
then , the scalar product @xmath214 is finite .
considering an arbitrary closed interval @xmath215 , we require a stronger convergence than ( [ hypf1 ] ) for the discrete case : + @xmath216 + + actually , this assumption should be fulfilled in most realistic physical situations with a discrete spectrum .
+ generally , the eigenmodes @xmath125 decay exponentially with respect to @xmath217 and there is a uniform bound on the localization length provided @xmath218 ( provided this interval does not contain the mobility edges ) .
if we assume the coupling constants @xmath219 fulfill , then @xmath220 decays exponentially as a function of @xmath221 , the distance from the `` center of mass '' of mode @xmath123 .
then , since @xmath126 is bounded from below by a positive nonvanishing number , series converges
. note , however , that even for a system with acoustic phonons the series may exist because the translational invariance implies at least @xmath222 for @xmath223 however , in principle , there is no need for exponential localization , the eigenmodes could decay with some algebraic law providing they are square summable . + the contribution of the discrete spectrum of the phonon bath to the fourier transform measure ( [ posmeas ] ) is nothing but the discrete part @xmath194 which is a sum of dirac functions + @xmath224 + this immediately leads to @xmath225 + because of , @xmath226 is an absolutely convergent discrete sum of cosine functions .
consequently , in general it is an almost periodic function which does not decay to zero at infinity but oscillates between a lower and an upper bound .
the situation of a pure discrete spectrum occurs in disordered systems , in @xmath227 or @xmath228 or in higher dimensions with strong enough disorder . + * _ singular continuous phonon spectrum _
+ by definition , the singular continuous spectrum is the rest of the phonon spectrum after removing the discrete and the absolutely continuous part .
if there is a singular continuous part in the phonon spectrum , it simply appears as the singular continuous component @xmath195 of @xmath179 . in standard models which are spatially periodic or random ,
there is no singular continuous spectrum but there are special models @xcite which are neither spatially periodic nor random ( for example quasiperiodic ) and where the whole spectrum is singular continuous . in such models , for @xmath229 ,
the eigenvectors @xmath125 are not square summable though are not uniformy extended as for example planewaves .
they have a behavior at infinity which is not universal .
for example , they might slowly decay with power laws . or they might not decay but with intermittent and sparse peaks with possibly fractal distribution and zero density in average .
we shall not much discuss this exceptional situation in this paper .
+ if the disorder in @xmath230 systems is weak enough a mixed situation occurs with an absolutely continuous part in a central interval and a discrete one in the two intervals above and below the mobility edges .
then , @xmath179 is the sum of a discrete part and an absolutely continuous component with supports which are the corresponding frequency intervals .
when there are acoustic phonons in a random model , the spectrum is still fully discrete in one and likely two dimensions but with a localization length which diverges at zero frequency . in three dimensions ,
the low frequency phonon modes are extended . at larger frequency beyond a mobility edge
, the spectrum is discrete . in all these cases ,
the density of states remains a smooth measure indistinctly the spectrum is absolutely continuous or discrete .
there are other situations we shall not consider here where the spectrum may be absolutely continuous but where the support is a fat cantor set @xcite(then @xmath231 is not a continuous function of @xmath188 ) or when the spectrum is discrete but is not dense on finite intervals ( for example in smooth quasiperiodic potentials ) or when the spectrum is purely singular continuous ( quasiperiodic square potentials @xcite ) .
we shall not discuss here these exceptional situations .
the markov approximation is the simplest approximation which consists in assuming that the spectrum of the phonon bath is only absolutely continuous with infinite support and the coupling constants @xmath232 are uniform .
thus @xmath233 and @xmath234 .
then , eq . becomes @xmath235 which contains a damping term proportional to the velocities and a langevin force .
the solution of this equation at zero temperature , ( @xmath236 ) , always converge to a static solution corresponding to the system at equilibrium .
this result is essentially the consequence of the absolutely continuous phonon spectrum which extends from frequency zero to infinity .
starting from an initial condition where the phonon bath is at rest , i.e. @xmath237 and @xmath238 the energy provided by the anharmonic subsystem to this phonon bath may be viewed as energy dissipation for this subsystem . in the general case ,
the rate of energy dissipation is found to be related to the fourier transform of the velocity - velocity correlation function of the trajectory instead of being proportional to its square velocity as in the original langevin case . at zero temperature
when the fluctuating random force in eq .
vanishes , the possible asymptotic solutions are those which could persist without any energy dissipation . since the dissipation rate depends on a time average , the role of the transient dynamics toward the asymptotic solution disappears . actually , the asymptotic solution might not be always a static solution when the structure of the phonon spectrum is not uniform .
we already know situations when the phonon spectrum is absolutely continuous but bounded with an upper bound ( and possibly with frequency gaps where the density of states vanishes , where ( interband ) dbs are asymptotic solutions
. then their fundamental frequency ( and harmonics ) necessarily belong to gaps of the phonon spectrum @xcite .
we also know that when the phonon spectrum is purely discrete that there are time - periodic solutions ( called intraband dbs ) with fundamental frequencies which belong to a certain fat cantor set ( with finite measure ) @xcite dense in the phonon spectrum but which does not contain any phonon eigenfrequencies . however , we do not know if they could be asymptotic solutions for other initial conditions which are not those of that solutions
. we would like to comment in that section on the relationship between vanishing energy dissipation and the possible asymptotic solutions for arbitrary initial conditions .
this relationship depends sensitively on the properties of the phonon spectrum .
let us first investigate the energy transferred after a long time to the phonon bath of a solution @xmath6 of eq.([effenp ] ) ( or of the equivalent equation ( [ effenp1 ] ) ) at zero temperature , i.e. for @xmath239 .
the consequence for the asymptotic solutions will be discussed in the final subsection .
+ for time @xmath240 given , let us first define functions + @xmath241 where @xmath242 for @xmath243 and @xmath244 for @xmath245 and @xmath246 , i.e. @xmath247 is the normalized characteristic function on @xmath248 $ ] .
its fourier transform is @xmath249 let us now consider the effective energy of the anharmonic part of the system defined as @xmath250 using eq.([effenp1 ] ) with @xmath251 , ( since zero temperature is assumed ) the time derivative of this energy becomes @xmath252 the average of this time derivative from time @xmath253 to @xmath9 is @xmath254 where we used @xmath255 and definition ( [ veldef ] ) . then , using the inverse fourier transform @xmath256 and @xmath257,we obtain the energy dissipation functional @xmath258 definition ( [ kernp2 ] ) of @xmath4 implies that its fourier transform @xmath259 is a positive semidefinite matrix for any @xmath188 , consequently we have @xmath260 .
this result is physically obvious because the energy variation @xmath261 of the anharmonic part of the system has to be negative at any time @xmath9 since the total energy must be conserved and there was initially no energy in the phonon bath .
@xmath262 is obviously bounded from below ( for physical stability of the model ) and from above by the total finite energy which is initially injected to the system .
consequently , we have @xmath263 .
thus a necessary condition for the asymptotic solutions is that @xmath264 the limit of the products in ( [ avdis3 ] ) for @xmath265 might not be a smooth function of @xmath188 anymore .
for example if @xmath6 is time periodic , it should exhibit dirac peaks .
considering the limit within the topology of weak convergence of measures , this limit can be defined as a positive measure matrix @xmath266 .
then , it is proven in appendix [ appa ] , that as a measure , this limit is the fourier transform of the velocity - velocity correlation functions @xmath267 @xmath268 for a given solution @xmath6 of eq .
with @xmath269 we have derived ( in subsection 4.1 ) an exact formula for the averaged energy dissipation rate .
we will apply this formula to a sinusoidal trajectory @xmath270 with frequency @xmath7 and arbitrary phase @xmath271 .
we have @xmath272 which can also be interpreted as a driving force .
the normalized characteristic function fulfills @xmath273 its fourier transform @xmath274 fulfills @xmath275 , @xmath276 ( for @xmath277 , we set @xmath278 ) and @xmath279 for @xmath280 . then , we can write eq.([velfor ] ) as @xmath281 where @xmath282 .
this yields : @xmath283 then the average frequency dependent damping over time @xmath9 for that time periodic function @xmath6 follows from @xmath284 d\omega\ , .
\label{freqd}\end{aligned}\ ] ] the average dissipation rate @xmath285 depends sensitively on the damping kernel @xmath286 which itself depends on the phonon spectrum .
we will discuss this dependence of @xmath285 for @xmath40 for an absolutely continuous and a discrete phonon spectrum . * _ absolutely continuous phonon spectrum . _
+ as discussed in ( i ) of subsection 3.2 , @xmath287 is an absolutely continuous measure where @xmath193 is a positive continuous function on its support which is the absolutely continuous spectrum @xmath94 .
it has been assumed to be the union of one or several intervals .
@xmath288 is vanishing elsewhere .
then , using the properties of @xmath289 from above , we get a result independent of the phase @xmath271 + @xmath290 + where @xmath291 has been used . + consequently , there is energy dissipation , only , if the driving frequency @xmath7 or ( @xmath292 belongs to the absolutely continuous spectrum @xmath94 . in that case
the rate of the energy dissipation is well - defined and the energy of the phonon bath linearly increases in average with respect to time .
+ if @xmath7 belongs to the edge of one of the intervals the dirac delta functions in eq .
will contribute only half to the integral , i.e. @xmath293 is still a constant .
accordingly , @xmath294 for @xmath40 .
however , this is not true anymore if in one - dimensional systems there is a van hove singularity . in that case @xmath295
diverges and a superdissipative behavior should be expected(see next subsection ) .
+ when @xmath296 we have @xmath297 and consequently @xmath298 .
then it comes out from eq . that @xmath299 where @xmath131 is some finite constant
. since @xmath300 , the effective energy @xmath262 is monotonously decaying towards @xmath301 , but remains bounded from below at infinite time . + * _ discrete phonon spectrum _ + this case has been discussed in ( ii ) of subsection 3.2 .
@xmath302 is a discrete measure ( [ measured ] ) . substituting this into eq . and taking eq . into account yields + @xmath303
+ where @xmath304 are _ positive _ coefficients .
+ since the positive series @xmath305 is absolutely convergent , we can label the countable set of eigenfrequencies by positive integers @xmath306 such that the sequence @xmath307 is monotone decaying ( @xmath308 for any @xmath123 ) .
thus , the resonances @xmath123 are simply ordered according to their strength .
then , @xmath307 can be supposed to converge exponentially to zero as in standard physical models with discrete spectrum .
the @xmath7-dependent energy dissipation will be more subtle , compared to case ( i ) of an absolutely continuous phonon spectrum .
+ when @xmath309 does not belong to the discrete spectrum ( which is the closure of the set of eigenvalues ) , the same proof as above for @xmath310 yields again that there is no energy dissipation and that the phonon bath energy remains bounded from below .
when @xmath311 and when there is a resonance that is there exists @xmath312 such that we have precisely @xmath313 , then series ( [ freqd3 ] ) contains a term @xmath314 which diverges when @xmath315 .
consequently , @xmath316 diverges as @xmath317 when @xmath315 while the energy absorbed by the phonon bath @xmath8 diverges proportionally to @xmath318 .
this result should be expected since we know that the energy of an harmonic oscillator driven at its resonant frequency diverge as a function of time as @xmath319 .
+ however when @xmath311 but @xmath320 , the situation is more subtle because series ( [ freqd3 ] ) though it does not contain any diverging term , contains terms with denominators @xmath321 which may be arbitrarily small and may cause divergency . inspired by the construction of the liouville numbers from rational numbers , we construct a subset of @xmath93 which has full measure and
where this series is absolutely convergent for all @xmath7 in that subset .
+ with @xmath322 from eq .
we get from eq .
+ @xmath323 + + @xmath324 is an infinite series of periodic functions of @xmath9 .
if this series is absolutely convergent , its sum is an almost periodic function of @xmath9 ( in the sense of h.bohr@xcite ) which is smooth and bounded .
a sufficient condition for absolute convergence of series ( [ freqd4 ] ) is that the positive series @xmath325 is convergent .
+ in appendix b we will prove that there exists a set @xmath13 of zero lebesgue measure such that the series @xmath326 is convergent for all @xmath327 .
this implies that there exists a set @xmath14 of full measure such that the energy @xmath328 transferred to the phonon bath remains finite and bounded for all times , provided @xmath7 belongs to @xmath329 . on the other side ,
it is also proven in appendix b that there exists a subset @xmath16 of @xmath13 of zero lebesgues measure such that @xmath324 diverges for @xmath40 , if @xmath17 .
in addition it is proven that under some conditions for @xmath330 an arbitrary eigenfrequency @xmath11 belongs to @xmath16 with probability zero , only .
this excludes @xmath331 for all @xmath17 .
for @xmath332 it is @xmath333 and @xmath334 for an _ ordinary _ dissipative langevin bath with an absolutely continuous spectrum .
therefore we may conjecture that @xmath335 if @xmath17 , with an exponent @xmath189 which may depend on @xmath7 .
compared to ordinary dissipation where @xmath336 , situations where @xmath337 may be called _ superdissipative _ and those where @xmath338 _
subdissipative_. + it has been shown that the classical dissipation kernel @xmath4 determines the nonlocal action occurring in the path integral representation of the quantum mechanical propagator , after elimination of the harmonic d.o.f .
the nonlocal action describes quantum dissipation .
depending on the low frequency behavior of the spectral density @xmath339 ( corresponding to @xmath340 , i.e. @xmath341 , for @xmath342 , qualitative different quantum behavior is found .
for instance , for a particle in a double well coupled to a harmonic bath there can be ohmic damping @xmath343 , superohmic @xmath344 and subohmic @xmath345 damping @xcite . since the @xmath7-dependent dissipation discussed above is not related to the low frequency behavior of @xmath41 , the implication of the subdissipative and superdisspative behavior for quantum dissipation is not obvious .
+ in conclusion , we have obtained here a rigorous proof that when the phonon spectrum is discrete and with the assumption of fast enough decay of coefficients @xmath307 ( which is generally fulfilled in physical models ) , there is no energy dissipation by the phonon bath when it is driven ( locally ) by a time periodic force for _ most _ choices of its frequency @xmath7 ( i.e. probability 1 ) , when @xmath7 belongs to the phonon spectrum @xmath93 ( the closure of the set of eigenvalues ) .
this result first sharply contrasts the properties of a langevin bath with absolutely continuous spectrum which is always dissipative and second , gives evidence for the existence of idb .
we also proved the existence of a zero measure subset @xmath16 of frequencies for the driving force where the energy absorbed by the phonon bath is unbounded as a function of time . in that case , the phonon bath may be superdissipative or subdissipative .
+ * _ mixed case _ + when the phonon spectrum contains both an absolutely continuous component and a discrete component , energy dissipation is simply given by the sum of these two contributions .
thus for having no energy dissipation , @xmath7 has not to belong to @xmath94 but may belong to the subset @xmath346 in @xmath93 defined above where there is no energy dissipation .
of course , there is no energy dissipation when @xmath7 does not belong to any components of the spectrum @xmath347 . if one chooses a general initial condition @xmath348 , @xmath349 @xmath350 , @xmath351 for the full system one may ask for the asymptotic behavior of the solution @xmath352 , @xmath353 of eq . for @xmath354 .
the asymptotic solutions of a trajectory @xmath355 generated from an arbitrary initial condition are defined as the trajectories in the intersection @xmath356 for all time @xmath9 of the closure of the trajectory @xmath357 starting from time @xmath9 .
this set could be a fixed point corresponding to a static solution .
it could be an invariant circle .
then the asymptotic trajectory would be time periodic ( discrete breathers )
. it could be also an invariant torus and then the asymptotic trajectory would be quasiperiodic .
we should also not exclude a priori situations were the limit set is a strange attractor with chaotic trajectories .
we wish to discriminate about the possible limits which are allowed in principle according to the properties of the phonon bath . in the following
we will discuss qualitatively which asymptotic solutions may be possible . from the average energy dissipation rate ( eq . )
we get a nonoverlapping criterion for the support of both positive measures @xmath358 and @xmath359 .
this criterion , together with the generation of higher harmonics , can be used to specify features of the asymptotic solutions . * _ nonoverlapping criteria _ + equation ( [ avdis3 ] ) relates the rate of energy dissipation after a finite time @xmath9 to the smooth function @xmath360 . for @xmath40 , @xmath361
is related to the fourier transform of the velocity - velocity correlation function ( [ corrf ] ) which does not depend on the possible transient regime near the initial time but essentially on the asymptotic solution . for @xmath362 , the rate of energy dissipation becomes + @xmath363 + the positive matrix of limit measures @xmath364 ( fourier transform of the matrix of velocity - velocity correlation functions ) can be decomposed as usual in three components ( absolutely continuous , discrete and singular continuous ) @xmath365 + then , strictly speaking , the integral in eq.([avdis5 ] ) of the product of two positive generalized functions may be mathematically undefined when both measures @xmath366 and @xmath367 are not purely absolutely continuous .
such a situation may occur for example when both of these measures exhibits a dirac component at the same frequency @xmath126 .
however , in that case , the physical situation is clear because the anharmonic system would drive the phonon bath with a resonant time periodic force component at frequency @xmath126 .
then , the energy which is provided to the resonant harmonic oscillator should grow as a function of time as @xmath319 , faster than @xmath24 .
this resonant situation would correspond to a superdissipative behavior where @xmath368 since the energy dissipation per unit time would diverges as @xmath24 for @xmath369 .
more generally , we have to ascribe an infinite positive value to the integral ( [ avdis5 ] ) if its integrand is a product of two singular continuous or discrete functions and their supports overlap . + since the support of a positive measure plays a crucial role in the following and not all readers may be familiar with it we give a brief definition .
a support @xmath370 for the absolutely continuous part @xmath371 can be chosen as the set of points where the derivative @xmath372 is well - defined ( for most @xmath188 ) and is nonvanishing ( it is positive ) .
it is a measurable set with nonzero measure .
a support @xmath373 ( the smallest ) of the discrete component @xmath374 is the countable set of points @xmath375 where @xmath376 .
the support @xmath377 of the singular continuous measure @xmath378 can be chosen as the set of points where @xmath379 diverges to infinity ( this limit is defined for most @xmath188 and is zero by definition ) .
it is a measurable set which is uncountable . + we assume that the support of @xmath380 is the same for all @xmath381 .
. shows that this holds if for given @xmath123 the coupling constants @xmath382 are nonzero for all @xmath189
. this may be not true in presence of special spatial symmetries .
the same assumption is made the for the support of @xmath383 in generic models all variables @xmath384 , @xmath385 are coupled one with each other .
consequently , we assume that the support of the fourier transform of the correlation functions @xmath386 does not depend on @xmath381 .
this implies that @xmath387 have a common support .
+ the main result is that since the initial energy of the anharmonic system is finite and the total energy is bounded from below ( for stability reasons ) , a necessary condition for the asymptotic solutions , is that the average rate of energy dissipation at time infinity given by the integral ( [ avdis5 ] ) be strictly zero .
this implies the support of the positive measures @xmath388 and @xmath367 do not overlap .
+ * _ conjectures for the limit solutions _ + since @xmath389 for @xmath40 is _ not _ the fourier transform of the velocity @xmath390 which itself is directly related to the fourier transform @xmath391 of the trajectory @xmath6 , the support of @xmath392 may not yield information on @xmath393 .
if @xmath6 is periodic , quasiperiodic or almost period then @xmath394 is periodic , quasiperiodic or almost periodic . in this case
the support of @xmath392 is identical with that of @xmath393 .
however , if @xmath6 is chaotic it is not clear how @xmath393 is related to @xmath395 + if @xmath7 is in the support of @xmath393 , the nonlinearities in eq . will generate all higher harmonics with frequencies @xmath396 , with @xmath106 an integer .
similarly , if @xmath397 and @xmath398 are in the support , the nonlinearities will generate higher harmonics with frequencies @xmath399 , with @xmath400 and @xmath401 integers .
we assume that this also holds for the support of @xmath395 therefore we conjecture that because of nonlinearities of the anharmonic system , the supports @xmath370 ( resp . @xmath373,@xmath402 @xmath377 ) have the following invariance properties . if @xmath403 and @xmath404 ( resp . @xmath373,@xmath402 @xmath377 ) , then we have @xmath405 ( resp .
@xmath377 ) for any pair of integers @xmath400 and @xmath401 . a measurable set which has this property is either the full real axis or has a zero measure ( proof in appendix [ appd ] ) .
consequently , @xmath370 should be the whole real axis and therefore should overlap the support of the phonon spectrum @xmath80 .
this generates energy dissipation .
consequently , there should not exist any absolutely continuous component in the fourier spectrum of the velocity - velocity correlation function but only a discrete component or a singular continuous one ( or both ) .
+ in case of a discrete component there are two possibilities .
first , @xmath406 , i.e. @xmath5 is periodic with period @xmath407 which is consistent with an asymptotically periodic solution @xmath408 for @xmath43 .
second , @xmath409 with @xmath410 for all integers @xmath411 , i.e. @xmath5 is quasiperiodic or almost periodic .
this again is consistent with an asymptotically quasiperiodic or almost periodic solution @xmath6 . + in order that an asymptotically periodic solution exists its frequency @xmath7 has not been in the support of @xmath412 .
this requires @xmath413 for all @xmath123 .
we have already shown ( cf .
( ii ) of subsection 4.2 ) that this condition is not sufficient .
this point will be discussed in great detail in section 5 . for an asymptotically quasiperiodic or almost periodic solution
a necessary condition is @xmath414 for and @xmath415 and all @xmath123 .
we do not discuss this case but mention that the existence of such solutions were proven for a special class of models with quenched disorder @xcite .
we expect that such solutions also exist for the present model .
+ since solutions which are neither periodic , nor quasiperiodic ( or almost periodic ) are chaotic , the case of a singular continuous component of @xmath416 can only be related to asymptotically chaotic solutions .
the fact that the component is singular and not absolutely continuous ( which has already been excluded ) may imply that the chaos is not fully developed .
this subtle point is also not discussed further .
+ finally , the asymptotic solution can also be a constant one , i.e. @xmath417 , where @xmath418 is a local minimum of @xmath419 . in that case
it is @xmath420 and @xmath421 , s.t .
@xmath422 . here
we used that @xmath423 exists .
+ this discussion shows that asymptotic solutions can be constant , periodic , quasiperiodic ( or almost periodic ) and may be weakly chaotic .
as far as we know the existence of asymptotically weak chaotic solutions of eq . has never been investigated .
we should also stress again that we have discussed _ possible _ asymptotic solutions , only .
we have no information which of these asymptotic solutions will occur for given initial conditions , e.g. @xmath348 , @xmath424 , @xmath425 , @xmath426 .
in the previous section we have argued that the solution @xmath6 of eq.([effenp ] ) ( or equivalently eq.([effenp1 ] ) ) may converge towards an asymptotic solution which might be periodic , quasiperiodic or weakly chaotic .
we prove now that indeed there may exist exact periodic solutions under some conditions .
when the phonon spectrum has an absolutely continuous component , there may exist exact time periodic solutions providing their frequency and higher harmonics are outside the absolutely continuous spectrum @xmath94 .
when there is a discrete component , those solutions may exist with frequency and higher harmonics inside the discrete spectrum @xmath93 but _ sufficiently _ different from the phonon frequencies @xmath11 . in the present section ,
we prove the existence of such periodic solutions . considering a time periodic solution of the uncoupled ( @xmath427 ) hamiltonian @xmath428 with frequency @xmath429 and period @xmath9
, we ask whether this solution persists as an exact time periodic solution at the same frequency when the coupling @xmath35 is turned on at least up to a nonvanishing value .
@xmath7 will be considered as a given number which has to be chosen in an appropriate set in order this solution persists .
we also need to assume some conditions concerning the time periodic solutions which are generally fulfilled in order to answer this question .
this time periodic solutions should belong to a one parameter family of time periodic solutions . thus , this family can be represented locally in action angle variables .
its energy @xmath430 only depends on its action @xmath431 ( at least defined in some interval ) , then @xmath432 is the frequency of the solution .
we need to assume that this solution is not marginal at the considered frequency @xmath7 that is @xmath433 when @xmath434 .
the possible existence of such time periodic solutions in complex systems is already known and for example may be based on earlier proofs of ( interband ) dbs ( for which the fundamental frequency and its harmonics are not resonant with the phonon band ) considered in @xcite . the equation of motion for a periodic solution @xmath435 with period @xmath436 of the anharmonic subsystem coupled to the collection of harmonic oscillators is ( see appendix [ appd ] ) @xmath437 with @xmath438 which should no be confused with @xmath439 .
@xmath440 is periodically continued for @xmath245 and @xmath441 .
a necessary condition for the existence of idb with frequency @xmath7 is that the periodic kernel @xmath442 is a bounded and smooth function in @xmath24 for given @xmath7 .
this condition will be discussed in the following . according to the hypothesis , eqs .
, , the absolutely continuous part becomes the integral of a smooth function of @xmath188 in the limit of @xmath25 to infinity : @xmath443 which diverges for @xmath444 if @xmath445 for some nonzero integer @xmath106 .
then , @xmath446 is not a smooth bounded function of time .
note , the divergence at @xmath277 ( @xmath447 ) is harmless because of the hypothesis eq . of finite reorganization energy .
thus , the contribution of the absolutely continuous spectrum , @xmath446 , is a smooth bounded function of time when @xmath448 this is the standard condition expected from earlier theories for the existence of dbs which assume the absence of resonance of the fundamental frequency and its higher harmonics with the absolutely ( continuous ) phonon spectrum .
it is consistent with the conclusions drawn from eq . in ( i ) of subsection 4.2 .
the contribution due to the discrete spectrum @xmath449 can not be transformed into the integral of a smooth function .
. clearly demonstrates the problem of small denominators which makes the proof of the existence of idb with frequency @xmath7 in presence of a dense discrete phonon spectrum nontrivial .
this problem of small denominator is quite similar to that we encountered in ( ii ) of subsection 4.2 . in the following
we will study this small denominator problem . for given constant @xmath450 and given @xmath7 , we define a closed interval or a finite union of closed intervals @xmath451 strictly included in the discrete spectrum @xmath93 such that for @xmath452 , @xmath453 @xmath454 is arbitrary except that it must be chosen small enough in order @xmath451 does not contain the edges of @xmath93 ( mobility edges ) .
this set @xmath451 contains the values of @xmath455 . then , for @xmath456 , the denominators @xmath457 are bounded from below by @xmath454 .
we split again the discrete kernel into two parts @xmath458 where @xmath459 is a smooth bounded function of time because of the absence of small denominators ( they are always larger than @xmath460 and therefore the sum converges because of the hypothesis of finite reorganization energy ( [ hypf1 ] ) ; the divergence at @xmath461 is harmless since @xmath462 for @xmath463 , as already mentioned before ) the small denominators are in the second part @xmath464 in order to insure the convergence of , we choose @xmath7 in a particular subset @xmath21 of the real positive axis defined such that for all @xmath465 , we have @xmath466 then , if @xmath467 , we have @xmath468 which is convergent according to the hypothesis . consequently , for @xmath469 , the discrete spectrum contribution @xmath470 to the kernel @xmath22 in eq . is a smooth bounded function of time .
we now have to prove that @xmath21 is not an empty set if we choose the parameter @xmath471 in an appropriate way .
first , we choose @xmath471 large enough in order @xmath472 can be defined ( uniquely ) for all @xmath473 by @xmath474 then , condition becomes @xmath475 which requires for all @xmath476 and all @xmath465 @xmath477\frac{\omega_{\nu}}{p+\epsilon_{\nu}},\frac{\omega_{\nu}}{p-\epsilon_{\nu}}\right [ \quad .
\label{nuset}\ ] ] @xmath21 is the complementary set @xmath478 . the total measure of the gaps of @xmath21 can be easily bounded from definition ( [ nuset ] ) for @xmath218 where @xmath451 is any interval or union of intervals strictly included in @xmath93 as @xmath479 since @xmath480 , we have @xmath481 , using definition and inequality . then using condition ,
we obtain @xmath482 where @xmath483 choosing @xmath471 large ,
the lebesgue measure of @xmath484 becomes small so that the measure of the cantor set @xmath21 tends to full measure in @xmath485 .
after having proved the existence of a set @xmath21 s.t .
@xmath486 is a bounded and smooth function of @xmath469 we now apply the implicit function theorem to prove the existence of idb .
we consider the banach space @xmath487 of t - periodic differentiable functions @xmath488 with continuous time derivatives which are time periodic @xmath489 and fulfill time reversibility : @xmath490 and @xmath491 .
the norm in @xmath487 is defined as @xmath492 + @xmath493 .
we consider the banach space @xmath494 of t - periodic continuous and time reversible functions @xmath495 with the norm @xmath496 .
let us consider for a given @xmath469 , the map @xmath497 defined as @xmath498 it maps @xmath487 in @xmath494 and is continuous and differentiable with respect to @xmath499 and @xmath35 .
the zeros of @xmath497 yield time periodic solutions of eq . .
at @xmath427 , let us assume that the anharmonic subsystem has a periodic and time reversible solution @xmath500 with period @xmath436 with @xmath469 . if @xmath501 is invertible at @xmath502 then this solution can be continued in @xmath35 in a neighborhood of zero up to a nonzero value .
we have @xmath503 the condition for invertibality of @xmath504 follows from the floquet analysis @xcite . setting the r.h.s .
of to zero yields the linearized equation of motion with respect to the periodic and time reversible solution @xmath505 of the anharmonic subsystem .
integration with respect to time over one period leads to the floquet matrix of the unperturbed db .
then , @xmath504 is invertible , if the floquet matrix has only a single eigenvalue which equals one .
since its corresponding eigenfunction is antisymmetric with respect to time reversion , this eigenmode is harmless .
we have assumed that a general macroscopic classical particle system with arbitrary interactions can be decomposed into a microscopic or macroscopic anharmonic subsystem with coordinates , @xmath506 , and anharmonic interaction energy @xmath507 , and a remaining macroscopic part with harmonic interactions . in order that the linear spectrum may possess a continuous part one has to take the thermodynamic limit of the harmonic subsystem .
the interaction between the harmonic displacement variables @xmath49 at site @xmath50 with @xmath508 is assumed to be bilinear with coupling constants @xmath509 .
@xmath35 is introduced for technical convenience . a realistic system for which these assumptions may apply is , e.g. a complex impurity embedded in a lattice where the latter can be treated in harmonic approximation .
the harmonic part can be diagonalized by introducing normal mode coordinates @xmath510 for eigenmodes with frequency @xmath11 .
then our model takes the form which we call a zwanzig - kac model @xcite for which the harmonic degrees of freedom can be eliminated .
this results in the integro - differential equation eq . for @xmath6 , which is of langevin type with memory kernels @xmath511 .
the properties of @xmath511 , or equivalently of the dissipation kernels @xmath512 , depend strongly on the spectrum of the harmonic part .
the elimination of a macroscopic number of degrees of freedom leads to fluctuating forces and ( as always believed ) to dissipation .
the former vanish at zero temperature .
then it is the common belief that @xmath6 for arbitrary initial conditions @xmath513 , @xmath514 converges to one of the local minima of the effective potential @xmath515 .
however , this is only true if the phonon spectrum of the harmonic part is absolutely continuous with infinite support . in general
, its support may be not infinite and in addition there may also exist a discrete spectrum , e.g. due to disorder .
it has been one of the major goals of our contribution to prove that energy dissipation strongly depends on both , the specific properties of the phonon spectrum and of the trajectories @xmath6 . for rather general conditions
we have succeeded to derive an exact formula for the energy dissipation rate @xmath285 of the anharmonic system averaged over a time interval of length @xmath9 . for @xmath516
it only involves @xmath517 , the product of the fourier transform of the dissipation kernels @xmath518 and of the velocity - velocity correlation functions @xmath519 . for a finite initial energy and a hamiltonian bounded from below it must be @xmath520
this is possible , only , if either @xmath521 , which holds for @xmath522 ( local minimum of @xmath523 for @xmath43 , or the support of the measures @xmath524 and @xmath525 have no overlap . besides the convergence of @xmath6 to one of the local minimum @xmath418 of @xmath419 , the nonoverlapping criteria allows _ possible _ asymptotic solutions which are periodic , quasiperiodic ( or almost periodic ) or may be even weakly chaotic .
if weakly chaotic asymptotic trajectories really exist , the corresponding measure @xmath416 must be singular continuous .
an absolutely continuous measure part of @xmath416 is ruled out , since due to a theorem ( proved in appendix c ) its support must be the full real line , which always has an overlap with the support of @xmath526 .
quasiperiodic ( or almost periodic ) aymptotic solutions can not exist as well , if @xmath527 has an absolutely continuous component . because , with @xmath397 and @xmath398 in the support of @xmath525 also @xmath528 with @xmath529 integers will be in that support . then there exist an infinite number of integer pairs ( @xmath400 , @xmath401 ) , such that @xmath399 has an overlap with the support of @xmath527 , provided @xmath530 is an irrational number .
quasiperiodic ( or almost periodic ) solutions may exist if @xmath526 has a discrete support , only .
we do not have a proof for this .
for a special class of models with disorder the existence of such solutions were proven rigorously @xcite .
the model studied in the present paper , may posses almost periodic solutions , as well .
periodic asymptotic solutions can occur when @xmath527 has a _ compact _ absolutely continuous , a singular continuous and/or a discrete component .
their existence is supported by the calculation of the dissipation rate @xmath285 for a sinusoidal force with frequency @xmath7 . in that case a zero measure set @xmath13 exists ( see appendix b ) , such that @xmath531 is bounded for all @xmath532 , the full measure complement of @xmath13 . in that case a fourier component of @xmath6 with frequency @xmath7 would be undamped .
that dissipation might be even more subtle has been demonstrated by the proof of the existence of a subset @xmath16 of @xmath13 ( see appendix b ) .
the zero measure set @xmath16 does not contain any of the phonon frequencies @xmath11 with probability one ( assuming @xmath11 and @xmath533 , @xmath534 as uncorrelated ) .
although @xmath17 is not in resonance with any phonon frequency there is dissipation .
depending on @xmath7 there may exist subdissipative and superdissipative behavior , i.e. @xmath8 increases , respectively , slower and faster than @xmath9 , where the linear @xmath9-dependence is the behavior of ordinary dissipation . whether @xmath8 follows a power law @xmath535 with @xmath536 ( subdissipative ) and @xmath537 ( superdissipative ) , as conjectured in ( ii ) of subsection 4.2 ,
is not obvious .
anyway , we have shown that dissipation can be rather complex .
+ the existence of nonrelaxing solutions also shows that the phonon bath can act as an anomalous thermostat .
depending on the initial conditions for @xmath6 , the perturbed anharmonic system may not relax to equilibrium , provided the initial energy is finite , i.e. of order @xmath538 in the total number of d.o.f .
there may even exist a sharp transition between relaxing and nonrelaxing situations .
as shown recently @xcite for a reduced fermi - pasta - ulam model with a single anharmonic bond embedded in an infinite harmonic chain there exists a critical excitation amplitude @xmath539 of the anharmonic bond such that the initial configuration relaxes to equilibrium for @xmath540 and to a discrete breather for @xmath541 . that such an incomplete energy spreading can occur
has also been proven for a general discrete nonlinear schrdinger equation in any spatial dimension and without and with disorder @xcite . another important goal of our paper has been to prove the existence of idbs . the existence of idbs with frequency @xmath7 has been shown by proving the smoothness and boundedness of the _ periodic _ kernel @xmath542 with respect to time @xmath24 . in that case , following reference @xcite , the implicit function theorem can be applied to the time periodic solutions at frequency @xmath7 of the uncoupled subsystem , i.e. at @xmath543 , in order to prove the existence of time periodic solutions at the same frequency @xmath7 for a nonzero value of @xmath35 .
the boundedness of @xmath544 is not obvious due to the occurrence of small denominators ( cf .
these small denominators make more explicit the resonance problem between the frequency @xmath7 and the harmonic mode frequencies @xmath11 ( which is also the basic problem , discussed in ref .
it is quite analogous to the small denominator problem in secular perturbation theory in classical mechanics .
we have proven that there exists a cantor set @xmath21 of finite lebesgue measure such that @xmath542 is smooth and bounded for all @xmath23 .
this implies that the periodic solutions of the isolated anharmonic subsystem with frequency @xmath23 can be continued up to a small but finite interaction strength between the anharmonic and harmonic subsystem
. therefore , there are idbs for all @xmath23 , provided @xmath35 is small enough .
the approach in ref .
@xcite leading to the same conclusion for a different class of models , is partly complementary to ours . the authors of ref .
@xcite consider a fully anharmonic system with quartic anharmonicity whereas our model consists of an anharmonic subsystem embedded in a harmonic lattice .
the anharmonicity can be rather general . for our class of model it is the elimination of the harmonic d.o.f . which finally leads to the small denominators in its explicit form of eq . .
there are several open problems which could be studied but which go beyond the present contribution .
_ first _ , the linear stability of the ibds is one of them . _
second _ , the existence of quasiperiodic or almost periodic solutions , which have already been proven for a special model @xcite , and particularly of weakly chaotic solutions of the present model could be investigated .
_ third _ , there is the question whether an arbitrary initial condition @xmath513 , @xmath514 with large enough energy does relax completely or not . if it does not , where does it converge to ?
since almost periodic solutions ( which correspond to the existence of kam tori ) should be linearly stable , i.e. all lyapunov exponents should be zero , one should exclude convergence to _ true _ tori .
thus , the limit solution could only be at the edge of the kam tori region with still discrete or perhaps singular continuous spectrum ? _ fourth _ , if both scenarios , i.e. complete and incomplete spreading ( depending on the initial condition ) exist , it will be interesting to study the existence of a sharp transition between both .
* acknowledgement : * this work has been started at the johannes gutenberg university in mainz ( germany ) .
one of us ( s.a . ) thanks for the financial support .
it has been extended and completed at the max - planck - institut fr physik komplexer systeme in dresden ( germany ) when both authors were visitors for a longer period of time .
we gratefully acknowledge the mpipks for financial support and its hospitality .
we also would like to thank jrg frhlich for his critical and helpful comments on this manuscript .
99 j.p .
hansen and i.r .
mcdonald , _ theory of simple liquids _ , third edition , academic press , london ( 2006 ) d. forster , _ hydrodynamic fluctuations , broken symmetry and correlation functions _ , benjamin , reading ( 1975 ) g.w .
ford , m. kac and p. mazur , _ j. math . phys . _ * 6 * , ( 1965 ) 504 - 515 r. zwanzig , _ j. stat
. phys . _ * 9 * , ( 1973 ) 215 - 220 a. j. sievers and s. takeno , _ phys.rev.lett . _ * 61 * ( 1988 ) 97 d.k .
campbell , s. flach and y.s .
kivshar , _ physics today _ * 57 * ( 2004 ) 43 - 59 s. aubry , physica * 103d * ( 1997 ) 201 - 250 s. flach and c.r .
willis , _ phys.rep .
_ * 295 * ( 1998 ) 181 s. flach , a. gorbach , _
phys.rep . _ * 467 * , ( 2008 ) 1 r.s . mackay and s. aubry ,
_ nonlinearity _ * 7 * ( 1994 ) 1623 - 1643 r.livi , m. spicci , r.s .
mackay , _ nonlinearity _ * 10 * ( 1997 ) 1421 - 1434 s. aubry , _ ann .
inst.h.poincar , phys .
* 68 * ( 1998 ) 381 - 420 s. aubry , g. kopidakis , v. kadelburg , _ discrete and continuous dynamical systems _ , * b1 * , ( 2001 ) 271 - 298 g. james , _
j.nonlinear sci .
_ * 13 * ( 2003 ) 27 - 63 c. albanse and j. frhlich , _ commun.math.phys . _ * 116 * ( 1988)475 - 502 c. albanse and j. frhlich , _ commun.math.phys . _
* 138 * ( 1991)193 - 205 c. albanse , j. frhlich and t. spencer , _ commun .
phys . _ * 119 * ( 1988 ) 677 - 699 g. kopidakis and s. aubry , _ physica _ * d130 * ( 1999 ) 155186 g. kopidakis and s. aubry , _ physica _ * d139 * ( 2000 ) 247275 s. gonzlez - prez - sandi , j. fujioka and b. a. malomed , physica * d179 * ( 2004 ) 86 - 100 k. yagasaki , a. r. champneys and b. a. malomed , nonlinearity * 18 * ( 2005 ) 2591 - 2613 j. frhlich , t. spencer and c.e .
wayne , _ j. stat .
_ * 42 * ( 1986 ) 247 - 274 t. cretegny and s. aubry _ phys.rev .
_ * b55 * ( 1997 ) r11929 h. kunz , _ commun.math.phys . _
* 78 * ( 1980 ) 201 ; j. frhlich , _
commun.math.phys . _
* 88 * ( 1983)151 ; ibid , _
commun.math.phys . _ * 101 * ( 1985 ) 21 ; f. delyon , _ commun.math.phys . _
* 100 * ( 1985 ) 463 w. rudin _
functional analysis_,2nd edit .
new york , mcgraw - hill ( 1991 ) p. hnggi , p. talkner and m. borkovec , _ review of modern physics _ * 62 * ( 1990 ) 251 - 341 m. kohmoto , b. sutherland and c. tang , _
phys.rev.lett . _
* 50 * ( 1983 ) 1870 s. ostlund , r. pandit , d. rand and h.j .
schellnhuber , _ phys.rev.lett . _
* 50 * ( 1983 ) 1873 d.r .
hofstadter , _ phys.rev . _ * b14 * ( 1976 ) 2239 e.a .
bredikhina , _ almost periodic functions _ in m. hazewinkel , _ encyclopaedia of mathematics _ , kluwer academic publishers ( 2001 ) a.j .
leggett , _ phys .
b _ * 30 * ( 1984 ) 1208 - 1218 a.j .
leggett , s. chakrovarty , a.t .
dorsey , m.p.a .
fisher , a. garg and w. zwerger , _ rev .
* 59 * ( 1987 ) 1 - 85 d. hajnal and r. schilling , phys .
* 101 * ( 2008 ) , 124101 g. kopidakis , s. komineas , s. flach and s. aubry , _ phys . rev
. lett _ * 100 * , 084103 ( 2008 ) i.s .
gradshteyn and i.m .
ryzhyk , _ tables of integrals , series and products _ ( academic , new york , 1992 )
we prove that ( if defined ) the limit of the square modulus of the fourier transform ( eq.([velfor ] ) ) of @xmath545 ( eq.([veldef ] ) ) is identical to the fourier transform of the matrix of autocorrelation functions of the velocities , but in the measure sense only .
since these fourier transforms might not be smooth functions , they should be considered as positive measures which may have a discrete or singular continuous part . in that case a direct analytical treatment would lead to convergence problems because the kernel of the involved integrals , may not vanish at infinity .
it is thus worthwhile to present a more rigorous treatment showing a relation not in terms of smooth functions , but of measure . for simplicity , we consider here only the case with a single component for the velocity vector but the proof would readily extend to cases with arbitrarily many components .
let us consider a smooth bounded function @xmath546 ( which could be for example the time dependant velocity we consider above ) and its fourier transform over a finite time @xmath9 weighted by @xmath547 @xmath548 this is a smooth function of @xmath188 for any finite @xmath9 .
its square modulus defines an absolutely continuous positive measure @xmath549 .
we must now assume that @xmath550 is defined in the sense of weak convergence of these measures @xcite .
the limit is a priori any kind of measure absolutely continuous , singular continuous or discrete .
this assumption means by definition that for any bounded continuous function @xmath551 ( for the uniform topology ) we have @xmath552 .
let us consider the subset of bounded continuous functions @xmath551 with a compact support for ensuring the existence of a smooth fourier transform @xmath553 which fulfills the condition @xmath554 and @xmath555 . then , @xmath556 @xmath557 since @xmath558 is bounded , this double integral is absolutely convergent in the limit @xmath315 , it readily comes out with the new integration variables @xmath24 and @xmath559 that for any @xmath560 smooth function @xmath561 , we have @xmath562 consequently , for most values of @xmath563 , the autocorrelation function @xmath564 is a well - defined function ( since @xmath565 is obviously bounded ) .
defining , the fourier transform measure @xmath566 , it comes out @xmath567 for all continuous bounded functions @xmath551 with compact support .
we can now take sequences of continuous function with compact support converging toward arbitrary continuous functions without compact support which proves the equality remains fulfilled for arbitrary continuous functions @xmath551 thus , we prove the equality of the fourier transform of the autocorrelation function @xmath568 considered as a measure with the limit at infinite time of the modulus square of finite time fourier transform of the trajectory averaged over time @xmath569 .
it appears from eq.[appa2 ] that assuming @xmath570 has a limit measure for @xmath315 or that the autocorrelation function ( [ autocorapp ] ) is defined are equivalent conditions .
a sufficient condition for the absolute convergence of series is the convergence of @xmath571 given an arbitrary convergent positive series @xmath572 with @xmath573 which decays slower than @xmath307 for example as a power law of @xmath123 such as @xmath574 ( with @xmath575 ) and given some positive integer @xmath471 , we can make that serie @xmath576 convergent if we choose @xmath7 such that @xmath577 and @xmath578 with @xmath579 .
this condition means @xmath580 where @xmath581 is the union of open intervals @xmath582 \omega_{\nu}-\frac{r_{\nu}}{k},\omega_{\nu}+\frac{r_{\nu}}{k } \right [ \right ) \bigcup \left ( \bigcup_{\nu } \left ] -\omega_{\nu}-\frac{r_{\nu}}{k},-\omega_{\nu}+\frac{r_{\nu}}{k } \right ] \right ) \label{setl}\ ] ] the lebesgue measure of this union of intervals is bounded as @xmath583 since @xmath307 is supposed to converge exponentially to zero and since @xmath584 has been chosen to go to zero as a power law , series @xmath585 is absolutely convergent for @xmath7 chosen in the complementary set of @xmath13 . does not have necessarily to decay exponentially but could also decay as a power law with an exponent @xmath190 strictly larger than @xmath586
this is sufficient for having @xmath587 . indeed
if we assume @xmath588 , then @xmath589 so that if @xmath590 this series can be made convergent with an appropriate choice of @xmath575 . ] consequently , we have @xmath591 .
@xmath581 is an open set since it is the union of open sets .
it is dense on the discrete spectrum @xmath93 since it contains @xmath126 for all @xmath123 which are dense in @xmath93 .
we have @xmath592 .
the set @xmath593 which is a countable intersection of open sets @xmath581 dense on @xmath93 , is not an open set but is also dense in @xmath93 ( according to baire theorem @xcite ) .
when @xmath594 , there exists @xmath471 such that @xmath595 , series ( [ freqd4 ] ) is absolutely convergent and remains bounded at infinite time thus , the energy transferred to the phonon bath remains finite and bounded at all time .
this set @xmath13 has zero lebesgue measure since @xmath596 for @xmath471 arbitrarily large .
thus , the complementary set @xmath14 has full lebesgue measure . by analogy with the theory of liouville numbers ,
we now prove that this set @xmath13 contains a ( zero measure ) subset @xmath597 for which series ( [ freqd4 ] ) surely diverges so that there is ( anomalous ) energy dissipation . we can also prove that this set @xmath598 does not contain any @xmath126 with probability one if we assume the distribution of eigenvalues @xmath126 is uniform enough on @xmath93 . actually , the probability distribution of these eigenvalues is nothing but the measure density of states @xmath599 .
we have to assume that this measure is absolutely continuous and moreover that the distribution probability of a pair of two eigenvalues @xmath126 and @xmath600 for @xmath601 is obtained as the product of probability distribution of single eigenvalues ( i.e. the eigenvalues are uncorrelated ) .
the energy transferred to the phonon bath given by series ( [ freqd4 ] ) diverges if its sequence of coefficients does not go to zero .
this situation surely occurs when for any given @xmath7 there is an infinite subsequence @xmath602 such that @xmath603 where @xmath471 is some positive constant .
this condition is equivalent to say @xmath604 where @xmath605 and @xmath606 is the union of open intervals @xmath607 \omega_{\nu } - k a_{\nu}^{1/2 } , \omega_{\nu } + k
a_{\nu}^{1/2}\right [ \right)\cup \left(\bigcup_{\nu \geq n } \left ] -\omega_{\nu } - k a_{\nu}^{1/2 } , -\omega_{\nu } + k a_{\nu}^{1/2}\right [ \right ) \label{setlp}\ ] ] @xmath606 is a dense open set on @xmath93 and @xmath608 for all @xmath25 .
thus the countable intersection @xmath609 is a dense set ( according again to baire theorem @xcite ) .
the lebesgue measure @xmath610 of this set @xmath598 is zero because @xmath611 which is the rest of an absolutely convergent series since @xmath307 decays faster than @xmath612 , since we assumed an exponential decay .
we have now to prove that giving an arbitrary @xmath126 , it could belong to this set @xmath598 only with probability @xmath253 . if @xmath613 we have @xmath614 , for any @xmath25
let us call @xmath615 the probability for @xmath126 to belong to @xmath616 . and
let us call @xmath617 , the probability that @xmath126 belongs to the unions of intervals @xmath606 .
we have @xmath618 .
we choose @xmath619 in order @xmath126 is not the center of an interval of @xmath606 .
then , since we assumed there is no correlation between the probability distribution of @xmath600 with that of @xmath126 the probability for @xmath126 to belong to @xmath606 is @xmath620 where @xmath621 is the density of states .
when @xmath25 goes to infinity , the lebesgue measure of @xmath606 goes to zero .
consequently @xmath622 has to be zero . since the set of @xmath126 is countable
, there is no @xmath613 with probability one .. and @xmath600 decay fast enough as @xmath623 diverges . ]
* lemma * _ let us consider a measurable subset @xmath624 of the real line with the following property : _
@xmath625 1 . _
then , either the lebesgue measure @xmath626 of @xmath627 is zero or @xmath628 is the whole real line . _ 2 .
_ there exists sets @xmath627 with property ( [ prop1 ] ) with zero lebesgue measure and which are uncountable _
since the measure of the real line is infinite , _ measurable _ means locally measurable , i.e. for any finite interval @xmath629 $ ] of width @xmath630 , the subset @xmath631 is measurable .
there exists many examples of sets having property ( [ prop1 ] ) : the set of integer multiples of a given number , the set of rational numbers . such sets are countable and thus have zero measure
. we shall give also examples of bigger sets which are not the real line , but uncountable .
this theorem proves that they must have at least zero lebesgue measure .
_ summary of the proof _ : we assume there exists a measurable set @xmath627 with the above property which is not the real line and such that there exists an interval in which it has a non zero lebesgue measure @xmath632 with @xmath633 . then
* we prove it has a uniform measure : @xmath634 only depends on the width of the interval @xmath630 and is proportional to this width ( statement 1 ) * for @xmath635 , we consider the sequence of translated sets @xmath636 for all integers @xmath637 and prove that there exists a finite integer @xmath106 such that @xmath638 .
then , @xmath639 with @xmath471 and @xmath640 integers and @xmath641 when @xmath642 and @xmath643 ( statement 2 ) * we consider the sequence of translated sets @xmath644 for @xmath637 and prove they are all disjoints .
then , @xmath645 should be infinite which is impossible since it is bounded by @xmath630 .
this contradiction proves the absence of measurable sets with the required properties and thus proves the theorem ( statement 3 ) .
let us assume @xmath627 is a measurable set with property ( [ prop1 ] ) and with a non vanishing measure .
we prove first that @xmath646 where @xmath189 is some non vanishing constant between @xmath253 and @xmath647 .
we choose two numbers @xmath648 and @xmath649 such that @xmath650 is an irrational number .
this is always possible because if all @xmath650 would be rational , then the set @xmath651 with @xmath652 would be a subset of the rational numbers which is countable .
thus , the measure @xmath653 would be zero which contradicts our initial assumption .
thus , we can find @xmath648 and @xmath649 with @xmath650 irrational .
then , the set of numbers @xmath654 is dense on the real axis ( weyl theorem ) .
thus for an arbitrarily chosen @xmath655 , we can choose a sequence @xmath656 in that dense set @xmath657 such that @xmath658 .
then , the sequence of translated sets @xmath659 is invariant .
we have obviously by translation @xmath660 .
then , for @xmath661 @xmath662 . consequently , @xmath663 does not depend on the origin @xmath664 of the interval @xmath665 but only on its width @xmath630 .
then , it comes out by splitting the intervals @xmath666 into arbitrary pieces that @xmath667 which readily implies that @xmath668 is proportional to @xmath630 independently of the origin @xmath664 of the interval .
_ nota _ : it is not trivial that the property @xmath669 for any @xmath670 and @xmath655 implies @xmath671 is linear . however , if @xmath671 is also monotonous ( for example increasing ) which is obviously true for our function @xmath672 as a function of @xmath630 , the proof that @xmath673 is linear is straightforward .
first , we prove @xmath673 when @xmath670 is rational and since this function is monotone and continuous , we also have @xmath674 for @xmath670 irrational .
let us assume now that @xmath675 which implies @xmath676 .
we consider the translated set @xmath677 .
we have @xmath678 . indeed ,
if we had a nonempty set as intersection , we should find @xmath648 and @xmath679 such that @xmath680 therefore , @xmath681 , in contradiction with the initial assumption on @xmath670 .
the measure of the translated set is identical to that of the original set since it has uniform measure : @xmath682 let us now consider the sequence of translated sets @xmath683 , ... , @xmath684 . if @xmath685 , there exists @xmath686 and @xmath679 , such that @xmath687 .
. then @xmath689 and @xmath690 are identical since @xmath691 .
conversely if @xmath692 , then @xmath693 . thus ,
if for any integer @xmath694 , @xmath695 , the series of sets @xmath696 are all disjoint .
the union @xmath697 , is a measurable set and @xmath698 is the sum of the measure of the disjoint components @xmath699 which is divergent .
this is impossible , because this measure must remain bounded by the measure of the whole interval , @xmath630 .
consequently , the assumption @xmath695 for any @xmath106 is wrong .
there are integers such that @xmath700 .
let us consider the smallest integer @xmath701 such that @xmath702 . then , @xmath703 and
if @xmath415 is written as the integer division @xmath704 with the rest @xmath705 , we have @xmath706 and for @xmath642we have @xmath707 . for @xmath708 and integer @xmath694
, we now consider the infinite sequence @xmath709 for @xmath710 .
we have @xmath711 since @xmath712 would implies @xmath713 which contradicts the assumption @xmath635 .
otherwise , we have @xmath714 for any @xmath715 .
it is @xmath716 .
if @xmath717 , we would have @xmath718 .
assuming @xmath719 , this is impossible because @xmath720 with @xmath721 and @xmath722 .
consequently the sequence of translated sets @xmath723 are all disjoint .
therefore , the measure @xmath724 is infinite and can not be smaller than @xmath630 this contradiction proves that @xmath627 can not have nonvanishing measure if it is not the whole real axis .
let us consider a sequence of positive numbers @xmath725 ( @xmath726 ) such that @xmath727 .
this property implies that @xmath728 goes to zero with @xmath729 faster than any exponential , that is for any @xmath730 , there exists @xmath731 such that @xmath732 @xmath733 is an example of such a sequence .
obviously , series @xmath734 is absolutely convergent .
then , let us now consider the set of real numbers @xmath627 defined as @xmath735 where @xmath411 is any bounded sequence of integers positive or negative that is there exists an integer @xmath25 such that @xmath736 . then
, we have @xmath737 where subsets @xmath738 are defined as @xmath739 with @xmath740 for all @xmath415 .
we have @xmath741 thus , assuming @xmath681 and @xmath742 , there exists @xmath743 and @xmath744 such that @xmath745 and @xmath746 . then for any integers @xmath747 and @xmath748
, we have @xmath749 which implies @xmath750 .
consequently , @xmath627 has property ( [ prop1 ] ) .
we now prove that @xmath627 has zero lebesgue measure , i.e. @xmath751 .
since @xmath627 is a countable union of subsets @xmath752 , it suffices to prove that @xmath753 for any @xmath25 . giving @xmath754 and
@xmath106 an arbitrary positive integer , the partial sum @xmath755 takes at most @xmath756 values .
the modulus of the difference @xmath757 is bounded as @xmath758 where @xmath759 .
then , @xmath752 is included in the union of @xmath756 intervals of width @xmath760 which have a total measure bounded as @xmath761 . since for any @xmath730
, there exists @xmath762 such that @xmath763 , we have @xmath764 .
choosing @xmath765 , the upper bound @xmath766 goes to zero for @xmath767 and thus can be made infinitely small .
consequently , the lebesgue measure @xmath768 of @xmath752 must be zero . which implies that @xmath769 has also lebesgue measure zero . for proving that @xmath627 is uncountable , it suffices to prove that one of its subset @xmath770 is uncountable
this set is defined as @xmath771 with @xmath772 for @xmath773 and @xmath774 or @xmath253 .
since @xmath775 has been assumed to go to zero for @xmath776 , integer @xmath106 can be chosen such that @xmath777 for any @xmath778 .
then for any @xmath779 , we have the strict inequalities @xmath780 .
this choice will ensure that all numbers in @xmath770 are surely distinct .
considering @xmath781 and @xmath782 where sequences @xmath783 are different , we define @xmath784 as the smallest @xmath415 such that @xmath785 .
then , we have @xmath786 .
since @xmath787 and similarly @xmath788 , we have @xmath789 . since @xmath790 and @xmath791 are different , we have @xmath792 which implies that @xmath670 and @xmath655 can not be equal . since , the set of infinite sequences @xmath793 with @xmath794 or @xmath253 for @xmath778 is obviously uncountable ( it can be mapped to the interval @xmath795 $ ] where numbers are represented in base @xmath586 ) , @xmath770 is uncountable as well as @xmath627 which proves the last statement of proposition ( 2 ) .
searching for periodic solutions with period @xmath407 | we investigate the dynamics of a macroscopic system which consists of an anharmonic subsystem embedded in an arbitrary harmonic lattice , including quenched disorder .
the coupling between both parts is bilinear .
elimination of the harmonic degrees of freedom leads to a nonlinear langevin equation with memory kernels @xmath0 and noise term @xmath1 for the anharmonic coordinates @xmath2 . for zero temperature , i.e for @xmath3
, we prove that the support of the fourier transform of @xmath4 and of the time averaged velocity - velocity correlations functions @xmath5 of the anharmonic system can not overlap . as a consequence , the asymptotic solutions can be constant , periodic , quasiperiodic or almost periodic , and possibly weakly chaotic . for a sinusoidal trajectory @xmath6 with frequency @xmath7
we find that the energy @xmath8 transferred to the harmonic system up to time @xmath9 is proportional to @xmath10 .
if @xmath7 equals one of the phonon frequencies @xmath11 , it is @xmath12 .
we prove that there is a zero measure set @xmath13 such that for @xmath7 in its full measure complement @xmath14 , it is @xmath15 , i.e. there is no energy dissipation . under certain conditions
@xmath13 contains a subset @xmath16 such that for @xmath17 the dissipation rate is nonzero and may be subdissipative @xmath18 or superdissipative @xmath19 , compared to ordinary dissipation @xmath20 .
consequently , the harmonic bath does act as an anomalous thermostat , in variance with the common belief that elimination of a macroscopically large number of degrees of freedom always generates dissipation , forcing convergence to equilibrium .
intraband discrete breathers are such solutions which do not relax .
we prove for arbitrary anharmonicity and small but finite coupling that intraband discrete breathers with frequency @xmath7 exist for all @xmath7 in a cantor set @xmath21 of finite lebesgue measure .
this is achieved by estimating the contribution of small denominators appearing for @xmath22 , related to @xmath0 .
for @xmath23 the small denominators do not lead to divergencies such that @xmath22 is a smooth and bounded function in @xmath24 . and |
the process of planet formation requires that small dust grains in the interstellar medium ( ism ) undergo a dramatic transformation , growing by more than 12 orders of magnitude in order to form terrestrial planets and the cores of gas and ice giants . as a very first step
, these @xmath7m - sized ism dust grains must increase their size and become macroscopic @xcite , a process that alters the optical properties of the dust grains considerably : as grains reach millimeter or larger sizes , the absolute value of the dust opacity , @xmath8 , decreases , and at the same time the power - law spectral index of the dust opacity , @xmath2 ( where @xmath9 ) , becomes smaller ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
since thermal dust emission at millimeter and centimeter wavelengths is ( mostly ) in the optically thin regime , the observed dust continuum emission will trace the bulk of the disk mass modulated by the dust opacity : @xmath10 , where @xmath11 is the dust mass surface density and @xmath12 is the planck function evaluated at the dust temperature @xmath13 . for optically thin dust ,
warm enough to be in the rayleigh - jeans ( r - j ) regime ( @xmath14 ) , it follows that @xmath15 .
thus , direct measurements of the dust emission spectrum ( i.e. , the spectral energy distribution , sed , at long wavelengths ) can be used to derive the value of @xmath2 , a method extensively employed in the literature .
disk - integrated measurements , from sub - mm to cm wavelengths , have shown that in most protoplanetary disks the value of @xmath2 tends to be lower than in the ism ( where @xmath16 , consistent with the presence of @xmath7m - sized dust grains or smaller in the interstellar medium , * ? ? ?
possible explanations for the low values of @xmath2 include the presence of regions of high optical depth , which will drive the dust emission spectrum to be close to @xmath17 , and thus a low @xmath2 would be inferred ( incorrectly ) .
however , @xcite has shown this would be only plausible for the brightest and most massive disks .
different composition and/or porosity could affect the optical properties of dust grains ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but for a sensible set of dust properties the most significant influence on the opacity spectral index arises from an increase in grain size @xcite
. thus , the difference in the dust opacity spectral index of disks , as compared to the ism , implies that growth of at least 4 orders of magnitude in size has taken place inside circumstellar disks ( for a review see : * ? ? ?
* ; * ? ? ?
but since most of these measurements are of the global disk properties , they provide limited knowledge of the evolution of the dust properties as a function of the distance to the central star ( hereafter referred to as radius ) .
theoretical analyses predict that the average dust grain size will change with radius , as dust grows and fragments while it is transported throughout the disk @xcite .
a fundamental limiting factor to the largest grain size is that of the radial drift problem . while solids move at keplerian velocities in the disk , the gaseous component moves at slightly sub - keplerian speed due to gas pressure support .
this velocity difference induces a drag in the large particles , which end up losing angular momentum and drifting radially to smaller and smaller orbits , until they are accreted onto the star and lost from the solid population @xcite .
this barrier will limit the maximum grain size allowed in a disk , with mm and cm - sized particles at tens of au from the star drifting inwards in timescale shorter than the disk lifetime @xcite .
thus , measuring changes in the dust properties as a function of radius is essential to investigate the effects of this barrier in the grain size distribution of disks .
only recently have measurements reached the required sensitivity and a sufficient lever arm in wavelength to characterize radial variations of the dust properties @xcite . with wavelength coverage that spanned an order of magnitude , from sub - mm to cm wavelengths , observational constraints in dust opacity were obtained for the disk around the young star as 209 @xcite . in this study ,
a gradient in the value of @xmath2 was found from @xmath18 at @xmath1920
au to @xmath20 at more than @xmath1980 au , while a constant value of the dust opacity throughout the disk failed to reproduce these observations .
furthermore , these results seem to agree with a population of dust grains limited by radial drift . here
we explore further this issue , by studying how the dust properties vary with radius in the protoplanetary disks that surround the young stars cytau , located in the taurus star forming region at a distance of @xmath21 pc @xcite , and doar25 , located in the l1688 dark cloud at a distance of @xmath22 pc @xcite .
doar25 and cytau are pre - main sequence stars of k5 and m1 spectral type , respectively .
both stars are quite young : cytau is @xmath23 myr old , while doar25 is about 4 myr old @xcite , and both present a significant emission excess over the stellar photosphere , from near - infrared to millimeter wavelengths .
doar25 has been imaged at sub - millimeter wavelengths with the submillimeter array @xcite , and cytau was imaged at the 1.3 and 2.8 mm bands with the plateau de bure interferometer @xcite . in this paper
, we triple the number of circumstellar disks for which our multiwavelength analysis has been used to measure any radial changes in the dust properties , particularly in the dust opacity .
the paper is structured as follows : first , we describe the observations and data calibration procedures in section [ obsdata ] , and we present observational results from these data in section [ observationalresults ] . in section [ modeling_emission ]
we discuss the model of the disk emission employed in this paper , followed by section [ results_modeling ] where observational constraints derived from this modeling are presented .
finally , in section [ discussion ] we discuss our results in the context of grain growth , comparing these observational constraints with the radial drift and fragmentation barriers for growth , as well as with other disks previously studied .
our findings and conclusions are presented in section [ conclusions ] .
we obtained multi - wavelength observations of the continuum emission from doar25 and cytau with three interferometers : the combined array for research in millimeter - wave astronomy ( carma ) , the sub - millimeter array ( sma ) , and the karl g. jansky very large array ( vla ) as part of the disks@evla collaboration .
cytau was observed at four different wavelengths : 1.3 mm , 2.8 mm , 7.1 mm , and 5.0 cm , with the highest angular resolution being @xmath24 ( 9.8 au ) at 7.1 mm .
doar25 was observed at five different wavelengths : 0.9 mm , 2.8 mm , 8.0 mm , 9.8 mm , and 5.0 cm , with the highest angular resolution being @xmath25 ( 12.5 au ) at 8.0 mm .
a summary of these observations and our data calibration procedure is presented below .
l c l c cy tau & 2007-nov-07 & c , 0.35 km & j0530 + 1331 + & 2009-jan-04 & b , 1.0 km & 3c 111 + & 2009-jan-05 & b , 1.0 km & 3c 111 + & 2010-sep-16 & d , 0.25 km & j0336 + 3218 + & 2010-sep-21 & d , 0.25 km & 3c 111 , j0336 + 3218 + & 2011-dec-12 & a , 2.0 km & j0336 + 3218 + cy tau & 2008-dec-11 & b , 1.0 km & 3c 111 + & 2010-feb-15 & a , 2.0 km & 3c 111,j0336 + 3218 + & 2011-dec-30 & b , 1.0 km & 3c 111 + & 2012-feb-08 & c , 0.35 km & 3c 111 + & 2012-feb-19 & c , 0.35 km & 3c 111 + doar 25 & 2010-jan-06 & b , 1.0 km & j1625 - 2527 + & 2010-jan-09 & b , 1.0 km & j1625 - 2527 + & 2010-nov-29 & a , 2.0 km & j1625 - 2527 + & 2010-nov-30 & a , 2.0 km & j1625 - 2527 + & 2010-dec-03 & a , 2.0 km & j1625 - 2527 + & 2010-dec-12 & a , 2.0 km & j1625 - 2527 + & 2011-apr-25 & c , 0.35 km & j1625 - 2527 + & 2012-jan-19 & c , 0.35 km & j1625 - 2527 + [ carma_table ] l c l c cy tau & 2010-nov-13 & c , 3.4 km & j0403 + 2600 + & 2011-apr-05 & b , 11.1 km & j0403 + 2600 + & 2012-oct-22 & a , 36.4 km & j0403 + 2600 + & 2012-oct-26 & a , 36.4 km & j0403 + 2600 + & 2012-oct-27 & a , 36.4 km & j0403 + 2600 + & 2012-oct-28 & a , 36.4 km & j0403 + 2600 + doar 25 & 2011-jan-23 & cnb , 11.1 km & j1625 - 2527 + & 2011-may-28 & bna , 36.4 km & j1625 - 2527 + & 2011-jun-15 & a , 36.4 km & j1625 - 2527 + cy tau & 2011-jul-23 & a , 36.4 km & j0403 + 2600 + doar 25 & 2011-jul-14 & a , 36.4 km & j1625 - 2527 + [ vla_table ] cytau and doar25 observations at 2.8 mm ( 107 ghz ) , and cytau observations at 1.3 mm ( 230 ghz ) were obtained with carma in the a , b , and c configurations , providing baseline lengths between 302000 m which cover spatial scales from @xmath26 down to @xmath27 at 2.8 mm , and between @xmath28 down to @xmath29 at 1.3 mm . a complete observing log of these observations can be found in table [ carma_table ] .
double - sideband single - polarization receivers were tuned to a frequency of 107 ghz in the case of 2.8 mm observations , and 230 ghz in the case of 1.3 mm observations . to optimize the continuum sensitivity , we configured all spectral windows in the correlator to the maximum possible bandwidth : 468.75-mhz - wide windows of 15 channels each , which provided 2.8 ghz of bandwidth for observations before the correlator upgrade in december 2009 , and 487.5-mhz - wide windows of 39 channels each , which provided 7.8 ghz of total bandwidth after the upgrade . the observing sequence interleaved observations of a complex gain calibrator ( generally 3 minutes in length ) with the science target . in c configuration
the calibrator - target cycle time was 1215 minutes , while in a and b configuration it was 510 minutes in order to track better the tropospheric phase fluctuations .
this observing sequence repeated throughout the track , usually from target rise to set .
in addition , a strong calibrator was observed to measure the complex bandpass , along with either a planet ( uranus , mars , or neptune ) or a secondary flux density calibrator ( either 3c273 or 3c84 ) that is monitored by the observatory .
the estimated fractional uncertainty in the absolute flux density scale is @xmath30% .
the carma data were calibrated using the multichannel image reconstruction , image analysis and display ( miriad ) software package @xcite .
each observation was calibrated separately .
malfunctioning antennas , receivers , and/or spectral windows were flagged , and updated antenna position corrections and line - length system corrections were applied .
observations of doar25 in the most extended ( a and b ) configurations used the carma paired antenna calibration system ( c - pacs ) to monitor the tropospheric delay fluctuations on eight of the outermost stations of the array , using adjacent 3.5 m antennas equipped with 1 cm receivers .
phase corrections derived from the c - pacs system are applied during post - processing . a complete description of c - pacs is presented by @xcite and @xcite .
cytau observations at 7.1 mm ( 42 ghz , q - band ) were made during a , b and c configurations of the vla , while doar25 observations at 8.0 and 9.8 mm ( 30.5 ghz and 37.5 ghz , ka - band ) were made during a , bna , and cnb configurations , providing baseline lengths for both targets between 60 m to 36 km .
these baseline lengths correspond to spatial scale coverage from about @xmath31 down to @xmath32 at 7.1 mm , and from about @xmath33 down to @xmath34 at 9.8 mm .
these observations , summarized in table [ vla_table ] , used dual - polarization receivers and two independently tunable basebands , where each baseband was configured to eight 128 mhz spectral windows of 64 channels , to provide the maximum continuum bandwidth per baseband ( 1 ghz ) . for the q - band observations , the two basebands were centered at 7.1 mm ( 42.5 ghz ) and 7.2 mm ( 41.5 ghz ) providing 2 ghz total bandwidth at 7.14 mm , while for the ka - band observations , the two 1 ghz basebands were centered at 8.0 mm ( 37.5 ghz ) and 9.8 mm ( 30.5 ghz ) . to distinguish the amount of free - free or non - thermal contamination present at millimeter and centimeter wavelengths , cytau and doar25 were observed at 5.0 cm ( 6.0 ghz ) in c - band .
these low - frequency vla observations were obtained during a configuration , with dual - polarization receivers and two independently tunable basebands , where each baseband was configured as for the high frequency observations with a total bandwidth per baseband of 1 ghz .
the two basebands were centered at 6.2 cm ( 4.8 ghz ) and 4.1 cm ( 7.3 ghz ) , in order to be able to infer the spectral slope of the emission at centimeter wavelengths .
the observing sequence interleaved observations of a complex gain calibrator ( 45 sec to a few minutes in length ) with science target observations , whose duration depended on the array configuration and observing frequency : target - calibrator cycle times in extended configurations were 1.55 minutes at high frequencies , 510 minutes in compact configurations .
the target - calibrator cycle time at c - band in a - configuration was 1015 minutes . a strong calibrator was used to determine the bandpass shape .
the absolute flux density calibration was determined from observations of a primary flux density calibrator ( 3c147 for cytau , 3c286 for doar25 ) .
models for the primary flux density calibrators are provided by the observatory .
the estimated uncertainty in the absolute flux density scale is 5% at c - band , and 10% at ka- and q - bands .
the vla data calibration was performed using the casa software package and a modified version of the vla calibration pipeline .
as with the carma observations , each observation was calibrated separately .
malfunctioning antennas , receivers , and/or spectral windows were flagged , as well as noisy channels at the edge of each spectral window .
the first few seconds at the beginning of each target observation were also flagged , along with any radio frequency interference ( especially an issue with the c - band data ) .
times of poor phase coherence , as measured on the gain calibrator scans , were also removed from the data . because the data were acquired over an extended time period , corrections for source proper motion or other systematic position offsets were applied prior to combining .
all vla data presented here were obtained as part of the disks@evla project ( ac982 ) .
doar25 observations at 0.9 mm ( 345 ghz ) were obtained with the sma between 2005 may and 2007 june ; these observations have already been presented by @xcite .
three different array configurations ( c , e , and v ) were used , providing baseline lengths between 8 - 500 m , corresponding to spatial scales from about @xmath35 down to @xmath27 .
double - sideband receivers were tuned to a local oscillator frequency of 340.755 ghz .
each sideband was configured to have 24 partially overlapping 104 mhz chunks , for a total continuum bandwidth of 4 ghz .
the observing sequence interleaved observations of a complex gain calibrator with science target observations twice as long .
the source - calibrator cycle was 8 minutes for the extended configuration , and @xmath3020 minutes for compact configurations .
bandpass and flux density calibrators were selected from different planets and satellites ( uranus , jupiter , saturn , titan , callisto ) , as well as strong quasars ( 3c454.3 , 3c279 ) , depending on their availability and array configuration .
the estimated uncertainty in the absolute flux density scale is @xmath36 .
the data were flagged and calibrated with the idl - based mir software package .
for each telescope we verified that the calibration from multiple observations obtained in different array configurations matched , to within the uncertainty in the absolute flux scale , by comparing the calibrated visibilities where they overlap in _ uv_-space .
the observations at 0.9 , 1.3 , 2.8 , 7.1 , 8.0 , and 9.8 mm were then averaged over the observing bandwidth , as follows . in the case of the 0.9 , 1.3 , 2.8 and 7 mm data , visibilities encompass a narrow range of frequencies over the entire bandwidth : @xmath37 at 0.9 mm , @xmath38 at 1.3 mm , @xmath39 at 2.8 mm , and @xmath40 at 7.1 mm .
these data were therefore averaged into a single wideband channel . in the case of ka - band vla observations ,
the two 1-ghz wide basebands are separated by @xmath41 ghz , so these data were averaged into two wideband channels : one centered at 30.5 ghz ( @xmath42 mm ) , the other centered at 37.5 ghz ( @xmath43 mm ) . calibrated observations from different array configurations
were combined to generate a single visibility file for each wavelength analyzed , in order to compare these interferometric observations with physical disk models .
synthesized maps at 1.3 , 2.8 , and 7.1 mm of the continuum emission from cytau are presented in figure [ maps_cytau ] .
each map extends @xmath44 , corresponding to 350 au at the adopted distance . at wavelengths of 1.3 mm , 7.1 mm , and 5.0 cm ,
natural weighting was used to maximize the sensitivity . at a wavelength of 2.8 mm , a briggs weighting scheme with a robust parameter of 0.5
was used in casa , to optimize a combination of resolution and sensitivity .
the resulting image properties and source photometry can be found in table [ table_image_cytau ] , from these measurements we infer an spectral index for cytau from 1.3 to 7.1 mm of @xmath45 .
the observations at 5 cm were used to estimate the contribution from processes other than thermal dust emission at 7.1 mm ( e.g. , chromospheric activity or thermal bremsstrahlung from photoevaporative disk winds driven by the central protostar ; * ? ? ? * ; * ? ? ?
* ) , which needs to be taken into account when modeling the dust emission from the disk .
however , cytau is not detected at 5 cm , thus we take the @xmath46 upper limit of 20 @xmath7jy and estimate the maximum possible contamination at 7.1 mm by assuming optically - thick free - free emission with spherical symmetry , for which the emission as a function of frequency , @xmath47 , is proportional to @xmath48 @xcite .
the maximum contamination at 7.1 mm thus corresponds to @xmath49 @xmath7jy , which represents only 4% of the total emission at this wavelength .
we note that the probable source of any such contamination will arise from very near the protostar , and would appear as an unresolved point source in the 7.1 mm data .
based on the 7.1 mm data themselves , using _
uv_-distances @xmath50 km ( corresponding to spatial scales smaller than @xmath51 ) , we constrain any unresolved point source to have a flux density of @xmath52 @xmath7jy ( some of which could be dust emission ) , consistent with our estimate of the potential contamination extrapolated from 5 cm .
synthesized maps at 0.9 , 2.8 , 8.0 and 9.8 mm of the continuum emission from cytau are presented in figure [ maps_doar25 ] .
each map extends @xmath53 , corresponding to 350 au at the adopted distance .
natural weighting was used in the imaging at wavelengths of 8.0 mm , 9.8 mm , and 5.0 cm , to maximize the sensitivity .
briggs weighting with robust parameters of 0.7 and 0.3 were used for imaging at wavelengths of 0.9 and 2.8 mm , respectively , to optimize the resolution and sensitivity at these wavelengths .
the resulting image properties and source photometry can be found in table [ table_image_doar25 ] , from these measurements we infer an spectral index for doar25 from 0.9 to 9.8 mm of @xmath54 .
c c c c c c c carma & 1.3 & 580 & @xmath55 & 83.3 & & @xmath56 + carma & 2.8 & 280 & @xmath57 & @xmath58 & & @xmath59 + vla & 7.1 & 7.8 & @xmath60 & @xmath61 & & @xmath62 + vla & 50 & 7.0 & @xmath63 & @xmath64 & & @xmath65 [ table_image_cytau ] c c c c c c c sma & 0.9 & 3500 & @xmath66 & 14.4 & & @xmath67 + carma & 2.8 & 270 & @xmath68 & 0.9 & & @xmath69 + vla & 8.0 & 11.0 & @xmath70 & @xmath71 & & @xmath72 + vla & 9.8 & 8.0 & @xmath73 & @xmath74 & & @xmath75 + vla & 50 & 5.9 & @xmath76 & @xmath77 & & @xmath78 [ table_image_doar25 ] as for cytau , the observations at 5 cm from doar25 were used to estimate the contribution from other than thermal dust emission at 8.0 and 9.8 mm . in the case of doar25
, @xmath79 cm emission is detected at a level of 6@xmath80 coincident with the star , with an integrated flux density of @xmath81 @xmath7jy .
assuming @xmath48 results in an estimated contamination of @xmath82 @xmath7jy at 8.0 mm , and @xmath83 @xmath7jy at 9.8 mm , which corresponds to 17% and 24% of the integrated emission at these wavelengths . a lower level of contamination is derived based only in the 8.0 and 9.8 mm data with _
uv_-distances @xmath50 km ( corresponding to spatial scales smaller than @xmath84 ) : an unresolved point source flux density of @xmath85 @xmath7jy is present in the 8.0 mm observations , while a flux density of @xmath86 @xmath7jy is constrained at 9.8 mm .
these estimates are consistent ( within the errorbars ) with the extrapolation from 5 cm .
however , we adopt the most conservative value ( the highest possible contamination ) in our analysis
. the brightness temperature of the emission at mm and cm wavelengths can be useful to discriminate between optically thin and optically thick emission .
for a medium at a physical temperature @xmath13 with no background radiation and an optical depth @xmath87 , the brightness temperature ( @xmath88 ) of the emission at a particular wavelength @xmath89 is related to @xmath13 as : @xmath90 ( see , e.g. , * ? ? ?
thus , in the optically thick limit ( @xmath91 ) the brightness temperature directly traces the temperature of the medium and is independent of wavelength : @xmath92 , while in the optically thin limit ( @xmath93 ) the brightness temperature will be much lower than the physical temperature of the medium : @xmath94 and will vary with wavelength according to the wavelength dependence of @xmath95 .
the data presented in 3.1 and 3.2 include high frequency observations and emission from the cold outer disk of cytau and doar25 , thus the rayleigh - jeans limit of @xmath14 might not be appropriate even at cm wavelengths . for this reason we calculate the brightness temperature of the emission using the full planck function without this approximation .
figures [ tb_profile_cytau ] and [ tb_profile_doar25 ] present azimuthally - averaged radial @xmath88 profiles for both disks at each observed wavelength , with the shaded region representing the @xmath96 constraint derived from the statistical uncertainty in the images . for a proper comparison , we convolved all observations to the same angular resolution ( i.e. , the lowest spatial resolution , which corresponds to the resolution of the 2.8 mm data ) .
in addition , we subtracted an unresolved point - source component for the long - wavelength observations at 7.1 mm ( cytau ) and at 8.0 and 9.8 mm ( doar25 ) , whose flux density is given by the amount of contamination from non - dust emission described in 3.1 and 3.2 . the observed brightness temperature profiles in figures [ tb_profile_cytau ] and [ tb_profile_doar25 ] decline with increasing distance from the star in both disks , becoming successively fainter as the wavelength of the emission
is increased .
if the observed emission were optically thick , the @xmath88 profiles would directly trace the physical temperature of the disk . for a vertically isothermal disk in this limit ,
the observed brightness temperature should then be the same at all wavelengths .
for both doar25 and cytau we observe a successively fainter @xmath88 profile with increasing wavelength , suggesting that for all but the shortest wavelength , the observed emission is consistent with being optically thin .
furthermore , in 5 we find that , for a flared disk in hydrostatic equilibrium , the midplane temperature is a factor of several higher than the observed brightness temperature of cytau and doar25 at _ all _ disk radii and at _ all _ wavelengths , thus indicating that the observed emission is optically thin rather than optically thick ( see section 4 for details of the modeling and 5.1 for the resulting midplane temperature profiles ) .
a colder midplane temperature could be found in the extreme case of a thin flat disk in the presence of no interstellar radiation .
this configuration results in the coldest midplane disk possible , since the opening angle of a thin disk is much smaller than for a flared disk , making the stellar heating less efficient .
note that a more settled disk interpretation is favored in the analysis of unresolved observations of cytau , as the sed from mid - infrared to far - infrared wavelengths is quite steep for this object ( see , e.g. , * ? ? ?
* ) . in the case of doar25 ,
spitzer irs spectra reveal a flat spectrum between 1030 microns consistent with a flared disk structure .
thus , we only consider the thin disk case for cytau .
we find a midplane temperature of @xmath97 k at 10 au , decreasing to a lower limit defined by interstellar radiation and cosmic ray heating ( assumed to be @xmath98 k ) at radii of 30 au and greater . only for this extreme case
is the observed @xmath88 profile of cytau consistent with the physical temperature of the disk midplane , and this is only for the shortest wavelength observations at 1.3 mm at radii @xmath1930 to 50 au . at longer wavelengths ( and at larger radii at 1.3 mm ) the brightness temperature is still below this cold midplane case , which indicates that the optical depth of the emission is less than 1 , even in this extreme case .
consequently , we will assume in the following analysis that the emission at 1.3 mm from cytau is optically thin , although there is a possibility of higher optical depth at 1.3 mm if the disk structure corresponds to a flat thin disk for this object .
, 56 au ) . ] , 80 au ) . ] in figures [ vis_profile_cytau ] and [ vis_profile_doar25 ] , we present the real and imaginary part of the visibility as a function of _ uv_-distance ( hereafter referred to as visibility profiles ) , for both cytau and doar25 , at each of the observed wavelengths .
each visibility was deprojected by the known inclination and position angle of the disk ( see section [ modeling_emission ] for details of the known disk geometry ) , before averaging into _
uv_-bins with a width of 40k@xmath89 . to compare the different wavelengths , each visibility profile is normalized by the measured flux density in the first _
uv_-bin , between 040 k@xmath89 . in figures [ vis_profile_cytau ] and [ vis_profile_doar25 ] , the black dotted line at a constant value of @xmath99 represents the observed visibility profile for an unresolved point source .
once a disk is resolved by interferometric observations , its visibility profile will decline from this reference line .
thus , the significant decline of the real part of the visibility profile at each wavelength demonstrates that the emission from both doar25 and cytau is resolved , at all the observed wavelengths .
furthermore , a compact source will have a shallower decline with _
uv_-distance than an extended source , whose decline will be quite steep .
thus , since the decline of the short - wavelength visibility profile ( 0.9 , 1.3 , and/or 2.8 mm ) is steeper than the long - wavelength visibility profile ( 7.1 , 8.0 and/or 9.8 mm ) , the disk emission observed at short - wavelengths is more extended than the disk emission observed at long - wavelengths , which has to be more compact .
hence , the differing visibility profiles demonstrate a wavelength - dependent structure in both the cytau and doar25 disks , similar to that found for the disks surrounding as 209 @xcite , cq tau @xcite , and several other stars in the taurus - auriga star - forming region @xcite . in the next section ,
we explain the wavelength - dependent structure of the emission from cytau and doar25 as radial variations of the dust properties across the circumstellar disk .
observations of cytau and doar25 were analyzed using a disk emission model that reproduces the radial brightness distribution at mm and cm wavelengths , equivalent to the disk model employed in the analysis of similar observations of the disk surrounding the young star as 209 @xcite . at @xmath100 , 8.0 , and 9.8 mm an additional point source at the center of the disk
is also included , to represent the potential contamination from sources other than thermal dust emission , as described in sections [ cytau_obs_res ] and [ doar_obs_res ] .
we adopt the two - layer disk approximation for the disk structure , first presented by @xcite . in this model , a flared , passively - heated disk is irradiated by the central star , with its structure characterized by two components : a disk surface layer that absorbs the stellar radiation , and a disk interior , which is opaque to the stellar photons .
dust in the surface layer will absorb the short - wavelength stellar emission and re - radiate it at longer wavelengths . as the disk interior absorbs the reprocessed emission , the inner regions of the disk can be efficiently heated .
we use the implementation of this radiative transfer problem as described by @xcite , where the temperature structure of the two - layer model is computed with the iterative method presented by @xcite .
the assumptions implicit in the two - layer disk model break down for the far outer regions of the disk , where the surface density of material drops significantly and even the disk interior becomes optically thin to the stellar radiation .
when this condition is met , we assume the same temperature for both the surface layer and disk interior .
the disk temperature smoothly decreases as a power - law function of radius down to a minimum of 10 k in the outer disk , where the dust is heated by the interstellar radiation field . and inclination of @xmath101 , derived from molecular line observations @xcite .
each bin has a width of 40 k@xmath89 , and has been normalized by the measured flux density at 040 k@xmath89 . filled circles and error bars of different color correspond to correlated real and imaginary part of the emission observed at different wavelengths ( 1.3 mm : blue circle , 2.8 mm : gray triangle , 7.1 mm : black square ) .
the continuous lines correspond to the best - fit disk emission model at each wavelength ( see modeling in 5.1 ) . ] and inclination of @xmath102 , as derived from the modeling of the shortest - wavelength emission .
each bin has a width of 40 k@xmath89 , and has been normalized by the measured flux density at 040 k@xmath89 . filled circles and error bars of different color correspond to the correlated real and imaginary parts of the emission observed at different wavelengths ( 0.9 mm : blue circle , 2.8 mm : gray triangle , 8.0 mm : black square , 9.8 mm : cyan diamond ) .
the continuous lines correspond to the best - fit disk emission model at each wavelength ( see modeling in 5.1 ) . ] for the mass surface density structure , @xmath103 , we adopt the similarity solution for a viscously - accreting keplerian disk subject to the gravity of a massive central object @xcite , which can be described by a power law combined with an exponential taper at large radii . if the disk viscosity , @xmath104 , is assumed to be a power - law with radius ( @xmath105 ) , then the solution for the disk surface density is time - independent @xcite .
we adopt the same parameterization of the similarity solution presented by @xcite : @xmath106 where @xmath107 corresponds to the surface density at radius @xmath108 , @xmath109 is the power - law exponent on the radial dependence of the viscosity , and @xmath110 is the characteristic scaling radius is directly related to the location in the disk where the mass accretion rate changes sign , the transition radius : @xmath111 . the transition radius defines a boundary for accretion or expansion , since for @xmath112 mass flows inward while for @xmath113 mass flows outward .
note that in the context of viscous disk evolution , the value of the transition radius will increase with time due to conservation of angular momentum : as matter is accreted onto the star the disk must expand to conserve total angular momentum .
we include its definition here since different authors will either use @xmath110 or @xmath114 in their preferred prescription for the disk surface density . ] .
this prescription for @xmath103 behaves as a power law for small radii , while at large radii it decreases smoothly in an exponential fashion .
the inner radius of the disk , @xmath115 , is defined as the radius at which the dust sublimates : @xmath116 which corresponds to @xmath117 au for cytau and @xmath118 au for doar25 , for a dust sublimation temperature of @xmath119 k. given that the angular resolution of our observations is more than an order of magnitude greater than @xmath115 , the value of this parameter does not affect our constraints on the disk structure . as in the analysis presented by @xcite , we compute the magnitude of the dust opacity , @xmath8 , using mie theory .
we adopt a grain population of compact spherical grains larger than @xmath120 m , in a power - law distribution of grain sizes : @xmath121 for @xmath122 .
grains in the disk interior are assumed to be composed of silicates , organic materials , and water ice , while grains in the disk surface are assumed to be depleted of ice .
we obtain optical constants for these grain materials from @xcite ( for silicates ) , @xcite ( for amorphous carbon ) , and @xcite ( for water ice ) .
recent studies have revised the solar abundance of oxygen @xcite , so although we assume the fractional abundances of these materials as given by @xcite , we reallocate the fractions of these astrophysical grains to account for the increased oxygen abundance .
this results in grains in the disk interior which are composed of 12% silicates , 44% organics and 44% water ice , while in the disk surface ( where water ice may not be present ) the fractional abundances correspond to 21% silicates and 79% organics . except for the normalization of the surface density at radius @xmath108 ( @xmath107 ) , the effect of dust grain composition in the disk modeling will produce a minimal effect in the derived parameters of the disk structure ( well within their uncertainties , as shown by
more specifically , although the dust composition does affect the absolute value of opacity , the fact that the dust emission is optically thin at long wavelengths ( 5.3 ) means that the derivation of the dust opacity spectral index @xmath2 is insensitive to the assumed absolute opacity value and hence dust composition .
this follows since the only frequency dependence when the emission is optically thin corresponds to @xmath123 . finally ,
the last parameters that determine the observed disk emission describe the geometry of the disk in the plane of the sky : inclination ( @xmath124 ) and position angle ( pa , measured from north to east ) . in the case of cytau , the disk geometry has been characterized from molecular line observations of co at @xmath125 resolution : @xmath126 , pa@xmath127 @xcite , which is what we adopted for this study .
for doar25 , there are no molecular line observations in the literature .
we therefore constrained its disk geometry from our modeling of the dust continuum observations at 0.9 mm , where the signal - to - noise ratio is the highest , finding @xmath128 and pa@xmath129 ( consistent with previous modeling by * ? ? ?
* ; * ? ? ?
we find the best - fit model to a single wavelength observation through @xmath130 minimization using a markov - chain monte carlo ( mcmc ) procedure ( see * ? ? ? * ) .
the @xmath130 probability distribution is sampled by varying the free parameters that define the surface density profile ( @xmath131 ) , since the inclination and position angle of the disk are fixed to the values reported above , and where @xmath108 has been fixed to a value of @xmath132 au .
a set of these parameters defines a model of the disk brightness distribution , with the addition of a point source at the disk center whose flux density corresponds to the estimated level of contamination from non - dust thermal emission , as inferred in section [ observationalresults ] . from this model emission
, we produce an image of the disk making sure that the significant spatial scales are covered .
we take the fourier transform of this image and obtain model visibilities sampled at the same locations in the _ uv_-domain as the original observations .
we compare model and observed visibilities by means of the @xmath130 , and construct 16 separate mcmc chains that sample the parameter space using a metropolis - hastings algorithm with a gibbs sampler .
these chains all converged to the best - fit model , which corresponds to the one that minimizes the @xmath130 and hence is the model that best reproduces our observations .
we note that this mcmc algorithm samples the posterior probability distribution function ( pdf ) of the parameters when chains have converged and reached equilibrium .
thus , to find best - fit values and confidence intervals for the parameters in our model we adopt a bayesian approach , marginalizing the resulting pdf over all but one parameter to obtain the probability distribution of the parameter of interest . from each parameter
pdf we find the region that contains 68.3%(@xmath96 ) , 95.5%(@xmath133 ) , and 99.7%(@xmath46 ) of all samples at equal probability to constrain each of the aforementioned parameters .
c c c c c & 1.3 & @xmath134 & @xmath135 & @xmath136 & 1.15 + 2.8 & @xmath137 & @xmath138 & @xmath139 & 1.14 + 7.1 & @xmath140 & @xmath141 & @xmath142 & 1.14 [ table_modeling_cytau ] c c c c c 0.9 & @xmath143 & @xmath144 & @xmath145 & 1.08 + 2.8 & @xmath146 & @xmath147 & @xmath148 & 1.07 + 8.0 & @xmath149 & @xmath150 & @xmath151 & 1.11 + 9.8 & @xmath152 & @xmath153 & @xmath154 & 1.12 [ table_modeling_doar25 ]
to model our cytau and doar25 observations , we first select the maximum grain size ( @xmath3 ) and grain size distribution slope ( @xmath155 ) that best reproduces the unresolved sed from sub - mm to cm wavelengths . for cytau
, these correspond to @xmath156 and @xmath157 mm , which results in an opacity spectral slope @xmath158 between 1.3 and 7.1 mm . for doar25 , the long - wavelength sed can be reproduced with @xmath156 and @xmath159 mm , resulting in an opacity spectral slope @xmath160 between 0.9 and 8.0 mm . note that while these values of @xmath3 and @xmath2 are consistent with the seds , they are relatively poorly constrained by the curvature of the long wavelength emission . in order to minimize the number of free parameters in this model , we assume a _ radially constant dust opacity _ throughout the disk and fit each wavelength separately .
our model then constrains the disk surface density ( @xmath103 ) and temperature profile ( @xmath161 ) at each wavelength ; these constraints will appear to be different at different wavelengths if the assumption of a constant dust opacity is not satisfied .
the same approach has been successfully employed to measure radial variations of the dust opacity in the disks surrounding ry tau , dg tau , and as 209 @xcite ; in the following sections we present our results for the circumstellar disks of cytau and doar25 .
the best - fit and @xmath96 constraints for each model parameter ( @xmath162 ) , obtained from their marginalized probability distribution function , are presented in tables [ table_modeling_cytau ] and [ table_modeling_doar25 ] . to demonstrate that these best - fit models are sensible representations of the observations
, we imaged the model and residual visibilities ( calculated by subtracting the model visibilites from the data ) in the same way as the original observations .
the resulting images are shown in figure [ maps_model_res_cytau ] and figure [ maps_model_res_doar25 ] for cytau and doar25 respectively .
additionally , for each of the best - fit models , we computed de - projected visibility profiles ( solid lines in figures [ vis_profile_cytau ] and [ vis_profile_doar25 ] ) . the adopted model is a good representation of the data at each separate wavelength , as no significant residual emission can be seen in the residual maps and the visibility profile for our best - fit model agrees with the observations at all wavelengths . .
contours start at @xmath163 , stepping by @xmath46 , where @xmath80 is the rms noise in each map ( table [ table_image_cytau ] ) . ] .
contours start at @xmath163 , stepping by @xmath46 , where @xmath80 is the rms noise in each map ( table [ table_image_doar25 ] ) . ] from each separate model fitting we characterize the disk surface density and temperature profile ( @xmath103 and @xmath161 ) at each wavelength .
figures [ cytau_structure ] and [ doar25_structure ] present the best - fit and @xmath46 constraints on the disk temperature , surface density , and optical depth ( @xmath164 ) , obtained from modeling each wavelength independently .
the temperature profiles inferred from observations at different wavelengths are very similar ( this is expected , since dust of different sizes in the disk should have reached radiative equilibrium and be at the same temperature ) . in particular ,
the best - fit temperature at each wavelength differs from the mean temperature profile by at most @xmath165 k for cytau and @xmath166 k for doar25 .
however , the surface density profiles that were inferred separately from each observation ( middle panels of figures [ cytau_structure ] and [ doar25_structure ] ) , are different at each observed wavelength .
this is clearly not physical ( the dust emission arises from the same disk at all wavelengths ! ) , and can not be reconciled through a global change in the dust opacity .
we conclude that we need to consider a change in the dust opacity as a function of location ( specifically , radius ) in the disk , in order to explain the observed dust emission .
this is explored further in [ beta_constraints ] below . finally , the right panels of figures [ cytau_structure ] and [ doar25_structure ] show the inferred optical depth , demonstrating that the dust emission is optically thin at all wavelengths for @xmath167 au ( well within our spatial resolution at all wavelengths for both disks ) . furthermore , the emission would be optically thin even if the mid - plane dust temperature were lower by a factor of a few .
since the dust emission is in the optically thin regime for both cytau and doar25 ( right panel of figures [ cytau_structure ] and [ doar25_structure ] ) , the observed emission will directly depend on three physical quantities : the dust opacity , the dust mass surface density , and the temperature , such that @xmath168 .
thus , our observations provide a constraint for the product @xmath169 , where the opacity , surface density , and temperature may vary with radius , but only the dust opacity is a function of the observed wavelength . our model fitting presented above which assumes a constant dust opacity with radius and where each wavelength is fitted separately
results in a wavelength - dependent @xmath103 , signifying that the assumption of a radially constant dust opacity is not warranted by the data . to reconcile these wavelength - dependent @xmath170 and @xmath171 ( figures [ cytau_structure ] and [ doar25_structure ] ) , as well as the wavelength - dependent disk structure seen in the normalized visibility profiles ( figures [ vis_profile_cytau ] and [ vis_profile_doar25 ] ) we require a change in the dust properties as a function of radial distance from the central star , for both the cytau and doar25 disks .
we focus on the interpretation that the spectral index of the dust opacity , @xmath2 , is changing with radius , which we obtain following the procedure described by @xcite and @xcite , and summarize here . since the constraints found in section [ sigma_temp_results ] reproduce the data well , these must be consistent with the _ true _ ( but unknown ) disk structure : @xmath172 where @xmath173 , @xmath174 , and @xmath175 are the true disk physical quantities , and the left - hand side of this equation corresponds to our multi - wavelength constraints for the disk structure .
thus , two different wavelengths , @xmath176 and @xmath177 ( and corresponding frequencies , @xmath178 and @xmath179 ) , can be related by @xmath180 assuming that at long wavelengths the dust opacity follows a power - law behavior ( @xmath181 ) , and @xmath182 corresponds to the spectral slope of the assumed constant dust opacity with radius , derived from the integrated sed used in the initial models . a useful prescription to infer radial variations of the opacity spectral slope ( @xmath183 ) in dual - wavelength observations
is derived from this last equation ( see also equation ( 9 ) from @xcite ) : @xmath184 \times \log \left [ \frac{\sigma_{\lambda_1 } b_{\nu_1}(t_{\lambda_1})/b_{\nu_1}(\overline t)}{\sigma_{\lambda_2 } b_{\nu_2}(t_{\lambda_2})/b_{\nu_2}(\overline t ) } \right ] \label{beta2}\ ] ] where the dust surface density and temperature depend on the distance from the star @xmath185 . equation [ beta2 ] implies that in logarithmic space @xmath186 is the slope of a straight line at points \{@xmath187}. note that if the emission is in the rayleigh - jeans ( r - j ) domain : @xmath188 , and thus the true dust temperature ( @xmath175 ) will not be needed to derive @xmath186 .
however , at cold temperatures in the outer disk and/or at short wavelengths , the r - j assumption breaks down and we require an estimate of @xmath175 . as mentioned in section [ sigma_temp_results ] , the temperature constraints from these multi - wavelength data are quite similar , thus we assume the true dust temperature to be the mean @xmath171 , shown as a dashed line on the left panels of figures [ cytau_structure ] and [ doar25_structure ] . note that a change in the estimate of @xmath175 will have a minimal impact in the @xmath186 constraint , as @xmath175 only appears in the ratio of planck functions evaluated at two different wavelengths in equation [ beta2 ] ( and close to
the r - j limit the dependence with @xmath175 is negligible ) . nevertheless , we tested two different assumptions where @xmath175 was equal to _ twice _ and _ half _ the mean of @xmath171 .
these substantial changes of @xmath189 only changed our constraints in @xmath190 by @xmath191% or less throughout the disk . ) as a function of radius , inferred from multi - wavelength observations of the cytau disk .
black line : best - fit @xmath192 ; colored areas : confidence interval constrained by our observations .
the vertical dashed - lines represent the smallest spatial scale probed at each wavelength , the arrow indicates that at 7.1 mm the smallest spatial scale is outside of the range plotted in this figure , corresponding to 42 mas or 6 au .
the errorbar at the bottom right indicates the systematic uncertainty on @xmath192 resulting from absolute flux scale uncertainty . ] ) as a function of radius , inferred from multi - wavelength observations of the doar25 disk .
black line : best - fit @xmath192 ; colored areas : confidence interval constrained by our observations .
the vertical dashed - lines represent the smallest spatial scale probed at each wavelength , the arrow indicates that at 8.0 and 9.8 mm the smallest spatial scale is outside of the range plotted in this figure , corresponding to 48 mas or 6 au at 8.0 mm , and 58 mas or 7 au at 9.8 mm .
the errorbar at the bottom right indicates the systematic uncertainty on @xmath192 resulting from absolute flux scale uncertainty . ]
since the result of the mcmc algorithm is a sampled posterior pdf for the parameter space , we can construct a pdf for the product @xmath193 at each radius and for each wavelength that has been modeled . to determine @xmath186 we use a monte carlo simulation , where the pdf of @xmath194 is randomly sampled at each wavelength
; a random sampling of @xmath186 will thus correspond to the slope of the line that intersects the points \{@xmath195 , @xmath196}. by obtaining a large number of random samples of @xmath194 we populate the pdf of @xmath190 at each radius .
we find the best - fit value of @xmath197 at a radius @xmath185 from the peak of this pdf , and we derive confidence intervals from the region of the distribution that contains 68.3% , 95.5% , and 99.7% of all samples at equal probability ( as described in section [ modeling_emission ] for the uncertainties of the model parameters ) .
these steps are performed for all radii sampled by the data , in order to obtain a best - fit value and inferred constraints on @xmath198 for cytau and doar25 .
figures [ beta_cytau ] and [ beta_doar25 ] present the observational constraint on @xmath2 as a function of radius , obtained for the circumstellar disks of cytau and doar25 .
the values of @xmath2 allowed by these multi - wavelength observations are significantly different from ( and below ) the ism value of the dust opacity slope ( @xmath199 ) for @xmath200 au in the case of cytau , and @xmath201 au for doar25 . furthermore
, a gradient on @xmath2 with radius is found in both disks .
this gradient can only be consistent with a constant value of @xmath2 when the uncertainties in @xmath192 are extended to @xmath202 .
we note that the uncertainty in the absolute flux density scale will potentially introduce an additional systematic offset in the constrained values of @xmath192 .
but because a fractional uncertainty in the flux density scale affects all radii equally at a particular wavelength , the overall shape and significance of the deviation in @xmath192 from a constant value are _ not _ affected by uncertainties in the flux density scale . given the 10 - 15% uncertainty in the absolute flux density scales for the three telescopes used , we estimate that the level of the systematic offset introduced to be @xmath203 for cytau and @xmath204 for doar25 , indicated by a vertical errorbar in the bottom - right corner of figures [ beta_cytau ] and [ beta_doar25 ] .
this systematic uncertainty is smallest for doar25 , because observations of this disk were obtained at wavelengths that are further separated than for cytau , reducing the uncertainty due to the increase in the lever arm used to infer @xmath2 .
as discussed in the introduction , the inferred changes in the dust opacity for cytau and doar25 may arise from changes in grain size , composition , or grain geometry .
@xcite explored the frequency dependence of the dust opacity for different materials which may have enhanced emissivity at long wavelengths and thus can create low values of @xmath2 with small grains .
he concluded that changes in composition are unlikely to explain @xmath205 and advocates for changes in the grain - size distribution to explain the common finding of @xmath206 in unresolved observations of protoplanetary disks ( for references see 1 ) . on the other hand , changes in grain porosity
can produce low values of @xmath2 ( see , e.g. , figure 11 of @xcite ) .
however , the low filling factor needed to reach @xmath205 requires quite large fluffy grains ( e.g. , larger than meter - size for filling factors of @xmath207 ) .
given the complexity of exploring changes in composition , geometry , and size of dust grains simultaneously , and given that the option of large grains can more simply reproduce the observed changes in @xmath8 , we assume that the radial variations of the dust opacity ( observed as changes in @xmath2 in the previous section ) are only caused by changes in @xmath3 with radius .
we now estimate the range of allowed maximum grain sizes at each location of the disk for cytau and doar25 .
as discussed by @xcite , the determination of @xmath192 depends on the assumption that the dust opacity follows a power - law with wavelength : @xmath181 .
this essential assumption might not be the case , as non - power - law dependence with wavelength in @xmath8 can be seen on dust populations which are limited to a maximum grain size ( @xmath3 ) of sub - millimeter and mm - sized particles ( see e.g. , * ? ? ?
thus , variations in the dust properties of these disks , particularly @xmath3 , should be directly constrained without this assumption . following the method outlined by @xcite , we find the best - fit dust opacity ( which depends on @xmath3 ) , along with the best - fit surface density , that is needed to reproduce the constraints on @xmath208 found in section [ sigma_temp_results ] . since the temperature profile at each wavelength ( @xmath209 )
is found to be similar at all wavelengths , we assume the dust temperature is equal to the mean @xmath209 at each radius ( see dashed line on left panel of figures [ cytau_structure ] and [ doar25_structure ] ) . at this point , we include the uncertainty in the absolute flux density scale in our modeling , which ranges between 10 - 15% for the three telescopes used .
this improvement to our method is important , as it introduces another source of error in the constraints obtained from the flux density at each wavelength , and impacts the error budget in the characterization of radial variations of the maximum grain size .
we note that the uncertainty in the flux density scale as a function of wavelength was not included in our previous analysis of as 209 ( e.g. , * ? ? ?
* ) . for each mcmc realization that reproduces the disk surface brightness as a function of radius ,
we draw a random number from a normal distribution centered at 1.0 and whose standard deviation ( @xmath80 ) corresponds to the flux scale uncertainty . depending on the wavelength
, this would correspond to @xmath210 or @xmath211 .
we then scale the surface density normalization of each mcmc realization by a different random number , ensuring that the uncertainty in the absolute flux density scale is reflected in the resulting probability distribution .
assuming the same dust composition as in section [ modeling_emission ] we place @xmath46 confidence limits for @xmath6 and @xmath174 , while exploring two representative grain size distributions : a _ steep _ distribution with @xmath156 and a _ shallow _ distribution with @xmath212 .
the chosen distributions are motivated by the resulting size - distribution of a collisional cascade @xcite and the measured grain size distribution in the ism @xcite , which are both characterized by @xmath213 , while a less steep distribution of @xmath212 is selected to observe the effects of a reduced total number of smaller particles .
these results are presented as the shaded regions in figures [ amax_cytau ] and [ amax_doar25 ] for cytau and doar25 , respectively .
we find that within @xmath214 au from the central protostar , grains have grown up to at least up to @xmath215 mm for cytau and @xmath216 mm for doar25 . the best - fit @xmath6 displays a gradient with radius , where smaller grains are present in the outer disk and larger grains
are present in the inner disk
. however , the rise of @xmath3 towards smaller radii strongly depends on the assumed slope of the grain size distribution .
this occurs because a _ shallow _ grain - size distribution will have more large grains ( by number ) than a _ steep _ distribution . since large grains
dominate the emission at these wavelengths , we can reproduce the observed emission with a maximum grain size that can be smaller for a shallow distribution than for a steep distribution .
this explains the difference in our @xmath3 constraints for the representative slopes @xmath212 and @xmath156 that we selected .
unfortunately , current observations do not have the necessary sensitivity to discern the value of the grain size distribution slope .
@xcite presented simple analytical forms for the variation of @xmath3 with radius for two different regimes : when a population of dust grains is limited by fragmentation of dust particles , whose differential velocity is driven by turbulence in the disk : @xmath217 and when a population of dust grains is limited by radial drift , as larger particles are removed from the population and accreted onto the star due to gas drag : @xmath218 here @xmath219 and @xmath220 are the gas and dust surface density profile , @xmath221 is the dust grain internal density , @xmath222 is the disk turbulence parameter as presented by @xcite , @xmath223 is the velocity at which a collision between dust grains will result in fragmentation rather than growth , @xmath224 is the local keplerian velocity , and @xmath225 is the local sound speed of the gas . employing our constraints on the disk surface density , which assume a constant gas - to - dust ratio of 100:1 over the entire disk , we compared our derived observational constraints on @xmath3 vs. radius with these grain growth models , adopting standard values for the turbulence ( @xmath226 ) , dust density ( @xmath227 g @xmath228 , from the @xcite abundances adopted in this paper ) , and fragmentation velocity ( @xmath229 m s@xmath230 * ? ? ? * ) .
also , we verify that mm - sized particles are in the epstein regime throughout the disk ( with a stokes number @xmath231 ) , which is a necessary condition for the use of equation [ drift_barrier ] as the size limit in radial drift . for this particular set of turbulence and fragmentation parameters , the theoretical @xmath3 curves for cytau and
doar25 are shown in figures [ amax_cytau ] and [ amax_doar25 ] , respectively , where the continuous line illustrates the maximum grain size for a drift - dominated population , and the dashed blue line correspond to a fragmentation - dominated population .
note that since our constraint on the surface density with radius depends on the adopted grain size distribution slope , @xmath155 , the drift and fragmentation barriers will depend on @xmath155 as well . in both disks and with this standard set of parameters , the fragmentation barrier under - predicts the maximum grain size achievable .
following equation [ frag_barrier ] , a better correspondence could be found if : ( 1 ) the disk turbulence were lowered to @xmath232 , or ( 2 ) the gas - to - dust ratio were higher than 100:1 by an order of magnitude , or ( 3 ) the critical velocity for fragmentation were higher than @xmath229 m s@xmath230 .
this last option seems unlikely if grains are compact , given that threshold velocities for fragmentation and/or bouncing of up to a few m s@xmath230 have been measured experimentally for compact centimeter and decimeter - sized grains ( see e.g. , @xcite for silicate grains , @xcite and for icy grains , and @xcite for a review ) . on the other hand ,
if grains are porous ( e.g. * ? ? ?
* ) , fragmentation could occur at higher threshold velocities .
however , porous grains have larger cross sections resulting in mass loss through collisional erosion rather than fragmentation , which similarly halts the growth of these grains @xcite .
now , the second option of a higher gas - to - dust ratio needs to be pursued with measurements of the total gas mass in these disks in particular , but this alternative also seems unlikely given the recent results in the taurus - auriga star forming region where low gas - to - dust ratios ( well below the ism standard of 100:1 ) are inferred for a large number of circumstellar disks by @xcite . finally , the first option of a quiescent disk of low turbulence ( @xmath233 ) for the disks around cytauand doar25(and also as 209 ) seems plausible , and would further augment grain growth .
we test this option by calculating the expected maximum grain size when the disk turbulence is an order of magnitude lower ( @xmath234 ) , and we find a better correspondence with our @xmath3 constraints on cytau and doar25 ( for this last disk we also increased the fragmentation velocity by 25% ) .
note that values of @xmath232 would still produce turbulent relative velocities high enough for fragmentation , a condition attained when @xmath235 , and confirmed for both cytau and doar25 .
the prediction of low turbulence should be tested with future observations of the turbulent linewidth in these circumstellar disks . at the same time
, the radial drift barrier roughly agrees with our @xmath3 inference in both disks .
for doar25 , the growth barrier imposed by a radial drift of macroscopic particles is within a factor of a few of the observational constraint , and the same occurs for cytau at least up to @xmath236 au .
most likely it is both mechanisms that are at play _ simultaneously _ in these disks .
finally , we note that at a radii @xmath237 au , the population of grains in cytau is above both barriers by a factor of few at least , and this is the case for both a shallow or a steep grain size distributions .
this result is readily appreciated in the cytau data , where millimeter - wave emission at 1.3 and 2.8 mm is observed at large distances from the star , indicating the presence of mm - sized grains at @xmath214 au .
if the gas surface density distribution were different from the dust surface density we measured , the pressure gradient term @xmath238 in equation [ drift_barrier ] would change , and thus a region of higher gas pressure , i.e. , a local maximum in @xmath239 , could trap mm - sized dust grains in the outer disk of cytau . in such a scenario , grains can overcome the radial drift barrier that is expected when the gas distribution just monotonically increases with radius .
the origin of these pressure enhancements is strongly debated : locally they may arise from turbulent eddies @xcite , while on a global disk scale they may emerge from rossby - wave instabilities @xcite , long - lived vortices , or from the gap opened by one or multiple planets . however , higher angular resolution observations of the distribution of small and large grains , as well as observations of tracers in the gas emission , are needed to understand what is the likely cause of these large grains still present in the outer disk of cytau .
over the past several years there have been multiple examples of circumstellar disks where the interpretation of radial variations of the dust properties is supported by resolved observations at mm and cm wavelengths @xcite . for example , in tw hya , @xcite found that a radially homogeneous distribution of small and large grains ( larger than @xmath240 @xmath7 m ) was not a suitable representation of the observations , and a steeper @xmath103 for the larger grains was required to fit these multi - wavelength dataset .
and recently @xcite studied a sample of disks and found that those with a steep midplane density gradient generally possess a smaller value of @xmath2 , indicative of dust growth and radial evolution .
the largest sample to date has been gathered by @xcite , where dual - wavelength observations of disks in taurus - auriga at angular resolutions between @xmath241 were presented .
this study found evidence for changes in the value of @xmath2 for several sources including cytau .
albeit limited by the angular resolution of the lowest frequency band , which corresponds to @xmath214 au resolution for their 2.7 mm observations of cytau , these authors infer a change from @xmath242 in the inner disk to high @xmath2-values in the outer disk of this young star .
however , given the closeness in frequency of the two wavelengths studied by @xcite , together with the lack of high spatial resolution at 2.7 mm , implies that their constraint is consistent with a constant value of @xmath2 at the @xmath46 level for this particular object . as shown in section [ beta_constraints ] , our increased wavelength coverage and high sensitivity allows us to discriminate radial variations of the dust emissivity index for cytau with high significance , where a constant value of @xmath2 is excluded at more than @xmath202 .
this arises because the uncertainty in the spectral index ( @xmath243 ) is directly related to the wavelength coverage and snr of the observations . in the case of dual - wavelength observations , simple error propagation results in an uncertainty of @xmath244 , which indicates that an efficient way to reduce the uncertainty in @xmath245 , and thus in @xmath2 , is to expand the wavelength coverage . to compare radial changes in the dust properties for different circumstellar disks
, we have gathered measurements from the literature that attempt to constrain @xmath192 , mostly from observations at the 1 and 3 mm bands .
since most of the available measurements barely resolve each circumstellar disk , we employ only two representative measurements of @xmath2 , one from the inner disk regions and another from the outer disk .
thus , for each published result we compile the value of @xmath2 and its @xmath96 uncertainty at the best resolution of the observations ( between @xmath246 au depending on each target ) to illustrate the inner disk constraint in @xmath2 , and at either the observed outer disk radius or at @xmath247 au , whichever is smallest , to represent the constraint of @xmath2 in the outer disk ( except for ry tau and dg tau , which at the outer disk radius have an unconstrained value of @xmath2 ; for these we select the outermost radii for which a @xmath2 constraint is still significant ) .
the compiled literature measurements are presented in the left panel of figure [ comparison_beta ] , where we have computed the slope in @xmath192 between the inner and outer disk for each star ( dashed lines ) , and the average slope in @xmath192 for all the disks analyzed in the literature ( grey shaded region ) .
the steepest change in the dust emissivity spectral index occurs for dg tau b ( where @xmath248 at 20au , increasing to @xmath249 at 120 au ; * ? ? ?
excluding the added uncertainty in @xmath2 due to the absolute flux scale uncertainty for the two millimeter - wave bands observed for dg tau b , the resulting change in @xmath192 for this object is comparable to the constraints found for cytau , doar25 , and as 209 , which are presented together on the left panel of figure [ comparison_beta ] .
interestingly , the mean @xmath192 profile from literature measurements is less steep than for the 3 disks analyzed in a consistent fashion .
this could be due to the fact that most previously published observations only include dust emission traced at millimeter wavelengths ( 1 and 3 mm bands ) , which may be optically thick and thus reduce the overall value of @xmath2 .
since the young stars presented in the left panel of figure [ comparison_beta ] boast some of the brightest disks observed in millimeter continuum emission , they represent a population of massive disks where high optical depth at millimeter wavelengths may be expected .
thus , including long - wavelength observations is critical for the calculation of @xmath192 , not only for the increased lever arm but also to avoid regions of high optical depth .
we note that the comparison between previous results in the literature ( @xmath250 ) and our multiwavelength analysis ( @xmath251 ) should ideally be performed over a similar wavelength range to avoid potential biases between different wavelengths that trace different grain sizes , and could explain some of the differences between @xmath252 and @xmath253 seen in figure [ comparison_beta ] .
finally , we note that cytau shows the steepest rise in @xmath2 with radius , indicative of a substantial change in dust properties over tens of au in this disk , while the smoothest rise in the emissivity spectral index occurs for as 209 in the inner disk and for doar25 in the outer disk .
however , these differences at not significant when the disks are compared at the @xmath46 level .
we present multi - wavelength observations of the protoplanetary disks surrounding the young stars cytau and doar25 .
observations at 0.9 , 1.3 , 2.8 , 8.0 and 9.8 mm spatially resolve the dust continuum emission from these disks , down to scales of tens of au . from observations obtained at 5 cm , we quantify the amount of continuum emission whose origin is not thermal dust emission .
we estimate the level of contamination for doar25 to be @xmath254% of the integrated emission at 8.0 and 9.8 mm , while for cytau we establish an upper limit of contamination of @xmath255% of the total emission at 7.1 mm .
although these levels of contamination may seem small , they need to be quantified when characterizing the dust emission spectrum .
the temperature and surface density profiles for both disks were constrained at all observed wavelengths shorter than 5 cm , and from this modeling , we find that a constant dust opacity with radius does not fit these observations .
this result is supported by the observational fact that the normalized visibility profiles of the dust continuum emission differ for different wavelengths : cm - wave observations trace emission from a compact disk structure , while mm - wave observations trace a more extended disk structure .
since our modeling indicates that the dust continuum emission is optically thin over the spatial scales explored ( and this is also warranted by the observed brightness temperature profiles in 3.4 ) , the observed emission spectrum will be directly proportional to the dust opacity spectrum , which allows us to infer radial variations of the dust opacity spectral index , @xmath192 .
we find that a constant value of @xmath2 is excluded by our observations at more than @xmath202 in both disks .
close to the central protostar , for @xmath256 au , the constraints on @xmath192 are of high significance and indicate a rapid change of the dust properties between the inner disk and the outer disk . in the outermost regions of the disk where the dust emission has lower snr , we can only set an lower limit for the dust opacity spectral index and we find that @xmath257 for cytau at @xmath5 au , while @xmath258 for doar25 at @xmath5 au .
we find that for the disks of as 209 , cytau , and doar25 , the profiles of @xmath192 are steeper than for the available @xmath192 constraints from the literature at the @xmath96 level @xcite , while these are all consistent at the @xmath46 level .
the increased wavelength coverage presented here and by @xcite is critical for obtaining constraints of high significance on @xmath192 . assuming that the changes in the dust properties discussed above arise only from changes in the maximum particle size of the disk grain size distribution , we derive radial variations of the maximum grain size , @xmath6 .
for the two different grain size distribution explored , @xmath212 and @xmath156 , we find a gradient in @xmath3 with increasing grain size at smaller radial distance from the star , in both the cytau and doar25 disks .
we compare the observational constraints in @xmath6 with theoretical expectations of the maximum grain size when a population of grains grows and evolves while being limited by fragmentation or by radial drift . for an assumed disk turbulence of @xmath259
, the fragmentation barrier does not seem to be the limiting factor in the growth of these grains , as we observe millimeter and centimeter dust particles at distances from the star where the fragmentation barrier should have depleted these large particles . however , our observations seem to be consistent with a low - turbulence disk midplane with @xmath260 in cytau and doar25 , as well as in as 209 @xcite . if disks are generally low - turbulence , then fragmentation may play a lesser role in shaping the grain size distribution of protoplanetary disks .
future constraints on disk turbulence will be necessary to assess this process . for the cytau disk ,
the population of grains inferred from our observations is above the growth barrier imposed by radial drift and fragmentation , with a difference of of a factor of a few at @xmath261 au radius .
we hypothesize that this difference could be readily explained if a region of higher gas pressure exists in the outer disk , and is keeping these millimeter - sized particles from drifting inwards .
future high resolution observations that trace millimeter dust grains , as well as the gas distribution , are needed .
observations with the atacama large sub - millimeter array ( alma ) will be extremely well suited for such investigations , as spatial resolutions matching the vla observations at 7 mm ( of @xmath24 or better ) are already possible @xcite .
finally , future observations of multiple disks at different stages of evolution ( classical vs. transitional ) , as well as different stellar mass or disk masses , will provide valuable information related to the first steps toward planet formation .
we thank the referee for valuable comments .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc . ongoing carma development and operations
are supported by the national science foundation under a cooperative agreement , and by the carma partner universities .
the sma is a joint project between the smithsonian astrophysical observatory and the academia sinica institute of astronomy and astrophysics , funded by the smithsonian institution and academia sinica .
part of this research was carried out at the jet propulsion laboratory , caltech , under a contract with the national aeronautics and space administration .
wilson , t. l. , rohlfs , k. , httemeister , s. 2009 , tools of radio astronomy , by thomas l. wilson ; kristen rohlfs and susanne httemeister .
isbn 978 - 3 - 540 - 85121 - 9
. published by springer - verlag , berlin , germany , 2009 | we present new results from the disks@evla program for two young stars : cytau and doar25 .
we trace continuum emission arising from their circusmtellar disks from spatially resolved observations , down to tens of au scales , at @xmath0 mm for doar25 and at @xmath1 mm for cytau .
additionally , we constrain the amount of emission whose origin is different from thermal dust emission from 5 cm observations . directly from interferometric data
, we find that observations at 7 mm and 1 cm trace emission from a compact disk while millimeter - wave observations trace an extended disk structure . from a physical disk model , where we characterize the disk structure of cytau and
doar25 at wavelengths shorter than 5 cm , we find that ( 1 ) dust continuum emission is optically thin at the observed wavelengths and over the spatial scales studied , ( 2 ) a constant value of the dust opacity is not warranted by our observations , and ( 3 ) a high - significance radial gradient of the dust opacity spectral index , @xmath2 , is consistent with the observed dust emission in both disks , with low-@xmath2 in the inner disk and high-@xmath2 in the outer disk . assuming that changes in dust properties arise solely due to changes in the maximum particle size ( @xmath3 )
, we constrain radial variations of @xmath3 in both disks , from cm - sized particles in the inner disk ( @xmath4 au ) to millimeter sizes in the outer disk ( @xmath5 au ) .
these observational constraints agree with theoretical predictions of the radial - drift barrier , however , fragmentation of dust grains could explain our @xmath6 constraints if these disks have lower turbulence and/or if dust can survive high - velocity collisions . |
this paper examines the structure of @xmath0-symmetric @xcite @xmath4 theory in @xmath5 dimensions and makes detailed comparisons between this theory and conventional @xmath6 theory .
this work is a natural sequel of our studies of @xmath0-symmetric @xmath7 theory @xcite .
we will see that the structure of the @xmath4 theory is much richer and more elaborate than that of conventional @xmath6 theory and also of the cubic theories .
this paper is organized as follows . in sec . [
s2 ] we review the main features of the conventional @xmath6 theory in @xmath5 dimensions in the presence of an external linear source @xmath8 .
the partition function for this theory is @xmath9 , where @xmath10 .
we analyze this theory using the method of steepest descents .
there is a critical value of @xmath11 : for @xmath12 all three saddle points are real , while for @xmath13 one saddle point is real and the other two form a complex - conjugate pair . in both cases
semiclassical approximations to the partition function and the corresponding green s functions are obtained . because @xmath1 , the one - point green s function vanishes even when @xmath13 .
( when @xmath14 , if @xmath13 symmetry breaking occurs . ) in sec .
[ s3 ] the corresponding analysis is performed for the @xmath0-symmetric theory in which @xmath15 and @xmath8 are replaced by @xmath16 and @xmath17 .
again , we find a critical value of @xmath11 . for @xmath12 , one of the three saddle points
is purely imaginary and the other two are @xmath0-conjugate . for @xmath13
all three saddle points lie on the imaginary-@xmath18 axis . when @xmath12 , new and surprising features appear .
we find that for @xmath19 only one saddle point contributes to @xmath20 $ ] , while for @xmath21 all three saddle points contribute . as a consequence ,
the partition function @xmath20 $ ] for @xmath19 differs from the partition function for @xmath21 and the green s functions obtained by taking the limit @xmath2 differ from those obtained in the limit @xmath3 .
thus , the @xmath0-symmetric quartic theory evades the properties of the bogoliubov limit for the construction of the green s functions .
( the bogoliubov calculation consists of taking derivatives of the vacuum persistence amplitude with respect to the external source @xmath8 , and then performing the @xmath22 limit , irrespective of the sign of @xmath8 . ) when @xmath13 , the green s functions obtained in the @xmath3 and @xmath2 limits are the same .
however , unlike the conventional quartic theory , the one - point green s function @xmath23 is nonvanishing even though @xmath5 ; the green s function has a purely imaginary value .
the nonvanishing of @xmath23 contradicts the mermin - wagner - coleman ( mwc ) theorem on the absence of symmetry breaking in @xmath1 dimensions @xcite .
we also find that when @xmath12 , the one - point green s function is nonvanishing .
section [ s4 ] gives a summary and conclusions .
in this section we review the ordinary @xmath6 theory in @xmath5 dimensions so that we can compare it later with the corresponding @xmath0-symmetric theory .
we begin by considering the generating functional for the green s functions , which is given by the ( normalized ) integral @xmath24=\sqrt{\frac{|m^2|}{2\pi}}\int^{+\infty}_{-\infty}d\varphi\,e^{-v(\varphi)}. \label{e1}\end{aligned}\ ] ] the potential is given by @xmath25 where @xmath15 is assumed to be positive to guarantee the convergence of ( [ e1 ] ) , but both @xmath11 and @xmath8 can take positive and negative values .
the normalization coefficient @xmath26 in ( [ e1 ] ) is chosen so that for the noninteracting theory @xmath27=\sqrt{\frac{|m^2|}{2\pi}}\int^{+\infty}_{-\infty}d\varphi\,e^{- \frac{m^2}{2}\varphi^2}=1.\ ] ] there is an implicit factor of @xmath28 in the exponent in the integral ( [ e1 ] ) . in this paper
we calculate the partition - function integral in the semiclassical approximation for which @xmath29 is small .
to do so we use the method of steepest descents @xcite .
we then identify the green s functions as coefficients of powers of @xmath8 in the expansion of the partition function @xmath20 $ ] . in order to evaluate the integral in ( [ e1 ] ) asymptotically in the limit @xmath30 , we consider the equation that determines the location of the saddle points : @xmath31 when solving ( [ e3 ] ) , we note that two different situations arise depending on the value of @xmath11 .
specifically , by defining the critical value @xmath32 , we see that the qualitative behavior of the three solutions of ( [ e3 ] ) depends on @xmath11 being larger or smaller than @xmath33 .
we begin by considering the case @xmath12 . for @xmath12
there is one real solution @xmath34 and there are two complex - conjugate solutions @xmath35 and @xmath36 to ( [ e3 ] ) . for small @xmath8 these solutions are @xmath37 for @xmath19 ( @xmath21 ) the solutions form a triangle with one vertex on the negative ( positive ) side of the real-@xmath18 axis and the other two vertices in the half - plane @xmath38 ( @xmath39 ) .
the integral in ( [ e1 ] ) is approximated by using laplace s method in the limit @xmath30 .
on the real axis , there is only one _ laplace point _ [ zero of @xmath40 , namely @xmath34 , which is a minimum of the potential @xmath41 .
thus , for all @xmath8 the partition function @xmath20 $ ] is approximated as @xmath42=\exp\left[-v(\varphi_0)-\frac{1}{2}\ln\left(v''(\varphi_0)/m^2\right ) \right].\ ] ] inserting @xmath34 into ( [ e4 ] ) and expanding in powers of @xmath8 , we get the formal taylor series @xmath43=1+\left(\frac{1}{m^2}-\frac{g}{2m^6}\right)\frac{h^2}{2}+\co(h^4).\ ] ] as a consequence of the normalization [ see ( [ e1 ] ) ] the coefficients of powers of @xmath8 in ( [ e5 ] ) give the green s functions of the theory . the first two connected green s functions @xmath44 and @xmath45 are @xmath46 these expressions are the usual perturbative result ( in powers of the coupling constant ) @xcite . for @xmath47
all solutions to ( [ e3 ] ) are real : @xmath48 where @xmath49 is the heaviside step function @xmath50 and @xmath51 note that @xmath52 always lies on the negative-@xmath18 axis while @xmath35 always lies on the positive-@xmath18 axis . for small values of @xmath8 in ( [ e6 ] ) we find that @xmath53 by examining the second derivative of the potential @xmath54 at these three saddle points , we find that on the real-@xmath18 axis , @xmath55 are two local minima and that @xmath34 is a local maximum of @xmath56 . for @xmath19
the absolute minimum is at @xmath52 while for @xmath21 the absolute minimum is at @xmath35 .
consequently , the contribution to the asymptotic behavior of @xmath20 $ ] in ( [ e1 ] ) from @xmath55 is given by @xmath57}{z[0]}=\exp\left[-v(\varphi_\pm)-\frac{1}{2 } \ln\left(v''(\varphi_\pm)/\mu^2\right)\right].\ ] ] observe that when @xmath19 ( @xmath21 ) the asymptotic behavior of @xmath20 $ ] is dominated by @xmath52 ( @xmath35 ) and that the contribution from @xmath35 ( @xmath52 ) is subdominant ( exponentially small ) . expanding the above formula in powers of @xmath8
, we obtain @xmath58}{z[0]}&= & \left(1+\delta_{h,0}\right ) \bigg[1 + \left(\theta(-h)\sqrt{\frac{6\mu^2}{g } } -\theta(h)\sqrt{\frac{6\mu^2}{g } } + \theta(h)\frac{1}{4}\sqrt{\frac{3 g}{2\mu^6 } } -\theta(-h)\frac{1}{4}\sqrt{\frac{3 g}{2\mu^6}}\right)h\nonumber\\ & + & \left(-\frac{1}{\mu^2}+\frac{6\mu^2}{g } + \frac{13}{32}\frac{g}{\mu^6}\right)\frac{h^2}{2}+\co(h^3)\bigg ] .
\label{e7}\end{aligned}\ ] ] this expression is discontinuous at @xmath59
. a doubling occurs at @xmath59 because at this special point there are no subdominant contributions ; that is , @xmath55 contribute equally to the semiclassical result ( [ e7 ] ) when @xmath59 . finally , we identify the connected green s functions @xmath60 and @xmath61 from ( [ e7 ] ) : @xmath62 note that for @xmath5 dimensions , the one - point green s function @xmath60 vanishes even when @xmath13 .
as is well known , when @xmath1 there is no spontaneous symmetry breaking because there is tunneling between the two minima @xmath35 and @xmath52 of the potential @xcite . in our case ( @xmath5 )
the vanishing of @xmath60 is due to the cancellation between the two contributions to @xmath20 $ ] coming from the vicinity of @xmath35 and @xmath52 .
moreover , @xmath61 does not have the usual perturbative behavior @xmath63 .
this is because there are contributions from two saddle points , which give a typical nonperturbative result @xcite .
let us compare the above result with the hypothetical @xmath64 case for which the volume factor suppresses the tunneling ( and so there is symmetry breaking ) .
under this supposition we would consider contributions from only one of the two vacua , @xmath35 or @xmath52 .
if we were to select @xmath52 , then @xmath65}{z[0]}$ ] would be replaced by @xmath66}{z^{(-)}[0]}$ ] .
thus , from ( [ e7 ] ) we would find that @xmath67 in ( [ e8 ] ) we observe the expected nonvanishing vacuum expectation value for @xmath18 , which is given by the tree - level result @xmath68 plus a perturbative correction .
correspondingly , in ( [ e9 ] ) we see that the propagator for the fluctuation around @xmath52 has the typical perturbative form @xmath63 . in the next section we will study the @xmath0-symmetric version of the theory and calculate the corresponding green s functions .
we will discover some intriguing new properties , different from those of the conventional @xmath6 theory .
in particular , we find that even though the formulas for the green s functions in ( [ e8 ] ) and ( [ e9 ] ) are only hypothetical , the correct formulas for the green s functions @xmath60 and @xmath61 of the corresponding @xmath0-symmetric theory can be obtained by analytic continuation of the formulas in ( [ e8 ] ) and ( [ e9 ] ) .
in this section we study the behavior of the massive @xmath6 theory in zero dimensions and with negative coupling @xmath16 ( @xmath69 ) .
we examine this theory in the presence of a linear source term @xmath70 in the potential @xmath41 : @xmath71 to this end , we calculate the partition function @xmath24=\sqrt{\frac{|m^2|}{2\pi}}\int_cd\varphi\,e^{-v(\varphi ) } , \label{e11}\end{aligned}\ ] ] where @xmath72 is a path of integration in the complex-@xmath18 plane for which the integral converges . in the present case
this means that @xmath72 terminates inside of two stokes wedges of angular opening @xmath73 , one centered about @xmath74 and the other centered about @xmath75 @xcite .
we also calculate the first two connected green s functions @xmath60 and @xmath61 . as in sec . [ s2 ] , we evaluate the integral in ( [ e11 ] ) in the steepest - descent approximation , which consists of localizing the integrand around a finite number of saddle points for which the real part of @xmath76 is maximal . on the steepest - descent contours
@xmath77 is constant . to locate the saddle points
we must solve the equation @xmath78 then , we must find the two steepest - descent paths in the complex-@xmath18 plane on which the saddle points are the maximum values of @xmath79 .
these paths are also constant - phase contours ( @xmath80 ) .
it is crucial to identify the constant - phase contour that can be deformed into the original integration path @xmath72 . equation ( [ e12 ] ) has three solutions : @xmath81 where @xmath82 and @xmath83 are real numbers given by @xmath84 and @xmath85 , and @xmath86 since the theory is @xmath0 symmetric , the solution @xmath35 , which has a positive - real component , is transformed into the @xmath52 solution , which has a negative - real part , by a reflection through the imaginary axis .
the solution @xmath34 lies on the imaginary axis .
we expand ( [ e13])([e15 ] ) for small @xmath8 to obtain much simpler expressions for the saddle points : @xmath87 note that for @xmath19 , @xmath35 and @xmath52 lie in the @xmath88 half - plane and for @xmath21 they lie in the @xmath89 half - plane ; for positive values of @xmath8 the solution @xmath34 lies on the negative - imaginary axis and for negative values of @xmath8 it lies on the positive - imaginary axis .
for @xmath59 the saddle points all lie on the real axis . to identify the steepest - descent curves , we let @xmath90 so that the real part of the potential in ( [ e10 ] ) is @xmath91 then , from the saddle - point equation ( [ e12 ] ) we have @xmath92 next , we construct the hessian matrix @xmath93 and evaluate it at each of the saddle points ( [ e13])([e15 ] ) .
the eigenvectors associated with the positive eigenvalues of this matrix provide the steepest - descent directions of @xmath94 . on a steepest - descent path
the contribution is localized at the saddle point , so we can use laplace s method to evaluate the integral . for the saddle point
@xmath34 , @xmath95 and @xmath96 [ see ( [ e13 ] ) ] and thus @xmath97 this matrix is already diagonal .
one can easily see that for @xmath12 the eigenvalue @xmath98 is positive , so the @xmath99 axis is the direction of steepest descent from @xmath34 .
for the other two saddle points @xmath55 , @xmath100 and @xmath101 and thus @xmath102 for @xmath103 and @xmath104 , the positive eigenvalue is @xmath105 and the corresponding unit eigenvector is @xmath106 the negative eigenvalue is @xmath107 with unit eigenvector @xmath108 for @xmath109 and @xmath110 the eigenvalues @xmath111 and @xmath112 and the corresponding unit eigenvectors @xmath113 and @xmath114 are obtained from the previous ones by making the replacement @xmath115 .
an inspection of ( [ e17 ] ) and ( [ e18 ] ) shows that for this case the positive eigenvalue is @xmath116 while the negative one is @xmath117 . for notational simplicity , in the following we label the two components of the unit vector @xmath118 using the letters @xmath119 and @xmath120 : @xmath121 .
also , we replace @xmath122 ( @xmath111 ) with @xmath123 . with this choice of parameters @xmath119 and @xmath120
we also get @xmath124 , @xmath125 , and @xmath126 .
as explained previously , for our calculation we are interested in the positive eigenvalues and the corresponding eigenvectors . with the notation
we have introduced the positive eigenvalue is @xmath123 in both cases ( @xmath35 and @xmath52 ) and the corresponding eigenvectors are @xmath127 and @xmath128 .
we now determine the constant - phase contours in the @xmath18 plane that are also the steepest - descent paths from the saddle points .
our goal is to deform the original path @xmath72 in ( [ e11 ] ) into a new set of steepest - descent contours passing through the saddle points ( [ e13])([e15 ] ) .
as explained earlier , on this new path @xmath77 is constant ( or , more precisely , piecewise constant ) , and the integral in ( [ e11 ] ) is approximated by a sum of gaussian integrals localized at the saddle points .
the constant - phase contours are determined by the equation @xmath129 the constant @xmath130 will take the values @xmath131 , @xmath132 , or @xmath133 when we evaluate ( [ e19 ] ) at @xmath35 , @xmath34 , or @xmath52 , respectively .
we now study the two cases @xmath19 and @xmath21 in turn . _
the @xmath19 case : _ we first consider the case @xmath19 .
figure [ f1 ] shows the saddle points @xmath34 , @xmath35 , and @xmath52 of ( [ e13])([e15 ] ) in the @xmath134 plane for a specific ( although irrelevant ) numerical choice of the parameters together with the constant - phase contours passing through these points .
the solid lines are the steepest - descent ( constant - phase ) contours associated with the three saddle points .
the phases are @xmath131 , @xmath133 , and @xmath132 , respectively .
-4 mm constant - phase contours in the @xmath134 plane for @xmath19 , with the arbitrary choice @xmath135 , @xmath136 , and @xmath137 .
the constant - phase contours associated with the saddle point @xmath34 are green , those associated with @xmath35 are blue , and those associated with @xmath52 are red .
the steepest - descent paths associated with these three saddle points are solid , and the steepest - ascent paths emanating from the saddle points are dashed .
the arrows indicate the _ down _ directions .
the original contour terminates in stokes wedges in the southwest and southeast quadrants of the complex plane .
therefore , for the case @xmath19 the only relevant saddle point is @xmath34 , and the original integration path must be deformed into the contour labeled @xmath72 . for this case the saddle points @xmath55 play no role in the asymptotic evaluation of the partition function.,title="fig:"]-4 mm in general , to evaluate the partition function @xmath20 $ ] of ( [ e11 ] ) in the saddle - point approximation
, we must identify the contour @xmath72 that passes through one or more of the saddle points and that at the same time is a steepest - descent path that terminates inside the stokes wedges in the southwest and southeast quadrants of the complex plane .
an inspection of fig .
[ f1 ] shows that when @xmath19 , of the three saddle points only one of them , namely @xmath34 , fulfills these conditions and that the solid curve labeled @xmath72 that passes through @xmath34 is the desired contour .
for @xmath138 we have @xmath139 [ see ( [ e19 ] ) ] , so the constant - phase contour @xmath72 is given by ( [ e19 ] ) with @xmath140 ; thus , the equation for the curve @xmath72 is @xmath141 this curve terminates in the middle of the stokes wedges . by evaluating the integral in ( [ e11 ] ) along this constant - phase contour @xmath72
, we obtain @xmath142=\sqrt{\frac{|m^2|}{2\pi}}\int_cd\varphi\,e^{-v(\varphi)}=\sqrt{\frac { \right)e^{-{\rm re}\,v[u , v(u)]}.\ ] ] expanding the real part of the potential around the saddle point ( @xmath95 , @xmath143 ) to second order in @xmath144 and @xmath145 , we get [ with @xmath146 @xmath147\sim{\rm re}\,v_0+\frac{1}{2}\left(u - u_0,v(u)-v_0\right ) \cdot h(\varphi_0)\cdot\left(\begin{array}{c } u - u_0\\ v(u)-v_0\end{array } \right).\ ] ] therefore , the saddle point approximation for @xmath20 $ ] is @xmath148=\exp\left(-{\rm re}\,v_0-\frac{1}{2}\ln\frac{\lambda_0}{m^2}\right),\ ] ] where @xmath149 and @xmath150 .
we expand the partition function @xmath20 $ ] in ( [ e20 ] ) in powers of @xmath8 ( @xmath151 ) and get @xmath152=1-\left(\frac{1}{m^2}+\frac{g}{2m^6}\right)\frac{h^2}{2}+\co(h^4).\ ] ] from this result we can obtain the one - point and two - point green s functions @xmath153}\left.\frac{d z[h]}{dh}\right|_{h=0}$ ] and @xmath154}\left.\frac{d^2 z[h]}{dh^2}\right|_{h=0}$ ] . finally , the connected green s functions are @xmath155 we emphasize that despite the negative sign of the quartic coupling constant ( @xmath156 ) ( which one might interpret as implying instability due to unboundedness below ) , the function @xmath157 is real and positive as a consequence of @xmath0 symmetry .
it is also worth noting that in this case the partition function ( [ e20 ] ) and the two - point green s function ( [ e23 ] ) has the form one would expect from a perturbative expansion in powers of the coupling constant @xmath15 . in other words ,
the saddle - point approximation to @xmath20 $ ] gives , as in the conventional positive - coupling constant theory , the usual perturbative expansion .
for this reason , we call the saddle point @xmath34 the _ perturbative _ saddle point .
let us compare the partition function ( [ e21 ] ) of the @xmath0-symmetric theory with the corresponding partition function ( [ e5 ] ) of the conventional theory for the @xmath12 case .
note first that the potential ( [ e10 ] ) of the @xmath0-symmetric theory is obtained from the potential ( [ e2 ] ) of the conventional theory by making the replacements @xmath158 and @xmath159 . by making the same replacements in the partition function ( [ e5 ] ) of the ordinary theory
, we obtain the partition function ( [ e21 ] ) of the @xmath0-symmetric theory .
thus , by analytically continuing the partition function of the conventional theory to imaginary values of @xmath8 and negative values of @xmath15 , the partition function of the corresponding @xmath0-symmetric theory is obtained .
we will see in the following that this simple connection between the two theories is lost when the partition function of one of the two theories gets contributions from more than one saddle point while the partition function of the other theory gets a contribution from only one saddle point . _ the @xmath21 case _ : when @xmath21 , the saddle points @xmath34 , @xmath35 , and @xmath52 and the constant - phase curves are obtained from those of the @xmath19 case by reflecting about the @xmath160 axis ( compare figs . [ f2 ] and [ f1 ] ) .
this has crucial consequences for the saddle - point approximation of @xmath20 $ ] in ( [ e11 ] ) .
indeed , fig .
[ f2 ] shows that the contour needed to evaluate the integral is quite different from the contour @xmath72 of fig .
as shown in fig . [ f2 ] , the integration path is now obtained by joining three constant - phase contours , @xmath161 , @xmath162 , and @xmath163 .
-4 mm constant - phase contours in the @xmath134 plane for the case @xmath21 with the arbitrary choice @xmath135 , @xmath136 , and @xmath164 .
as in fig .
[ f1 ] , the constant - phase contours associated with the saddle points @xmath34 , @xmath35 , and @xmath52 are colored green , blue , and red .
the steepest - descent paths associated with these three saddle points are solid , and the steepest - ascent paths emanating from the saddle points are dashed and the arrows indicate the _ down _ directions . as in the case @xmath19 ,
the original integration contour terminates in stokes wedges in the southwest and southeast quadrants of the complex plane .
however , unlike fig .
[ f1 ] , all three saddle points are relevant .
the original integration path must be deformed into three constant - phase contours labeled @xmath161 , @xmath162 , and @xmath163 , which are joined end - to - end.,title="fig:"]-4 mm for @xmath21 ( unlike the @xmath19 case ) all three saddle points contribute to the integral in ( [ e11 ] ) .
thus , along the path @xmath165 we decompose the integral as follows : @xmath142=\sqrt{\frac{|m^2|}{2\pi}}\left(\int_{c_-}\,d\varphi \,e^{-v(\varphi)}+ \int_{c_0}\,d\varphi\,e^{-v(\varphi)}+\int_{c_+}\,d\varphi\,e^{-v(\varphi ) } \right ) .
\label{e24}\ ] ] we consider each of these integrals in turn : for the integral along @xmath162 , the calculation is the same as for the @xmath19 case , and we obtain @xmath166 the curve @xmath163 is given by ( [ e19 ] ) , and using ( [ e14 ] ) we get @xmath167 we parametrize @xmath163 in terms of the variable @xmath168 , @xmath169 and @xmath170 and note that the vector @xmath171 is the unit vector tangent to @xmath163 in @xmath172 [ see ( [ e14 ] ) ] . expanding @xmath173 and @xmath174 to first order in @xmath168 , we get @xmath175 , @xmath176 , so the expansion of @xmath177 around @xmath35 to second order in @xmath168 is @xmath178\cdot h(\varphi_+)\cdot\left(\begin{array}{c}u_+(t)-u_+\\ v_+(t)-v_+ \end{array}\right)\nonumber\\ & = & { \rm re}\,v(\varphi_+)+\frac{1}{2}\lambda t^2,\nonumber\end{aligned}\ ] ] where we have used @xmath179 .
the integral along @xmath163 is then approximated by @xmath180.\end{aligned}\ ] ] last , we calculate the contribution to @xmath20 $ ] in ( [ e24 ] ) coming from the integral along the curve @xmath161 . inserting ( [ e15 ] ) into ( [ e16 ] ) and ( [ e19 ] ) , we find that @xmath181 and @xmath182 . around @xmath52 ,
the curve @xmath161 is approximated as @xmath183 and @xmath184 .
following the same steps as for the evaluation of the integral along @xmath163 , we find that @xmath185.\end{aligned}\ ] ] combining ( [ e25 ] ) , ( [ e26 ] ) , and ( [ e27 ] ) , and using the notation @xmath186 and @xmath187 , the final result for the saddle - point approximation to @xmath20 $ ] is @xmath188=\exp\left(-{\rm re}\,v_0-\frac{1}{2}\ln\frac{\lambda_0}{m^2}\right)+2 \left(a\cos \overline c+b\sin \overline c\right ) \exp\left(-{\rm re}\,\overline v-\frac{1}{2}\ln\frac{\lambda}{m^2}\right).\ ] ] the contribution to @xmath20 $ ] for @xmath21 that appears in the first term of ( [ e28 ] ) agrees with the result for @xmath20 $ ] in ( [ e20 ] ) for @xmath19 . however , in ( [ e28 ] ) an additional term appears , which comes from the contribution of the two additional nontrivial saddle points @xmath35 and @xmath52 .
these saddle points _ do not _ contribute to @xmath20 $ ] in the @xmath19 case .
we emphasize that the @xmath19 and the @xmath21 cases are different because the saddle - point contour passes only through the @xmath34 saddle point when @xmath19 ( see fig . [ f1 ] ) , while the saddle - point contour passes through all three saddle points , @xmath52 , @xmath34 , @xmath35 , when @xmath21 ( see fig . [ f2 ] ) .
thus , the @xmath19 and the @xmath21 cases can not be obtained from one another by simply changing the sign of @xmath8 in the partition function . the peculiar discontinuity in the partition functions for the @xmath19 case ( [ e20 ] ) and the @xmath21 case ( [ e28 ] ) only occurs in the @xmath0 symmetric theory and does not appear in the conventional @xmath6 theory . as a consequence , the green s functions obtained from ( [ e20 ] ) , which are the limit as @xmath3 of derivatives of @xmath20 $ ] , are different from the green s functions obtained from ( [ e28 ] ) , which appear in the limit @xmath2 .
to see this difference explicitly , we expand ( [ e28 ] ) in powers of @xmath8
. keeping only terms up to @xmath189 we get @xmath190}{z[0]}=1+\left(\sqrt{\frac{6 m^2}{g}}+\frac{1}{8m^2}\sqrt{\frac{3 g } { 2m^2}}\right)e^{-\frac{3m^4}{2g}}h-\left(\frac{1}{m^2}+\frac{g}{2m^6}\right ) \frac{h^2}{2},\ ] ] from which we read off the connected green s functions @xmath60 and @xmath61 : @xmath191 note that the two - point green s functions @xmath61 in ( [ e23 ] ) and ( [ e30 ] ) for the cases @xmath19 and @xmath21 are the same .
however , the one - point green s functions @xmath60 in ( [ e22 ] ) and ( [ e29 ] ) are different .
this difference is a violation of the usual bogoliubov limit @xcite for constructing the green s functions .
for the conventional theory , the green s functions are obtained as the @xmath22 limit of @xmath20 $ ] derivatives , independently of the sign of @xmath8 .
let us consider now the evaluation of @xmath20 $ ] in ( [ e11 ] ) for the case @xmath192 ( where @xmath193 ) .
as before , we first look for the saddle points and then determine the constant - phase contours for the saddle - point approximation of @xmath20 $ ] .
the saddle points are given by the solutions to ( [ e12 ] ) and are purely imaginary : @xmath194 where @xmath195 for any @xmath8 , @xmath52 lies on the negative - imaginary axis in the complex-@xmath18 plane , @xmath35 lies on the positive - imaginary axis , and depending on the sign of @xmath8 , @xmath34 lies on the positive - imaginary or negative - imaginary axis . expanding the saddle - point solutions in ( [ e31 ] ) in @xmath8 , we get @xmath196 next , letting @xmath90 , we find that the real part of the potential ( [ e10 ] ) is @xmath197 we then calculate the hessian matrix , which at the three saddle points ( [ e31 ] ) is diagonal : @xmath198 where @xmath199 is @xmath200 or @xmath201 , depending on which of the saddle points we consider . as before ,
the local directions of the steepest - descent paths from each of the saddle points are determined by finding the eigenvectors of the hessian matrix with @xmath99 and @xmath199 at the saddle points . because @xmath202 has a diagonal form , the eigendirections are parallel to the @xmath99 and the @xmath199 axes .
the signs of the two eigenvalues @xmath203 and @xmath204 are found by substituting @xmath205 , @xmath206 , and @xmath207 [ see ( [ e31 ] ) ] in ( [ e32 ] ) .
we find that @xmath208 is negative for @xmath34 and positive for @xmath35 and @xmath52 .
the opposite is true for @xmath204 . to find the constant - phase contours
we must solve the equation @xmath209 where @xmath56 is the potential in ( [ e10 ] ) and the constant @xmath130 is determined by evaluating @xmath77 at each of the saddle points @xmath35 , @xmath34 , and @xmath52 .
the real part of all three saddle points vanishes , so from ( [ e33 ] ) , we have @xmath140 for each saddle point .
factoring ( [ e33 ] ) gives a linear equation , namely @xmath210 , and a cubic equation for the constant - phase contours : @xmath211 the @xmath210 contour is just the @xmath199 axis and passes through all three saddle points @xmath35 , @xmath52 , and @xmath34 .
the cubic equation ( [ e34 ] ) provides the remaining three constant - phase contours , one for each of the three saddle points .
the constant - phase contours are plotted in fig .
[ f3 ] for a specific choice of the parameters . in the left ( right )
panel we consider the @xmath212 ( @xmath21 ) case .
observe that in both cases , the only path that terminates in the stokes wedges is the path that passes through @xmath52 .
this path is labeled @xmath161 .
-4 mm plane for the potential @xmath56 in ( [ e10 ] ) for the parameters @xmath213 , @xmath214 .
the left panel portrays the case @xmath19 for the specific value @xmath215 and the right panel portrays the case @xmath21 for the specific value @xmath216 .
one contour lies on the imaginary axis and the other three contours emerge horizontally from each of the three saddle points @xmath35 , @xmath52 , and @xmath34 .
the only relevant contour is the one that passes through @xmath52 because this contour terminates in the stokes wedges , which lie in the southwest and southeast quadrants of the @xmath134 plane .
this contour is labeled @xmath161 and is represented as a solid line .
all other contours are indicated by dashed lines . ]
plane for the potential @xmath56 in ( [ e10 ] ) for the parameters @xmath213 , @xmath214 .
the left panel portrays the case @xmath19 for the specific value @xmath215 and the right panel portrays the case @xmath21 for the specific value @xmath216 .
one contour lies on the imaginary axis and the other three contours emerge horizontally from each of the three saddle points @xmath35 , @xmath52 , and @xmath34 .
the only relevant contour is the one that passes through @xmath52 because this contour terminates in the stokes wedges , which lie in the southwest and southeast quadrants of the @xmath134 plane .
this contour is labeled @xmath161 and is represented as a solid line .
all other contours are indicated by dashed lines . ]
-4 mm laplace s method along the path @xmath161 approximates @xmath20 $ ] as @xmath217 , where @xmath218 and @xmath219 .
thus , the normalized partition function expanded in powers of @xmath8 is @xmath190}{z[0]}=1-\left(\sqrt{\frac{6\mu^2}{g}}+\frac{1}{4}\sqrt{\frac{3 g } { 2\mu^6}}\right)h+\left(\frac{1}{\mu^2}+\frac{6\mu^2}{g}+\frac{13g}{32\mu^6 } \right)\frac{h^2}{2}+\co(h^3).\ ] ] this equation determines the first two connected green s functions : @xmath220 note that the connected two - point green s function @xmath61 is precisely what one obtains from perturbation theory . in particular
, the squared mass of the fluctuation above the vacuum is @xmath221 , as in the conventional theory .
in contrast , @xmath23 is nonvanishing .
evidently , the @xmath0-symmetric theory evades the mwc theorem @xcite ( according to which there can not be spontaneous symmetry breaking below @xmath222 dimensions ) .
we have studied the zero - dimensional euclidean @xmath0-symmetric @xmath6 theory in the presence of an external source @xmath8 and have compared it with the corresponding conventional @xmath6 theory .
we calculated the partition function and the green s functions using the saddle - point approximation . as is the case for the conventional theory ,
the @xmath0-symmetric theory possess distinct phases depending on whether the squared mass @xmath11 is greater than or less than the critical value @xmath223 .
these phases are characterized by different green s functions .
furthermore , when @xmath12 , the @xmath0-symmetric theory exhibits two distinct subphases depending on whether @xmath3 or @xmath2 .
the subphases can be described as follows : for @xmath12 and for @xmath19 , the partition function and the green s functions of the @xmath0-symmetric theory can be obtained from the corresponding functions of the conventional theory by analytic continuation to imaginary values of the external source @xmath8 ( @xmath224 ) and to negative values of the coupling @xmath15 ( @xmath158 ) .
this connection between the two theories arises because in both cases the partition function receives contributions from one saddle point only ; the saddle point of the @xmath0-symmetric theory can be obtained from the saddle point of the conventional theory by an anticlockwise rotation of @xmath225 in the complex-@xmath18 plane . on the other hand , for @xmath12 and @xmath21 , while the partition function of the conventional theory is still dominated by the perturbative saddle point , the partition function of the @xmath0-symmetric theory receives contributions from the rotated perturbative saddle point and _ also from two nonperturbative saddle points_. this subphase behavior of the @xmath0-symmetric theory is a new and unexpected feature .
the green s functions obtained by performing the two limits @xmath3 and @xmath2 are different , and this is a clear violation of the usual bogoliubov calculation of the green s functions .
the @xmath0-symmetric theory differs from the conventional theory in other respects .
in addition to the behavior of the @xmath12 phase , new phenomena appear in the @xmath13 phase . spontaneous symmetry breaking in the conventional quartic theory is forbidden for @xmath1 due to tunneling between the degenerate minima ; this absence of symmetry breaking is due to the mwc theorem @xcite .
the mwc theorem certainly holds for @xmath5 and it implies that for the conventional theory the one - point green s function @xmath60 vanishes even when @xmath13 . in the context of the saddle - point approximation ,
the vanishing of @xmath60 is due to the cancellation of contributions to the partition function coming from the two nonperturbative saddle points .
thus , the @xmath0-symmetric theory evades the mwc theorem in much the same way that supersymmetry evades the coleman - mandula theorem , namely , by circumventing the assumptions needed to prove it .
however , for the @xmath0-symmetric quartic theory the dominant contribution to @xmath20 $ ] comes from just one saddle point .
thus , the @xmath13 phase of the @xmath0-symmetric theory is characterized by a nonvanishing @xmath23 when @xmath5 .
hence , we would expect that even when @xmath226 , in the @xmath0-symmetric @xmath6 theory a phase transition will occur that is not triggered by the usual spontaneous - symmetry - breaking mechanism .
moreover , we have demonstrated the surprising result that while @xmath227 in the conventional theory , @xmath228 when @xmath12 in the @xmath0-symmetric theory .
let us make a further comparison between the conventional and the @xmath0-symmetric quartic theories .
we have seen that when @xmath12 and @xmath21 , the partition function of the @xmath0-symmetric theory is not an analytic continuation of the corresponding partition function of the conventional theory .
this is because the @xmath0-symmetric @xmath20 $ ] receives contributions from three saddle points , while only one saddle point contributes to the partition function of the conventional theory . in the @xmath13 case ,
the situation is similar although the roles of the ordinary and the @xmath0-symmetric theories are exchanged .
the partition function of the conventional theory receives contributions from two saddle points , while @xmath20 $ ] of the corresponding @xmath0-symmetric theory gets contributions from only the one saddle point @xmath52 ( see fig .
we stress that a cancellation leading to @xmath227 occurs when there are two saddle points , but this is impossible when there is only one saddle point .
we have pointed out in sec . [
s2 ] in our discussion of the conventional theory that we can artificially induce spontaneous symmetry breaking by discarding the contribution to @xmath20 $ ] from @xmath35 . for @xmath1
this procedure is artificial because both the left and right saddle points contribute to @xmath20 $ ] so the theory can not exhibit spontaneous symmetry breaking .
of course , one of the two saddle points must be discarded only when @xmath64 .
we have entertained this possibility for the @xmath5 conventional theory in sec . [
s2 ] only to compare the green s functions in ( [ e35 ] ) with the analytic continuation ( a rotation in the complex-@xmath18 plane ) of the green s functions in ( [ e8 ] ) and ( [ e9 ] ) . to conclude , the structure of the quartic @xmath0-symmetric theory in zero dimensions is much richer than that of conventional quartic theory .
it is also far more elaborate than that of the cubic @xmath7 theory @xcite .
our immediate objective now is to extend the results for the @xmath0-symmetric quartic theory to the @xmath229 case , just as we did for the cubic theory . 100 for background on @xmath0 symmetry
see c. m. bender , rept .
phys . * 70 * , 947 ( 2007 ) ; c. m. bender , d. c. brody , j .- h .
chen , h. f. jones , k. a. milton , and m. c. ogilvie , phys .
d * 74 * , 025016 ( 2006 ) . c. m. bender , v. branchina , and e. messina , phys .
d * 85 * , 085001 ( 2012 ) . c. m. bender , v. branchina , and e. messina , phys .
d * 87 * , 085029 ( 2013 ) . the mermin - wagner - coleman theorem in statistical physics is proved in n. d. mermin and h. wagner , phys .
* 17 * , 1133 ( 1966 ) and in quantum field theory is proved in s. coleman , comm .
phys . * 31 * , 259 ( 1973 ) . c. m. bender and s. a. orszag , _ advanced mathematical methods for scientists and engineers _ ( mcgraw hill , new york , 1977 ) , chap
see , for instance , m. peskin and d. schroeder , _ an introduction to quantum field theory _
( addison - wesley , new york , 1995 ) .
see c. domb and m. s. green , _ phase transitions and critical phenomena _ ( academic , new york , 1976 ) ; s .- k .
ma , _ modern theory of critical phenomena _
( benjamin , reading , 1976 ) .
r. j. rivers , z. phys .
c * 22 * , 137 ( 1984 ) .
a. s. whitmann , in proc . of the 9th coral gables conference on fundamental interactions at high energy , coral gables , florida , 1972 , ed . by g. iverson , a. perlmutter , s. mintz ( plenum , new york , 1973 ) ; f. strocchi , _ elements of quantum mechanics of infinite systems _
( world scientific , singapore , 1986 ) . | a detailed study of a @xmath0-symmetric zero - dimensional quartic theory is presented and a comparison between the properties of this theory and those of a conventional quartic theory is given .
it is shown that the @xmath0-symmetric quartic theory evades the consequences of the mermin - wagner - coleman theorem regarding the absence of symmetry breaking in @xmath1 dimensions .
furthermore , the @xmath0-symmetric theory does not satisfy the usual bogoliubov limit for the construction of the green s functions because one obtains different results for the @xmath2 and the @xmath3 limits .
& pdflatex |
in 2006 , the babar collaboration reported the first observation of the @xmath3 baryon @xmath4 in the radiative decay @xmath5 , where the @xmath6 baryon @xmath7 was reconstructed in decays to the final states @xmath8 , @xmath9 , @xmath10 and @xmath11 @xcite .
the @xmath4 lies about @xmath12 above the @xmath7 , and is the last singly - charm baryon with zero orbital momentum observed experimentally @xcite . in 2008 , the d0 collaboration reported the first observation of the doubly strange baryon @xmath13 in the decay channel @xmath14 ( with @xmath15 and @xmath16 ) in @xmath17 collisions at @xmath18 tev @xcite .
the experimental value @xmath19 is about @xmath20 larger than the existing theoretical calculations ( see ref .
@xcite for a short review on the relevant literatures ) ; however , the cdf collaboration did not confirm the measured mass @xcite , i.e. they observed the mass of the @xmath21 is about @xmath22 , which is consistent with the existing theoretical calculations . by now , the @xmath6 antitriplet states ( @xmath23 , @xmath24 , and the @xmath6 and @xmath3 sextet states ( @xmath25 ) and ( @xmath26 ) have been well established ; while the corresponding @xmath27-wave bottom baryons are far from complete , only the @xmath28 , @xmath29 , @xmath30 , @xmath31 , @xmath32 have been observed @xcite .
those heavy baryons are particularly interesting for studying dynamics of the light quarks in the presence of a heavy quark . in the heavy quark limit
, the three light quarks form an @xmath33 flavor triplet , @xmath34 , two light quarks can form diquarks of a symmetric sextet and an antisymmetric antitriplet @xcite .
the light - cone qcd sum rules are a powerful theoretical tool in studying the ground state heavy baryons , they carry out the operator product expansion near the light - cone @xmath35 instead of the short distance @xmath36 , while the nonperturbative hadronic matrix elements are parameterized by the light - cone distribution amplitudes instead of the vacuum condensates @xcite .
the nonperturbative parameters in the light - cone distribution amplitudes are calculated with the conventional qcd sum rules and the values are universal .
in ref.@xcite , we assume the charm mesons @xmath37 and
@xmath38 with the spin - parity @xmath39 and @xmath40 respectively are the conventional @xmath41 states , and calculate the strong coupling constants @xmath42 and@xmath43 with the light - cone qcd sum rules , then take the vector meson dominance of the intermediate @xmath1 , study the radiative decays @xmath44 and @xmath45 . in previous works @xcite ( see also refs.@xcite ) , we have calculated the masses and the pole residues of the @xmath46 heavy baryons @xmath47 and the @xmath48 heavy baryons @xmath49 with the qcd sum rules . in this article , we extend our previous works to study the vertexes @xmath0 with the light - cone qcd sum rules , then assume the vector meson dominance of the intermediate @xmath1 , and calculate the radiative decays @xmath2 . in ref.@xcite ,
aliev et al study the radiative decays @xmath50 , @xmath51 and @xmath52 with the light cone qcd sum rules .
there have been many works dealing with the strong coupling constants of the pseudoscalar ( scalar ) octet mesons and vector nonet mesons with the baryons .
the @xmath53 , @xmath54 , @xmath55 and other strong coupling constants of the nonet vector mesons with the octet baryons have been calculated using the light cone qcd sum rules @xcite . in refs.@xcite ,
aliev et al study the strong coupling constants of the pseudoscalar octet mesons with the octet ( and decuplet ) baryons comprehensively . in refs.@xcite
, we study the strong decays @xmath56 , @xmath57 and @xmath58 using the light - cone qcd sum rules .
moreover , the coupling constants of the vector mesons @xmath59 and @xmath60 with the baryons are studied with the external field qcd sum rules @xcite .
the article is arranged as : in section 2 , we derive the strong coupling constants @xmath61 , @xmath62 and @xmath63 of the vertexes @xmath0 with the light - cone qcd sum rules ; in section 3 , the numerical result and discussion ; and section 4 is reserved for conclusion .
we parameterize the vertexes @xmath0 with three tensor structures due to lorentz invariance and introduce three strong coupling constants @xmath61 , @xmath62 and @xmath63 @xcite , @xmath64u^\mu(p ) \nonumber \\ & = & \epsilon_\mu\overline{u}(p+q ) \gamma^{\mu\nu}u_\nu(p ) \ , , \end{aligned}\ ] ] where the @xmath65 and @xmath66 are the dirac spinors of the heavy baryons @xmath47 and @xmath49 respectively , the @xmath67 is the polarization vector of the meson @xmath1 , and @xmath68 . in the following , we write down the two - point correlation function @xmath69 , @xmath70 where @xmath71 and @xmath72 , the @xmath73 are color indexes , the ioffe type heavy baryon currents @xmath74 and @xmath75 interpolate the @xmath46 baryons @xmath47 and the @xmath48 baryons @xmath49 respectively @xcite , the external vector state @xmath1 has the four momentum @xmath76 with @xmath77 .
basing on the quark - hadron duality @xcite , we can insert a complete set of intermediate hadronic states with the same quantum numbers as the current operators @xmath74 and @xmath75 into the correlation function @xmath78 to obtain the hadronic representation . after isolating the ground state contributions from the pole terms of the baryons @xmath47 and @xmath49
, we get the following result , @xmath79\left[m_{\omega_q^*}^2-p^2\right ] } + \cdots \nonumber \\ & = & \frac{\lambda_{\omega_q}\lambda_{\omega_q^ * } } { \left[m_{\omega_q}^2-(q+p)^2\right]\left[m_{\omega_q^*}^2-p^2\right ] } \left\{g_1 \left[m_{\omega_q}+m_{\omega_q^ * } \right]\!\not\!{\epsilon}\!\not\!{p } \gamma_5 q_\mu \right.\nonumber\\ & & -g_1\left[m_{\omega_q}+m_{\omega_q^ * } \right]\!\not\!{q}\!\not\!{p } \gamma_5\epsilon_\mu + g_2\!\not\!{q}\!\not\!{p } \gamma_5 p\cdot\epsilon q_\mu - g_2\!\not\!{q}\!\not\!{p } \gamma_5 q \cdot p \epsilon_\mu \nonumber\\ & & \left.-\frac{g_2}{2}\!\not\!{q}\!\not\!{p } \gamma_5 q^2 \epsilon_\mu - g_3 \!\not\!{q}\!\not\!{p } \gamma_5 q^2 \epsilon_\mu+\cdots\right\}+\cdots \ , , \end{aligned}\ ] ] where the following definitions have been used , @xmath80 the current @xmath75 couples not only to the spin - parity @xmath81 states , but also to the spin - parity @xmath82 states . for a generic @xmath83 resonance @xmath84 , @xmath85 where @xmath86 is the pole residue and
@xmath87 is the mass .
the spinor @xmath88 satisfies the usual dirac equation @xmath89 .
if we choose the tensor structures @xmath90 , @xmath91 , @xmath92 , the baryon @xmath84 has no contamination . in the following ,
we briefly outline the operator product expansion for the correlation function @xmath93 in perturbative qcd theory .
the calculations are performed at the large space - like momentum regions @xmath94 and @xmath95 , which correspond to the small light - cone distance @xmath35 required by the validity of the operator product expansion approach .
we write down the `` full '' propagator of a massive quark in the presence of the quark and gluon condensates firstly @xcite , @xmath96 + \cdots \ , , \nonumber\\ s_q^{ij}(x)&=&\frac{i}{(2\pi)^4}\int d^4k e^{-ik \cdot x } \left\ { \frac{\delta_{ij}}{\!\not\!{k}-m_q } -\frac{g_sg^{\alpha\beta}_{ij}}{4}\frac{\sigma_{\alpha\beta}(\!\not\!{k}+m_q)+(\!\not\!{k}+m_q ) \sigma_{\alpha\beta}}{(k^2-m_q^2)^2}\right.\nonumber\\ & & \left.+\frac{\pi^2}{3 } \langle \frac{\alpha_sgg}{\pi}\rangle \delta_{ij}m_q \frac{k^2+m_q\!\not\!{k}}{(k^2-m_q^2)^4 } + \cdots\right\ } \ , , \end{aligned}\ ] ] where @xmath97 and @xmath98 , then contract the quark fields in the correlation function @xmath69 with wick theorem , and obtain the result : @xmath99 \right.\nonumber \\ & & \left .
+ \gamma_5\gamma^\alpha s_q^{kk'}(-x)tr\left [ \gamma_\alpha s_{jj'}(-x)\gamma_\mu c\langle 0|s_i(0)\bar{s}_{i'}(x ) |\phi(q)\rangle^t c\right]\right\}\ , .\end{aligned}\ ] ] perform the following fierz re - ordering to extract the contributions from the two - particle and three - particle @xmath100-meson light - cone distribution amplitudes respectively , @xmath101 and replace the hadronic matrix elements ( such as the @xmath102 , etc . ) with the corresponding @xmath100-meson light - cone distribution amplitudes , then substitute the full @xmath103 and @xmath104 quark propagators into above correlation function and complete the integral in the coordinate space , finally integrate over the variable @xmath105 , we can obtain the correlation function @xmath69 at the level of quark - gluon degree of freedom . in calculation , the two - particle and three - particle @xmath100-meson light - cone distribution amplitudes have been used @xcite , the explicit definitions are given in the appendix .
the parameters in the light - cone distribution amplitudes are scale dependent and are estimated with the qcd sum rules @xcite . in this article ,
the energy scale @xmath106 is chosen to be @xmath107 . taking double borel transform with respect to the variables @xmath108 and @xmath109 respectively
, then subtract the contributions from the high resonances and continuum states by introducing the threshold parameter @xmath110 ( i.e. @xmath111}\int_0^{s_0 } ds s^{n-1}e^{-\frac{s}{m^2}}$ ] ) , finally we obtain six sum rules for the strong coupling constants @xmath112 , @xmath113 and @xmath114 @xmath115 respectively , the explicit expressions are presented in the appendix .
the input parameters are taken as @xmath116 , @xmath117 , @xmath118 , @xmath119 @xcite , @xmath120 , @xmath121 , @xmath122 , @xmath123 , @xmath124 @xcite , @xmath125 , @xmath126 , @xmath127 , @xmath128 , @xmath129 , @xmath130 , @xmath131 , @xmath132 , @xmath133 , @xmath134 , @xmath135 , @xmath136 , @xmath137 , @xmath138 , @xmath139 , @xmath140 , @xmath141 , @xmath142 , @xmath143 , @xmath144 , @xmath145 @xcite , @xmath146 , @xmath147 , @xmath148 @xcite , @xmath149 , @xmath150 , @xmath151 , @xmath152 , and @xmath153 at the energy scale @xmath154 @xcite .
the threshold parameters and the borel parameters are taken as @xmath155 and @xmath156 in the charm channels , and @xmath157 and @xmath158 in the bottom channels , which are determined by the two - point qcd sum rules to avoid possible contaminations from the high resonances and continuum states @xcite . in refs.@xcite ,
melikhov et al study the ground - state form - factor in an exactly solvable harmonic - oscillator model to illustrate the exact effective continuum threshold for vacuum - to - hadron correlation function is very difficult to obtain , as the effective continuum threshold maybe depend on the borel parameter .
we show the values of the strong coupling constants @xmath61 , @xmath62 and
@xmath159 with variation of the threshold parameters @xmath110 in fig.1 . from the figure we can see that in the present case the numerical results are insensitive to the threshold parameters . ,
@xmath62 and @xmath159 with variation of the threshold parameters @xmath110 , the borel parameters are taken to be the central values ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the threshold parameters @xmath110 , the borel parameters are taken to be the central values ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively . ,
title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the threshold parameters @xmath110 , the borel parameters are taken to be the central values ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the threshold parameters @xmath110 , the borel parameters are taken to be the central values ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively . ,
title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the threshold parameters @xmath110 , the borel parameters are taken to be the central values ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the threshold parameters @xmath110 , the borel parameters are taken to be the central values ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the borel parameter @xmath160 ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the borel parameter @xmath160 ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the borel parameter @xmath160 ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the borel parameter @xmath160 ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the borel parameter @xmath160 ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] , @xmath62 and @xmath159 with variation of the borel parameter @xmath160 ; the ( i ) and ( ii ) correspond to the charm and bottom channels respectively .
, title="fig:",width=226 ] the main uncertainties come from the six parameters @xmath161 , @xmath162 and @xmath163 , the variations of those parameters can lead to relatively large changes for the numerical values , refining those parameters are of great importance .
although there are many parameters in the light - cone distributions amplitudes @xcite , the uncertainties originate from those parameters are rather small . taking into account all the uncertainties of the relevant parameters ,
finally we obtain the numerical results of the strong coupling constants @xmath61 , @xmath62 and @xmath159 , which are shown in fig.2 , @xmath164 and @xmath165 in the charm and bottom channels respectively . in this article
, we calculate the uncertainties @xmath166 with the formula @xmath167 where the @xmath168 denote strong coupling constants @xmath61 , @xmath62 and @xmath159 , the @xmath169 denote the relevant parameters @xmath170 , @xmath171 , @xmath172 , @xmath173 . as the partial derivatives @xmath174 are difficult to carry out analytically , we take the approximation @xmath175 ^ 2 $ ] in the numerical calculations .
the light - cone qcd sum rules approaches have been applied to determine the strong coupling constant @xmath176 in the strong decay @xmath177 both in the leading approximation @xcite and the next - to - leading order approximation @xcite .
the discrepancy between the experimental data from the cleo collaboration and the theoretical predictions is rather large .
the upper bound @xmath178 ( @xmath179 @xcite ) is too small to account for the experimental data , @xmath180 @xcite .
there have been several explanations , for example , becirevic et al take into account the contribution from an explicit radial excitation to the hadronic spectral density to improve the value of @xmath176 @xcite ; kim tries to subtract the term @xmath181 which is supposed to come from a mathematically spurious term and should not be a part of the final sum rules to smear the discrepancy @xcite ; while duraes et al resort to the intermediate hadronic loops to improve the predictive ability @xcite . or the simple quark - hadron duality ansatz which works in the one - variable dispersion relation might be too crude for the double dispersion relation @xcite . irrespective of the possible reasons
, the light - cone sum rules can not give satisfactory value to account for the experimental data in the channel @xmath182 , while the light - cone qcd sum rules are rather successful in calculating the strong coupling constants among the baryons , baryons and mesons , for example , we study the strong decays @xmath56 , @xmath57 and @xmath183 using the light - cone qcd sum rules , and observe that the numerical values of the widths are in agreement with the experimental data within uncertainties @xcite .
the present predictions for the values of the strong coupling constants @xmath61 , @xmath62 and @xmath159 are reasonable .
the radiative decays @xmath2 can be described by the following electromagnetic lagrangian @xmath184 , @xmath185 where the @xmath186 is the electromagnetic field . from the lagrangian @xmath184 , we can obtain the decay amplitude with the assumption of the vector meson dominance , @xmath187 where the @xmath188 is the polarization vector of the photon . in the heavy quark limit
, the matrix elements @xmath189 and can be neglected , so we consider only the contribution of the intermediate @xmath1 . from the strong coupling constants @xmath112 and @xmath62 , we can obtain the decay widths @xmath190 , @xmath191 the numerical values are @xmath192 here we take the value @xmath193 from the cdf collaboration @xcite , if we take the value @xmath194 from the d0 collaboration @xcite , the radiative decay @xmath195 is kinematically forbidden . comparing with the values from the consitutent quark model @xmath196 @xcite , the hyper central model @xmath197 @xcite , and the non - relativistic potential model @xmath198 @xcite , the present prediction @xmath199 is rather good , though the uncertainty is somewhat large .
in this article , we parameterize the vertexes @xmath200 with three tensor structures due to lorentz invariance , study the corresponding three strong coupling constants with the light - cone qcd sum rules , then assume the vector meson dominance of the intermediate @xmath1 as the contributions from the @xmath201 and @xmath202 are negligible in the heavy quark limit , and calculate the radiative decay widths @xmath203 .
the predictions can be compared with the experimental data in the future .
the strong coupling constants in the vertexes @xmath0 are basic parameters in describing the interactions among the heavy baryon states , once reasonable values are obtained , we can use them to perform phenomenological analysis .
this work is supported by national natural science foundation , grant number 10775051 , and program for new century excellent talents in university , grant number ncet-07 - 0282 , and a foundation of ncepu .
the light - cone distribution amplitudes of the @xmath1 meson are defined by @xcite , @xmath204\nonumber\\ & & + \left [ \epsilon_\mu - q_\mu \frac{\epsilon \cdot x}{q \cdot x } \right]f_\phi m_\phi \int_0 ^ 1 du e^{-i \bar{u}q \cdot x } g_{\perp}^{(v)}(u ) \nonumber\\ & & -\frac{1}{2}x_\mu \frac{\epsilon \cdot x}{(q \cdot x)^2 } f_\phi m_\phi^3 \int_0 ^ 1 du e^{-i\bar{u}q \cdot x}c(u ) \ , , \nonumber\\ \langle 0| { \bar s } ( x ) \sigma_{\mu\nu } s(0 ) |\phi(q)\rangle & = & \left[\epsilon_\mu q_\nu-\epsilon_\nu q_\mu\right]i f_\phi^\perp \int_0 ^ 1 du e^{-i \bar{u } q\cdot x } \left[\phi_{\perp}(u)+\frac{m_\phi^2x^2}{16 } a_\perp(u)\right]\nonumber\\ & & + \left[q_\mu x_\nu - q_\nu x_\mu\right]\frac{\epsilon \cdot x}{(q \cdot x)^2}i f_\phi^\perp m_\phi^2 \int_0 ^ 1 du e^{-i \bar{u } q \cdot x } b_{\perp}(u ) \nonumber\\ & & + \frac{\epsilon_\mu x_\nu-\epsilon_\nu x_\mu}{2q \cdot x}i f_\phi^\perp m_\phi^2 \int_0 ^ 1 du e^{-i\bar{u}q \cdot x}c_\perp(u)\ , , \nonumber \\
\langle 0| { \bar s } ( x)\gamma_\mu \gamma_5 s(0 ) |\phi(q)\rangle & = & -\frac{1 } { 4}\epsilon_{\mu\nu\alpha\beta}\epsilon^\nu q^\alpha x^\beta\widetilde{f}_\phi m_\phi \int_0 ^ 1 du e^{-i \bar{u } q \cdot x } g_{\perp}^{(a)}(u ) \ , , \nonumber\\ \langle 0| { \bar s } ( x ) s(0 ) |\phi(q)\rangle & = & -\frac{i } { 2}\epsilon \cdot x\widetilde{f}_\phi^\perp m_\phi^2 \int_0 ^ 1 du e^{-i \bar{u } q \cdot x } h_{\parallel}^{(s)}(u ) \ , , \end{aligned}\ ] ] @xmath205 f_\phi m_\phi \int \mathcal{d}\alpha e^{-i ( \alpha_{\bar{s}}+v\alpha_g ) q\cdot x } \mathcal{a}(\alpha_i ) \ , , \nonumber\\ \langle 0| { \bar s } ( x)i\gamma_\alpha g_{\mu\nu}(vx ) s(0 ) q_\mu\right ] if_\phi^\perp m_\phi^2 \int \mathcal{d}\alpha e^{-i ( \alpha_{\bar{s}}+v\alpha_g ) q\cdot x } \widetilde{\mathcal{s}}(\alpha_i ) \ , , \nonumber\\ \langle 0| { \bar s } ( x ) g_{\mu\nu}(vx ) s(0 ) |\phi(q)\rangle & = & \left[\epsilon_\mu q_\nu-\epsilon_\nu q_\mu\right ] if_\phi^\perp m_\phi^2 \int \mathcal{d}\alpha e^{-i ( \alpha_{\bar{s}}+v\alpha_g ) q\cdot x } \mathcal{s}(\alpha_i ) \ , , \nonumber\\ \langle 0| { \bar s } ( x)\sigma_{\alpha\beta } g_{\mu\nu}(vx ) s(0 ) g_{\alpha\nu } -q_\alpha q_\nu g_{\beta\mu } + q_\beta q_\nu g_{\alpha\mu}\right ] \nonumber\\ & & f_\phi^\perp m_\phi^2 \frac{\epsilon\cdot x}{2 q \cdot x } \int\mathcal{d}\alpha e^{-i ( \alpha_{\bar{s}}+v\alpha_g ) q\cdot x } \mathcal{t}(\alpha_i ) \ , , \end{aligned}\ ] ] where @xmath206 @xmath207 , @xmath208 , @xmath209 , the lengthy expressions of the light - cone distribution amplitudes @xmath210 , @xmath211 , @xmath212 , @xmath213 , @xmath214 , @xmath215 , @xmath216 , @xmath217 , @xmath218 , @xmath219 , @xmath220 , @xmath221 , @xmath222 , @xmath223 , @xmath224 can be found in refs .
@xcite .
@xmath227}{\pi^2}m^2e_0(x)\int_0 ^ 1 dt t(1-t ) e^{-\frac{\widetilde{m}_q^2}{m^2}}\right.\nonumber\\ & & - \frac{u_0m_q^2 f_\phi m_\phi \left[\widetilde{\phi}_{\parallel}(1-u_0)-\widetilde{g}_{\perp}^{(v)}(1-u_0)\right]}{18m^4}\langle\frac{\alpha_sgg}{\pi}\rangle\int_0 ^ 1 dt \frac{1-t}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & -\frac{u_0f_\phi m_\phi^3 \widetilde{a}(1-u_0)}{4\pi^2 } \int_0 ^ 1 dt t e^{-\frac{\widetilde{m}_q^2}{m^2 } } + \frac{u_0m_q^2f_\phi m_\phi^3 \widetilde{a}(1-u_0)}{72m^6 } \langle \frac{\alpha_sgg}{\pi } \rangle\int_0 ^ 1 dt \frac{1}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2 } } \nonumber\\ & & -\frac{2u_0m_sf_\phi^{\perp}m_\phi^2 \widetilde{\widetilde{b}}_{\perp}(1-u_0)}{\pi^2}\int_0 ^ 1 dt t e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & + \frac{u_0m_sm_q^2f_\phi^{\perp}m_\phi^2 \widetilde{\widetilde{b}}_{\perp}(1-u_0)}{9m^6}\langle\frac{\alpha_sgg}{\pi}\rangle\int_0 ^ 1 dt \frac{1}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & -\frac{u_0\widetilde{f}_\phi m_\phi } { 4\pi^2}m^2e_0(x ) g_\perp^{(a)}(1-u_0)\int_0 ^ 1 dt t(1-t ) e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & + \frac{u_0m_q^2\widetilde{f}_\phi m_\phi } { 72m^4}\langle \frac{\alpha_sgg}{\pi}\rangle g_\perp^{(a)}(1-u_0)\int_0 ^ 1 dt \frac{1-t}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & \left.+\frac{f_\phi m_\phi}{2\pi^2 } m^2e_0(x ) \int_0 ^ 1 dt t\int_0^{u_0 } d\alpha_{\bar{s } } \int_{u_0-\alpha_{\bar{s}}}^{1-\alpha_{\bar{s } } } d\alpha_g \frac{\mathcal{a}(\alpha_i)+(1 - 2v)\mathcal{v}(\alpha_i ) } { \alpha_g } e^{-\frac{\widetilde{m}_q^2}{m^2 } } \right\ } \nonumber \end{aligned}\ ] ] @xmath228}{3m^2 } \right.\nonumber\\ & & - \frac{u_0m_s\langle \bar{s } s\rangle
f_\phi m_\phi^3 \widetilde{a}(1-u_0)}{6m^4}\left(1+\frac{m_q^2}{m^2}\right ) \nonumber\\ & & - \frac{u_0m_s\langle \bar{s}g_s
\sigma gs\rangle f_\phi m_\phi \left[\widetilde{\phi}_\parallel(1-u_0)-\widetilde{g}_{\perp}^{(v)}(1-u_0)\right]}{9m^4}\left(1+\frac{m_q^2}{m^2}\right ) \nonumber\\ & & + \frac{8u_0\langle\bar{s}s\rangle f_\phi^{\perp}m_\phi^2 \widetilde{\widetilde{b}}_{\perp}(1-u_0 ) } { 3m^2 } - \frac{2u_0\langle \bar{s}g_s \sigma gs\rangle f_\phi^{\perp}m_\phi^2\widetilde{\widetilde{b}}_{\perp}(1-u_0)}{3m^4}\left(1+\frac{m_q^2}{m^2}\right)\nonumber\\ & & \left.-\frac{u_0m_s\langle \bar{s}s\rangle\widetilde{f}_\phi m_\phi g_\perp^{(a)}(1-u_0)}{6m^2 } \right\ } \ , , \end{aligned}\ ] ] @xmath229}{2\pi^2}m^4e_1(x)\int_0 ^ 1 dt t(1-t ) e^{-\frac{\widetilde{m}_q^2}{m^2}}\right.\nonumber\\ & & - \frac{m_q^2f_\phi m_\phi \left[\widetilde{\phi}_{\parallel}(1-u_0)-\widetilde{g}_{\perp}^{(v)}(1-u_0)\right]}{36m^2}\langle\frac{\alpha_sgg}{\pi}\rangle\int_0 ^ 1 dt \frac{1-t}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & -\frac{f_\phi m_\phi^3 \widetilde{a}(1-u_0)}{8\pi^2 } m^2e_0(x)\int_0 ^ 1 dt t e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & + \frac{m_q^2f_\phi m_\phi^3 \widetilde{a}(1-u_0)}{144m^4 } \langle \frac{\alpha_sgg}{\pi } \rangle\int_0 ^ 1 dt \frac{1}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2 } } \nonumber\\ & & + \frac{m_sf_\phi^{\perp } \phi_{\perp}(1-u_0)}{2\pi^2}m^4e_1(x)\int_0 ^ 1 dt t e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & - \frac{m_sm_q^2f_\phi^{\perp } \phi_{\perp}(1-u_0)}{36m^2}\langle\frac{\alpha_sgg}{\pi}\rangle\int_0 ^ 1 dt \frac{1}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & -\frac{m_sf_\phi^{\perp}m_\phi^2\widetilde{\widetilde{b}}_\perp ( 1-u_0)}{\pi^2}m^2e_0(x)\int_0 ^ 1 dt
t e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & + \frac{m_sm_q^2f_\phi^{\perp}m_\phi^2 \widetilde{\widetilde{b}}_{\perp}(1-u_0)}{18m^4}\langle\frac{\alpha_sgg}{\pi}\rangle\int_0 ^ 1 dt \frac{1}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & + \frac{\widetilde{f}_\phi m_\phi } { 8\pi^2}m^4e_1(x ) g_\perp^{(a)}(1-u_0)\int_0 ^ 1 dt t(1+t ) e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & -\frac{m_q^2\widetilde{f}_\phi m_\phi } { 144m^2}\langle \frac{\alpha_sgg}{\pi}\rangle g_\perp^{(a)}(1-u_0)\int_0
^ 1 dt \frac{1+t}{t^2 } e^{-\frac{\widetilde{m}_q^2}{m^2}}\nonumber\\ & & \left.-\frac{u_0f_\phi m_\phi^3 } { 2\pi^2}m^2e_0(x ) \int_0 ^ 1 dt t\int_0^{u_0 } d\alpha_{\bar{s } } \int_{u_0-\alpha_{\bar{s}}}^{1-\alpha_{\bar{s } } } d\alpha_g v\frac{\mathcal{a}(\alpha_i)-\mathcal{v}(\alpha_i ) } { \alpha_g } e^{-\frac{\widetilde{m}_q^2}{m^2 } } \right\ } \nonumber \end{aligned}\ ] ] @xmath230}{3 } \right.\nonumber\\ & & - \frac{m_s\langle \bar{s}s\rangle f_\phi m_\phi^3 \widetilde{a}(1-u_0)}{12m^2}\left(1+\frac{m_q^2}{m^2}\right ) \nonumber\\ & & - \frac{m_s\langle \bar{s}g_s \sigma gs\rangle
f_\phi m_\phi \left[\widetilde{\phi}_\parallel(1-u_0)-\widetilde{g}_{\perp}^{(v)}(1-u_0)\right]}{18m^2}\left(1+\frac{m_q^2}{m^2}\right ) \nonumber\\ & & -\frac{2\langle\bar{s}s\rangle f_\phi^{\perp } \phi_{\perp}(1-u_0)}{3}m^2e_0(x ) + \frac{\langle \bar{s}g_s \sigma gs\rangle f_\phi^{\perp}\phi_{\perp}(1-u_0)}{6}\left(1+\frac{m_q^2}{m^2}\right)\nonumber\\ & & - \frac{m_q^4 f_\phi^{\perp } m_\phi^2 \langle \bar{s}g_s\sigma g s\rangle a_{\perp}(1-u_0)}{24m^6 } - \frac{m_s f_\phi^{\perp } m_\phi^2 a_{\perp}(1-u_0)}{8\pi^2 } m^2 e_0(x ) \nonumber\\ & & + \frac{\langle\bar{s}s\rangle f_\phi^{\perp } m_\phi^2 a_{\perp}(1-u_0)}{6 } \left(1+\frac{m_q^2}{m^2}\right ) + \frac{4\langle\bar{s}s\rangle f_\phi^{\perp}m_\phi^2 \widetilde{\widetilde{b}}_{\perp}(1-u_0 ) } { 3}\nonumber\\ & & - \frac{\langle \bar{s}g_s \sigma
gs\rangle f_\phi^{\perp}m_\phi^2\widetilde{\widetilde{b}}_{\perp}(1-u_0)}{3m^2}\left(1+\frac{m_q^2}{m^2}\right)\nonumber\\ & & \left.+\frac{m_s\langle \bar{s}s\rangle\widetilde{f}_\phi m_\phi g_\perp^{(a)}(1-u_0)}{12}\left(1+\frac{2m_q^2}{m^2}\right ) \right\ } \ , , \end{aligned}\ ] ] where @xmath231 and @xmath232 as @xmath233 , @xmath234 , @xmath235 , @xmath236 , @xmath237 , and @xmath238 , @xmath239 , the @xmath240 denote the light - cone distribution amplitudes . for some technical details concerning the three particle @xmath100-meson light - cone distribution amplitudes , one can consult ref.@xcite . | in this article , we study the vertexes @xmath0 with the light - cone qcd sum rules , then assume the vector meson dominance of the intermediate @xmath1 , and calculate the radiative decays @xmath2 .
+ zhi - gang wang + department of physics , north china electric power university , baoding 071003 , p. r. china + pacs numbers : 11.55.hx , 12.40.vv , 13.30.ce , 14.20.lq , 14.20.mr * key words : * heavy baryons ; light - cone qcd sum rules |
the combination of light weight , strong covalent bonds , and low dimensionality gives carbon nanostructures , such as graphene and nanotubes , superior mechanical and thermal properties @xcite , making them interesting candidate materials for thermal management @xcite and phononics applications @xcite .
extremely high , possibly divergent , thermal conductivity ( @xmath0 ) of a two - dimensional ( 2d ) phonon gas was predicted by klemens and pedraza ten years before single layer graphene was isolated for the first time @xcite .
the first measurements in suspended graphene at about room temperature confirmed klemens and pedraza s predictions , reporting values of @xmath0 in the range @xmath1 wm@xmath2k@xmath2 @xcite .
later experiments found @xmath3 wm@xmath2k@xmath2at @xmath4 k , and @xmath5 wm@xmath2k@xmath2at @xmath6 k @xcite .
recent measurements on suspended graphene in vacuum yield @xmath0 in the range @xmath7 wm@xmath2k@xmath2at @xmath4 k @xcite . in spite of the efforts to refine these measurements , an accurate determination of @xmath0 remains a tough experimental challenge @xcite .
a fundamental reason for such difficulty is that heat transport in graphene is very sensitive to defects and experimental conditions .
for example , when graphene is supported on a substrate @xmath0 is reduced to @xmath8 wm@xmath2k@xmath2at room temperature due to phonon scattering from the substrate @xcite .
@xmath0 can vary by as much as 50@xmath9 at room temperature as a function of the isotopic composition of graphene @xcite , and was found to be sensitive also to the lateral dimension of measured patches @xcite and to the presence of wrinkles , which may lower @xmath0 by @xmath10 @xcite .
this high sensitivity can be advantageous because it offers the possibility to manipulate @xmath0 in graphene - based devices either by tuning the concentration of defects and mass disorder during growth or by imposing controlled external conditions .
high and tunable thermal conductivity for a single sheet of atoms opens up the possibility of application in a range of thermal management devices , from high - power electronics @xcite all the way to thermoelectric applications @xcite . in pristine graphene at room temperature
, heat is conducted almost exclusively by phonons @xcite , so we can focus on lattice thermal conductivity . both lattice dynamics ( ld ) calculations @xcite and molecular dynamics ( md ) studies @xcite
indicate that @xmath0 of suspended graphene converges with system size , in contrast with ideal 2d models , for which @xmath0 diverges logarithmically @xcite .
flexural phonons ( out - of - plane vibrational modes ) play a decisive role both as heat carriers and as scatterers .
nevertheless , some of these works report very diverse numerical results , stemming from the use of different methods relying on different approximations , and from the choice of various interatomic potentials .
theoretical studies suggest that it is possible to control @xmath0 in graphene by applying mechanical ( tensile ) strain , however simulations have given contradictory results .
recent _ ab initio _ ld calculations showed that , while @xmath0 is finite for unstrained suspended graphene , it diverges when tensile strain is applied @xcite .
in contrast , md results point in the opposite direction , indicating a reduction of @xmath0 upon strain @xcite . even though discrepancies between ld and md results are expected , as the two methods rely upon different approximations @xcite
, it is unusual to get such differences in trends .
in fact , in ld calculations anharmonic interactions are usually truncated at the first order , while in md simulations quantum effects can not be taken into account , so phonon populations obey to classical statistics .
both approximations conspire to make @xmath11 larger than @xmath12 .
in this work we report the results of equilibrium molecular dynamics ( emd ) simulations of heat transport in suspended graphene as a function of strain and isotopic mass disorder .
we begin by investigating size convergence of @xmath0 in isotopically pure unstrained graphene , and then we study the effects of mechanical strain and isotopic disorder .
our goal is to verify whether the divergence , predicted by ld calculations on isotropically strained graphene , also occurs upon uniaxial strain at finite temperature .
we also probe how the combination of strain and mass disorder affects @xmath0
. a microscopic interpretation of the results is provided in terms of phonon populations and lifetimes computed at finite temperature .
we compute the thermal conductivity of graphene by emd simulations in models with periodic boundary conditions .
we use the tersoff empirical potential @xcite recently re - parametrized to accurately reproduce the vibrational properties of carbon nanostructures @xcite .
anharmonic ld calculations employing this set of parameters result in a thermal conductivity @xmath13 wm@xmath2k@xmath2for a @xmath14 @xmath15 m graphene flake at @xmath16 k , well within the range of experimental measurements .
md production runs are performed in the microcanonical ensemble , starting from initial configurations equilibrated at the target temperature @xcite .
temperatures between 300 and 1000 k are considered .
the equations of motion are integrated with a 1 fs time step .
the cell parameters are optimized at the simulation temperature to achieve zero stress in the ( zig - zag ) direction , perpendicular to the strained one .
following linear response theory , @xmath0 is computed from the integral of the autocorrelation function of the heat flux @xmath17 in a microcanonical simulation , according to the green - kubo formula @xcite @xmath18 where @xmath19 is boltzmann s constant , @xmath20 is the temperature and @xmath21 the volume , which is here defined as the surface area of the graphene foil times a nominal thickness of @xmath22 . in practice , @xmath0 is taken as the stationary value of eq .
( [ eq : greenkubo ] ) before it drifts due to accumulated statistical noise .
although @xmath0 is in general a tensor , the hexagonal symmetry of graphene yields @xmath23 and @xmath24 .
the limits to infinite time and infinite volume in eq .
( [ eq : greenkubo ] ) indicate that size convergence and phase space sampling have to be carefully considered .
this aspect is particularly important for low - dimensional systems , for which transport coefficients usually diverge @xcite .
therefore , investigating size and time convergence is not merely a technical aspect , but it brings important physical insight . in order to effectively sample the phase space and achieve statistical accuracy in evaluating eq .
( [ eq : greenkubo ] ) , each reported value of @xmath0 is obtained by averaging over at least @xmath25 independent simulations of at least @xmath26 ns .
to check size convergence we perform simulations with approximately square supercells of increasing size .
we consider systems made of from @xmath27 to @xmath28 atoms .
the smallest supercell is @xmath29 @xmath30 and the largest one @xmath31 @xmath30 . fig .
[ fig : gk1 ] shows the calculated thermal conductivity as a function of the number of atoms in the simulation cell .
the anisotropy between the in - plane thermal conductivities seen for the smallest cell ( @xmath27 atoms ) is due to an uneven and insufficient sampling of the vibrational modes in the two directions .
as the cell size is increased a better sampling is achieved , and the anisotropy vanishes .
we find that a @xmath32 @xmath30 simulation cell , containing @xmath33 atoms , is required to obtain a converged value @xmath34 wm@xmath2k@xmath2 .
our estimate of @xmath0 is lower than the values reported in recent works , in which smaller systems were simulated @xcite .
the inset in fig .
[ fig : gk1 ] displays the normalized heat flux autocorrelation functions ( hfacf ) for several simulations with different cell sizes , showing that in all cases the time decay is faster than @xmath35 , which guarantees convergence of eq .
( [ eq : greenkubo ] ) .
we can conclude that @xmath0 of unstrained graphene at finite temperature is finite and converges with size , confirming the prediction of former _ ab initio _ ld calculations @xcite .
classical calculations of @xmath0 far below the debye temperature ( @xmath36 for graphene ) may yield large differences with respect to calculations taking into account the proper quantum statistics for phonons .
two effects contribute to such differences , yet in opposite directions : classical calculations give shorter phonon lifetimes than quantum calculations , but classical phonon heat capacities are always larger than quantum ones .
ld calculations showed that in graphene at room temperature these two effects compensate to the point that classical @xmath0 underestimates quantum @xmath0 only by about 10@xmath9 @xcite .
wm@xmath2k@xmath2is achieved for a cell with @xmath33 atoms .
inset : time decay of the normalized heat flux autocorrelation functions ( hfacf ) for @xmath37 , @xmath33 , and @xmath38 atom cells .
all hfacf decay faster than @xmath35 . ] in the zig - zag direction ( red ) .
dispersion curves are computed in the strained armchair direction ( @xmath39-m ) and in the stress - free zigzag direction ( @xmath39-k ) .
wavevectors are in units of the inverse lattice constant @xmath40 , rescaled according to applied strain . ] the observed size convergence of @xmath0 from above , provides an insight into the contribution of flexural phonons to thermal transport in graphene .
convergence trends can be interpreted by referring to the dispersion relations of phonons in graphene , computed in the harmonic approximation by diagonalizing the dynamical matrix @xcite ( fig .
[ fig : dispersions ] ) .
accurate md calculations of the thermal conductivity of graphene or carbon nanotubes require a converged sampling of the low - frequency acoustic flexural ( za ) modes @xcite .
whereas the contribution to @xmath0 of in - plane acoustic modes converges relatively fast , good sampling of the za modes is achieved only for large simulation cells because of their quadratic dispersion relation near the @xmath39 point .
za modes are expected to provide a significant contribution to heat transport , and have been identified as the majority heat carriers @xcite .
however , close to the @xmath39 point their group velocity vanishes and their main role in thermal transport is to scatter other heat carriers .
our convergence trends indeed show that the overall effect of low - frequency za modes is to lower the in - plane thermal conductivity of graphene .
in fact , performing simulations on ` 2d graphene ' , i.e. , a graphene sheet in which atoms move only in plane , we observe logarithmic divergence of @xmath41 , as shown in fig .
[ fig : k.vs.time.2d3d ] , in accordance with theoretical and numerical studies on 2d model systems @xcite .
our results demonstrate the dual role of za modes , which is to provide an important reservoir of heat carriers , as well as the main scattering channel that prevents the divergence of @xmath0 @xcite . for the 2d case , yielding @xmath42 divergence of @xmath0 .
] strain affects the vibrational properties of materials , as it modifies phonon dispersion relations .
speed of sound , frequency range , scattering rates , and therefore thermal conductivity , are all altered .
we apply uniaxial tensile strain along the armchair direction and relax the simulation cell to achieve zero stress in the perpendicular ( zigzag ) direction .
the dispersion relations of strained graphene ( strain @xmath43 ) are compared to the unstrained ones in fig .
[ fig : dispersions ] . in the low - frequency range
the most significant changes is the linearization of the dispersion relation of the za mode along the strain axis . in the direction perpendicular to the strain axis
the za branch remains unchanged .
this implies non - vanishing group velocity for the za modes propagating along the strain axis .
in addition , in - plane acoustic modes are slightly softened in both directions . in the high - frequency range
the degeneracy of the zone center optical phonon is broken , in accordance with raman measurements.@xcite fig .
[ fig : kappa.strain ] shows the thermal conductivity of graphene as a function of time , calculated as the argument of the time limit in eq .
( [ eq : greenkubo ] ) , along the strained direction for strain up to @xmath44 . for small strain ( @xmath45 )
, the thermal conductivity in the strained direction still converges , yet to a larger value than in the unstrained case . as strain
is increased ( @xmath46 in the figure ) , the thermal conductivity along the strained direction tends to diverge .
the same behavior is observed for a larger simulation cell containing more than @xmath47 atoms .
the inset in fig .
[ fig : kappa.strain ] displays the time decay of the respective hfacf . for @xmath48
, the hfacf decays faster than @xmath35 .
however , at larger strain the hfacf decays as @xmath35 , such that @xmath0 diverges as @xmath42 , following the standard behavior of transport coefficients in 2d systems @xcite .
meanwhile , the thermal conductivity in the stress - free direction does not diverge for thermal conductivity parallel and perpendicular to strain direction , thermal conductivity of isotropically strained graphene at @xmath49 k , vdos of strained and unstrained graphene , and a comparison of phonon populations in the presence of strain and isotopic mass disorder . ] .
in fact , the thermal conductivity perpendicular to the strain direction is slightly reduced , due to a mild softening of the in - plane acoustic modes as shown in fig .
[ fig : dispersions ] .
it is important to point out that a logarithmic divergence of @xmath0 with size in finite 2d model systems under stationary non - equilibrium conditions implies a logarithmic divergence of @xmath0 as a function of time in periodic systems at equilibrium and _ vice versa _ @xcite . in other words , a @xmath35 decay of the heat flux autocorrelation function in a periodic system at equilibrium ( in the absence of a temperature gradient ) , implies a logarithmic divergence of @xmath0 with system size under non - equilibrium conditions ( in the presence of a finite temperature gradient ) .
therefore , our predictions can ( in principle ) be probed experimentally by measuring the size dependence of the thermal conductivity in strained graphene samples . _ ab initio _
ld calculations predict divergence of @xmath0 in isotropically strained graphene for any amount of applied strain @xcite .
md simulations of isotropically strained graphene at @xmath49 k suggest that @xmath0 diverges already for @xmath50 , confirming the predictions from ld @xcite .
in contrast , uniaxial strain and finite temperature limit divergence to relatively large strain @xmath46 , whereas at lower strain @xmath0 remains finite .
simulations at @xmath51 k confirm the divergence of @xmath0 along the direction of strain persists for @xmath46 , evidencing no significant difference with respect to the trends observed at room temperature . in this higher temperature regime , even though far below the debye temperature of graphene , classical phonon populations approach quantum populations , and quantum effects are mitigated .
since phonon lifetimes computed with classical statistics are usually underestimated with respect to those obtained with the correct quantum statistics @xcite , we can safely argue that the divergent nature of @xmath0 in strained graphene is not an artifact of classical md .
\(a ) + k and ( b ) @xmath52 k , calculated from eq .
( [ eq : greenkubo ] ) .
the dashed horizontal lines indicate the average @xmath0 for unstrained graphene .
the inset shows the time decay of the normalized hfacf . for high strain
, the hfacf decays as @xmath35 such that @xmath0 diverges as @xmath42 as observed in the main panel.,title="fig : " ] + ( b ) + k and ( b ) @xmath52 k , calculated from eq .
( [ eq : greenkubo ] ) .
the dashed horizontal lines indicate the average @xmath0 for unstrained graphene .
the inset shows the time decay of the normalized hfacf . for high strain
, the hfacf decays as @xmath35 such that @xmath0 diverges as @xmath42 as observed in the main panel.,title="fig : " ] from both analytical models and ld calculations , it appears that the cause for the divergence of @xmath0 lies in the linearization of long - wavelength za modes ( fig .
[ fig : dispersions ] ) .
such alteration of the dispersion relations affects phonon populations , which are probed by computing the vibrational density of states ( vdos ) of strained and unstrained samples .
our calculations indeed show a depletion of the vdos for the za modes in the frequency range from @xmath53 thz up to @xmath54 thz @xcite .
-m ) are shown .
lifetimes at small wavevectors @xmath55 diverge for the 4% strained model .
the 1/q decay is shown ( dashed line ) for reference . ]
these changes have dramatic effects on phonon lifetimes ( fig .
[ fig : lifetimes ] ) , which diverge for large strain and lead to divergence of the thermal conductivity . using the solution of the linearized boltzmann transport equation in the relaxation time approximation
, one can express @xmath0 for a periodic system of finite size as the sum of the contributions of each phonon mode , as @xmath56 .
here @xmath57 is the specific heat of mode @xmath58 , @xmath59 its group velocity , and @xmath60 its lifetime .
since @xmath61 and @xmath57 are always finite , divergence of @xmath0 implies divergence of @xmath60 for some of the modes .
phonon lifetimes are computed here as the decay time of the autocorrelation function of the energy of the normal modes in microcanonical md simulations @xcite .
we indeed observe the presence of za modes with slowly decaying correlation functions when @xmath62 is larger than 2@xmath9 , which imply diverging @xmath63 .
the trends of @xmath64 ( fig .
[ fig : lifetimes ] ) permit the extrapolation of our results to extended systems , for which @xmath0 is expressed in integral form : @xmath65 where the integral is taken over the two - dimensional brillouin zone ( bz ) .
the limit for @xmath66 determines whether @xmath0 diverges . in
the classical case @xmath67 is a constant , and in general @xmath68 .
the group velocity of the za modes along @xmath69 is @xmath70 . for unstrained graphene @xmath71 , whereas when strain is applied along @xmath69 , @xmath72 tends to a constant value . in unstrained graphene @xmath0 would diverge for @xmath73 , while when strain is applied @xmath0 diverges for @xmath74 .
our calculations show that @xmath75 increases with strain , reaching @xmath76 for @xmath77 , which is consistent with the observed threshold for @xmath0 divergence .
so far we have presented results for isotopically pure graphene ( @xmath78% @xmath79c ) .
given the demonstration of graphene growth with customized isotopic composition @xcite , it is also worth investigating the combined effect of strain and controlled isotopic composition on the thermal conductivity of graphene , checking whether its divergence can be suppressed by mass disorder .
we consider pure @xmath79c , natural composition ( @xmath80 @xmath81c ) , and @xmath81c enriched graphene models ( 10@xmath9 , 50@xmath9 and 99.2@xmath9 ) . in absence of defects
the lattice thermal conductivity is limited by phonon - phonon scattering @xcite , therefore @xmath82 .
meanwhile , the scattering of phonons by defects is temperature independent @xcite , so that @xmath83 trends are modified . in fig .
[ fig : kappa.vs.t ] we indeed observe @xmath84 for isotopically pure graphene , natural graphene and @xmath85% @xmath81c isotopically enriched graphene , indicating that for the natural isotopic composition the effect of mass disorder is almost negligible .
however , as the amount of @xmath81c increases to 10@xmath9 and 50@xmath9 , @xmath0 decreases more slowly with @xmath20 indicating that mass disorder becomes the primary source of phonon scattering .
this aspect could be exploited in thermal management devices that operate over wide temperature ranges , where it would be undesirable to have large variations in @xmath0 with @xmath20 .
the ratio between @xmath0 for isotopically pure and 50@xmath9 @xmath81c - enriched graphene is about 2 , and agrees well with recent experimental measurements @xcite . c. solid lines are fitted to the data in order to extract the temperature dependence , as reported in the plot . ] even though large variations of @xmath0 as a function of isotopic composition are found , when uniaxial tensile strain is applied the general behavior of @xmath0 for isotopically modified graphene is not qualitatively different from the isotopically pure case .
@xmath0 along the strained direction increases for low strain and diverges for @xmath46 even in samples with the highest isotopic disorder ( 50@xmath9 @xmath81c ) , as shown in fig .
[ fig : kappa.vs.timec13 ] .
the inset shows that above the strain threshold the hfacf decay as @xmath35 and thus @xmath86 .
the ratio between phonon populations shows that in isotopically enriched graphene , as in the isotopically pure case , tensile strain induces similar reductions to the population of za modes @xcite .
therefore , mass disorder is not sufficient to suppress the divergence of the lifetimes of low - frequency za modes .
in fact , these modes have a wavelength of several tens of nm , thus even in the presence of isotopic mass disorder they propagate as in a continuous medium , and are not significantly affected by scattering centers at the atomic scale .
it is worth noting that mass disorder can not suppress divergence in non - linear models as well @xcite .
k , calculated from eq .
( [ eq : greenkubo ] ) , with a 50@xmath9 concentration of @xmath81c .
the dashed horizontal lines indicate the average @xmath0 for the unstrained case .
the inset shows the normalized heat flux autocorrelation functions , which decay like @xmath35 for @xmath87% , yielding @xmath42 divergence of @xmath0 . ]
in conclusion , we have shown that heat transport in suspended graphene is controlled by za modes , which contribute the essential scattering channels to limit the thermal conductivity in unstrained samples .
in fact , in absence of za modes @xmath0 would diverge with size , as in ideal 2d models .
uniaxial tensile strain reduces the population of za phonon modes at low frequency , makes their zone center group velocity finite and increases their lifetime , thus causing divergence of the thermal conductivity in the strained direction .
we then predict that @xmath0 of strained samples would diverge logarithmically with the size of the samples .
it is important to point out that our predictions based on computer simulations are accessible to experiments . in an experimental setup
, @xmath0 will always be finite and limited by boundary and defect scattering .
nonetheless , by performing measurements in strained samples of increasing size it should be possible to see a logarithmic dependence of @xmath0 as a function of size .
the amount of strain required to observe divergence , @xmath88 , might also be within reach of current experimental techniques .
we also predict that the presence of isotopic mass disorder does not suppress the divergence of @xmath0 that should be then expected to occur also in samples with natural composition , which are easier to grow than isotopically pure ones .
42ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1038/nmat3064 [ * * , ( ) ] link:\doibase 10.1063/1.2907977 [ * * , ( ) ] link:\doibase 10.1063/1.3183587 [ * * , ( ) ] link:\doibase 10.1021/nl901231s [ * * , ( ) ] link:\doibase 10.1103/revmodphys.84.1045 [ * * , ( ) ] link:\doibase 10.1016/0008 - 6223(94)90096 - 5 [ * * , ( ) ] link:\doibase 10.1021/nl0731872 [ * * , ( ) ] link:\doibase 10.1021/nl9041966 [ * * , ( ) ] link:\doibase 10.1021/nn102915x [ * * , ( ) ] link:\doibase 10.1126/science.1184014 [ * * , ( ) ] link:\doibase 10.1038/nmat3207 [ * * , ( ) ] link:\doibase 10.1021/nl301230 g [ * * , ( ) ] link:\doibase 10.1088/0957 - 4484/23/36/365701 [ * * , ( ) ] link:\doibase 10.1038/ncomms1828 [ * * , ( ) ] link:\doibase 10.1021/nn3031595 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.205441 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.115427 [ * * , ( ) ] link:\doibase 10.1063/1.3665226 [ * * , ( ) ] link:\doibase 10.1021/nl202694 m [ * * , ( ) ] link:\doibase 10.1063/1.3246155 [ * * , ( ) ] link:\doibase 10.1063/1.3435465 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.045413 [ * * , ( ) ] http://www.springerlink.com/index/mh60x378752gv73m.pdf [ * * , ( ) ] link:\doibase 10.1016/s0370 - 1573(02)00558 - 6 [ * * , ( ) ] link:\doibase 10.1103/physreve.86.040101 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.245318 [ * * , ( ) ] link:\doibase 10.1103/physrevb.79.064301 [ * * , ( ) ] link:\doibase 10.1039/c2cp42394d [ * * , ( ) ] http://link.aps.org/doi/10.1103/physrevlett.61.2879 [ * * , ( ) ] link:\doibase 10.1103/physrevb.69.134103 [ * * , ( ) ] link:\doibase 10.1063/1.1740082 [ * * , ( ) ] link:\doibase 10.1143/jpsj.12.570 [ * * , ( ) ] link:\doibase 10.1063/1.1854281 [ * * , ( ) ] link:\doibase 10.1021/nl202118d [ * * , ( ) ] link:\doibase 10.1021/nn200114p [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.115460 [ * * , ( ) ] @noop _ _ ( , , ) link:\doibase 10.1103/physrevlett.99.255502 [ * * , ( ) ] link:\doibase 10.1103/physrevb.79.205433 [ * * , ( ) ] link:\doibase 10.1103/physrevb.34.5058 [ * * , ( ) ] @noop _ _ ( , , ) | we investigate the effect of strain and isotopic disorder on thermal transport in suspended graphene by equilibrium molecular dynamics simulations .
we show that the thermal conductivity of unstrained graphene , calculated from the fluctuations of the heat current at equilibrium is finite and converges with size at finite temperature .
in contrast , the thermal conductivity of strained graphene diverges logarithmically with the size of the models , when strain exceeds a relatively large threshold value of 2% .
an analysis of phonon populations and lifetimes explains the divergence of the thermal conductivity as a consequence of changes in the occupation of low - frequency out - of - plane phonons and an increase in their lifetimes due to strain . |
let @xmath0 and @xmath1 be two polynomials with complex coefficients and degrees less than or equal to @xmath2 .
the only joint invariant under affine substitutions which is linear in the coefficients is @xmath3_n = a_0 b_n - \frac{1}{n } a_1 b_{n-1 } + \frac{1}{\binom{n}{2 } } a_2 b_{n-2 } + \ldots + ( -1)^n a_n b_0.\ ] ] the two polynomials are called apolar if @xmath4_n = 0 $ ] .
apolarity provides the ground for the study of the simplest type of invariant , generalizing in higher degree the notion of harmonic quadrics .
the geometric implications of apolarity are surprising and multifold , see for instance @xcite
. a celebrated result due to grace and heawood asserts that the complex zeros of two apolar polynomials can not be separated by a circle or a straight line . as a matter of fact ,
this zero location property is equivalent to apolarity @xcite .
an array of independent proofs of grace s theorem , as the result is named nowadays , are known ( see @xcite ) .
the common technical ingredient of these proofs is a lemma of laguerre and induction .
notable is also the coincidence , up to a conjugation in the second argument , of the apolarity invariant and fischer s inner product : @xmath5 well known for identifying the adjoint of complex differentiation with the multiplication by the variable @xcite .
it is natural and convenient to symmetrize the polynomials @xmath6 and @xmath7 and interpret both apolarity and fischer inner products in terms of the roots of these polynomials . in this direction
an observation due to walsh establishes a powerful equivalent statement to grace s theorem , known as the walsh coincidence theorem : if a monic polynomial @xmath6 of degree @xmath2 has no zeros in a circular domain @xmath8 ( that is a disc , complement of a disk , a half - space , open or closed ) , then its symmetrization has no zeros in @xmath9 , @xcite . in its turn
, walsh s theorem explains and offers a natural framework for a series of polynomial inequalities , in one or several variables @xcite .
the aim of the present note is to study the pull - back of the apolarity invariant by a polynomial mapping @xmath10 of degree @xmath11 .
it turns out that the pull - back form can be represented on the original space of polynomials of degree @xmath2 by an invertible linear operator @xmath12 : @xmath13_{nd } = [ t_{q , n } f , g]_n.\ ] ] at the geometric level , the pull back is transformed into push - forward by the ramified cover map @xmath10 . for a circular domain @xmath8 , we distinguish between the set theoretic image @xmath14 and the full push - forward @xmath15 which consists of all points in @xmath14 having the full fiber contained in @xmath8 .
relative @xmath10-versions of the theorems of grace and walsh follow easily , for instance : _ if two polynomials @xmath6 and @xmath7 are @xmath16 and all zeros of @xmath6 are contained in @xmath17 , then @xmath7 has a zero in @xmath14_. as expected , the @xmath10-version of the walsh coincidence theorem has non - trivial consequences in the form of polynomial inequalities for symmetric polynomials of several variables . to be more specific
we prove below that a majorization on the diagonal of @xmath14 is transmitted , with the same constant , to the polydomain @xmath18 .
the operator @xmath12 representing the @xmath10-apolarity form is complex symmetric , in the sense of @xcite , and as a consequence the eigenfunctions of its modulus are doubly orthogonal , both in the fischer norm , and with respect to the apolarity bilinear form .
in particular the zeros of these polynomials can not be separated by higher degree algebraic sets deduced from the boundaries of @xmath14 , respectively @xmath17 .
the idea of providing a more flexible version of grace s theorem also relates to recent work of b. and h. sendov in @xcite , which is concerned with finding minimal domains satisfying the statement of grace s theorem for a given polynomial .
this paper is structured as follows : section [ sec : prelim ] contains notation , preliminaries and some of the classical results . in section [ sec : polyimages ] , we discuss polynomial images of circular domains .
section [ sec : symmetrization ] contains the core results on symmetrization and pullback and the relative versions of the theorems of grace and walsh .
section [ sec : skeweigenfunctions ] concerns the study of skew - eigenfunctions of the symmetrization operator @xmath12 in the sense of @xcite .
_ acknowledgements .
_ we would like to thank one of the referees for helpful comments .
daniel plaumann was partially supported by a research fellowship of the nanyang technological university .
we write @xmath19_n$ ] for the space of complex polynomials in @xmath20 of degree at most @xmath2 . for @xmath21_n$ ]
, the _ symmetrization _ of @xmath6 in degree @xmath2 is the polynomial in @xmath2 variables @xmath22 defined by @xmath23 where @xmath24 is the elementary symmetric polynomial of degree @xmath25 in @xmath22 .
the symmetrization is the unique multiaffine symmetric polynomial of degree at most @xmath2 in @xmath22 with the property @xmath26 to verify uniqueness , just note that a multiaffine symmetric polynomial of degree at most @xmath2 in @xmath22 is necessarily of the form @xmath27 .
[ sharpcheck ] for @xmath21_n$ ] , we write @xmath28 observe that @xmath29 and @xmath30 . note also that the definition of @xmath31 depends on @xmath2 .
even if @xmath32 , so that @xmath6 is of degree less than @xmath2 , it is understood that @xmath31 is defined as above whenever we take @xmath33_n$ ] .
[ fischerip ] the _ fischer inner product _ is the hermitian inner product on @xmath19_n$ ] given by @xmath34 for @xmath35 and @xmath36 .
the following hold for all @xmath33_n$ ] . 1 .
[ fip : orthogonality ] @xmath37 for @xmath38 ; + 2 .
[ fip : sharpcheck ] @xmath39 for all @xmath40_n$ ] ; + 3 .
@xmath41 for all @xmath40_{n-1}$ ] ; + 4 .
@xmath42 for all @xmath43 ; + 5 .
[ fip : symmetrization ] @xmath44 .
( 1)(4 ) are checked directly . to verify ( 5 )
, expanding the product @xmath45 shows that @xmath46 [ bracket ] for @xmath47_n$ ] , we define @xmath3_n = { \langle f , g^{\#\vee}\rangle}_n.\ ] ] note that if @xmath7 is monic of degree @xmath2 with zeros @xmath48 , then @xmath3_n = \operatorname{sym}_n(f)(\mu_1,\dots,\mu_n).\ ] ] by [ fischerip]([fip : symmetrization ] ) .
the form @xmath49_n$ ] is bilinear and non - degenerate and , by [ fischerip]([fip : sharpcheck ] ) , satisfies @xmath3_n = ( -1)^n[g , f]_n.\ ] ] if @xmath4_n = 0 $ ] , then @xmath6 and @xmath7 are called _
apolar_. a _
circular domain _ is an open or closed disk or halfspace in @xmath50 , or the complement of any such set .
we use @xmath51 to denote the open unit disk .
[ thm : walsh ] if @xmath52 is a circular domain and @xmath33_n$ ] is monic without zeros in @xmath8 , then @xmath53 has no zeros in @xmath9 .
since the complement of a circular domain is again a circular domain , an equivalent statement of walsh s theorem is let @xmath52 be a circular domain .
if @xmath33_n$ ] is monic with all zeros in @xmath8 and @xmath54 is a zero of @xmath53 , then @xmath55 for some @xmath56 .
see hrmander @xcite ( or @xcite ) for the proof .
[ cor : grace ] let @xmath52 be a circular domain and let @xmath47 $ ] be apolar polynomials of the same degree . if all zeros of @xmath6 are contained in @xmath8 , then @xmath7 has at least one zero in @xmath8
. let @xmath57 .
without loss of generality , we may assume that @xmath7 is monic and write @xmath58
. then @xmath3_n = \operatorname{sym}_n(f)(\mu_1,\dots,\mu_n)\ ] ] by [ bracket ] . by walsh s theorem ,
at least one of @xmath48 must lie in @xmath8 , as claimed .
let @xmath52 be a circular domain and let @xmath59 $ ] be a monic polynomial of degree @xmath11 .
we want to understand how walsh s and grace s theorem transform under the map @xmath60 .
we will use the following notation : @xmath61 the set @xmath17 is semialgebraic and its boundary is contained in the real algebraic curve @xmath62 . since the complement of a circular domain is a circular domain ,
the set @xmath17 is the complement of the image of a circular domain , namely @xmath63 let @xmath64 .
two points @xmath65 have the same image under @xmath10 if and only if @xmath66 .
this implies @xmath67 and @xmath17 is the image of that region under @xmath10 .
in particular , if @xmath68 is the open unit disk , then @xmath69 is non - empty if and only if @xmath70 .
for example , take @xmath71 .
the image of the unit circle under @xmath10 is the real quartic curve @xmath72 in the complex plane , where @xmath73 the preimage of this curve under @xmath10 consists of the unit circle and the shifted unit circle @xmath74 .
the region @xmath69 and its preimage are shown in fig .
[ fig : quarticcurve ] . and
its preimage for @xmath75,height=151 ] @xmath76 and its preimage for @xmath75,height=188 ] [ example : cubic ] consider the cubic polynomial @xmath77 .
the image of the unit circle under @xmath10 is the real sextic curve @xmath78 given by @xmath79 shown on the right - hand side of figure [ fig : sexticcurve ] .
the preimage of this curve under @xmath10 has two real components , the unit circle and a curve of degree @xmath80 given by the vanishing of the polynomial @xmath81 this curve is shown on the left of figure [ fig : sexticcurve ] . together with the unit circle
, it bounds the region @xmath82 . and
its preimage for @xmath83,height=188 ] @xmath76 and its preimage for @xmath83,height=188 ] note that the region @xmath17 need not be connected in general . to construct an example in which it is not ,
consider a conformal map @xmath84 from the unit disc onto an annular sector @xmath85 such a map @xmath86 exists by the riemann mapping theorem and , by caratheodory s theorem @xcite , extends continuously to a map @xmath87 between the closures .
now @xmath88 can be approximated uniformly by a sequence of polynomials .
consequently , if @xmath89 $ ] satisfies @xmath90 the image @xmath91 has hausdorff - distance at most @xmath92 from @xmath93 . put @xmath94 , then , by construction , the complement of the image @xmath95 is disconnected , the origin being contained in a bounded connected component of @xmath96 .
thus if we consider the circular region @xmath97 , then @xmath98 is not connected . in order to test whether @xmath69 is non - empty for a given @xmath10
, we may proceed as follows .
given a monic polynomial @xmath59_d$ ] , write @xmath99 by the schur - cohn criterion ( see @xcite ) , all roots of @xmath10 are contained in @xmath51 if and only if the hermitian form defined by the @xmath100-matrix @xmath101 is positive definite .
this proves the following .
the region @xmath69 is empty if and only if the matrix @xmath102 has a non - positive eigenvalue , for every @xmath43 .
similar criteria exist for the case of a halfplane instead of a disk .
consider the family of cubic polynomials @xmath103 for @xmath104 .
we compute the schur - cohn matrix and find @xmath105 thus we see that if @xmath106 , the matrix @xmath107 is not positive definite for any @xmath108 , so that @xmath109 .
conversely , if @xmath110 , then @xmath111 is positive definite and hence @xmath112 . for real @xmath113 , the picture in the complex plane will essentially look the same as in the case @xmath114 in example [ example : cubic ] above . for @xmath115
, the image curve will degenerate as in figure [ fig : emptyregion ] and the region @xmath69 will be empty .
, the region @xmath69 is empty.,height=188 ] if @xmath10 is viewed as a rational function on the riemann sphere @xmath116 , then @xmath117 .
it follows that if @xmath8 is an unbounded domain , so that @xmath118 is contained in the closure of @xmath8 inside @xmath116 , then @xmath118 is also in the closure of @xmath17 , by continuity .
in particular , @xmath17 is non - empty .
+ more generally , we may consider @xmath119 where @xmath120 is the cardinality of the fiber , counted with multiplicities .
clearly , the regions @xmath121 are pairwise disjoint for different values of @xmath25 and the boundary of @xmath121 is contained in the curve @xmath62 .
furthermore , we have @xmath122 the regions @xmath121 can also be characterized in terms of a fiber - counting integral .
for example , let @xmath97 be the complement of the unit disk
. for @xmath123 , let @xmath124 then @xmath125 is the number of preimages of @xmath126 under @xmath10 contained in @xmath8 , counted with multiplicities , so that @xmath127 . for our purposes , it would be best if we could have @xmath128 . unfortunately
, this case does not occur in a non - trivial way , as the following theorem shows .
the equality @xmath129 occurs if and only if @xmath130 for some @xmath131 and @xmath132 .
let @xmath59 $ ] be monic of degree @xmath11 and assume @xmath129 , which is equivalent to @xmath133 .
write @xmath134 , then @xmath135 by continuuity .
we show first that this implies @xmath136 .
note that @xmath10 is non - constant and hence @xmath60 is a surjective open map , which implies @xmath137 .
suppose this inclusion is strict , which means that @xmath138 contains interior points of @xmath139 .
since @xmath138 is a closed curve , it follows that @xmath140 is disconnected . on the other hand
, @xmath135 implies @xmath141 , hence @xmath95 is disconnected , a contradiction . from this
we see that @xmath95 is an open subset of @xmath50 with connected boundary and hence it is simply connected .
fix a point @xmath142 . by the riemann mapping theorem
, there exists a biholomorphic map @xmath143 with @xmath144 .
let @xmath145 .
since @xmath95 is simply connected and bounded by the jordan curve @xmath138 , the holomorphic map @xmath146 extends continuously to a map @xmath147 , by carathodory s theorem @xcite .
this implies that @xmath146 has a representation as a finite blaschke product of degree @xmath11 , i.e. there exists a monic polynomial @xmath148 $ ] of degree @xmath11 and @xmath149 such that @xmath150 for all @xmath151 ( see for example @xcite ) .
let @xmath152 be the zeros of @xmath153 , then @xmath154 implies that @xmath155 are also zeros of @xmath156 , so that @xmath157 .
factoring @xmath86 as @xmath158 for some holomorphic map @xmath159 , we can write @xmath160 for @xmath161 . dividing both sides by @xmath156 shows @xmath162 therefore ,
@xmath163 is a polynomial of degree at most @xmath11 that is constant along the fibers of @xmath10 . by lemma [ lemma : constantonfibers ] below
, this implies that there exist constants @xmath164 such that @xmath165 .
but the point @xmath142 can be chosen arbitrarily , so we obtain constants @xmath166 depending on @xmath142 and identities @xmath167 expanding @xmath168 and comparing leading coefficients on both sides leads to @xmath169 . after cancelling leading coefficients ,
we are then left with @xmath170 for this to hold for all @xmath142 , we must have @xmath171 , so that @xmath172 is constant , as claimed .
[ lemma : constantonfibers ] if @xmath173 $ ] are two polynomials of the same degree such that @xmath174 is constant along the fibers of @xmath10 , i.e. @xmath175 implies @xmath176 for all @xmath65 , then there are constants @xmath164 such that @xmath177 .
if @xmath178 are constant , there is nothing to show . otherwise , consider the algebraic curve @xmath179 . by hypothesis ,
the polynomial @xmath180 $ ] vanishes identically on @xmath181 .
since @xmath182 is square - free , hilbert s nullstellensatz gives an identity @xmath183 for some @xmath184 $ ] .
let @xmath185 be any zero of @xmath174 , then @xmath186 .
since @xmath174 and @xmath10 are of the same degree , @xmath187 must have degree @xmath188 in @xmath20 , so that putting @xmath189 and @xmath190 yields the desired identity .
let @xmath33_n$ ] , @xmath59_d$ ] , with @xmath10 monic , and consider @xmath191 , the symmetrization of @xmath192 in the @xmath193 variables @xmath194 .
let @xmath195 since @xmath10 has degree @xmath11 , the fibers of @xmath196 can be identified with points in @xmath197 .
let @xmath198 and put @xmath199 in other words , @xmath200 are the zeros of the polynomial @xmath201 .
we compute the restriction of @xmath202 to this fiber . by [ fischerip]([fip : symmetrization ] ) , we have @xmath203 since @xmath204 , we find @xmath205 by [ sharpcheck ] , hence @xmath206 we introduce the following notation .
[ prop : propertiest ] a. @xmath208 for all @xmath33_n$ ] .
b. @xmath209 for all @xmath33_n$ ] . c. the leading coefficient of @xmath210 is @xmath211 these leading coefficients are exactly the eigenvalues of @xmath12 .
d. the linear operator @xmath212_n\to{{\mathbb}{c}}[z]_n$ ] is invertible .
\(a ) by construction , @xmath213 is symmetric and multiaffine of degree @xmath2 in @xmath214 and satisfies @xmath215 . by the uniqueness of the symmetrization
, it therefore coincides with @xmath216 .
\(b ) and ( c ) for @xmath217 , we compute @xmath218 if @xmath219 ( where @xmath220 ) , then @xmath221 since monomials in @xmath20 of different degree are orthogonal and every term of @xmath222 has degree at most @xmath223 in @xmath20 , we see that @xmath224 is a polynomial of degree at most @xmath25 in @xmath126 . since @xmath225 , it follows that @xmath210 has degree at most @xmath25 in @xmath20 . to find the leading coefficient , isolate all terms of degree @xmath25 in @xmath226 on the right - hand side of the inner product above : these are of the form @xmath227 for some polynomial @xmath228 $ ] .
since the left - hand side is of degree @xmath223 in @xmath20 with leading term @xmath229 , we have @xmath230 .
thus the coefficient of @xmath231 is found to be equal to @xmath232 , as claimed .
the equality @xmath233 is equivalent to the fact that the matrix representing @xmath12 in the basis @xmath234 is upper - triangular .
its diagonal entries and thus its eigenvalues are exactly the leading coefficients of @xmath210 .
\(d ) follows from ( c ) .
we say that two polynomials @xmath47_n$ ] are _
@xmath10-apolar _ if @xmath235_{nd}=0.\ ] ] the notion of @xmath10-apolarity is strongly related to the operator @xmath12 , as the following proposition shows .
[ prop : qapolarity ] we have @xmath235_{nd } = [ t_{q , n}(f),g]_n\ ] ] for all @xmath47_n$ ] . if @xmath236 , let @xmath237 . using prop .
[ prop : propertiest](a ) and [ fischerip ] , we find @xmath238_{nd } & = \beta{\langle f\circ q,\prod\nolimits_{j=1}^n ( q-\mu_j)^{\#\vee}\rangle}_{nd } = \beta s_{q , n}(f)(\mu_1,\dots,\mu_n)\\ & = \beta \operatorname{sym}_n(t_{q , n}(f))(\mu_1,\dots,\mu_n ) = [ t_{q , n}(f),g]_n.\end{aligned}\ ] ] if @xmath239 , the identity also holds , by continuity .
let @xmath240 following the above computation , we find @xmath241 where @xmath242 is the discriminant of @xmath10 .
hence @xmath243 with respect to the basis @xmath244 , the operator @xmath245 is therefore represented by the uppper - triangular matrix @xmath246 for a general cubic polynomial @xmath247 direct computation shows @xmath248 with respect to the basis @xmath249 , the operator @xmath250 is therefore represented by the upper - triangular matrix @xmath251 let @xmath252 in the basis @xmath244 the operator @xmath253_2\to{{\mathbb}{c}}[z]_2 $ ] is represented by the matrix @xmath254 where @xmath255 is the discriminant of @xmath10 and @xmath256 .
the operator @xmath250 in the basis @xmath249 is represented by the matrix @xmath257 we are now ready for the generalized versions of the theorems of grace and walsh .
let @xmath8 be a circular domain and @xmath59 $ ] a monic polynomial of degree @xmath11 . if @xmath47 $ ] are @xmath10-apolar polynomials of the same degree and all zeros of @xmath6 lie in @xmath17 , then @xmath7 has at least one zero in @xmath14 .
if @xmath258 are the zeros of @xmath259 , then @xmath260_{nd}=\operatorname{sym}_{nd}(f\circ q)(y_{11},\dots , y_{nd})$ ] by [ bracket ] .
so if @xmath260_{nd}=0 $ ] , then @xmath261 is a zero of @xmath202 .
since all zeros of @xmath6 are in @xmath17 , all zeros of @xmath192 are in @xmath8 , and therefore @xmath262 for some @xmath263 by walsh s theorem ( thm .
[ thm : walsh ] ) .
hence @xmath264 is a zero of @xmath7 .
[ thm : genwalsh ] let @xmath8 be a circular domain and @xmath59 $ ] a monic polynomial .
let @xmath33 $ ] be a polynomial of degree @xmath2 . 1 .
assume that all zeros of @xmath6 lie in @xmath17 .
then all zeros of @xmath265 lie in @xmath14 and if @xmath198 is a zero of @xmath266 , then @xmath267 for some @xmath25 .
2 . assume that @xmath6 does not vanish in @xmath14 .
then @xmath265 does not vanish in @xmath17 and @xmath213 does not vanish in @xmath268 .
\(1 ) assume that @xmath269 . by definition
, @xmath213 is the restriction of @xmath202 to the fibers of @xmath10 , so @xmath270 , where @xmath271 for @xmath272
. since all zeros of @xmath6 are in @xmath17 , the zeros of @xmath192 are in @xmath8 . by walsh
s theorem ( thm .
[ thm : walsh ] ) , this implies @xmath262 for some @xmath263 and hence @xmath273 , as claimed .
\(2 ) apply ( 1 ) to the circular domain @xmath274 and use the identities @xmath275 and @xmath276 .
[ cor : generalizedbernstein ] let @xmath47 $ ] be monic of degree @xmath2 without common zeros in @xmath14 .
given @xmath277 with @xmath278 , there exists @xmath279 with @xmath280 such that @xmath281 if @xmath282 on @xmath14 , then @xmath283 \(1 ) suppose that @xmath149 is not in @xmath284 . since @xmath6 and @xmath7 have no common zeros in @xmath14 , this implies that @xmath285 does not vanish anywhere in @xmath14 .
then @xmath286 does not vanish anywhere in @xmath268 by thm .
[ thm : genwalsh](2 ) , so @xmath287 is not assumed by the rational function @xmath288 on @xmath268 .
( 2 ) follows immediately from ( 1 ) .
let @xmath103 where @xmath289 is real and positive ( c.f .
example [ example : cubic ] ) .
the region @xmath69 contains the disc @xmath290 , while @xmath14 is contained in the disc @xmath291 . now a weaker version of cor .
[ cor : generalizedbernstein ] says that given @xmath47_n$ ] such that @xmath292 for all @xmath20 with @xmath293 , we must have @xmath294 for all @xmath22 with @xmath295 for @xmath296 . + for instance , if @xmath297 are such that @xmath298 it follows that @xmath299 whenever @xmath300 .
let @xmath301 be a complex hilbert space . recall that a map @xmath302 is called _ antilinear _
if @xmath303 and @xmath304 hold for all @xmath305 , @xmath149 . if @xmath306 holds for all @xmath305 , then @xmath88 is called _ isometric_. [ lemma : antilinearisometry ] let @xmath301 be a complex hilbert space of finite dimension @xmath2 and let @xmath309 be an antilinear isometry with @xmath310 .
then there exists an orthonormal basis @xmath311 such that @xmath312 for @xmath296 .
take any vector @xmath313 with @xmath314
. then @xmath315 satisfies @xmath316 .
if @xmath317 , we put @xmath318 .
if @xmath319 , this means @xmath320 .
we put @xmath321 , so that @xmath322 . now
since @xmath323 is an isometry , the orthogonal complement @xmath324 is @xmath323-invariant and the claim follows by induction . [ thm : skew_eigenvalues ] let @xmath301 be a complex hilbert space of finite dimension @xmath2 equipped with an antilinear conjugation @xmath325 .
let @xmath326 be an invertible linear map satisfying @xmath327 1 .
there exists an antilinear conjugation @xmath323 with @xmath328 that commutes with @xmath329 , where @xmath330 , and satisfies @xmath331 .
2 . there exists an orthonormal basis @xmath311 of @xmath301 such that @xmath332 where @xmath333 are the singular values of @xmath334 .
\(1 ) write @xmath335 as above and let @xmath336 be the polar decomposition of @xmath334 , where @xmath337 is unitary
. then @xmath338 and @xmath339 .
put @xmath340 and @xmath341 . since @xmath325 and @xmath337 are both unitary , so is @xmath342 .
furthermore , since @xmath329 is positive definite , we have @xmath343 for any @xmath344 in @xmath301 , which shows that @xmath345 is also positive definite . by the uniqueness of the polar decomposition
, @xmath346 implies @xmath347 and @xmath348 .
the first equality means @xmath349 , hence @xmath350 .
put @xmath351 , then @xmath352 . also , @xmath323 is antilinear and @xmath353 , hence @xmath323 is isometric .
thus @xmath323 is an antilinear conjugation .
it also commutes with @xmath329 , since @xmath354 .
\(2 ) let @xmath355 be the singular values of @xmath334 , i.e. the eigenvalues of @xmath329 .
applying lemma [ lemma : antilinearisometry ] for the restriction of @xmath323 to each eigenspace of @xmath329 , we choose an orthonormal basis @xmath311 of corresponding eigenvectors of @xmath329 each of which is fixed under @xmath323 .
we then have @xmath356 for @xmath296 , as desired .
if @xmath59 $ ] is monic and of even degree , there exists a basis @xmath358 of @xmath19_n$ ] which is orthonormal with respect to the fischer inner product and satisfies @xmath359_{nd } = \lambda_j\delta_{jk},\ ] ] for all @xmath360 , where @xmath361 are the singular values of the operator @xmath12 .
let @xmath362 and put @xmath363_n\to{{\mathbb}{c}}[z]_n,\ p\mapsto p^{\#\vee}.\ ] ] and @xmath364 .
it is easily checked that @xmath325 is indeed an antilinear conjugation with @xmath365 .
the identity @xmath366 in the hypothesis of thm .
[ thm : skew_eigenvalues ] then follows from prop .
[ prop : qapolarity ] : for all @xmath47_n$ ] , we have @xmath367_n = [ f\circ q,({\varepsilon}cg)\circ q]_{nd } \\ & = [ ( { \varepsilon}c g)\circ q , f\circ q]_{nd}=[t({\varepsilon}c g),f]_n = { \varepsilon}[f ,
t({\varepsilon}c g)]_n = \\ & = { \langle f , ctcg\rangle}_n , \end{aligned}\ ] ] hence @xmath366 , as claimed .
( note that @xmath49_{nd}$ ] is symmetric , since @xmath11 is even . ) now the @xmath358 of @xmath19_n$ ] given by thm .
[ thm : skew_eigenvalues](2 ) indeed satisfy @xmath368_n & = [ t f_j , f_k]_n = { \langle tf_j , cf_k\rangle}_n\\ & = { \langle \lambda_j cf_j , cf_k\rangle}_n=\lambda_j\delta_{jk } \end{aligned}\ ] ] by prop .
[ prop : qapolarity ] , since @xmath325 is isometric . | we study the pullback of the apolarity invariant of complex polynomials in one variable under a polynomial map on the complex plane . as a consequence ,
we obtain variations of the classical results of grace and walsh in which the unit disk , or a circular domain , is replaced by its image under the given polynomial map . |
cooper s theorem @xcite implies that if there is an attractive interaction in a cold fermi sea , the system is unstable with respect to formation of a particle - particle condensate . in qcd , single - gluon exchange between quarks of different color generates an attractive interaction in the color - antitriplet channel @xcite .
thus , it appears unavoidable that color superconductivity occurs in dense quark matter which is sufficiently cold @xcite .
how a dense quark phase matches onto hadronic matter is difficult to address @xcite . in particular , while a quark - quark condensate may form , such condensation competes with the tendency of a quark - quark pair to bind with a third quark , to form a color - singlet hadron .
one way of understanding color superconductivity is to compute at very high densities , where by asymptotic freedom , perturbation theory can be used . at nonzero temperature , but zero quark density , it is well known that perturbation theory is a particularly bad approximation @xcite .
if @xmath4 is the coupling constant for qcd , the free energy is not an expansion in @xmath5 , but only in @xmath4 , with a series which is well behaved only for @xmath6 . in contrast , at zero temperature and nonzero quark density , the free energy is an expansion in @xmath7 , and appears to be well behaved for much larger values of the coupling constant , up to values of @xmath8 @xcite .
similar conclusions can be reached by comparing the gluon `` mass '' , @xmath9 or @xmath10 to the chemical potential , @xmath11 , or the temperature , @xmath12 @xcite .
thus for cold , dense quark matter , perturbation theory might give us information which not even lattice qcd calculations can provide .
color superconductivity is rather different from ordinary superconductivity , as in the model of bardeen , cooper , and schrieffer ( bcs ) @xcite . in bcs - like theories ,
superconductivity is determined by infrared divergences which arise in the scattering between two fermions close to the fermi surface : the initial fermions , with momenta @xmath13 and @xmath14 , scatter into a pair with momenta @xmath15 and @xmath16 .
summing up bubble diagrams generates an instability which is only cured by a fermion - fermion condensate .
if the fermions interact through a point - like four - fermion coupling , though , there is no correlation between the initial and outgoing momenta , @xmath13 and @xmath15 . in the gap equation
, this implies that the gap function is constant with respect to momentum , as long as the momenta are near the fermi surface . in qcd ,
however , scattering through single - gluon exchange strongly correlates the direction of the in- and out - going quarks : there is a logarithmic divergence for forward - angle scattering , @xmath17 .
this extra logarithm from forward - scattering implies that the gap is not an exponential in @xmath18 , as in bcs - like theories , but only in @xmath19 @xcite . as a consequence ,
the gap function is no longer constant as a function of momentum , even about the fermi surface .
the logarithmic divergence for forward - angle scattering arises because in cold , dense quark matter , static , magnetic interactions are not screened through a `` magnetic mass '' .
this is very different from a system of hot quarks and gluons . over large distances ,
a hot system is essentially three - dimensional ; gluons in three dimensions have power - like infrared divergences , which screen static , magnetic fluctuations through a magnetic mass @xmath20 .
in contrast , loop corrections in cold , dense quark matter are essentially four - dimensional ; infrared divergences are at worst logarithmic , so that any magnetic mass is at best @xmath21 , which is much smaller than the scale for color superconductivity , @xmath22 @xcite .
the dependence of the zero - temperature color - superconducting spin - zero condensate @xmath23 on the qcd coupling constant @xmath4 was first computed by son @xcite , who used a beautiful renormalization - group analysis to show that @xmath24 here @xmath25 arises naturally from the solution , assuming three colors . in son
s result , @xmath26 is a pure number of order one . in this paper
we derive the gap equations for a condensate with total spin @xmath0 and @xmath27 massless quark flavors , at an arbitrary temperature @xmath12 .
our derivation is perturbative , _
i.e. _ , @xmath28 . in this case
there are three scales in cold , dense qcd , the chemical potential @xmath11 , the gluon mass @xmath9 , and the color - superconducting condensate @xmath29 , and they are naturally ordered , @xmath30 .
we solve the gap equations to `` leading logarithmic accuracy '' , by which we mean the following . in the gap equations , the leading terms are @xmath31 .
these terms generate the exponential in @xmath19 in eq .
( [ eqphi0 ] ) , therefore @xmath32 is of order @xmath19 .
there are also leading logarithmic terms @xmath33 , which are @xmath34 plus a constant .
the @xmath35 gives rise to the prefactor @xmath36 , while the constant contributes to @xmath37 .
for @xmath38 flavors of massless quarks we find @xmath39 there are other terms in the gap equations , which do not arise from @xmath40 .
these terms are of order one , and thus of the same order as the constant term originating from the logarithm @xmath40 .
hence they contribute in the same way to @xmath37 .
we do not compute these , so that in ( [ b0 ] ) there is an undetermined constant @xmath41 .
our results were described previously @xcite , and agree completely with an independent analysis by schfer and wilczek @xcite ; they also overlap with those of hong _ et al .
the factor @xmath42 in eq .
( [ b0 ] ) originates from the @xmath38 dependence of @xmath43 ; however , @xmath41 will also depend on @xmath38 @xcite .
even though we do not determine @xmath41 , it is very interesting that the numerical value of @xmath44 is large .
this implies that , for chemical potentials of order @xmath45 gev , the gap can be of order @xmath46 mev @xcite .
such large values of the gap are in accord with previous estimates obtained within nambu jona - lasinio models @xcite , and are much larger than original estimates by bailin and love , @xmath47 mev @xcite .
we then solve the gap equations at nonzero temperature .
we find a surprising result : while the detailed form of the gap function is very different in qcd versus bcs - like theories , the temperature dependence of the condensate the dimensionless ratio of the condensate at a temperature @xmath12 to that at zero temperature , @xmath48 , is _ identical _ to bcs - type theories . in particular , the ratio of the critical temperature , @xmath49 , is as in bcs @xcite .
our result for @xmath48 is valid to leading order in weak coupling , even though we can not compute the overall magnitude of @xmath23 , _
i.e. _ , @xmath41 . in our mean - field approximation
the transition is of second order , but the transition can be driven first order by critical fluctuations near the would - be critical point @xcite .
we then classify the condensates with total spin @xmath1 .
following our classification of condensates with @xmath0 , which employed projectors for chirality , helicity , and energy , we show that there are two types of spin - one condensates , longitudinal and transverse . as an example
, we solve the gap equations for @xmath50 and find @xmath51 that is , not only is the parametric dependence of the spin - one condensates on @xmath4 the same as for spin zero , as argued originally by son @xcite , but to leading logarithmic accuracy , even the constant in front is the same .
we do not expect that @xmath52 , the undetermined constant analogous to @xmath41 , is the same .
surely @xmath52 is smaller than @xmath41 , and depends on whether the condensate is longitudinal or transverse .
this is very different from bcs - like theories , where condensates with higher spin are typically exponentially suppressed . in qcd ,
higher - spin condensates are only suppressed by a pure number .
nevertheless , one dramatic implication of our results is that superconductivity in qcd may be very unlike one s intuition from non - relativistic systems . instead of higher - spin gaps being much smaller
, they may be relatively large , which is important for phenomenology .
consider , for instance , a quark star with @xmath53 , @xmath54 , and @xmath55 quarks . in the limit of high densities , when the strange quark mass @xmath56 is negligible , the number of @xmath53 , @xmath54 , and @xmath55 quarks are equal .
then charge neutrality is automatic , and the preferred color - superconducting condensates are of spin @xmath0 , with color - flavor locking @xcite .
for realistic densities , however , the chemical potentials for @xmath53 , @xmath54 , and @xmath55 will not be equal .
the strange quark chemical potential differs from the up and down quark chemical potentials due to @xmath57 .
the up and down quark chemical potentials differ on account of charge neutrality .
these effects suppress the color - antitriplet @xmath0 condensates , because they are composed of quarks with different flavor . on the other hand
, the color - antitriplet @xmath1 condensates may form between quarks of the same flavor , and are not suppressed if the chemical potentials of the various flavors differ .
this paper is organized as follows . in section [ ii ] , we derive the gap equations for a spin @xmath0 condensate of @xmath58 massless flavors at an arbitrary temperature @xmath12 . in section [ iii ] we solve these equations to leading logarithmic accuracy , first at @xmath59 , and then at nonzero @xmath12 . in section [ j1 ]
we classify the possible spin @xmath1 condensates , and solve the gap equations for @xmath60 .
section [ v ] concludes this work with a discussion of higher - order effects which contribute to @xmath41 and @xmath52 .
our units are @xmath61 , the metric tensor is @xmath62 .
4-vectors are denoted by capital letters , @xmath63 , and @xmath64 .
in general , a color - superconducting condensate @xmath65 is a @xmath66 matrix in fundamental color space ( @xmath67 ) , a @xmath68 matrix in flavor space ( @xmath69 ) , and a @xmath70 matrix in dirac space @xcite . as shown in @xcite , for @xmath27
a color - antitriplet condensate is a flavor singlet , @xmath71 by a global color rotation , we can always choose @xmath72 to point in 3direction in color space , @xmath73 . for @xmath74 ,
even if we assume that the dominant condensate is a color antitriplet , there is always a small admixture of a color sextet @xcite .
the color - antitriplet condensate is a flavor antitriplet , and the color sextet a flavor sextet @xcite . in this paper , for simplicity we consider only the case @xmath27 . the gap equation for a color - superconducting condensate of massless fermions was derived in @xcite [ cf .
( a35 ) of @xcite ] . at a nonzero temperature @xmath12 ,
this equation reads ( we suppress fundamental color and flavor indices for the moment ) : @xmath75 here , @xmath76 in the infinite - volume limit ( @xmath77 labels the matsubara frequencies @xmath78 ) , and a summation over lorentz indices @xmath79 as well as adjoint color indices @xmath80 is implied ; @xmath81 is the number of colors .
@xmath82 is the gluon propagator , @xmath83 is the propagator for _ free _ , massless particles ( upper sign ) or charge - conjugate particles ( lower sign ) , @xmath84^{-1 } - \sigma^\pm \right\}^{-1}\ ] ] is the propagator for _ quasiparticles _ ( upper sign ) or charge - conjugate quasiparticles ( lower sign ) , and @xmath85 their self energy arising from the interaction with the condensate .
the charge - conjugate condensate is @xmath86 and the vertices are @xmath87 where @xmath88 are the dirac matrices and @xmath89 the gell - mann matrices , @xmath90 is the charge conjugation matrix . in the following ,
we analyze the flavor , color , and dirac structure of the gap equation ( [ gapeq ] ) . for @xmath58 , the color - flavor structure of the condensate ( [ cond ] )
does not mix color and flavor indices , and thus the analysis of flavor and color can be done separately .
this is different for @xmath74 , where color rotations are locked to flavor rotations .
we first discuss the flavor structure of eq .
( [ gapeq ] ) .
fundamental color indices will be suppressed for the moment .
the free propagator ( [ freeprop ] ) is diagonal in flavor , @xmath91 so is the self energy , @xmath92 and thus @xmath93 .
therefore , both the left- and right - hand sides of ( [ gapeq ] ) are simply proportional to @xmath94 : the flavor structure of the gap equation is trivial in qcd with @xmath27 flavors , and will thus not be explicitly denoted in the following .
the free propagator ( [ freeprop ] ) is diagonal in the color indices for the fundamental representation , @xmath95 the self energy @xmath96 is also diagonal , but not all diagonal elements are equal : @xmath97 the self energy for quarks with color 3 vanishes ! this is easy to understand .
let us first note that , according to eq .
( a17 ) of @xcite , the condensate @xmath98 is actually proportional to @xmath99 , while @xmath100 .
( @xmath101 is the charge - conjugate fermion field . ) thus , according to eq .
( [ selfenergy ] ) , the self energy @xmath102 arises from the following process : a quark with , let s say , color 1 annihilates with a corresponding antiquark in @xmath98 , creating a charge - conjugate quark with color 2 .
this quark is propagated with the charge - conjugate propagator @xmath103 , and annihilates with a charge - conjugate antiquark of color 2 in @xmath104 , whereby a quark with color 1 is emitted .
as only quarks with colors 1 and 2 condense , it is not possible to annihilate and create quarks with color 3 in this process ; thus the latter do not attain a self energy .
one can now compute the color structure of the quasiparticle propagator , @xmath105^{-1 } - \left(\delta_{ij } - \delta_{i3}\ , \delta_{j3 } \right)\ , \sigma^+ \right\}^{-1 } = \left(\delta_{ij } - \delta_{i3}\ , \delta_{j3 } \right)\ , g^+ + \delta_{i3}\ , \delta_{j3}\ , g_0^+\,\ , , \ ] ] where @xmath106^{-1 } - \sigma^+ \right\}^{-1}$ ] . when inserted into the gap equation ( [ gapeq ] ) , the terms @xmath107 in ( [ quasiprop2 ] ) vanish , as @xmath108 .
the gap equation becomes @xmath109 from the explicit form of the gell - mann matrices we now infer that only gluons with adjoint colors 1 , 2 , 3 , and 8 participate in the gap equation .
a color - superconducting condensate @xmath110 breaks @xmath111 to @xmath112 ; gluons 1 , 2 , and 3 then correspond to the generators of the unbroken subgroup , and thus remain massless , while the 8th gluon attains a mass through the anderson
higgs effect . denoting @xmath113 and @xmath114
, we obtain @xmath115 \
, g_0^{-}(q ) \ , \phi^+(q ) \ , g^{+}(q)\ , \gamma^\nu\,\ , , \ ] ] where a common factor @xmath116 has been dropped from both sides of the gap equation . in a complete treatment of the gap equations
the effect of the condensate on the gluon propagator has to be included @xcite .
in contrast , we use the gluon propagator in the `` hard dense loop '' ( hdl ) approximation @xcite .
the hdl propagator introduces a gluon mass @xmath117 . since in perturbation
theory the scale of the condensate @xmath23 is much smaller than @xmath43 , it is reasonable to expect that , to first approximation , we can neglect the effects of the condensate on the gluon propagator .
a more detailed explanation is given below .
we consequently assume @xmath118 , and obtain @xmath119 in @xcite we have shown that the gap matrix @xmath98 for a condensate with total spin @xmath0 has the form @xmath120 the @xmath70 matrices @xmath121 are defined as @xmath122 where @xmath123 are projectors for chirality , @xmath124 , @xmath125 while @xmath126 are projectors for helicity , and @xmath127 are projectors for energy , @xmath128 . while the @xmath129 are composed of projectors , they are not projectors themselves [ cf .
( b29 ) in @xcite ] , and thus were termed `` quasiprojectors '' in @xcite . in the ultrarelativistic limit , @xmath130
, the energy projectors simplify to @xmath131 this has three major consequences .
first , the quasiprojectors ( [ quasiprojectors ] ) become true projectors [ cf .
( b29 ) of @xcite ] .
second , @xmath132 , expressing the fact that right - handed , positive - helicity particles can not have negative energy , _ etc .
( b30 ) of @xcite ] .
third , _ either _ the chirality , _ or _ the helicity , _ or _ the energy projector in ( [ quasiprojectors ] ) becomes superfluous [ cf .
( b31 ) of @xcite ] . in the following ,
we omit the helicity projector and use just chirality and energy projectors , @xmath133 then , ( [ generalj0 ] ) simplifies to [ cf .
( 8) of @xcite ] @xmath134 the quasiparticle propagator assumes the form [ cf .
( 15 ) of @xcite ] @xmath135 ^ 2 } \ , \left[g_0 ^ -(q ) \right]^{-1 } \,\
, , \ ] ] where @xmath136 . from eq .
( 26 ) of @xcite we then infer that the gap equation ( [ gapeq2 ] ) can be written in the form @xmath137 ^ 2 } \ , { \cal p}_{-h}^{-e}({\bf q } ) \ ,
\gamma^\nu\,\ , , \ ] ] where @xmath138 , if @xmath139 , and @xmath140 , if @xmath141 . with the help of the projectors @xmath142
one can derive gap equations for the individual gap functions @xmath143 , @xmath144 ^ 2 } \ , { \rm tr } \left[{\cal p}_h^e({\bf k})\ , \gamma^\mu \ , { \cal p}_{-h}^{-e}({\bf q})\ , \gamma^\nu \right ] \right .
\nonumber \\ & & \hspace*{3.6 cm } \left .
+ \ , \frac{\phi_h^{-e}(q)}{q_0 ^ 2- \left[\epsilon^{-e}_q(\phi_h^{-e})\right]^2 } \ , { \rm tr } \left[{\cal p}_h^e({\bf k } ) \ , \gamma^\mu \ , { \cal p}_{-h}^e({\bf q } ) \ , \gamma^\nu \right ] \right\ } \,\ , .\end{aligned}\ ] ] to obtain this result we have used @xmath145 and @xmath146 , so that the gap equations for right- and left - handed condensates decouple .
this is a consequence of the @xmath147 symmetry of the qcd lagrangian , which is expected to be effectively restored at asymptotically high densities @xcite . in coulomb gauge ,
the hdl propagator is @xcite @xmath148 we set the gauge parameter @xmath149 .
we shall show later that our results are manifestly gauge - invariant to leading logarithmic accuracy .
the longitudinal and transverse propagators @xmath150 are defined in eqs .
( [ longtransprops ] ) below . the traces in ( [ projectedgapeq ] ) are readily evaluated .
we need the terms [ traces ] @xmath151 & = & \frac{1 \pm \hat{\bf k } \cdot \hat{\bf q}}{2 } \,\ , , \\ \sum_i { \rm tr } \left [ { \cal p}_h\ , \lambda^e({\bf k } ) \ , \gamma_i\ , \lambda^{\mp e}({\bf q})\ , \gamma_i \right ] & = & - \ , \frac{3 \mp \hat{\bf k } \cdot \hat{\bf q}}{2 } \,\ , , \\ { \rm tr } \left [ { \cal p}_h\ , \lambda^e({\bf k } ) \ , \bbox{\gamma }
\cdot \hat{\bf p } \ , \lambda^{\mp e}({\bf q})\ , \bbox{\gamma } \cdot \hat{\bf p } \right ] & = & - \ , \frac{1\pm \hat{\bf k } \cdot \hat{\bf q}}{2}\ , \frac{(k\mp q)^2}{p^2 } \,\ , , \end{aligned}\ ] ] where @xmath152 .
obviously , the final result is independent of the chirality projector .
we therefore conclude that the gap equations ( [ projectedgapeq ] ) for right- and left - handed gap functions are _ identical_.
this means that right- and left - handed gaps are equal up to a complex phase factor @xmath153 .
condensation fixes the value of @xmath154 and breaks @xmath147 spontaneously .
as discussed in @xcite this leads to spontaneous breaking of parity .
the gap equations for either right- or left - handed gap functions read @xmath155 ^ 2 } \ , \left [ \delta_l(k - q ) \ , \frac{1 + \hat{\bf k } \cdot \hat{\bf q}}{2 } \right .
\nonumber \\ & & \left .
\hspace*{4.1 cm } + \,\delta_t(k - q ) \left ( - \frac{3-\hat{\bf k } \cdot \hat{\bf q}}{2 } + \frac{1 + \hat{\bf k } \cdot \hat{\bf q}}{2 } \ , \frac{(k - q)^2}{({\bf k } - { \bf q})^2 } \right ) \right ] \nonumber \\ & & \hspace*{1.6 cm } + \ , \frac{\phi_h^{-e}(q)}{q_0 ^ 2- \left[\epsilon^{-e}_q(\phi_h^{-e})\right]^2 } \ , \left [ \delta_l(k - q )
\ , \frac{1 - \hat{\bf k } \cdot \hat{\bf q}}{2 } \right . \nonumber \\ & & \left . \left .
cm } + \,\delta_t(k - q ) \left ( - \frac{3+\hat{\bf k } \cdot \hat{\bf q}}{2 } + \frac{1 - \hat{\bf k } \cdot \hat{\bf q}}{2 } \ , \frac{(k+q)^2}{({\bf k } - { \bf q})^2 } \right ) \right ]
\right\ } \,\ , .\end{aligned}\ ] ] the gap equations involve singularities from both the quark and gluon propagators .
the poles of @xmath156 ^ 2\right\}$ ] give a residue @xmath157 .
remember , though , that the quasiparticle energy @xmath158 is very small near the fermi surface , @xmath159 .
the quasi - antiparticle energy @xmath160 , however , is always larger than @xmath11 . therefore ,
in weak coupling , the dominant terms arise from the quasiparticle poles @xmath161 @xcite , and the contribution from quasi - antiparticle poles can be neglected .
consequently , we do not need to solve eq .
( [ projectedgapeq2 ] ) for the quasi - antiparticle gaps , @xmath162 , in order to determine the solution for the quasiparticle gaps , @xmath163 . in the following , we drop the subscript @xmath164 and superscript @xmath165 to simplify the notation , and denote @xmath166 . to perform the matsubara sum over quark energies @xmath167
, we introduce spectral representations . for the gluon propagators
@xcite , [ longtransprops ] @xmath168 \ , e^{-\omega \tau } + n_b(\omega / t ) \ , e^{\omega \tau } \right\ } \,\ , , \end{aligned}\ ] ] where @xmath169 is the bose
einstein distribution function .
the term @xmath170 in the longitudinal propagator cancels the contribution of @xmath171 at @xmath172 @xcite .
the spectral densities are given by @xcite [ specdens ] @xmath173 + \rho^{\rm cut}_{l , t}(\omega,{\bf p } ) \
, \theta(p-\omega ) \,\ , , \\ \rho_l^{\rm pole}(\omega,{\bf p } ) & = & \frac{\omega\ , ( \omega^2 - p^2 ) } { p^2\,(p^2 + 3m_g^2-\omega^2 ) } \,\ , , \\
\rho_l^{\rm cut}(\omega,{\bf p } ) & = & \frac{2m^2}{\pi}\ , \frac{\omega}{p}\ , \left\ { \left [ p^2 + 3\ , m_g^2 \ , \left ( 1 -\frac{\omega}{2p } \ , \ln \left| \frac{p+\omega}{p-\omega } \right| \right ) \right]^2 + \left(2 m^2 \,\frac{\omega}{p } \right)^2 \right\}^{-1 } \,\ , , \\
\rho_t^{\rm pole}(\omega,{\bf p } ) & = & \frac{\omega\ , ( \omega^2 - p^2 ) } { 3m_g^2\,\omega^2-(\omega^2-p^2)^2 } \,\ , , \\
\rho_t^{\rm cut}(\omega,{\bf p } ) & = & \frac{m^2}{\pi}\ , \frac{\omega}{p}\ , \frac{p^2}{p^2-\omega^2}\ , \left\ { \left [ p^2+\frac{3}{2}\ , m_g^2 \ , \left ( \frac{\omega^2}{p^2-\omega^2 } + \frac{\omega}{2p}\ , \ln \left| \frac{p+\omega}{p-\omega } \right| \right ) \right]^2 + \left(m^2\ , \frac{\omega}{p } \right)^2 \right\}^{-1 } \ , \ , .\end{aligned}\ ] ] the basic parameter of the hdl propagators is the gluon mass @xmath174 we also found it convenient to introduce @xmath175 the functions @xmath176 are the solutions of the equations [ dispersion ] @xmath177 & = & 0 \,\ , , \\
p^2 \left ( \omega_t^2 -p^2\right ) - \frac{3}{2}\,m_g^2\ , \left [ \omega_t^2 + \frac{\omega_t}{2p } \left(p^2-\omega_t^2\right)\ , \ln \left ( \frac{\omega_t+p}{\omega_t - p } \right ) \right ] & = & 0\,\ , , \end{aligned}\ ] ] and define the dispersion relations for longitudinal and transverse gluons , respectively .
they satisfy @xmath178 .
we argue later that the phase - space region which dominates the gap integrals is the nearly static , small - momentum limit .
the gluon energies are on the order of the gap , @xmath179 , while the gluon momenta @xmath180 are much larger than @xmath181 , but much smaller than @xmath43 . in this limit , @xmath182 , @xmath183
we also introduce a spectral representation for the quantity @xmath184 ^ 2 } \equiv \int_0^{1/t } d\tau \ , e^{q_0 \tau}\ , \xi(\tau,{\bf q } ) \,\,.\ ] ] neglecting the singularities of @xmath185 in the complex @xmath186 plane , and assuming @xmath187 to be an even function of @xmath186 , we find @xmath188\ , e^{-\omega \tau } - n_f(\omega / t ) \ , e^{\omega \tau } \right\ } \,\ , , \ ] ] where @xmath189 and @xmath190 is the fermi
dirac distribution function . by neglecting the singularities of @xmath185 , the only contribution to the spectral representation of @xmath191 is from the poles of @xmath192 ^ 2\right\}$ ] , which generates the delta function in the spectral density ( [ fermionspec ] ) .
this forces the energy in the gap function @xmath187 to lie on the quasiparticle mass shell , @xmath193 .
the matsubara sums over @xmath186 can now be computed as ( @xmath194 [ matsu ] @xmath195 \right\ } \,\
, , \label{matsu1 } \\ \lefteqn{t \sum_{q_0 } \delta_t(k - q ) \ , \xi(q ) = - \frac{\phi(\epsilon_q,{\bf q})}{2\ , \epsilon_q } \int_0^\infty d \omega\ , \rho_t(\omega,{\bf p } ) } \nonumber\\ & \times & \left [ \;\ ; \frac{1}{2 } \ , \tanh \left ( \frac{\epsilon_q}{2t}\right ) \ , \left ( \frac{1}{k_0+\omega+\epsilon_q } - \frac{1}{k_0-\omega-\epsilon_q } - \frac{1}{k_0-\omega+\epsilon_q } + \frac{1}{k_0+\omega-\epsilon_q } \right ) \right .
\nonumber \\ & & + \left .
\frac{1}{2 } \ , \coth \left ( \frac{\omega}{2t}\right ) \ , \left ( \frac{1}{k_0+\omega+\epsilon_q } - \frac{1}{k_0-\omega-\epsilon_q } + \frac{1}{k_0-\omega+\epsilon_q } - \frac{1}{k_0+\omega-\epsilon_q } \right ) \right]\,\ , , \label{matsu2 } \end{aligned}\ ] ] where we have made use of ( [ fermionspec ] ) and @xmath196 , since @xmath197 . in field theories ,
the only physical quantities are those on the mass shell , such as s - matrix elements . in our case
, this implies that we need the gap function not for euclidean values of the energy , @xmath197 , but on the quasiparticle mass shell , for real values of @xmath198 .
this is achieved by analytically continuing @xmath199 @xcite .
the analytic continuation of the gap equation ( [ projectedgapeq2 ] ) introduces a term @xmath200 in the energy denominators in ( [ matsu ] ) . as @xmath201 ( where @xmath202 stands for the principal value ) , this generates an imaginary part for the function @xmath203 .
as shown below , this imaginary part is down by @xmath4 relative to the real part .
this justifies neglecting the singularities of @xmath185 in deriving ( [ matsu ] ) . in the following ,
implicitly we take the principal value of all energy denominators . finally , since the gap function @xmath204 does not depend on the orientation of the 3-vector @xmath13 , but only on its magnitude , @xmath205 , cf .
( [ projectedgapeq2 ] ) , the on - shell gap function is a function of @xmath205 only , @xmath206 . the contribution from longitudinal gluons is given by eq .
( [ matsu1 ] ) . in general , this equation can only be evaluated numerically , using the full expressions ( [ specdens ] ) for the spectral densities .
we can proceed analytically by noting that , due to the factor @xmath207 , the integral over momenta @xmath208 is dominated by the region close to the fermi surface . in the following , we shall therefore assume that the momenta @xmath13 and @xmath208 are close to the fermi surface , such that @xmath209 . now rewrite @xmath210 \,\ , , \label{relat}\end{aligned}\ ] ] and make use of the sum rule @xcite @xmath211 to obtain @xmath212 + { \cal k}_l({\bf k},{\bf q } ) \right\ } \,\,,\ ] ] where @xmath213\,\ , , \ ] ] and @xmath214 \,\ , .\ ] ] the integral @xmath215 consists of two parts , one from the pole term in the spectral density , and one from the cut term .
the pole term is @xmath216\,\ , .\ ] ] as @xmath217 , we may expand the energy denominators around @xmath218 , @xmath219 this contribution is quadratic in @xmath220 and thus negligible .
the cut term is estimated using the approximate form of the spectral density ( [ approxspec ] ) , @xmath221 \nonumber \\ & = & \frac{2m^2}{\pi p}\ , \frac{1}{\left(p^2 + 3m_g^2\right)^2 } \left [ ( \epsilon_q+\epsilon_k ) \
, \ln \left|\frac{\epsilon_q+\epsilon_k - p } { \epsilon_q+\epsilon_k+p}\right| + ( \epsilon_q-\epsilon_k)\ , \ln \left| \frac{\epsilon_q-\epsilon_k - p}{\epsilon_q-\epsilon_k+p}\right| \right ] \,\ , .\end{aligned}\ ] ] this integral is proportional to factors @xmath222 , which parametrically cancel the prefactor @xmath207 in ( [ long ] ) .
( this is true except for @xmath223 , where these factors cancel when expanding the logarithms .
however , the jacobian of the angular integration in the gap equation , @xmath224 , suppresses this contribution . )
it is thus negligible compared to the term @xmath225 in ( [ long ] ) .
we now turn to evaluate @xmath226 . after expanding the energy denominators in ( [ kl ] ) around @xmath217 ,
the pole part reads @xmath227 the range of temperatures of interest are limited by @xmath228 , the critical temperature for the onset of color superconductivity . as @xmath229 , @xmath230 .
the factor @xmath231 cancels the prefactor in ( [ long ] ) , such that @xmath232 is negligible compared to the term @xmath225 in ( [ long ] ) . to estimate the cut term , we use the approximate expression ( [ approxspec ] ) , and employ a relation similar to ( [ relat ] ) , @xmath233\,\ , , \label{relat2}\end{aligned}\ ] ] to obtain @xmath234\,\ , .\end{aligned}\ ] ] for @xmath181 of order @xmath12 or larger , this contribution is small , since then @xmath235 , and thus all terms are proportional to factors @xmath236 . for @xmath237 , however , the large ( classical ) occupation number density of gluons can enhance this contribution . in the classical limit ,
one approximates @xmath238 .
simultaneously , as the range of validity of this approximation is @xmath237 , the upper limit of the integral should now be replaced by @xmath239 . reverting the step ( [ relat2 ] ) , we obtain @xmath240 for @xmath241 , we may expand the logarithm to show that this contribution is quadratically small in @xmath12 .
on the other hand , for @xmath242 , close to the fermi surface @xmath243 , and the logarithm can be expanded for @xmath244 .
the result is proportional to @xmath236 , which cancels the prefactor @xmath207 , and thus suppresses this contribution .
we shall therefore neglect it in the following . the final result for
the longitudinal contribution is thus @xmath245 this result could also have been obtained by simply taking the static limit of the longitudinal gluon propagator , @xmath246 , on the left - hand side in ( [ longfinal ] ) , and performing the matsubara sum directly .
we conclude that the contribution of static electric gluons dominates over that of non - static electric gluons .
we note that , while the individual terms ( [ jlpole ] ) and ( [ klpole ] ) exhibit an apparent infrared divergent prefactor @xmath247 , the sum of all terms is infrared finite , as can be shown by computing the contribution of electric gluons without using ( [ relat ] ) .
the contribution from transverse gluons is written in a form similar to ( [ long ] ) : @xmath248 \,\,,\ ] ] where @xmath249\,\ , , \ ] ] and @xmath250 \,\ , .\ ] ] the function @xmath251 consists of a pole and a cut term . in the first , we expand the energy denominators around @xmath252 , to obtain to leading order in @xmath236 : @xmath253 for the cut term , we again employ ( [ approxspec ] ) to obtain @xmath254 \,\ , .\ ] ] the @xmath181 integral can be performed analytically .
denote @xmath255 .
then @xmath256 \nonumber \\ & = & \frac{p}{\pi\ , m^2 } \left\ { 2 \left[\frac{a}{a^2+(\epsilon_q + \epsilon_k)^2 } + \frac{a}{a^2+(\epsilon_q - \epsilon_k)^2}\right ] \arctan \left ( \frac{p}{a } \right ) \right .
\nonumber \\ & & \left .
\hspace*{1.2 cm } - \frac{\epsilon_q + \epsilon_k}{a^2+(\epsilon_q + \epsilon_k)^2}\ , \ln \left| \frac{p+\epsilon_q + \epsilon_k}{p-(\epsilon_q + \epsilon_k)}\right| - \frac{\epsilon_q - \epsilon_k}{a^2+(\epsilon_q - \epsilon_k)^2}\ , \ln \left| \frac{p+\epsilon_q - \epsilon_k}{p-(\epsilon_q - \epsilon_k)}\right| \right\ } \,\ , .
\label{52}\end{aligned}\ ] ] the terms proportional to @xmath236 are of higher order and can be neglected , such that @xmath257 \frac{2}{\pi}\,\arctan\left(\frac{m^2}{p^2 } \right ) \,\ , .\ ] ] the @xmath258 cuts @xmath251 off for momenta @xmath180 larger than @xmath259 . to make further progress , in the following we replace the more complicated @xmath258 with a simple @xmath154 function cutoff , @xmath260 .
however , to retain consistency with the sum rule @xcite @xmath261 we also have to modify the result for the pole term ( [ complicated ] ) . to enact the modification , note that @xmath262 is identical to the left - hand side of the sum rule ( [ sumrule ] ) , when we set @xmath263 in ( [ jt ] ) .
thus , the most simple choice is @xmath264 since then , for @xmath263 , @xmath265 the function @xmath266 also consists of a pole and a cut term . in the first , we expand the energy denominators around @xmath267 , @xmath268 as the factor @xmath231 cancels the prefactor @xmath207 in ( [ trans ] ) , this contribution is of higher order and thus negligible . the cut term is estimated with ( [ approxspec ] ) to be ( as before , @xmath269 ) @xmath270 \,\,.\end{aligned}\ ] ] for @xmath181 of order @xmath12 or larger , @xmath271 , similar steps that led to ( [ 52 ] ) now show that the final result is proportional to @xmath236 , and thus of higher order . for @xmath237 , however , @xmath272 \nonumber \\ & \simeq & \frac{m^2}{\pi } \left [ \frac{p\ , t}{p^6 + m^4 ( \epsilon_q + \epsilon_k)^2 } \ , \ln \left| \frac{p^ * + \epsilon_q + \epsilon_k}{p^*- ( \epsilon_q + \epsilon_k ) } \right| + \frac{p\ , t}{p^6 + m^4 ( \epsilon_q - \epsilon_k)^2 } \ , \ln \left| \frac{p^ * + \epsilon_q - \epsilon_k}{p^*- ( \epsilon_q - \epsilon_k ) } \right|\right]\,\ , , \end{aligned}\ ] ] where as before @xmath273 , and terms proportional to @xmath236 have been neglected in the last line . for small @xmath12 ,
an expansion of the logarithms shows that this contribution is quadratic in @xmath12 . on the other hand , for @xmath242 , an expansion of the logarithms for @xmath244 shows that it is proportional to @xmath236 ; it thus parametrically cancels the prefactor @xmath207 .
we shall thus neglect this contribution in the following .
our final result for the transverse contribution is therefore @xmath274 \right\ } \,\ , .\end{aligned}\ ] ] collecting the results from the previous subsections , and replacing the angular integration @xmath275 by an integral over @xmath276 , the gap equation ( [ projectedgapeq2 ] ) reads @xmath277 \left ( 1 + \frac{p^2}{4 k q } - \frac{(k^2-q^2)^2}{4 k q p^2 } \right ) \right\}\,\ , .
\label{particlegap2}\end{aligned}\ ] ] here , we restricted the @xmath278 integration to a region @xmath279 around the fermi surface , since we assumed that @xmath280
. therefore , we may set @xmath281 .
also , to leading order we may neglect @xmath282 terms with respect to @xmath283 terms .
then , the gap equation ( [ particlegap2 ] ) simplifies to @xmath284 \,\,.\ ] ] the first term in brackets , @xmath285 , arises from static electric gluons [ proportional to @xmath286 in eq.([particlegap2 ] ) ] , the second , @xmath287 , from non - static magnetic gluons [ proportional to @xmath288 in eq .
( [ particlegap2 ] ) ] , and the last term , @xmath289 , from soft , landau - damped magnetic gluons [ proportional to @xmath290 in eq .
( [ particlegap2 ] ) ] .
these terms can be combined to give @xmath291 where we exploited the symmetry of the integrand in ( [ gapeqalmostfinal ] ) around the fermi surface to restrict the @xmath292 integration to the interval @xmath293 , _
i.e. _ , we only integrate over momenta @xmath278 from the fermi surface up to @xmath294 .
moreover , we have defined @xmath295 the temperature dependence of @xmath43 , cf .
( [ gluonmass ] ) , has been neglected , as @xmath296 .
the result ( [ b ] ) for @xmath297 agrees with the analysis of schfer and wilczek @xcite
. it does not agree with that of hong _ et al . _
@xcite , because they neglected the contribution of electric gluons .
note that our gap equation ( [ gapeqfinal ] ) differs from that of son @xcite and of schfer and wilczek @xcite in that we integrate over momenta @xmath278 , while they integrate over euclidean energies @xmath186 .
the energy dependence of our gap function is always given by the quasiparticle mass shell @xmath298 .
one advantage of our approach is that the extension to nonzero temperature is immediate . from eq .
( [ particlegap2 ] ) , the gluon momenta which dominate the contribution of nearly static , transverse gluons , are @xmath299 , or for energies @xmath300 close to the fermi surface , @xmath301 . while these momenta are small relative to the gluon mass @xmath43 , they are much larger than the condensate @xmath302 .
this is why we can use an hdl propagator in the gap equations , neglecting the effects of the condensate .
the approximations which lead to ( [ gapeqfinal ] ) are only valid to leading logarithmic accuracy , in that we neglect terms of order 1 relative to @xmath303
. possible contributions of order 1 , which contribute to @xmath41 in eq .
( [ b0 ] ) , are discussed in section [ v ] .
the gap equation ( [ gapeqfinal ] ) is an integral equation for the function @xmath304 . in general , it can only be solved numerically @xcite . in this section ,
we first discuss the parametric dependence of the solution to ( [ gapeqfinal ] ) on the qcd coupling constant . to this end , it is instructive to understand which terms in ( [ gapeqalmostfinal ] ) determine the exponent and the prefactor in eq .
( [ eqphi0 ] ) .
we then derive an approximate analytical solution by converting the integral equation ( [ gapeqfinal ] ) into a differential equation .
the solution is discussed in detail at zero and at nonzero temperature .
let us introduce the variable @xmath305 at @xmath59 , the general structure of the gap equation ( [ gapeqalmostfinal ] ) is @xmath306 \ ; \phi_q \ ; .\ ] ] the first term , @xmath307 , arises from nearly static , transverse gluons . as compared to ( [ gapeqalmostfinal ] ) we have factored out a term @xmath308 .
the second term , @xmath309 $ ] , combines this term with the contribution from electric and non - static magnetic gluons .
the ratio @xmath44 is a well - defined number and given in eq .
( [ b0 ] ) .
these first two terms comprise the leading logarithmic approximation .
the last term , @xmath310 , represents contributions that go beyond leading logarithmic accuracy , such as terms of order one in @xmath311 , or of higher order , for instance @xmath312 .
we have neglected these terms in our derivation of eq .
( [ gapeqalmostfinal ] ) . in eq .
( [ b0 ] ) , they were written in terms of the undetermined constant @xmath41 . we are therefore permitted to choose @xmath313 in the following . in order to understand which terms are responsible for the exponent and the prefactor in eq .
( [ eqphi0 ] ) let us solve ( [ generalstruc ] ) assuming that the gap function @xmath314 is constant as a function of momentum .
this will give the wrong coefficient in the exponent , but suffices to understand the parametric dependence of ( [ eqphi0 ] ) on @xmath4 .
for @xmath315 , the @xmath316 integral in eq .
( [ generalstruc ] ) can be performed to yield ( @xmath317 ) @xmath318\,\ , .\ ] ] the first term arises from the leading term @xmath319 in ( [ generalstruc ] ) , while the second comes from the leading logarithmic term @xmath309 $ ] . the quadratic equation ( [ quadraticeq ] ) for @xmath320
has the solution @xmath321 in weak coupling , the term @xmath322 dominates the right - hand side , so that we can expand the square root .
this term then gives rise to the exponent in @xmath323 for @xmath23 . the term @xmath324 $ ] in ( [ quadsol ] ) gives rise to the prefactor of the exponential . in this way we obtain @xmath325
\,\ , .\ ] ] note that , to leading order in @xmath4 , the dependence on the cutoff @xmath326 cancels in the final result .
equation ( [ phi0wrong ] ) is rather similar to eq .
( [ eqphi0 ] ) .
the difference is that , due to our ( erroneous ) assumption of a constant gap function , the coefficient in the exponent is incorrect , @xmath327 , instead of @xmath328 .
the gap equation ( [ gapeqfinal ] ) can be solved analytically by approximating the logarithm under the integral . as in @xcite
, we replace @xmath329 and the gap equation ( [ gapeqfinal ] ) becomes @xmath330 \,\,.\ ] ] upon differentiation with respect to @xmath205 , we see that the gap function @xmath331 monotonously decreases from its maximum at the fermi surface @xmath332 ( we assume @xmath333 for @xmath334 ) .
let us introduce the variables @xmath335 where @xmath336 is the value of the gap at the fermi surface . obviously , @xmath337 . as @xmath338 [ cf .
( [ b ] ) ] , both @xmath339 and @xmath340 are much larger than 1 . in terms of these new variables , @xmath341 \equiv \epsilon(y )
\epsilon_k & = & b\mu\ , e^{-x } \left[1 + \left(\frac{\phi(x)}{\phi}\ , e^{-(x^*-x ) } \right)^2 \right ] = \epsilon(x)\,\ , , \label{ek}\end{aligned}\ ] ] where we defined a new function @xmath342 .
the transformation to the new variable @xmath340 is natural , because @xmath343 is the measure for integration , without any further jacobian .
it is similar to son s variable @xmath344 . for son s variable
, however , @xmath345 , so that in the gap equation , it is necessary to include a jacobian for the transformation from @xmath186 to @xmath346 .
this jacobian does not affect his results for the parametric dependence of the gap function on @xmath4 , but to leading logarithmic accuracy , it does affect the prefactor and the shape of the gap function , cf .
discussion at the end of subsection [ iiic ] . in the new variables
@xmath347 the approximate gap equation reads @xmath348 \,\ , , \ ] ] with @xmath342 given by ( [ eq ] ) . here , we have neglected terms @xmath349 \ll y$ ] and @xmath350\ll x$ ] .
this is a good approximation , as the value of the logarithms is bounded from above by @xmath351 , while @xmath352 .
differentiating with respect to @xmath339 yields @xmath353 this is the generalization of son s equation for the gap function to nonzero temperature .
the approximation ( [ sonstrick ] ) has succeeded to convert the original integral equation into the differential equation ( [ diffeq ] ) . for nonzero @xmath12 , however , because @xmath354 is a complicated function of @xmath339 , this equation still requires a numerical solution . on the other hand , for @xmath59
the solution is simply a trigonometric function . determining its phase and amplitude from the values of @xmath355 and @xmath356 at @xmath357
, we obtain @xmath358 \,\ , .\ ] ] the value of the zero - temperature gap function at the fermi surface , @xmath23 , can now be obtained by inserting the solution ( [ solution ] ) into the approximate gap equation for @xmath359 .
this yields @xmath360\,\ , .\ ] ] the dependence on the momentum cutoff @xmath326 is spurious : in weak coupling , @xmath361 , we may expand the @xmath258 with the result @xmath362 \,\ , .\ ] ] this is identical to eq .
( [ eqphi0 ] ) , except that here the undetermined constant @xmath41 is equal to one , since we computed only to leading logarithmic accuracy and dropped terms of order one in the gap equation .
while our analysis is strictly valid only for small values of @xmath4 , it is instructive to extrapolate to strong coupling .
the behavior of @xmath23 as function of @xmath4 is shown in fig.[phi0 ] .
one observes that @xmath23 has a maximum at a coupling constant @xmath363 .
the maximum value is quite large , @xmath364 , due to the large prefactor @xmath37 , eq .
( [ b0 ] ) , which has important implications for phenomenology @xcite . to leading order in @xmath365 , @xmath366 , such that @xmath367 the gap function peaks at the fermi surface , @xmath368 , and varies over a region @xmath369 .
this implies that the gap integral is dominated by momenta exponentially close to the fermi surface , @xmath370 .
for example , when @xmath371 , and consequently @xmath372 , @xmath373 , and the gap function is @xmath365 times smaller than at the fermi surface .
physically , this dependence of the gap function on momentum reflects the fact that single - gluon exchange is dominated by forward - angle scattering @xcite . the effect of using the variable @xmath340 , eq .
( [ defvary ] ) , instead of a variable @xmath374 analogous to son s variable @xmath375 @xcite is the appearance of a prefactor @xmath376 in eq .
( [ eqresult ] ) , as well as a factor of @xmath376 under the logarithm in @xmath377 .
these factors of 2 were found empirically in a numerical analysis by schfer and wilczek @xcite . at nonzero temperature , the gap function depends on both @xmath339 and @xmath12 . in this subsection
we denote this dependence by @xmath378 .
the value of @xmath378 at the fermi surface is denoted by @xmath379 .
as before , @xmath380 . in weak coupling , we can compute @xmath378 by assuming @xmath381 .
in other words , the @xmath339 dependence of the gap function is taken to be the same as at zero temperature , so that the only effect of nonzero temperature is to change the overall magnitude of the gap function .
we first present this calculation , and then discuss why , in weak coupling , it is reasonable to assume that the gap function as function of @xmath339 does not change when the temperature is varied .
finally , we verify this assumption by numerical calculations .
for reference , we review how the solution to a bcs - type gap equation changes with temperature .
taking the bcs coupling constant to be @xmath382 , the gap equation is of the form @xmath383 in bcs - type theories , the momentum dependence of the gap equation can be ignored , so that we just have a fixed - point equation which determines @xmath379 . besides the trivial solution , @xmath384 , at sufficiently small temperatures there is also a non - trivial solution @xmath385 .
it is determined by solving the equation @xmath386 at zero temperature , the solution is @xmath387 the gap equation requires us to introduce a cutoff , @xmath388 , but all that matters is that the cutoff is much larger than the value of the gap at zero temperature , @xmath389 . with increasing @xmath12 , the thermal factor @xmath390 reduces the integrand in ( [ bcsgap0 ] ) , so that @xmath379 has to decrease in order to achieve equality of right- and left - hand sides of eq .
( [ bcsgap0 ] ) . above a critical temperature , @xmath228 ,
this is no longer possible and we only have the trivial solution , @xmath391 .
let us investigate in more detail how the balance between the left- and right - hand sides of eq .
( [ bcsgap0 ] ) is achieved at nonzero temperature .
we divide the integration region in ( [ bcsgap0 ] ) into two pieces , for @xmath292 smaller or larger than @xmath392 , where @xmath393 is some pure number , which is assumed to be large .
the first region corresponds to momenta near the fermi surface , @xmath394 .
although @xmath395 , because the gap is _ exponentially small _ in the coupling constant , eq . ( [ solutionbcs ] ) , in weak coupling this region constitutes only a rather small contribution to the complete gap integral in ( [ bcsgap0 ] ) .
the bulk of the integral comes from the region of momenta away from the fermi surface , @xmath396 .
the quintessential point of our argument , which we shall apply to the qcd case shortly , is that , for temperatures on the order of @xmath23 , the thermal factor can be neglected in this region , up to corrections @xmath397 .
the gap equation ( [ bcsgap0 ] ) can therefore be written in the form : @xmath398 \,\ , .\ ] ] in the small region very near the fermi surface , for momenta @xmath399 , the thermal factor must be retained . in weak coupling , @xmath400 , the region away from the fermi surface
almost saturates the @xmath401 on the left - hand side of the gap equation ( [ bcsgap000 ] ) . to see this
, we compute the contribution of this region to the right - hand side , @xmath402 where we used the solution ( [ solutionbcs ] ) at @xmath59 .
hence , in order to fulfill the gap equation ( [ bcsgap000 ] ) , the first integral in ( [ bcsgap000 ] ) has to balance the term @xmath403 , @xmath404 this equation can be written in a more concise form .
note that @xmath405 where @xmath406 is the quasiparticle excitation energy at zero temperature .
we therefore replace the term @xmath407 in ( [ balance ] ) by this integral , @xmath408 \,\,.\ ] ] the same result could have been obtained directly , by subtracting eq .
( [ bcsgap0 ] ) at zero temperature from ( [ bcsgap0 ] ) at @xmath409 , and using the fact that , for momenta @xmath410 , @xmath411 , so that the contributions to the integrals away from the fermi surface cancel in the subtraction process .
the integral in ( [ bcsgap00 ] ) is finite in the infrared , even for @xmath412 , since @xmath413 as @xmath414 . dividing all quantities with dimension of energy by @xmath23
, one realizes that eq .
( [ bcsgap00 ] ) determines the ratio @xmath48 as a function of @xmath415 .
note that this ratio is independent of the cutoff @xmath388 , as well as the coupling constant @xmath382 , since @xmath416 is just an overall constant in ( [ bcsgap00 ] ) which we can divide out .
equation ( [ bcsgap00 ] ) has the following interpretation . as mentioned above , the @xmath401 on the left - hand side of the gap equation ( [ bcsgap000 ] ) is almost completely saturated by momenta away from the fermi surface , @xmath417 , where thermal corrections are negligible .
thermal effects become important in a small region near the fermi surface , @xmath418 . only in this region
, the change of the gap function with temperature has to compensate for the presence of the thermal factor .
the ratio @xmath48 is therefore _
completely _ determined by investigating how the gap equation changes with @xmath12 in a small region around the fermi surface ! in eq .
( [ bcsgap00 ] ) , we are allowed to send @xmath419 , as done in eq .
( 10 ) of @xcite , because @xmath420 for @xmath421 , and @xmath422 for @xmath423 .
the ratio @xmath48 is therefore not only independent of the cutoff @xmath388 , and the coupling constant @xmath382 , but also of @xmath393 .
however , sending @xmath419 somewhat obscures the fact that only a small region around the fermi surface controls how @xmath379 changes with temperature .
equation ( [ bcsgap00 ] ) can not be evaluated analytically for arbitrary @xmath12 , although it is easy to solve numerically ; as is well known , the transition is of second order , with @xmath424 as @xmath242 . at the critical temperature @xmath228 , however
, one can evaluate ( [ bcsgap00 ] ) analytically : @xmath425 \simeq
\ln \left(\frac{\zeta\ , \phi_0}{2\ , t_c}\right ) = 0 \;\;\ ; , \;\;\ ; \zeta = \frac{2 e^\gamma}{\pi } \simeq 1.13 \ ; . \label{bcsgapt4}\
] ] here @xmath426 is the euler - mascheroni constant .
the solution is @xmath427 which is the usual result in bcs - theory @xcite .
we now go through a similar computation for qcd , neglecting the change in the @xmath339 dependence of the gap function with temperature , @xmath381 .
consider the gap equation ( [ approxgapeq ] ) for @xmath359 , _ i.e. _ , for the gap function at the fermi surface , @xmath379 .
assuming @xmath428 , and taking the non - trivial solution @xmath385 , we can divide both sides by @xmath379 .
we obtain @xmath429 this is the equation analogous to ( [ bcsgap0 ] ) .
we now assume that the critical temperature is on the order of the value of the condensate at zero temperature , @xmath23 . in analogy to the treatment for bcs - like theories ,
we divide the region of integration into one in which @xmath399 , and one in which @xmath430 . as before , for momenta away from the fermi surface
the thermal factor is neglected .
using @xmath431 , we obtain @xmath432 \,\ , , \label{qcdgapt2}\ ] ] which is analogous to ( [ bcsgap000 ] ) .
following the definition of the variable @xmath339 in ( [ defvarx ] ) , here we introduced @xmath433 $ ] .
again , in weak coupling the @xmath401 on the left - hand side is almost completely saturated by the integration region away from the fermi surface . to see this , we compute the respective integral in a power series expansion in @xmath365 : @xmath434 in order to fulfill ( [ qcdgapt2 ] ) , the integral over the region very near the fermi surface has to compensate the terms of order @xmath435 in eq .
( [ qcdgapt2a ] ) , @xmath436 to see how this compensation works , we expand the various terms on the left - hand side in powers of @xmath365 . as @xmath437 , and @xmath438 ,
@xmath340 is of order @xmath357 in the whole integration region , up to corrections of order @xmath365 .
note that this is equivalent to approximating the factor @xmath439 by @xmath440 .
for the momentum dependence of the gap function , this has the consequence that @xmath441 , _
i.e. _ , to leading order in weak coupling , the gap function can be taken to be constant , @xmath442 . in this way one
obtains @xmath443 where we reverted the integration variable @xmath340 into @xmath292 . with ( [ 2kappa ] )
, we can write this as @xmath444 + o(\bar{g}^2 ) \,\,.\ ] ] apart from the prefactor ( which can be divided out ) , to leading order in weak coupling , eq .
( [ qcdgapt4a ] ) is the same as in bcs - type theories , eq .
( [ bcsgap00 ] ) ! thus , also the ratio @xmath48 is _ identical _ to that in bcs . given the very different nature of the gap equation , this is a remarkable result .
how can we claim that we can reliably compute @xmath48 to leading order in weak coupling , although we can not compute the overall magnitude of the condensate at zero temperature , the constant @xmath41 in ( [ b0 ] ) ? to understand this , consider the counting of powers of @xmath4 in the gap equation ; for this , it suffices to use the approximations of subsection [ iiia ] . from eq .
( [ quadraticeq ] ) we have seen that the exponential in @xmath19 arises from terms @xmath445 . since @xmath446
, these terms are of order @xmath401 , and therefore of the same order as the left - hand side of ( [ quadraticeq ] ) .
analogously , in the nonzero - temperature gap equation ( [ qcdgapt2 ] ) , terms of order @xmath401 arise from the region of momenta away from the fermi surface , and thus balance the @xmath401 on the left - hand side . as temperature effects are negligible in this region , we conclude that the exponential behavior @xmath447 of the gap can not change with temperature .
the prefactor of the exponential is determined by terms @xmath448 in the gap equation ( [ quadraticeq ] ) .
these include _ all _ terms of this order , _
i.e. _ , the leading logarithmic terms as well as terms of order one , which give rise to @xmath41 . from eq .
( [ qcdgapt3 ] ) we realize that temperature effects also enter at order @xmath4 .
these effects therefore change the prefactor of the exponential . in the above discussion
, we have determined the change of the gap equation with temperature to leading order in @xmath4 , or in other words , we have determined the prefactor at nonzero temperature . to leading order in @xmath4 ,
the result is that the prefactor changes precisely in the same way as in bcs - like theories . to understand how the gap function changes with _ both _ @xmath339 and @xmath12 , consider first the region just below the critical temperature . even though the overall magnitude of the gap , @xmath379 , is small
, we can still consider @xmath449 ; even as @xmath242 , this ratio remains of order one .
then near the fermi surface , the gap function _ must _ change due to thermal effects : after all , the variable @xmath450 $ ] diverges as @xmath451 , when @xmath424 . to understand the change in the gap function , consider energies small relative to the temperature , @xmath452 . in this limit , the thermal factor @xmath453 \simeq \epsilon/(2t)$ ] . using the definition of the variable @xmath339 , eq .
( [ diffeq ] ) is approximately @xmath454 about small @xmath455 , the solution is given as a power series in @xmath456 : @xmath457 \phi(t ) \;\;\ ; , \;\;\ ; t \rightarrow t_c \;\;\ ; , \;\;\ ; \epsilon \ll t \ ; . \label{qcdgapt6}\ ] ] thus when the energy is much less than the temperature , although @xmath458 , the gap function is not @xmath459 , as for the zero - temperature gap function , but approaches a constant .
that is , due to the thermal factor in the equation for the gap function , the momentum dependence is cut off , and the gap function `` flattens '' .
this is why in our previous calculation , we could neglect the momentum dependence of the gap function when the energy is less than @xmath392 .
our analytical results are confirmed by numerical solutions of eq .
( [ diffeq ] ) at nonzero temperature . in fig .
[ phioft ] we show the gap function at the fermi surface as a function of temperature .
we considered two values of the coupling constant , @xmath460 and @xmath461 .
the first is safely in the weak - coupling regime , while the latter is the value where @xmath23 has a maximum as a function of @xmath4 , cf .
[ phi0 ] . in weak coupling ,
@xmath48 is indistinguishable from bcs .
surprisingly , even for large @xmath462 , while @xmath228 is slightly larger than in bcs theory , @xmath463 , @xmath48 is rather similar to the behavior in bcs - like theories .
this is not unlike the situation in strongly coupled bcs , _
i.e. _ eliashberg theory @xcite , where @xmath464 also changes slightly , albeit in the opposite direction . in fig .
[ gapfunction ] ( a ) we show @xmath378 as a function of @xmath339 at @xmath462 for two temperatures , @xmath59 and @xmath465 which is close to @xmath228 . as the temperature increases , the overall scale of @xmath378 decreases , because the condensate is evaporating , cf.fig .
[ phioft ] .
simultaneously , @xmath466 $ ] increases . in fig .
[ gapfunction ] ( b ) we plot @xmath449 as a function of @xmath467 at the same two temperatures . as can be seen ,
once we divide out the overall scale @xmath379 , the ratio @xmath449 is relatively insensitive to temperature . at @xmath465 , we observe that the gap function does `` flatten '' as it approaches the fermi surface , @xmath468 , as we argued previously on the basis of eq .
( [ qcdgapt6 ] ) .
the gap matrix @xmath98 for a condensate with total spin @xmath0 has the form ( [ generalj0 ] ) .
generalizing this equation to a condensate with total spin @xmath1 is straightforward . for a @xmath1 condensate , the individual gap functions @xmath469 in eq .
( [ generalj0 ] ) are no longer scalar functions in coordinate space , but 3-vectors , @xmath470 .
these vectors have a component along the direction of @xmath13 , @xmath471 and two components transverse to this direction , @xmath472 such that @xmath473 the gap matrix @xmath98 is , however , still a scalar in coordinate space .
one therefore has to contract the vector indices of @xmath474 with the other independent 3-vector at our disposal , @xmath475 , @xmath476 where we have defined @xmath477 and used the projector property @xmath478 .
suppressing color and flavor indices for the moment , the most general ansatz for a gap matrix describing @xmath1 condensates is therefore @xmath479 \ , { \cal p}_{hs}^e({\bf k } ) \,\ , .\ ] ] the scalar product @xmath480 has simple commutation properties with the quasiprojectors @xmath129 : [ commutations ] @xmath481\,\ , .
\label{commutations3}\end{aligned}\ ] ] in the massless limit , @xmath482 , and @xmath483 commutes with the energy projectors . instead of contracting the _ transverse _ components ( [ transverse ] ) of @xmath484 with @xmath475 ,
one could have alternatively also used @xmath484 itself .
the commutation properties for @xmath485 , however , are more complicated than for @xmath486 , with additional terms proportional to @xmath487 .
the advantage of using @xmath488 is that these terms are cancelled by the projector @xmath489 .
finally , an important question that arises is why we choose @xmath475 to contract the transverse components of @xmath470 in eq .
( [ generalj1 ] ) .
why not @xmath490 , or @xmath491 ?
the answer lies again in the commutation properties ( [ commutations ] ) . in the commutation property for @xmath492 or @xmath493 corresponding to ( [ commutations3 ] ) , the sign of the energy on the right - hand side would be flipped .
physically , this means that such terms represent pairing of particles with antiparticles , and not of particles with particles , cf .
discussion of eq .
( [ decomposition ] ) in subsection [ ssurlimit ] .
this is clearly not what happens in a superconductor , where particles form pairs near the fermi surface .
the @xmath1 condensates can be decomposed according to their transformation properties under parity .
let us first write the interaction term @xmath494 in the effective action @xcite in the form @xmath495 let us define new condensate fields @xmath496 in terms of these fields , the interaction term ( [ interaction ] ) assumes the form @xmath497 \frac{1+e\,\gamma_0\ , \bbox{\gamma } \cdot \hat{\bf k}}{2 } \
, \lambda^e ( { \bf k } ) \right .
\nonumber \\ & & \hspace*{1.7 cm } \left .
+ \left [ \bbox{\pi}_+^e(k ) \cdot \left ( \hat{\bf k } + \bbox{\gamma}_\perp ( { \bf k } ) \right ) + \bbox{\pi}_-^e(k ) \cdot \left(\hat{\bf k } + \bbox{\gamma}_\perp ( { \bf k } ) \right ) \gamma_5 \ , \right ] \frac{1-e\,\gamma_0\ , \bbox{\gamma } \cdot \hat{\bf k}}{2 } \ , \lambda^e ( { \bf k } ) \right\ } \psi(k ) \,\ , .
\label{int2}\end{aligned}\ ] ] in the limit @xmath498 , @xmath499 , such that the terms proportional to @xmath500 vanish . under a parity transformation , @xmath501 , @xmath502 , @xmath503 , and @xmath504 .
thus , the terms @xmath505 and the energy projectors do not change under a parity transformation .
the parity of @xmath506 is opposite to that of @xmath507 @xcite .
a parity transformation leaves the effective action invariant , since the qcd lagrangian ( from which the effective action is derived ) does not violate parity explicitly .
thus , if we perform a parity transformation of the interaction term ( [ int2 ] ) , the gap functions @xmath508 must have the same parity as the accompanying dirac matrices .
consequently , we conclude that @xmath509 are parity - even , while @xmath510 are parity - odd .
two limiting cases are of special interest . if all right - handed , positive - helicity gaps are equal to the left - handed , negative - helicity gaps , @xmath511 , and all right - handed , negative - helicity gaps are equal to the left - handed , positive - helicity gaps , @xmath512 , then @xmath513 , _
i.e. _ , the odd - parity gaps vanish , and condensation occurs exclusively in the @xmath514 channel . [ it is proven in appendix [ a ] that in this case eq . ( [ int2 ] )
agrees with the ansatz of bailin and love for the @xmath515 gaps @xcite . ] on the other hand , if they are equal in magnitude , but different in sign , condensation occurs exclusively in the @xmath516 channel .
this is different for the @xmath0 gaps . in that case
, one can write down an equation analogous to ( [ int2 ] ) .
the difference is that all vectors @xmath517 are replaced by scalar functions @xmath518 , and the terms @xmath519 are absent , too . for the parity transformation properties , this has the consequence that if @xmath520 , @xmath521 , condensation occurs in the @xmath522 channel . in the following , we exclusively consider massless fermions . in this case , the helicity projectors ( and corresponding indices ) become superfluous , cf .
( [ urlimit ] ) , @xmath523 and the quasiprojectors become true projectors , cf .
discussion above [ and eq .
( b29 ) @xcite ] .
the general ansatz ( [ generalj1 ] ) simplifies to @xmath524 \
, { \cal p}_h^e ( { \bf k } ) \,\ , .\ ] ] in the effective action , the interaction term @xmath525 decomposes into four parts on account of the projectors @xmath526 [ cf .
( b34 ) of @xcite ] , @xmath527\,\,.\ ] ] here we defined @xmath528 , and used the commutation properties ( [ commutations ] ) of @xmath483 . from ( [ decomposition ] ) it is obvious that the longitudinal gaps correspond to condensation of quarks with the _ same _ chirality , @xmath529 or @xmath530 . in this respect
the longitudinal @xmath1 gaps are similar to the @xmath0 gaps @xcite .
on the other hand , the transverse @xmath1 gaps correspond to condensation of quarks with _ different _ chiralities , @xmath531 or @xmath532 .
equation ( [ decomposition ] ) shows why we do not choose @xmath533 or @xmath534 to contract the transverse components of @xmath470 in ( [ generalj1 ] ) , cf .
discussion at the end of subsection [ classi ] .
the extra factors @xmath535 or @xmath536 flip the sign of the energy in ( [ commutations3 ] ) , and the transverse gaps would describe pairing of particles with antiparticles of the same chirality . with the help of the symmetry property [ cf .
( b4 ) of @xcite ] @xmath537^t\,\ , , \ ] ] we derive @xmath538^t = - \varphi_h^e ( k ) \;\;\;\ ; , \;\;\;\ ; \left[\bbox{\chi}_h^e(-k)\right]^t = - \bbox{\chi}_{-h}^e(k ) \,\ , .\ ] ] the symmetry properties ( [ symmetryprops ] ) allow us to classify the possible color and flavor representations of the @xmath1 condensates for massless fermions .
we assume there are @xmath38 flavors of massless fermions with a global flavor symmetry @xmath539 and , of course , a local color symmetry @xmath111 . from group theory , @xmath540 where the subscripts @xmath541 and @xmath55 denote antisymmetric or symmetric representations , respectively . for single - gluon exchange , the color - antitriplet channel @xmath542 is attractive , and the color - sextet channel @xmath543 is repulsive .
we first discuss the longitudinal condensates @xmath544 . from eq .
( [ decomposition ] ) , the longitudinal @xmath1 condensates only couple quarks of the same chirality , and so do not break the @xmath539 flavor symmetry . by ( [ symmetryprops ] ) , the longitudinal condensates must correspond to a color - flavor representation which is overall antisymmetric . for @xmath50 ,
the flavor representation is trivial , and condensation must occur in the @xmath542 channel .
this is most likely the favored channel for condensation of quarks of the same flavor , since the @xmath0 gaps are overall symmetric @xcite , and consequently must be in the repulsive @xmath543 channel . for @xmath545 ,
the allowed color - flavor representations are either @xmath546 , which is favored , or @xmath547 .
this is in contrast to the @xmath0 gaps , which come in @xmath548 or @xmath549 . here
the flavor representation refers to either @xmath550 or @xmath551 . for @xmath74 , there are @xmath552 or @xmath553 . on the other hand ,
the @xmath0 gaps are @xmath554 or @xmath555 . if only longitudinal @xmath1 gaps are present , the parity properties are analogous to the @xmath0 gaps . in particular , since condensates of different chiralities do not mix , the magnitude of the longitudinal @xmath1 gaps will be equal , while their relative phase represents the spontaneous breaking of parity @xcite .
the symmetry relation ( [ symmetryprops ] ) relates transverse @xmath1 condensates of different chirality .
therefore , we can not draw general conclusions about the possible color - flavor representations of the transverse @xmath1 gaps . _ a priori _ , both the symmetric as well as the antisymmetric color and flavor representations are allowed .
two special cases are of interest . if @xmath556 , so that condensation is in the @xmath515 channel , the color - flavor representations of the transverse @xmath1 gaps are identical to those of the longitudinal @xmath1 gaps . on the other hand , if @xmath557 , so that condensation is in the @xmath516 channel , they are equal to those of the @xmath0 gaps . when @xmath558 , the appearance of transverse @xmath1 gaps must necessarily break the @xmath539 symmetry to a vector - like @xmath559 symmetry .
this striking feature arises because the transverse gaps are proportional to @xmath483 , and not , say , @xmath560 . in general
, condensation can occur in channels with arbitrary total spin @xmath561 .
therefore , the gap matrix @xmath562 will not simply be of the @xmath0 form ( [ generalj0 ] ) , or the @xmath1 form ( [ generalj1 ] ) , but it will be a sum of ( [ generalj0 ] ) and ( [ generalj1 ] ) , as well as contain contributions from higher spin @xmath563 .
we do not attempt to solve this problem in full generality .
instead , we consider the simpler case where the gap matrix contains just @xmath0 and @xmath1 contributions . we also take the fermions to be massless .
the gap matrix then assumes the form @xmath564 \ , { \cal p}_h^e({\bf k } ) \,\ , .\ ] ] where @xmath143 denote the @xmath0 gaps , and @xmath544 the longitudinal and @xmath565 the transverse components of the @xmath1 gaps .
the quasiparticle propagator can be computed from eqs .
( [ quasiprop ] ) , ( [ selfenergy ] ) , and ( [ chargeconjcond ] ) , and the commutation property ( [ commutations3 ] ) . ignoring the color and flavor structure for the moment , one obtains @xmath566^\dagger + [ \varphi_h^e(q)]^\dagger - [ \bbox{\chi}_{-h}^e(q)]^\dagger \cdot \bbox{\gamma}_\perp ( { \bf q } ) \right ) \left ( \frac { } { } \phi_h^e(q ) + \varphi_h^e(q ) + \bbox{\chi}_h^e(q ) \cdot \bbox{\gamma}_\perp ( { \bf q } ) \right ) \right ] { \cal p}_h^e({\bf q } ) \right\}^{-1 } \left[g_0 ^ -(q ) \right]^{-1}. \nonumber\end{aligned}\ ] ] to invert the term in curly brackets , we first have to specify the color - flavor structure of the various condensates . in the following , we just consider the most simple case @xmath50 .
let us assume that condensation occurs exclusively in the attractive color - antitriplet channel @xmath542 , _
i.e. _ , @xmath567 is an antisymmetric @xmath66 matrix in color space , @xmath568 . in this case , the @xmath0 gaps vanish , @xmath569 , on account of fermi statistics @xcite .
since the individual gap functions @xmath570 are also in the @xmath542 representation , we conclude from ( [ symmetryprops ] ) that @xmath571 .
in contrast , @xmath572 and @xmath573 remain _ a priori _ unrelated .
the term in curly brackets in ( [ quasipropj01 ] ) has an off - diagonal contribution in color space .
since its contribution to @xmath574 is quadratic in @xmath575 , we shall neglect it in the following , and consider only the diagonal part , @xmath576 .
even so , the inversion of the term in curly brackets is still cumbersome , due to the presence of terms @xmath577 . to simplify the treatment , we assume that the @xmath1 gaps are _ real - valued_. with eq .
( [ symmetryprops ] ) this leads to @xmath578^\dagger = \left [ \varphi_h^e ( -k ) \right]^t \equiv - \varphi_h^e ( k ) \;\;\;\ ; , \;\;\;\ ; \left [ \bbox{\chi}_{-h}^e ( k ) \right]^\dagger = \left [ \bbox{\chi}_{-h}^e ( -k ) \right]^t \equiv - \bbox{\chi}_h^e ( k ) \,\ , .\ ] ] this has the consequence that all cross terms @xmath577 between the gaps @xmath544 and @xmath565 vanish in ( [ quasipropj01 ] ) .
the quasiparticle propagator can now be explicitly computed : @xmath579 ^ 2}\ , \left[g_0 ^ - ( q ) \right]^{-1 } \,\ , .
\label{prop}\ ] ] the quasiparticle energies are @xmath580^{1/2 } \,\ , .\ ] ] inserting ( [ prop ] ) into the gap equation ( [ gapeq ] ) , the same steps that led to ( [ 19 ] ) now lead to @xmath581 ^ 2 } \ , { \cal p}_{-h}^{-e}({\bf q})\ , \gamma^\nu \,\ , .\ ] ] the gap equations for different fundamental colors @xmath582 decouple , therefore we omit the color index in the following .
taking projections we arrive at the following equation for the longitudinal gap functions : @xmath583 ^ 2 } \ , { \rm tr } \left[{\cal p}_h^e({\bf k})\ , \gamma^\mu \ , { \cal p}_{-h}^{-e}({\bf q})\ , \gamma^\nu \right ] \right .
\nonumber \\ & & \hspace*{3.6 cm } \left .
+ \ , \frac{\varphi_h^{-e}(q)}{q_0 ^ 2- \left[\epsilon^{-e}_q(\bbox{\phi}_h^{-e})\right]^2 } \ , { \rm tr } \left[{\cal p}_h^e({\bf k } ) \ , \gamma^\mu\ , { \cal p}_{-h}^e({\bf q } ) \
, \gamma^\nu \right ] \right\ } \,\ , .\end{aligned}\ ] ] this equation is rather similar to eq .
( [ projectedgapeq ] ) .
the only difference is the appearance of the transverse gaps in the quasiparticle and quasi - antiparticle energies @xmath584 , cf.eq .
( [ quasipenergy ] ) . for the transverse gaps , we use
the fact that @xmath585 to arrive at @xmath586 ^ 2 } \cdot \frac{1}{2}\ , { \rm tr } \left[\bbox{\gamma}\ , \lambda^{-e}({\bf q } ) \ , \gamma^\nu \ , \lambda^e({\bf k})\ , \bbox{\gamma}_\perp({\bf k})\ , \gamma^\mu \right ] \right .
\nonumber \\ & & \hspace*{3.6 cm } \left .
+ \ , \frac{\bbox{\chi}_h^{-e}(q)}{q_0 ^ 2 - \left[\epsilon^{-e}_q(\bbox{\phi}_h^{-e})\right]^2 } \cdot \frac{1}{2}\ , { \rm tr } \left[\bbox{\gamma}\ , \lambda^e({\bf q } ) \ , \gamma^\nu \ , \lambda^e({\bf k})\ , \bbox{\gamma}_\perp({\bf k } ) \
, \gamma^\mu \right ] \right\ } \,\ , .\end{aligned}\ ] ] taking coulomb gauge for the gluon propagator , eq.([coulombgauge ] ) , and computing the traces similar to eqs .
( [ traces ] ) , we obtain @xmath587 ^ 2 } \cdot \left ( { \bf 1}\ , \frac{1 + \hat{\bf k } \cdot \hat{\bf q}}{2 } - \hat{\bf k}\ , \frac{\hat{\bf k } + \hat{\bf q}}{2 } \right ) \right .
\nonumber \\ & & \hspace*{4.3 cm } \times \left [ \delta_l(k - q ) - \delta_t(k - q)\ , \left(1- \frac{(k- q)^2}{({\bf k } - { \bf q})^2 } \right ) \right ] \nonumber \\ & & \hspace*{1.6 cm } \left . + \frac{\bbox{\chi}_h^{-e}(q)}{q_0 ^ 2 - \left[\epsilon^{-e}_q(\bbox{\phi}_h^{-e})\right]^2 } \cdot \left ( { \bf 1}\ , \frac{1 - \hat{\bf k } \cdot \hat{\bf q}}{2 } - \hat{\bf k}\ , \frac{\hat{\bf k } - \hat{\bf q}}{2 } \right ) \right .
\nonumber \\ & & \hspace*{4.3 cm } \left
. \times \left [ \delta_l(k - q ) - \delta_t(k - q)\ , \left(1- \frac{(k+ q)^2}{({\bf k } - { \bf q})^2}\right ) \right ] \right\ } \,\ , .\end{aligned}\ ] ] neglecting the antiparticle contribution , and taking @xmath588 , as well as @xmath589 ( the collinear limit ; in eq .
( [ particlegap2 ] ) this corresponds to neglecting terms of order @xmath282 ) , we obtain for the transverse quasiparticle gap @xmath590 ^ 2 } \ , \left [ \delta_l(k - q ) - \delta_t(k - q ) \right ] \,\ , .\ ] ] to leading logarithmic order , the gap equation for the transverse gap is _ identical _ to that for the longitudinal gap . to see this directly
, one computes the traces in eq .
( [ projectedgapeqj1 + ] ) .
one obtains an equation very similar to eq .
( [ projectedgapeq2 ] ) .
now one computes the angular factors for @xmath589 , @xmath591 : @xmath592 which proves our assertion .
consequently , to this order all @xmath1 gaps are of equal magnitude .
further , the sum of the squares of the @xmath1 gaps fulfills the same gap equation as the square of the @xmath0 gap .
this conclusion agrees with son s renormalization - group analysis @xcite , who argued that the parametric dependence on @xmath4 of any spin @xmath561 gap is the same .
we find that even the prefactor is the same to leading logarithmic accuracy .
our results differ from those of hsu and schwetz @xcite , who argued that @xmath0 gaps are favored over those for higher spin . beyond leading logarithmic order
we suggest that , if @xmath0 gaps are allowed by color - flavor symmetry , they are probably favored over the @xmath1 gaps . if only a @xmath1 gap is allowed , as for @xmath60 , we believe that either the longitudinal gaps , or the transverse gaps with a definite color - flavor representation , will be favored .
which one is favored will be determined by the constants @xmath41 and @xmath52 in eqs .
( [ b0 ] ) and ( [ phi1 ] ) .
we conclude by addressing effects which can contribute to the constants @xmath41 and @xmath52 in the condensate , eqs.([b0 ] ) and ( [ phi1 ] ) .
we suggest that our lengthy calculations which led to the result ( [ b0 ] ) for @xmath37 may be done much more efficiently by constructing an effective theory for quarks near the fermi surface , as initiated by hong @xcite .
this is presumably the easiest way to calculate @xmath41 and @xmath52 as well .
nevertheless , we can estimate what kind of effects could contribute to these as of yet undetermined constants .
these include : * ( i ) gauge dependence of the condensate : * + if calculated properly , any physical quantity must be independent of the choice of gauge . for color superconductivity , what is physical is the gap on the quasiparticle mass shell . at nonzero temperature and zero quark density , general arguments due to kobes , kunstatter , and rebhan @xcite
indicate that the mass shells for quarks and gluons are gauge invariant .
their proof does not extend obviously to nonzero density , but we shall assume that to be the case . at higher orders , the quasiparticle self energy , @xmath102 ,
contains not only the interaction with the condensate , eq .
( [ selfenergy ] ) , but also wave - function renormalization . any apparent gauge dependence in the former
must be cancelled by the gauge dependence of wave - function renormalization , leaving the quasiparticle mass shell gauge invariant .
it is easy to see that such subtleties do not enter at the order to which we have computed .
the gluon propagator in coulomb gauge , with gauge fixing parameter @xmath593 , is given in ( [ coulombgauge ] ) .
alternatively , one could have taken covariant gauge ; with gauge fixing parameter @xmath594 , the covariant gauge propagator was given in eq .
( 37 ) of @xcite .
we have seen , however , that to the accuracy to which we compute , only the gluon propagator in the static limit , @xmath595 , matters . in the static limit , the gauge dependent terms are identical for either coulomb or covariant gauges , and appear only as spatially longitudinal terms , @xmath596 .
consider , for example , how the gauge dependent part of the gluon propagator affects the quasiparticle contribution , @xmath597 ^ 2\right\ } $ ] , to the gap equation for @xmath163 . from the gap equation ( [ projectedgapeq ] ) , and using ( [ traces ] ) , this becomes @xmath598 \sim - \xi_c \ , \frac{1 + \hat{\bf k } \cdot \hat{\bf q}}{2}\ , \frac{(k - q)^2}{p^2 } \rightarrow 0 \;\ ; , \;\
; k , q \rightarrow \mu \ ; .\ ] ] that is , there are gauge dependent terms , but they only contribute to the antiparticle gaps , @xmath162 .
( further , the antiparticle gaps must be computed on their proper mass shell . at the fermi surface ,
@xmath599 is not small , in contrast to @xmath600 . )
consequently , gauge dependent terms in the particle gaps do not appear to even affect the prefactor in the condensate , the constant @xmath41 .
these conclusions about gauge invariance agree with the results of schfer and wilczek @xcite .
in contrast , hong _ et al . _
@xcite argue that landau gauge is preferred , as in other approximate treatments of schwinger - dyson equations .
we insist that in the present example , direct calculation demonstrates gauge invariance without further ado . *
( ii ) wave - function and vertex renormalization : * + as noted by son @xcite , one can have infrared singular factors for wave - function renormalization . for non - relativistic fermions ,
this was noted long ago by holstein , norton , and pincus @xcite .
( this wave - function renormalization is not the hdl correction discussed by schfer and wilczek @xcite ; such corrections involve two hard lines , and are down by @xmath5 . ) the dominant corrections involve a very soft transverse gluon on a quark line ; this produces gauge - invariant terms of the form @xmath601 when @xmath602 .
this correction was computed by brown , liu , and ren @xcite , who find that it is a large effect . besides such wave - function renormalization
, one might expect that the slavnov
taylor identities would also generate similar corrections for the gluon , and for the quark - quark - gluon vertex .
this was not found , however , by the authors of @xcite . *
( iii ) effects of the condensate : * + we have computed the gap using an hdl - resummed gluon propagator .
this is possible because the momenta which generate the gap are much larger than the scale of the gap . to see this , note that in the transverse gluon propagator , ( [ approxspec ] ) , landau damping contributes to the gluon propagator when the momentum @xmath603 ; since the frequency @xmath179 , the dominant momenta are @xmath604 , which for small @xmath302 is much larger than @xmath302 .
effects of the condensate on the gluon propagator can be estimated by power counting at large momentum , @xmath605 .
one would naturally expect that they are @xmath606 , but due to an infrared divergence , they are larger , @xmath607 @xcite .
these terms are important when @xmath608 , or @xmath604 .
this is exactly the same scale at which landau damping operates .
these effects will not alter the coefficient of the logarithmic divergence ( and hence the exponent ) , but they will produce terms of order one in the gap equation , which contribute to @xmath41 and @xmath52 . *
( iv ) damping of the condensate : * + in the above , we neglected the fact that the gap function has an imaginary part . to understand this imaginary part ,
consider first the self energy for a quark in a fermi sea .
as computed by le bellac and manuel , and by vanderheyden and ollitrault @xcite , away from the edge of the fermi sea , the quark can decay into another quark and a very soft gluon .
this is only possible with a hdl - resummed gluon , whose spectral representation has support from landau damping in the space - like region .
the damping rate of the quark behaves as @xmath609 , vanishing at the fermi surface . from a similar physical process ,
the gap function acquires a nonzero imaginary part when its momentum is away from the fermi surface .
a quark can scatter into a quark with a different momentum through a very soft gluon .
we can estimate the resulting imaginary part of the gap function as follows @xcite . if we had not restricted our analysis to the principal value part of the energy denominators arising in ( [ matsu ] ) , instead of @xmath610 in eq .
( [ gapeqfinal ] ) we would have obtained @xmath611 .
this logarithm has a cut for @xmath612 , giving rise to an imaginary part for @xmath331 , @xmath613 taking @xmath614 , cf .
( [ defvarx ] ) , momenta exponentially close to the fermi surface occur when @xmath615 ( in weak coupling ) . in this region ,
the imaginary part of the gap function , @xmath616 , is _ down _ by @xmath365 relative to the real part , @xmath617 . at the fermi surface itself , @xmath359 ,
the imaginary part vanishes .
away from the fermi surface , @xmath618 , so @xmath619 , and @xmath331 is strongly damped , with the real and imaginary parts of comparable magnitude , @xmath620 .
a gap function with a nonzero imaginary part is actually well known from strongly coupled superconductors , as studied in eliashberg theory @xcite .
damping occurs for a similar reason as here , due to a nonzero imaginary part for the plasmon .
there is no problem in principle with including the damping of the gap function .
a spectral representation for the gap function is introduced , analogous to that of the quark and gluon propagators .
the resulting gap integrals are more involved ( especially at nonzero temperature ) , but can be treated in the manner which we employed above . * ( v ) magnetic mass : * + as argued in section i , at zero temperature the scale for the magnetic mass is @xmath21 .
it is therefore negligible compared to the scale of the condensate . at nonzero temperature ,
the scale is no larger than @xmath621 . for @xmath622 ,
this is down by @xmath5 , and will only affect the prefactor of the gap to higher order in @xmath4 .
we conclude by stressing that the determination of the prefactor is not merely an interesting problem in its own right , but because it truly determines the physics of color superconductivity ( at least in weak coupling ) . to the order at which we compute
, there is absolutely _ no _ preference for the condensate to favor spin - zero over spin - one ( or spin - two , _ etc . _ ! ) .
surely the spin - one condensate is less favored than spin - zero ; the ratio of the two condensates is , in weak coupling , a pure number which can be uniquely computed , once one knows how to compute the prefactor in the condensate . indeed , perhaps one should entertain a more speculative hypothesis . even if a @xmath0 gap is favored , maybe there is always some small admixture of higher - spin gaps , and rotational invariance is inevitably broken in the true ground state of color superconductivity .
+ + + * acknowledgment * + + we acknowledge discussions with d. blaschke , w. brown , v.j .
emery , d.k .
hong , s.d.h .
hsu , m. laine , j.t .
liu , v.n .
muthukumar , k. rajagopal , h.c .
ren , t. schfer , and d.t .
we especially thank t. schfer and d.t .
son for discussions on the ratio @xmath464 .
r.d.p . was supported in part by doe grant de - ac02 - 98ch10886 .
thanks riken , bnl and the u.s . dept . of energy for providing the facilities essential for the completion of this work , and columbia university s nuclear theory group for continuing access to their computing facilities .
according to the results of subsection [ parity ] , if condensation occurs exclusively in the @xmath515 channel , the ansatz for the gap matrix reads ( we suppress the dependence of the gap functions on @xmath623 in the following ) @xmath624 \ , \lambda^e ( { \bf k})\,\ , .\ ] ] now use @xmath625 \nonumber \\ & + & \bbox{\gamma } \ , \frac{1}{4}\ , \left[(1 \pm \beta_{\bf k})\ , ( 1 \pm e \ , \gamma_0\ , \bbox{\gamma } \cdot \hat{\bf k } ) \mp \alpha_{\bf k}\ , \bbox{\gamma } \cdot \hat{\bf k } + e\ , \alpha_{\bf k } \ , \gamma_0 \right ] \end{aligned}\ ] ] to obtain @xmath626 \cdot \hat{\bf k } \nonumber \\ & + & \left [ - \frac{1 + \beta_{\bf k } - \alpha_{\bf k}}{4 } \ , ( \bbox{\phi}_+^+ - \bbox{\phi}_+^- ) + \frac{1- \beta_{\bf k } + \alpha_{\bf k}}{4 } \ , ( \bbox{\pi}_+^+ - \bbox{\pi}_+^- ) \right]\cdot \hat{\bf k } \ , \gamma_0 \nonumber \\ & + & \left [ - \frac{1 + \beta_{\bf k } + \alpha_{\bf k}}{4 } \ , ( \bbox{\phi}_+^+ + \bbox{\phi}_+^- ) - \frac{1- \beta_{\bf k } - \alpha_{\bf k}}{4 } \ , ( \bbox{\pi}_+^+ + \bbox{\pi}_+^- ) \right]\cdot \hat{\bf k } \ , \bbox{\gamma } \cdot \hat{\bf k } \nonumber \\ & + & \left
[ \frac{1 + \beta_{\bf k } + \alpha_{\bf k}}{4 } \ , ( \bbox{\phi}_+^+ - \bbox{\phi}_+^- ) - \frac{1- \beta_{\bf k } - \alpha_{\bf k}}{4 } \ , ( \bbox{\pi}_+^+ - \bbox{\pi}_+^- ) \right]\cdot \hat{\bf k } \ , \gamma_0
, \bbox{\gamma } \cdot \hat{\bf k } \nonumber \\ & + & \left [ \frac{1 + \beta_{\bf k}}{4 } \ , ( \bbox{\phi}_+^+ + \bbox{\phi}_+^- ) + \frac{1- \beta_{\bf k } } { 4 } \ , ( \bbox{\pi}_+^+ + \bbox{\pi}_+^- ) \right ] \cdot \bbox{\gamma } \nonumber \\ & + & \frac{\alpha_{\bf k}}{4 } \ , \left ( \bbox{\phi}_+^+ - \bbox{\phi}_+^- + \bbox{\pi}_+^+ - \bbox{\pi}_+^- \right)\cdot \bbox{\gamma } \ , \gamma_0 \nonumber \\ & + & \frac{\alpha_{\bf k}}{4 } \ , \left ( - \bbox{\phi}_+^+ - \bbox{\phi}_+^- + \bbox{\pi}_+^+ + \bbox{\pi}_+^- \right)\cdot \bbox{\gamma } \ , \bbox{\gamma } \cdot \hat{\bf k } \nonumber \\ & + & \left [ \frac{1 + \beta_{\bf k}}{4 } \ , ( \bbox{\phi}_+^+ - \bbox{\phi}_+^- ) - \frac{1- \beta_{\bf k}}{4 } \ , ( \bbox{\pi}_+^+ - \bbox{\pi}_+^- ) \right]\cdot \bbox{\gamma } \ , \gamma_0 \ , \bbox{\gamma } \cdot \hat{\bf k } \,\ , .
\label{explicitj1}\end{aligned}\ ] ] bailin and love s ansatz for a @xmath515 gap reads [ cf .
( 4.24 ) of @xcite ; we again suppress the momentum dependence of the gap functions ] @xmath627 [ we added an @xmath582 in the last term as compared to @xcite .
this simplifies the notation in the following , but is not essential , as the gap functions are in general complex - valued . ] with the definition of @xmath628 one computes @xmath629 and rewrites ( [ bl ] ) as @xmath630 direct comparison of ( [ explicitj1 ] ) and ( [ bl2 ] ) reveals these relations exhibit a redundancy in bailin and love s ansatz ( [ bl ] ) : only four of the eight 3-vector gap functions @xmath632 are independent .
the reason for this redundancy is that in eq .
( [ bl ] ) the transverse components of @xmath633 and @xmath634 never appear , as these gap functions are projected in the longitudinal direction @xmath487 .
furthermore , only the transverse components of @xmath635 and @xmath636 appear on account of ( [ spatprodukt ] ) .
finally , the longitudinal components of the remaining two gap functions @xmath637 and @xmath638 can be absorbed by redefining @xmath639 and @xmath640 , thus only their transverse components contribute . in this manner ,
only half of the original 24 gap functions are independent .
physically , this can be understood from the restriction to the positive - parity channel @xmath515 .
thus , only the four 3-vectors @xmath641 appear on the right - hand side of ( [ relations ] ) .
the other four 3-vectors @xmath642 do not contribute , as they correspond to pairing in the @xmath643 channel .
this said , one can readily derive a more efficient form of bailin and love s ansatz ( [ bl ] ) , which utilizes both longitudinal and transverse components of the truly independent gap functions . choosing the latter to be @xmath644 , and @xmath636 , we obtain from ( [ bl ] ) with ( [ relations ] ) @xmath645 it is now also easy to interpret the results obtained by bailin and love in @xcite , where they studied @xmath514 condensation in the ultrarelativistic limit . in this limit , @xmath646 , thus @xmath647 on account of ( [ relations ] ) .
bailin and love discuss two order parameters for condensation in the @xmath515 channel , the first being the longitudinal component of @xmath648 , the second being @xmath649 . from ( [ relations ] ) , we identify the first with the longitudinal component of @xmath650 . from ( [ spatprodukt ] ) , we realize that the second is identical to @xmath651 , _
i.e. _ , essentially the transverse components of @xmath650 .
thus , bailin and love discuss two separate gap equations , one for the longitudinal , the other for the transverse components of the independent gap function @xmath650 .
although we also find two gap equations for the longitudinal and the transverse gaps , cf .
section [ j1 ] , we do _ not _ find that they decouple , as the excitation energies ( [ quasipenergy ] ) contain _ both _ longitudinal as well as transverse gap functions .
fetter and j.d .
walecka , _ quantum theory of many - particle systems _ ( mcgraw hill , new york , 1971 ) ; a.a .
abrikosov , l.p .
gorkov , and i.e. dzyaloshinski , _ methods of quantum field theory in statistical physics _ ( dover , new york , 1963 ) .
m. alford , k. rajagopal , and f. wilczek , phys .
lett . * b422 * , 247 ( 1998 ) ; r. rapp , t. schfer , e.v .
shuryak , and m. velkovsky , phys .
lett . * 81 * , 53 ( 1998 ) ; hep - ph/9904353 ; n. evans , s.d.h .
hsu , and m. schwetz , nucl .
b551 * , 275 ( 1999 ) ; phys . lett . *
b449 * 281 , ( 1999 ) ; j. berges and k. rajagopal , nucl .
* b538 * , 215 ( 1999 ) ; t. schfer and f. wilczek , phys .
* b450 * , 325 ( 1999 ) ; g.w .
carter and d. diakonov , phys .
d * 60 * , 016004 ( 1999 ) ; k. langfeld and m. rho , hep - ph/9811227 ; m. alford , j. berges , and k. rajagopal , hep - ph/9903502 ; e. shuster and d.t .
son , hep - ph/9905448 ; d.k .
hong , m. rho , and i. zahed , hep - ph/9906551 ; v.a .
miransky , i.a .
shovkovy , and l.c.r .
wijewardhana , hep - ph/9908212 ; r. casalbuoni and r. gatto , hep - ph/9908227 ; hep - ph/9909419 ; b.y .
park , m. rho , a. wirzbad , and i. zahed , sunysb - preprint ( unpublished ) . | we derive perturbatively the gap equations for a color - superconducting condensate with total spin @xmath0 in dense qcd . at zero temperature ,
we confirm the results of son for the dependence of the condensate on the coupling constant , and compute the prefactor to leading logarithmic accuracy . at nonzero temperature
, we find that to leading order in weak coupling , the temperature dependence of the condensate is identical to that in bcs - like theories .
the condensates for total spin @xmath1 are classified ; to leading logarithmic accuracy these condensates are of the same order as those of spin @xmath0 .
16.5 cm -1.5 cm * color superconductivity in weak coupling * + robert d. pisarski @xmath2 and dirk h. rischke @xmath3 + + _ @xmath2department of physics _ + _ brookhaven national laboratory , upton , new york 11973 , u.s.a . _
+ _ email : [email protected]_ + + _ @xmath3riken - bnl research center _
+ _ brookhaven national laboratory , upton , new york 11973 , u.s.a . _
+ _ email : [email protected]_ + + + |
the original idea of kaluza and klein is based on the hypothesis that the input space - time is a @xmath1-dimensional space @xmath2 which can be represented as a tensor product of the visible four - dimensional world @xmath3 with a compact internal @xmath4-dimensional space @xmath5 _ ( 4+d ) = m_4 k_d . [ 1 ] the compact internal space @xmath5 is space - like one i.e. it has only spatial dimensions which may be considered as extra spatial dimensions of @xmath3 . in according with the tensor product structure of the space @xmath6 the metric may be chosen in a factorizable form .
this means that if @xmath7 , ( @xmath8 ) , are local coordinates on @xmath6 then the factorizable metric looks like @xmath9 where @xmath10 is the metric on @xmath3 . in the year 1921 , kaluza proposed a unification of the einstein gravity and the maxwell theory of electromagnetism in four dimensions starting from einstein gravity in five dimensions .
he assumed that the five - dimensional space @xmath11 had to be a product of a four - dimensional space - time @xmath3 and a circle @xmath12 : @xmath13 .
after that the metric @xmath14 in @xmath11 can be decomposed into 10 components describing einstein gravity tensor @xmath15 , four components @xmath16 forming an electromagnetic gauge field , and one component @xmath17 representing a scalar field .
it was shown that the zero mode sector of the kaluza model ( the model is actually a five - dimensional einstein gravity ) is equivalent to the four - dimensional theory which describes the einstein gravity with a four - dimensional general coordinate transformations and the maxwell theory of electromagnetism with a gauge transformations . in his original work
, kaluza assumed the zero mode of scalar field had to be positive constant in order to insure a positive energy .
recently some models with extra dimensions have been proposed to attack the electroweak quantum instability of the standard model known as hierarchy problem between the electroweak and gravity scales . to illustrate the main idea how the hierarchy problem may be solved in a theory with extra dimensions let us consider the einstein @xmath1-dimensional gravity with the action
@xmath18 where @xmath19 , and the ricci scalar curvature @xmath20 is defined by the metric @xmath21 . by the mode expanding and integrating over @xmath22 one
obtains the four - dimensional action @xmath23 where @xmath24 , @xmath25 is einstein gravity tensor on @xmath3 , and the ricci scalar curvature @xmath26 is defined by the metric @xmath25 , @xmath27 is four - dimensional gravitational newton constant related with the fundamental constant @xmath28 by the following equation g_n = g_(4+d ) , [ g ] where @xmath29 is the volume of the compact internal space @xmath5 .
we can rewrite eq .
( [ g ] ) in terms of the four - dimensional planck mass @xmath30 and a fundamental mass scale of the @xmath1-dimensional gravity @xmath31 m_pl^2 = v_d m^d+2 .
[ m ] the latter formula is often cited as the reduction formula .
if @xmath32 is a characteristic size of @xmath5 then @xmath33 , i.e. we suppose @xmath34 , where @xmath35 is some constant .
( [ m ] ) gives m_pl = c_d^1/2m(mr)^d/2.[m2 ] now , it s clear that , if the size @xmath32 of the compact internal space @xmath5 is large compared to the fundamental length @xmath36 , the planck mass is much larger than the fundamental gravity scale .
going further on , if we suppose that the fundamental gravity scale is of the same order as the electroweak scale , @xmath37 , then a huge gap between @xmath38 and @xmath39 is resulted from the large size of the internal extra space @xmath5 .
so , the hierarchy problem replaces by the hierarchy @xmath40 and becomes the problem to explain why the size @xmath32 of extra space @xmath5 is large . from eq .
( [ m2 ] ) , assuming that @xmath37 , it follows r~m^-1()^2/d~10
^ 32/d 10 ^ -17cm.[r ] or @xmath41 an exceptional case @xmath42 gives @xmath43 , ( @xmath44 ) and this is certainly a quite interesting observation . in the case @xmath45 we have @xmath46 , i.e. this case is excluded by unacceptable large value of @xmath32 . for @xmath47
one obtains @xmath48 @xmath49 @xmath50 it is obviously that the basic idea of the kaluza - klein scenario may be applied to any model in quantum field theory .
as example , let us consider the simplest case of ( 4+d)-dimensional model of scalar field with the action s = d^4+dz , [ s ] where @xmath19 , @xmath51 is the metric on @xmath52 , @xmath3 is pseudo - euclidean minkowski space - time , @xmath5 is a compact internal @xmath4-dimensional space with the characteristic size @xmath32 .
let @xmath53 be the laplace operator on the internal space @xmath22 , and @xmath54 are ortho - normalized eigenfunctions of the laplace operator _
k_d y_n(y ) = - y_n(y ) , [ yn ] and @xmath55 is a ( multi)index labeling the eigenvalue @xmath56 of the eigenfunction @xmath54 .
@xmath4-dimensional torus @xmath57 with equal radii @xmath32 is an especially simple example of the compact internal space of extra dimensions @xmath5 .
the eigenfunctions and eigenvalues in this special case look like y_n(y ) = ( i _ m=1^d n_my^m / r ) , [ t ] @xmath58 where @xmath59 are integer numbers , @xmath60 is the volume of the torus . to reduce the multidimensional theory to the effective four - dimensional one we wright a harmonic expansion for the multidimensional field @xmath61 ( z ) = ( x , y ) = _ n ^(n)(x ) y_n(y ) .
[ h ] the coefficients @xmath62 of the harmonic expansion ( [ h ] ) are called kaluza - klein ( kk ) excitations or kk modes , and they usually include the zero - mode @xmath63 , corresponding to @xmath64 and the eigenvalue @xmath65 .
substitution of the kk mode expansion into action ( [ s ] ) and integration over the internal space @xmath22 gives s = d^4x \ { ( _ ^(0 ) ) ^2 - ( ^(0))^2 .
+ ( ^(0))^4 + @xmath66 + \frac{g}{4 ! } ( \phi^{(0)})^{2 } \sum_{n\neq 0 } \phi^{(n ) } \phi^{(n)*}\right\ } + \ldots.\ ] ] for the masses of the kk modes one obtains m_n^2 = m^2 + , [ m ] and the coupling constant @xmath67 of the four - dimensional theory is related to the coupling constant @xmath28 of the initial multidimensional theory by the equation g = , [ g ] where @xmath29 is the volume of the compact internal space of extra dimensions @xmath5 .
the fundamental coupling constant @xmath28 has dimension @xmath68^{-d}$ ] .
so , the four - dimensional coupling constant @xmath67 is dimensionless one as it should be .
( [ m],[g ] ) represent the basic relations of kaluza - klein scenario .
similar relations take place for other types of multidimensional quantum field theoretical models . from four - dimensional point of view
we can interpret each kk mode as a particle with the mass @xmath69 given by eq .
( [ m ] ) .
we see that in according with kaluza - klein scenario any multidimensional field contains an infinite set of kk modes , i.e. an infinite set of four - dimensional particles with increasing masses , which is called the kaluza - klein tower .
therefore , an experimental observation of series kk excitations with a characteristic spectrum of the form ( [ m ] ) would be an evidence of the existence of extra dimensions .
so far the kk partners of the particles of the standard model have not been observed . in the kaluza - klein scenario
this fact can be explained by a microscopic small size @xmath32 of extra dimensions ( @xmath70 ) ; in that case the kk excitations may be produced only at super - high energies of the scale @xmath71 .
below this scale only homogeneous zero modes with @xmath64 are accessible ones for an observation in recent high energy experiments .
that is why , there is a hope to search the kk excitations at the future lhc and other colliders .
however , as we have mentioned above , the solution of hierarchy problem motivated vice versa to introduce the internal extra space with a large size ( @xmath43 ) . to resolve this obvious contradiction
recently it has been proposed a remarkable idea of brane world picture " according to which all matter fields ( except gravity ) are localized on a three - dimensional submanifold
brane embedded in fundamental multidimensional space . in the brane world scenario
extra dimensions may have large and even very large size ; they may also have experimentally observable effects .
even though the models with the brane world scenario may rather seem as exotic ones , nevertheless , they provide a base for a nontrivial phenomenological issues related to the fundamental problems in particle physics and cosmology .
we refer with a pleasure the interested reader to the excellent review articles @xcite and many references therein .
in this note we would like to argue in favour of that the extra dimensions have been observed for a long time in the experiments at very low energies where the nucleon - nucleon dynamics has been studied . in this respect , we would like to show that the structure of proton - proton total cross section at very low energies has a clear signature of the existence of the extra dimensions .
recently a simple theoretical formula describing the global structure of proton - proton total cross - section in the whole range of energies available up today has been derived .
the fit to the experimental data with the formula was made , and it was shown that there is a very good correspondence of the theoretical formula to the existing experimental data obtained at the accelerators @xcite .
moreover it turned out there is a very good correspondence of the theory to all existing cosmic ray experimental data as well .
the predicted values for @xmath72 obtained from theoretical description of all existing accelerators data are completely compatible with the values obtained from cosmic ray experiments @xcite .
the global structure of proton - proton total cross section is shown in fig . 1 extracted from paper @xcite . ( 288,200 ) ( 15,10 ) ( 144,0)@xmath73 ( -5,95 ) the theoretical formula describing the global structure of proton - proton total cross section has the following form @xmath74,\ ] ] @xmath75(gev^{-2}),\ ] ] @xmath76 @xmath77 @xmath78 @xmath79 @xmath80}.\ ] ] for the numerical values of the parameters @xmath81 see original paper @xcite .
the mathematical structure of the formula is very simple and physically transparent : the total cross section is represented in a factorized form ; one factor describes high energy asymptotics of total cross section and it has the universal energy dependence predicted by the general theorems in local quantum field theory ( froissart theorem ) ; the other factor is responsible for the behaviour of total cross section at low energies and it has a complicated resonance structure . the nontrivial feature of the formula is the presence of the new threshold " @xmath82 which is near the elastic one . some information concerning the diproton resonances is collected in table 1 .
the positions of resonances and their widths , listed in table 1 , were fixed in our fit , and only relative contributions of the resonances @xmath83 have been considered as free fit parameters .
fitted parameters @xmath83 obtained by the fit are listed in table 1 too .
diproton resonances . [ cols="<,^,^,>",options="header " , ] the constraint ( [ dio ] ) corresponds to the special ( diophantus ! ) selection of the states .
it s clear that in general case of generic extra compact manifold we would have a significantly more wealthy spectrum of kk - excitations .
we could imagine that there exist such extra compact manifold with a suitable geometry where kk - excitations of a few input fundamental entities ( proton , electron , photon , etc . )
would provide the experimentally observed spectrum of all particles , their resonances and nuclei states . as we hope , it would be possible to find in this way the _ global _ solution of the spectral problem . anyhow ,
we believe that such perfect extra compact manifold with a beautiful geometry and its good - looking shapes exist .
in this short article we have presented some arguments in favour of kaluza and klein ideas genius which are waiting their time in high - energy experiments at future colliders .
in fact , we have shown that kaluza - klein picture of the world has been confirmed in the experiments at very low energies where the nucleon - nucleon dynamics has been studied .
geniusly simple formula
( [ kk ] ) provided by kaluza - klein approach so accurately describes the mass spectrum of diproton system , it s very nice ; certainly , it is not an accidental coincidence .
here we have concerned the simplest model where the protons were considered as a scalar particles .
it is well known that account of fermionic degrees of freedom may result the nontrivial problems related to both the index and the kernel of dirac operator on a generic compact manifold .
however , since the kernel of dirac operator is equal to the kernel of its square , we can say with confidence that account of fermionic degrees of freedom for a proton will not change our main conclusion .
this conclusion is that the existence of the extra dimensions was experimentally proved for a long time , but we did not understand it .
now , seems we understand it . * * v.a .
rubakov , _ large and infinite extra dimensions _ , sov .
j. uspekhi * 171 * , 913 ( 2001 ) ; e - print hep - ph/0104152 .
kubyshin , e - print hep - ph/0111027 .
arkhipov , _ what can we learn from the study of single diffractive dissociation at high energies ? _ in proceedings of viiith blois workshop on elastic and diffractive scattering , protvino , russia , june 28july 2 , 1999 , world scientific , singapore , 2000 , pp .
109 - 118 ; preprint ihep 99 - 43 , protvino , 1999 ; e - print hep - ph/9909531 .
a.a . arkhipov , _ on global structure of hadronic total cross sections _ , preprint ihep 99 - 45 , protvino , 1999 ; e - print hep - ph/9911533 .
arkhipov , _ proton - proton total cross sections from the window of cosmic ray experiments _ , preprint ihep 2001 - 23 , protvino , 2001 ; e - print hep - ph/0108118 ; in proceedings of ixth blois workshop on elastic and diffractive scattering , pruhonice near prague , czech republic , june 9 - 15 , 2001 , eds . v. kundrat , p. zavada , institute of physics , prague , czech republic , 2002 , pp .
293 - 304 .
troyan , v.n .
pechenov , sov .
phys . * 56 * , 191 ( 1993 ) ; yu.a .
troyan , sov .
j. physics of element .
and atomic nuclei * 24 * , 683 ( 1993 ) .
b. tatischeff et al . , phys .
* 2005 ( 1992 ) .
arkhipov , _ on a manifestation of dibaryon resonances in the structure of proton - proton total cross section at low energies _ , preprint ihep 2001 - 44 , protvino , 2001 ; e - print hep - ph/0110399 ; in proceedings of the ninth international conference on hadron spectroscopy , protvino , russia , 25 august-1 september , 2001 , eds .
d. amelin , a.m. zaitsev , melville , new york , 2002 , aip conference proceedings , vol .
771 - 776 .
abramov et al . , sov .
* 57 * , 850 ( 1994 ) .
filkov et al . , e - prints nucl - th/0009044 ; hep - ex/0006029 .
b. tatischeff et al . , z. phys . a*328 * 147 ( 1987 ) .
b. tatischeff et al . , phys .
rev . c*59 * 1878 ( 1999 ) .
b. tatischeff et al . , phys .
rev . c*36 * 1995 ( 1987 ) .
b. tatischeff et al . , in proceedings of the xth international seminar on high energy physics problems , dubna , 24 - 29 september , 1990 , ussr , eds .
a.m. baldin , v.v .
burov , l.p .
kaptari , world scientific , 1990 , p. 177 ; in proceedings of the xiith international seminar on high energy physics problems , vol .
ii , dubna , 1997 , p. 62 .
y. ohashi et al . , phys .
c*36 * 2432 ( 1987 ) .
andreev et al . , z. phys . a*327
* 363 ( 1987 ) .
gareev , g.s .
kazacha , yu.l .
ratis , sov .
j. physics of element .
part . and atomic nuclei * 27 * , 97 ( 1996 ) . | we present some arguments in favour of that the kaluza - klein picture of the world has been confirmed in the experiments at very low energies where the nucleon - nucleon dynamics has been studied . early predicted @xmath0-particle is related to the new scale of strong nucleon - nucleon forces in the spirit of kaluza - klein approach .
it is shown that kk excitations remarkably describe the experimentally observed mass spectrum of diproton system . * @xmath0-particle and kaluza - klein world * + a.a .
arkhipov + _ state research center
institute for high energy physics " + 142280 protvino , moscow region , russia _ + |
in this section we give some basic definitions , and prove a couple of useful facts about algebras and free subsets .
we refer the reader to @xcite for the definitions of set - theoretic terms . throughout this paper we are working in zfc set theory .
let @xmath5 be a set .
@xmath6 is a * finitary function on @xmath5 * if and only if @xmath7 for some finite @xmath8 .
@xmath9 is an * algebra * if and only if @xmath10 where @xmath5 is a set , and @xmath11 is a set of finitary functions on @xmath5 which is closed under composition and contains the identity function . in this case
we say @xmath9 is an * algebra on @xmath5 .
* let @xmath12 be an algebra and let @xmath13 .
then @xmath14 is a * subalgebra of @xmath15 * if and only if @xmath14 is closed under all the functions in @xmath11 . in this case
there is a natural algebra structure on @xmath14 given by @xmath16 .
let @xmath12 be an algebra .
then @xmath17 let @xmath12 be an algebra . if @xmath18 then @xmath19 @xmath20 is the least subalgebra of @xmath9 containing @xmath1 .
let @xmath12 be an algebra . if @xmath18 then @xmath1 is * free * if and only if @xmath21 in the context of boolean algebras , free subsets are more usually referred to as * irredundant .
* we can use free sets to generate large numbers of subalgebras using the following well - known fact .
[ freefact ] if @xmath18 is free for @xmath9 then @xmath22 .
for each @xmath23 let @xmath24 .
it follows from the freeness of @xmath1 that @xmath25 , so that the map @xmath26 is an injection from @xmath27 into @xmath28 .
the next fact ( also well - known ) gives us one way of generating irredundant subsets in a boolean algebra .
[ bafact ] if @xmath29 is a strong limit cardinal and @xmath0 is a boolean algebra with @xmath30 then @xmath0 has an irredundant subset of cardinality @xmath29
. we can build by induction a sequence @xmath31 such that @xmath32 is not above any element in the subalgebra generated by @xmath33 .
the point is that a subalgebra of size less than @xmath29 can not be dense , since @xmath29 is strong limit .
it follows from the property we have arranged that we have enumerated an irredundant subset of cardinality @xmath29 .
we make the remark that the last fact works even if @xmath34 .
the next fact is a technical assertion which we will use when we discuss free subsets and subalgebras in the general setting .
[ singularfact ] let @xmath35 be a singular strong limit cardinal .
let @xmath12 be an algebra and suppose that @xmath36 and @xmath37 .
then @xmath28 has cardinality at least @xmath38 .
let @xmath39 , then @xmath40 since @xmath35 is singular and strong limit .
define @xmath41 from the assumptions on @xmath5 and @xmath35 , @xmath42 so that @xmath43 .
observe that if @xmath44 then @xmath45 . so if we define an equivalence relation on @xmath46 by setting @xmath47 then the classes each have size at worst @xmath48 . hence there are @xmath38 classes and so we can generate @xmath38 subalgebras by closing representative elements from each class .
let @xmath29 be a cardinal .
@xmath49 is the following property of @xmath29 : for all algebras @xmath12 with @xmath50 and @xmath51 there exists @xmath18 free for @xmath15 such that @xmath52 . in some contexts
we might wish to make a more complex definition of the form `` @xmath53 iff for every algebra on @xmath29 with at most @xmath35 functions there is a free set in the @xmath54-complete filter @xmath55 '' but @xmath49 is sufficient for the arguments here .
we collect some information about @xmath49 . if @xmath29 is ramsey then @xmath49 .
let @xmath12 be an algebra with @xmath50 , @xmath51 .
we lose nothing by assuming @xmath56 .
we can regard @xmath15 as a structure for a first - order language @xmath57 with @xmath58 . by a standard application of ramseyness
we can get @xmath59 of order type @xmath29 with @xmath1 a set of order indiscernibles for @xmath15
. now @xmath1 must be free by an easy application of indiscernibility .
[ cc - fact ] if @xmath49 holds and @xmath60 is a @xmath61-c.c .
forcing for some @xmath62 then @xmath49 holds in @xmath63 .
let @xmath64 name an algebra on @xmath29 with @xmath65 functions , @xmath66 .
for each @xmath67 from @xmath29 and @xmath68 we may ( by the chain condition of @xmath60 ) enumerate the possible values of @xmath69 as @xmath70 .
now define in @xmath71 an algebra @xmath72 and use @xmath49 to get @xmath73 free for @xmath74 .
clearly it is forced that @xmath75 be free for @xmath76 in @xmath63 .
in this section we will define a combinatorial principle @xmath77 and show that if @xmath77 holds then we get the desired conclusion about boolean algebras .
we will also consider a limitation on the possibilities for generalising the result .
the principle @xmath77 is the conjunction of the following two statements : 1 . for all infinite cardinals
@xmath29 , @xmath78 is weakly inaccessible and @xmath79 2 . for all infinite cardinals @xmath29
, @xmath80 holds .
[ startoba ] @xmath77 implies that every infinite boolean algebra @xmath0 has an irredundant subset @xmath1 such that @xmath2 .
observe that a boolean algebra can be regarded as an algebra with @xmath81 functions ( just take the boolean operations and close under composition ) .
define a closed unbounded class of infinite cardinals by @xmath82 there is a unique @xmath83 such that @xmath84 . by the assumption s1 , @xmath85 , so it will suffice to find a free subset of size @xmath35 .
since @xmath35 is infinite we can find @xmath86 a subalgebra of @xmath0 with @xmath87 , and it suffices to find a free subset of size @xmath35 in @xmath86 .
case a : @xmath88 . in this case @xmath89 , and s2 implies that @xmath90 holds , so that @xmath86 has a free ( irredundant ) subset of cardinality @xmath35 .
case b : @xmath35 is strong limit . in this case fact [ bafact ] implies that there is an irredundant subset of cardinality @xmath35 . at this point we could ask about the situation for more general algebras .
could every algebra with countably many functions have a large free subset , or a large number of subalgebras ? there is a result of shelah ( see chapter iii of @xcite ) which sheds some light on this question .
if @xmath29 is inaccessible and not mahlo ( for example if @xmath29 is the first inaccessible cardinal ) then @xmath91 ^ 2_{\kappa}$ ] . from this
it follows immediately that if @xmath29 is such a cardinal , then we can define an algebra @xmath15 on @xmath29 with countably many functions such that @xmath92 . in the next section
we will prove that in the absence of inaccessibles some information about general algebras can be extracted from @xmath77 .
the model of @xmath77 which we eventually construct will in fact contain some quite large cardinals ; it will be a set model of zfc in which there is a proper class of cardinals @xmath93 which are @xmath94-supercompact .
in this section we prove that in the absence of inaccessible cardinals @xmath77 implies that certain algebras have a large number of subalgebras . by the remarks at the end of the last section
, it will follow that the restriction that no inaccessibles should exist is essential .
suppose that @xmath77 holds and there is no inaccessible cardinal .
let @xmath29 and @xmath65 be infinite cardinals . if @xmath12 is an algebra such that @xmath95 and @xmath96 , and @xmath97 , then @xmath28 has cardinality @xmath48 .
let @xmath15 , @xmath29 and @xmath65 be as above .
define a class of infinite cardinals @xmath98 by @xmath99 there are no inaccessible cardinals , so @xmath98 is closed and unbounded in the class of ordinals .
let @xmath35 be the unique element of @xmath98 such that @xmath100 , where we know that @xmath35 exists because @xmath101 and @xmath98 is a closed unbounded class of ordinals .
since @xmath102 , @xmath103 .
now let @xmath14 be a subalgebra of @xmath15 of cardinality @xmath35 ; such a subalgebra can be obtained by taking any subset of size @xmath35 and closing it to get a subalgebra .
s1 implies that @xmath104 , so it suffices to show that the algebra @xmath105 has @xmath38 subalgebras .
we distinguish two cases : case a : @xmath35 is singular strong limit . in this case
we may apply fact [ singularfact ] to conclude that @xmath106 has @xmath38 subalgebras .
case b : @xmath88 .
it follows from s2 that @xmath90 is true , hence there is a free subset of size @xmath35 for @xmath106 .
but now by fact [ freefact ] we may generate @xmath38 subalgebras . this concludes the proof . for algebras with countably many functions
the last result guarantees that there are many subalgebras as long as @xmath107 .
in fact we can do slightly better here , using another result of shelah ( see @xcite ) . if @xmath98 is a closed unbounded subset of @xmath108^{\aleph_0}$ ] then @xmath109 . if @xmath77 holds and there are no inaccessibles , then for every algebra @xmath15 with @xmath110 and countably many functions @xmath111 . if @xmath107 we already have it .
if @xmath112 then apply the theorem we just quoted and the fact that @xmath113 .
in this section we sketch the argument of @xcite and indicate how we modify it to get a model in which @xmath77 holds .
we begin with a brief review of our conventions about forcing . for us
@xmath114 means that @xmath115 is stronger than @xmath116 , a forcing is @xmath65-closed if every decreasing sequence of length less than @xmath65 has a lower bound , a forcing is @xmath65-dense if it adds no @xmath117-sequence of ordinals , and @xmath118 is the forcing for adding @xmath65 cohen subsets of @xmath119 .
foreman and woodin begin the construction in @xcite with a model @xmath71 in which @xmath29 is supercompact , and in which for each finite @xmath8 they have arranged that @xmath120 is weakly inaccessible and @xmath121 .
they force with a rather complex forcing @xmath60 , and pass to a submodel of @xmath63 which is of the form @xmath122 for @xmath123 a projection of @xmath60 .
@xmath60 here is a kind of hybrid of magidor s forcing from @xcite to violate the singular cardinals hypothesis at @xmath124 and radin s forcing from @xcite .
just as in @xcite the forcing @xmath60 does too much damage to @xmath71 and the desired model is an inner model of @xmath63 , but in the context of @xcite it is necessary to be more explicit about the forcing for which the inner model is a generic extension .
the following are the key properties of @xmath123 . 1 .
@xmath29 is still inaccessible ( and in fact is @xmath125-supercompact ) in @xmath126 .
2 . @xmath123 adds ( among other things ) a generic club of order type @xmath29 in @xmath29 .
in what follows we will assume that a generic @xmath127 for @xmath123 is given , and enumerate this club in increasing order as @xmath128 .
3 . * in @xmath129 * each cardinal @xmath130 reflects to some extent the properties of @xmath29 .
in particular in @xmath71 each @xmath130 is measurable , and for each @xmath8 we have that @xmath131 is weakly inaccessible and @xmath132 .
4 . for each @xmath93 , if @xmath133 is a condition which determines @xmath130 and @xmath134 then @xmath135 ( the suborder of conditions which refine @xmath115 ) factors as @xmath136 where 1 .
@xmath137 is @xmath138-c.c . if @xmath93 is limit and @xmath139-c.c .
if @xmath140 .
all bounded subsets of @xmath141 which occur in @xmath142 have already appeared in the extension by @xmath143 .
given all this it is routine to check that if we truncate @xmath126 at @xmath29 then we obtain a set model of zfc in which gch fails everywhere , and moreover exponentiation follows the pattern of clause 1 in @xmath77 .
the reader should think of @xmath123 as shooting a club of cardinals through @xmath29 , and doing a certain amount of work between each successive pair of cardinals .
things have been arranged so that gch will fail for a long region past each cardinal @xmath130 on the club , so that what needs to be done is to blow up the powerset of a well - chosen point in that region to have size @xmath134 . for our purposes
we need to change the construction of @xcite slightly , by changing the forcing which is done between the points of the generic club to blow up powersets .
the reason for the change is that we are trying to get @xmath144 to hold whenever @xmath65 is of the form @xmath145 , and in general it will not be possible to arrange that @xmath144 is preserved by forcing with @xmath146 .
the solution to this dilemma is to replace the cohen forcing defined in @xmath71 by a cohen forcing defined in a well chosen inner model .
we need to choose this inner model to be small enough that cohen forcing from that model preserves @xmath144 , yet large enough that it retains some degree of closure sufficient to make the arguments of @xcite go through . for technical reasons
we will be adding subsets to @xmath147 , rather than @xmath148 as in the case of @xcite . to be more precise we will want to replace clauses p@xmath149 and p@xmath150 above by 1 . for each @xmath151 it is the case in @xmath71 that @xmath130 is measurable , @xmath131 is weakly inaccessible , and @xmath152 .
also @xmath153 holds for each such @xmath8 .
+ there exists for all possible values of @xmath130 , @xmath134 a notion of forcing @xmath154 ( which we will denote by @xmath155 in what follows ) such that 1 .
@xmath155 is @xmath148-closed , @xmath156-c.c . and @xmath147-dense .
2 . @xmath155 adds @xmath134 subsets to @xmath147 .
3 . in @xmath157
the property @xmath153 still holds , density guarantees that the property survives for @xmath158 .
] for @xmath151 .
if @xmath115 determines @xmath130 and @xmath134 then @xmath135 factors as @xmath159 where 1 .
@xmath137 is @xmath138-c.c . if @xmath93 is limit , and @xmath160-c.c .
if @xmath161 .
2 . @xmath162 is as above .
3 . all bounded subsets of @xmath163 in @xmath142 are already in the extension by @xmath164 .
we ll show that given a model in which p@xmath165 , p@xmath166 , p@xmath167 and p@xmath168 hold we can construct a model of @xmath77 .
let @xmath127 be generic for a modified version of @xmath123 obeying p@xmath165 , p@xmath166 , p@xmath167 and p@xmath168 . if we define @xmath169 $ ] , then @xmath170 is a model of zfc in which @xmath77 holds .
s1 is proved just as in @xcite .
it follows from our assumptions that in @xmath170 we have the following situation : * @xmath35 is strong limit if and only if @xmath171 for some limit @xmath65 . *
cardinal arithmetic follows the pattern that for @xmath172 @xmath173 while @xmath174 we need to show that for all infinite @xmath65 we have @xmath175 .
the proof divides into two cases . 1 .
@xmath48 is of the form @xmath176 where @xmath177 and @xmath178 is a limit ordinal . by the factorisation properties in p@xmath168 above
it will suffice to show that @xmath175 holds in the extension by @xmath179 .
+ certainly @xmath175 holds in the extension by @xmath180 , because p@xmath167 says just that .
also we have that @xmath181 , and @xmath182 is @xmath183-c.c . so that applying fact [ cc - fact ] we get that @xmath175 holds in the extension by @xmath179 .
@xmath48 is of the form @xmath184 where @xmath185 .
again it suffices to show that @xmath175 holds in the extension by @xmath186 .
+ @xmath175 holds in @xmath187 , @xmath188 and @xmath189 is @xmath160-c.c .
so that as in the last case we can apply fact [ cc - fact ] to get that @xmath175 in the extension by @xmath186 .
most of the hard work in this paper comes in preparing the model over which we intend ( ultimately ) to do our version of the construction from @xcite .
we ll make use of laver s `` indestructibility '' theorem from @xcite as a labour - saving device . at a certain point below
we will sketch a proof , since we need a little more information than is contained in the statement .
[ laverology ] let @xmath29 be supercompact , let @xmath190 .
then there is a @xmath29-c.c . and @xmath178-directed closed forcing @xmath60 , of cardinality @xmath29 , such that in the extension by @xmath60 the supercompactness of @xmath29 is indestructible under @xmath29-directed closed forcing .
now we will describe how to prepare the model over which we will do radin forcing .
let us start with six supercompact cardinals enumerated in increasing order as @xmath191 in some initial model @xmath192 .
we ll force to make each of the @xmath193 indestructibly supercompact under @xmath193-directed closed forcing . to do this
let @xmath194 be laver s forcing to make @xmath195 indestructible under @xmath195-directed closed forcing , and force with @xmath196 . in @xmath197
the cardinal @xmath195 is supercompact by construction , and the rest of the @xmath193 are still supercompact because @xmath198 and supercompactness survives small forcing .
now let @xmath199 be ( as guaranteed by fact [ laverology ] ) a forcing which makes @xmath200 indestructible and is @xmath195-directed closed . in @xmath201 we claim that all the @xmath193 are supercompact and that both @xmath195 and @xmath200 are indestructible .
the point for the indestructibility of @xmath195 is that if @xmath202 is @xmath195-directed closed then @xmath203 is @xmath195-directed closed in @xmath197 , so that @xmath195 has been immunised against its ill effects and is supercompact in @xmath204 . repeating in the obvious way
, we make each of the @xmath193 indestructible .
let the resulting model be called @xmath71 .
now we will force over @xmath71 with a rather complicated @xmath195-directed closed forcing .
broadly speaking we will make @xmath195 have the properties that are demanded of @xmath130 in p@xmath167 from the last section
. of course @xmath195 will still be supercompact after this , and we will use a reflection argument to show that we have many points in @xmath195 which are good candidates to become points on the generic club . to be more precise
we are aiming to make a model @xmath205 in which the following list of properties holds : 1 .
@xmath195 is supercompact .
2 . for each @xmath206 ,
@xmath207 and @xmath208 is a weakly inaccessible cardinal with @xmath209 .
3 . for any @xmath210
there is a forcing @xmath211 such that ( if we define @xmath212 for @xmath178 with @xmath213 ) 4 .
@xmath214 , @xmath215 is @xmath216-directed closed , @xmath217-dense and @xmath218-c.c .
5 . in @xmath219 , @xmath220 and
@xmath221 holds for @xmath222 .
6 . if @xmath223 then @xmath215 is a complete suborder of @xmath224 . for each @xmath206 let @xmath225 be the cohen conditions ( as computed by @xmath71 ) for adding @xmath208 subsets of @xmath193 .
we will abbreviate this by @xmath226 .
let @xmath227 .
let @xmath228 $ ] .
in @xmath170 we know that @xmath207 for @xmath229 , and that @xmath217 is still supercompact since @xmath230 . in @xmath170
define a forcing as in fact [ laverology ] for making @xmath217 indestructible .
more precisely choose @xmath231 such that * in @xmath170 , @xmath232 is @xmath217-c.c . and @xmath233-directed closed with cardinality @xmath217 .
* in @xmath234 $ ] the supercompactness of @xmath217 is indestructible under @xmath217-directed closed forcing .
let @xmath235 $ ] .
define @xmath236 , and observe that for every @xmath178 @xmath237 .
this is really the key point in the construction , to choose @xmath238 in exactly the right inner model ( see our remarks in the last section just before the definition of p@xmath167 ) .
finally let @xmath239 $ ] , where we stress that @xmath240 is cohen forcing as computed in @xmath71 .
we ll prove that we have the required list of facts in @xmath205 . 1 .
@xmath195 is supercompact .
2 . clearly @xmath241 is @xmath195-directed closed in @xmath71 .
@xmath240 is @xmath216-directed closed in @xmath71 so is still @xmath195-directed closed in @xmath242 , hence @xmath243 is @xmath195-directed closed in @xmath71 . by the indestructibility of @xmath195 in @xmath71 ,
@xmath195 is supercompact in @xmath205 .
3 . for each @xmath206 ,
@xmath207 and @xmath208 is a weakly inaccessible cardinal with @xmath209 .
4 . in @xmath170
we have this for @xmath229 .
@xmath240 is @xmath216-closed and @xmath218-c.c . in @xmath71
, @xmath244 is @xmath245-c.c . even in @xmath246
$ ] so by the usual arguments with easton s lemma @xmath240 is @xmath216-dense and @xmath218-c.c . in @xmath170 .
@xmath232 is @xmath233-closed in @xmath170 so that @xmath240 is @xmath216-dense and @xmath218-c.c . in @xmath242 .
moreover , by the closure of @xmath232 we still have the claim for @xmath229 in @xmath242 .
@xmath217 is inaccessible ( indeed supercompact ) in @xmath242 so that after forcing with @xmath240 we have the claim for @xmath206 .
recall that @xmath237
@xmath214 , @xmath215 is @xmath216-directed closed , @xmath217-dense and @xmath218-c.c .
the cardinality statement is clear . @xmath215
is @xmath217-directed closed in @xmath242 , so that ( by what we proved about the properties of @xmath240 in @xmath242 during the proof of k2 ) @xmath215 is @xmath216-directed closed and @xmath217-dense in @xmath205 .
since @xmath215 is @xmath217-closed in @xmath242 , @xmath240 is @xmath218-c.c . in
@xmath247 $ ] .
@xmath215 is @xmath248-c.c . in @xmath242 so that @xmath249 is @xmath248-c.c . in @xmath242 , so that finally @xmath215 is @xmath248-c.c . in @xmath205 .
8 . in @xmath219 , @xmath220 and
@xmath221 holds for @xmath222 .
easily @xmath220 .
we break up the rest of the proof into a series of claims . + for @xmath229 , @xmath221 holds in @xmath250 $ ] .
+ @xmath193 is supercompact in @xmath251 $ ] , because @xmath252 is @xmath193-directed closed .
@xmath253 has @xmath254-c.c . in that model so that @xmath221 holds in @xmath170 .
also we know that @xmath250 = v_1[\dot{{\bbb r}}][\dot{{\bbb p}}_4][\dot{{\bbb q}}_{\eta}]$ ] is an extension of @xmath170 by @xmath216-dense forcing , so @xmath221 is still true in @xmath250 $ ] .
+ @xmath255 holds in @xmath250 $ ] .
+ @xmath256 = v[\dot { { \bbb p}}_{0,4}][\dot{{\bbb r}}]$ ] . arguing as in the previous claim , @xmath255 holds in @xmath257 $ ] .
@xmath240 is @xmath218-c.c . in @xmath170
so @xmath232 is @xmath218-dense in @xmath258$],hence @xmath255 holds in @xmath205 .
finally @xmath215 is @xmath217-dense in @xmath205 , so @xmath255 holds in @xmath259 $ ] .
+ @xmath260 holds in @xmath259 $ ] .
+ @xmath259 = v_2[\dot{{\bbb p}}_4][\dot{{\bbb q}}_{\eta } ] = v_2[\dot{{\bbb q}}_{\eta}][\dot{{\bbb p}}_4]$ ] .
as @xmath215 is @xmath217-directed closed in @xmath242 , @xmath217 is supercompact in @xmath261 $ ] .
@xmath240 is @xmath218-c.c . in @xmath262 $ ] , so that @xmath255 is still true in @xmath259 $ ] .
+ this finishes the proof of k4 .
10 . if @xmath223 then @xmath215 is a complete suborder of @xmath224
this is immediate by the uniform definition of cohen forcing .
we finish this section by proving that in @xmath205 we can find a highly supercompact embedding @xmath263 with critical point @xmath195 such that in @xmath5 the cardinal @xmath195 has some properties resembling k1k5 above .
the point is that there will be many @xmath264 which have these properties in @xmath205 , and eventually we ll ensure that every candidate to be on the generic club added by the radin forcing has those properties . for the rest of this section we will work in the model @xmath205 unless otherwise specified .
a pair of cardinals @xmath265 is _ sweet _ iff ( setting @xmath266 ) 1 .
@xmath93 is measurable .
2 . for each @xmath222 ,
@xmath267 is weakly inaccessible and @xmath268 .
there is a forcing @xmath269 such that ( setting @xmath270 ) 4 .
@xmath271 , @xmath272 is @xmath273-directed closed , @xmath274-dense and @xmath275-c.c .
5 . in the generic extension by @xmath272 , @xmath276 and @xmath277 holds for @xmath222 . 6 . if @xmath278 then @xmath272 is a complete suborder of @xmath279 . let @xmath280 . in @xmath205
there is an embedding @xmath263 such that @xmath281 , @xmath282 , and in the model @xmath5 the pair @xmath283 is sweet .
recall that we started this section with a model @xmath192 , and forced with some @xmath196 to make @xmath195 indestructible .
let @xmath284 be the forcing that we do over @xmath197 to get to the model @xmath242 , and recall that we force over @xmath242 with @xmath240 to get @xmath205 .
@xmath285 is @xmath195-directed closed in @xmath197 .
fix a cardinal @xmath35 much larger than @xmath65 , such that @xmath286 . as in @xcite , fix in @xmath192 an embedding @xmath287 such that 1 .
@xmath288 , @xmath289 .
2 . @xmath290 .
@xmath291 , where @xmath232 is @xmath61-closed in @xmath292 .
let @xmath293 be @xmath196-generic over @xmath192 , let @xmath294 be @xmath285-generic over @xmath295 $ ] .
let @xmath296[g_1][g_2]$ ] and let @xmath297[g_1][g_2]$ ] .
now @xmath298 is closed under @xmath35-sequences in @xmath205 so that the factor forcing @xmath232 is @xmath61-closed in @xmath205 .
let @xmath299 be @xmath232-generic over @xmath205 , then it is easy to see that @xmath300 lifts to @xmath301 { \longrightarrow}n^+[h].\ ] ] now we know that @xmath302 $ ] , and @xmath303 $ ] , so that @xmath304
\in n^+[h]$ ] . by the directed closure of @xmath305 in @xmath306
$ ] we can find a `` master condition '' @xmath115 ( a lower bound in @xmath305 for @xmath304 $ ] ) and then force over @xmath307 $ ] with @xmath308 to get a generic @xmath309 and an embedding @xmath310[x].\ ] ] of course this embedding is not in @xmath205 but a certain approximation is .
observe that @xmath311 is @xmath61-closed in @xmath205 .
now factor @xmath312 through the ultrapower by @xmath313 \in j_2(a ) \ ; \ } } } , \ ] ] to get a commutative triangle @xmath314{e , t}{j_2 } \arrow{se , b}{j } \node[2]{n^ * } \\ \node[2]{m } \arrow{ne , r}{k } \end{diagram}\ ] ] by the closure @xmath315 , so @xmath316 is an * internal * ultrapower map .
it witnesses the @xmath35-supercompactness of @xmath195 in the model @xmath205 , because ( as can easily be checked ) @xmath317 is a fine normal measure on @xmath318 .
it remains to check that @xmath283 is sweet in @xmath5 .
we ll actually check that @xmath319 is sweet in @xmath320 , this suffices because @xmath321 .
clauses 1 and 2 are immediate by the closure of @xmath5 inside the model @xmath205 .
now to witness sweetness let us set @xmath322[g_1 ] } = add({\kappa}_5 , j_2({\kappa}_0))_{n[g_0][g_1]}.\ ] ] clause 3 is true in @xmath298 by the same arguments that we used for @xmath205 above .
@xmath320 is an extension of @xmath298 by highly closed forcing so that clause 3 is still true in @xmath320 .
clause 5 is easy so we are left with clause 4 .
let @xmath323 be @xmath324-generic over @xmath320 for @xmath324 .
then @xmath323 is generic over @xmath298 . by the chain condition of @xmath324 and the closure of @xmath298 in @xmath205 , @xmath205 and
@xmath298 see the same set of antichains for @xmath324 .
hence @xmath323 is @xmath324-generic over @xmath205 , and we showed that @xmath221 holds in @xmath325 $ ] .
but then by closure again @xmath221 holds in @xmath326 $ ] , by easton s
lemma @xmath327 is generic over @xmath326 $ ] for highly dense forcing so that finally @xmath221 holds in @xmath328 $ ] .
in this section we will indicate how to modify the construction of @xcite , so as to force over the model @xmath205 obtained in the last section and produce a model of @xmath77 .
the modification that we are making is so minor that it seems pointless to reproduce the long and complicated definitions of @xmath60 and @xmath123 from @xcite ; in this section we just give a sketch of what is going on in @xcite , an indication of how the construction there is to be modified , and then some hints to the diligent reader as to how the proofs in @xcite can be changed to work for our modified forcing . as we mentioned in section [ starsection ] ,
the construction of @xcite involves building a forcing @xmath60 and then defining a projection @xmath123 ; the idea is that the analysis of @xmath60 provides information which shows that @xmath123 does not too much damage to the cardinal structure of the ground model .
we start by giving a sketchy account of @xmath123 as constructed in @xcite .
recall the the aim of @xmath123 is to add a sequence @xmath329 such that : 1 .
@xmath330 enumerates a closed unbounded subset of @xmath29 .
@xmath331 is @xmath332-generic over @xmath71 .
3 . in the generic extension ,
cardinals are preserved and @xmath29 is still inaccessible
. a condition will prescribe finitely many of the @xmath130 , and for each @xmath130 that it prescribes it will give some information about @xmath331 ; it will also place some constraint on the possibilities for adding in new values of @xmath130 and for giving information about the corresponding @xmath331 .
the idea goes back to prikry forcing , where a condition prescribes an initial segment of the generic @xmath333-sequence and puts constraint on the possibilities for adding further points . to be more specific @xmath123
is built from a pair @xmath334 where @xmath335 and @xmath336 is a measure on @xmath337 for @xmath338 .
@xmath336 will concentrate on pairs @xmath339 that resemble @xmath340 ; in particular if we let @xmath341 , @xmath342 will be a cardinal that resembles @xmath29 .
@xmath343 will be an ultrafilter on a boolean algebra @xmath344 , which consists of functions @xmath345 such that @xmath346 and @xmath347 , with the operations defined pointwise .
conditions in @xmath123 will have the form @xmath348 where 1 .
@xmath349 , @xmath350 .
2 . @xmath351 , @xmath352 .
3 . @xmath353 , @xmath354 .
4 . @xmath355 for @xmath356 , @xmath357 .
the aim of this condition is to force that the cardinals @xmath358 are on the generic club , and that the conditions @xmath359 are in the corresponding generics .
it is also intended to force that if we add in a new cardinal @xmath93 and add condition @xmath360 between @xmath358 and @xmath361 then there is a pair @xmath362 in some @xmath363 such that @xmath364 and @xmath365 .
this motivates the ordering ( which we do not spell out in detail here ) ; a condition is refined either by refining the `` constraint parts '' or by adding in new quintuples that obey the current constraints and which impose constraints compatible with the current ones .
the key fact about @xmath123 is that given a condition @xmath115 , any question about the generic extension can be decided by refining the constraint parts of @xmath115 ( this is modelled on prikry s well - known lemma about prikry forcing ) .
it will follow from this that bounded subsets of @xmath29 are derived from initial segments of the generic as in p4 b ) from section [ starsection ] , which in turn will imply that @xmath123 has the desired effect of causing the gch to fail everywhere below @xmath29 .
it remains to be seen that forcing with @xmath123 preserves the large cardinal character of @xmath29 ; this is done via a master condition argument which succeeds because @xmath366 is quite long ( it has a so - called _ repeat point _ ) and the forcing @xmath367 is @xmath368-directed - closed .
we can now describe how we modify the construction of @xcite . to bring our notation more into line with @xcite ,
let @xmath369 , let @xmath370 , let @xmath371 be the embedding constructed at the end of the last section . for each @xmath372
such that @xmath265 is sweet let us fix @xmath373 witnessing this .
our modification of @xcite is simply to restrict attention to those pairs @xmath374 such that @xmath375 is sweet ( there are sufficiently many because @xmath376 is sweet in @xmath5 ) and then to replace the forcing @xmath377 by the forcing @xmath378 .
it turns out that the proofs in @xcite go through essentially unaltered , because they only make appeal to rather general properties ( chain condition , distributivity , directed closure ) of cohen forcing .
there is one slightly subtle point , which is that is we will need a certain uniformity in the dependence of @xmath272 on @xmath178 , as expressed by the equation @xmath379 .
the real point of this is that when we consider a condition in @xmath380 which puts onto the generic club a point @xmath372 followed by the point @xmath29 , the forcing which is happening between @xmath93 and @xmath29 is @xmath381 ; it will be crucial that @xmath382 , that is the same forcing that would be used in @xmath123 between @xmath93 and @xmath29 .
we conclude this section with a short discussion of the necessary changes in the proofs .
we assume that the reader has a copy of @xcite to hand ; all references below are to theorem and section numbers from that paper .
section three : we will use @xmath371 to construct the master sequence @xmath383 . recall that we chose @xmath316 to witness @xmath384-supercompactness of @xmath195 , it can be checked that this is enough to make all the arguments below go through .
when we build @xmath383 we will choose each @xmath385 so that @xmath386 contains only @xmath339 with @xmath375 sweet , and then let @xmath387 . as in @xcite , if @xmath388 is the ultrapower by @xmath389 then @xmath390 .
if we let @xmath6 be the function given by @xmath391 then @xmath392_{m_{\alpha}}$ ] is @xmath216-closed in @xmath393 , hence is @xmath245-closed in @xmath71 .
this closure will suffice to make appropriate versions of 3.3 , 3.4 and 3.5 go through .
section four : 4.1 is just as in @xcite .
a version of 4.2 goes through because @xmath394 .
section five : in the definition of `` suitable quintuple '' we demand that @xmath395 and @xmath396 .
the definition of `` addability '' is unchanged .
clauses d ) and e ) of that definition still make sense because ( e.g. ) @xmath397 is a complete suborder of @xmath398 .
versions of 5.1 and 5.2 go through without change . in the definition of `` condition in @xmath60 ''
we demand that @xmath399 , @xmath400 .
the ordering on @xmath60 is as before .
the definitions of `` canonical representatives '' and of `` upper and lower parts '' are unchanged , as is 5.3 .
we can do a `` local '' version of our argument for 3.3 above to show 5.4 goes through .
similarly 5.5 , 5.6 and 5.8 go through .
section six : we need to make the obvious changes in the definitions of `` suitable quintuple '' and of the projected forcing @xmath123 .
the proofs all go through because the forcing @xmath272 has the right chain condition and closure .
section seven : the master condition argument of this section goes through because the forcing notion @xmath401 is @xmath216-directed closed .
putting together the results of the preceding sections we get the following result .
99 u. abraham , m. rubin and s. shelah , on the consistency of some partition theorems for continuous colorings and the structure of @xmath403-dense real order types , annals of pure and applied logic 29 ( 1985 ) , 123206 . | we show that it is consistent with zfc ( relative to large cardinals ) that every infinite boolean algebra @xmath0 has an irredundant subset @xmath1 such that @xmath2 .
this implies in particular that @xmath0 has @xmath3 subalgebras .
we also discuss some more general problems about subalgebras and free subsets of an algebra .
the result on the number of subalgebras in a boolean algebra solves a question of monk from @xcite .
the paper is intended to be accessible as far as possible to a general audience , in particular we have confined the more technical material to a `` black box '' at the end .
the proof involves a variation on foreman and woodin s model in which gch fails everywhere .
pre amssym.def at=
= 11 # 1#1 # 1#2#3#4#5@ # 2=ne@@ # 2=@@@ # 1="#3@#4#5 @#1#2#3 @#1#1 012345678 9abcdef = msam10 scaled 1 = msam7 scaled 1 = msam5 scaled 1 = = = @@
@"0@39 = msbm10 scaled 1 = msbm7 scaled 1 = msbm5 scaled 1 = = = @@ # 1#1 # 1@h@xmath4 @>@ em``0@5b#1 ' ' 0362#1 # 1@h@xmath4 @>@ em``0@5d#1 ' ' 0365#1 = eufm10 = eufm7 = eufm5 = = = # 1#1 = pre amssym.def at pre amssym.tex atpre amssym.tex at= = 11 1200 1201 1202 1003 1004 1205 1006 1007 1308 1309 130a 130b 120c 130d 130e 130f 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 131a 131b 131c 131d 131e 131f 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 132a 132b 132c 132d 132e 132f 1330 1331 1332 1333 1334 1335 1336 1337 1038 133a 133b 133c 133d 133e 133f 1340 1341 1342 1343 1344 1345 1046 1347 1048
1349 134a 134d 104e 104f 1350 1351 1352 1353 1354 1356 1357 1259 125a 125b 105c 105d 105e 135f 1360 1361 1362 1363 1264 1265 1266 1267 1268 1269 136a 136b 136c 136d 136e 136f 1073 1374 1275 1376 1377 107b 127c 127d 127e 127f 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 230a 230b 230c 230d 230e 230f 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 231a 231b 231c 231d 231e 231f 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 232a 232b 232c 232d 232e 232f 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 233a 233b 233c 233d 223e 203f 2040
2060 2061 2066 2067 2368 2069 206a 206b 236c 236d 226e 226f 2370 2371 2272 2373 2374 2375 2376 2377 2378 2379 207a 207b 207c 207d 207e 237f = pre amssym.tex at = 11 begintheorem#1#2 opargbegintheorem#1#2#3 = 12 [ claim ] = 10000 = 10000 |
at the tevatron , @xmath1 and @xmath0 bosons are predominantly produced through quark - antiquark annihilation within the colliding proton and anti - proton beams .
the bosons decay into two fermions which , due to their large mass , typically have a high momentum in the direction transverse to the beam .
the @xmath1 and @xmath0 boson decay channels , where the charged lepton is an electron or muon , have clean experimental signatures with low background contamination and can therefore be utilized for @xmath1 and @xmath0 boson precision measurements .
events with a @xmath0 boson interference are indistinguishable from pure z exchange . for the remainder of this document we use @xmath0 boson to represent z/@xmath2 , including interference . ]
are identified by the detection of two charged leptons , each with high transverse momentum .
the electron energy is measured with the electromagnetic calorimeter and the muon momentum with the cdf central drift chamber .
@xmath1 boson events are identified by one high transverse momentum charged lepton and one high transverse momentum neutrino .
since the neutrino does not interact with the detector , its existence is inferred by the transverse momentum imbalance in the detector .
we describe four recent measurements from cdf .
two that constrain the parton distribution function of the proton , the @xmath1 boson charge asymmetry and the @xmath0 boson rapidity measurements and two @xmath1 boson precision measurements of the @xmath1 boson mass and width .
on average the @xmath3(@xmath4 ) quark carries a higher fraction of the ( anti)proton s momentum than the @xmath5(@xmath6 ) quark , thus a @xmath7(@xmath8 ) produced via a @xmath9(@xmath10 ) annihilation will tend to be boosted along the ( anti)proton beam .
this results in a non - zero forward - backward @xmath1 boson asymmetry , defined as : @xmath11 where @xmath12 is the @xmath1 boson rapidity and @xmath13 is the differential cross section for @xmath7 or @xmath8 boson production .
a measurement of @xmath14 is sensitive to the ratio of @xmath3 and @xmath5 quark components of the protons parton distribution functions ( pdf ) .
since the longitudinal neutrino momentum is not known , a measurement of the electron or muon charge asymmetry , rather than @xmath14 , has traditionally been made .
this distribution is a convolution of @xmath14 and the @xmath15 asymmetry from @xmath1 boson decays and the two asymmetries tend to cancel at large pseudorapidities . in this analysis @xcite , using 1 fb@xmath16 of @xmath17 data , @xmath14 is extracted by constraining @xmath18 to its measured value , giving two possible solutions , each receiving a probability weight according to the decay structure and @xmath19 . the measured @xmath1 boson charge asymmetry is compared to the prediction from pdfs in figure [ wcharge ] using nlo cteq6.1 m @xcite ( left ) and nnlo mrst 2006 @xcite ( right ) .
the respective pdf uncertainties are shown with shaded bands .
given the small experimental uncertainties this measurement will help to constrain pdfs in future fits .
boson charge asymmetry and theoretical predictions from nlocteq6.1 ( left ) and nnlo mrst 2006 ( right ) , with their associated pdf uncertainties . ]
boson charge asymmetry and theoretical predictions from nlocteq6.1 ( left ) and nnlo mrst 2006 ( right ) , with their associated pdf uncertainties . ]
@xmath0 boson production at the tevatron proceeds predominantly through annihilation of a @xmath3 ( or @xmath5 ) quark in the proton and a @xmath4 ( or @xmath6 ) quark in the anti - proton .
the two partons carry momentum fractions @xmath20 and @xmath21 . at leading order ( lo ) ,
the momentum fraction @xmath20 ( @xmath21 ) carried by the two partons are related to the rapidity ( @xmath22 ) of the @xmath0 boson via the relation : @xmath23 where @xmath24 is the center of mass energy and @xmath25 is the mass of the @xmath0 bosons .
thus a measurement of @xmath26 places constraints on the proton pdfs .
in particular , the high rapidity region probes both the high and low @xmath27 regions of the pdfs .
cdf has made a measurement of @xmath26 using 2.1 fb@xmath16 of @xmath28 data with @xmath29@xmath30@xmath29 @xmath31 2.8 .
this measurement has recently been included in fits by the mstw collaboration @xcite and boson rapidity distribution @xcite . ]
boson rapidity distribution @xcite . ]
figure [ zrapidity ] shows the results of the nnlo fit .
the data are well fitted , particularly when the systematic uncertainties are taken into account .
there is evidence of a slight excess of data over theory at high rapidity .
the @xmath1 boson mass receives self - energy corrections due to vacuum fluctuations involving virtual particles .
thus the @xmath1 boson mass probes the particle spectrum in nature , including particles that have yet to be observed directly . in order to constrain the higgs boson mass
, we need to account for the radiative corrections to @xmath18 due to the dominant top - bottom quark loop diagrams . for fixed values of other inputs , the current uncertainty on the top quark mass , @[email protected] gev/@xmath34 @xcite corresponds to an uncertainty in its @xmath1 boson mass correction of 8 mev/@xmath34 .
the current world average from direct measurements at lep @xcite and tevatron @xcite yield a world average boson mass measurement @xmath35 @xmath36 80401@xmath3343 mev/@xmath34 @xcite .
the gfitter collaboration subsequently computed a new world average of 80399@xmath3323 mev/@xmath34 @xcite .
] of 80398@xmath3325 mev/@xmath34 .
it is clearly profitable to reduce the @xmath1 boson mass uncertainty further as a means of constraining the higgs boson mass .
the current one - sided 95% c.l .
upper limit from indirect electroweak constraints is @xmath37@xmath31163 gev / c@xmath38 @xcite . at the tevatron ,
the @xmath1 boson mass is extracted from a template fit to the jacobian edge of the transverse mass distribution , defined as @xmath39)}$ ] , where @xmath40 is the transverse momentum of the charged lepton and @xmath41 is the inferred neutrino transverse momentum .
a fast monte carlo simulation is used to model the lineshape of the template distributions , accounting for the detector response , resolution and acceptance effects .
we constrain these important detector and physics effects by control samples and calculations which determine the precision of the measurement of @xmath35 .
examples of the effects that need to be rigorously studied , extracted and modeled , include the tracker momentum scale and resolution , the calorimeter energy response , scale and resolution , the intrinsic @xmath1 boson transverse momentum , internal qed radiation and the proton parton distribution functions ( pdfs ) . using 200 pb@xmath16 of data , cdf combines the measurement in the muon and electron decay channel and obtains @xmath35=80413@xmath3334(stat)@xmath3334(syst ) mev/@xmath34 @xcite @xcite .
figure [ wmass ] on the left shows the transverse mass fit for the muon decay channel .
right : invariant mass fit in the di - muon decay channel using 2.3 fb@xmath42 . ] measurement .
right : invariant mass fit in the di - muon decay channel using 2.3 fb@xmath42 . ] many of the systematic uncertainties in this measurement scale with the statistics of the control samples used to calibrate the detector and can therefore be improved with an analysis of the larger datasets in hand .
additionally , improvements in the production and decay model such as improved pdf fits are likely to further reduce the overall systematic uncertainty on future measurements of @xmath35 .
cdf has begun analyzing an over ten times larger data sample with an integrated luminosity of 2.3 fb@xmath16 , as can be seen in the @xmath0 boson mass fit on the right side of figure [ wmass ] .
studies in progress confirm that many of the systematic uncertainties scale with luminosity as expected and we look forward to an updated @xmath35 measurement with a precision better than the current world average of 25 mev / c@xmath38 .
the @xmath1 boson width is known within the standard model to an extremely high precision of 0.1% , thus an accurate experimental measurement is desirable to test this prediction .
the measurement technique is very similar to the extraction of the @xmath1 boson mass , but instead of fitting the jacobian edge , the fit is performed in the tail region 90 @xmath31 @xmath43 @xmath31 200 gev / c@xmath38 of the transverse mass distribution , where the gaussian detector resolution falls off more quickly compared to the @xmath1 boson breit - wigner lineshape .
cdf uses 350 @xmath44 analyzing the electron and muon decay channels to measure the @xmath1 boson width @xcite .
figure [ wwidth ] shows the transverse mass fits for @xmath17 events .
right : comparison with other tevatron measurements and lep-2 . ]
right : comparison with other tevatron measurements and lep-2 . ]
the results of the muon and electron decay channel are combined to give the final result @xmath45 = 2032 @xmath33 73 mev / c@xmath38 , the world s most precise measurement , which is in good agreement with the standard model prediction of 2091 @xmath33 2 mev / c@xmath38 @xcite .
the large tevatron datasets of @xmath1 and @xmath0 bosons with their clean decays to leptonic final states allow for a rich electroweak physics program .
the cdf @xmath1 boson charge asymmetry and @xmath0 boson rapidity measurements improve the constraints on proton pdfs , which in turn are important inputs to precision measurements of the @xmath1 boson mass and @xmath1 boson width .
the anticipated more precise determination of the @xmath1 boson mass from cdf is essential for a further constraint of the higgs boson mass and will test the standard model with higher precision .
i would like to thank my cdf collaborators , especially those whose work went into the results presented here at the beautiful lake louise winter institute . | we present results from electroweak precision measurements at cdf .
these include the measurement of the @xmath0 boson rapidity distribution using di - electron events , the @xmath1 boson charge asymmetry measurement in the electron - neutrino decay and the direct measurement of the @xmath1 boson mass and width in both the electron - neutrino and muon - neutrino decay channels . |
in the past decade quadcopters have been studied due to their relative simple fabrication in comparison to other aerial vehicles , which turns them into ideal platforms for modeling , simulation and implementation of control algorithms .
the fact that they are unmanned vehicles naturally invites developers to explore tasks that require a high degree of autonomy .
+ past works such as @xcite have set the base for developing quadcopter platforms from their construction to the automation techniques necessary to control the highly non - linear dynamics that characterize these vehicles .
+ the scope of the quadcopter technology has changed over the years .
the cost and sizes have been reduced , it is now a platform affordable for a broad type of public , from researchers to hobbyists .
but beyond the economic revenue these vehicles generate , the manufacturers are searching for more autonomy , longer flight time , high data processing capabilities and adaptation to changing environments , hence the active research on quadcopters . + a fairly new type of quadcopters are the so called `` nanoquads '' that are of considerably low size and weight , making them an ideal platform for indoor usage .
the project proposed here considers the study of a commercial platform of a nanoquadcopter called `` crazyflie 2.0 '' developed by bitcraze company @xcite .
weighting only 27 grams and having 9.2 cm of length and width , this nanoquad has rapidly become one of the preferred platforms for quadcopter research .
+ for indoor control of quadcopters different localization techniques can be employed , for example the vicon motion capture system @xcite is one of the preferred systems for precise localization and it has been used widely in recent quadcopter studies @xcite .
a recent low - cost technology based on ultra - wide band radio modules has proven effective for indoor localization in robotics systems and specially in quacopters @xcite .
its low costs are inviting developers to create their own implementations and the system is getting more precise and robust . in a few words ,
the system measures the distance between two ultra - wide band modules , normally called anchor and tag , by measuring the time of flight of an electromagnetic wave .
thus , by the simple relationship between time , distance and velocity ( in this case , the speed of light ) , then the distance can be easily determined . by having at least three anchors constantly calculating the distance between them and a certain tag , a triangulation allows to calculate the position in space of the tag , knowing beforehand the fixed position of each anchor with respect to a frame .
+ the uwb system can be implemented using a two - way ranging protocol or a one - way ranging protocol . in two - way ranging , the tag communicates with each anchor individually following a sequence to go through all the anchors and calculate each distance . on the other hand , in one - way ranging the tag constantly broadcasts messages that are received by every anchor and by precisely synchronising the clocks of the anchors then the distance between each of them and the tag are calculated .
one - way ranging is particular useful for multi - robot localization applications as there exists no bottle - neck in the number of tags the system can support . in particular , for this project the two - way ranging system developed in @xcite was used to test the control loop behavior using different localization techniques .
this system was developed using the decawave dmw1000 ultra - wide band module @xcite which offers an accuracy of 10 - 20 centimeters in distance measurements .
the main objectives of the research project were : 1 . develop the mathematical model that describes the dynamics of the crazyflie 2.0 quadcopter .
2 . create a simulation environment for testing position and trajectory tracking control algorithms .
implement , test and compare different control architectures .
4 . evaluate the performance of a low - cost uwb - based localization system when integrated in the control loop .
a set of small milestones were defined to help achieve the main objectives of the project : 1 .
investigate past works to identify the physical and aerodynamical parameters of the crazyflie 2.0 . 2 .
linearize the quadcopter s dynamics around hover state .
3 . study and identify the control architecture inside the crazyflie s firmware .
4 . design , simulate and implement an off - board position controller using data from the vicon positioning system .
5 . conceive a second control system , from simulation to implementation , to track more demanding trajectories .
compare the performance of both controllers with in - flight data .
compare the performance of the lqt controller using both the vicon and the uwb systems .
in this section a mathematical model of the crazyflie 2.0 is proposed .
this study was the basis on which the simulation environment was built and an important component in the design of controllers .
thus , it was important to dedicate enough time to understand how the system works and identify correctly some physical parameters that were relevant for the simulation to be useful in the real case scenario .
before any dynamic study of the quadcopter begins , it is necessary to define the coordinate frames of the body of the quadcopter ( non - inertial frame ) as well as the inertial frame , also called `` world frame '' , which in the case of this project refers to the coordinate frame set by the external positioning system ( vicon / uwb ) . following the conventions set by the `` bitcraze '' company when designing their quadcopter , as seen in the body - fixed frame
is defined . body - fixed frame and inertial frame . ] in the aeronautic systems , a popular axes convention is to define a positive altitude downwards , the y axis pointing towards the east and the x axis pointing towards the true north .
these types of frames are called ned frames ( north , east , down ) .
it was decided to follow the convention used in the crazyflie 2.0 firmware , meaning a positive altitude upwards , which defines an enu frame ( east , north , up ) .
another remark is that the origin of the body - fixed frame matches with the center of gravity of the quadcopter .
+ another important remark is knowing the flight configuration of the quadcopter as there are two of them : configuration `` + '' or configuration `` x '' .
the difference between them is the orientation of the x - y frame in terms of the arms of the quadcopter , as shown in taken from the manufacturer s website @xcite and modified accordingly .
`` + '' configuration at the left and `` x '' configuration at the right . ] in the modern conceptions of quadcopters the `` x '' configuration is prefered over the `` + '' configuration , mainly because in `` x '' it is easier to add a camera functionality as the quadcopter s arms will not be interfering with the images captured . by default
the crazyflie 2.0 is in x mode , so for the rest of this project and during the mathematical modeling it will be considered that the quadcopter is in this configuration .
the dynamic equations of the quadcopter proposed here take into account certain physical properties that are not necessarily perfectly valid in the real platform that is being used in this work , but they are good approximations that simplify greatly the study and comprehension of this type of vehicles . here
are the hypothesis : 1 .
the quadcopter is a rigid body that can not be deformed , thus it is possible to use the well - known dynamic equations of a rigid body .
the quadcopter is symmetrical in its geometry , mass and propulsion system .
the mass is constant ( i.e its derivative is 0 )
. the mechanical classic laws of motion are valid in inertial systems , so to be able to translate these equations into the body frame it is necessary to define a rigid transformation matrix from the inertial frame to the body - fixed frame , in which only the rotational part is meaningful to the discussion and is given by three successive rotations : first a rotation of an angle @xmath0 around the @xmath1 axis , then a rotation of an angle @xmath2 around the intermediate @xmath3 axis and finally a rotation of an angle @xmath4 around the intermediate @xmath5 axis .
once these three rotations are calculated , the resulting transformation matrix is defined as : @xmath6$}\label{eq : euler}\ ] ] where @xmath4 , @xmath2 and @xmath0 represent the roll , pitch and yaw angles of the quadcopter s body .
shows the direction of said angles in the crazyflie 2.0 body - fixed frame defined previously .
euler angles in the quadcopter s body . ]
the notation convention used during the mathematical analysis of the quadcopter s dynamics is exhibited in , where the state variables are defined .
.[tab : notation]notation for vectors and states . [ cols="^,^,^",options="header " , ] the experimental data confirms the robustness of the control system while using a positioning system with around 100 times greater standard deviation noise than the initial system .
the tuning of the kalman filter to adapt to this new source of noise was vital to ensure a good compromise between filtering and precision in the estimations . even though there is a clear advantage on using more precise technology , such as the vicon , the system could still track the desired trajectories with an acceptable degree of precision while using a more cheaper system such as the uwb .
+ the fact that the position integral gains for the x and y positions had to be lowered while using the uwb system , was a necessary compromise to ensure less oscillations while following the desired trajectory .
such a compromise did not arise while using the vicon system , with which the quadcopter remained stable and followed smoothly the trajectories for a wide range of gain values .
+ if the control system could only use the uwb system , then a more detailed study of how to compensate the different sources of noise and biases should be made .
for instance , the uwb system looses precision when the tag is close to one of the anchors , or if the tag is facing away from one of the anchors .
all this subtleties , if taken in account while designing the control system , could lead to a better performance than the one obtained and presented in this work .
+ finally , the video found in @xcite shows a summary of the project s simulation and experimental results .
the study was set out to explore the dynamics of an open source nanoquadcopter named crazyflie 2.0 , as well as creating a simulation environment for control design and then testing it in the real platform .
this type of unmanned aerial vehicles is becoming the preferred platform for testing control algorithms of diverse natures , thus the inherent importance of conceiving a mathematical model of the vehicle that can predict , up to some extent , how the system will evolve over time .
hence , the project started by a modeling of the nanoquadcopter and an identification of certain physical parameters , based in previous work .
working in parallel with the literature and the quadcopter s embedded firmware was the main key in describing the system behavior just as it is in the real platform , an important milestone for future work as the dynamics of a system is the heart of every simulation environment .
+ the second phase of the project was building the simulation that served as the first test - bench of the control architectures proposed . using both the non - linear dynamics and the linearised state space realisation of the system ,
the simulation created is a solid testing environment to conceive all types of control systems .
it was incredibly useful during the first stages of the project to get a better understanding of how the system worked .
in addition , the simulation was used for designing both the pid position controller and the lqt trajectory tracker .
+ an important conclusion is that the initial belief that all dynamics were decoupled as suggested during the linear modeling was not entirely true in the non - linear system .
as observed in the simulations , there exists some interference between movements that , for instance , does not allow the quadcopter to describe a perfectly straight line trajectory when there are more than one movement involved ( a yaw rotation for example ) . + the position pid tracker was tested for time - varying trajectories , such as circles and helices .
even though the system could described these trajectories , there were some drawbacks and performance issues , for example not getting fast enough to the desired points which lead to errors in the desired trajectory .
the fact that the task at hand was managed by a position tracker and not a trajectory tracker was the main reason of these discrepancies . to address the deficiencies of the pid controller ,
a new control system was conceived using the lqt algorithm , which proved to have interesting characteristics while following step responses , mainly that it started moving before the command was asked in order to reduce the tracking error .
the feature was possible thanks to the off - line calculation of the algorithm and the knowledge of the trajectory beforehand .
+ the comparisons between the pid and the lqt controller indicate a clear superiority of the lqt in terms of reducing the trajectory tracking error , specially in the more demanding trajectories , in which the lqt algorithm reduced up to 4 times the rms errors obtained with the pid controller .
directly related to the better tracking , the lqt incurred in higher levels of control effort than the pid , but it also eliminated the great command peaks seen in the motor time plots of the pid , thus getting rid of the undesired motor saturations that could lead to unstable states .
+ there are two main drawbacks of the lqt algorithm with respect to the pid : the first one is the inability to specify trajectories for the heading ( yaw angle ) and the second one is the need to know the trajectory before its execution .
taking in account these shortcomings , it is proposed as future work for this research to incorporate a method to control the yaw angle while keeping the good performance in the lqt algorithm , the author proposes a gain - scheduling method being the yaw angle the scheduling variable as a possible solution for this problem . as for the second drawback of the lqt algorithm , more research
should be directed towards an on - line implementation thus making the controller useful in more complex tasks such as planning and execution missions in real - time .
+ the gui created for trajectory generation proved to be a valuable asset to quickly test different types of trajectories , with varying difficulty .
but the tool can be improved by adding physical constraints to the trajectory generation , as to assure the trajectory is feasible for the quadcopter to follow .
future work in this area should explore feasible trajectory generation as proposed in works such as @xcite .
+ the simulation versus experimental comparative time plots show that the simulation environment developed in this project was accurate to some extent , serving its purpose as a useful design tool for the controllers synthesized , but it had its limitations mainly due to unmodeled phenomena , which lead to the need of introducing high integral gains in the controllers to compensate the model errors and other perturbations of the system . as future work
, it is suggested a more thorough model identification for the quadcopter , for example using numerical methods such as the closed - loop `` black box '' identification proposed in @xcite . + the kalman filter approach for estimating the linear velocities from the position data proved to be successful using both the vicon and the uwb , specially with the latter in which the data had 100 times greater standard deviation noise .
the vicon versus uwb experiments suggest that in both cases the lqt tracked the desired position , but with obvious different levels of smoothness and precision .
even though both performances were satisfactory in terms of the scope of this work , future research into improving the control system while using the uwb position system would be ideal .
starting from an identification of different sources of added noise and biases of the uwb system , upto different filtering techniques that are more appropriate than the classic kalman filter proposed in this project are the author s recommendations to improve the control system performance .
+ this work represents a solid base for future research using this platform , with enough explanation in the calculus for newcomers in the area to understand the basic functioning of the system .
the simulation environment was developed in a fashion that corresponds exactly with the equations shown in the mathematical model , which helps in the quick understanding of how everything works and saves time in comprehending an otherwise complex system , plus it is easily customizable for future users to develop their own controllers .
the project successfully fulfilled its ultimate goal of characterizing the provided quadcopter platform and doing all the steps needed to develop an efficient control system for trajectory tracking .
10 hanna , w. ( 2014 ) . _ modelling and control of an unmanned aerial vehicle _
( b.eng thesis , charles darwin university ) .
subramanian , g. p. ( 2015 ) .
_ nonlinear control strategies for quadrotors and cubesats _
thesis , university of illinois at urbana - champaign ) .
greitzer , e. m. , spakovszky , z. s. , & waitz , i. a. ( 2006 ) .
_ thermodynamics and propulsion . _ mechanical engineering , mit .
corke , p. ( 2011 ) .
_ robotics , vision and control : fundamental algorithms in matlab ( vol .
_ springer .
hartman , d. , landis , k. , mehrer , m. , moreno , s. , & kim , j.(2014 ) _ quadcopter dynamic modeling and simulation ( quadsim ) v1.00 _ ( senior design project , drexel university ) hoenig , w. , milanes , c. , scaria , l. , phan , t. , bolas , m. , & ayanian , n. ( 2015 ) .
_ mixed reality for robotics_. in intelligent robots and systems ( iros ) , 2015 ieee / rsj international conference on ( pp .
5382 - 5387 ) .
ieee elruby , a. y. , el - khatib , m. m. , el - amary , n. h. , & hashad , a. i. ( 2012 ) .
_ dynamic modeling and control of quadrotor vehicle_. in fifteenth international conference on applied mechanics and mechanical engineering , amme ( vol .
karwoski , k. ( 2011 ) .
_ quadrocopter control design and flight operation_. ( internship final report , nasa usrp ) sonnevend , i. ( 2010 ) . _
analysis and model based control of a quadrotor helicopter . _
( bsc diploma work , pter pzmny catholic university , faculty of information technology , budapest , hungary ( supervisor : g. szederknyi ) ) habib , m. k. , abdelaal , w. g. a. , & saad , m. s. ( 2014 ) .
_ dynamic modeling and control of a quadrotor using linear and nonlinear approaches_. ( m.s .
thesis , the american university in cairo ) .
landry , b. ( 2015 ) . _ planning and control for quadrotor flight through cluttered environments _
( master s degree thesis , massachusetts institute of technology ) .
dunkley , o. , engel , j. , sturm , j. , & cremers , d. ( 2014 ) .
_ visual - inertial navigation for a camera - equipped 25 g nano - quadrotor_. in iros2014 aerial open source robotics workshop .
xu , d. , wang , l. , li , g. , & guo , l. ( 2012 , august ) .
_ modeling and trajectory tracking control of a quad - rotor uav . _ in proceedings of the 2012 international conference on computer application and system modeling .
atlantis press .
meyer , j. , sendobry , a. , kohlbrecher , s. , klingauf , u. , & von stryk , o. ( 2012 ) .
_ comprehensive simulation of quadrotor uavs using ros and gazebo .
_ in simulation , modeling , and programming for autonomous robots ( pp .
400 - 411 ) .
springer berlin heidelberg .
suimez , e. c. ( 2014 ) .
trajectory tracking of a quadrotor unmanned aerial vehicle ( uav ) via attitude and position control ( master s degree thesis , middle east technical university ) .
oh , s. m. ( 2012 ) .
_ modeling and control of a quad - rotor helicopter_. ( m.s .
thesis , university of florida ) pounds , p. e. i. ( 2007 ) .
_ design , construction and control of a large quadrotor micro air vehicle . _
( doctoral dissertation , australian national university . ) tamami , n. , pitowarno , e. , & astawa , i. g. p. ( 2014 ) . _ proportional derivative active force control for x configuration quadcopter . _
journal of mechatronics , electrical power , and vehicular technology , 5(2 ) , 67 - 74 .
roo , m. ( 2015 ) .
_ optimal event handling by multiple uavs . _
( m.s . report , university of twente ) lehnert , c. , & corke , p. ( 2013 ) .
_ av - design and implementation of an open source micro quadrotor_. ac on robotics and automation , eds .
sabatino , f.(2015 ) .
_ quadrotor control : modeling , nonlinear control design , and simulation . _
( master s degree project , kth royal institute of technology ) .
kader , s. a. , el - henawy , a. e. , & oda , a. n. ( 2014 ) .
_ quadcopter system modeling and autopilot synthesis . _ in international journal of engineering research and technology ( vol .
3 , no . 11 ( november-2014 ) ) .
esrsa publications .
naidu , d. s. ( 2002)._optimal control systems . _ crc press .
mathworks(2015)._state estimation using time - varying kalman filter .
_ retrieved may 16 , 2016 , from http://www.mathworks.com/help/control/getstart/estimating-states-of-time-varying-systems-using-kalman-filters.html hoffmann , g. m. , waslander , s. l. , & tomlin , c. j. ( 2008)._quadrotor helicopter trajectory tracking control .
_ in aiaa guidance , navigation and control conference and exhibit ( pp .
1 - 14 ) .
mueller , m. w. , & dandrea , r. ( 2013)._a model predictive controller for quadrocopter state interception .
_ in european control conference ( pp
. 1383 - 1389 ) .
mu , s. , zeng , y. , & wu , p. ( 2008)._multivariable control of anaerobic reactor by using external recirculation and bypass ratio .
_ journal of chemical technology and biotechnology , 83(6 ) , 892 - 903 .
huang , h. , hoffmann , g. m. , waslander , s. l. , & tomlin , c. j. ( 2009)._aerodynamics and control of autonomous quadrotor helicopters in aggressive maneuvering .
_ in robotics and automation , 2009 .
ieee international conference on ( pp .
3277 - 3282 ) .
sujit , p. b. , saripalli , s. , & sousa , j. b. ( 2014)._unmanned aerial vehicle path following : a survey and analysis of algorithms for fixed - wing unmanned aerial vehicles_. ieee control systems , 34(1 ) , 42 - 59 .
bouabdallah , s. , noth , a. , & siegwart , r. ( 2004)._pid vs lq control techniques applied to an indoor micro quadrotor_. in intelligent robots and systems , 2004.(iros 2004 ) . proceedings .
2004 ieee / rsj international conference on ( vol .
2451 - 2456 ) .
peraire , j. , & widnall , s. ( 2009 ) _ lecture l28 - 3d rigid body dynamics_. mit opencourseware ,
dynamics fall 2009 .
available online : http://ocw.mit.edu .
bitcraze(2015 ) .
_ crazyflie 2.0 assembly instructions_. retrieved august 3 , 2016 , from https://wiki.bitcraze.io/projects:crazyflie2:userguide:assembly bitcraze(2015 ) .
_ crazyflie 2.0 main page_. retrieved august 5 , 2016 . from https://www.bitcraze.io/crazyflie-2/ vicon motion systems ( 2015 ) .
_ vicon motion capture system main page_. retrieved august 5 , from https://www.vicon.com/ mueller , m. w. , hamer , m. , & dandrea , r. ( 2015 ) .
_ fusing ultra - wideband range measurements with accelerometers and rate gyroscopes for quadrocopter state estimation_. in 2015 ieee international conference on robotics and automation ( icra ) ( pp .
1730 - 1736 ) .
rafrafi , w. , & le ny , j. ( 2016 ) .
_ intgration dun systme radio bande ultra - large pour la navigation de robots mobiles_. ( master s degree thesis , cole polytechnique de montral ) .
decawave(2015 ) .
_ scensor dwm1000 module product page_. retrieved august 11 , from http://www.decawave.com/products/dwm1000-module luis , c. ( 2016 ) .
_ trajectory tracking of a crazyflie 2.0 nanoquadcopter _ [ video file ] . retrieved august 13 , from https://youtu.be/c-sxovcyhjq
the firmware used during this project was `` release 2016.02 '' found in https://github.com/bitcraze/crazyflie-release/releases , with the following changes : .... # ifdef quad_formation_x int16_t r = control->roll / 2.0f ; int16_t p = control->pitch / 2.0f ; motorpower.m1 = limitthrust(control->thrust - r - p - control->yaw ) ; motorpower.m2 = limitthrust(control->thrust - r + p + control->yaw ) ; motorpower.m3 = limitthrust(control->thrust + r + p - control->yaw ) ; motorpower.m4 = limitthrust(control->thrust + r - p + control->yaw ) ; .... .... attitudecontrollercorrectattitudepid(state->attitude.roll , -state->attitude.pitch , state->attitude.yaw , setpoint->attitude.roll , setpoint->attitude.pitch , attitudedesired.yaw , & ratedesired.roll , & ratedesired.pitch , & ratedesired.yaw ) ; //bypass attitude controller if rate mode active if ( setpoint->mode.roll = = modevelocity ) { ratedesired.roll = setpoint->attituderate.roll ; } if ( setpoint->mode.pitch = = modevelocity ) { ratedesired.pitch = setpoint->attituderate.pitch ; } if ( setpoint->mode.yaw = = modevelocity ) { ratedesired.yaw = setpoint->attituderate.yaw ; } | the primary purpose of this study is to investigate the system modeling of a nanoquadcopter as well as designing position and trajectory control algorithms , with the ultimate goal of testing the system both in simulation and on a real platform .
the open source nanoquadcopter platform named crazyflie 2.0 was chosen for the project . the first phase consisted in the development of a mathematical model that describes the dynamics of the quadcopter .
secondly , a simulation environment was created to design two different control architectures : cascaded pid position tracker and lqt trajectory tracker . finally , the implementation phase consisted in testing the controllers on the chosen platform and comparing their performance in trajectory tracking .
our simulations agreed with the experimental results , and further refinement of the model is proposed as future work through closed - loop model identification techniques .
the results show that the lqt controller performed better at tracking trajectories , with rms errors in position up to four times smaller than those obtained with the pid .
lqt control effort was greater , but eliminated the high control peaks that induced motor saturation in the pid controller .
the lqt controller was also tested using an ultra - wide band two - way ranging system , and comparisons with the more precise vicon system indicate that the controller could track a trajectory in both cases despise the difference in noise levels between the two systems .
cole polytechnique de montral electrical engineering department automation section + + + + * * + + + + luis , c. , & le ny , j. ( august , 2016 ) .
_ design of a trajectory tracking controller for a nanoquadcopter_. technical report , mobile robotics and autonomous systems laboratory , polytechnique montreal .
+ + + author : supervisor : carlos luis jrme le ny |
the aim of this paper is to discuss the statistical physics of turbulent advection of passive scalars @xcite .
we are interested in scalar fields @xmath0 which are advected by a velocity field @xmath1 such that together they solve the set of equations @xmath2 in these equations @xmath3 , @xmath4 and @xmath5 are the pressure field , the kinematic viscosity and the scalar diffusivity respectively . in this paper
the navier - stokes equations will be always forced with a time - stationary forcing @xmath6 .
the scalar field will be forced or unforced ( decaying ) , depending on our interests below .
we focus on the case of high reynolds number re and high peclet number pe , i.e. @xmath7 , where the turbulence of the velocity field is fully developed and where both fields display a significant range of scaling behavior at scales that are sufficiently far from the forcing scales . a major theoretical question that had been answered recently has to do with the scaling properties of the correlation functions of the advected field @xcite .
define the simultaneous many - point correlation function @xmath8 in the forced case by @xmath9 with @xmath10 denoting an average with respect to realizations of the stationary forcing @xmath11 and of the velocity field @xmath12 .
it had been known for a long time that for high re and pe these functions are homogeneous functions of their arguments , i.e. @xmath13 with @xmath14 being a scaling exponent that in general is _ anomalous _ ,
i.e. can not be guessed from dimensional considerations .
but only recently it became clear how these exponents are determined by the dynamical processes .
to understand the progress made , we rewrite the dynamical equation for @xmath15 in the shorthand form @xmath16 where in the present case @xmath17 . in recent work @xcite it was clarified why and how passive fields exhibit anomalous scaling , when the velocity field is a generic turbulent field .
the key is to consider a _ decaying problem _ associated with eq .
( [ defl ] ) , in which the forcing @xmath11 is put to zero .
the problem becomes then a linear initial value problem , @xmath18 with a formal solution @xmath19 with the operator @xmath20\big|_{{{\bm{r } } } , { { \bm{r}}}'}\ , \label{defr}\ ] ] and @xmath21 being the time ordering operator .
define next the _ time dependent _ correlation functions of the decaying problem : @xmath22 here pointed brackets without subscript @xmath23 refer to the decaying object in which averaging is taken with respect to realizations of the velocity field and initial conditions . as a result of eq .
( [ oper ] ) the decaying correlation functions are evolved by a propagator @xmath24 , ( with @xmath25 and @xmath26 ) : @xmath27 in writing this equation we made explicit use of the fact that the _ initial _ distribution of the passive field @xmath28 is statistically independent of the advecting velocity field .
thus the operator @xmath24 can be written explicitly @xmath29 the key finding @xcite is that the operator @xmath24 possesses a _ left _
eigenfunction of eigenvalue 1 , i. e. there exists ( for each @xmath30 ) a time - independent function @xmath31 satisfying @xmath32 the functions @xmath33 are referred to as `` statistically preserved structures '' , being invariant to the dynamics , even though _ the operator is strongly time dependent and decaying_. how to form , from these functions , infinitely many conserved variables in the decaying problem was shown in @xcite , and is discussed again in sect .
iii . the functions @xmath31 are homogeneous functions of their arguments , with anomalous scaling exponents @xmath14 : @xmath34 where `` @xmath35 '' stand for sub - leading scaling terms . since eq .
( [ zeromode ] ) contains @xmath31 on both sides , the scaling exponent @xmath14 can not be determined from dimensional considerations , and it can be anomalous .
more importantly , it was shown that the correlation functions of the forced case , @xmath36 eq .
( [ deffn ] ) , have exactly the same scaling exponents as @xmath31 @xcite . in the scaling sense
@xmath37 this is how anomalous scaling in passive fields is understood .
besides exactly soluble examples in which the advecting velocity field is non - generic ( i.e. @xmath38-correlated in time ) the existence of eigenfunctions of eigenvalue 1 of the time dependent propagator was demonstrated fully only in shell models of turbulence . while the present authors believe that shell models contain a lot of the robust properties of real turbulence ,
this belief is not universally accepted in the community .
it is therefore necessary to demonstrate that the mechanism sketched above exists indeed in the full problem eqs .
( [ ns])-([ps ] ) .
this had been done for 3rd order correlations within the lagrangian formulation in @xcite .
the aim of this paper is to demonstrate this in the eulerian frame , and for correlation functions of order 2 , 4 and 6 . in section
ii we describe the simulations of eqs .
( [ ns ] ) and ( [ ps ] ) and the type of measurements that we performed . sect .
iii is a theoretical digression , in which we analyze an exactly soluble simple model to guide ourselves as to how to analyze the numerical results to find the scaling forms of the @xmath39th oder propagators and their eigenfunctions . sect .
iv describes the analysis of the data , and sect .
v offers a discussion and a summary of the paper .
we performed a direct numerical simulation ( dns ) of eqs .
( [ ns ] ) and ( [ ps ] ) on a @xmath40 2-d array .
the forcing @xmath41 in eq .
( [ ns ] ) is random , @xmath38 correlated in time , isotropic and homogeneous .
its @xmath42 depdendence is @xmath43\ .\ ] ] this forcing is biased towards the small scales ; this is done because of the inverse energy cascade that characterizes 2-dimensional turbulence . the fluid dissipation is modeled by a hyper - viscosity term proportional to @xmath44 . in addition
we employed a `` friction '' term proportional to @xmath45 in order to stabilize the velocity field on the largest scales .
the passive field @xmath46 dissipates normally as shown in eq .
( [ ps ] ) .
the simulations were performed for a decaying passive field @xmath46 , that is , the forcing @xmath23 in eq .
( [ ps ] ) was put to zero .
the initial conditions for the @xmath46 field were of the form @xmath47 where @xmath48 is the fourier transform of the real space variable defined as : @xmath49 ( from now on we will omit the tilde above the functions and denote the k space functions only by their variables ) .
@xmath50 is a random variable in the interval @xmath51 $ ] where @xmath52 , insuring that @xmath53 , and therefore that @xmath0 is real . as the initial conditions for @xmath46 and the forcing of the @xmath45 fields are both homogeneous and isotropic so are the scalar correlation functions @xmath54 defined in eq .
( [ defg ] ) , at all times @xmath55 .
we measured the @xmath56 space correlation functions : @xmath57 where again the average @xmath58 is over initial conditions and over realizations of the @xmath45 field .
the delta function appears due to translational invariance . in accordance to eq .
( [ defprop ] ) we define the @xmath56 space propagator : @xmath59 the only two point correlator that is not zero is @xmath60 .
because of the isotropy of the initial conditions and the driving field , the correlator depends only on the magnitude of @xmath56 .
we can therefore consider the second order structure function @xmath61 , and it s propagator @xmath62 . in discrete @xmath42-space ( as in our simulation on a grid )
the propagator @xmath62 has a matrix representation . for the choice of initial conditions as in eq .
( [ init ] ) , @xmath63 is simply the @xmath64th column of the propagator : @xmath65 in figure [ fig1 ] we plot such a column of the 2-point propagator , at 10 different times .
two properties of the propagator should be noticed : as time progresses the overall amplitude decreases due to the decay , while the maximum moves from the initial @xmath64 to lower values of @xmath42 . in sects .
iii and iv we will find the scaling form of this propagator . for 10 different times .
the first time is @xmath66 , i.e. the eddy turn over time at scale @xmath67 , in lightest grey .
later times , in darker and darker greys , are at 2@xmath66 , 3@xmath66 , etc . until 10@xmath66 . ] in the case of the multi - point correlator the overall @xmath38 function and the isotropy of the fields do not reduce the dependence to a one variable function .
the propagators are therefore functions of many @xmath42-vectors .
it turns out however that measurement of the statistics for objects depending on many @xmath42-vectors are very taxing .
we opted therefore the extract from the dns partial information on the dynamics of the @xmath68 order structure functions with @xmath69 : @xmath70 accordingly we define the reduced propagators : @xmath71 in figure [ fig2 ] we present the numerical results for of the fourth and sixth order reduced propagators , at the same ten different times .
these objects display similar qualitative behavior to that of the 2-point propagator .
we will learn in the next section how to think about the scaling properties of these objects and how to re - plot the numerical results in proper rescaled variables .
( upper panel ) and @xmath72 ( lower panel ) for the same 10 different times as in fig .
[ fig1],title="fig : " ] ( upper panel ) and @xmath72 ( lower panel ) for the same 10 different times as in fig .
[ fig1],title="fig : " ]
in order to motivate the data analysis presented in sect .
iv we turn now to the kraichnan model for passive scalar advection .
the model is exactly solvable , and examining the analytic forms of the propagators offers clues to what should be expected in the generic case . in principle
we could perform this analysis in terms of the kraichnan model for a 2-dimensional passive scalar .
in fact , it is sufficient to consider the shell - model version ; the latter is very transparent to analytic manipulations , and for our purposes the results throw equally useful light on how to perform the analysis of the dns results . the kraichnan shell model for passive scalar advection @xcite , as all other shell models of turbulent flows ,
is written in terms of fourier components of the field .
the fourier components are restricted to shells denoted by the index @xmath73 , and the equation takes on the form @xmath74 here @xmath75 are the shell @xmath42 vector , @xmath76 for some @xmath64 and @xmath77 .
the shell components of the velocity field , @xmath78 , are gaussian random variables , @xmath38-correlated in time , satisfying : @xmath79 here @xmath80 is the scaling exponent of the @xmath81 field .
the second order correlator satisfies the explicit equation ( see @xcite ) : @xmath82 @xmath83 is the shell equivalent of the 2nd order structure function in eq .
( [ defpropd2 ] ) , @xmath84 the time evolution operator @xmath85 has the form @xmath86 where @xmath87 and @xmath88 is the dimensional scaling of the 2-point correlation function .
the operator @xmath85 has the following useful scaling property : @xmath89 the second order propagator has , in the limit of vanishing viscosity , the explicit form : @xmath90 note that the time evolution operator and the propagator are both hermitian , and thus admit an eigenvector decomposition .
the time evolution operator has two types of eigen - vectors @xmath91 and @xmath92 which we can regard as slow modes and fast modes respectively . here
the index @xmath42 stands for the eigen - mode index .
the fast modes are dominated by the viscous term , their support is in the viscous range , and they can essentially be taken to be @xmath93 for some shell @xmath94 with the shell index @xmath95 , above which viscosity dominates eq .
( [ kri ] ) .
the transition shell @xmath96 is determined by the condition @xmath97 . for the fast modes we have the following approximate equation : @xmath98 therefore their eigen - values are @xmath99 . for slow modes , @xmath91 , which have their support on shells smaller than @xmath96 ,
the dissipative term can be neglected and they satisfy an eigen - function equation of the form : @xmath100 for a sufficiently large inertial range we can use the scaling property ( [ propm ] ) to get : @xmath101 shifting indices we rewrite , @xmath102 define now a vector @xmath103 we have : @xmath104 we can thus define @xmath105 , an eigen - function of the time evolution operator @xmath85 with eigen - value @xmath106 .
we can thefore conclude that the eigen - vectors and eigen - values may be obtained from each other by shift of the indices : @xmath107 fig .
[ multi1 ] demonstrates that indeed two different eigen - functions coincide once shifted with respect to each other .
it follows from ( [ beta_shift ] ) that @xmath108 in fig .
[ multi2 ] the spectrum of the eigen - values is plotted , showing the two expected regions : a region for which the scaling of the eigen - values is @xmath109 for the slow modes , and @xmath110 for the fast modes .
( squares ) , and @xmath111 ( triangles ) , for @xmath112 $ ] ] ( dashed line ) for the slow modes , and by @xmath113 ( dotted line ) for the fast modes .
the transition to viscous range occur at shell @xmath114 ] the lowest eigenvalue is proportional to @xmath115 .
the eigen - function @xmath116 associated with it can be calculated explicitly , and it exhibits a normal , dimensional , scaling in the inertial range : @xmath117 using eqs .([psi_shift ] ) and ( [ zmod ] ) , that imply that @xmath91 has a scailing tail with exponent @xmath118 , starting after shell @xmath94 we establish : @xmath119 we now use these results to learn how to re - scale our numerical data .
suppose that we started with an initial condition in some shell @xmath39 in the inertial range , @xmath120 .
the fast modes do not contribute significantly to the time dependence , since they decay fast and anyhow have no support in the inertial range .
thus , using only the slow modes and eqs .
( [ krim1]),([krip ] ) , we have for @xmath121 , @xmath122 in going from the first to the second line we inserted the @xmath38-function initial conditions , getting thus a column of the propagator @xmath123
. remember that in our dns calculations we fixed the column index , and are interested in the scaling behavior with respect to the row index and time .
therefore we compute now ( up to an overall dimensional constant ) , @xmath124 where we have used the explicit form of the propagator in eq .
( [ phi1 ] ) and eq .
( [ alphalam ] ) . for @xmath73
much smaller than @xmath125 we can neglect the second term . then we conclude that the propagator is a homogeneous function of the variables @xmath126 and @xmath55 .
explicitly , multiplying by @xmath55 , @xmath127 for fixed @xmath128 this is a homogeneous function of two arguments which can be always written in the form @xmath129 for some function @xmath130 .
the symmetry between @xmath125 and @xmath39 is restored by realizing that the asymptotic form of @xmath130 is @xmath131 for large @xmath132 .
this scaling form was also found in the study of the generic shell models of passive scalar advection ( arbitrary time dependence in the velocity correlations ) @xcite and we will show that the same scaling form survives when we go back to the generic model studied by dns . the invariance of eq .
( [ pshell ] ) under a uniform phase change @xmath133 dictates that the only non - zero correlation functions will have an equal number of variables @xmath134 and conjugated variables @xmath135 .
we define : @xmath136 as in the 2-point case we can define the respective multi - point differential time derivative operator , analogous to eq .
( [ kri ] ) . in the limit of vanishing viscosity : @xmath137 the respective propagator is defined by : @xmath138 because of the time reversibility of the statistics of the @xmath45 fields , and the anti - hermiticity of the un - averaged differential operator @xmath139 in eq .
( [ ldef ] ) : @xmath140 , both the time derivative operator @xmath141 and the propagator itself are hermitian , and therefore admit eigenvalue decomposition . furthermore the operator @xmath141 has the following scaling property : @xmath142 as in the two point case the dynamics within the scaling range is determined by slow modes .
@xmath143 in this equation the index @xmath144 stands for a family of eigen - modes , and @xmath42 for their index within the family .
each @xmath144 family can be obtained from any one of its members by shifting the indices .
the eigen - values of the modes within a given family can be obtained also by a shift : @xmath145 note that in the 2-point case we had only one family of eigen - modes .
here we added the index @xmath144 to the eigen - values and eigen - modes to distinguish the different families .
note also that in eq .
( [ scalesim ] ) the scaling exponent of the eigenvalues is @xmath118 ; this stems from the scaling properties of the differential operator @xmath141 , cf .
( [ defm2 m ] ) .
however , the eigen - modes display in general anomalous scaling which can be represented by @xmath146 where @xmath147 is the ( anomalous ) scaling exponent of the @xmath144th family of eigen - modes .
since the dns were analyzed in terms of fused objects , we focus on initial conditions in eq.([prop2 m ] ) for which all the @xmath68 indices are the same .
we also measure the resulting structure functions @xmath148 this procedure will extract a fused propagator for which there are only two indices , and we denote it bellow as @xmath149 .
this propagator is not hermitian , it has in general no eigen - function decomposition , but we can understand the scaling form of any of its columns from the knowledge of the full propagator and its eigen - functions . the equivalent of eq .
( [ phi1 ] ) for @xmath150 , @xmath121 , summing over all families of slow modes ( i.e. over the index @xmath144 ) is : @xmath151 in going from first to second line we have used the fact that @xmath152 this follows directly from eqs .
( [ gamshift ] ) and ( [ shift ] ) . using the same argumentation as in eq .
( [ phi2 ] ) we get : @xmath153 here again the @xmath154 s are dimensional constants and the functions @xmath155 are some functions with the asymptotic form @xmath156 for @xmath132 large .
we note that for sufficiently long times and if the scaling exponents are well separated , we can expect the sum to be dominated by the leading scaling exponent .
the detailed analysis that was possible for the simple model of sect .
iii is not available for the generic model eqs.([ns ] ) , ( [ ps ] ) .
our strategy is to assume that the scaling forms derived in the last section are still valid for the generic model , and to demonstrate this by replotting the data accordingly .
we will see that these predictions are born out by the data . in light of eq .
( [ prp2 ] ) , we expect the 2nd order propagator @xmath157 to be , for a fixed @xmath64 , a homogeneous function of the variable @xmath158 , and to decay as @xmath159 : @xmath160 where @xmath161 is some function .
we test the correctness of this form in fig .
[ fig4 - 1 ] .
to this aim we re - plot that data shown in fig .
[ fig1 ] in different coordinates , multiplied by @xmath55 and as a function of @xmath162 the quality of the data collapse appears to strongly support the proposed scaling form .
we note that the data collapse is superior on the right of the maximum , and less convincing at its left .
we believe that this stems from two reasons .
first , there is better statistics for the right part of the curve , simply because it belongs to larger @xmath42 vectors where the angular average is more extensive .
second , the left part of the curve is more sensitive to the finite size effects , particularly to the fact that the driving velocity field loses its scaling form close to @xmath163 .
with all the limitations of the numerical simulations we consider the data collapse as very satisfactory . for 10 different times ( light grey earliest dark grey last ) ] if the prediction for the 2-point function holds , then we will have for the time dependent integral : @xmath164 in fig .
[ fig4 - 2 ] , we show that indeed after an initial period the integral settles on a scaling form consistent with eq .
( [ t2 ] ) . , ( solid line ) as a function of time . the dotted line is the expected @xmath165 .
( time is measured in this figure , and in the ones that follow in units of @xmath66 the eddy turnover time of @xmath67 ) ] using the form of the propagator we can establish the existence of a left zero mode of the 2-point propagator . integrating over the two sides of eq .
( [ zeromode ] ) we have a time independent expression on the left hand side . we therefore expect that the weighed integral of the propagator with the function @xmath166 will be constant : @xmath167 where we have used the fact that in an isotropic 2-dimensional systen we have @xmath168 .
we note that the constancy of this integral should be judged on the background of the decaying function , as done in fig .
[ fig4 - 3 ] .
we see that while the 2nd order structure function decays over three orders of magnitude , the constant " objects changes by a factor of 2 .
the lack of constancy can be attributed to the sensitivity to the outer scale as seen in the data collapse in fig .
( [ fig4 - 1 ] ) . if the collapse on the left side of the curve were perfect , so would be the constancy of the weighted integral .
note that in the calculation we have employed @xmath169 in agreement with @xcite weighed by the left zero mode @xmath170 ( solid line ) , compared to the integral of the un - weighed second order structure function in dotted line . ]
( upper panel ) and @xmath171 ( lower panel ) for 10 different times ( light grey earliest dark grey last),title="fig : " ] ( upper panel ) and @xmath171 ( lower panel ) for 10 different times ( light grey earliest dark grey last),title="fig : " ] examining eq.([phimult ] ) , we expect that for long times the functional form of the propagator for the fused correlation function defined in eq.([defpropdmult ] ) is : @xmath172 where @xmath173 is the leading scaling exponent for the @xmath68th correlation function . in fig .
[ fig4 - 4 ] we demonstrate the data collapse obtained by assuming this form for the 4th and 6th order propagators .
we notice the same excellent collapse on the right hand side of the function as in the 2nd order propagator .
also the problems with the outer scale show up in a similar manner , giving less than impressive collapse of the left hand part of the function . for the present data collapse we employed simple scaling @xmath174 ; our data does not support strongly anomalous exponents . using the form of the fused multi - point correlators
we can again predict the time behavior of their integral : @xmath175 in fig .
[ fig4 - 5 ] the time dependence of the integrals is plotted , showing agreement of similar quality to fig .
[ fig4 - 2 ] . ,
compared with the expected @xmath176 , in the dotted line .
, title="fig : " ] , compared with the expected @xmath176 , in the dotted line . , title="fig : " ] as in the 2-point case , weighing the fused multi - point functions by the appropriate multi point fused steady state correlators should yield a constant : @xmath177 where @xmath178 .
the quality of the constancy is demonstrated in fig .
[ fig4 - 6 ] .
again finite size effects lead to some decrease in time of these weighted objects , which nevertheless is very much reduced compared to the decaying correlation functions .
( upper panel ) and the 6-point propagator @xmath179 weighed by the left zero modes @xmath180 and @xmath181 respectively in solid line , compared to the integral of the un - weighed object in dotted line , title="fig : " ] ( upper panel ) and the 6-point propagator @xmath179 weighed by the left zero modes @xmath180 and @xmath181 respectively in solid line , compared to the integral of the un - weighed object in dotted line , title="fig : " ]
we have demonstrated in this paper that the generic advection of a passive scalar by a velocity field that obeys the navier - stokes equations can be discussed in terms of eulerian statistically preserved structures . by initiating a decay with delta - function initial conditions ( concentrated on @xmath182 )
we have found numerically the corresponding columns of the time - dependent propagators for the 2nd , 4th and 6th order correlation functions ( where for the 4th and 6th order objects we considered partial ( fused " ) information ) .
note that in contrast to the lagrangian formulation of statistically preserved structures @xcite , for which there is no preserved structure corresponding to the 2nd order correlation , in the eulerian formulation such an object exists and had been analyzed explicitly .
we have used a simple ( non - generic ) model of passive scalar advection to guess the analytic scaling form of the propagators in the generic problem .
the test for the relevance of this guessed form is the data collapse shown in fig . 5 and 8 .
the guessed time dependence appears to be in close correspondence with the data as shown in figs .
6 and 9 .
the analytic forms of the propagators predict the existence of eigen - modes with eigenvalue 1 ( statistically preserved structures ) .
we believe that this is the first demonstration of eulerian statistically preserved structures in a generic flow .
the numerical evidence for the constancy of the latter is encouraging , if not fully conclusive , as seen in figs . 7 and 10 .
we attributed the ( relatively small ) decrease in amplitude of the putative statistically preserved structures to the less than perfect data collapse at the largest scales ( smallest @xmath42-vectors ) that are seen in figs . 5 and 8 .
these in turn stem from the intervention of the outer scale in our scaling range , a boundary effect that we did not succeed to eliminate in our modest - size simulations .
it would be interesting to see whether larger 2-d simulations could remove this finite - size effect to demonstrate conclusively the constancy of the statistically preserved structures . in concluding we wish to point out an additional benefit to the present formulation .
usually in modeling turbulent advection it is customary to resort to dubious concepts such as ` turbulent diffusion ' in order to write a diffusion equation for the correlation functions .
the present approach indicates a much better procedure , i.e. to find , for a given turbulent field , the form of the propagator which can be then used to provide analytic predictions for any initial ( t=0 ) correlation function
. this procedure may be quite attractive for more complex hydrodynamic flows with unusual boundaries or coherent structures .
we believe that when the turbulence is sufficiently well developed scaling forms for the propagator will exist , and once found can be used to solve efficiently any statistical initial value problem .
we thank antonio celani for providing us with his code for dns of 2-dimensional turbulent advection .
this work had been supported in part by the european commission under a tmr grant , the minerva foundation , munich , germany , and the naftali and anna backenroth - bronicki fund for research in chaos and complexity . | we analyze numerically the time - dependent linear operators that govern the dynamics of eulerian correlation functions of a decaying passive scalar advected by a stationary , forced 2-dimensional navier - stokes turbulence .
we show how to naturally discuss the dynamics in terms of effective compact operators that display eulerian statistically preserved structures which determine the anomalous scaling of the correlation functions .
in passing we point out a bonus of the present approach , in providing analytic predictions for the time - dependent correlation functions in decaying turbulent transport . |
let @xmath0 be an i.i.d . sequence with @xmath1 , where we denote @xmath2 .
the filtration generated by @xmath3 will be denoted by @xmath4 .
an @xmath5-adapted complex valued process @xmath6 is called * parisian * ( walk ) if ( i ) it is a martingale starting from a point in @xmath7 \equiv \ { a + b \zeta : a , b \in \mathbb{z } \ } $ ] , and ( ii ) @xmath8 for all @xmath9 .
thus , a parisian walk is a random walk on @xmath10 $ ] .
note that there are a lot of parisian walks as functions of @xmath11 , but the law is unique up to the initial point .
$ ] .,title="fig:",width=207,height=185 ] $ ] .,title="fig:",width=215,height=185 ]
we classify the points in @xmath10 $ ] as @xmath12 @xmath13 further we define _ closure _ of @xmath14 and @xmath15 , respectively , as @xmath16 in this paper .
_ local time _ of a parisian walk is defined to be the number of exit from @xmath17 , @xmath18 and @xmath19 .
more precisely , we set @xmath20 for @xmath21 and * local time * @xmath22 is defined by @xmath23 . in addition
, an @xmath24-martingale @xmath25 with @xmath26 and @xmath27 will be called * simple * ( walk ) .
of course , the law of simple walks is unique up to the initial points . in this paper
, we present the following [ 0304152131p ] for a parisian walk @xmath28 , @xmath29 is a simple walk . here , @xmath30 is the graph distance from @xmath31 in @xmath32 $ ] , namely it is the length of the shortest path from @xmath31 . * remark .
* our results heavily rely on the arguments found in fujita s paper @xcite on 1-dimensional random walks .
we begin with the following lemma . [ 0304301706p ] let @xmath28 be a parisian walk .
then the two dimensional process @xmath33 enjoys martingale representation property ; every complex valued @xmath24-martingale is represented as a stochastic integral with respect to @xmath33 .
denote @xmath34 for @xmath35 .
fix @xmath9 and set @xmath36 for @xmath37 .
then we have @xmath38 = 1 $ ] if @xmath39 and @xmath40 otherwise because of the martingale property and of the fact that @xmath41 .
therefore @xmath42 forms an orthonormal basis ( onb ) of @xmath43 since @xmath44 .
for an adapted @xmath25 , expanding @xmath45 with respect to this onb and denoting @xmath46 = x_s $ ] , we have @xmath47 by summing up the above equation , we obtain the doob decomposition of @xmath48 , and this completes the proof . * remark . *
the above lemma can be easily extended to general unit root cases .
the point here is that parisian walk is the right discrete analogue of planer brownian motion .
let @xmath6 be a parisian walk , and let @xmath49 be a complex valued function on @xmath7 $ ] .
then we have the following formula , which would correspond to an it s formula in @xmath24 . for @xmath50 , we have @xmath51 as in the expression ( [ 0304112126p ] ) , @xmath52 for some @xmath53-measurable @xmath54 and @xmath55 . on the set of @xmath56 , @xmath57 , and @xmath58
respectively , we have @xmath59 solving ( [ 0304112135p ] ) in terms of @xmath60 , we obtain ( [ 0304112136p ] ) .
for a parisian walk @xmath28 , set @xmath61 then , we have the following * remark . * as we have seen in the proof of lemma [ 0304301706p ] , @xmath28 and @xmath73 are mutually orthogonal martingale so that , after taking appropriate scaling limit , @xmath28 converges to a so - called _ planer _ brownian motion .
then , we can prove the standard it s formula from our it s formula ( [ 0304112136p ] ) , noting that ( [ 0304112136p ] ) can be rewritten as follows if @xmath49 is real valued .
@xmath74 where we have set @xmath75 so that @xmath76 corresponds to the _ discrete laplacian _ operator in the context of random walks on graphs , and @xmath77 , which corresponds to a _ differential _ operator on @xmath78 $ ] . | in the present paper , ito formula and tanaka formula for a special kind of symmetric random walk in the complex plane are studied .
the random walk is called _ parisian walk _ , and its _ local time _ is defined to be the number of exit from some regions . |
the study of the fundamental theory of the strong interaction quantum chromo dynamics ( qcd ) in extreme conditions of temperature , density and small parton momentum fraction ( low-@xmath0 ) has attracted an increasing experimental and theoretical interest during the last 20 years . indeed ,
qcd is not only a quantum field theory with an extremely rich dynamical content ( asymptotic freedom , infrared slavery , ( approximate ) chiral symmetry , non trivial vacuum topology , strong cp violation problem , @xmath1 axial - vector anomaly , ... ) but also the only sector of the standard model whose full _ collective _ behaviour phase diagram , phase transitions , thermalization of fundamental fields is accessible to scrutiny in the laboratory .
the study of the many - body dynamics of high - density qcd covers a vast range of fundamental physics problems ( fig .
[ fig : qcd_facets ] ) : 0.8 cm ) plane .
the arrows indicate the expected crossing through the deconfinement transition during the expansion phase in heavy - ion collisions at different accelerators .
the dashed freeze - out curve indicates where hadrochemical equilibrium is attained in the latest stage of the collision @xcite.,title="fig:",width=245 ] ) plane .
the arrows indicate the expected crossing through the deconfinement transition during the expansion phase in heavy - ion collisions at different accelerators .
the dashed freeze - out curve indicates where hadrochemical equilibrium is attained in the latest stage of the collision @xcite.,title="fig:",width=309 ] * * deconfinement and chiral symmetry restoration * : lattice qcd calculations @xcite predict a new form of matter at energy densities ( well ) above @xmath2 1 gev/@xmath3 consisting of an extended volume of deconfined and bare - mass quarks and gluons : the quark gluon plasma ( qgp ) @xcite .
the scrutiny of this new state of matter equation - of - state ( eos ) , order of the phase transition , transport properties , etc . promises to shed light on basic aspects of the strong interaction such as the nature of confinement , the mechanism of mass generation ( chiral symmetry breaking , structure of the qcd vacuum ) and hadronization , which still evade a thorough theoretical description @xcite due to their highly non - perturbative nature . * * early universe cosmology * : the quark - hadron phase transition took place some 10 @xmath4s after the big - bang and is believed to have been the most important event in the universe between the electro - weak ( or susy ) transition ( @xmath5 s ) and nucleosynthesis ( @xmath6 200 s ) . depending on the order of the transition - order in pure su(3 ) gluodynamics and of a fast cross - over type for @xmath7 = 2 + 1 quarks @xcite , is still sensitive to lattice extrapolations to the continuum limit . ] , several cosmological implications have been postulated @xcite such as the formation of strangelets and cold dark - matter ( wimp ) clumps , or baryon fluctuations leading to inhomogeneous nucleosynthesis . * * parton structure and evolution at small-@xmath0 * : hera data @xcite indicate that when probed at high energies , hadrons consist of a very dense system of gluons with small ( bjorken ) momentum @xmath8 . at low @xmath0 ,
the probability to emit an extra gluon is large , proportional to @xmath9 , and @xmath10 fusion processes will play an increasing role in the parton evolution in the hadronic wavefunctions . at high virtualities @xmath11 and moderately low @xmath0 , such evolution is described by linear dglap @xcite or bfkl @xcite equations , suitable for a dilute parton regime . at @xmath12 and below an energy - dependent `` saturation momentum '' @xmath13 ,
hadrons are however more appropriately described as dense , saturated parton systems in the context of the `` colour glass condensate '' ( cgc ) @xcite effective theory with the corresponding non - linear bk / jimwlk @xcite evolution equations . since the growth of the gluon density depends on the transverse size of the hadron , saturation effects are expected to set in earlier for ultrarelativistic heavy nuclei ( for which @xmath14 , with @xmath15 the number of nucleons ) than for free nucleons . *
* gauge / string duality * : theoretical applications of the anti - de - sitter / conformal - field - theory ( ads / cft ) correspondence provide results in strongly coupled ( i.e. large t hooft coupling @xmath16 ) su(@xmath17 ) gauge theories in terms of a weakly - coupled dual gravity theory @xcite .
recent applications of this formalism for qcd - like ( @xmath18 super yang - mills ) theories have led to the determination of transport properties of experimental relevance such as the qgp viscosity @xcite , the `` jet quenching '' parameter @xmath19 @xcite , or the heavy - quark diffusion coefficient @xcite from black hole thermodynamics calculations .
such results provide valuable insights on _ dynamical _ properties of strongly - coupled qcd that can not be directly treated by perturbative or lattice methods , and open new phenomenological and experimental leads . * * compact object astrophysics * : at high baryon densities and not too high temperatures , the attractive force between ( colour antisymmetric ) quarks can lead to the formation of bound @xmath20 cooper pairs .
cold dense matter is thus expected to behave as a colour super - conductor with a non trivial quark pairing structure due to the combination of the various quantum numbers involved ( spin , colour , flavour ) @xcite .
this regime , currently beyond the direct reach of accelerator - based research ( except indirectly in the region of baryon densities around the qcd tri - critical point , fig .
[ fig : qcd_facets ] right ) , may be realised in the core of compact ( neutron , hybrid or other exotic ) stars and , thus , open to study through astronomical observation .
the only experimental means available so far to investigate the ( thermo)dynamics of a multi - parton system involves the use of large atomic nuclei collided at ultrarelativistic energies .
figure [ fig : livingston_plot ] left , shows the total center - of - mass energy available for particle production ( i.e. subtracting the rest mass of the colliding hadrons ) at different accelerators as a function of the first operation year ( `` livingston plot '' ) @xcite .
the exponential increase in performance translates into an energy doubling every 2 ( 3 ) years for the ion ( @xmath21 ) beams .
head - on collisions of heavy ions ( @xmath22 ) can produce extremely hot and dense qcd matter by concentrating a substantial amount of energy ( @xmath23(1 tev ) at mid - rapidities at the lhc , see fig .
[ fig : livingston_plot ] right ) in an _ extended _ cylindrical volume @xmath24 150 @xmath3 for a typical large nucleus with radius @xmath25 = 6.5 fm , at thermalization times of @xmath26 = 1 fm/@xmath27 . -0.5 cm = 0 , and corresponding bjorken energy density @xmath28 = 1 fm/@xmath29 @xcite , in central heavy - ion collisions at various c.m . energies @xcite fitted to a logarithmic parametrization.,title="fig:",width=294,height=226 ] 0.25 cm = 0 , and corresponding bjorken energy density @xmath28 = 1 fm/@xmath29 @xcite , in central heavy - ion collisions at various c.m . energies @xcite fitted to a logarithmic parametrization.,title="fig:",width=302,height=226 ] -0.20 cm the hot and dense systems produced in high - energy @xmath22 collisions are not prepared under controlled thermodynamical conditions but they follow a dynamical trajectory along the phase diagram shown e.g. in fig .
[ fig : qcd_facets ] , right .
after the collision , the system ( with a temperature profile decreasing from the center ) expands with relativistic longitudinal ( transverse ) velocities @xmath30 and cools at rates @xmath31 ( e.g. @xmath32 = 3 for a longitudinal - only expansion @xcite ) .
when @xmath33 reaches @xmath34 190 mev , the quark matter undergoes a phase transition into hadrons .
the produced hadronic gas stops self - interacting collectively at freeze - out times @xmath35 10 20 fm/@xmath27 @xcite . at the initial stages of the reaction ( 1 fm/@xmath27 after impact ) , the commonly used `` bjorken estimate '' @xcite gives energy densities attained at mid - rapidity of @xmath36 5 , 10 gev/@xmath3 at rhic and lhc ( fig .
[ fig : livingston_plot ] right ) .
although these values can only be considered as a lower limit since they are obtained in a simple 1 + 1d expansion scenario ignoring any effects from longitudinal work , they are already about 5 and 10 times larger , respectively , than the qcd critical energy density for deconfinement .
high - energy collisions in heavy - ion colliders provide therefore the appropriate conditions for the study of highly excited quark - gluon matter . +
this review is organised as follows .
sections [ sec : probes ] introduces the experimental probes of subhadronic matter used in high - energy @xmath22 collisions .
section [ sec : exp ] briefly reviews the experimental apparatus used in heavy - ion collisions at rhic and sps . in section [ sec : pp_dau_ref ] various hard qcd results from proton - proton collisions at rhic are presented as the `` free space '' baseline to which one compares the heavy - ion ( `` qcd medium '' ) data .
sections [ sec : saturation][sec : chiral ] each discuss a different physics observable in the context of the latest experimental results available at rhic and sps .
due to space limitations , the list of topics chosen covers but a fraction of the substantial amount of data collected in the last years . those observables that provide direct information on the _ partonic _ phases of the reaction have been given preference over `` soft '' or bulk observables from the late hadronic stages .
thus , important results relative to the freeze - out phase of the collision e.g. chemical and kinetic equilibrium from light ( and especially strange @xcite ) hadron abundances @xcite , `` femtoscopy '' measurements from hbt radii @xcite , or possible signatures of ( prehadronic ) divergent susceptibilities from final net charge and @xmath37 fluctuations @xcite are not treated here .
the selected experimental results are mostly from the comprehensive reviews of the four rhic experiments ( phenix @xcite , star @xcite , phobos @xcite , and brahms @xcite ) from @xmath38 , @xmath39 and @xmath40 collisions up to a maximum center - of - mass energy of @xmath41 = 200 gev , as well as recent results from na60 @xmath42 reactions at sps ( @xmath41 = 17.3 gev ) @xcite .
direct information on the thermodynamical and transport properties of the strongly interacting medium produced in @xmath22 collisions is commonly obtained by comparing the results for a given observable @xmath43 to those measured in proton(deuteron)-nucleus ( @xmath44 , `` cold qcd matter '' ) and in proton - proton ( @xmath40 , `` qcd vacuum '' ) collisions as a function of c.m .
energy , transverse momentum , rapidity , reaction centrality ( impact parameter @xmath45 ) , and particle type ( mass ) .
schematically : @xmath46 the observed _ enhancements _ and/or _ suppressions _ in the @xmath47 ratios can then be directly linked to the properties of the strongly interacting matter after accounting for a realistic modeling of the space - time evolution of the @xmath22 expansion process .
+ among the observables whose _ suppression _ is expected to provide information on the produced system , we will discuss : * the * total particle multiplicity * which , related via local parton - hadron duality @xcite to the number of initially produced partons , will be suppressed if the initial parton flux is reduced due to low-@xmath0 saturation effects in the colliding nuclei @xcite ( section [ sec : saturation ] ) . * * high-@xmath48 leading hadrons * are expected to be produced in reduced yields due to medium - induced energy loss via gluonsstrahlung of the parent partons in a system with a large number density of colour charges @xcite ( sections [ sec : jet_quench ] and [ sec : dijets ] ) .
* dissociation of the * heavy quarkonia * bound states has long since proposed @xcite as a sensitive signature of debye screening effects above @xmath49 ( section [ sec : jpsi ] ) .
* the * mass of light vector mesons * is expected to drop in some scenarios of chiral symmetry restoration @xcite which directly link the in - medium meson mass to a temperature- ( and baryon density- ) dependent chiral condensate : @xmath50 ( section [ sec : chiral ] ) .
likewise , the following observables discussed hereafter are expected to be _ enhanced _ in a strongly - interacting multi - parton system compared to the results in @xmath40 collisions : * the * soft hadron spectra * ( @xmath51 2 gev/@xmath27 ) , both inclusive and relative to the reaction plane , are expected to be boosted due to the development of collective radial and elliptical ( hydrodynamical ) flows in the early partonic phases of the reaction @xcite ( section [ sec : sqgp ] ) .
* semihard ( @xmath52 25 gev/@xmath27 ) * yields and flows of baryons * are predicted to be enhanced in the context of parton recombination ( or coalescence ) models @xcite which take into account modifications of the hadronization process in a dense deconfined medium ( section [ sec : recomb ] ) . *
the measurement in @xmath22 collisions of a * thermal photon * excess over the expected prompt @xmath53 yield @xcite would provide direct access to the thermodynamical properties ( temperature , eos ) of the produced system ( section [ sec : photons ] ) .
-0.5 cm collisions provide direct information on properties of qcd matter such as the @xmath19 transport coefficient , the initial gluon rapidity density @xmath54 , and the critical temperature and energy density.,title="fig:",width=302,height=245 ] -0.2 cm among all available experimental observables , particles with large transverse momentum @xmath48 and/or high mass ( `` hard probes '' ) @xcite are of crucial importance for several reasons ( fig .
[ fig : qgp_probes ] ) : ( i ) they originate from partonic scatterings with large momentum transfer @xmath11 and thus are directly coupled to the fundamental qcd degrees of freedom ; ( ii ) their production time - scale is very short , @xmath55 0.1 fm/@xmath27 , allowing them to propagate through ( and be potentially affected by ) the medium , ( iii ) their cross sections can be theoretically predicted using the perturbative qcd ( pqcd ) framework .
hard processes thus constitute experimentally- and theoretically - controlled self - generated `` tomographic '' probes of the hottest and densest phases of the reaction . the pqcd _ factorization theorem _
@xcite allows one to determine the production cross section of a given hard probe as the convolution of long - distance parton distribution ( pdfs , @xmath56 ) and fragmentation ( ffs , @xmath57 ) functions and the ( perturbatively computable up to a given order in @xmath58 ) parton - parton scattering cross section : @xmath59 the validity of eq .
( [ eq : factorization ] ) holds on the possibility to separate long- and short - distance effects with independent qcd time ( length ) scales as well as on the assumption of _ incoherent _ parton - parton scatterings . in @xmath22 collisions ,
the incoherence condition for hard processes implies : @xmath60 , i.e. that , in the _ absence _ of medium effects , the parton flux in a nucleus @xmath15 should be the same as that of a superposition of @xmath15 independent nucleons .
thus , @xmath61 and minimum - bias hard cross sections in @xmath62 collisions are expected to scale simply as @xmath63 . in the most general case , for a given @xmath62 reaction with arbitrary impact parameter
@xmath45 the yield can be obtained by multiplying the cross sections measured in @xmath40 collisions with the ratio of the incident parton flux of the two nuclei : @xmath64 , where @xmath65 ( normalised to @xmath66 ) is the nuclear overlap function at @xmath45 determined within a purely geometric glauber eikonal model using the measured woods - saxon distribution for the colliding nuclei @xcite . the standard method to quantify the effects of the medium on the yield of a hard probe produced in a @xmath22 reaction
is thus given by the _ nuclear modification factor _
: @xmath67 which measures the deviation of @xmath22 at @xmath45 from an incoherent superposition of @xmath68 collisions .
this normalization is usually known as `` binary collision scaling '' .
the relativistic heavy ion collider ( rhic ) @xcite at brookhaven national laboratory is a 3.8-km circumference accelerator composed of two identical , quasi - circular rings of superconducting magnets ( @xmath69400 dipoles and @xmath69500 quadrupoles ) with six crossing - points .
the machine , which started operation in 1999 , can accelerate nuclei ( protons ) up to a maximum of 100 ( 250 ) gev/@xmath27 per nucleon .
the center - of - mass energies in @xmath22 and @xmath44 collisions , @xmath41 = 200 gev ( @xmath70 = 500 gev for @xmath40 ) , are more than an order of magnitude larger than those at the cern sps ( @xmath41 = 17.3 gev ) .
the currently attained average @xmath22 luminosities luminosity '' for hard processes , obtained scaling by the number of nucleon - nucleon collisions , is a much larger value : @xmath71 = 15 @xmath4b@xmath72s@xmath72 .
] , @xmath73 @xmath74s@xmath72 = 0.4 mb@xmath72s@xmath72 are twice the design luminosity thanks mainly to an improved vacuum system @xcite .
there are four dedicated experiments at rhic : two large multi - detector systems ( phenix and star , fig .
[ fig : star_phenix ] ) and two smaller specialised spectrometres ( brahms and phobos ) .
+ brahms @xcite has two movable magnetic spectrometer arms with hadron identification ( @xmath75 ) capabilities up to very large rapidities ( @xmath76 4 for charged pions ) .
the two spectrometers consist of a total of 5 tracking chambers and 5 dipole magnets ( maximum field of 1.7 t ) , plus 2 time - of - flight systems and a ring imaging erenkov detector ( rich ) for particle identification .
the angular coverage of the forward spectrometer ( fs ) goes from 2.3@xmath77 @xmath78 30@xmath77 ( solid angle of 0.8 msr ) up to momenta of 35 gev/@xmath27 .
the midrapidity spectrometer ( msr ) covers 30@xmath77 @xmath78 95@xmath77 ( acceptance 6.5 msr ) .
the phobos experiment @xcite is based on silicon pad detectors and covers nearly the full solid angle ( 11 units of pseudorapidity ) for charged particles , featuring excellent global event characterization at rhic .
phobos consists of 4 subsystems : a multiplicity array ( `` octagon '' , @xmath793.2 , and `` rings '' , @xmath805.4 ) , a finely segmented vertex detector , a two - arm magnetic spectrometer ( with dipole magnet of strength 1.5 tm ) at midrapidity including a time - of - flight wall , and several trigger detectors .
particle identification ( pid ) is based on time - of - flight and energy loss in the silicon .
+ phenix @xcite is a high interaction - rate experiment specifically designed to measure hard qcd probes such as high-@xmath48 hadrons , direct @xmath53 , lepton pairs , and heavy flavour .
phenix ( fig .
[ fig : star_phenix ] , left ) achieves good mass and pid resolution , and small granularity by combining 13 detector subsystems ( @xmath69350,000 channels ) divided into : ( i ) 2 central arm spectrometers for electron , photon and hadron measurement at mid - rapidity ( @xmath81 , @xmath82 ) ; ( ii ) 2 forward - backward spectrometers for muon detection ( @xmath83 = 1.15 - 2.25 , @xmath84 ) ; and ( iii ) 4 global ( inner ) detectors for trigger and centrality selection .
two types of electromagnetic calorimeters , pbsc and pbgl , measure @xmath53 and @xmath85 .
additional electron identification is possible thanks to the rich detector .
charged hadrons are measured in the axial central magnetic field ( strength 1.15 tm ) by a drift chamber ( dc ) and 3 layers of mwpc s with pad readout ( pc ) .
hadron identification ( @xmath86 , @xmath87 , and @xmath88 ) is achieved by matching the reconstructed tracks to hits in a time - of - flight wall ( tof ) . triggering is based on two beam - beam counters ( bbc , @xmath83 = 3.0 - 3.9 ) and the zero degree calorimeters ( zdc , with @xmath89 2 mrad ) .
phenix features a state - of - the - art daq system capable of recording @xmath69300 mb / s to disk with event sizes of @xmath69100 kb ( i.e. coping with event rates of @xmath693 khz ) .
+ the star experiment @xcite is based around a large - acceptance time projection chamber ( tpc ) inside a solenoidal magnet with radius 260 cm and maximum field strength 0.5 t ( fig .
[ fig : star_phenix ] , right ) .
the tpc with radius 200 cm and full azimuthal acceptance over @xmath791.4 provides exceptional charged particle tracking and pid via ionization energy loss @xmath90 in the tpc gas and reconstruction of secondary vertices for weakly decaying particles .
additional tracking is provided by inner silicon drift detectors at midrapidity and forward radial - drift tpcs at 2.5@xmath914 .
photons and electrons are measured in the barrel and endcap electromagnetic calorimeters ( emc ) covering [email protected] and full @xmath93 .
the large star coverage allows for multi - particle correlation studies , jet reconstruction in @xmath40 , and also measurement of strange and charm hadrons . triggering is done with the zdcs , forward scintillators ( beam - beam counters ) , a barrel of scintillator slats surrounding the tpc , and the emc .
+ na60 @xcite is a high interaction - rate _ fixed - target _ experiment focused on the study of dimuon , vector meson and open charm production in @xmath94 and @xmath22 collisions at the cern sps .
the 17 m long muon spectrometer previously used by na38 and na50 , is composed of 8 multi - wire proportional chambers , 4 scintillator trigger hodoscopes , and a toroidal magnet . this spectrometer is separated by a @xmath695 m long hadron absorber ( mostly carbon ) from the vertex region , where a silicon tracker made of pixel detectors @xcite placed in a 2.5 t dipole field , measures the produced charged particles .
the matching between the muons and the vertex tracks leads to an excellent dimuon mass resolution .
proton - proton collisions are the baseline free - space reference to which one compares the @xmath22 results in order to identify initial- or final - state effects which modify the expectation of equation ( [ eq : factorization2 ] ) and can thus provide direct information on the underlying qcd medium .
figure [ fig : pp_rhic_vs_nlo ] collects six different @xmath48-differential inclusive cross sections measured at rhic in @xmath40 collisions at @xmath70 = 200 gev : jets @xcite , charged hadrons @xcite , neutral pions @xcite , direct photons @xcite , @xmath95 mesons ( indirectly measured via @xmath85 from their semileptonic decay ) @xcite all at central rapidities ( @xmath96 ) , and negative hadrons at forward rapidities ( @xmath97 = 3.2 ) @xcite .
the measurements cover 9 orders of magnitude in cross section ( from 10 mb down to 10 pb ) , and broad ranges in transverse momentum ( from zero momentum for @xmath98 mesons up to @xmath6945 gev/@xmath27 , a half of the kinematical limit , for jets ) and rapidity ( from @xmath97 = 0 up to @xmath97 = 3.2 for @xmath99 and even , not shown here , @xmath100 = 3.8 for @xmath101 s
@xcite ) . whenever there is a concurrent measurement of the same observable by two or more rhic experiments , the data are consistent with each other ( the only exception being the heavy - flavour electron cross sections , which are a factor 2 - 3 larger in star @xcite than in phenix @xcite ) .
collisions at @xmath70 = 200 gev ( 10%-30% syst .
uncertainties not shown for clarity purposes ) measured by star @xcite , phenix @xcite , and brahms @xcite compared to nlo @xcite and nll @xcite pqcd predictions ( yellow bands ) .
right : nuclear modification factors for @xmath102 versus rapidity @xcite ( top ) , and for high-@xmath48 @xmath101 at mid - rapidity @xcite ( bottom ) in central @xmath39 collisions at @xmath41 = 200 gev compared to pqcd calculations @xcite with eks98 @xcite nuclear shadowing.,title="fig:",width=325 ] collisions at @xmath70 = 200 gev ( 10%-30% syst .
uncertainties not shown for clarity purposes ) measured by star @xcite , phenix @xcite , and brahms @xcite compared to nlo @xcite and nll @xcite pqcd predictions ( yellow bands ) .
right : nuclear modification factors for @xmath102 versus rapidity @xcite ( top ) , and for high-@xmath48 @xmath101 at mid - rapidity @xcite ( bottom ) in central @xmath39 collisions at @xmath41 = 200 gev compared to pqcd calculations @xcite with eks98 @xcite nuclear shadowing.,title="fig:",width=260 ] standard next - to - leading - order ( nlo ) @xcite or resummed next - to - leading log ( nll ) @xcite perturbative qcd calculations with modern proton pdfs @xcite , hadron fragmentation functions @xcite , and with varying factorization - renormalization scales ( @xmath103 ) show an overall good agreement with the available @xmath40 data at @xmath104 = 200 gev ( yellow bands in fig .
[ fig : pp_rhic_vs_nlo ] , left ) .
this is true even in the semi - hard range @xmath105 gev/@xmath27 , where a perturbative description would be expected to give a poorer description of the spectra .
these results indicate that the hard qcd cross sections at rhic energies are well under control both experimentally and theoretically in their full kinematic domain .
this is at variance with measurements at lower ( fixed - target ) energies where several data - theory discrepancies still remain @xcite . +
not only the proton - proton hard cross sections are well under control theoretically at rhic but the hard yields in deuteron - nucleus collisions do not show any significant deviation from the perturbative expectations .
figure [ fig : pp_rhic_vs_nlo ] right , shows the nuclear modification factors measured in @xmath39 collisions at @xmath41 = 200 gev for @xmath102 , @xmath106 @xcite , and for leading @xmath101 , @xmath107 at @xmath96 @xcite . at _ mid - rapidity _
, the maximum deviation from the @xmath108 = 1 expectation for hard processes without initial - state effects is of the order of @xmath6920% .
both @xmath109 ratios are well accounted for by standard pqcd calculations @xcite that include dglap - based parametrizations of nuclear shadowing @xcite and/or a mild amount of initial - state @xmath48 broadening @xcite to account for a modest `` cronin enhancement '' @xcite .
these data clearly confirm that at mid rapidities , the parton flux of the incident gold nucleus can be basically obtained by geometric superposition of the nucleon pdfs , and that the nuclear @xmath110 modifications of the pdfs are very modest .
since no final - state dense and hot system is expected to be produced in @xmath39 collisions , such results indicate that any deviations from @xmath111 = 1 larger than @xmath6940% potentially observed for hard probes in @xmath38 collisions ( at central rapidities ) can only be due to _ final_-state effects in the medium produced in the latter reactions .
the bulk hadron multiplicities measured at mid - rapidity in central @xmath38 at @xmath41 = 200 gev @xmath113 700 , are comparatively lower than the @xmath113 1000 expectations of `` minijet '' dominated scenarios @xcite , soft regge models @xcite ( without accounting for strong shadowing effects @xcite ) , or extrapolations from an incoherent sum of proton - proton collisions @xcite ( fig .
[ fig : dndeta ] , left ) . however , colour glass condensate ( cgc ) approaches @xcite which effectively take into account a reduced _ initial _ number of scattering centers in the nuclear pdfs , @xmath114 , agree well with experimental data . in the saturation models non - linear effects become important and start to saturate the parton densities when the area occupied by the partons becomes similar to that of the hadron , @xmath115 . for a nucleus with @xmath15 nucleons and total gluon distribution @xmath116
this condition translates into the following saturation momentum @xcite : @xmath117 with @xmath118 0.20.3 @xcite .
the mass number dependence implies that , at comparable energies , non - linear effects will be @xmath119 6 times larger in a heavy nucleus ( @xmath120 200 for @xmath121 or @xmath122 ) than in a proton .
based on the general expression ( [ eq : qs ] ) , cgc - based models can describe the centrality and c.m .
energy dependences of the bulk @xmath22 hadron production ( fig .
[ fig : dndeta ] , right ) .
+ in central @xmath38 at @xmath41 = 200 gev @xcite ( the saturation model prediction is identified as mclv @xcite ) .
right : energy- and centrality- ( in terms of the number of nucleons participating in the collision , @xmath123 ) dependences of @xmath124 ( normalised by @xmath123 ) : phobos data @xcite versus the saturation prediction @xcite .
, title="fig:",width=253,height=264 ] in central @xmath38 at @xmath41 = 200 gev @xcite ( the saturation model prediction is identified as mclv @xcite ) .
right : energy- and centrality- ( in terms of the number of nucleons participating in the collision , @xmath123 ) dependences of @xmath124 ( normalised by @xmath123 ) : phobos data @xcite versus the saturation prediction @xcite .
, title="fig:",width=332 ] the second manifestation of cgc - like effects in the rhic data is the brahms observation of suppressed yields of moderately high-@xmath48 hadrons ( @xmath125 gev/@xmath27 ) in @xmath39 relative to @xmath40 collisions at forward rapidities ( @xmath126 3.2 , fig .
[ fig : rda ] ) @xcite .
hadron production at such small angles is theoretically sensitive to partons in the @xmath121 nucleus with @xmath127(10@xmath128 ) @xcite .
the observed nuclear modification factor , @xmath129 0.8 , can not be reproduced by pqcd calculations that include the same _ leading - twist _ nuclear shadowing @xcite that describes the @xmath39 data at @xmath130 0 ( fig .
[ fig : pp_rhic_vs_nlo ] , right ) but can be described by cgc approaches that parametrise the @xmath121 nucleus as a saturated gluon wavefunction @xcite . for negative hadrons measured at forward rapidities by brahms in @xmath39 at @xmath41 = 200 gev @xcite compared to predictions of leading - twist shadowing pqcd @xcite ( two upper curves ) and cgc @xcite ( lower curves).,width=294 ]
it is worth noting , however , that at rhic energies the saturation momentum is in the transition between the soft and hard regimes ( @xmath131 2 gev@xmath132 ) and that the results consistent with the cgc predictions are in a kinematic range with relatively low momentum scales : @xmath133 0.5 gev for the bulk hadron multiplicities and @xmath133 2 gev for forward inclusive hadron production .
in such a kinematic range non - perturbative effects can blur a simple interpretation based on partonic degrees of freedom alone .
following eq .
( [ eq : qs ] ) , the relevance of low-@xmath0 qcd effects will certainly be enhanced at the lhc due to the increased : center - of - mass energy , nuclear radius ( @xmath134 ) , and rapidity of the produced partons @xcite . at the lhc
, the saturation momentum @xmath131 5 10 gev@xmath132 @xcite will be more clearly perturbative and the relevant @xmath0 values in @xmath22 and @xmath94 collisions will be 3070 times lower than at rhic : @xmath135 at central ( forward ) rapidities for processes with a hard scale @xmath13610 gev .
the bulk hadron production ( @xmath137 2 gev/@xmath27 ) in @xmath38 reactions at rhic shows strong collective effects known as radial and elliptic flows .
first , the measured single hadron @xmath48 spectra have an inverse slope parameter @xmath138 larger than that measured in @xmath40 collisions , increasing with reaction centrality and with hadron mass as expected if collective expansion blue - shifts the hadron spectra .
empirically , one finds : @xmath139 , with @xmath33 and @xmath140 being the freeze - out temperature and average collective flow velocity of the `` fireball '' , and @xmath141 the mass of the hadron .
phenomenological fits of the spectra to `` blast wave '' models yield transverse flow velocities @xmath142 0.6 @xcite .
full hydrodynamical calculations which start with a partonic phase very shortly after impact ( @xmath143 1 fm/@xmath27 ) develop the amount of collective radial flow needed to accurately reproduce all the measured hadron spectra ( fig .
[ fig : hydro_spec ] ) . 3 gev/@xmath27 in 0 - 10% central @xmath38 collisions at @xmath41 = 200 gev compared to hydrodynamics(+pqcd ) calculations @xcite.,width=302 ] below @xmath144 2 gev/@xmath27 for a variety of hadrons @xcite compared to hydrodynamic predictions @xcite.,width=302 ] secondly , the azimuthal distributions @xmath145 of hadrons emitted relative to the reaction plane ( @xmath146 ) show a strong harmonic modulation with a preferential `` in - plane '' emission in non - central collisions .
such an azimuthal flow pattern is a truly collective effect ( absent in @xmath40 collisions ) consistent with an efficient translation of the initial coordinate - space anisotropy in @xmath22 reactions with non - zero impact parameter ( i.e. with a lens - shaped overlap zone , see fig . [
fig : v2 ] left ) into a final `` elliptical '' asymmetry in momentum - space .
rescattering between the produced particles drives collective motion along the pressure gradient , which is larger for directions parallel to the smallest dimension of the lens .
the strength of this asymmetry is quantified via the second fourier coefficient @xmath147 of the azimuthal decomposition of single inclusive hadron spectra with respect to the reaction plane @xcite @xmath148\right).\ ] ] the large @xmath149 0.16 measured in the data ( fig .
[ fig : v2 ] , right ) indicates a strong degree of collectivity ( pressure gradients ) building up in the first instants of the collision .
indeed , elliptic flow develops in the initial phase of the reaction and quickly self - quenches beyond @xmath35 5 fm/@xmath27 as the original spatial eccentricity disappears @xcite .
two additional experimental observations support the existence of an efficient hydrodynamical response with very short thermalization times .
first , not only light hadrons but also @xmath98 mesons ( indirectly measured via their semileptonic decays into @xmath85 ) show momentum anisotropies with @xmath150 as large as 10% @xcite .
the fact that the heavy @xmath27 and @xmath45 quarks participate in the common flow of the medium is clearly suggestive of a robust collective response during the early _ partonic _ phase .
in addition , the @xmath150 values measured for different centralities , at different center - of - mass energies ( 200 and 62 gev ) and for different colliding systems ( @xmath38 and @xmath151 ) are found to show simple scaling laws with the reaction eccentricity @xcite also in agreement with hydrodynamics expectations . + as a function of the hadron rapidity density @xmath124 normalised by the reaction overlap area @xmath152 , measured at sps @xcite and rhic @xcite compared to the `` hydrodynamical limit '' expectations for a fully thermalised system with hard ( hrg - like ) or soft ( qgp - like ) eos @xcite . adapted from @xcite ( 10% errors have been added to approx .
account for @xmath150 , @xmath153 syst .
uncertainties).,width=359,height=245 ] these results have attracted much attention for several reasons .
first , the strong @xmath150 seen in the data is not consistent with the much lower values , @xmath154 6% , expected by transport models of hadronic matter @xcite or for a partonic system rescattering with perturbative cross sections ( @xmath155 3 mb ) @xcite .
the magnitude , and the @xmath48 and hadron mass dependences of the radial and elliptic flows below @xmath52 2 gev/@xmath27 are , however , well described by _
ideal _ hydrodynamics models whose space - time evolution starts with a realistic qgp equation of state ( eos ) with initial energy densities @xmath156 30 gev/@xmath3 at thermalization times @xmath157 0.6 fm/@xmath27 @xcite ( fig .
[ fig : v2 ] ) .
second , such a degree of accord between relativistic hydrodynamics and the data was absent at lower cern - sps energies @xcite .
[ fig : v2_hydro_limit ] shows the particle - density dependence of the @xmath150 parameter ( scaled by the eccentricity of the reaction @xmath153 to remove centrality - dependent geometrical effects ) measured in semicentral collisions at different c.m .
the fact that sps data lie below the `` hydrodynamical limit '' curves @xcite estimated for a system completely thermalised , suggests that equilibration is only partially achieved at the top sps energy @xcite .
rhic @xmath150 data in the range @xmath158 62 200 gev @xcite are , however , close to the full thermalization expectations .
third , inclusion of viscous ( i.e. `` internal dissipation '' ) corrections to the ideal fluid dynamics equations spoils the reproduction of @xmath159 , especially above @xmath52 1 gev/@xmath27 where even a modest viscosity brings @xmath150 towards zero @xcite .
estimates of the maximum amount of viscosity allowed by the @xmath159 data @xcite give a value for the dimensionless viscosity / entropy ratio close to the conjectured universal lower bound , @xmath160 , obtained from ads / cft calculations @xcite .
similarly , approaches @xcite that describe simultaneously the large @xmath159 and quenching factors @xmath161 of heavy - flavour @xmath85 require small heavy - quark diffusion coefficients ( @xmath162 ) which correspond to very small shear viscosities , @xmath163 , and/or very short thermalization times .
+ the fast ( local ) thermalization times supported by the robust collective flow generated in the first instants of the reaction , and the good agreement of the @xmath48- and mass - differential spectra and @xmath150 with _ ideal _ relativistic hydrodynamics models which assume a fluid evolution with zero viscosity ( i.e. with negligible internal shear stress ) , have been presented as evidences that the matter formed at rhic is a _ strongly interacting _ qgp ( sqgp ) @xcite .
this new state of matter with _
liquid_-like properties e.g. a coulomb coupling parameter @xmath164 3 for @xmath165 at @xmath166 200 mev @xcite challenges the anticipated paradigm @xcite of a weakly interacting gas of relativistic partons ( with @xmath167 1 ) , lending support to the application of strongly - coupled - gauge / weakly - coupled - gravity duality techniques @xcite to compute relevant sqgp parameters . + it is worth noting , however , that recent lattice results predict a qcd transition temperature , @xmath34 190 mev @xcite , which is @xmath693040 mev larger than the freeze - out temperature extracted from observed particle yields in heavy ion experiments ( dotted - dashed curve in fig .
[ fig : qcd_facets ] ) @xcite .
thus , an intermediate regime between the qcd transition and freeze - out could exist during which the system created in heavy ion collisions persists in a very dense strongly - interacting hadronic phase . at lhc energies
the contribution from the qgp phase to the collective particle flow(s ) will be much larger than at rhic or sps and , therefore , the @xmath150 will be less dependent on the details of the subsequent hadronic phase
. the measurement of the differential elliptic flow properties in @xmath22 collisions at the lhc will be of primary importance to confirm or not the sqgp interpretation as well as to search for a possible weakening of the @xmath150 indicative of the existence of a weakly interacting qgp phase at higher temperatures than those of the liquid - like state found at rhic @xcite .
among the most exciting results of the rhic physics programme is the observed strong suppression of high-@xmath48 leading hadron spectra in central @xmath22 @xcite consistent with the predicted attenuation of the parent quark and gluon jets in a dense qcd medium ( `` jet quenching '' ) @xcite . above @xmath52 5 gev/@xmath27 ,
neutral mesons ( @xmath101 , @xmath97 ) @xcite and inclusive charged hadrons @xcite all show a common factor of @xmath695 suppression compared to an incoherent superposition of @xmath40 collisions ( fig .
[ fig : r_aa_rhic_200 ] , left ) .
such a significant suppression was not observed at sps where , after reevaluating the @xmath40 baseline spectrum @xcite , the central-@xmath22 meson spectra show a @xmath111 around unity ( fig .
[ fig : raa_data_vs_models ] , right ) probably due to the cancellation of a @xmath6950% final - state suppression by initial - state cronin broadening @xcite . at rhic ,
the @xmath168 1 perturbative expectation which holds for other hard probes such as `` colour blind '' photons @xcite and for high-@xmath48 hadrons in @xmath39 reactions ( fig .
[ fig : pp_rhic_vs_nlo ] , right ) is badly broken ( @xmath168 0.2 ) in central @xmath38 collisions .
this strongly supports the picture of partonic energy loss in _ final - state _ interactions within the dense matter produced in the reaction .
+ for @xmath53 , @xmath101 , @xmath97 @xcite , @xmath169 @xcite , and @xmath170 @xcite ( left ) and for `` non - photonic '' @xmath85 from d , b mesons @xcite ( right ) in central @xmath38 at @xmath41 = 200 gev compared to various parton energy loss model predictions @xcite.,title="fig:",width=321,height=213 ] for @xmath53 , @xmath101 , @xmath97 @xcite , @xmath169 @xcite , and @xmath170 @xcite ( left ) and for `` non - photonic '' @xmath85 from d , b mesons @xcite ( right ) in central @xmath38 at @xmath41 = 200 gev compared to various parton energy loss model predictions @xcite.,title="fig:",height=198 ] the dominant contribution to the energy loss is believed to be of non - abelian radiative nature ( i.e. due to gluon radiation ) as described in the glv @xcite and bdmps @xcite ( or lpci @xcite ) formalisms . in the glv approach , the initial gluon density @xmath54 of the expanding plasma ( with transverse area @xmath152 and length @xmath171 ) can be estimated from the measured energy loss @xmath172 : @xmath173 where @xmath174 is the casimir colour factor of the parton ( 4/3 for quarks , 3 for gluons ) . in the bdmps framework ,
the transport coefficient parameter can be identified with the coefficient in the exponential of an adjoint wilson loop averaged over the medium length : @xmath175 $ ] @xcite . ]
@xmath176 , characterizing the squared average momentum transfer of the hard parton per unit distance , can be derived from the average energy loss according to : @xmath177 from the general eqs .
( [ eq : glv ] ) and ( [ eq : bdmps ] ) , very large initial gluon rapidity densities , @xmath178 1100 @xmath179 300 @xcite , or equivalently , transport coefficients @xmath180 11 @xmath179 3 gev@xmath132/fm @xcite , are required in order to explain the observed amount of hadron suppression at rhic .
the corresponding values for sps are @xmath178 400 @xmath179 100 and @xmath180 3.5 @xmath179 1 gev@xmath132/fm @xcite . 7 gev/@xmath29 for @xmath181 in @xmath38 and @xmath151 collisions at @xmath41 = 200 gev compared to the bdmps - based pqm model @xcite for different values of the @xmath19 coefficient .
right : excitation function of the nuclear modification factor , @xmath182 , for @xmath101 in central @xmath22 reactions at a fixed @xmath48 = 4 gev/@xmath27 @xcite , compared to predictions of a jet - quenching model with canonical non - abelian ( solid line ) energy loss @xcite.,title="fig:",width=283,height=217 ] 7 gev/@xmath29 for @xmath181 in @xmath38 and @xmath151 collisions at @xmath41 = 200 gev compared to the bdmps - based pqm model @xcite for different values of the @xmath19 coefficient .
right : excitation function of the nuclear modification factor , @xmath182 , for @xmath101 in central @xmath22 reactions at a fixed @xmath48 = 4 gev/@xmath27 @xcite , compared to predictions of a jet - quenching model with canonical non - abelian ( solid line ) energy loss @xcite.,title="fig:",width=283 ] -0.cm most of the empirical properties of the quenching factor for light - flavour hadrons
magnitude , @xmath48- , centrality- , @xmath41- dependences of the suppression are in quantitative agreement with the predictions of non - abelian parton energy loss models ( fig .
[ fig : raa_data_vs_models ] ) .
however , the fact that the high-@xmath48 @xmath85 from semi - leptonic @xmath183 and @xmath184 decays is as suppressed as the light hadrons in central @xmath38 ( fig .
[ fig : r_aa_rhic_200 ] , right ) @xcite is in apparent conflict with the robust @xmath185 prediction of radiative energy loss models . since the gluonsstrahlung probability of quarks and gluons
is completely determined by the gauge structure ( casimir factors ) of su(3 ) , the colour octet gluons ( which fragment predominantly into _ light _ hadrons ) are expected to lose energy at @xmath186 = 9/4 times the rate of quarks .
in addition , massive @xmath187 quarks are expected to lose less energy than light ones due to their suppressed small - angle gluon radiation already in the vacuum ( `` dead - cone '' effect ) @xcite . in order to reproduce the high-@xmath48 open charm / bottom suppression ,
jet quenching models require either initial gluon rapidity densities ( @xmath178 3000 ) @xcite inconsistent with the total hadron multiplicities , @xmath188 @xcite or with the @xmath54 needed to describe the quenched light hadron spectra , or they need a smaller relative contribution of @xmath184 relative to @xmath183 mesons than theoretically expected in the measured decay electron @xmath48 range @xcite .
this discrepancy reference ( fig .
[ fig : pp_rhic_vs_nlo ] , left ) is not as good as for the light hadron spectra @xcite . ]
may point to an additional contribution from elastic ( i.e. non - radiative ) energy loss @xcite for heavy - quarks @xcite which was considered negligible so far @xcite .
the unique possibility at the lhc to fully reconstruct jets @xcite , to tag them with prompt @xmath53 @xcite or @xmath189 @xcite and to carry out detailed studies in the @xmath190 quark sector @xcite will be very valuable to clarify the response of strongly interacting matter to fast heavy - quarks , and will provide accurate information on the transport properties of qcd matter @xcite .
full jet reconstruction in @xmath22 collisions with standard jet algorithms @xcite is unpractical at rhic energies due to low cross - sections for high-@xmath191 jets and the overwhelming background of soft particles in the underlying event ( only above @xmath6930 gev are jets above the background ) . instead , jet - like
correlations at rhic are conventionally measured on a statistical basis by selecting high-@xmath48 _ trigger _ particles and measuring the azimuthal ( @xmath192 ) and rapidity ( @xmath193 ) distributions of _ associated _ hadrons ( @xmath194 ) relative to the trigger : @xmath195 combinatorial background contributions , corrections for finite pair acceptance , and the superimposed effects of global azimuthal modulations ( elliptic flow ) are taken into account with different techniques @xcite . in @xmath40 or @xmath39
collisions , a dijet signal appears as two distinct back - to - back gaussian - like peaks around @xmath196 = 0 ( near - side ) and @xmath82 ( away - side ) ( @xmath39 panel in fig . [ figs : star_dijets ] ) . at variance with this standard dijet topology in the qcd vacuum ,
early star results for semihard jets in central @xmath38 reactions @xcite showed a complete disappearance of the opposite side peak for 3 @xmath197 6 gev/@xmath27 while the near - side correlation remained unchanged ( fig .
[ figs : star_dijets ] , leftmost panel ) .
such a monojet - like topology confirmed a jet - quenching picture where a @xmath198 hard scattering takes place near the surface of the system with the trigger parton being unaffected and the away - side parton losing energy while traversing a medium opaque to coloured probes . for rising @xmath199 ,
the away - side parton is seen to increasingly `` punchthrough '' the medium @xcite , although the azimuthally - opposite correlation strength is still significantly reduced compared to @xmath39 ( fig .
[ figs : star_dijets ] ) . charged hadron pairs measured in : 0 - 5% central @xmath38 events in various @xmath199 and @xmath200 ( gev/@xmath27 ) ranges ( left ) , and in @xmath39 and 0 - 5%-central @xmath38 for fixed @xmath199 = 8 15 gev/@xmath27 and two @xmath200 ranges @xcite ( right).,title="fig:",width=287,height=257 ] charged hadron pairs measured in : 0 - 5% central @xmath38 events in various @xmath199 and @xmath200 ( gev/@xmath27 )
ranges ( left ) , and in @xmath39 and 0 - 5%-central @xmath38 for fixed @xmath199 = 8 15 gev/@xmath27 and two @xmath200 ranges @xcite ( right).,title="fig:",width=287,height=255 ] the estimated energy loss , eqs .
( [ eq : glv ] ) and ( [ eq : bdmps ] ) , of the quenched partons is very large , up to @xmath201 3 gev / fm for a 10 gev parton @xcite and most of the ( mini)jets , apart from those close to the surface , dump a significant fraction of their energy and momentum in a cell of about 1 @xmath3 in the rest frame of the medium .
since energy and momentum are conserved , the fragments of the quenched parton are either shifted to lower energy ( @xmath512 gev/@xmath27 ) and/or scattered into a broadened angular distribution . both softening and broadening of the away - side distribution
are seen in the data @xcite when the @xmath48 threshold of the away - side associated hadrons is _
lowered_. = 1 2.5 gev/@xmath27 ) relative to a higher @xmath48 trigger hadron measured at rhic . left : phenix data in central ( top ) and peripheral ( bottom ) @xmath38 @xcite ( the arrows indicate the local maxima in the away - side hemisphere ) .
right : star data in central @xmath38 ( squares ) and @xmath39 ( circles ) for two ranges of @xmath199 @xcite.,title="fig:",width=283 ] = 1 2.5 gev/@xmath27 ) relative to a higher @xmath48 trigger hadron measured at rhic . left : phenix data in central ( top ) and peripheral ( bottom ) @xmath38 @xcite ( the arrows indicate the local maxima in the away - side hemisphere ) .
right : star data in central @xmath38 ( squares ) and @xmath39 ( circles ) for two ranges of @xmath199 @xcite.,title="fig:",width=317,height=207 ] fig . [
fig : dndphi ] shows the dihadron azimuthal correlations @xmath202 in central @xmath38 collisions for @xmath200 = 1 2.5 gev/@xmath27 @xcite . in this semi - hard range
, the away - side hemisphere shows a very unconventional angular distribution with a `` dip '' at @xmath203 and two neighbouring local maxima at @xmath204 1.1 .
such a non - gaussian `` volcano''-like profile has been explained as due to the preferential emission of energy from the quenched parton at a finite angle with respect to the jet axis .
this could happen in a purely radiative energy loss scenario @xcite but more intriguing explanations for the conical - like pattern have been put forward based on the dissipation of the lost energy into a collective mode of the medium which generates a wake of lower energy gluons with mach- @xcite or erenkov - like @xcite angular emissions . in the mach - cone scenario , the _ speed of sound_. ] of the traversed matter , @xmath205 , can be determined from the characteristic supersonic angle of the emitted secondaries : @xmath206 , where @xmath207 is the mach shock wave angle and @xmath208 the time - averaged value of the speed of sound of the medium traversed by the parton . the resulting preferential emission of secondary partons from the plasma at a _ fixed _
angle @xmath209 , yields a value @xmath210 0.45 which is larger than that of a hadron - resonance gas ( @xmath211 0.35 ) @xcite , and not far from that of a deconfined qgp30% deviations from the ideal - gas limit in the @xmath212,@xmath213 and @xmath214 dependences up to very high @xmath33 s @xcite , the ideal - gas relation @xmath215 ( as well as other ratios of thermodynamical potentials ) approximately holds above @xmath216 . ] ( @xmath217 ) .
experimental confirmation of the mach - cone picture or other alternative emission mechanisms for the associated particles in the quenched jet requires more detailed differential studies @xcite such as the ongoing analysis of three- and many - particle azimuthal correlations @xcite .
the increasingly suppressed production of mesons above @xmath52 2 gev/@xmath27 in central @xmath38 reactions at rhic contrasts with the simultaneous _ unsuppressed _
@xmath88 @xcite and @xmath218 @xcite yields in the range @xmath52 2 4 gev/@xmath27 .
the intermediate-@xmath48 range in @xmath38 reactions features an anomalous baryon / meson@xmath69 0.8 ratio which is roughly four times higher than in more elementary @xmath40 or @xmath219 interactions ( fig .
[ fig : baryons ] , left ) .
semihard ( anti)protons show an enhancement with respect to the `` @xmath220 scaling '' expectation for the ratio of perturbative cross sections at different c.m .
energies @xcite , whereas @xmath220 scaling holds for _ all _ hadrons ( mesons ) measured in @xmath40 ( @xmath38 ) collisions at rhic @xcite .
only above @xmath52 6 gev/@xmath27 @xcite ( fig .
[ fig : baryons ] , left ) the meson / baryon ratio is again consistent with the expected yields obtained from universal fragmentation functions .
not only their spectra are enhanced , but at @xmath48 = 2 gev/@xmath27 the @xmath150 elliptic flow parameter of baryons exceeds that of mesons ( @xmath221 0.16 , see fig .
[ fig : v2 ] , right ) and keeps increasing up to @xmath136 4 gev/@xmath27 when it finally saturates at @xmath222 0.22 @xcite .
all those observations clearly indicate that standard hadron production via ( mini)jet fragmentation is not sufficient to explain the rhic data for baryons at transverse momenta of a few gev/@xmath27 . and @xmath38 collisions at @xmath41 = 200 gev @xcite compared to the ratio in light quark fragmentation in @xmath219 at @xmath70 = 91.2 gev ( dotted - dashed line ) and to parton coalescence predictions @xcite .
right : elliptic flow parameter @xmath150 for all hadrons at rhic normalised by the number of constituent quarks of each species ( @xmath223 = 2,3 for mesons , baryons ) @xmath224 the transverse kinetic energy @xmath225 normalised also by @xmath223 @xcite.,title="fig:",width=325 ] and @xmath38 collisions at @xmath41 = 200 gev @xcite compared to the ratio in light quark fragmentation in @xmath219 at @xmath70 = 91.2 gev ( dotted - dashed line ) and to parton coalescence predictions @xcite .
right : elliptic flow parameter @xmath150 for all hadrons at rhic normalised by the number of constituent quarks of each species ( @xmath223 = 2,3 for mesons , baryons ) @xmath224 the transverse kinetic energy @xmath225 normalised also by @xmath223 @xcite.,title="fig:",width=287,height=207 ] -0.cm the observed baryon nuclear modification factors close to unity , the high baryon / meson ratios , and their large elliptic flow have lent support to the existence of an extra mechanism of baryon production in @xmath38 based on quark coalescence in a dense partonic medium @xcite .
recombination models compute the spectrum of hadrons as a convolution of wigner functions with single parton thermal distributions leading to an exponential distribution which dominates over the standard power - law fragmentation regime below @xmath136 6 gev/@xmath27 .
coalescence models can thus reproduce the enhanced baryon production and predict also that the elliptic flow of any hadron species should follow the underlying partonic flow scaled by the number @xmath32 of ( recombined ) constituent quarks in the hadron : @xmath226 , @xmath227 for mesons and baryons , respectively @xcite .
such a quark - scaling law for the elliptic flow is confirmed by the data @xcite .
figure [ fig : baryons ] right , shows a recent variation of the quark - number scaling law for @xmath150 that uses the transverse kinetic energy ( @xmath225 , @xmath228 ) rather than the @xmath48 and seems to account perfectly for the scaled @xmath150 of all measured species also in the soft hydrodynamical regime below @xmath1362 gev/@xmath27 @xcite .
the overall success of valence quark coalescence models to explain hadron production in the semi - hard regime highlights the role of _ thermalised _ degrees of freedom in the produced system with _
( as opposed to hadronic ) quantum numbers .
in order to describe the transient systems produced in @xmath22 collisions in terms of _ thermodynamical _ variables ( @xmath33 , @xmath229 , @xmath230 , etc . ) linked by an eos which can be compared to lattice qcd expectations , it is a prerequisite to establish that the underlying degrees of freedom form , at some stage of the reaction , a statistical ensemble . proving that _ local _ 7 fm can only occur for time - scales @xmath231 7 fm/@xmath27 . ]
thermalization has been attained in the course of the collision is thus a crucial issue both experimentally and theoretically @xcite .
the large elliptic flow signal observed in the data strongly supports the idea of fast thermalization as discussed in section [ sec : sqgp ] .
the identification of real and/or virtual @xmath53 radiation from the produced `` fireball '' with properties consistent with a thermal distribution would in addition allow us to determine the underlying temperature and eos of the system .
hydrodynamical @xcite and parton transport @xcite calculations indicate that the same cascade of secondary parton - parton collisions that drives the system towards equilibration in the first tenths of fm/@xmath27 results in an identifiable emission of thermal radiation above the prompt perturbative yield in @xmath22 reactions in a window @xmath52 1 3 gev/@xmath27 .
spectra measured by phenix @xcite in central @xmath38 collisions at @xmath41 = 200 gev compared to nlo pqcd @xcite and different hydrodynamics predictions @xcite including a qgp phase with @xmath232 = 350 - 600 mev.,title="fig:",width=377,height=257 ] -0.3 cm the preliminary @xmath53 spectrum measured in central @xmath38 by phenix @xcite is consistent with the sum of such perturbative ( prompt ) plus thermal ( secondary ) contributions ( fig .
[ fig : thermal_photons ] ) .
different hydrodynamical calculations of thermal photon production @xcite can reproduce the phenix data assuming the formation of a radiating qgp with temperatures around @xmath233 ( @xmath232 = 400 600 mev , corresponding to average temperatures @xmath234 350 mev over the source profile ) at times @xmath26 = 2@xmath25/@xmath235 0.2 fm/@xmath27 right after the two colliding nuclei pass - through each other . a word of caution should be noted in that the baseline ( @xmath236-scaled ) prompt @xmath53 spectrum at @xmath48 = 1 4 gev/@xmath27 is obtained from nlo calculations , because no measurement is yet available in this @xmath48 range .
the confirmation of the existence of a thermal enhancement over the prompt component , will require a direct measurement of the @xmath40 photon spectrum down to @xmath144 1 gev/@xmath27 .
+ after subtracting the prompt @xmath53 component from the inclusive direct photon spectrum , the local inverse slope parameter @xmath138 in the range @xmath52 2 4 gev/@xmath27 can be used as a relatively good surrogate of the initial medium temperature @xcite .
the combination of @xmath138 with another global observable directly related to the entropy density @xmath230 would therefore allow one to determine the effective number of degrees of freedom @xmath237 of the system via the stefan - boltzmann ratio @xmath238 . since the final charged particle rapidity density @xmath124 is directly correlated with the initial entropy density of the system , by empirically studying the evolution of @xmath239 versus @xmath138 in different @xmath22 centralities one can effectively study the evolution of the number of degrees of freedom and look for any threshold behaviour related to the sudden increase at the transition temperature and/or a flattening of @xmath240 for temperatures above @xmath49 .
such an approach has been tested in the context of a 2d+1 hydrodynamical model which effectively reproduces the hadron and photon data at rhic @xcite .
we found that one can clearly distinguish between the equation of state of a weakly interacting qgp and that of a system with hadron - resonance - gas - like eos ( i.e. with rapidly rising number of mass states with @xmath33 ) .
however , direct evidence of the parton - hadron phase change itself as a jump in @xmath240 around @xmath241 can only be potentially visible in @xmath38 reactions at _ lower _ center - of - mass energies ( @xmath242 20 65 gev ) @xcite . at the lhc
, the expected temperatures @xmath23(1 gev ) reached in central @xmath243 will also produce a significant thermal photon signal up to @xmath52 6 gev/@xmath27 @xcite which can be used to determine the thermodynamical conditions in the _ plateau _ regime of the @xmath244 eos , closer to the ideal - gas limit than at rhic .
the study of heavy - quark bound states in high - energy @xmath22 collisions has been long since proposed as a sensitive probe of the thermodynamical properties of the produced medium @xcite .
analysis of quarkonia correlators and potentials in finite-@xmath33 lattice qcd indicate that the different charmonium and bottomonium states dissociate at temperatures for which the colour ( debye ) screening radius of the medium falls below their corresponding @xmath245 binding radius .
recent lattice analyses of the quarkonia spectral functions @xcite indicate that the ground states ( @xmath102 and @xmath246 ) survive at least up to @xmath247 whereas the less bounded @xmath248 and @xmath249 melt near @xmath49 .
experimental confirmation of such a threshold - like dissociation pattern would provide a direct means to determine the transition temperature reached in the system and their comparison to _ ab initio _
lattice qcd predictions .
a significant amount of experimental data on @xmath102 production in different @xmath44 and @xmath22 collisions has been collected at sps @xcite and rhic @xcite .
the corresponding nuclear modification factors compiled in @xcite are shown in fig .
[ fig : raa_jpsi ] as a function of @xmath250 .
the surprisingly similar amount of @xmath102 suppression observed at sps and rhic energies ( with expected temperature differences of a factor of @xmath692 ) has been interpreted in a sequential - dissociation scenario @xcite where the @xmath102 survives up to @xmath247 in agreement with the lattice predictions , and the observed suppression at both c.m .
energies is just due to the absence of ( 30% and 10% ) feed - down decay contributions from @xmath251 and @xmath252 resonances which melt already at @xmath253 .
the confirmation of such an interpretation would set an upper limit of @xmath254 400 mev for the temperatures reached at rhic .
nuclear modification factor versus centrality @xcite ( given by the number of participant nucleons in the collision ) measured in @xmath22 and @xmath44 collisions at sps @xcite and rhic @xcite.,title="fig:",width=377,height=264]-0.1 cm -0.4 cm other explanations of the comparatively low depletion of @xmath102 yields at rhic have been put forward based on a much stronger direct @xmath102 suppression ( at temperatures close to @xmath49 ) combined with @xmath255 pairs regeneration from the abundant charm quarks collision at the top rhic energy . ] in the dense medium @xcite .
the lhc measurements will be crucial to resolve this issue .
a strongly suppressed @xmath102 yield in @xmath243 at 5.5 tev where the expected initial temperatures will be well above 2@xmath49 would support the sequential - screening scenario , whereas recombination models predict a strong enhancement due to the larger density of @xmath255 pairs in the medium .
in addition , the abundant production of the @xmath256 states at lhc energies will open up a unique opportunity to study the threshold dissociation behaviour of the whole bottomonium family .
the @xmath246 is expected to survive up to @xmath257 and , therefore , _
direct _ suppression of the @xmath258 ground - state would be indicative of medium temperatures around 1 gev at the lhc .
in - medium modifications of the spectral function ( mass , width ) of the light vector mesons ( @xmath259 , @xmath260 , and @xmath93 ) have been proposed as a promising signature of the ( approximate ) restoration of chiral symmetry in the @xmath261 quark sector @xcite . in the qcd vacuum
the spontaneous breaking of chiral symmetry manifests itself in the hadron spectrum through the mass - splitting of `` chiral partners '' i.e. between states of opposite parity but equal quantum numbers .
chiral symmetry breaking leads , in the mesonic sector , to the non - degeneracy of the pseudo-/scalar ( @xmath262 ) and axial-/vector ( @xmath263 ) channels . for temperatures above the chiral transition
, massless left- and right - handed quarks will decouple and one expects to observe a gradual disappearance of the mass - splitting , leading to a shift of the masses of vector mesons and their chiral partners .
the @xmath259 meson is an excellent candidate for the experimental study of in - medium spectral functions in @xmath22 collisions due to ( i ) its short lifetime ( @xmath26 = 1.3 fm/@xmath27 ) allowing it to decay before regaining its vacuum spectral shape , and ( ii ) its ( rare but detectable ) dilepton decay branching ratio ( @xmath264 ) which is unaffected by final - state interactions with the surrounding environment .
+ mass spectrum in semi - central @xmath42 collisions at @xmath41 = 17.3 gev @xcite compared to the expected @xmath259 line - shape in different theoretical scenarios : vacuum @xmath259 , collisional broadening @xcite , and mass drop @xcite.,title="fig : " ] -0.cm the na60 experiment has recently studied a high - statistics sample of low mass muon pairs in @xmath42 collisions at @xmath41 = 17.3 gev @xcite .
the measured @xmath259 spectrum , after subtraction of all other sources of opposite - sign muon pairs , is peaked at the nominal ( free ) mass ( @xmath265 = 0.77 gev/@xmath266 ) but its width is substantially broadened for increasing centralities ( fig .
[ fig : rho_mass ] ) .
the @xmath259 spectral shape is better reproduced by models which consider in - medium width broadening due to interactions in a hot and dense _
medium close to the expected phase boundary ( @xmath267 190 mev ) @xcite , than by a long - standing prediction based on a downwards mass - shift coupled directly to the melting of the chiral condensate @xcite .
such a result sets new constraints on the possible realization of chiral symmetry in qcd matter . at rhic energies ,
the factor of two larger temperatures reached compared to sps prefigure that the @xmath259 should exhibit a completely `` melted '' line - shape if , as expected from the lattice , the chiral and deconfinement transitions occur at the same @xmath49 .
the @xmath259 measurement in the dielectron channel will be possible in phenix with the recently installed hadron - blind - detector @xcite which will allow to suppress by two orders of magnitude the large combinatorial background arising from light - meson dalitz decays and photon conversions , while preserving 50% of the signal .
high - energy collisions of heavy ions provide the only existing method today to explore empirically the phase diagram of qcd at extreme values of temperature , density and low-@xmath0 .
we have reviewed the experimental and phenomenological progress in nucleus - nucleus collisions at bnl - rhic ( @xmath158 20 200 gev ) and cern - sps ( @xmath158 20 gev ) in recent years with a particular emphasis on the modifications suffered by different hard probes in the qcd medium that is produced , compared to baseline measurements in proton - proton or proton(deuteron)-nucleus collisions .
the observed modifications allow us to obtain direct information on fundamental ( thermo)dynamical properties of strongly interacting matter .
+ the reduced @xmath38 hadron multiplicities and the depleted yields of semihard hadrons ( @xmath52 1 4 gev/@xmath27 ) at forward rapidities in @xmath39 collisions indicate that the initial conditions of the nuclei accelerated at rhic energies are consistent with colour - glass - condensate approaches that model the hadronic parton distributions as a low-@xmath0 saturated gluon wavefunction evolving according to non - linear qcd equations . these new data
complementary to the rich hera results on the partonic structure of the proton
shed new light on the high - energy limit of qcd , a physics topic not only appealing in its own right but an essential ingredient for any serious attempt to compute a large variety of hadron- , photon- and neutrino- scattering cross sections at increasingly large energies .
+ the robust radial and elliptic flows seen for all identified hadron species up to @xmath52 2 gev/@xmath27 in @xmath38 reactions at @xmath41 = 200 gev are remarkably well described by _ ideal _ relativistic hydrodynamics calculations which model the expanding system starting with a realistic qgp equation - of - state with initial energy densities @xmath156 30 gev/@xmath3 at thermalization times @xmath157 0.6 fm/@xmath27 .
detailed hydro - data comparisons for various differential observables in particular the elliptic flow parameter @xmath159 for different light and heavy hadron species indicate that the system exhibits very small viscosities ( i.e. very short mean free paths ) and is strongly coupled at variance with the anticipated qgp paradigm of a weakly interacting gas of relativistic partons . in the intermediate @xmath52 2
4 gev/@xmath27 range , baryons have enhanced yields and flows compared to mesons pointing to a novel channel for hadronization based on constituent - quark coalescence in a dense partonic medium .
the existence of a new mechanism of hadron formation in heavy - ion reactions at transverse momenta of a few gev/@xmath27 , apart from standard parton fragmentation , offers new ways to probe the space - time dynamics of confinement in different qcd environments .
+ in the high-@xmath48 sector , leading hadrons ( but not colour - blind prompt @xmath53 ) are suppressed by up to a factor of @xmath695 in the range @xmath52 4 20 gev/@xmath27 compared to perturbatively - scaled proton - proton spectra .
this result is in agreement with non - abelian energy loss models that assume that the produced partons traverse a medium with very large parton rapidity densities @xmath178 1100 and transport coefficients @xmath180 10 gev@xmath132/fm .
the energy lost by the quenched parton in the medium apparently shows up in a very unconventional conical - like azimuthal profile of secondary hadrons ( @xmath52 1 3 gev/@xmath27 ) in the away - side hemisphere of high-@xmath48 trigger hadrons .
interpretations of this preferential azimuthal emission at @xmath204 1.1 , as caused by the generation of a mach - cone boom by a supersonic parton propagating through the dense system , yield average speeds of sound @xmath268 0.45 not far from those expected from lattice qcd for deconfined quark - gluon matter .
+ in the electromagnetic sector , preliminary real and virtual @xmath53 spectra in central @xmath38 in the range @xmath48 = 1 14 gev/@xmath27 can be described as the sum of a perturbative ( prompt ) photon contribution plus a secondary component of thermal origin .
hydrodynamics calculations can reproduce the data assuming the formation of a radiating qgp with average temperatures @xmath269 350 mev .
such temperature values seem to be consistent with the observation that the amount of @xmath102 suppression at rhic and sps is about the same as expected from lattice - based calculations that predict a survival of the @xmath255 ground state up to @xmath247 but a `` melting '' of the @xmath248 and @xmath249 ( which feed - down at a @xmath6940% level to the @xmath102 ) near @xmath49 . at the cern sps ,
recent na60 results on the @xmath259 spectral function in central @xmath42 collisions at @xmath41 = 17.3 gev indicate that the width of the vector meson is substantially broadened in the medium .
theoretical calculations indicate that most of the broadening can be accounted for by collisional effects in a hot and dense hadronic medium with initial temperatures close to @xmath34175 mev ( note that such a result is also consistent with the observed hadron abundances at sps which indicate that the system reaches chemical equilibrium at temperatures around 160 mev ) .
+ the overall scenario taking form from the wealth of recent experimental data suggests , on the one hand , that heavy - ion collisions at rhic energies produce a strongly interacting liquid - like qgp with very large initial parton rapidity densities @xmath178 1100 , temperatures @xmath269 350 mev and very low shear viscosities . on the other hand ,
systems produced at sps seem to be only partially thermalised according to the lower measured collective anisotropic flow compared to rhic , have initial @xmath178 400 and temperatures around the phase boundary at @xmath34 175 mev . however , it is fair to acknowledge that among the existing signals there is yet no incontrovertible `` textbook '' figure _ proving _ the formation of a thermalised extended medium consisting of deconfined and chirally - symmetric quarks and gluons .
one such evidence would be a direct empirical observation of a jump in the number of effective degrees of freedom at the phase change as expected by eos calculations in the lattice . since rhic top energies seem to produce systems at twice @xmath49 and sps data point to conditions just at the predicted phase change , the expected jump is likely to be observed ( e.g. via precise studies of the correlation of thermal photon slopes with the global hadron multiplicities ) in a next phase of rhic running at intermediate energies ( @xmath158 20 62 gev ) and high luminosities .
likewise , confirmation of the concurrent chiral and deconfinement transitions in qcd matter will require e.g. precise measurements of the @xmath259 spectral shape ( and , ideally , that of its chiral partner @xmath270 ) at rhic energies .
lower - energy runs at rhic , as well as at the projected cbm facility @xcite , will access also the region of large baryon densities around the qcd critical point . finding signs of the tri - critical point at relatively high temperatures
would indicate that the smooth cross - over changes to a first order phase transition at higher baryon densities , a result of relevance for the conditions prevailing in the core of neutron ( and other compact ) stars .
direct validation of the strongly - coupled interpretation of the medium formed at rhic and potential observation of the anticipated _ weakly _ interacting quark - gluon plasma require key measurements in @xmath243 at 5.5 tev at cern - lhc where the initial temperatures @xmath23(1 gev ) should be large enough to observe the direct melting of ground - state quarkonia resonances .
in addition , at the lhc , the longer duration of the qgp phase and the much abundant production of other hard probes ( in particular parton energy loss results for _ fully _ reconstructed , @xmath53- or @xmath189-tagged , and flavour - identified jets ) thermal photons , @xmath150 flow parameter , etc . will likely result in indisputable probes of the deconfined medium much less dependent on details of the later hadronic phase . +
the experimental advances in the last years have been paralleled by significant progresses in the theoretical description of high - density qcd matter .
lattice methods are increasingly more refined and powerful to describe not only the infrared collective dynamics ( eos , critical parameters ) but also the in - medium correlators ( quarkonia ) .
effective field theories have been developed in specific domains such as the colour - glass - condensate which effectively describes the high - energy ( low-@xmath0 ) limit of qcd .
perturbative calculations have substantially improved the description of the interaction of hard probes with hot and dense quark - gluon matter .
last but not least , duality approaches based on the application of ads / cft correspondence between weakly coupled gravity and strongly coupled qcd - like systems , are providing new powerful insights on dynamical properties that can not be directly treated by either perturbation theory or lattice methods while simultaneously opening novel directions for phenomenological studies and experimental searches .
+ the impressive experimental and theoretical advances triggered by the wealth of high - statistics , high - quality data collected in ultrarelativistic nucleus - nucleus collisions at rhic and sps , have significantly expanded the knowledge of many - body qcd at extreme conditions of temperature , density and low-@xmath0 . those studies
which will be substantially extended in the upcoming lhc ( and likely rhic - ii ) nucleus - nucleus and proton - nucleus programme go beyond the strict realm of the strong interaction and shed light on a vast ramification of fundamental physics problems .
knowledge of the collective behaviour of many - parton systems is of primary importance not only to address basic aspects of the strong interaction such as the nature of confinement or the mechanism of mass generation via chiral symmetry breaking , but to ascertain the high - energy limit of all scattering cross sections involving hadronic objects , the inner structure of compact stellar objects , or the evolution of the early universe between the electroweak transition and primordial nucleosynthesis .
special thanks due to f. antinori , n. armesto , w. busza , a. de roeck , d. denegri , j. harris , b. jacak , p. jacobs , c. loureno , g. rolandi , c. salgado , j. schukraft , y. schutz , w. wyslouch , and b. zajc for a careful reading of the manuscript , informative discussions and useful suggestions .
this work is supported by the 6th eu framework programme contract meif - ct-2005 - 025073 .
100 see e.g. karsch f and laermann e 2003 in hwa r c ( ed ) _ et al .
vol 3 '' 1 - 59 , world scientific shuryak e v 1978 sov .
jetp * 47 * 212 [ zh . eksp .
* 74 * 408 ] ; also collins j c and perry m j 1975 _ phys .
lett . _ * 34 * 1353 , and freedman b a and mclerran l d 1977 _ phys .
d _ * 16 * 1169 jaffe a m and witten e 2000 , `` quantum yang - mills theory '' , clay maths inst . millennium prize , ` http://www.claymath.org/millennium/yang-mills_theory ` ; and gross d 2000 , in `` ten problems in fundamental physics '' , ` http://feynman.physics.lsa.umich.edu/strings2000/millennium.html ` schaefer t 2005 `` hugs 2005 lectures '' world scientific _ preprint _ hep - ph/0509068 cleymans j , oeschler h , redlich k and wheaton s 2006 32 s165 schwarz d j 2003 , ann .
12 220 h1 collaboration adloff c _ et al .
_ 2001 33 ; zeus collaboration breitweg j _ et al .
_ 1999 609 gribov v n and lipatov l n 1972 , sov .
j. nucl .
15 * 438 ; altarelli g and parisi g 1977 298 ; dokshitzer yu l 1977 , sov . phys .
jetp * 46 * 641 lipatov l n 1976 , sov .
j. nucl .
phys.*23 * 338 ; kuraev e a , lipatov l n and fadin v s 1977 , zh .
fiz * 72 * 3 ; balitsky ya ya , lipatov l n 1978 , sov .
j. nucl .
* 28 * 822 see e.g. iancu e and venugopalan r 2003 in hwa r c ( ed ) _ et al .
vol 3 '' 249 - 3363 , world scientific jalilian - marian j , kovner a , leonidov a and weigert h 1997 , 415 ; jalilian - marian j , kovner a , leonidov a and weigert h 1999 , 014014 ; iancu e , leonidov a and mclerran l 2001 583 ; kovchegov y v 1999 034008 maldacena j m 1998 adv .
phys . 2 231 [ int .
j. theor .
38 1113 ] ; witten e 1998 adv .
theor . math .
2 505 kovtun p , son d t and starinets a o 2005 , _ phys .
lett . _ * 94 * 111601 liu h , rajagopal k and wiedemann u a 2006 _ phys .
lett . _ * 97 * 182301 herzog c p _
_ 2006 jhep * 0607 * 013 , gubser s s 2006 _ phys .
d _ * 74 * 126005 , casalderrey - solana j and teaney d 2006 085012 rajagopal k and wilczek f 2000 in shifman m ( ed . ) `` at the frontier of particle physics vol . 3 '' , 2061 - 2151 ; alford m g , rajagopal k and wilczek f 1998 _ phys .
b _ * 422 * 247 schukraft j 2006 _ preprint _ nucl - ex/0602014 bjorken j d 1983 140phenix collaboration adler s s _ et al .
_ 2005 71 034908 star collaboration sahoo r _
_ 2006 _ rom .
phys . _ * 58 * 055 kolb p f and heinz u in hwa r c ( ed ) _ et al . _ `` qgp .
vol 3 '' 634 - 714 , world scientific letessier j and rafelski j 2000 _ int .
j. mod .
e _ * 9 * 107 antinori j 2004 _ j. phys .
g _ * 30 * s725 braun - munzinger p , redlich k and stachel j 2003 in hwa r c ( ed ) _ et al . _ `` qgp .
vol 3 '' , world scientific ; lisa m a , pratt s , soltz r and wiedemann u a 2005 ann .
nucl . part .
sci . * 55 * 357 jeon s and koch v 2003 in hwa r c ( ed ) _ et al .
vol 3 '' , world scientific , 430 - 490 phenix collaboration adcox k _ et al .
_ 2005 757 184 star collaboration adams j _ et al .
_ 2005 757 102 phobos collaboration back b b _ et al .
_ 2005 757 28 brahms collaboration arsene i _ et al .
_ 2005 757 1 na60 collaboration arnaldi r _
_ 2006 _ phys .
lett . _ * 96 * 162302 azimov y i , dokshitzer y l khoze v a and troian s i 1985 _ z. phys .
c _ * 27 * 65 kharzeev d and nardi n 2001 _ phys . lett . b _ * 507 * 121 ; kharzeev d , levin e and nardi m 2005 _ nucl .
a _ * 747 * 609 baier r , dokshitzer y l , mueller a h , peign s and schiff d 1997 265 ; baier r , schiff d , zakharov b g 2000 ann .
nucl . part .
sci . * 50 * 37 gyulassy m , vitev i , wang x n and zhang b w 2003 in hwa r c ( ed . ) _ et al .
. vol . 3 '' 123 - 191 world scientific , kovner a and wiedemann u a in hwa r c ( ed .
) _ et al . _ : `` qgp .
3 '' 192 - 248 world scientific matsui t and satz h 1986 178 416 see e.g. rapp r and wambach j 2000 _ adv .
_ * 25 * 1 brown g e and rho m 2002 , _ phys
. rept . _ * 363 * 85 huovinen p and ruuskanen p v 2006 _ ann .
sci . _ * 56 * hwa r c and yang c b 2003 034902 ; fries r j , muller b , nonaka n and bass s a 2003 , 044902 ; greco v , ko c m and levai p 2003 202302 stankus p 2005 ann .
nucl . part .
sci . * 55 * 517 and refs .
therein mangano m , satz h and wiedermann u `` hard probes in heavy - ion collisions at the lhc '' cern-2004 - 009 see detailed refs .
@xcite jacobs p and wang x n 2005 _ prog . part .
phys . _ * 54 * 443 collins j c , soper d e and sterman g 1985 _ nucl .
b _ * 261 * 104 see e.g. denterria d in cern yellow report on hard probes in heavy ion collisions at the lhc _ preprint _ nucl - ex/0302016 hahn h _ et al . _
2003 499 245 fischer w _
2004 epac-2004-moplt165 , lucerne , switzerland .
brahms collaboration j. adamczyk et.al .
2003 437 phobos collaboration b. b. back 2003 603 phenix collaboration adcox k 2003 _ et al .
_ 469 star collaboration ackermann k h _ et al . _
624 loureno c 2006 _ czech .
j. phys . _
* 56 * a13 ; na60 collaboration usai g _
_ 2005 43 415 keil n _
_ 2005 * a565 * 55 star collaboration abelev b i _ et al .
_ 2006 252001 star collaboration adams j _ et al . _
2003 172302 phenix collaboration adler s s _ et al .
_ 2003 _ phys .
lett . _ * 91 * 241803 phenix collaboration adler s s _ et al .
_ 2007 _ phys .
lett . _ * 98 * 012002 phenix collaboration adare a _ et al .
_ 2006 252002 brahms collaboration arsene i _ et al .
_ 2004 _ phys .
* 93 * 242303 star collaboration adams j _ et al .
_ 2004 _ phys .
lett . _ * 92 * 171801 star collaboration abelev b i 2006 _ preprint _ nucl - ex/0607012 aversa f _ et al .
_ 1989 _ nucl .
b _ * 327 * 105 , jager b _ et al .
_ 2003 _ phys .
d _ * 67 * 054005 , and vogelsang w ( private communication ) gordon l e and vogelsang w 1993 _ phys . rev .
d _ * 48 * 3136 , 1994 _ phys .
d _ * 50 * 1901 , and vogelsang w ( private communication ) ; aurenche p _
1984 87 ; and 1988 661cacciari m , nason p and vogt r 2005 _ phys .
_ * 95 * 122001 pumplin j _ et al .
_ 2002 jhep * 0207 * 012 kniehl b a , kramer g and potter b 2001 _ nucl . phys .
b _ * 597 * 337 kretzer s 2000 054001 bourrely c and soffer j 2004 _ eur
. phys .
j. c _ * 36 * 371 aurenche p
_ et al _ 2006 _ phys .
d _ * 73 * 094007 phenix collaboration adler s s _ et al .
_ 2006 * 96 * 012304 phenix collaboration adler s s _ et al .
_ 2003 , 072303 ; phenix collaboration adler s s _ et al .
_ 2006 _ preprint _ nucl - ex/0610036 vogt r 2005 _ phys .
c _ * 71 * 054902 guzey v , strikman m and vogelsang w 2004 _ phys .
b _ * 603 * 173 eskola k j , kolhinen v j and salgado c 1999 _ eur . phys .
j. c _ * 9 * 61 accardi a _ et al .
_ `` hard probes in heavy ion collisions at the lhc : pdfs , shadowing and @xmath94 collisions '' _ preprint _ hep - ph/0308248 cronin j w _ et al . _
1975 _ phys .
_ d*11 * 3105 ; antreasyan d _
et al._1979 764 gyulassy m and wang x n 1994 _ comput .
phys . commun .
_ 83 , 307 capella a , sukhatme u , tan c i and tran thanh van j 1994 _ phys . rept . _
* 236 * 225 armesto n and pajares c 2000 _ int .
j. mod .
a _ * 15 * 2019 eskola k j 2002 _ nucl .
a _ * 698 * 78 armesto n , salgado c a and wiedemann u a 2005 _ phys .
lett . _ * 94 * 022002 gribov l v , levin e m and ryskin m g 1983 _ phys .
rept . _ * 100 * 1 mueller a h and qiu j w 1986 _ nucl .
b _ * 268 * 427 mclerran l d and venugopalan r 1994 _ phys . rev .
d _ * 50 * 2225 , phys .
d * 49 * 3352 , phys .
d * 49 * 2233 accardi a 2005 _ acta phys . hung .
a _ * 22 * 289 de florian d and sassot r 2004 _ phys .
d _ * 69 * 074028 kharzeev d , kovchegov y and tuchin k 2004 _ phys .
b _ * 599 * 23 jalilian - marian j 2005 _ nucl . phys .
a _ * 748 * 664 dumitru a , hayashigaki a and jalilian - marian j 2006 , _ nucl . phys .
a _ * 765 * 464 denterria d 2007 155 denterria d and peressounko d 2006 , _ eur .
j. c _ * 46 * 451 ollitrault j y 1992 _ phys .
d _ * 46 * , 229 voloshin s and zhang y 1996 _ z. phys .
c _ * 70 * 665 kolb p f , sollfrank j and heinz u w 2000 _ phys .
c _ * 62 * 054909 akiba y [ phenix collaboration ] 2006 _ nucl
. phys .
a _ * 774 * 403 snellings r 2005 _ aip conf .
proc . _ * 756 * 390 phobos collaboration manly s _
_ 2006 _ nucl .
a _ * 774 * 523 phenix collaboration adare a _ et al _ 2006 _ preprint _ nucl - ex/0608033 zhu x i , bleicher m and stoecker h 2005 064911 molnar d and gyulassy m 2002 495 [ errat .- ibid . a * 703 * 893 ] teaney d , lauret j and shuryak e v 2001 _ preprint _ nucl - th/0110037 hirano t and tsuda k 2002 054905 na49 collaboration alt c _ et al . _ 2003 034903 voloshin s a 2006 _ aip conf.proc._ 870 691 phenix collaboration adler s s _ et al .
_ 2005 232302 phenix collaboration adcox k _ et al .
_ 2002 212301 voloshin s a and poskanzer a m 2000 27 teaney d 2003 034913
hirano t and gyulassy m 2006 71 moore g d and teaney d 2005 064904 van hees , greco v and rapp r 206 73 034913 shuryak e v 2004 _ prog . part .
phys . _ * 53 * 273 lee t d 2005 1 gyulassy
m and mclerran l 2005 30 peshier a and cassing w 2005 172301 heinz u w 2005 _ preprint _ nucl - th/0512051 thoma m h 2005 l7 cheng m _ et al .
_ 2006 74 054507 phenix collaboration adcox k _ et al .
_ 2002 022301 ; star collaboration adler c _ et al .
_ 2002 202301 ; phenix collaboration adler
s s _ et al . _ 2003 072301 bjorken j
d 1982 fermilab - pub-82 - 059-thy gyulassy g and plmer m 1990 432 ; wang x n and gyulassy m 1992 1480 phenix collaboration adler s s _ et al .
_ 2006 202301 ; shimomura m [ phenix collaboration ] 2006 457 denterria d 2004 phys .
b * 596 * , 32 wa98 collaboration aggarwal m m _ et al .
_ 2002 225 na57 collaboration antinori _
_ 2005 17 blume c 2007 65 phenix collaboration adler s s _ et al .
_ 2005 _ phys .
lett . _ * 94 * 232301 gyulassy m , levai p and vitev i 2000 5535 ; 2001 371 vitev i and gyulassy m 2002 252301 wiedemann u a 2000 30 ; salgado c a and wiedemann u a 2003 014008 zakharov b g 1997 jetp lett . * 65 * 615 ; djordjevic m , gyulassy m and wicks s 2005 112301 ; wicks s , horowitz s , djordjevic m and gyulassy m 2005 _ preprint _ nucl - th/0512076 dainese a , loizides c and paic g 2005 461 eskola k j , honkanen h , salgado c a wiedemann u a 2005 511 armesto n , dainese a , salgado c a and wiedemann u a 2005 054027 ; armesto
n _ et al .
_ 2006 362 loizides c 2007 339 denterria d 2005 295 phenix collaboration adler s s _ et al . _
2006 032301dokshitzer and y l and kharzeev d e 2001 199 wang q and wang x n 2005 71 014903 mustafa m g and thoma m h 2005 _ acta phys . hung .
a _ * 22 * 93 peshier a 2007 9 accardi a _ et al .
_ `` hard probes in heavy ion collisions at the lhc : jets '' _ preprint _ hep - ph/0310274 arleo f _ et al .
_ 2004 jhep * 0411 * 009 lokhtin i p , sherstnev a v and snigirev a m 2004 260 bedjidian m _
_ `` hard probes in heavy ion collisions at the lhc : heavy quarks '' _ preprint _ hep - ph/0311048dainese a 2007 417 see e.g blazey g c _ et al . _ 2000
_ preprint _ hep - ex/0005012 star collaboration adler c _ et al .
_ 2003 082302 star collaboration adams j _ et al .
152301 phenix collaboration adler s s _ et al . _
2006 * 97 * 052301 ajitanand n
_ 2005 phys .
c * 72 * 011902 star collaboration adams j _ et al .
_ 2006 * 97 * 162301 peitzmann t [ star collaboration ] 2006 proceeds .
hard probes06 .
polosa a d and salgado c a 2006 _ preprint _ hep - ph/0607295 stoecker h 2005 121 casalderrey - solana j , shuryak e and teaney d 2005 _ j. phys .
* 27 * 22 ruppert j and muller b 2005 123 koch v , majumder a and wang x n 2006 _ phys .
lett . _ * 96 * 172302 dremin i m 2006 nucl .
a * 767 * 233 mohanty b and alam j e 2003 _ phys .
c _ * 68 * 064903 lacey r a 2006 _ preprint _ nucl - ex/0608046.v leeuwen m 2007 109 ajitanand n n 2007 519 phenix collaboration adler s s _ et al . _
2003 172301 phenix collaboration adler s s _ et al .
_ 2004 , 034910 star collaboration adams j _ et al .
_ 2006 152301 star collaboration adams j _ et al .
_ 2004 052302 tannenbaum m j 2006 rept .
phys . * 69 * 2005 and refs .
therein star collaboration adams j _ et al .
_ 2006 161 phenix collaboration adler s s _ et al . _
2003 182301 star collaboration adams j _ et al .
_ 2005 014904 issah m and taranenko a [ phenix collaboration ] _ preprint _ nucl - ex/0604011 molnar d and voloshin s a 2003 092301 baier r , mueller a h , schiff d and son d t 2001 51 renk t , bass s a and srivastava d k 2005 632 bathe s [ phenix collaboration ] 2006 731 rasanen s s 2003 717 srivastava d k and sinha b 2001 034902 ; srivastava d k 2001 pramana * 57 * 235 turbide s , rapp r and gale c 2004 014903 alam j e , sarkar s , hatsuda t , nayak t k and sinha b 2001 021901 denterria d and peressounko d 2007 , in preparation arleo f _ et al .
_ 2003 in cern yellow report on hard probes in heavy ion collisions at the lhc _ preprint _ hep - ph/0311131 .
umeda t , nomura k and matsufuru h 2005 9 , asakawa m and hatsuda t 2004 012001 datta s , karsch f , petreczky p and wetzorke i 2004 094507 , mocsy a and petreczky p 2005 77 na38 collaboration baglin c _ et al .
_ 1995 617 na50 collaboration abreu m c _ et al .
_ 2000 28 na60 collaboration arnaldi r _
_ 2006 711 phenix collaboration a. adare _ et al .
_ 2006 _ preprint _ nucl - ex/0611020 loureno c 2007 451 karsch f , kharzeev d and satz h 2006 75 ; satz h 2007 249 thews
r l , schroedter m and rafelski j 2001 054905 ; grandchamp l and rapp r 2002 415 van hees h and rapp r 2006 102301 fraenkel z _
_ 2005 466 friese v 2006 377 | a concise review of the experimental and phenomenological progress in high - energy heavy - ion physics over the past few years is presented .
emphasis is put on measurements at bnl - rhic and cern - sps which provide information on fundamental properties of qcd matter at extreme values of temperature , density and low-@xmath0 . the new opportunities accessible at the lhc , which may help clarify some of the current open issues ,
are also outlined . |
the b5 star hr 7409 ( 7 vul , bd+19 4039 , td1 24807 , hd 183537 , hic / hip 95818 ) was noted for its broad helium lines and other unspecified spectroscopic peculiarities @xcite .
the broad helium and magnesium line profiles show that hr 7409 has a high rotation velocity @xmath2kms@xmath3 @xcite
. @xcite found it to be photometrically variable ( @xmath4 mag ) but with an unknown period , although hipparcos photometry only shows possible variations of 0.009 mag and an `` unsolved '' periodicity of 0.59 d during the survey time line @xcite . using the same data @xcite measured variations of 0.013 mag with a period of 2.72 d. the ultraviolet ( uv ) spectrum obtained with the td-1 satellite @xcite shows a b5-type star intermediate to main - sequence and supergiant @xcite .
the uv photometric measurements obtained with the astronomical netherlands satellite ( ans , * ? ? ?
* ) are consistent with td-1 flux measurements @xcite and show the effect of interstellar reddening @xcite consistent with its proximity to the galactic plane ( @xmath5 ) .
the ans photometry also shows evidence of variability .
hr 7409 originally figured among stars assembled as a likely open cluster around 4 and 5 vul @xcite and later known as collinder 399 @xcite .
after the collinder 399 membership was reduced to six objects that excluded hr 7409 @xcite , the increased precision of parallax and proper motion measurements ( hipparcos , tycho , tycho2 ) allowed @xcite and @xcite to conclude that collinder 399 is not a real cluster . in summary ,
hr 7409 is not part of a cluster , but it may belong to a local system or association referred to as the gould belt @xcite .
@xcite initially listed hr 7409 as a new spectroscopic binary but additional radial velocity measurements @xcite contradicted their initial assessment .
@xcite noted the `` nebulous '' helium lines while broad hydrogen lines were also noted by @xcite .
both observations were evidence of a high rotation velocity .
a few additional radial velocity measurements that differ from these early measurements are available in the literature @xcite underlining difficulties in measuring radial velocities in fast rotating b stars . on the other hand
, @xcite did not consider hr 7409 to be in a binary .
although hr 7409 was known to be peculiar , a detailed spectroscopic investigation of a possible link to the be phenomenon is , so far , lacking .
@xcite examined the role played by rotation on the be phenomenon ( see the review by * ? ? ?
* ) and concluded that although be stars are apparently rotating near critical velocity , other mechanisms are possibly responsible for episodes of mass loss .
in fact , @xcite found that the average rotation rate is 88% of the critical velocity , while @xcite concluded that be stars are sub - critical rotators .
we present in section 2 a series of h@xmath6 high - dispersion spectra showing that hr 7409 is a be star in a single - lined spectroscopic binary .
the h@xmath6 emission appears variable on a time - scale of years .
we constrain the parameters of the be star using the ultraviolet - to - optical spectral energy distribution ( section 3.1 ) and balmer line profiles ( section 3.2 ) .
next , we measure the binary parameters ( section 4.1 ) and mass function of the companion ( section 4.2 ) while attempting to retrace the prior evolution of the system
. we summarize and conclude in section 5 .
table [ tbl-1 ] lists astrometric and photometric measurements of hr 7409 . the ultraviolet ( ans ) , optical ( ubv ) , and infrared ( 2mass ) photometry and the hipparcos parallax help calculate the extinction - corrected absolute luminosity .
the ans magnitudes are converted to flux measurements using @xmath7 ccc parameter & value & ref .
+ + @xmath8 ( mas ) & [email protected] , [email protected] & 1,2 + @xmath10 ( masyr@xmath3 ) & [email protected] , [email protected] & 1,2 + @xmath11 ( masyr@xmath3 ) & @[email protected] , @[email protected] & 1,2 + + @xmath13[155@xmath915 nm ] ( mag ) & [email protected] & 3 + @xmath13[180@xmath915 nm ] ( mag ) & [email protected] & 3 + @xmath13[220@xmath920 nm ] ( mag ) & [email protected] & 3 + @xmath13[250@xmath915 nm ] ( mag ) & [email protected] & 3 + @xmath13[330@xmath910 nm ] ( mag ) & [email protected] & 3 + + @xmath14 ( mag ) & @[email protected] , @[email protected] & 4,5 + @xmath15 ( mag ) & @[email protected] , @[email protected] & 4,5 + @xmath16 ( mag ) & [email protected] , [email protected] & 4,5 + + @xmath17 ( mag ) & [email protected] & 6 + @xmath18 ( mag ) & [email protected] & 6 + @xmath19 ( mag ) & [email protected] & 6 + + @xmath20 ( mag ) & 0.06 , 0.07 , 0.09 , 0.08 & 7,8,9,10
+ we observed hr 7409 using the coud spectrograph attached to the 2 m telescope at ondejov observatory @xcite .
we obtained the spectroscopic series using the 830.77 lines per mm grating with a site @xmath21 ccd , with the slit width set at @xmath22 , resulting in a spectral resolution of @xmath23 and a spectral range from 6254 to 6763 .
we verified the stability of the wavelength scale by measuring the wavelength centroids of oi sky lines .
the velocity scale remained stable within 1 km s@xmath3 .
table [ tbl-2 ] presents our observation log .
[ cols= " > , > , > , > , > , > , > , > , > , > , > , > " , ]
s.v . and a.k .
are supported by ga av grant numbers iaa300030908 and iaa301630901 , respectively , and by ga r grant number p209/10/0967 . a.k . also acknowledges support from the centre for theoretical astrophysics ( lc06014 ) .
the visit of s.j . and i.p . at ondejov observatory was supported by the department of astronomy , faculty of mathematics , university of belgrade .
we thank the referee , j. zorec , for a prompt and informative review .
we also thank p. koda , p. koubsk , m. netolick , j. polster , b. kuerov , d. korkov , and p. hadrava for obtaining some of the spectra used in the present study . this research has made use of the simbad database , operated at cds , strasbourg , france .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
abt h. a. , cardona o. , 1984 , apj , 285 , 190 abt h. a. , levato h. , grosso m. , 2002 , apj , 573 , 359 baumgardt h. , 1998 , a&a , 340 , 402 bekki k. , 2009 , mnras , 398 , l36 berghoefer t. w. , schmitt j. h. m. m. , cassinelli j. p. , 1996
, a&as , 118 , 481 cardelli j. a. , clayton g. c. , mathis j. s. , 1989 , apj , 345 , 245 castelli f. , kurucz r. l. , 2003 , iau symposium 210 , eds
. n. piskunov , w.w .
weiss , and d.f .
gray , asp , a20 clark j. s. , tarasov a. e. , panko e. a. , 2003 , a&a , 403 , 239 collinder p. , 1931
lund , 2 , 1 cranmer s. r. , 2005 , apj , 634 , 585 crawford d. l. , barnes j. v. , golson j. c. , 1971 , aj , 76 , 1058 cucchiaro a. , jaschek m. , jaschek c. , macau - hercot d. , 1980 , a&as , 40 , 207 dias w. s. , lpine j. r. d. , alessi b. s. , 2001 , a&a , 376 , 441 duflot m. , figon p. , meyssonnier n. , 1995 , a&as , 114 , 269 fehrenbach c. , duflot m. , genty v. , amieux g. , 1996 , bull .
centre donnees stellaires , 48 , 11 frmat y. , zorec j. , hubert a .- m . , floquet m. , 2005 , a&a , 440 , 305 fresneau a. , monier r. , 1999 , aj , 118 , 421 friedemann c. 1992 , bull .
centre donnees stellaires , 40 , 31 gies d. r. , et al . , 2008 ,
apj , 682 , l117 gies d. r. , bagnuolo w. g. , jr . , ferrara e. c. , kaye a. b. , thaller m. l. , penny l. r. , peters g. j. , 1998 , apj , 493 , 440 guarinos j. , 1992 , eso conference and workshop proceedings , 43 , 301 hall d. s. , vanlandingham f. g. , 1970 , pasp , 82 , 640 harmanec p. , et al . , 2000 ,
a&a , 364 , l85 heber u. , 2009 , ara&a , 47 , 211 hill g. , hilditch r. w. , pfannenschmidt e. l. , 1976 , publ .
dominion astrophys .
15 , 1 hoogerwerf r. , de bruijne j. h. j. , de zeeuw p. t. , 2001 , a&a , 365 , 49 hubert a.m. , 2007 , in active ob - stars : laboratories for stellar and circumstellar physics , asp conference series , vol . 361 , 27 jamar c. , macau - hercot d. , monfils a. , thompson g. i. , houziaux l. , wilson r. , 1976 , ultraviolet bright - star spectrophotometric catalogue , esa sr-27 koen c. , eyer l. , 2002 , mnras , 331 , 45 koubsk p. , et al .
, 1997 , a&a , 328 , 551 k s. , harmanec p. , 1975
astron . inst . of czechoslovakia , 26 , 65 lamers h. j. g. l. m. , harzevoort j. m. a. g. , schrijver h. , hoogerwerf r. , kudritzki r. p. , 1997 ,
a&a , 325 , l25 lanz t. , hubeny i. , 2007 , apjs , 169 , 83 lesh j. r. , 1968 , apjs , 17 , 371 lopes de oliveira r. , motch c. , haberl f. , negueruela i. , janot - pacheco e. , 2006 , a&a , 454 , 265 lovekin c. c. , deupree r. g. , short c. i. , 2006 , apj , 643 , 460 mermilliod j. c. 2006 , homogeneous means in the ubv system ( mermilliod 1991 ) , vizier online data catalog 2168 meyer h. , 1903 , mitt .
kniglichen sternw .
breslau , 2 , 49 meyer h. , 1905 , astron .
nachri . , 167 , 321 molenda - akowicz j. , 2002 , phd thesis , wrocaw university neckel t. , klare g. , 1980 , a&as , 42 , 251 packet w. , 1981 , a&a , 102 , 17 perryman m. a. c. , et al .
, 1997 , a&a , 323 , l49 plaskett j. s. , harper w. e. , young r. k. , plaskett h. h. , 1921 , publ .
dominion astrophys .
obs . , 1 , 287 plaskett j. s. , pearce j. a. , 1931 , publ .
dominion astrophys .
, 5 , 1 poeckert r. , 1981 , pasp , 93 , 297 pols o. r. , cote j. , waters l. b. f. m. , heise j. , 1991 , a&a , 241 , 419 porter j. m. , rivinius t. , 2003 , pasp , 115 , 1153 raguzova n. v. , 2001 , a&a , 367 , 848 rappaport s. , podsiadlowski p. , horev i. , 2009 , apj , 698 , 666 savage b. d. , massa d. , meade m. , wesselius p. r. , 1985 , apjs , 59 , 397 schaller g. , schaerer d. , meynet g. , maeder a. , 1992 , a&as , 96 , 269 shipman h. l. , 1979 , apj , 228 , 240 sion e. m. , fritz m. l. , mcmullin j. p. , lallo m. d. , 1988 , aj , 96 , 251 skrutskie m. f. , et al . , 2006 ,
aj , 131 , 1163 lechta m. , koda p. , 2002
czech republic , 90 , 1 slettebak a. , 1976 , iaus , 70 , 123 steele i. a. , negueruela i. , clark j. s. , 1999 , a&as , 137 , 147 thompson g. i. , nandy k. , jamar c. , monfils a. , houziaux l. , carnochan d. j. , wilson r. , 1978 , catalogue of stellar ultraviolet fluxes ( td-1 ) .
the science research council , uk van bever j. , vanbeveren d. , 1997 , a&a , 322 , 116 waters l. b. f. m. , cote j. , pols o. r. , 1991 , a&a , 250 , 437 waters l. b. f. m. , pols o. r. , hogeveen s. j. , cote j. , van den heuvel e. p. j. , 1989 , a&a , 220 , l1 willems b. , kolb u. , 2004 , a&a , 419 , 1057 van leeuwen f. , 2007 , a&a , 474 , 653 wesselius p. r. , van duinen r. j. , de jonge a. r. w. , aalders j. w. g. , luinge w. , wildeman k. j. , 1982 , a&as , 49 , 427 wolff s. c. , edwards s. , preston g. w. , 1982 , apj , 252 , 322 | hr 7409 ( 7 vul ) is a newly identified be star possibly part of the gould belt and is the massive component of a 69-day spectroscopic binary . the binary parameters and properties of the be star measured using high - dispersion spectra obtained at ondejov observatory and at rozhen observatory imply the presence of a low mass companion ( @xmath0 ) . if the pair is relatively young ( @xmath1 myr ) , then the companion is a k v star , but , following another , older evolutionary scenario , the companion is a horizontal - branch star or possibly a white dwarf star . in the latter scenario , a past episode of mass transfer from an evolved star onto a less massive dwarf star would be responsible for the peculiar nature of the present - day , fast - rotating be star .
[ firstpage ] stars : emission - line , be stars : individual : hr 7409 |
although several interpretations of the glass transition exist ( see e.g. ref .
@xcite ) , the only microscopic approach leading to a variety of predictions is based upon the mode coupling theory ( mct ) .
this theory , which was proposed first by bengtzelius , gtze and sjlander @xcite and leutheusser @xcite , provides an equation of motion for the normalized density correlator @xmath15 of a simple liquid .
note that @xmath16 depends on @xmath17 only , due to the isotropy of the system . in its idealized version ,
mct predicts the existence of a dynamical transition at a critical temperature @xmath0 ( or a critical density @xmath18 ) from an ergodic to a nonergodic phase , corresponding to a liquid and a glass , respectively .
the nonergodicity parameter ( nep ) @xmath19 serves as an order parameter for that transition .
@xmath20 may change either continuously ( type - a transition ) or discontinuously ( type - b transition ) at @xmath0 . for the structural glass transition
only the latter is relevant .
the test of the predictions of the ideal mct has challenged both , experimental work and computer simulations .
most of the experiments have determined the intermediate scattering function @xmath21 or its time - fourier transform , the dynamical structure factor ( coherent and incoherent part ) .
this yields information on the dynamics of the translational degrees of freedom ( tdof ) , only .
a great part of these investigations was reviewed by gtze and sjgren @xcite and in refs .
@xcite . since most of the glass formers are molecular systems , it is also important to study the role of the orientational degrees of freedom ( odof ) and their coupling to the tdof .
a convenient way for this is , e.g. , the use of dielectric spectroscopy .
recent results from dielectric measurements @xcite are partly consistent with the predictions of mct for simple liquids .
the interpretation of light scattering experiments is not so obvious . since tdof and odof
may both contribute to light scattering , it is not so easy to separate the orientational part .
( see e.g. ref .
@xcite ) . concerning computer simulations ,
only a few molecular liquids have been studied so far @xcite . in a recent molecular dynamics [ md]-simulation
@xcite the present authors investigated a liquid of diatomic , rigid molecules with lennard - jones interactions .
apart from linear molecules with head - tail symmetry , this is the simplest choice for a molecular system . because dielectric and light scattering measurements yield essentially only information for @xmath22 , we restricted ourselves to orientational correlators with @xmath23 and to the translational and rotational diffusion constant @xmath24 and @xmath25 , respectively @xcite .
the main result we have found is the existence of a power law dependence of all the orientational relaxation times @xmath26 , @xmath27 ( @xmath28 ) and of @xmath24 on temperature with a single critical temperature @xmath29 , in agreement with ideal mct .
slightly above @xmath0 the numerical results for @xmath30 , @xmath27 and @xmath24 deviate from a power law , due to ergodicity restoring processes .
these processes , which are not considered by the ideal mct , can be accounted for by its extended version @xcite .
the observation that the temperature dependence of @xmath25 does not fit into this scheme , but bifurcates from @xmath24 at a temperature significantly above @xmath0 and then obeys an arrhenius law , rather than a power law , has been a further important result .
the main purpose of the present and the following paper @xcite is the extension of our recent results @xcite to finite @xmath5 and to perform a thorough test of the predictions of the ideal mct .
the present paper is restricted to the tdof only , whereas the following paper @xcite considers orientational correlators , which explicitly contain the coupling between odof and tdof .
although our simulation investigates a molecular system , we make a comparison with the predictions of mct for _ simple _ liquids .
the ideal mct for simple liquids has been recently extended to a dumbbell molecule in a simple liquid @xcite and to molecular liquids of linear @xcite and arbitrary shaped molecules @xcite .
so far , the investigations of these molecular mct - equations were restricted to the calculation of the nep @xcite . for linear molecules without head - tail symmetry , it follows from these equations @xcite that the tdof and odof freeze at a single critical temperature @xmath0 , consistent with our recent results for @xmath23 @xcite .
the investigation of the time dependent molecular mct - equations , which would allow comparison with our md - results , will be done in future .
the outline of this paper is as follows .
the next section will review the predictions of the ideal mct ( for simple liquids ) which will be tested . in section [ sec3 ] the model as well as some details of the computer simulation
are discussed . in section [ sec4 ]
we present our md - results and the final section contains a discussion of these results and our main conclusions .
in this section we will give a short summary of those predictions of the ideal mct ( for simple liquids ) , which will be compared with our md - results .
for details the reader may consult the review papers @xcite .
for completeness we also would like to mention the mct - approach by use of nonlinear fluctuating hydrodynamics @xcite .
ideal mct predicts the existence of a dynamical transition at @xmath0 from an ergodic to a nonergodic phase which is an ideal glass transition . for temperatures close to @xmath0 ,
mct predicts the existence of two scaling laws for @xmath21 with time scales @xmath31 and @xmath32 , where @xmath33 is the @xmath6-relaxation time . in the first scaling regime ,
i.e. for @xmath34 , the density correlator takes the form : @xmath35 with @xmath36 the nep at @xmath0 and @xmath37 the critical amplitude .
@xmath38 is a microscopic time of the order of @xmath39 sec .
the so - called @xmath1-correlator ( or critical correlator ) @xmath40 obeys the first scaling law : @xmath41 with the correlation scale @xmath42 and the separation parameter @xmath43 , where @xmath44 .
the @xmath45-independent master functions @xmath46 are solutions of a certain scaling equation .
the corresponding @xmath1-relaxation time scale @xmath47 is given by : @xmath48 for @xmath40 one obtains the following asymptotic power laws : @xmath49 the critical law , upper equation of ( 6 ) , holds above and below @xmath0 , whereas the von schweidler law , lower equation of ( 6 ) , is valid only above the transition point . below @xmath0 , @xmath50 decays to a constant for @xmath51 .
both exponents @xmath52 and @xmath53 are related to the exponent parameter @xmath54 by : @xmath55 with @xmath56 the gamma function . from ( 7 ) one gets that @xmath57 and @xmath58 .
@xmath59 is determined by the static correlators at @xmath0 .
the result ( 3 ) which states that in the first scaling regime the @xmath5- and @xmath60-dependence factorizes is one of the most important predictions of the ideal mct . due to this factorization ,
( 3 ) can easily be transformed to real space : @xmath61 where @xmath62 , @xmath63 and @xmath64 are the fourier transform of @xmath21 , @xmath36 and @xmath37 , respectively .
it is important to realize that the result ( 6 ) only holds asymptotically for @xmath65 .
the next order correction to both of these asymptotic laws was recently derived and calculated for a system of hard spheres by franosch _
corrections to the von schweidler law @xcite @xmath66 have already been studied for hard spheres by fuchs _
there it was demonstrated that these corrections may be important .
the expansion ( 9 ) is valid for @xmath67 , where the @xmath6-relaxation time scale @xmath33 is given by @xmath68 with @xmath69 the translational diffusion constant @xmath24 is predicted to scale like @xmath70 . therefore it is : @xmath71 corrections to the critical law lead to @xcite : @xmath72 in the second scaling regime , i.e. for @xmath60 of the order of @xmath33 , a master function @xmath73 exists such that @xmath74 in glass science the result ( 14 ) is called time - temperature superposition principle . the expansion of the r.h.s . of ( 14 ) with respect to @xmath75 yields ( 9 ) .
equation ( 14 ) represents the second scaling law .
finally we mention that all these mct - results hold for the self part of the density correlator as well .
the model we investigate is a one - component system of rigid diatomic molecules .
each molecule is composed of two different lennard - jones particles , in the following denoted by @xmath76 and @xmath77 , which are separated by a distance @xmath78 and each of which has the same mass @xmath79 .
the interaction between two molecules is given by the sum of the interaction between the four particles which is given by the lennard - jones potential @xmath80 where @xmath81 .
the lennard - jones parameters are given by : @xmath82 , @xmath83 , @xmath84 and @xmath85 . in the following
we will use reduced units and use @xmath86 as the unit of length , @xmath87 as the unit of energy ( setting @xmath88 ) and @xmath89 as the unit of time .
if the atoms are argon - like this time unit corresponds to approximately 0.3 ps . in order to make the simulation more realistic we did it at constant external pressure @xmath90=1.0 .
the length of the equilibration runs always exceeded the typical relaxation time of the system at the temperature considered , which allows us to conclude that in the subsequent production runs we were investigating the _ equilibrium _ dynamics of the system .
the temperatures we investigated are @xmath91 , 3.0 , 2.0 , 1.4 , 1.1 , 0.85 , 0.70 , 0.632 , 0.588 , 0.549 , 0.520 , 0.500 , 0.489 and 0.477 .
the total number of molecules was 500 and , in order to improve the statistics of the results , we averaged at each temperature over at least eight independent runs . for more details
see @xcite .
for a clearer presentation of our results , this section is divided into two subsections , where the first is devoted to the static and the second to the dynamical properties .
the latter is divided into two parts again , which are real space and @xmath5-space representation .
first of all we remark that thermodynamic quantities like , e.g. the average density and the enthalpy do not exhibit any signature of a singular behavior for the full temperature range @xmath92 , we have investigated . from this
we conclude that the observed slowing down is probably not related to an approach of the system to a critical point of a second order phase transition .
the structural properties are one of the most interesting static features of a supercooled liquid .
figure 1 shows the static structure factor @xmath3 of the center of mass positions for different temperatures as a function of @xmath5 .
its @xmath5-dependence has the typical behavior expected for a liquid , with a main peak at @xmath93 and a first minimum at @xmath94 for the lowest temperature .
with increasing temperature the peak positions shift to smaller @xmath5 values , due to an increase of the average distance between the molecular centers , and the peaks become broader . a comparison of @xmath3 for @xmath95 with the partial structure factors @xmath96 , where @xmath97 refers to atom @xmath76 and atom @xmath77 , is given in fig . 2 .
whereas the main peak in @xmath3 and @xmath96 is essentially at the same position , the remaining @xmath5-dependence is quite different for the various correlators . for the pre - peak of @xmath98 and @xmath99 at @xmath100 , which could indicate a `` medium '' range order , we have not found a conclusive interpretation . concerning the microscopic structural properties it is more convenient to consider the atomic radial pair distribution functions @xmath101 and the corresponding function @xmath102 for the center of mass positions , which are shown in figs . 3 and 4 , respectively .
there is no significant difference between the @xmath101 probably due to the fact that the choice of our model parameters @xcite does not introduce a strong asymmetry between particles of type @xmath76 and particles of type @xmath77 .
besides a first and a splitted second nearest neighbor peak at about @xmath103 and @xmath104 , @xmath105 , respectively ( fig .
3 ) , an intermediate peak occurs at @xmath106 .
the peak at @xmath107 is also found for @xmath108 , but its main peak is at @xmath109 ( cf .
the peaks in @xmath110 at @xmath111 and @xmath112 are missing for @xmath108 .
the splitted _ first _
nearest neighbor peak of @xmath108 can be attributed to cross- and t - configurations , as also found for other molecular liquids ( see e.g. @xcite ) .
we do not dwell more on these results , since our main concern are the dynamical features in the strongly supercooled regime .
the main reason we have presented @xmath3 is to compare its @xmath5-dependence with the static correlation functions of the orientational correlators presented in part ii .
we already anticipate that this @xmath5-dependence can be rather different for the various correlators .
to start we discuss the self motion of a molecule .
one of the important transport coefficient is the translational diffusion constant @xmath24 which is obtained from the mean squared displacement by : @xmath113 where @xmath114 is the number of molecules .
the temperature - dependence of @xmath24 is presented in fig . 5 on a double - logarithmic scale .
the diffusion constant follows a power law , as predicted by ideal mct [ cf .
( 12 ) ] , over a surprisingly large range of four decades in @xmath24 . as critical temperature
we obtain @xmath29 and for the corresponding exponent @xmath115 . with use of eqs .
( 7 ) and ( 11 ) , this yields the von schweidler exponent @xmath116 , the critical exponent @xmath117 and the exponent parameter @xmath118 . at the two lowest temperatures the deviations of the numerical values for @xmath24 from the power law , which has been mentioned in the introduction , can clearly be seen .
the inset of fig .
5 depicts the time dependence of the mean squared displacement for all temperatures , from which @xmath119 was deduced .
more information on the self motion is contained in the self part of the van hove correlator : @xmath120 the @xmath121-dependence of @xmath122 is shown in fig .
6 for times which are equidistant on a logarithmic time axis and for the lowest temperature . with time , the @xmath123-peak at @xmath124 broadens and for @xmath125 and @xmath126 , i.e. in the hydrodynamic limit , it approaches a gaussian distribution : @xmath127 in the time span @xmath128 , where the mean squared displacement for the lowest temperature exhibits a plateau ( cf .
the inset of fig .
5 ) , the @xmath121-dependence of @xmath122 varies only weakly with time , which indicates the validity of the factorization [ eq . ( 8) ] . to check this ,
we have plotted in fig .
7 @xmath129 as a function of @xmath121 for fixed @xmath130 and @xmath131 and various @xmath60-values .
( 8) is valid , this ratio must be independent of @xmath60 and @xmath131 .
for the times @xmath60 and @xmath131 we have chosen in fig . 7 , this is indeed reasonably fulfilled . that this is not true for the full time span ,
can be inferred from the inset of fig .
another interesting conclusion can be drawn from the @xmath121-dependence of @xmath132 for the largest time . as can be seen from fig .
6 , at no time there is any indication for the presence of a secondary peak at @xmath133 , the nearest neighbor distance . from the absence of such a peak it is usually concluded , see e.g. ref .
@xcite , that no hopping processes are present .
but for the present system we have strong evidence , see fig . 5 , that at the lowest temperature hopping processes are indeed present .
thus we conclude that hopping processes do not necessarily lead to a secondary peak in @xmath132 .
the @xmath121-dependence of @xmath132 is not gaussian for a large time regime .
the deviation from a gaussian can be quantified by the non - gaussian parameters @xmath134 @xcite .
8 shows @xmath135 in case of a gaussian process @xmath136 vanishes . for @xmath137 and
@xmath125 @xmath136 goes to zero .
but in between there exists a time regime for all temperatures where @xmath138 is substantially different from zero .
the increasing part of @xmath138 seems to fall onto a master curve , as already found earlier @xcite .
the time @xmath139 , where @xmath136 obtains its maximum , obeys a power law @xmath140 ( see inset of fig .
8) with @xmath141 which coincides with @xmath142 , derived from @xmath119 , in agreement with the results of sciortino _ et al . _ @xcite .
this suggests that @xmath136 for @xmath143 is determined by the @xmath6-relaxation as it follows from ideal mct .
exactly the same conclusions also hold for @xmath144 .
the collective dynamics can be studied by use of the distinct part of the van hove correlator @xmath145 which is shown in fig . 9 for various times and again for the lowest temperature .
its @xmath121- and @xmath60-dependence is quite similar to that found for a binary liquid @xcite . the test of the factorization [ eq . ( 8) ]
can be done as for the self part .
the corresponding ratio is presented in fig .
10 for the same time regime as in fig .
a reasonable collapse onto one master curve occurs , in agreement with ideal mct .
although the visualization of local structural properties favors the use of a real space representation of the correlators , most experimental results are obtained in @xmath5-space .
in addition the translational invariance of the interactions also suggests a theoretical description by use of @xmath5-dependent correlators .
for instance the memory part of the mct - equation for the density correlator @xmath21 is `` diagonal '' in @xmath5-space but involves a convolution in real space .
let us start with the self part @xmath146 of the collective density correlator @xmath21 [ cf .
( 1 ) ] , which is presented in fig . 11 for @xmath147 , the position of the main peak in @xmath3 ( cf .
fig . 1 ) .
we remind the reader that @xmath5 is measured in units of @xmath86 , the position of the lennard - jones minimum for the @xmath76-particles .
the small bump in @xmath146 at @xmath148 , which is also present in @xmath21 ( cf .
14 and 15 ) , is due to a sound wave and @xmath149 is the time for the sound wave to traverse the box size of our sample .
this effect was also observed by lewis _
_ @xcite and has recently been shown to be much more pronounced in strong glass formers @xcite .
for the highest temperature , @xmath91 , the decay of @xmath146 is essentially exponential .
with decreasing temperature the relaxation crosses over into a two - step process , as predicted by mct .
for the lowest temperature , @xmath150 , a `` quasi - plateau '' can be seen .
its height is a measure of the critical nep @xmath151 .
the reader should note that @xmath152 decays to zero for large times , indicating that the length of the runs is large enough to equilibrate the system . in order to test mct ,
we have determined the @xmath6-relaxation times @xmath153 from the condition @xmath154 .
its temperature dependence is shown in fig .
12 ( squares ) for @xmath147 and @xmath155 .
using @xmath0 = 0.475 , as determined from @xmath119 , both relaxation times follow a power law over about three decades .
the corresponding @xmath10-values are practically the same @xmath156 , but differ significantly from @xmath157 .
such a discrepancy between @xmath142 and @xmath158 has been observed already before @xcite and indicates that the mct prediction , that the two exponents should be equal , is not valid for these systems .
@xmath156 yields the von schweidler exponent @xmath159 and the exponent parameter @xmath160 .
we also note that a power law fit with @xmath0 as a free parameter yields a value for @xmath0 which coincides with @xmath29 to within @xmath161 .
the relaxation time @xmath153 can now be used to rescale time in order to test the time - temperature superposition principle [ eq .
( 14 ) ] , which is done in fig .
we find that the relaxation curves fall indeed onto one master curve for the lowest temperatures . the late stage relaxation can be fitted well by a kohlrausch - williams - watts - law ( kww ) , i.e. @xmath162 $ ] , dashed curve . for higher temperatures ( cf .
e.g. the curve for @xmath163 ) no such scaling exists .
we also note that such a scaling was not possible for certain types of _ orientational _ correlation functions @xcite , which shows that this prediction of mct is not a trivial one .
having demonstrated the validity of the second scaling law [ eq .
( 14 ) ] , we can test whether the von schweidler law , including the next order correction [ eq .
( 9 ) ] fits the master curve well in the late @xmath1-relaxation regime . as can be seen from fig .
11b this type of fit works very well ( dotted line ) . in practice , this has been done for @xmath146 at the lowest temperature by using @xmath159 ( deduced from @xmath158 ) .
the wave vector dependence of @xmath151 , @xmath164 and @xmath165 is shown in fig . 13 . here
a comment is in order .
the fit of the data with the von schweidler law including the next order correction yields @xmath166 and @xmath167 [ cf .
since the @xmath5-independent @xmath6-relaxation time scale @xmath168 can only be determined up to a @xmath169-independent factor , the same is true for @xmath170 and @xmath171 . in fig .
13 we therefore show @xmath172 and @xmath173 .
the @xmath5-variation of these quantities is quite similar to that for hard spheres @xcite .
particularly we also find a zero of @xmath165 at a finite q - value @xmath174 with @xmath175 for @xmath176 .
taking into account that @xmath5 for the hard sphere system is given in units of the diameter @xmath177 , we can deduce @xmath178 from ref .
that this is about three times larger than our value , is partly due to a larger effective diameter @xmath179 of our molecules , compared to @xmath177 .
we now turn to the collective dynamics as obtained from the density correlator @xmath21 , which is presented , respectively , for @xmath2 and @xmath4 in figs .
14a and 14b .
the time - dependence looks similar to that of @xmath146 , but the height of the `` quasi - plateau '' is much lower and the stretching of the relaxation is more pronounced for the case of @xmath4 . the @xmath6-relaxation time @xmath180 has been determined from the condition @xmath181 , since using the @xmath182 definition of @xmath33 would lead to an underestimation of the @xmath6relaxation time for @xmath4 , since the plateau is so low .
( we note that choosing @xmath182 instead of @xmath183 , for the case where both definitions can be used , or even using a kww - fit to deduce @xmath180 , leads to essentially the same temperature - dependence ) .
the temperature dependence of @xmath8 is shown in fig .
12 . taking
again @xmath0 = 0.475 as given , @xmath180 obeys a power law , with @xmath184 and @xmath185 .
it is remarkable that the range for the power law for @xmath4 is almost one decade less than for @xmath2 , which shows that different correlators reach the asymptotic regime at different temperatures .
15 shows that also for these correlation functions the second scaling law holds and that it holds better for @xmath186 than for @xmath187 .
the late stage relaxation can again be fitted well by a kww - law ( dashed line ) . since the curves for the different temperatures fall onto a master curve , it is sufficient to focus on the curve for the lowest temperature in order to test whether the first scaling law holds .
the results of our analysis are shown in fig .
16 and 17 for @xmath2 and @xmath4 , respectively .
let us discuss @xmath2 first .
as can be seen from fig .
16 , the von schweidler law , with @xmath188 , the same value we used to fit the von schweidler law to @xmath146 , fits the data over about 2.5 decades in time ( long dashed line ) .
taking into account the next order correction ( cf .
9 ) leads to a significant improvement of the fit for @xmath189 ( solid line ) .
we have also fitted the numerical data with the @xmath1-correlator with ( dotted line ) and without ( short dashed line ) correction .
this was done by solving the equation for @xmath190 ( see ref .
@xcite ) and making use of eqs .
( 3 ) and ( 4 ) . here
we encounter the same problem as we did for the fit with the von schweidler law , because the @xmath1-relaxation time @xmath31 can only be determined up to a @xmath169-independent factor . in order to proceed
we have chosen for @xmath31 the position of the inflection point of @xmath191 , indicated in fig .
16 by a filled circle .
the optimum values @xmath160 and @xmath192 stemming from that fit were used for the similar fit of all the other correlators of the present paper and of ref .
@xcite , except for the orientational correlators with @xmath28 @xcite . from the figure we recognize that in the late @xmath1-relaxation regime these fits are identical to the ones of the von schweidler law , as it should be . in the early @xmath1-relaxation regime
, however , the @xmath1-correlator fits the data much better than the von schweidler law , since part of the approach to the plateau is fitted well also .
nevertheless , the critical law is not really observed , even if one includes the corrections to it [ cf .
( 13 ) ] . for @xmath4
it is not possible to obtain a good fit with the von schweidler law alone , if @xmath53 is kept fixed at 0.55 .
if @xmath53 is used as a free parameter , a satisfactory fit is obtained but at the cost of a @xmath5-dependent von schweidler exponent @xmath193 , in contradiction to mct .
since the critical nep obtained from this fit were rather structureless and did not qualitatively agree with the nep obtained from solving the molecular mct - equations @xcite , we decided to keep @xmath53 = 0.55 ( obtained from @xmath158 ) fixed and to take the next order correction into account . with this approach we obtained good agreement between all nep from our md - simulation and those from the mct - equations @xcite .
but even if the first correction to the von schweidler law ( cf .
9 ) is taken into account , the fit with eq .
9 works well for only about one decade in time . on the other hand , the critical correlator ( with @xmath194 ) including the correction ( on the von schweidler side ) fits the data over two decades .
nevertheless , this range is significantly smaller than the one found for @xmath2 which shows that corrections to the scaling laws might be more important for @xmath187 than for @xmath186 .
the quantities @xmath36 , @xmath195 and @xmath196 , depicted in fig .
18 were determined in a similar way as for the self part , i.e. by keeping the value of @xmath53 fixed to 0.55 . comparing these quantities with @xmath3 ( fig .
1 ) we find that @xmath36 is in phase and both , @xmath195 and @xmath196 in anti - phase with @xmath3 .
we note that the peak at @xmath197 in @xmath36 does not exist for @xmath3 .
however , whether this peak is real or just a statistical fluctuation can presently not be decided for sure . since the static correlator @xmath198 has a pronounced maximum at @xmath199 ( see ref .
@xcite ) , one might be tempted to relate this prepeak of @xmath200 to the translation - rotation coupling .
the variation of @xmath36 and @xmath195 with @xmath5 resembles the one found , e.g. , for hard spheres @xcite , for a binary liquid @xcite and that for water molecules @xcite .
interestingly the @xmath5-dependence of @xmath196 , which is in phase with that of @xmath195 , qualitatively agrees with that found by fuchs _
_ for hard spheres @xcite
. however , in contrast to the hard sphere system , @xmath196 does not change sign for the @xmath5-regime we have studied , but vanishes at @xmath2 at which it has a minimum .
this fact explains why the ( asymptotic ) von schweidler law fits the data rather well for @xmath2 , where @xmath201 , but not for @xmath4 . to conclude
we show in fig .
19 that also the variation of the @xmath6-relaxation time @xmath8 with @xmath5 is in phase with @xmath3 , which is in close analogy with , e.g. , the hard sphere system @xcite .
in this paper we have performed a detailed test of the predictions of ideal mct ( for simple liquids ) for the translational degrees of freedom ( tdof ) for a supercooled liquid of rigid diatomic molecules without head - tail symmetry .
the lennard - jones interactions we used for the molecular system are not so different from those of the binary liquid @xcite with @xmath202 a- and @xmath203 b - atoms .
the main difference between both systems of course is , that for the present case the @xmath204 a- and @xmath205 b - atoms are pairwise connected in order to form diatomic molecules .
therefore the comparison of the dynamical behavior of both systems , allows to discuss the influence of the orientational degrees of freedom ( odof ) on tdof .
our test was mainly concerned with ( i ) the existence of a single critical temperature @xmath0 and ( ii ) the validity of the two scaling laws of ideal mct .
this single critical temperature indicates a strong coupling between tdof and odof .
although the mathematical structure of the mct - equations for molecular liquids @xcite differs from the one of simple liquids , we expect that the two scaling laws hold for molecular liquids , too .
this belief is based upon the fact that the first scaling law is always valid in a so - called type b transition ( i.e. the nep changes _ discontinuously _ at the transition ) , the type of transition that is relevant for structural glasses @xcite .
furthermore the existence of a single transition temperature for linear and arbitrary molecules without additional symmetry can also be proved for the molecular mct - equation @xcite .
taking @xmath0 and @xmath10 as free parameters , the @xmath6-relaxation times @xmath7 and @xmath8 for @xmath2 and @xmath4 can be fitted with a power law .
the resulting @xmath0 s differ from @xmath0 = 0.475 ( deduced from @xmath119 ) by less than @xmath161 .
the same value of @xmath0 was found for the orientational correlation functions in ref .
@xcite and we therefore conclude that our data is compatible with the existence of a single transition temperature . the sharp transition at @xmath0 , as predicted by _ ideal _ mct , is however smeared out , due to ergodicity restoring processes which can be accounted for by the extended mct @xcite .
these processes are often associated with a hopping of the atoms or molecules , as it has been demonstrated for a binary liquid @xcite .
an evidence for such hopping processes is the occurrence of a second peak in @xmath122 at @xmath206 for large times @xcite .
but no such peak is observed in our results , even beyond the @xmath121-range shown in fig .
the absence of hopping processes with respect to the tdof was also found in the md - simulation of ckn @xcite .
therefore only orientational jumps remain , which indeed have been shown to be present @xcite . comparing the different values of @xmath10
, one finds that those determined from @xmath146 and @xmath21 fluctuate around 2.55 .
the same is more or less true for the purely orientational correlators @xmath207 and @xmath208 [ 17 ] for @xmath209 , but not for @xmath28 where @xmath210 and @xmath211 was found .
this result demonstrates that , contrary to the prediction of ideal mct , @xmath10 is not a `` universal '' exponent for our system .
such a conclusion was already drawn for binary liquids @xcite .
we note however , that this discrepancy does not seem to exist for the md - simulation of water @xcite .
for all the correlators related to the tdof , the second scaling law is reasonably well fulfilled .
this is in contrast to our findings for @xmath13 and @xmath14 @xcite . only for relatively large values of @xmath212
do the @xmath207 show the second scaling law .
regarding the first scaling law we can say that it works well for @xmath146 and @xmath21 . for @xmath4
we have demonstrated that the next correction to the von schweidler law becomes important .
we stress that a consistent description of our data in the framework of mct is only possible by taking this correction into account . if this is not done , a @xmath5-dependent von schweidler exponent and a rather structureless @xmath5-variation of the orientational nep ( which will be discussed in the following paper @xcite ) results .
the critical law is not observed for any of the correlators , even if the next correction is taken into account .
this is probably due to a strong interference of the critical dynamics with the microscopic motion . in summary
, we conclude that , with respect to ideal mct , the tdof of our molecular system behave quite similar as found for a binary liquid @xcite . hence , the qualitative features of the dynamics of tdof is not altered by its coupling to odof .
although a single transition temperature for tdof and odof can be specified , part of the orientational dynamics as measured , e.g. by @xmath13 , does not fit into the framework of ideal mct @xcite . whether this is only due to the @xmath213-reorientational jumps @xcite , is presently not clear .
j. jckle , rep .
phys . * 49 * , 171 ( 1986 ) ; j. p. sethna , j. d. shore and m. huang , phys .
b44 , 4943 ( 1991 ) ; s. a. kivelson , x. zhao , d. kivelson , t. m. fischer and c. m. knobler , j. chem .
phys * 101 * , 2391 ( 1994 ) ; d. kivelson , s. a. kivelson , x. zhao , z. nussinov and g. tarjus , physica a * 219 * , 27 ( 1995 ) .
u. bengtzelius , w. gtze and a. sjlander , j. phys .
c * 17 * , 5915 ( 1984 ) .
e. leutheusser , phys .
a * 29 * , 2765 ( 1984 ) .
w. gtze and l. sjgren , rep .
* 55 * , 241 ( 1992 ) .
w. gtze in _ liquids , freezing and the glass transition _ , eds .
j. p. hansen , d. levesque and j. zinn - justin ( north - holland , amsterdam , 1991 ) ; r. schilling in `` disorder effects on relaxation processes '' , eds .
r. richert and a. blumen ( springer , berlin , 1994 ) ; theme issue on relaxation kinetics in supercooled liquids - mode coupling theory and its experimental tests , ed .
s. yip , volume * 24 * , no .
6 - 8 ( 1995 ) of _ transport theory and statistical physics _ ; w. kob , p. 28 in _ experimental and theoretical approaches to supercooled liquids : advances and novel applications _ ,
j. fourkas , d. kivelson , u. mohanty , and k. nelson ( acs books , washington , 1997 ) .
p. lunkenheimer , a. pimenow , b. schiener , r. bhmer and a. loidl , europhys . lett . * 33 * , 611 ( 1996 ) ; p. lunkenheimer , a. pimenow , m. dressel , y. g. gonchunov , r. bhmer and a. loidl , phys . rev . lett . * 77 * , 318 ( 1996 ) ; p. lunkenheimer , a. pimenow , m. dressel , y. g. gonchunov , u. schneider , b. schiener r. bhmer and a. loidl , proc .
mrs fall meeting , boston 1996 . h. z. cummins , g. li , w. du , r. pick , and c. dreyfus , phys .
e * 53 * , 896 ( 1996 ) .
g. f. signorini , j .- l .
barrat and m. l. klein , j. chem . phys . * 92 * , 1294 ( 1990 ) .
p. sindzingre , m. l. klein , j. chem . phys . *
96 * , 4681 ( 1992 ) .
l. j. lewis and g. wahnstrm , phys .
rev * e 50 * , 3865 ( 1994 ) . g. wahnstrm and l. j. lewis , prog .
( in press ) .
s. r. kudchadkar and j. m. wiest , j. chem .
phys . * 103 * , 8566 ( 1995 ) .
p. gallo , f. sciortino , p. tartaglia and s .- h .
chen , phys .
* 76 * , 2730 ( 1996 ) ; f. sciortino , p. gallo , p. tartaglia and s. h. chen , phys .
e * 54 * , 6331 ( 1996 ) .
f. sciortino , p. tartaglia and p. gallo , proc .
mrs fall meeting , boston 1996 . w. j. ma and s. k. lai , phys .
e * 55 * , 2026 ( 1997 ) .
s. kmmerer , w. kob and r. schilling , phys .
e * 56 * , xxxx ( 1997 ) .
w. gtze and l. sjgren , z. phys .
b * 65 * , 415 ( 1987 ) .
s. kmmerer , w. kob and r. schilling ( following paper ) .
t. franosch , f. fuchs , w. gtze , m. r. mayr and a. p. singh ( unpublished ) .
r. schmitz ( unpublished ) .
t. scheidsteger and r. schilling , 6 .
workshop on disordered systems , 1997 ; r. schilling and t. scheidsteger , phys .
e ( in press ) . c. theis , diploma thesis , johannes gutenberg - universitt ( 1997 ) .
s. p. das and g. f. mazenko , phys .
a * 34 * , 2265 ( 1986 ) .
r. schmitz , j. w. dufty and p. de , phys .
* 71 * , 2066 ( 1993 ) .
t. franosch , m. fuchs , w. gtze , m. r. mayr and a. p. singh , phys .
e * 55 * , xxxx , ( 1997 ) .
w. gtze in _ amorphous and liquid materials _ , ed .
e. lscher ( nijhoff , dordrecht , 1987 ) .
m. fuchs , i. hofacker and a. latz , phys .
a * 45 * , 898 ( 1992 ) .
j. p. hansen and i. r. mcdonald , `` theory of simple liquids '' , 2nd edn . ,
( academic press , london , 1986 ) ; w. b. street , d. j. tildesley , proc .
a. * 348 * , 485 ( 1976 ) and * 355 * , 239 ( 1977 ) .
w. kob and h. c. andersen , phys .
e * 51 * , 4626 ( 1995 ) ; _ ibid . _ * 52 * , 4134 ( 1995 ) .
a. rahman , phys .
136 , a * 405 * ( 1964 ) ; j. p. boon and s. yip , `` molecular hydrodynamics '' ( dover , new york , 1980 ) .
j. horbach , w. kob , k. binder and c.a .
angell , phys .
e. * 54 * , r5897 ( 1996 ) .
m. nauroth and w. kob , phys .
e * 55 * , 657 ( 1997 ) .
f. sciortiono , private communication .
r. schilling ( unpublished ) .
barrat , j .-
roux and j .- p .
hansen , chem .
phys . * 149 * , 197 ( 1990 ) . | a molecular dynamics simulation is performed for a supercooled liquid of rigid diatomic molecules .
the time - dependent self and collective density correlators of the molecular centers of mass are determined and compared with the predictions of the ideal mode coupling theory ( mct ) for simple liquids .
this is done in real as well as in momentum space .
one of the main results is the existence of a unique transition temperature @xmath0 , where the dynamics crosses over from an ergodic to a quasi - nonergodic behavior .
the value for @xmath0 agrees with that found earlier for the orientational dynamics within the error bars . in the @xmath1-regime of mct the factorization of space- and time
dependence is satisfactorily fulfilled for both types of correlations .
the first scaling law of ideal mct holds in the von schweidler regime , only , since the validity of the critical law can not be confirmed , due to a strong interference with the microscopic dynamics . in this first scaling regime a consistent description within ideal mct emerges only , if the next order correction to the asymptotic law is taken into account .
this correction is almost negligible for @xmath2 , the position of the main peak in the static structure factor @xmath3 , but becomes important for @xmath4 , the position of its first minimum .
the second scaling law , i.e. the time - temperature superposition principle , holds reasonably well for the self and collective density correlators and different values for @xmath5 .
the @xmath6-relaxation times @xmath7 and @xmath8 follow a power law in @xmath9 over 2 3 decades .
the corresponding exponent @xmath10 is weakly @xmath5-dependent and is around @xmath11 .
this value is in agreement with the one predicted by mct from the value of the von schweidler exponent but at variance with the corresponding exponent @xmath12 obtained for the orientational correlators @xmath13 and @xmath14 , studied earlier . |
classical cepheids ( cepheids hereafter ) are arguably the most important class of variable stars . from the _ leavitt _ ( _ period - luminosity _ @xmath7 _
p - l _ ) _ law _ , these stars serve as crucial `` standard candles '' for determining the cosmic distance scale and measuring the hubble constant ( @xmath8 ) .
also , cepheids play a fundamental role in the calibration of type ia supernovae , which indicate that the expansion of the universe is accelerating , and also infer the existence of dark energy @xcite . in the near future , with the advent of missions such as _ gaia _ , _ jwst _ and _ wfirst _ , cepheids will be employed to empirically measure the hubble constant to @xmath9 @xcite . at this precision , and combining this directly measured @xmath8 with the results from _ planck _ @xcite and _ wmap _ @xcite and those expected from _ wfirst _ , the study of cepheids will tightly constrain fundamental cosmological parameters .
cepheids are also important astrophysical laboratories for probing the internal structure , stellar atmospheres and evolution of moderate mass ( @xmath10 @xmath11 ) stars ( and type ii sn progenitors , for cepheids above @xmath12 ) .
for example , the study of cepheids provides valuable information about convective core overshooting @xcite , mass loss @xcite , rotation @xcite and helium core burning @xcite for moderate mass , evolved stars .
studies of changes in their pulsation periods and amplitudes reveal evolutionary changes too subtle to detect directly , and understanding these various characteristics of cepheids is crucial to their use as high - precision standard candles . to this end , the _ secret lives of cepheids _ ( slic )
program was initiated : a comprehensive study of cepheid behavior , evolution , pulsations , atmospheres , heating dynamics , shocks and winds .
the program started with the initially limited goal of photometrically monitoring the period and amplitude changes of polaris @xcite , and has since expanded to include 16 bright cepheids , covering a wide range of pulsation properties ( e.g. pulsation periods , amplitudes , modes of pulsation ) .
the program now spans almost the entire electromagnetic spectrum , from recently acquired x - ray observations ( _ xmm - newton _ and _ chandra _ ) , to fuv / uv spectra ( _ hst _ , _ iue _ and _ fuse _ ) , to the ground - based photoelectric photometry we continue to gather , to high - resolution ir spectra ( _ sst _ ) .
taken together , these data reveal cepheids to be complex objects , with surprising levels of activity and behaviors . at optical wavelengths ,
we continue to monitor cepheid light curves for possible signs of real - time stellar evolution , manifested by changes in light curve morphology , average brightness and times of maximum light . for example , in an increasing number of cepheids , the pulsation period is known to change significantly , by as much as 200 sec yr@xmath1 or more @xcite .
our photometry first indicated the recovery of polaris light amplitude ( from its minimum value of @xmath20.02-mag during the 1990s ) in the early 2000s @xcite and subsequent studies have confirmed our initial findings @xcite .
in addition to the evolutionary studies , we are investigating the atmospheric dynamics and heating of cepheids using x - ray uv data ( the emphasis of this paper ) .
@xcite ( and references therein ) have shown variations in the atmospheric turbulence of cepheids due to pulsation - driven shocks , and have discussed the effects ( e.g. pulsation phase - dependent photospheric line broadening ) . during the 1980s and 90s ,
studies such as @xcite , gathered numerous uv spectra of cepheids with the _ international ultraviolet explorer _ ( _ iue _ ) , finding variable uv line emissions likely also due to shocks .
@xcite further calculated that this shock heating could produce x - ray emitting plasmas .
these data , combined with the current study , all show cepheid atmospheres to contain plasmas with temperatures of 10@xmath13
10@xmath14 k @xcite that vary in phase with the pulsations of the cepheids , indicating that the stellar pulsations are responsible for either the very presence , or at least the excitation , of the hot plasma emissions .
the most likely mechanism , as discussed later , is pulsation - driven shocks propagating through the cepheids outer atmospheres .
the detection and characterization of shocks in cepheid atmospheres has become rather important in recent years as they may play a role in cepheid mass loss @xcite .
the long exposures ( @xmath15 50 ksec ) usually required for good x - ray characterizations have slowed progress in that wavelength range , but due to the excellent _ hst cosmic origins spectrograph _ ( _ cos _ ) sensitivity , the uv fuv situation is rather different . typically , very good uv
fuv spectra are obtained in a single _ hst _ orbit .
is an important calibrator for the _ period - luminosity law _ because it has one of the most precisely determined distances ( @xmath16 pc from _ hst fine guidance sensor _
( _ fgs _ ) parallax determination ) for a cepheid @xcite .
this distance agrees well with the revised _ hipparcos _
parallax distance of @xmath17-pc @xcite .
selected properties of are given in table [ tbl1 ] , including two separate masses derived from either stellar pulsation theory or evolution models .
the difference between the two values is a prime example of the _ cepheid mass discrepancy _ , which has existed for decades ( see @xcite ) .
photometry of dates back to its discovery light curve by @xcite .
it is also the second nearest cepheid ( only is nearer ) , and the member of a wide binary system with a hotter , a0v - type companion ( ) at a projected distance of 40
away from the cepheid @xcite .
further , is a member of the cep ob6 star cluster @xcite .
@xcite determined a cluster - derived distance of @xmath18 pc , which agrees very well with distances derived from parallax measures .
recently , a bow shock has also been found around the cepheid , where the stellar wind is colliding with the surrounding interstellar medium , importantly constraining the mass loss rate to 10@xmath1910@xmath20 m@xmath21 yr@xmath1 @xcite .
this was confirmed by @xcite , using _ very large array _ ( _ vla _ ) h i 21 cm line observations to calculate a mass loss rate of ( @xmath22 ) @xmath23 m@xmath21 yr@xmath1 .
typical of many cepheids , displays the `` saw tooth '' light curve , with a quicker rise to maximum brightness , and a much slower decline to minimum brightness , as shown in fig .
because of its proximity ( with respect to other cepheids ) , has been extensively studied at all wavelengths , including earlier uv studies ( e.g. @xcite ) carried with the _ international ultraviolet explorer _ ( _ iue _ ) satellite .
photoelectric _ ubvri _ photometry for this program was carried out with the _ four college automatic photoelectric telescope _ ( _ fcapt _
@xcite ) housed at _
fairborn observatory_. the full data set was split to produce two well - covered light curves , centered at epochs 2008.9 and 2011.4 .
fourier analyses of the resulting light curves yielded two times of maximum light , which were added to those found in the literature @xcite .
our analysis of the complete timings dataset ( fig .
[ fig2 ] ) shows the period to be decreasing over time . from a quadratic fit to the o - c data , the rate of period change is calculated to be dp / dt = @xmath24 sec yr@xmath1 , indicating that is currently making its second crossing of the instability strip @xcite . according to theory , the star contracts and heats up during the second crossing , and its surface temperature rises .
thus , the average density of the star increases , resulting in a decrease in pulsation period , according to the _ period - density relation _
@xmath25 where _ p _ is the pulsation period and @xmath26 is the average stellar density .
residuals of the quadratic o - c fit show no significant , additional ( cyclic or otherwise ) period variability . _
b- _ and _ v_-band light amplitudes of @xmath27- and @xmath28-mag , respectively , were also obtained from the individual _ fcapt _ light curves , and compared to archival observations . from this comparison
, the light amplitude of appears to be increasing over time .
if simple linear trends are assumed , the light amplitude of has been increasing at a rate of @xmath20.011-mag in the _ v_-band and @xmath20.012-mag in the _ b_-band , per decade , over the last @xmath3 years ( total changes of @xmath29 mag in _ v _ and @xmath30 mag in _ b _ fig .
[ fig3 ] ) .
it is interesting to note that many of the early observations of , from the discovery of its variability in the late-1700s through to the early-1900s , display still smaller visual light amplitudes of @xmath31-mag @xcite .
known as a very capable and meticulous observer , @xcite observed with a polarizing photometer and reported a visual amplitude of 0.76-mag ( though a fourier analysis of the data returned an amplitude of 0.73-mag ) , based on observations made in 1906 .
these values would extend the amplitude trend observed in the past @xmath3 years ( fig .
[ fig3 ] ) . however , these visual datasets require modern , careful calibrations before a conclusive determination of continued amplitude change can be achieved .
in addition , the bottom panel of fig .
[ fig3 ] shows that the average brightness of has also increased over the same time span .
constructing a theoretical model from spectroscopic observations , @xcite concluded that cepheids could produce x - ray activity via pulsation - induced shock - heating of their atmospheres .
however , pointed _ einstein
_ and _ rntgen satellite _ ( _ rosat _ ) observations failed to detect any such activity , and indicated that cepheids were not ( at least significant ) x - ray sources . even with _
log l@xmath32 _ @xmath33 29 erg s@xmath1 ( as determined by @xcite ) , the problem with detecting x - rays from cepheids is that these stars ( except for at @xmath34 pc @xcite ) are far away , at @xmath35 pc .
thus , they are expected to have relatively weak x - ray fluxes ( @xmath36 erg s@xmath1 @xmath37 ) .
though was detected , at the @xmath38 level , in a _
rosat_/_high resolution imager _
( _ hri _ ) archival image , the detection was not noticed until several years after the observation was carried out ( see * ? ? ?
confirmation of x - ray activity and definitive detections of x - rays from other nearby cepheids would not occur until the arrival of more powerful x - ray observatories such as _ xmm - newton _ and _
chandra_. we successfully obtained _ chandra _ ( pi : evans ) and _ xmm - newton _ ( pis : guinan & engle ) observations of five cepheids , , , and have been observed so far .
the _ chandra _
data reduction for is discussed in @xcite .
_ xmm - newton _ data were reduced in the usual fashion , making use of the _ science analysis system _ ( _ sas v13.0.0 _ ) routines .
modeling of the _ chandra _ and _ xmm - newton _ data were then carried out using the _ sherpa _ modeling and fitting package [ distributed as part of the _ chandra interactive analysis of observations _ ( _ ciao _ ) _ v4.5 _ suite ] .
_ mekal _ models @xcite were used for the final two - temperature ( 2-t ) fitting and flux calculations .
the nearest three cepheids ( and so far the only cepheids detected ) , & have x - ray luminosities of @xmath39 erg s@xmath1 . neither the short period cepheid ( @xmath40 d ; @xmath41 pc ) nor the long period , luminous cepheid ( @xmath42 d ; @xmath43 pc : @xcite ) were detected .
upper x - ray luminosity limits of @xmath44 and @xmath45 ergs s@xmath1 were estimated for and , respectively , based on exposure times , background count rates and stellar distances .
therefore , it is still possible that and are x - ray sources with similar levels of activity to the cepheids detected thus far , but are too distant to be detected above the background of the _ xmm - newton _ exposures .
however , the failure to detect two of our targets underscores a long - standing ambiguity present in the x - ray studies of cepheids . since cepheids are young stars ( @xmath46 myr ) , any coeval main - sequence g - k - m companions ( if present ) would be coronal x - ray sources with x - ray luminosities similar to that of the cepheids ( see * ? ? ? * and references therein ) .
therefore , caution must be exercised as to whether unresolved companions are producing the x - ray activity . however , the _ hst - cos _ results ( @xcite , and this study ) show plasmas with temperatures up to @xmath47 k that vary ( in phase ) with the cepheids pulsation periods .
this indicates that plasmas approaching soft x - ray emitting temperatures do exist in cepheid atmospheres . as part of this study ,
five separate _
xmm - newton _ visits were carried out for . to provide better pulsation phase coverage , three of the visits were split during reduction into @xmath48 ksec sub - exposures , resulting in 8 individual exposures for analysis .
background - subtracted count rates were @xmath49 counts s@xmath1 for all exposures . a neutral hydrogen ( @xmath50 ) column density value of @xmath51 ( @xmath52 ) @xmath37 was adopted , based on distance relationships determined by @xcite and corroborated by target reddening @xcite .
the relevant two - temperature _ mekal _ model fitted parameters are given in table [ tbl2 ] .
all x - ray energy distributions peak in the @xmath53 kev range ( @xmath54 k [ mk ] ) and have measured flux values in the range of @xmath55 erg s@xmath1 @xmath37 ( x - ray luminosities of @xmath56 erg s@xmath1 ) .
the phased x - ray fluxes are shown in fig .
the cepheid s x - ray activity appears to reach a maximum near @xmath57 .
this is different than what is found for the uv emission lines , which peak in the phase - range @xmath58 ( see fig . [ fig4 ] ) .
the nearby a0-type companion of is also detected in the x - ray data , impying that this star may have a cooler , unresolved companion .
for comparison , the x - ray flux of the companion star remains essentially constant in all exposures , while the x - ray flux of increases by @xmath59 in the exposure near @xmath57 .
although our analysis has ruled out a possible flare or other transient events , this observation still represents a single ( 37 ksec ) sub - exposure . a new ( recently approved by _
chandra _ ) observation at similar phase will further investigate the high flux level .
x - ray observations of ( _ p _ = 9.84 days ) , the subject of a follow - up paper , show a similar flux increase at @xmath57 . possibly related to the x - ray variations reported here for is the recent discovery by @xcite of small , pulsation - phased x - ray variations of the @xmath60 cep - type variable star .
the periodic x - ray variations for ( hd 46328 ) have the same period as the fundamental stellar pulsation of this hot , magnetic b0.5 iv star .
the x - ray fluxes peak near the optical maximum brightness that occurs near the minimum stellar radius and higest temperature .
the relatively strong x - ray emissions observed for @xmath60 cep - type variables as well as other hot , massive stars arise from shock - heated winds ( see * ? ? ? * ) .
@xcite suggest that the periodic x - ray variations arise from small pulsation - induced changes in the wind structure , possibly coupled with changes in the magnetic field .
prior to the installation of _ hst - cos _ , the most thorough uv studies of cepheids were carried out in the 1980s90s with the _ international ultraviolet explorer _ ( _ iue _ ) .
studies of particular note are the series of papers by @xcite involving _ iue _ data ( @xmath61 ) for multiple cepheids , including , and @xcite which focused on the large available _ iue _ dataset for the 35 day cepheid .
these studies found uv emission lines , including o i 1305 , c iv 1550 , and he ii 1640 , to vary with the cepheid pulsation periods evidence that the cepheids ( and not unseen companions ) were responsible for the emissions , and pulsations may be the formation mechanism .
the emissions are indicative of hot plasmas up to @xmath62 k ( also , he ii can be partly formed via photoionization by coronal radiation
@xcite ) .
one of the major aims of this study is to improve upon previous _ iue _ results , since what appeared to be photospheric continuum flux introduced uncertainties in both emission line identifications and flux measures .
_ cos _ , with finer spectral resolution , higher sensitivity and lower noise , can return superior spectra for which much more precise studies could be carried out . in addition , closely - spaced emission lines that were blended together in _ iue _
spectra could be individually resolved in _
cos _ data and perhaps even less - prominent emission lines would be detected . since the only reliable data in the @xmath63 region available for cepheids ( at the beginning of this project ) was low - resolution ( @xmath64 ) _ iue - swp_(_r _ ) spectra , it was quite frankly unknown exactly what the _ cos _ observations would show .
the _ hst - cos _ spectra of used in this study , and measured fluxes , are given in table [ tbl3 ] .
_ iue _ was known to suffer from scattered optical light contamination @xcite , though it turned out to be much more of an issue than was anticipated .
[ fig5 ] shows a comparison of representative _ iue _ and _ cos _ spectra for .
the improvement is dramatic .
the much higher resolution was expected , but as the figure shows , a great deal of what could have been continuum flux from the photosphere turned out to be scattered light .
this was the reason that _ iue _ spectra of the program cepheids could unambiguously show only the strongest emission lines ( if any ) . in the case of
, there was uncertainty as to whether any emission lines were present in the spectra ; there was only a possible detection of the strong , but blended , oxygen / sulfur lines near @xmath65 .
the scattered light is not present in the _ cos _ spectra of cepheids , however , which display a wealth of emission lines not detected in the archival _ iue _ spectra .
these lines define rich and complex cepheid atmospheres and , as is well known , offer excellent atmospheric diagnostic potential ( @xcite and references therein ) , since different line species originate in plasmas of specific temperatures .
also , emission line strengths and ratios , as well as line broadening and radial velocities ( rvs ) , when measured over the stars pulsation cycles , offer important atmospheric diagnostics and can also help distinguish cepheid supergiant atmospheric emissions from those of possible unresolved main sequence companions ( if present ) .
the flux values reported here are integrated fluxes for the emission lines with the continuum flux subtracted . to determine rvs for the emission lines , gaussian fits were made although many important emission lines are present in the _ cos _ spectra , a few of importance were selected for the current study .
these lines were selected because they represent a wide range of formation temperatures , and are relatively free from contamination via blending with nearby lines , offering a `` pure '' measurement of the emission line in question .
the lines selected are : * o i 1358 selected in favor of the well - known o i triplet at @xmath66 because the triplet are fluorescent lines excited by h lyman-@xmath60 radiation @xcite , suffer from blending with ( primarily ) nearby s i lines and heavy airglow contamination .
therefore , the flux of the o triplet is not necessarily indicative of plasmas at the o i peak equilibrium formation temperature of @xmath67 k @xcite . *
si iv 1393/1403 a doublet with a peak formation temperature of @xmath68 k @xcite .
this doublet represents an important link between the cooler , o i emitting plasmas and the hotter plasmas responsible for the n v and o v emission lines . *
n v 1239/1243 another doublet , but with a higher peak formation temperature of @xmath69 k @xcite .
this doublet is very important because the lines occur at shorter wavelengths where photospheric continuum flux is essentially negligible .
the n v doublet is the best measure of higher - temperature plasma variability in the uv spectra of the cepheids . *
additional lines although not selected for a variability study due to its location on the red wing of the lyman-@xmath70 geocoronal - contaminated emission line , the o v 1218 line is present in certain spectra and has a peak formation temperature of @xmath71 k @xcite
. this places it among the hottest lines observable in the uv spectra of cool stars like the cepheids , and its detection provides an important link to higher - temperature x - ray emissions .
additionally , measures have been made of the si iii 1206 and 1298 features , and the 1298/1206 flux ratio was used to carry out atmospheric density estimates .
data accumulation and results in the satellite - based high - energy studies have been much slower than for the optical studies . however , currently has the most complete phase coverage with _ cos_. as an illustration of the results of the uv program thus far , the current _ cos _ light curves of are given in fig .
[ fig4 ] . in looking at these plots ,
the most informative features are : 1 .
the phase at which line emissions begin to increase .
the phase coverage of has been illustrative in terms of the rise in emission flux . in comparing the emission line
fluxes to the included photospheric rv plot for , we see good correlation between the phase where uv line emissions begin to peak and a `` pre - piston phase '' as given by the rvs . during this pre - piston phase
, the photosphere has almost reached its minimum radius , its recession is decelerating , and it is about to begin expanding again . @xcite
have shown that a shock should be propagating through the atmosphere of at this phase ( @xmath72 ) .
the phasing of the flux increase is strong evidence in favor of a shock heating mechanism .
2 . the abruptness of the increase in uv emissions from is indicative of a sudden heating or excitation mechanism . over a span of @xmath73 hours ( from @xmath74 ) , the o i 1358 flux increases by @xmath75 ( numbered points 4 7 in fig . [ fig4 ] ) and
the si iv flux increases by @xmath76 .
such an abrupt , strong excitation also points to a shock - related mechanism .
the decrease in flux is also rapid ( though not as rapid as the increase ) , as one would expect for a sudden heating event such as a shock .
the presence of an apparent quiescent level of uv emissions during @xmath77 .
this indicates a persistent atmospheric heating , possibly magnetic in origin , as is the case for numerous other classes of cool stars .
another feature of note is the phase - difference between the peak flux of the most energetic ( highest peak formation temperature ) emission feature observed
n v @xmath21240 and the peak fluxes of the two cooler emission features .
unfortunately , for , this aspect of the program requires a number of _ very _ narrowly phase - spaced observations that we simply do not have .
thus , a strict , quantitative conclusion can not yet be drawn .
however , the lower temperature plasma emission lines do appear to peak earlier than n v. in fig .
[ fig4 ] , spectrum 7 is clearly the most active spectrum in terms of o i 1358 and si iv 1393 emissions . qualitatively speaking , spectrum 7 looks to represent the peak phase of emissions for these two features .
however , for n v emissions , spectrum 7 has nearly the same flux as spectrum 8 .
since the overall shapes of the emission curves for these three features are so similar , it can be concluded that n v emissions are still rising in spectrum 7 , where o i and si iv emissions have essentially peaked .
thus , n v must have a peak emission slightly later ( @xmath78 ) in phase than o i and si iv .
as discussed by @xcite , emissions from the hottest plasmas are expected to peak first in the case of shock - heating , followed by line emissions from cooling plasmas in the post - shock regions . however , it s worth noting that the similarity of the rise times in uv line fluxes may result from the plasma exciting line emissions of various temperatures as it begins ionizing through to higher ionization states , and the process continues as the shock front moves through more neutral and singly ionized plasma .
this is an interesting behavior , but one that will require more in - depth modeling and analysis before firm conclusions are drawn .
in addition to what is learned from emission line fluxes about the heating mechanism(s ) at work in cepheid atmospheres , the emission line profiles and rvs can provide valuable , complementary information .
[ fig6 ] gives the profiles of the o i 1358 and si iv 1393 emission lines observed in several spectra of .
the most potentially informative characteristic is the asymmetry present in spectra 68 , where the lines show a strong , additional blue - shifted emission component .
this component is likely caused by an expanding shock emerging from the cepheid photosphere .
on the `` near '' side of the cepheid atmosphere , the shock is approaching , producing the blue - shifted emission . in spectrum 8 ,
the o i line still shows heavy asymmetry , but the si iv line shows an extremely broad and even emission profile , indicative of a large velocity distribution but no additional blue - shifted feature . at this phase ( @xmath79 )
we are likely observing si iv emission from a very turbulent post - shock region , where the high turbulence is responsible for the velocity distribution , meaning that the shock has `` passed by '' the si iv emitting region .
the difference in o i and si iv line profiles at this phase indicates that they are likely originating from different regions ( heights ) within the cepheid atmosphere .
the line profiles offer up further evidence in favor of shock - heating and compression being responsible for the enhanced emissions .
the rvs of the uv emission lines can give information on the workings of the atmosphere , although we note there are _ cos _ wavelength calibration issues affecting their absolute accuracy @xcite . as such , the velocities can show a larger than normal uncertainty , but the agreement in overall velocity trends between the three lines plotted gives confidence in the measures . in fig .
[ fig7 ] , the rvs ( from top to bottom panels ) of the o i , si iv , n v emission lines and photosphere are plotted . as indicated in the figure
, the emission line rvs have had the phase - specific photospheric rv removed .
for the three emission line rv plots , the dashed gray horizontal line indicates zero velocity : when the line emitting region and the photosphere have identical velocities .
thus , line rvs above this line indicate that the line emitting region is compressing on to the photosphere , and line rvs below this line indicate the region is expanding away from the photosphere . in the bottom ( photospheric rv )
plot , the dashed gray horizontal line represents the average velocity of the star . for spectra where the si iv and/or o i line showed asymmetry ,
two gaussian profiles were fit to the line , and the rv of the broad atmospheric emission is plotted , as opposed to the blue - shifted emission component discussed in the previous paragraph .
the agreement between the si iv and o i velocity behaviors and that of the n v line , which maintained a symmetric single - gaussian profile throughout the observed phases , gives confidence in the double - gaussian approach .
[ fig7 ] shows that , from @xmath80 ( spectra 48 in fig .
[ fig4 ] ) , the atmosphere of is compressing on top of the photosphere .
this compression begins just as the cepheid photosphere is starting to expand again ( the piston phase ) .
this compression would result in further plasma excitation at these phases , contributing to the increased emission lines fluxes ( fig .
[ fig4 ] ) and line broadening ( fig . [ fig6 ] ) that are observed .
attempts have also been made to find suitable electron density - sensitive emission line ratios , to gain further physical insights into the cepheid s atmosphere .
a handful of well - studied ratios exist in the literature , making use of such emission lines as , e.g. c iii ( 1909 ) , si iii ( 1892 ) and o iv ( @xmath21400 ) . unfortunately
, either the spectra available for do not cover the wavelengths of these emission lines ( c iii 1909 or si iii 1892 ) , or the lines are not strong enough to allow an unambiguous measurement ( o iv @xmath21400 ) . thus , none of these density measures could successfully be applied . making use of the latest chianti atomic database ( http://www.chiantidatabase.org/ ) available at the time of writing ( @xmath81 ) ,
investigations were carried out using a ratio of si iii line fluxes ( 1298/1206 ) , as mentioned previously .
measurements of almost all spectra for give densities below the diagnostic range of the si iii 1298/1206 ratio ( @xmath82 @xmath83 ) .
spectrum 6 of ( as numbered in figs .
[ fig4 ] , [ fig6 ] ) , where the flux is steeply rising , gives a much higher density of @xmath84 @xmath83 . for reference , at similar plasma temperatures to those probed by the si iii ratio above , quiet regions of the sun have measured densities of @xmath85 @xmath83 and active regions have densities of @xmath86 @xmath83 @xcite , and other studies ( * ? ? ?
* e.g. ) have also found solar densities to match that measured for spectrum 6 of .
however , issues have been raised with using the si iii 1206 line in density diagnostics .
@xcite calculated that , for the sun , the 1206 line would be far too optically thick to give an accurate density . although we believe that , in a supergiant such as , the line would essentially be optically thin and suitable in that regard , there is also the temperature sensitivity to take into account when using a ground state transition , such as si iii 1206 , with other subordinate features from higher levels . as such
, we view the result as confirmation of increased atmospheric density during the phase of rising flux , but are still investigating the usefulness of the si iii ratio in returning a precise , numerical density measure .
all previously - mentioned features of the spectra support the conclusion that s atmospheric plasmas are heated / excited via pulsation - driven compression and shock propagation .
the prototype cepheid is undergoing some very interesting and complex behaviors .
optical photometry confirms that the pulsation period of the cepheid is steadily decreasing over time at a rate of dp / dt = @xmath24 sec yr@xmath1 .
the decrease in the pulsation period indicates the star s radius is slowly decreasing as it contracts and evolves toward the blue ( hotter ) side of the h - r diagram .
also , standardized _ bv _ photoelectric photometry ( from 19582012 ) provide evidence that the light amplitude and mean brightness are increasing . the observed changes in the light amplitude and average brightness of the star may also arise from slow evolutionary changes , indicating that we are witnessing stellar evolution in `` real time . ''
multiple x - ray observations and uv spectra have been taken of with _ xmm - newton _ and _ hst - cos_. from these data , it is clear that possesses an outer atmosphere of heated plasmas in the temperature range @xmath87 k. soft x - ray emissions are observed ( @xmath88 erg s@xmath1 @xmath37 ) with evidence of pulsation phase - dependent variability .
this variability will need to be confirmed through an approved ( _ chandra _ ) , deep exposure taken at a similar phase .
if the increased x - ray activity near @xmath89 is confirmed through additional data , the large difference between the phases of enhanced uv and x - ray emission will imply that perhaps competing ( and at times , complementary ) plasma - excitation mechanisms are at play in the cepheid atmosphere .
the denser phasing of the _ hst - cos _ uv spectra have provided a clearer picture of the @xmath90 k plasmas . though a full astrophysical modeling of the data is underway and will have to be the topic of a future uv - dedicated paper , initial results from the spectra give a clear , general picture of the atmosphere .
uv emission line fluxes vary by a factor of @xmath91 , depending on line species and plasma temperature ( see fig . [ fig4 ] ) , and the rise in emission line flux begins very near the phase of minimum stellar radius , just before the photosphere begins expanding again .
this is the cepheid `` piston '' phase , when a shock is expected to emerge from the photosphere and begin propagating through the atmosphere . shortly after this phase ,
as the flux is increasing , the line profiles become asymmetric due to a blue - shifted emission feature ( fig .
[ fig6 ] ) .
this blue - shifted feature is caused by the additional excitation of plasma at the expanding shock front . during the phases of enhanced activity
, the line profiles also noticeably broaden .
this is a combination of the emergence of the blue - shifted feature , in addition to growing emissions from turbulent plasma in the post - shock regions and the overall compression of the atmosphere .
this compression is seen in fig .
[ fig7 ] during @xmath92 : while the photospheric recession slows , eventually turning to outward expansion , the atmosphere ( as traced by emission line rvs ) continues to recede , compressing around the photosphere .
the phasing and strength of the flux increase , the line asymmetry and broadening of the line profiles and the photospheric vs. atmospheric rv behaviors all confirm that the atmospheric plasmas around undergo regular excitation from the cepheid s radial pulsations via atmospheric compression and shock propagation .
the persistence of a quiescent atmosphere , outside of the shock - excited phase - range , indicates that a second heating mechanism , such as a stellar magnetic field , is also present in .
the combined x - ray
uv optical study of , carried out under the slic program , shows the inherent complexity of this prototype cepheid that underscores our ( still incomplete ) understanding of this important class of variable stars .
further , analyses of the long - term optical photometry provide strong evidence that relatively rapid evolutionary changes are taking place .
this study demonstrates the high value of historical photometric data , in this case dating back to the time of @xcite .
our study of time - series observations , from x - ray to optical wavelengths , are changing our perspective of these stars and have provided an initial step to understand not only cepheid structures but also the details of their evolution and dynamic photospheres .
we plan to continue high precision photometry to track the period change and also ascertain the observed small changes in the star s light amplitude and luminosity .
the _ xmm - newton _ x - ray and _ hst - cos _ fuv spectrophotometry confirm the presence of variable , hot @xmath93 k@xmath94 k plasmas in the outer atmosphere of this pulsating cool supergiant , most likely arising from pulsation - driven shocks .
the similarity of the strengths of uv emissions ( as well as x - ray luminosities ) to some non - pulsating `` hybrid '' f k giants and supergiants ( * ? ? ?
* and references therein ) indicate that magnetic fields could also play a role in the origin of these high - energy emission lines .
although the presence of phase - dependent fuv line emissions was expected from prior _ iue _ studies of cepheids , the x - ray emissions ( especially the possible _ _ l__@xmath32 variability ) is very surprising and is being followed - up .
the pulsation shock heating of the atmosphere of ( and , by extension , of other cepheids ) has yet undetermined effects on the energetics , heating , structure and dynamics of cepheid atmospheres that could effect the _
p - l law_. as discussed previously , if cepheids are to be used to significantly improve the cosmic distance scale and precisely determine the hubble constant ( _ _ h__@xmath95 ) to an uncertainty of @xmath21% , it will be necessary to have a clearer understanding of the effects of the pulsation - induced shocks on cepheid atmospheres .
the deposition of shock - induced energies in the outer atmosphere could affect the luminosity and pulsations of the stars .
as additional targets are analyzed , the slic program will address these problems with a larger sample of stars .
in forthcoming papers we will present the results for additional representative cepheids that include , , and . finally , we will continue to carry out ground - based photometry and to request additional space - based observations to help unlock the secret lives of these stars .
the authors wish to thank the tireless efforts of lou boyd , _ fairborn observatory _
director , in maintaining the _
fcapt_. the authors also wish to thank the financial support of nsf grant ast05 - 07542 , and nasa grants hst - go11726 , hst - go12302 , hst - go13019 and xmm - newton : nasa nnx14af12 g .
support for _ hst _ programs 11726 , 12302 and 13019 was provided by nasa through grants from the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas 5 - 26555 .
_ facilities : _ , , .
lc spectral type & f5ib
g1ib@xmath96 + @xmath97 ( k ) & @xmath98 + mass ( pulsational ) ( @xmath11 ) & @xmath99 + mass ( evolutionary ) ( @xmath11 ) & @xmath100 + mean luminosity ( @xmath101 ) & @xmath102 + mean radius ( @xmath103 ) & @xmath104 + distance ( pc ) & @xmath105 + _ @xmath106v@xmath107_-mag & @xmath108 * + _ v_-band amplitude & @xmath28 * + + + + + cccccccccccc & 2006/01/19 18:04 & 2006/01/20 12:37 & 0.054 & 0.12 & & 0.254 & 1.00 & 6.070 & 3.000 & 5.382e+28 & 28.73 + & 2455217.253 & 2455218.026 & & & & 0.704 & 0.63 & & & & + & & & & & & & & & & & + & 2006/01/21 18:05 & 2006/01/22 14:17 & 0.43 & 0.51 & & 0.618 & 0.29 & 15.040 & 2.000 & 1.334e+29 & 29.13 + & 2455219.254 & 2455220.095 & & & & 2.115 & 1.00 & & & & + & & & & & & & & & & & + & & & 0.51 & 0.58 & & 0.353 & 0.48 & 8.172 & 2.000 & 7.246e+28 & 28.86 + & & & & & & 1.345 & 1.00 & & & & + & & & & & & & & & & & + & 2004/06/4 14:26 & 2004/06/4 21:53 & 0.332 & 0.39 & & 0.995 & 1.00 & 5.130 & 1.500 & 4.549e+28 & 28.66 + & 2454623.101 & 2454623.412 & & & 3.5e+20 & 0.401 & 0.64 & & & & + & & & & & & & & & & & + & 2013/06/28 6:34 & 2013/06/29 13:49 & 0.84 & 0.96 & & 0.321 & 0.97 & 4.340 & 1.000 & 3.848e+28 & 28.59 + & 2456471.774 & 2456473.076 & & & & 1.328 & 1.00 & & & & + & & & & & & & & & & & + & & & 0.96 & 0.08 & & 0.408 & 1.00 & 4.350 & 2.200 & 3.857e+28 & 28.59 + & & & & & & 1.378 & 0.65 & & & & + & & & & & & & & & & & + & 2013/07/2 6:17 & 2013/07/3 7:51 & 0.58 & 0.68 & & 0.613 & 1.00 & 3.710 & 0.700 & 3.290e+28 & 28.52 + & 2456475.762 & 2456476.827 & & & & 0.741 & 0.01 & & & & + & & & & & & & & & & & + & & & 0.68 & 0.78 & & 0.361 & 1.00 & 5.944 & 1.500 & 5.270e+28 & 28.72 + & & & & & & 0.956 & 0.42 & & & & + ccccccccccccc & 2010/10/19 & 2010/10/19 & 0.610 & 924.992 & g130 m & 1291 & 1.379e-15 & 3.203e-16 & 1.085e-15 & 1.885e-16 & 1.765e-15 & 4.519e-16 + & 00:12 & 00:30 & & & & & & & & & & + & 2010/10/19 & 2010/10/19 & & 923.968 & g160 m & 1589 & & & & & & + & 01:30 & 01:48 & & & & & & & & & & + & 2010/12/12 & 2010/12/12 & 0.715 & 924.992 & g130 m & 1291 & 1.687e-15 & 3.675e-16 & 1.163e-15 & 2.658e-16 & 2.239e-15 & 4.311e-16 + & 06:12 & 06:29 & & & & & & & & & & + & 2010/12/12 & 2010/12/12 & & 923.904 & g160 m & 1589 & & & & & & + & 06:34 & 06:52 & & & & & & & & & & + & 2010/10/30 & 2010/10/30 & 0.782 & 924.96 & g130 m & 1291 & 1.482e-15 & 3.015e-16 & 1.198e-15 & 2.405e-16 & 1.979e-15 & 4.326e-16 + & 16:24 & 16:41 & & & & & & & & & & + & 2010/10/30 & 2010/10/30 & & 923.968 & g160 m & 1589 & & & & & & + & 16:54 & 17:11 & & & & & & & & & & + & 2010/10/31 & 2010/10/31 & 0.872 & 925.024 & g130 m & 1291 & 1.680e-15 & 2.978e-16 & 3.068e-15 & 3.159e-16 & 3.674e-15 & 6.304e-16 + & 04:04 & 04:22 & & & & & & & & & & + & 2010/10/31 & 2010/10/31 & & 924.032 & g160 m & 1589 & & & & & & + & 04:26 & 04:44 & & & & & & & & & & + & 2010/12/13 & 2010/12/14 & 0.041 & 925.024 & g130 m & 1291 & 8.000e-15 & 4.657e-16 & 4.461e-15 & 4.089e-16 & 1.831e-14 & 1.142e-15 + & 23:42 & 00:00 & & & & & & & & & & + & 2010/12/14 & 2010/12/14 & & 924.032 & g160 m & 1589 & & & & & & + & 00:04 & 01:13 & & & & & & & & & & + & 2010/10/22 & 2010/10/22 & 0.180 & 925.024 & g130 m & 1291 & 4.063e-15 & 3.777e-16 & 1.681e-15 & 2.540e-16 & 4.906e-15 & 5.904e-16 + & 01:41 & 01:59 & & & & & & & & & & + & 2010/10/22 & 2010/10/22 & & 924.032 & g160 m & 1589 & & & & & & + & 02:56 & 03:14 & & & & & & & & & & + & 2010/10/31 & 2010/10/31 & 0.860 & 925.024 & g130 m & 1291 & 1.848e-15 & 3.016e-16 & 1.703e-15 & 2.452e-16 & 3.063e-15 & 5.278e-16 + & 02:28 & 02:46 & & & & & & & & & & + & 2010/10/31 & 2010/10/31 & & 924.032 & g160 m & 1589 & & & & & & + & 02:50 & 03:08 & & & & & & & & & & + & 2010/10/29 & 2010/10/29 & 0.647 & 924.928 & g130 m & 1291 & 1.492e-15 & 3.303e-16 & 1.198e-15 & 2.363e-16 & 2.088e-15 & 4.451e-16 + & 22:59 & 23:28 & & & & & & & & & & + & 2010/10/29 & 2010/10/29 & & 923.936 & g160 m & 1589 & & & & & & + & 23:32 & 23:50 & & & & & & & & & & + & 2011/06/13 & 2011/06/13 & 0.901 & 1152.032 & g130 m & 1291 & 2.845e-15 & 3.200e-16 & 6.053e-15 & 3.722e-16 & 1.197e-14 & 6.869e-16 + & 16:50 & 17:12 & & & & & & & & & & + & 2011/06/13 & 2011/06/13 & & 1024.032 & g160 m & 1589 & & & & & & + & 17:17 & 17:37 & & & & & & & & & & + & 2013/01/18 & 2013/01/18 & 0.933 & 767.008 & g130 m & 1291 & 7.496e-15 & 6.732e-16 & 1.099e-14 & 7.579e-16 & 3.460e-14 & 1.696e-15 + & 18:29 & 18:49 & & & & & & & & & & + & 2013/01/18 & 2013/01/18 & & 763.072 & g160 m & 1589 & & & & & & + & 18:52 & 20:12 & & & & & & & & & & +
the _ ubvri _ photoelectric data for , carried out with the _ fcapt _ for this study , are given in table [ tbl4 ] .
the data were obtained between december 2007 and january 2012 .
lrrrlrrrlrrrlrrrlrrr 2454437.5956 & -0.2357 & 4.3136 & 0.0026 & 2454437.5926 & -0.2362 & 0.6518 & 0.0060 & 2454437.5941 & -0.2360 & 0.8588 & 0.0062 & 2454623.9116 & 0.4846 & 0.8475 & 0.0133 & 2454623.9118 & 0.4846 & 0.4449 & 0.0182 + 2454439.5928 & 0.1365 & 3.6690 & 0.0056 & 2454439.5941 & 0.1367 & 0.3887 & 0.0216 & 2454439.5928 & 0.1365 & 0.5839 & 0.0104 & 2454624.9074 & -0.3299 & 0.8906 & 0.0133 & 2454624.9076 & -0.3298 & 0.4568 & 0.0213 + 2454440.5858 & 0.3215 & 3.9369 & 0.0015 & 2454440.5856 & 0.3215 & 0.5173 & 0.0081 & 2454440.5857 & 0.3215 & 0.7411 & 0.0058 & 2454625.9097 & -0.1431 & 0.8115 & 0.0306 & 2454625.9099 & -0.1431 & 0.3999 & 0.0298 + 2454447.5794 & -0.3752 & 4.2614 & 0.0029 & 2454447.5806 & -0.3750 & 0.6512 & 0.0043 & 2454447.5793 & -0.3752 & 0.8752 & 0.0035 & 2454626.9008 & 0.0416 & 0.6316 & 0.1148 & 2454626.9010 & 0.0416 & 0.3114 & 0.1183 + 2454449.5797 & -0.0024 & 3.4627 & 0.0206 & 2454449.5795 & -0.0025 & 0.3408 & 0.0429 & 2454449.5797 & -0.0024 & 0.4522 & 0.0270 & 2454628.8955 & 0.4133 & 0.8242 & 0.0368 & 2454628.8957 & 0.4134 & 0.4223 & 0.0388 + 2454450.5797 & 0.1839 & 3.7435 & 0.0051 & 2454450.5795 & 0.1839 & 0.4299 & 0.0150 & 2454450.5797 & 0.1839 & 0.6358 & 0.0064 & 2454630.8917 & -0.2147 & 0.8686 & 0.0130 & 2454630.8919 & -0.2147 & 0.4471 & 0.0157 + 2454452.6001 & -0.4396 & 4.1927 & 0.0008 & 2454452.6000 & -0.4396 & 0.6393 & 0.0033 & 2454452.6001 & -0.4396 & 0.8575 & 0.0028 & 2454631.8928 & -0.0281 & 0.6187 & 0.0376 & 2454631.8930 & -0.0281 & 0.3017 & 0.0322 + 2454454.5902 & -0.0687 & 3.6357 & 0.0080 & 2454454.5914 & -0.0685 & 0.3344 & 0.0223 & 2454454.5901 & -0.0687 & 0.5130 & 0.0192 & 2454634.8827 & -0.4710 & 0.8436 & 0.0104 & 2454634.8829 & -0.4709 & 0.4573 & 0.0132 + 2454458.6295 & -0.3160 & 4.3210 & 0.0038 & 2454458.6293 & -0.3160 & 0.6698 & 0.0066 & 2454458.6294 & -0.3160 & 0.8739 & 0.0075 & 2454636.8755 & -0.0996 & 0.7132 & 0.0159 & 2454636.8757 & -0.0996 & 0.3449 & 0.0187 + 2454459.6007 & -0.1350 & 4.0087 & 0.0061 & 2454459.6005 & -0.1350 & 0.4212 & 0.0188 & 2454459.6007 & -0.1350 & 0.6829 & 0.0087 & 2454637.9467 & 0.1000 & 0.6852 & 0.0240 & 2454637.9469 & 0.1001 & 0.3425 & 0.0154 + 2454460.6012 & 0.0514 & 3.5319 & 0.0044 & 2454460.6010 & 0.0514 & 0.3427 & 0.0241 & 2454460.6011 & 0.0514 & 0.4928 & 0.0098 & 2454638.8722 & 0.2725 & 0.7897 & 0.0339 & 2454638.8724 & 0.2725 & 0.4029 & 0.0381 + 2454465.5856 & -0.0197 & 3.4726 & 0.0097 & 2454465.5854 & -0.0197 & 0.3114 & 0.0393 & 2454465.5855 & -0.0197 & 0.4432 & 0.0161 & 2454640.8665 & -0.3559 & 0.8592 & 0.0332 & 2454640.8667 & -0.3558 & 0.4639 & 0.0394 + 2454466.5860 & 0.1667 & 3.7179 & 0.0030 & 2454466.5858 & 0.1667 & 0.4135 & 0.0243 & 2454466.5860 & 0.1667 & 0.6152 & 0.0090 & 2455098.6397 & -0.0492 & 0.6392 & 0.0147 & 2455098.6399 & -0.0492 & 0.3083 & 0.0179 + 2454623.9115 & 0.4845 & 4.0934 & 0.0069 & 2454623.9125 & 0.4847 & 0.5818 & 0.0075 & 2454623.9114 & 0.4845 & 0.8290 & 0.0102 & 2455098.8062 & -0.0182 & 0.6168 & 0.0112 & 2455098.8064 & -0.0182 & 0.2922 & 0.0067 + 2454624.9073 & -0.3299 & 4.2869 & 0.0101 & 2454624.9096 & -0.3295 & 0.6965 & 0.0045 & 2454624.9072 & -0.3299 & 0.8948 & 0.0110 & 2455099.6797 & 0.1446 & 0.7060 & 0.0267 & 2455099.6799 & 0.1446 & 0.3626 & 0.0325 + 2454625.9096 & -0.1431 & 4.0377 & 0.0132 & 2454625.9119 & -0.1427 & 0.4678 & 0.0181 & 2454625.9095 & -0.1431 & 0.6995 & 0.0203 & 2455099.8099 & 0.1688 & 0.7150 & 0.0091 & 2455099.8101 & 0.1689 & 0.3659 & 0.0079 + 2454626.9007 & 0.0416 & 3.5028 & 0.0736 & 2454626.9005 & 0.0415 & 0.3168 & 0.1136 & 2454626.9006 & 0.0416 & 0.4855 & 0.1044 & 2455100.6798 & 0.3309 & 0.7976 & 0.0147 & 2455100.6800 & 0.3310 & 0.4048 & 0.0113 + 2454628.8954 & 0.4133 & 4.0524 & 0.0255 & 2454628.8952 & 0.4133 & 0.5968 & 0.0100 & 2454628.8953 & 0.4133 & 0.7961 & 0.0272 & 2455100.8050 & 0.3543 & 0.7972 & 0.0098 & 2455100.8052 & 0.3543 & 0.4187 & 0.0084 + 2454630.8915 & -0.2147 & 4.2666 & 0.0064 & 2454630.8914 & -0.2147 & 0.6220 & 0.0067 & 2454630.8915 & -0.2147 & 0.8435 & 0.0091 & 2455101.8010 & -0.4601 & 0.8341 & 0.0063 & 2455101.8012 & -0.4601 & 0.4423 & 0.0075 + 2454631.8927 & -0.0282 & 3.4735 & 0.0338 & 2454631.8925 & -0.0282 & 0.3129 & 0.0547 & 2454631.8926 & -0.0282 & 0.4578 & 0.0472 & 2455102.6711 & -0.2980 & 0.8716 & 0.0120 & 2455102.6713 & -0.2979 & 0.4458 & 0.0148 + 2454634.8826 & -0.4710 & 4.1342 & 0.0059 & 2454634.8837 & -0.4708 & 0.6428 & 0.0085 & 2454634.8825 & -0.4710 & 0.8499 & 0.0103 & 2455102.7988 & -0.2742 & 0.8655 & 0.0043 & 2455102.7989 & -0.2742 & 0.4518 & 0.0037 + 2454636.8754 & -0.0996 & 3.8125 & 0.0075 & 2454636.8752 & -0.0997 & 0.3607 & 0.0195 & 2454636.8753 & -0.0996 & 0.5974 & 0.0164 & 2455106.7865 & 0.4689 & 0.8297 & 0.0105 & 2455106.7867 & 0.4690 & 0.4382 & 0.0095 + 2454637.9466 & 0.1000 & 3.5949 & 0.0221 & 2454637.9464 & 0.1000 & 0.3873 & 0.0198 & 2454637.9465 & 0.1000 & 0.5456 & 0.0281 & 2455131.7560 & 0.1220 & 0.6839 & 0.0093 & 2455131.7562 & 0.1221 & 0.3430 & 0.0087 + 2454638.8721 & 0.2725 & 3.8548 & 0.0283 & 2454638.8719 & 0.2724 & 0.4832 & 0.0229 & 2454638.8720 & 0.2724 & 0.6989 & 0.0339 & 2455133.7728 & 0.4979 & 0.8295 & 0.0112 & 2455133.7730 & 0.4979 & 0.4465 & 0.0096 + 2454640.8664 & -0.3559 & 4.2744 & 0.0198 & 2454640.8712 & -0.3550 & 0.6568 & 0.0127 & 2454640.8688 & -0.3554 & 0.8617 & 0.0228 & 2455134.7811 & -0.3142 & 0.8777 & 0.0105 & 2455134.7813 & -0.3142 & 0.4600 & 0.0121 + 2455098.6395 & -0.0493 & 3.5439 & 0.0031 & 2455098.6394 & -0.0493 & 0.3202 & 0.0300 & 2455098.6395 & -0.0493 & 0.4703 & 0.0115 & 2455135.7812 & -0.1279 & 0.7493 & 0.0051 & 2455135.7814 & -0.1278 & 0.3802 & 0.0063 + 2455098.8061 & -0.0182 & 3.4698 & 0.0098 & 2455098.8059 & -0.0183 & 0.3162 & 0.0282 & 2455098.8060 & -0.0182 & 0.4468 & 0.0147 & 2455136.7803 & 0.0583 & 0.6346 & 0.0162 & 2455136.7805 & 0.0584 & 0.3161 & 0.0164 + 2455099.6796 & 0.1445 & 3.6712 & 0.0174 & 2455099.6794 & 0.1445 & 0.3933 & 0.0283 & 2455099.6795 & 0.1445 & 0.5873 & 0.0251 & 2455143.7568 & 0.3584 & 0.7916 & 0.0094 & 2455143.7570 & 0.3584 & 0.4198 & 0.0095 + 2455099.8098 & 0.1688 & 3.7231 & 0.0083 & 2455099.8096 & 0.1688 & 0.3957 & 0.0177 & 2455099.8097 & 0.1688 & 0.6079 & 0.0111 & 2455144.7440 & -0.4576 & 0.8426 & 0.0040 & 2455144.7442 & -0.4576 & 0.4436 & 0.0044 + 2455100.6797 & 0.3309 & 3.9503 & 0.0098 & 2455100.6795 & 0.3309 & 0.4961 & 0.0188 & 2455100.6796 & 0.3309 & 0.7481 & 0.0099 & 2455145.7525 & -0.2697 & 0.8528 & 0.0057 & 2455145.7527 & -0.2697 & 0.4529 & 0.0047 + 2455100.8049 & 0.3542 & 3.9894 & 0.0057 & 2455100.8047 & 0.3542 & 0.5408 & 0.0110 & 2455100.8049 & 0.3543 & 0.7592 & 0.0064 & 2455146.7368 & -0.0863 & 0.6751 & 0.0133 & 2455146.7370 & -0.0862 & 0.3376 & 0.0091 + 2455101.8009 & -0.4601 & 4.1662 & 0.0051 & 2455101.8007 & -0.4602 & 0.6152 & 0.0068 & 2455101.8008 & -0.4602 & 0.8436 & 0.0056 & 2455151.7215 & -0.1574 & 0.7938 & 0.0089 & 2455151.7217 & -0.1573 & 0.4094 & 0.0075 + 2455102.6697 & -0.2982 & 4.3076 & 0.0037 & 2455102.6695 & -0.2983 & 0.6399 & 0.0111 & 2455102.6696 & -0.2983 & 0.8748 & 0.0098 & 2455468.8626 & -0.0577 & 0.6324 & 0.0083 & 2455468.8628 & -0.0577 & 0.3048 & 0.0098 + 2455102.7986 & -0.2742 & 4.3196 & 0.0036 & 2455102.7984 & -0.2743 & 0.6272 & 0.0052 & 2455102.7986 & -0.2742 & 0.8799 & 0.0044 & 2455469.8525 & 0.1268 & 0.6848 & 0.0109 & 2455469.8527 & 0.1268 & 0.3427 & 0.0078 + 2455106.7864 & 0.4689 & 4.0959 & 0.0071 & 2455106.7862 & 0.4689 & 0.5972 & 0.0116 & 2455106.7863 & 0.4689 & 0.8225 & 0.0090 & 2455470.8497 & 0.3126 & 0.7778 & 0.0119 & 2455470.8499 & 0.3126 & 0.4063 & 0.0100 + 2455131.7559 & 0.1220 & 3.6400 & 0.0047 & 2455131.7557 & 0.1220 & 0.3726 & 0.0261 & 2455131.7558 & 0.1220 & 0.5682 & 0.0131 & 2455471.8479 & 0.4986 & 0.8250 & 0.0085 & 2455471.8481 & 0.4987 & 0.4364 & 0.0092 + 2455133.7727 & 0.4978 & 4.1171 & 0.0079 & 2455133.7725 & 0.4978 & 0.6057 & 0.0094 & 2455133.7726 & 0.4978 & 0.8312 & 0.0087 & 2455476.8329 & 0.4276 & 0.8102 & 0.0108 & 2455476.8331 & 0.4276 & 0.4323 & 0.0127 + 2455134.7810 & -0.3143 & 4.3066 & 0.0058 & 2455134.7808 & -0.3143 & 0.6630 & 0.0104 & 2455134.7810 & -0.3143 & 0.8879 & 0.0103 & 2455477.8318 & -0.3863 & 0.8584 & 0.0070 & 2455477.8320 & -0.3862 & 0.4547 & 0.0083 + 2455135.7811 & -0.1279 & 3.9710 & 0.0016 & 2455135.7809 & -0.1279 & 0.4214 & 0.0131 & 2455135.7810 & -0.1279 & 0.6597 & 0.0042 & 2455478.8279 & -0.2007 & 0.8286 & 0.0038 & 2455478.8281 & -0.2006 & 0.4406 & 0.0048 + 2455136.7802 & 0.0583 & 3.5396 & 0.0107 & 2455136.7800 & 0.0583 & 0.3263 & 0.0237 & 2455136.7801 & 0.0583 & 0.5041 & 0.0139 & 2455479.8261 & -0.0146 & 0.6120 & 0.0090 & 2455479.8263 & -0.0146 & 0.2867 & 0.0101 + 2455143.7567 & 0.3584 & 3.9866 & 0.0033 & 2455143.7565 & 0.3583 & 0.5265 & 0.0077 & 2455143.7566 & 0.3584 & 0.7666 & 0.0068 & 2455480.8225 & 0.1710 & 0.7142 & 0.0151 & 2455480.8227 & 0.1711 & 0.3588 & 0.0147 + 2455144.7438 & -0.4577 & 4.1803 & 0.0026 & 2455144.7449 & -0.4575 & 0.6064 & 0.0024 & 2455144.7438 & -0.4577 & 0.8490 & 0.0032 & 2455481.8207 & 0.3571 & 0.7908 & 0.0090 & 2455481.8209 & 0.3571 & 0.4168 & 0.0099 + 2455145.7524 & -0.2697 & 4.3230 & 0.0037 & 2455145.7535 & -0.2695 & 0.6619 & 0.0052 & 2455145.7523 & -0.2697 & 0.8749 & 0.0049 & 2455482.8168 & -0.4573 & 0.8476 & 0.0022 & 2455482.8170 & -0.4573 & 0.4427 & 0.0055 + 2455146.7367 & -0.0863 & 3.7385 & 0.0118 & 2455146.7378 & -0.0861 & 0.3430 & 0.0261 & 2455146.7366 & -0.0863 & 0.5586 & 0.0177 & 2455488.8023 & -0.3419 & 0.8648 & 0.0057 & 2455488.8025 & -0.3419 & 0.4536 & 0.0050 + 2455151.7214 & -0.1574 & 4.1278 & 0.0068 & 2455151.7212 & -0.1574 & 0.4719 & 0.0093 & 2455151.7214 & -0.1574 & 0.7403 & 0.0095 & 2455492.8072 & 0.4044 & 0.8116 & 0.0069 & 2455492.8074 & 0.4044 & 0.4269 & 0.0051 + 2455468.8625 & -0.0577 & 3.5880 & 0.0058 & 2455468.8623 & -0.0578 & 0.2636 & 0.0332 & 2455468.8624 & -0.0577 & 0.4836 & 0.0086 & 2455493.7962 & -0.4113 & 0.8532 & 0.0073 & 2455493.7964 & -0.4113 & 0.4506 & 0.0078 + 2455469.8524 & 0.1267 & 3.6395 & 0.0091 & 2455469.8522 & 0.1267 & 0.3709 & 0.0277 & 2455469.8523 & 0.1267 & 0.5750 & 0.0101 & 2455495.7965 & -0.0385 & 0.6257 & 0.0089 & 2455495.7967 & -0.0385 & 0.2863 & 0.0159 + 2455470.8496 & 0.3126 & 3.9216 & 0.0090 & 2455470.8506 & 0.3128 & 0.4550 & 0.0090 & 2455470.8495 & 0.3126 & 0.7337 & 0.0107 & 2455498.7914 & -0.4804 & 0.8334 & 0.0042 & 2455498.7916 & -0.4804 & 0.4434 & 0.0021 + 2455471.8477 & 0.4986 & 4.1325 & 0.0074 & 2455471.8476 & 0.4986 & 0.5523 & 0.0085 & 2455471.8477 & 0.4986 & 0.8203 & 0.0075 & 2455499.7862 & -0.2950 & 0.8576 & 0.0082 & 2455499.7864 & -0.2950 & 0.4483 & 0.0096 + 2455476.8328 & 0.4275 & 4.0543 & 0.0066 & 2455476.8326 & 0.4275 & 0.5565 & 0.0078 & 2455476.8328 & 0.4276 & 0.7966 & 0.0079 & 2455501.7760 & 0.0758 & 0.6490 & 0.0141 & 2455501.7761 & 0.0758 & 0.3168 & 0.0154 + 2455477.8317 & -0.3863 & 4.2534 & 0.0037 & 2455477.8315 & -0.3863 & 0.6369 & 0.0067 & 2455477.8317 & -0.3863 & 0.8640 & 0.0050 & 2455502.7737 & 0.2617 & 0.7618 & 0.0091 & 2455502.7739 & 0.2617 & 0.3878 & 0.0057 + 2455478.8278 & -0.2007 & 4.2579 & 0.0035 & 2455478.8288 & -0.2005 & 0.5306 & 0.0030 & 2455478.8277 & -0.2007 & 0.8274 & 0.0043 & 2455503.7705 & 0.4474 & 0.8195 & 0.0057 & 2455503.7707 & 0.4475 & 0.4345 & 0.0047 + 2455479.8259 & -0.0147 & 3.4598 & 0.0058 & 2455479.8258 & -0.0147 & 0.2990 & 0.0374 & 2455479.8259 & -0.0147 & 0.4438 & 0.0124 & 2455504.7678 & -0.3667 & 0.8593 & 0.0076 & 2455504.7679 & -0.3667 & 0.4492 & 0.0054 + 2455480.8224 & 0.1710 & 3.7180 & 0.0116 & 2455480.8222 & 0.1710 & 0.3776 & 0.0215 & 2455480.8223 & 0.1710 & 0.6115 & 0.0154 & 2455506.7644 & 0.0054 & 0.6110 & 0.0125 & 2455506.7633 & 0.0051 & 0.2903 & 0.0178 + 2455481.8206 & 0.3570 & 3.9785 & 0.0029 & 2455481.8204 & 0.3570 & 0.5003 & 0.0195 & 2455481.8205 & 0.3570 & 0.7636 & 0.0049 & 2455508.7577 & 0.3768 & 0.8131 & 0.0051 & 2455508.7579 & 0.3768 & 0.4182 & 0.0051 + 2455482.8167 & -0.4573 & 4.1735 & 0.0006 & 2455482.8165 & -0.4574 & 0.6105 & 0.0096 & 2455482.8166 & -0.4574 & 0.8388 & 0.0063 & 2455510.7524 & -0.2515 & 0.8705 & 0.0049 & 2455510.7526 & -0.2514 & 0.4505 & 0.0052 + 2455488.8022 & -0.3419 & 4.2813 & 0.0050 & 2455488.8033 & -0.3417 & 0.6661 & 0.0053 & 2455488.8034 & -0.3417 & 0.8781 & 0.0051 & 2455511.7500 & -0.0656 & 0.6559 & 0.0093 & 2455511.7502 & -0.0655 & 0.3113 & 0.0094 + 2455492.8071 & 0.4044 & 4.0341 & 0.0058 & 2455492.8069 & 0.4044 & 0.5646 & 0.0185 & 2455492.8070 & 0.4044 & 0.7843 & 0.0097 & 2455512.7464 & 0.1201 & 0.6888 & 0.0080 & 2455512.7466 & 0.1201 & 0.3422 & 0.0100 + 2455493.7961 & -0.4113 & 4.2234 & 0.0036 & 2455493.7959 & -0.4113 & 0.6438 & 0.0093 & 2455493.7960 & -0.4113 & 0.8642 & 0.0054 & 2455513.7467 & 0.3065 & 0.7853 & 0.0058 & 2455513.7469 & 0.3065 & 0.4056 & 0.0047 + 2455495.7963 & -0.0386 & 3.5080 & 0.0040 & 2455495.7962 & -0.0386 & 0.3188 & 0.0224 & 2455495.7963 & -0.0386 & 0.4529 & 0.0059 & 2455515.7440 & -0.3213 & 0.8583 & 0.0084 & 2455515.7442 & -0.3213 & 0.4594 & 0.0066 + 2455498.7913 & -0.4805 & 4.1397 & 0.0038 & 2455498.7911 & -0.4805 & 0.5898 & 0.0089 & 2455498.7913 & -0.4804 & 0.8403 & 0.0063 & 2455516.7383 & -0.1360 & 0.7643 & 0.0068 & 2455516.7384 & -0.1360 & 0.3856 & 0.0085 + 2455499.7861 & -0.2951 & 4.3109 & 0.0060 & 2455499.7872 & -0.2949 & 0.6147 & 0.0060 & 2455499.7860 & -0.2951 & 0.8810 & 0.0075 & 2455532.6063 & -0.1790 & 0.8205 & 0.0040 & 2455532.6065 & -0.1789 & 0.4245 & 0.0019 + 2455501.7758 & 0.0757 & 3.5584 & 0.0083 & 2455502.7734 & 0.2616 & 0.4184 & 0.0107 & 2455501.7758 & 0.0757 & 0.5080 & 0.0131 & 2455532.6854 & -0.1642 & 0.8072 & 0.0039 & 2455532.6856 & -0.1642 & 0.4113 & 0.0024 + 2455502.7736 & 0.2617 & 3.8492 & 0.0084 & 2455503.7702 & 0.4474 & 0.5650 & 0.0081 & 2455502.7735 & 0.2616 & 0.6962 & 0.0085 & 2455533.5941 & 0.0051 & 0.6192 & 0.0055 & 2455533.5943 & 0.0051 & 0.2939 & 0.0052 + 2455503.7704 & 0.4474 & 4.0703 & 0.0041 & 2455504.7687 & -0.3666 & 0.6450 & 0.0064 & 2455503.7703 & 0.4474 & 0.8109 & 0.0073 & 2455862.6718 & 0.3292 & 0.7970 & 0.0059 & 2455862.6720 & 0.3292 & 0.4151 & 0.0039 + 2455504.7676 & -0.3668 & 4.2629 & 0.0068 & 2455506.7628 & 0.0051 & 0.3137 & 0.0261 & 2455504.7676 & -0.3668 & 0.8764 & 0.0087 & 2455865.6113 & -0.1231 & 0.7487 & 0.0054 & 2455865.6114 & -0.1230 & 0.3780 & 0.0073 + 2455506.7656 & 0.0056 & 3.4644 & 0.0101 & 2455508.7574 & 0.3767 & 0.4885 & 0.0094 & 2455506.7642 & 0.0053 & 0.4493 & 0.0161 & 2455866.5963 & 0.0605 & 0.6474 & 0.0064 & 2455866.5964 & 0.0605 & 0.3171 & 0.0048 + 2455508.7576 & 0.3768 & 4.0027 & 0.0015 & 2455510.7521 & -0.2515 & 0.6259 & 0.0038 & 2455508.7575 & 0.3768 & 0.7787 & 0.0063 & 2455867.5955 & 0.2467 & 0.7607 & 0.0056 & 2455867.5957 & 0.2467 & 0.3860 & 0.0058 + 2455510.7523 & -0.2515 & 4.3192 & 0.0031 & 2455511.7497 & -0.0656 & 0.3313 & 0.0237 & 2455510.7522 & -0.2515 & 0.8726 & 0.0042 & 2455868.5951 & 0.4330 & 0.8311 & 0.0035 & 2455868.5952 & 0.4330 & 0.4313 & 0.0053 + 2455511.7498 & -0.0656 & 3.6148 & 0.0078 & 2455512.7461 & 0.1200 & 0.3340 & 0.0193 & 2455511.7498 & -0.0656 & 0.5041 & 0.0247 & 2455881.5652 & -0.1500 & 0.7904 & 0.0045 & 2455881.5653 & -0.1500 & 0.4069 & 0.0046 + 2455512.7463 & 0.1201 & 3.6349 & 0.0040 & 2455513.7464 & 0.3065 & 0.4628 & 0.0073 & 2455512.7462 & 0.1201 & 0.5566 & 0.0100 & 2455881.5800 & -0.1473 & 0.7815 & 0.0030 & 2455881.5802 & -0.1472 & 0.4031 & 0.0031 + 2455513.7465 & 0.3065 & 3.9130 & 0.0042 & 2455515.7437 & -0.3213 & 0.6444 & 0.0078 & 2455513.7465 & 0.3065 & 0.7305 & 0.0070 & 2455881.5949 & -0.1445 & 0.7834 & 0.0067 & 2455881.5950 & -0.1445 & 0.3968 & 0.0064 + 2455515.7439 & -0.3213 & 4.2962 & 0.0067 & 2455516.7379 & -0.1361 & 0.3991 & 0.0162 & 2455515.7438 & -0.3213 & 0.8734 & 0.0084 & 2455881.6098 & -0.1417 & 0.7768 & 0.0053 & 2455881.6099 & -0.1417 & 0.3948 & 0.0038 + 2455516.7381 & -0.1360 & 4.0154 & 0.0058 & 2455532.6060 & -0.1790 & 0.5079 & 0.0084 & 2455516.7381 & -0.1360 & 0.6895 & 0.0087 & 2455881.6246 & -0.1390 & 0.7713 & 0.0041 & 2455881.6248 & -0.1389 & 0.3922 & 0.0049 + 2455532.6062 & -0.1790 & 4.1985 & 0.0037 & 2455532.6851 & -0.1643 & 0.4582 & 0.0140 & 2455532.6061 & -0.1790 & 0.7854 & 0.0039 & 2455881.6396 & -0.1362 & 0.7630 & 0.0063 & 2455881.6397 & -0.1361 & 0.3914 & 0.0066 + 2455532.6853 & -0.1643 & 4.1480 & 0.0034 & 2455533.5938 & 0.0050 & 0.3076 & 0.0636 & 2455532.6852 & -0.1643 & 0.7549 & 0.0081 & 2455881.6542 & -0.1334 & 0.7657 & 0.0067 & 2455881.6543 & -0.1334 & 0.3871 & 0.0052 + 2455533.5940 & 0.0051 & 3.4631 & 0.0044 & 2455862.6716 & 0.3291 & 0.5273 & 0.0073 & 2455533.5939 & 0.0051 & 0.4426 & 0.0056 & 2455881.6688 & -0.1307 & 0.7597 & 0.0032 & 2455881.6689 & -0.1307 & 0.3864 & 0.0047 + 2455862.6718 & 0.3292 & 3.9454 & 0.0048 & 2455865.6121 & -0.1229 & 0.4116 & 0.0161 & 2455862.6717 & 0.3291 & 0.7361 & 0.0058 & 2455881.6834 & -0.1280 & 0.7515 & 0.0079 & 2455881.6835 & -0.1280 & 0.3776 & 0.0049 + 2455865.6113 & -0.1231 & 3.9479 & 0.0047 & 2455866.5960 & 0.0604 & 0.3642 & 0.0151 & 2455865.6112 & -0.1231 & 0.6473 & 0.0085 & 2455881.6981 & -0.1253 & 0.7551 & 0.0035 & 2455881.6982 & -0.1252 & 0.3733 & 0.0042 + 2455866.5962 & 0.0605 & 3.5297 & 0.0046 & 2455867.5953 & 0.2467 & 0.4690 & 0.0202 & 2455866.5962 & 0.0605 & 0.4819 & 0.0056 & 2455881.7128 & -0.1225 & 0.7531 & 0.0084 & 2455881.7129 & -0.1225 & 0.3776 & 0.0074 + 2455867.5955 & 0.2467 & 3.8237 & 0.0048 & 2455868.5949 & 0.4329 & 0.5862 & 0.0089 & 2455867.5954 & 0.2467 & 0.6749 & 0.0078 & 2455881.7274 & -0.1198 & 0.7378 & 0.0052 & 2455881.7276 & -0.1198 & 0.3727 & 0.0050 + 2455868.5950 & 0.4330 & 4.0575 & 0.0014 & 2455881.5649 & -0.1501 & 0.4587 & 0.0050 & 2455868.5950 & 0.4330 & 0.8013 & 0.0030 & 2455882.7387 & 0.0687 & 0.6529 & 0.0120 & 2455882.7389 & 0.0687 & 0.3096 & 0.0122 + 2455881.5651 & -0.1500 & 4.0841 & 0.0023 & 2455881.5798 & -0.1473 & 0.4551 & 0.0141 & 2455881.5651 & -0.1500 & 0.7125 & 0.0035 & 2455888.7209 & 0.1835 & 0.7202 & 0.0049 & 2455888.7210 & 0.1835 & 0.3609 & 0.0049 + 2455881.5800 & -0.1473 & 4.0662 & 0.0026 & 2455881.5946 & -0.1445 & 0.4413 & 0.0064 & 2455881.5799 & -0.1473 & 0.7114 & 0.0043 & 2455892.6969 & -0.0756 & 0.6567 & 0.0080 & 2455892.6971 & -0.0756 & 0.3329 & 0.0099 + 2455881.5948 & -0.1445 & 4.0545 & 0.0033 & 2455881.6095 & -0.1418 & 0.4362 & 0.0135 & 2455881.5947 & -0.1445 & 0.7006 & 0.0052 & 2455893.7074 & 0.1127 & 0.6821 & 0.0075 & 2455893.7075 & 0.1127 & 0.3441 & 0.0047 + 2455881.6097 & -0.1417 & 4.0402 & 0.0042 & 2455881.6244 & -0.1390 & 0.4287 & 0.0077 & 2455881.6096 & -0.1417 & 0.6987 & 0.0069 & 2455900.6875 & 0.4134 & 0.8189 & 0.0075 & 2455900.6877 & 0.4135 & 0.4316 & 0.0063 + 2455881.6246 & -0.1390 & 4.0255 & 0.0036 & 2455881.6393 & -0.1362 & 0.4173 & 0.0093 & 2455881.6245 & -0.1390 & 0.6893 & 0.0044 & 2455911.6610 & 0.4584 & 0.8345 & 0.0057 & 2455911.6611 & 0.4584 & 0.4278 & 0.0082 + 2455881.6395 & -0.1362 & 4.0084 & 0.0047 & 2455881.6539 & -0.1335 & 0.4129 & 0.0020 & 2455881.6394 & -0.1362 & 0.6861 & 0.0073 & 2455912.6203 & -0.3629 & 0.8583 & 0.0074 & 2455912.6204 & -0.3628 & 0.4601 & 0.0050 + 2455881.6541 & -0.1335 & 3.9959 & 0.0056 & 2455881.6685 & -0.1308 & 0.4136 & 0.0092 & 2455881.6541 & -0.1335 & 0.6806 & 0.0058 & 2455919.5649 & -0.0687 & 0.6601 & 0.0076 & 2455919.5650 & -0.0687 & 0.3246 & 0.0104 + 2455881.6687 & -0.1307 & 3.9808 & 0.0021 & 2455881.6832 & -0.1280 & 0.4107 & 0.0057 & 2455881.6687 & -0.1307 & 0.6726 & 0.0056 & 2455919.5806 & -0.0658 & 0.6615 & 0.0101 & 2455919.5796 & -0.0660 & 0.3160 & 0.0112 + 2455881.6833 & -0.1280 & 3.9622 & 0.0076 & 2455881.6978 & -0.1253 & 0.4064 & 0.0085 & 2455881.6833 & -0.1280 & 0.6657 & 0.0087 & 2455919.5947 & -0.0632 & 0.6539 & 0.0117 & 2455919.5948 & -0.0632 & 0.3128 & 0.0101 + 2455881.6980 & -0.1253 & 3.9496 & 0.0025 & 2455881.7125 & -0.1226 & 0.3926 & 0.0164 & 2455881.6980 & -0.1253 & 0.6482 & 0.0061 & 2455919.6069 & -0.0609 & 0.6510 & 0.0114 & 2455919.6070 & -0.0609 & 0.3162 & 0.0140 + 2455881.7127 & -0.1225 & 3.9351 & 0.0048 & 2455881.7272 & -0.1198 & 0.3862 & 0.0083 & 2455881.7127 & -0.1225 & 0.6463 & 0.0061 & 2455919.6215 & -0.0582 & 0.6572 & 0.0092 & 2455919.6217 & -0.0581 & 0.2988 & 0.0087 + 2455881.7274 & -0.1198 & 3.9171 & 0.0030 & 2455882.7385 & 0.0686 & 0.3482 & 0.0101 & 2455881.7273 & -0.1198 & 0.6407 & 0.0066 & 2455922.6360 & -0.4964 & 0.8478 & 0.0065 & 2455922.6361 & -0.4964 & 0.4338 & 0.0039 + 2455882.7387 & 0.0687 & 3.5467 & 0.0059 & 2455888.7206 & 0.1834 & 0.4150 & 0.0201 & 2455882.7386 & 0.0686 & 0.5016 & 0.0065 & 2455926.6212 & 0.2462 & 0.7505 & 0.0086 & 2455926.6214 & 0.2463 & 0.3917 & 0.0088 + 2455888.7209 & 0.1834 & 3.7327 & 0.0025 & 2455892.6967 & -0.0757 & 0.3329 & 0.0126 & 2455888.7208 & 0.1834 & 0.6285 & 0.0045 & 2455927.6224 & 0.4328 & 0.8066 & 0.0098 & 2455927.6225 & 0.4328 & 0.4227 & 0.0108 + 2455892.6969 & -0.0756 & 3.6782 & 0.0051 & 2455893.7072 & 0.1127 & 0.3696 & 0.0104 & 2455892.6968 & -0.0756 & 0.5418 & 0.0128 & 2455931.6091 & 0.1757 & 0.7170 & 0.0067 & 2455931.6092 & 0.1757 & 0.3588 & 0.0071 + 2455893.7074 & 0.1127 & 3.6233 & 0.0062 & 2455900.6873 & 0.4134 & 0.5663 & 0.0066 & 2455893.7073 & 0.1127 & 0.5509 & 0.0072 & & & & & & & & + 2455900.6875 & 0.4134 & 4.0396 & 0.0048 & 2455911.6608 & 0.4583 & 0.5911 & 0.0063 & 2455900.6874 & 0.4134 & 0.7927 & 0.0068 & & & & & & & & + 2455911.6610 & 0.4584 & 4.0806 & 0.0023 & 2455912.6201 & -0.3629 & 0.6577 & 0.0056 & 2455911.6609 & 0.4584 & 0.8116 & 0.0054 & & & & & & & & + 2455912.6203 & -0.3629 & 4.2621 & 0.0061 & 2455919.5647 & -0.0688 & 0.3442 & 0.0165 & 2455912.6202 & -0.3629 & 0.8832 & 0.0077 & & & & & & & & + 2455919.5649 & -0.0687 & 3.6346 & 0.0062 & 2455919.5793 & -0.0660 & 0.3317 & 0.0183 & 2455919.5648 & -0.0687 & 0.5022 & 0.0125 & & & & & & & & + 2455919.5816 & -0.0656 & 3.6206 & 0.0063 & 2455919.5944 & -0.0632 & 0.3344 & 0.0118 & 2455919.5805 & -0.0658 & 0.5072 & 0.0098 & & & & & & & & + 2455919.5946 & -0.0632 & 3.6030 & 0.0111 & 2455919.6066 & -0.0610 & 0.3326 & 0.0126 & 2455919.5946 & -0.0632 & 0.4904 & 0.0152 & & & & & & & & + 2455919.6068 & -0.0609 & 3.5938 & 0.0103 & 2455919.6213 & -0.0582 & 0.3053 & 0.0177 & 2455919.6067 & -0.0609 & 0.4928 & 0.0126 & & & & & & & & + 2455919.6215 & -0.0582 & 3.5778 & 0.0056 & 2455922.6358 & -0.4965 & 0.5951 & 0.0057 & 2455919.6214 & -0.0582 & 0.4911 & 0.0056 & & & & & & & & + 2455922.6360 & -0.4964 & 4.1270 & 0.0054 & 2455926.6210 & 0.2462 & 0.4516 & 0.0113 & 2455922.6359 & -0.4964 & 0.8262 & 0.0065 & & & & & & & & + 2455926.6211 & 0.2462 & 3.8209 & 0.0048 & 2455927.6222 & 0.4328 & 0.5737 & 0.0143 & 2455926.6211 & 0.2462 & 0.6892 & 0.0083 & & & & & & & & + 2455927.6223 & 0.4328 & 4.0375 & 0.0060 & 2455931.6089 & 0.1757 & 0.3997 & 0.0097 & 2455927.6223 & 0.4328 & 0.8152 & 0.0090 & & & & & & & & + 2455931.6091 & 0.1757 & 3.7166 & 0.0021 & & & & & 2455931.6090 & 0.1757 & 0.6286 & 0.0046 & & & & & & & & + davis , j. j. , tracey , j. c. , engle , s. g. , & guinan , e. f. 2002 , bulletin of the american astronomical society , 34 , # 117.11 de zeeuw , p. t. , hoogerwerf , r. , de bruijne , j. h. j. , brown , a. g. a. , & blaauw , a. 1999 , , 117 , 354 drake , s. a. , singh , k. p. , white , n. e. , mewe , r. , & kaastra , j. s. 1996 , cool stars , stellar systems , and the sun , 109 , 263 doyle , j. g. , oshea , e. , erdelyi , r. , et al .
1997 , , 173 , 243 engle , s. g. , & guinan , e. f. 2012 , journal of astronomy and space sciences , 29 , 181 engle , s. g. , guinan , e. f. , & koch , r. h. 2004 , bulletin of the american astronomical society , 36 , 744 evans , n. r. , guinan , e. , engle , s. , et al .
2010 , , 139 , 1968 evans , n. r. , schaefer , g. , bond , h. e. , et al .
2007 , iau symposium , 240 , 102 fokin , a. b. , gillet , d. , & breitfellner , m. g. 1996 , , 307 , 503 goodricke , j. , & bayer , j. 1786 , royal society of london philosophical transactions series i , 76 , 48 guinan , e. f. , & engle , s. g. 2009 , iau symposium , 258 , 395 kiss , l. l. 1998 , , 297 , 825 koncewicz , r. , & jordan , c. 2007 , , 374 , 220 linsky , j. l. , wood , b. e. , judge , p. , et al . 1995 , , 442 , 381 majaess , d. , turner , d. , & gieren , w. 2012 , , 747 , 145 marengo , m. , evans , n. r. , matthews , l. d. , et al . 2013 ,
stellar pulsations : impact of new instrumentation and new insights , 31 , 99 matthews , l. d. , marengo , m. , evans , n. r. , & bono , g. 2012 , , 744 , 53 neilson , h. r. , langer , n. , engle , s. g. , guinan , e. , & izzard , r. 2012 , , 760 , l18 neilson , h. r. , cantiello , m. , & langer , n. 2011 , , 529 , l9 pagano , i. , linsky , j. l. , carkner , l. , et al .
2000 , , 532 , 497 prugniel , p. , soubiran , c. , koleva , m. , & le borgne , d. 2007 , vizier online data catalog , 3251 , 0 sasselov , d. d. , & lester , j. b. 1994 , , 423 , 795 schmidt , e. g. , & parsons , s. b. 1984 , , 279 , 202 schmidt , e. g. , & parsons , s. b. 1984 , , 279 , 215 schmidt , e. g. , & parsons , s. b. 1982 , , 48 , 185 spreckley , s. a. , & stevens , i. r. 2008 , , 388 , 1239 turner , d. g. , & berdnikov , l. n. 2004 , , 423 , 335 turner , d. g. , bryukhanov , i. s. , balyuk , i. i. , et al .
2007 , , 119 , 1247 van leeuwen , f. , feast , m. w. , whitelock , p. a. , & laney , c. d. 2007 , , 379 , 723 | over the past decade , the _ secret lives of cepheids _ ( slic ) program has been carried out at villanova university to study aspects and behaviors of classical cepheids that are still not well - understood . in this , the first of several planned papers on program cepheids , we report the current results for , the cepheid prototype .
ongoing photometry has been obtained to search for changes in the pulsation period , light curve morphology and amplitude . combining our photometry with the times of maximum light compilation by @xcite returns a small period change of dp / dt @xmath0 sec yr@xmath1 .
there is also evidence for a gradual light amplitude increase of @xmath20.011-mag ( _ v_-band ) and @xmath20.012-mag ( _ b_-band ) per decade over the last @xmath3 years .
in addition , _ hst - cos _ uv spectrophotometry and _ xmm - newton _ x - ray data were carried out to investigate the high - temperature plasmas present above the cepheid photospheres . in total , from the five visits ( eight exposures ) with _ xmm - newton _ , is found to be a soft x - ray source ( @xmath4 [ 0.32 kev ] @xmath5 erg sec@xmath1 ) with peak flux at @xmath6 = 0.60.9 kev .
the x - ray activity is found to vary , possibly in phase with the stellar pulsations . from 20102013 , nine observations of were carried out with _ hst - cos_. the uv emissions are also variable , and well - phased with the stellar pulsations .
maximum uv line emissions occur near , or slightly before , maximum optical light , varying by as much as twenty times .
this variability shows that pulsation - induced shock - heating plays a significant role in cepheid atmospheres , possibly in addition to a quiescent , magnetic heating .
the results of this study show cepheid atmospheres to be rather complex and dynamic . |
multicarrier communications have recently attracted much attention in wireless and mobile applications .
the orthogonal frequency division multiplexing ( ofdm ) has been employed as a multiplexing and a multiple access technique in a variety of wireless communication standards such as ieee802.11 wireless lan @xcite , ieee802.16 mobile wimax @xcite , and 3gpp - lte @xcite .
also , the multicarrier code - division multiple access ( mc - cdma ) , a combined scheme of ofdm and cdma @xcite@xmath0@xcite , has been proposed to enjoy the benefits of ofdm and cdma by allocating the spread data symbols to subcarriers .
the popularity of multicarrier communications is mainly due to the robustness to multipath fading channels and the efficient hardware implementation employing fast fourier transform ( fft ) techniques .
however , multicarrier communications have the major drawback of the high peak - to - average power ratio ( papr ) of transmitted signals , which may nullify all the potential benefits @xcite .
a number of techniques have been developed for papr reduction of ofdm signals .
in particular , a constructive and theoretical approach is to employ a coding scheme @xcite@xcite that provides low papr and good error correction capability for transmitted ofdm signals .
the golay complementary sequences @xcite , which belong to a coset of the first - order reed - muller code @xcite , are a good example of the coding scheme .
paterson @xcite also discussed several coding schemes for papr reduction of _ multicode _ cdma . in @xcite
, he summarized the algebraic coding approaches for peak power control in ofdm and multicode cdma . for a summary of the other papr reduction techniques for ofdm ,
see @xcite . to reduce the papr of multicarrier cdma ( mc - cdma ) signals , on the other hand , numerous studies have been focused on the power characteristics of spreading sequences .
ochiai and imai @xcite presented statistical results of the papr in downlink mc - cdma , where multiple users are supported by walsh - hadamard or golay complementary spreading sequences . considering a single user mc - cdma , popovi @xcite presented the basic criteria for the selection of spreading sequences by studying the crest factors ( cf ) @xmath1 of various binary and polyphase sequences .
similar studies can be found in @xcite with multiple access interference ( mai ) minimization . in mc - cdma supporting multiple users or code channels , the crest factors of various spreading sequences have been compared in @xcite and @xcite , where the walsh - hadamard spreading sequences showed the best papr properties , provided that a large number of spreading sequences are combined for the transmitted mc - cdma signals .
more studies can be found in @xcite@xmath0@xcite on the papr of various spreading sequences in mc - cdma .
if mc - cdma assigns multiple spreading sequences to a single user , the multicode mc - cdma can be equivalently treated as the _ spread ofdm _ ( s - ofdm )
@xcite@xmath0@xcite , where a data symbol of the user is spread across a set of subcarriers to enjoy frequency diversity .
if the number of used spreading sequences is large , the walsh - hadamard spread ofdm can be viewed as a papr reducing scheme @xcite@xcite , compared to a conventional ofdm .
also , an error correction code may be applied prior to walsh - hadamard spreading for improving the error rate performance @xcite@xcite or controlling the peak power @xcite of s - ofdm . to the best of our knowledge ,
most of the efforts on papr reduction of mc - cdma and s - ofdm have been verified mainly by statistical experiments , not by thorough theoretical analysis . through the experiments ,
the papr of the multicarrier signals has been statistically observed , but it has never been addressed whether it is theoretically bounded . in this paper , we propose a binary reed - muller coded mc - cdma system and study its papr properties . in the coded mc - cdma , the information data multiplexed from users is encoded by a reed - muller subcode and the codeword is then fully - loaded to walsh - hadamard spreading sequences .
the coding scheme plays a role of reducing the papr of transmitted mc - cdma signals as well as providing the error correction capability .
we first establish the polynomial representation of a coded mc - cdma signal for theoretical analysis of the papr . a recursive construction of boolean functions is then presented for the reed - muller subcodes , where the papr of the mc - cdma signal encoded by the subcode is proven to be theoretically bounded .
the author of @xcite pointed out that the construction is equivalent to type - iii sequences in @xcite where he made a general and mathematical study for boolean functions with bounded papr , not considering the application to mc - cdma .
we also discuss a connection between the code rate of the subcode and the maximum papr .
simulation results show that the papr of the coded mc - cdma signal is not only theoretically bounded , but also statistically reduced .
in particular , the coded mc - cdma solves the major papr problem of uncoded mc - cdma by dramatically reducing its papr for the small number of users . in conclusion
, the reed - muller codes can be effectively utilized for peak power control in mc - cdma with small and moderate numbers of users , subcarriers , and spreading factors .
the papr properties of the coded mc - cdma equivalently address those of the reed - muller coded s - ofdm which supports multiple data from a single user .
throughout this paper , mc - cdma abbreviates _ multicarrier _ cdma @xmath0 not multicode cdma .
this paper discusses a coded mc - cdma system employing binary codewords , binary spreading sequences , and bpsk modulation .
hence , we focus our description on binary cases .
the following notations will be used throughout this paper .
* @xmath2 is a spreading factor or spreading sequence length . *
@xmath3 and @xmath4 are the actual and the maximum numbers of users supported by a coded mc - cdma system , respectively , where @xmath5 . in the rest of this paper
, we will use the context of @xmath3 _ users _ , where the @xmath3 users can be treated as @xmath3 data bits of a single user in s - ofdm or multicode mc - cdma .
* @xmath6 is the codeword length of a @xmath7 code , where @xmath8 .
* @xmath9 is the number of information bits that each user transmits in an ofdm symbol .
* @xmath10 denotes the @xmath9-bit information of the @xmath11th user , @xmath12 , while @xmath13 denotes the @xmath4-bit uncoded data multiplexed from @xmath3 users and zero - tailed at the @xmath14th spreading process , @xmath15 .
note @xmath16 .
* @xmath17 denotes the coded output of @xmath18 by a @xmath7 code at the @xmath14th spreading process .
note @xmath19 for @xmath20 .
* @xmath21 denotes the bpsk modulation output of @xmath22 that experiences the @xmath14th spreading process .
hence , @xmath23 . *
@xmath24 denotes the @xmath2-chip spreading sequence assigned for the @xmath25th coded bit of @xmath26 , while @xmath27 is a set of the @xmath28th spreading chips across all @xmath6 spreading sequences , where @xmath29 .
@xmath30 is a @xmath31 orthogonal spreading matrix with @xmath32 , where @xmath33 is the @xmath25th row vector and @xmath34 is the @xmath28th column vector .
* @xmath35 denotes the output data of length @xmath2 of the @xmath14th spreading process .
figure [ fig : coded_sys ] illustrates a coded mc - cdma transmitter proposed in this paper .
assume that @xmath3 users access to the coded mc - cdma system , where the @xmath11th user , @xmath12 , is actively transmitting the @xmath9-bit information @xmath36 over an ofdm symbol .
the @xmath14th information bit of @xmath36 is multiplexed across all @xmath11 s and then if @xmath5 , zeros are attached to form the @xmath4-bit uncoded data @xmath18 , @xmath15 .
then , @xmath18 is encoded by a @xmath7 code to generate the @xmath6-bit codeword @xmath22 and its bpsk modulation @xmath26 . the @xmath25th coded bit of @xmath26
is then spread by the @xmath2-bit spreading sequence @xmath33 , @xmath20 , where a pair of the spreading sequences is mutually orthogonal . at the @xmath14th spreading process
, the spread bits of length @xmath2 are linearly combined over @xmath6 spreading sequences to produce @xmath37 , where each element of @xmath37 can take an arbitrary value .
obviously , the spreading process is equivalent to a _ transform _ of @xmath26 by the orthogonal spreading matrix @xmath30 , i.e. , @xmath38 , where @xmath33 is the @xmath25th row vector of @xmath30 .
the @xmath9 blocks of the spread data @xmath37 of length @xmath2 experience an @xmath39 block interleaver for frequency diversity , and total @xmath40 bits are allocated to @xmath40 subcarriers by inverse fft ( ifft ) .
the mc - cdma receiver accomplishes the reverse operation to recover the original information @xmath36 for the @xmath11th user , where the despreading process is equivalent to a transform by @xmath41 , the transpose of @xmath30 . from figure
[ fig : coded_sys ] , the baseband transmission signal over an ofdm symbol duration @xmath42 is given by @xmath43 where @xmath44 .
note that if the zero - tail processes and the @xmath7 encoders are removed from figure [ fig : coded_sys ] , then @xmath45 with @xmath46 is equivalent to a conventional uncoded mc - cdma signal in @xcite .
the normalization factor @xmath47 is used in ( [ eq : s_t ] ) to ensure that the average power of @xmath45 is equal to that of the uncoded mc - cdma signal for @xmath3 users , which will be shown in section ii - d .
let @xmath48 be a binary vector where @xmath49 , @xmath50 .
a _ boolean function _
@xmath51 is defined by @xmath52 where @xmath53 and @xmath54 is obtained by a binary representation of @xmath55 , @xmath56 .
note that the addition in a boolean function is computed modulo-@xmath57 . in ( [ eq : gbool ] )
, the order of the @xmath11th monomial with nonzero @xmath58 is given by @xmath59 , and the highest order of the monomials with nonzero @xmath58 s is called the _ ( algebraic ) degree _ of the boolean function @xmath60 .
associated with a boolean function @xmath60 , a binary codeword of length @xmath61 is defined by @xmath62 where @xmath63 . in other words ,
the _ associated _ codeword @xmath64 of length @xmath61 is obtained by the boolean function @xmath65 while @xmath66 runs through @xmath67 to @xmath68 in the increasing order .
the _ @xmath69th - order reed - muller code _
@xmath70 is defined by a set of binary codewords of length @xmath61 where each codeword is generated by a boolean function of degree at most @xmath69 . in other words ,
each codeword in @xmath70 is the associated codeword @xmath64 of length @xmath61 in ( [ eq : assoc_gen ] ) where the boolean function @xmath60 has the degree of at most @xmath69 .
the @xmath69th - order reed - muller code @xmath71 has the dimension of @xmath72 and the minimum hamming distance of @xmath73 . for more details on boolean functions and reed - muller codes ,
see @xcite .
the _ walsh - hadamard matrix _ is recursively constructed by @xmath74 $ ] and @xmath75 then , it is easy to see that @xmath76 where @xmath77 denotes a transpose and @xmath78 the @xmath79 identity matrix .
( [ eq : wh_prop ] ) shows that the walsh - hadamard matrix is symmetric and orthogonal , where the rows ( or columns ) are orthogonal vectors of length @xmath61 , called _
walsh - hadamard sequences_. a theoretically defined walsh - hadamard matrix @xcite has no normalization factor @xmath80 in ( [ eq : wh ] ) .
however , we introduce it so that each walsh - hadamard sequence has the unit energy .
the walsh - hadamard sequences are described by the algebraic structure of boolean functions and the first - order reed - muller code . in the walsh - hadamard matrix @xmath81 ,
let @xmath82 be the @xmath28th column vector of length @xmath61 , i.e. , @xmath83 .
let @xmath84 and @xmath85 , where @xmath86 .
then , @xmath87 is given by @xmath88 where the addition in the exponent is the modulo-@xmath57 addition . without the normalization factor @xmath89 ,
the @xmath28th column vector of length @xmath61 is a ` @xmath90'-codeword of length @xmath61 associated with the boolean function @xmath91 of @xmath92 variables .
precisely , @xmath91 generates @xmath82 through ( [ eq : h_col ] ) while the row index @xmath25 runs through @xmath67 to @xmath68 , where @xmath82 is equivalent to a ` @xmath90'-codeword in the first - order reed - muller code @xmath93 .
since each column vector corresponds to a codeword of @xmath93 of the minimum hamming weight @xmath94 , it is straightforward that the sum of the column elements is either @xmath95 or @xmath67 , i.e. , @xmath96 from ( [ eq : wh ] ) , a @xmath97 walsh - hadamard matrix is given by @xmath98 we have @xmath99 , @xmath100 , @xmath101 , and @xmath102 .
then , it is easily checked that each boolean function @xmath91 generates the @xmath28th column vector @xmath82 , @xmath103 , through ( [ eq : h_col ] ) while the row index @xmath104 runs through @xmath105 to @xmath106 .
also , ( [ eq : exp_sum ] ) is true for each column vector . the _ walsh - hadamard transform _
@xcite of a vector @xmath107 is defined by @xmath108 where @xmath84 and @xmath85 . from the algebraic structure of the walsh - hadamard matrix
described above , it is straightforward that the walsh - hadamard transform of @xmath109 is given as @xmath110 the peak - to - average power ratio ( papr ) @xcite of a multicarrier signal @xmath45 is defined by @xmath111}\ ] ] where @xmath42 is an ofdm symbol duration and @xmath112 $ ] denotes the ensemble average . using the orthogonality of spreading sequences ,
the approach made in @xcite implies that the average power of the mc - cdma signal @xmath45 in ( [ eq : s_t ] ) is determined by @xmath113 = \frac{w}{k } \cdot \sum_{n=0 } ^{n-1 } \sum_{l=0 } ^{l-1 } \sum_{k=0 } ^{k-1 }
|d_n ^{(k)}|^2 |c_l ^{(k)}|^2 .\ ] ] in particular , if @xmath114 and @xmath33 has the unit energy , i.e. , @xmath115 , then ( [ eq : avg_pwr ] ) becomes @xmath116 = \frac{w}{k } \cdot n \cdot k = nw\ ] ] which is equal to the average power of an uncoded mc - cdma signal where each of @xmath3 users transmits the @xmath9-bit information over an ofdm symbol duration . in the following ,
we define a polynomial @xmath117 _ associated with _
@xmath45 , similar to @xcite .
[ def : assoc_poly ] in general , the coded mc - cdma signal @xmath45 in ( [ eq : s_t ] ) has a form of @xmath118 , where @xmath40 is the number of subcarriers over an ofdm symbol and @xmath119 takes an arbitrary value . with @xmath120 , the _
associated _ polynomial @xmath117 is defined by @xmath121 from ( [ eq : avg_pwr2 ] ) and ( [ eq : s_z_org ] ) , the papr of @xmath45 is translated into @xmath122
we establish the polynomial representation of a coded mc - cdma signal by presenting the associated polynomial introduced in definition [ def : assoc_poly ] . for simplicity , we first study the polynomial representation for @xmath123 , where each user transmits a single information bit with a single spreading process in an ofdm symbol . the general representation with @xmath124 is then discussed .
with @xmath123 , a coded mc - cdma signal is denoted by @xmath125 then , the polynomial representation of @xmath126 is established by the following theorem .
[ th : matrix_s ] the polynomial @xmath127 associated with @xmath126 in ( [ eq : s_t_0 ] ) is given by @xmath128 where @xmath129 .
in particular , if @xmath130 and @xmath30 is a @xmath79 walsh - hadamard matrix , i.e. , @xmath131 , then @xmath132 where @xmath133 is the walsh - hadamard transform of @xmath134 . _ proof . _ in ( [ eq : s_t_0 ] ) , let @xmath135 where @xmath136 . then , @xmath137 with @xmath120 , the associated polynomial is then given by @xmath138 , which derives ( [ eq : s_z ] ) .
if @xmath131 , then ( [ eq : s_z_had ] ) is immediate .
@xmath139 [ co : s_z_sqr ] with @xmath123 , if a coded mc - cdma signal @xmath126 has the papr of at most @xmath140 , then @xmath141 from ( [ eq : s_z_papr ] ) . also , we have from ( [ eq : s_z ] ) @xmath142 where @xmath129 with @xmath143 . in ( [ eq : u_0 ] ) , if @xmath131 , the spread output @xmath144 is the walsh - hadamard transform of @xmath134 with a scaling factor @xmath145 in a walsh - hadamard spread mc - cdma , where @xmath130 .
a similar property has been noticed in a spreading process of multicode cdma @xcite .
in addition , @xmath146 has the following structure .
[ lemma : cz ] let @xmath81 be a @xmath79 walsh - hadamard matrix and @xmath147 .
then , @xmath148 where @xmath149 , where @xmath150 and @xmath151 .
note that @xmath152 if @xmath153 , or @xmath154 if @xmath155 . _ proof . _ if @xmath156 , then @xmath157 thus , ( [ eq : cz ] ) is true for @xmath156 .
assume ( [ eq : cz ] ) also holds for @xmath158 , i.e. , @xmath159 where @xmath160 .
from the recursive construction of @xmath161 , we have @xmath162 where @xmath163 .
thus , from ( [ eq : m_k-1 ] ) and ( [ eq : m_k ] ) , @xmath164 from ( [ eq : m1 ] ) and ( [ eq : m_k_true ] ) , ( [ eq : cz ] ) is true by induction .
@xmath139 in definition 6 of @xcite , parker and tellambura defined @xmath165 , a set of normalized complex sequences of length @xmath61 by a tensor product , where @xmath166 .
in fact , @xmath146 in lemma [ lemma : cz ] is a special case of @xmath165 with @xmath150 and @xmath151 , @xmath167 . in ( [ eq : s_t_0 ] ) , replacing @xmath168 by @xmath169 leads us to @xmath170 and its associated polynomial @xmath171 , i.e. , @xmath172 where @xmath129 . obviously , @xmath170 is also a coded mc - cdma signal where @xmath26 is loaded with a single spreading process over an ofdm symbol . in particular , if @xmath173 , then the spreading process is equivalent to the walsh - hadamard transform ( wht ) , which enables the efficient implementation of spreading and despreading processes .
[ th : s_z_n ] with @xmath171 in ( [ eq : s_n_2 ] ) , the associated polynomial of a coded mc - cdma signal @xmath45 in ( [ eq : s_t ] ) is determined by @xmath174 in other words , @xmath117 is a polynomial obtained by _ interleaving _
@xmath171 s for @xmath15
. _ proof . _ in ( [ eq : s_t ] ) , @xmath175 for a given @xmath14 , @xmath176 is a signal assigned to the @xmath177th subcarriers while @xmath28 runs through @xmath67 to @xmath178 . from ( [ eq : s_n ] ) , it is straightforward that the associated polynomial of @xmath179 is given by @xmath180 where @xmath181 . compared to ( [ eq : s_n_2 ] ) , @xmath182 , and thus the associated polynomial @xmath117 is @xmath183 @xmath139 [ rm : cor1 ] since @xmath170 is a coded mc - cdma signal with a single spreading process over an ofdm symbol , corollary [ co : s_z_sqr ] is also valid for @xmath171 . precisely , if @xmath170 has the papr of at most @xmath140 , then @xmath184 and @xmath185 for @xmath15 , where @xmath129 with @xmath143 . using its associated polynomial @xmath117 in theorem [ th : s_z_n ] , we determine the papr bound of a coded mc - cdma signal @xmath45 with @xmath124 .
[ th : papr_s_n ] in ( [ eq : s_n_2 ] ) , assume the maximum papr of @xmath170 is @xmath140 , i.e. , @xmath186 .
then , the coded mc - cdma signal @xmath45 in ( [ eq : s_t ] ) has the papr of at most @xmath187 , i.e. , @xmath188 _ proof .
_ from the associated polynomial @xmath117 in ( [ eq : s_z_n ] ) , @xmath189 where @xmath143 . from remark [ rm : cor1 ] , @xmath186 implies @xmath190 for every @xmath14 .
thus , @xmath191 .
therefore , the papr of @xmath45 is bounded by @xmath192 @xmath139 although the proof is straightforward and the bound seems not so tight , theorem [ th : papr_s_n ] gives us an insight that the maximum papr of coded mc - cdma signals increases as each user transmits more data bits ( @xmath9 ) in an ofdm symbol .
therefore , @xmath9 should be as small as possible to remove the probability that the mc - cdma signal has the high papr .
in this section , we develop a variety of subcodes of @xmath70 for a @xmath7 coding scheme of a coded mc - cdma in figure [ fig : coded_sys ] , where the codeword of length @xmath193 is associated with a boolean function of degree @xmath69 .
we assume that the @xmath6-bit codeword is _ fully - loaded _ to all the available walsh - hadamard spreading sequences of length @xmath61 , so @xmath130 .
we analyze the papr properties of the fully - loaded , reed - muller coded , and walsh - hadamard spread mc - cdma signals .
first of all , we study the papr for @xmath123 . then , the papr for @xmath194 is investigated . in the mc - cdma system with @xmath123 , @xmath195 where @xmath196 is a codeword of a @xmath197 reed - muller subcode . in this section ,
we denote @xmath198 for simplicity .
let @xmath198 be a codeword of the first - order reed - muller code @xmath199 .
when it is employed as a @xmath7 coding scheme in a coded mc - cdma , the dimension is @xmath200 and the codeword length is @xmath201 . each codeword is associated with a boolean function of @xmath202 where the addition is computed modulo-@xmath57 . the papr of the mc - cdma signal encoded by a codeword in @xmath203 is determined in the following .
[ th : papr_1 ] with @xmath130 and @xmath123 , let @xmath126 be a walsh - hadamard spread mc - cdma signal encoded by @xmath204 . then
, the papr of @xmath126 is @xmath205 _ proof . _ from theorem
[ th : matrix_s ] , the associated polynomial of @xmath126 is given by @xmath206 where @xmath133 is the walsh - hadamard transform of @xmath134 , and @xmath147 . by definition ,
@xmath207 where @xmath208 in ( [ eq : bool_1 ] ) , @xmath209 . from ( [ eq : exp_sum ] ) , @xmath210 where @xmath211 for given @xmath58 s .
therefore , @xmath212 . for any @xmath213 ,
the papr of @xmath126 is therefore @xmath214 @xmath139 from theorem [ th : papr_1 ] , we see that the first - order reed - muller code is a simple and effective coding scheme that provides the uniform power for the coded mc - cdma signals . however , it has a relatively low code rate @xmath215 , which vanishes as the code length increases . therefore , we need to develop high - rate coding schemes at the expense of the papr increases . from a seed pair of codes ,
we present how to recursively construct a new code using the associated boolean functions .
we also analyze the papr of the coded mc - cdma signals .
[ th : recur ] let @xmath60 and @xmath216 be boolean functions of @xmath217 variables , where @xmath218 and @xmath219 are the codewords of length @xmath94 associated with @xmath60 and @xmath216 , respectively .
assume that the code rate of each code is @xmath220 and @xmath221 , respectively .
let @xmath222 and @xmath223 be the coded mc - cdma signals encoded by @xmath224 and @xmath225 , respectively , each of which has a form of ( [ eq : s_t_0 ] ) where @xmath226 and @xmath123 .
assume that each signal has the papr of at most @xmath140 .
consider a boolean function @xmath227 of @xmath92 variables defined by @xmath228 then , a codeword @xmath229 of length @xmath61 associated with @xmath230 has the code rate @xmath231 .
let @xmath126 be a coded mc - cdma signal of ( [ eq : s_t_0 ] ) encoded by @xmath229 , where @xmath130 and @xmath123 .
then , the papr of @xmath126 is @xmath232 _ proof . _ obviously , @xmath233 and @xmath234 , respectively .
thus , the number of codewords @xmath229 is @xmath235 and the code rate is @xmath236 . in particular , if @xmath237 , then we keep the code rate @xmath238 while the codeword length doubles .
let @xmath239 and @xmath240 be the bpsk modulation outputs of length @xmath94 from @xmath224 and @xmath225 , respectively . from ( [ eq : recur ] ) , it is straightforward that @xmath241 and @xmath242 , where ` @xmath243 ' denotes a concatenation .
then , the associated polynomial @xmath127 is determined by @xmath244 where @xmath147 , @xmath245 , @xmath246 , and @xmath247 .
let @xmath248 and @xmath249 .
then , ( [ eq : recur_s_z ] ) becomes @xmath250 replacing the boolean function @xmath216 by @xmath251 leads us to the change of the above associated polynomial to @xmath252 , i.e. , @xmath253 then , latexmath:[\[\label{eq : recur_s_z_sum } if @xmath222 and @xmath223 have the papr of at most @xmath140 , then corollary [ co : s_z_sqr ] implies @xmath255 and @xmath256 , respectively .
thus , @xmath257 ) and ( [ eq : b_f_b_g ] ) , @xmath258 where @xmath259 from the definition of @xmath260 and @xmath261 .
thus , the papr of @xmath126 is @xmath262 @xmath139 the recursive construction of a boolean function has been originally discussed in @xcite for the papr of multicode cdma . in theorem
[ th : recur ] , we showed that the construction of ( [ eq : recur ] ) also provides the bounded papr for multicarrier cdma .
in general , if there is a seed code @xmath263 of length @xmath94 and size @xmath264 , then we can construct a new code @xmath265 of length @xmath61 and size @xmath266 by _ concatenating _ a pair of codewords from the seed .
if the papr of each coded mc - cdma signal for the seed @xmath263 is at most @xmath140 , then each coded mc - cdma signal encoded by the new code @xmath265 provides the papr of at most @xmath267 .
if each codeword in @xmath263 is associated with a boolean function of degree at most @xmath268 , then @xmath265 is a subcode of @xmath70 defined by a boolean function of degree @xmath69 , where the minimum hamming distance of @xmath265 is at least @xmath73 .
construction [ cst : gen_code ] summarizes a recursive code construction for the application to mc - cdma .
[ cst : gen_code ] for positive integers @xmath269 and @xmath92 , @xmath270 , let @xmath271 and @xmath272 , where @xmath273 . starting with @xmath274 and @xmath275 ,
the boolean function @xmath276 of degree @xmath69 is constructed by the @xmath277 successive recursions of @xmath278 while @xmath279 runs through @xmath57 to @xmath69 . in ( [ eq : b_r_bool ] ) , the boolean function @xmath280 has the same form as @xmath281 , but may have different coefficients .
let @xmath198 be a codeword of @xmath282 , associated with a boolean function of @xmath283 .
then , @xmath284 has total @xmath285 codewords through the @xmath277 recursions . in a walsh - hadamard spread mc -
cdma with @xmath130 and @xmath123 , the papr of @xmath126 encoded by @xmath229 is at most @xmath286 from theorems [ th : papr_1 ] and [ th : recur ] .
finally , the code parameters of @xmath284 are summarized as follows . * dimension @xmath287 and code length @xmath193 , * code rate @xmath288 , * minimum hamming distance @xmath289 , * @xmath290 .
first of all , we present a specific code example @xmath291 of length @xmath61 through a single recursion , where a pair of codewords in @xmath292 is employed as the seed .
[ cst : b2 ] let @xmath198 be a codeword of @xmath293 that is associated with a boolean function @xmath294 defined by @xmath295 in a walsh - hadamard spread mc - cdma with @xmath130 and @xmath123 , the code parameters including the papr of a coded mc - cdma signal @xmath126 encoded by @xmath229 are summarized as follows . * dimension @xmath296 and code length @xmath193 , * code rate @xmath297 , * minimum hamming distance @xmath298 , * @xmath299 . with a single recursion ( @xmath300 ) in construction
[ cst : gen_code ] , ( [ eq : bool_2 ] ) is straightforward by @xmath301 where @xmath302 and @xmath303 .
[ rem : b2 ] by generalizing ( [ eq : bool_2 ] ) to @xmath304 we obtain a code @xmath305 associated with @xmath306 .
obviously , the coding scheme @xmath291 in construction [ cst : b2 ] is a special case of @xmath305 for @xmath307 .
it is not so hard to prove that a coded mc - cdma signal @xmath126 encoded by a codeword associated with @xmath306 has the papr of at most @xmath57 for any @xmath308 .
while @xmath308 runs through @xmath309 to @xmath310 , we have @xmath311 distinct codewords in @xmath305 , more than the number of codewords in @xmath291 .
if we compare the code rates of @xmath305 and @xmath291 , however , the code rate difference of @xmath312 is very small and approaches to @xmath67 as @xmath92 increases .
meanwhile , the encoding and the decoding complexities of the generalized coding scheme @xmath305 are obviously larger than those of @xmath291 . for the little contribution to the code rate and the increase of the complexities from @xmath305
, we therefore consider @xmath291 as our coding scheme for low papr . by employing the coding scheme @xmath291 , the walsh - hadamard spread mc - cdma system with @xmath130 and @xmath123 is able to support maximum @xmath313 users or @xmath313 information bits from a single user in an ofdm symbol , providing the papr of at most @xmath57 .
[ cst : b3 ] let a boolean function @xmath314 be defined by @xmath315 where @xmath316 .
let @xmath198 be a codeword of @xmath317 that is associated with @xmath314 . in a walsh - hadamard spread mc
- cdma with @xmath130 and @xmath123 , the code parameters including the papr of a coded mc - cdma signal @xmath126 encoded by @xmath229 are summarized as * dimension @xmath318 and code length @xmath193 , * code rate @xmath319 , * minimum hamming distance @xmath320 , * @xmath321 .
similar to construction [ cst : b2 ] , the boolean function @xmath314 is immediate from the twice recursions of ( [ eq : recur ] ) for @xmath322 and @xmath106 with @xmath323 . with the coding scheme @xmath324 , the walsh - hadamard spread mc - cdma system with @xmath130 and @xmath123 is able to support maximum @xmath325 users or @xmath325 information bits from a single user in an ofdm symbol , providing the papr of at most @xmath326 .
table [ tb : gen_recur ] lists the parameters of @xmath327 of length @xmath61 from construction [ cst : gen_code ] , where @xmath328 and @xmath329 .
it elucidates a connection between the code rates of @xmath327 and the maximum papr of mc - cdma signals encoded by @xmath327 .
obviously , we obtain a high rate reed - muller subcode at the cost of high papr for the coded mc - cdma signals .
.the parameters of @xmath284 of length @xmath61 for some @xmath69 s [ cols="^,^,^,^,^ " , ] [ tb : code_rate ] in figure [ fig : coded_sys ] , the reed - muller subcode introduced in construction [ cst : gen_code ] is applied at each spreading process for encoding the information from a single or multiple users in the reed - muller coded and walsh - hadamard spread mc - cdma transmitter ( @xmath130 ) .
precisely , a @xmath197 reed - muller subcode @xmath284 encodes a @xmath4-bit input data @xmath330 at the @xmath14th spreading process , @xmath15 , to produce a codeword @xmath22 of length @xmath61 , which goes through walsh - hadamard spreading , interleaving , and ifft in the sequel . in the encoding process , the codeword @xmath331 is obtained by @xmath332 where @xmath333 is the @xmath334 generator matrix of @xmath284 , where @xmath335 .
the recursion of boolean functions in ( [ eq : b_r_bool ] ) equivalently derives the recursion of generator matrices of @xmath336 where @xmath337 of length @xmath338 and @xmath339 is a @xmath340 matrix . by elementary row operations ,
it is equivalent to @xmath341 while @xmath279 runs through @xmath57 to @xmath69 , the generator matrix @xmath333 is constructed by the @xmath277 recursions of ( [ eq : g_r_org ] ) or ( [ eq : g_r ] ) , where the initial matrix @xmath342 is the @xmath343 generator matrix of @xmath344 given by @xmath345 for the notations @xmath346 s of the generator matrix of the first - order reed - muller codes , see @xcite . in particular , we are able to determine @xmath347 and @xmath348 directly from the boolean expressions in ( [ eq : bool_2 ] ) and ( [ eq : bool_3 ] ) , respectively .
the generator matrices of @xmath349 and @xmath350 are @xmath351 and @xmath352 matrices , respectively . with each @xmath346 of length @xmath61
, we have @xmath353 note that @xmath347 and @xmath348 in ( [ eq : g_23 ] ) have the different orders of rows with those generated by the recursions of ( [ eq : g_r_org ] ) or ( [ eq : g_r ] ) . in this paper , we use @xmath347 and @xmath348 in ( [ eq : g_23 ] ) .
let @xmath354 . then
, the generator matrices of @xmath349 and @xmath350 are @xmath355 and @xmath356 matrices , respectively .
@xmath357 we briefly introduce some decoding techniques for reed - muller subcodes .
the first - order reed - muller code can be decoded by the _ fast hadamard transform ( fht ) _ technique described in @xcite . in general ,
the @xmath69th - order reed - muller code is decoded by the _ reed decoding algorithm _ @xcite . in particular ,
if we consider @xmath291 or @xmath324 as a supercode of the union of cosets of @xmath93 , then we can accomplish the soft decision decoding by removing each possible coset representative from the received codeword and then applying the fht @xcite@xcite . for the encoding and decoding of golay complementary sequences ,
see @xcite . in what follows
, we discuss the papr of coded mc - cdma signals in a general case of @xmath124 .
we restrict our attention to a walsh - hadamard spread mc - cdma employing @xmath291 or @xmath324 that provides the acceptable code rate as well as the low papr for the coded mc - cdma signals .
we show that the maximum papr depends on the actual number of users supported by the mc - cdma .
[ th : user ] assume that @xmath291 is employed in a walsh - hadamard spread mc - cdma system in figure [ fig : coded_sys ] , where @xmath130 .
the maximum papr of the coded mc - cdma signal @xmath45 is then determined by @xmath358 similarly , if the system employs @xmath324 , then the maximum papr of @xmath45 is @xmath359 _ proof . _ in theorem [ th : papr_s_n ] , it is easy to see that @xmath360 , where @xmath126 and @xmath170 are given in ( [ eq : s_t_0 ] ) and ( [ eq : s_n_2 ] ) , respectively
. therefore , @xmath361 , and @xmath326 when @xmath203 , @xmath291 , and @xmath324 are employed as the coding scheme , respectively . in @xmath347 of ( [ eq : g_23 ] ) , if @xmath362 , the first @xmath363 rows participate in the encoding process , while the other rows are ignored by zero tailing . since
a linear combination of the first @xmath363 rows generates a codeword of @xmath203 , it is obvious that if @xmath362 , then @xmath364 from theorems [ th : papr_s_n ] and [ th : papr_1 ] .
if @xmath365 , on the other hand , @xmath366 from construction [ cst : b2 ] and theorem [ th : papr_s_n ] .
therefore , ( [ eq : papr2_w ] ) is true for @xmath291 .
similar to this approach , ( [ eq : papr3_w ] ) is also true for @xmath324 from the generator matrix @xmath348 of ( [ eq : g_23 ] ) and theorem [ th : papr_s_n ] .
@xmath139 in general , if @xmath284 is employed as the coding scheme , the walsh - hadamard spread mc - cdma signals have the papr of at most @xmath367 from construction [ cst : gen_code ] and theorem [ th : papr_s_n ] . however , theorem [ th : user ] is not true for the mc - cdma signals if the generator matrix @xmath333 is recursively constructed by ( [ eq : g_r_org ] ) or ( [ eq : g_r ] ) .
we need to reorder the rows of @xmath333 to achieve the maximum papr depending on the number of actual users as in theorem [ th : user ] . in section iv - b , we developed various reed - muller subcodes of length @xmath61 to control the peak power of mc - cdma signals with @xmath61 subcarriers in a systematic way .
in fact , we may employ the coding scheme for a codeword of length @xmath368 , where @xmath369 , which encodes @xmath370 , a concatenation of @xmath9 uncoded data block .
then , the codeword covers the entire @xmath368 subcarriers to control the peak power and to ultimately reduce the maximum papr of the coded mc - cdma signal . in this case , however , the code rate may be dramatically reduced for such a long codeword because the coding scheme is a subcode of the reed - muller code .
this also enlightens a connection between the code rates and the maximum papr of coded mc - cdma signals . virtually treating a single user s data as multiple users one ,
a coded mc - cdma system can be considered as an equivalent _ spread ofdm _
@xcite , where the single user s data is spread across a set of subcarriers to enjoy frequency diversity . by applying the coding schemes introduced in this section
, the spread ofdm additionally has the benefits of low papr and good error correction capability .
this section provides simulation results to confirm our theoretical analysis and presents some discussions on statistical results of papr of mc - cdma signals .
the papr properties of a reed - muller coded and walsh - hadamard spread mc - cdma system are compared to those of a pair of uncoded systems . in the uncoded systems , the one employs walsh - hadamard ( wh ) spreading sequences ,
while the other uses golay complementary ( gc ) spreading sequences each of which forms a row of a recursively constructed golay complementary spreading matrix @xcite . in our coded mc - cdma , we employ @xmath324 as the coding scheme which we believe is a good coding solution providing the acceptable code rates , the moderate complexity , and the low papr for the coded mc - cdma signals . for a fair comparison , we assume that all the mc - cdma systems transmit the same number of information bits in an ofdm symbol from @xmath3 active users .
if the uncoded systems transmit @xmath371 information bits in an ofdm symbol , our coded system of code rate @xmath372 then needs to transmit @xmath373 _ coded _
bits for the transmission of @xmath371 information bits . therefore
, while the uncoded ones have @xmath374 subcarriers , the coded system needs to use @xmath375 subcarriers in an ofdm symbol , where @xmath2 is a spreading factor used in the uncoded systems . in the following ,
our simulations employ @xmath324 of code rate @xmath376 or @xmath309 , where the coded mc - cdma uses @xmath377 or @xmath374 subcarriers in an ofdm symbol by employing the spreading sequences of length @xmath378 or @xmath2 . in our simulations , we measure the discrete - time papr @xcite of each mc - cdma signal from the idft ( inverse discrete fourier transform ) of the oversampling factor @xmath379 .
also , we statistically measure the papr over @xmath380 ofdm symbols for @xmath381 randomly generated information bits .
in figures [ fig : ccpd12 ] and [ fig : papr12 ] , @xmath3 users access to each mc - cdma system to transmit @xmath382 information bits per each user in an ofdm symbol .
the uncoded mc - cdma systems use the spreading factor @xmath383 and @xmath384 subcarriers , where maximum @xmath385 users are supported , i.e. , @xmath386 . as the coded mc - cdma system also needs to support up to @xmath385 users ,
we choose a @xmath387 code @xmath388 as its coding scheme , where @xmath389 , @xmath390 , and @xmath376 .
thus , our coded mc - cdma uses the spreading factor @xmath391 and @xmath392 subcarriers to transmit @xmath371 information bits in an ofdm symbol from the @xmath3 active users . in the coded mc - cdma system ,
each @xmath393-bit codeword is fully - loaded to all the available @xmath393 walsh - hadamard spreading sequences regardless of @xmath3 . on the other hand ,
the uncoded mc - cdma systems assign the @xmath3 spreading sequences to @xmath3 users _ on demand _ , so they are fully - loaded only if @xmath394 . or @xmath395 information bits in an ofdm symbol .
the code rate of the coded mc - cdma is @xmath396.,scaledwidth=100.0% ] figure [ fig : ccpd12 ] shows the complementary cumulative distribution functions ( ccdf ) of @xmath397 of each mc - cdma signal for @xmath398 and @xmath385 .
it reveals that the coded mc - cdma is superior to the others when the number of active users is small .
precisely , if @xmath398 , it reduces the papr @xmath399 achieving @xmath400 by more than @xmath57 db , compared to the uncoded systems .
moreover , theorem [ th : user ] ensures that there exists no coded mc - cdma signal with papr @xmath401 db for @xmath398 , which implies that the coded mc - cdma also outperforms the uncoded ones in theoretical aspects . if @xmath394 , the coded mc - cdma has almost the same papr @xmath399 as the uncoded walsh - hadamard spread mc - cdma for achieving @xmath400 . even in this case
, it is theoretically guaranteed that no coded mc - cdma signal has @xmath402 db , which may not be true in the uncoded systems .
figure [ fig : ccpd12 ] also shows that most of the coded mc - cdma signals in the statistical experiments have much smaller papr than the theoretical maximum predicted by theorem [ th : user ] . and active number of users for mc - cdma systems where @xmath403 .
the code rate of the coded mc - cdma is @xmath396.,scaledwidth=100.0% ] figure [ fig : papr12 ] displays the papr @xmath399 of each mc - cdma achieving @xmath400 according to the number of active users @xmath404 .
it is well known @xcite that the uncoded walsh - hadamard ( wh ) spread mc - cdma shows the high papr when the number of active users is small .
the papr then decreases as the number of users increases .
on the other hand , the uncoded golay complementary ( gc ) spread mc - cdma has the low papr for the small number of users . however , the papr gets higher than that of the uncoded walsh - hadamard spread mc - cdma as the number of users increases .
figure [ fig : papr12 ] shows that the coded mc - cdma is a good alternative to the two uncoded systems by providing the smallest papr @xmath399 for almost all user numbers .
moreover , theorem [ th : user ] assures that the maximum papr of the coded system is theoretically limited to @xmath405 db for @xmath406 , @xmath407 db for @xmath408 , and @xmath409 db for @xmath410 , respectively , where figure [ fig : papr12 ] provides the numerical evidences .
therefore , it is theoretically guaranteed that there exists no coded mc - cdma signal with the papr higher than the maximum values for each user , which is not generally true in the uncoded systems .
the theoretical and statistical results show that the coded mc - cdma dramatically reduces its papr for the small number of users , which effectively solves the high papr problem in the uncoded mc - cdma .
figure [ fig : papr12 ] also shows that if the number of active users is large ( @xmath411 ) , the statistical papr @xmath399 is much smaller than the theoretical maximum predicted by theorem [ th : user ] . as a result
, we claim that the coded mc - cdma provides the best statistical and theoretical solution for papr reduction for any number of users . in figures [ fig : ccpd1 ] and [ fig : papr1 ] , the mc - cdma systems support @xmath3 users where each user transmits @xmath413 information bits in an ofdm symbol .
the uncoded mc - cdma systems use the spreading factor @xmath414 and @xmath384 subcarriers to transmit @xmath371 information bits in an ofdm symbol . to support up to @xmath379 users , the coded mc - cdma system employs @xmath415 , a @xmath416 code with @xmath417 and @xmath418 . in this case , @xmath415 is used as a mapping scheme mentioned in remark [ rem : rm_map ] . in our mc - cdma , each @xmath379-bit uncoded data @xmath18 , @xmath419 , is _ transformed _ by the reed - muller mapping scheme for papr reduction .
thus , it uses the spreading factor @xmath379 and @xmath395 subcarriers to transmit @xmath371 data bits in an ofdm symbol , which is the same as the uncoded systems .
however , note that each @xmath379-bit codeword of our mc - cdma is fully - loaded to all the available spreading sequences of length @xmath379 regardless of @xmath3 . or @xmath395 information bits in an ofdm symbol employing the spreading factor @xmath379 and @xmath395 subcarriers .
the code rate of the coded mc - cdma is @xmath420.,scaledwidth=100.0% ] figure [ fig : ccpd1 ] shows the results of @xmath397 of each mc - cdma for @xmath421 and @xmath379 .
we observed that if @xmath421 , our reed - muller mapped mc - cdma system reduces the papr @xmath399 by about @xmath106 db to achieve @xmath422 , compared to the uncoded walsh - hadamard ( wh ) spread mc - cdma employing the same spreading factor and the same number of subcarriers .
moreover , theorem [ th : user ] ensures that our system has no signal with papr @xmath401 db for @xmath421 .
note that theorem [ th : user ] determines the maximum papr of @xmath423 db ( if @xmath424 ) , @xmath425 db ( if @xmath426 ) , and @xmath427 db ( if @xmath428 ) , respectively .
thus , even if the statistical papr property of our mc - cdma is almost identical to that of the uncoded walsh - hadamard spread mc - cdma for @xmath398 , our system has no probability of signals with the papr higher than @xmath427 db , which is however unclear in the uncoded systems . and active number of users for mc - cdma systems where @xmath403 .
all the mc - cdma systems use the spreading factor @xmath379 and @xmath395 subcarriers .
the code rate of the coded mc - cdma is @xmath420.,scaledwidth=100.0% ] similar to figure [ fig : papr12 ] , figure [ fig : papr1 ] displays the papr @xmath399 of each mc - cdma achieving @xmath400 according to the number of users @xmath429 .
it shows that our mc - cdma provides the smallest papr @xmath399 for almost all user numbers .
also , it numerically confirms that the theoretical maximums of papr in theorem [ th : user ] hold for each user number .
figure [ fig : papr1 ] shows that if the number of active users is large , the statistical papr @xmath399 is much smaller than the theoretical maximum predicted by theorem [ th : user ] .
for the small number of users , on the other hand , the coded mc - cdma solves the high papr problem of the uncoded mc - cdma by dramatically reducing its papr .
along with figure [ fig : papr12 ] , the coded mc - cdma can be the best statistical and theoretical solution for papr reduction for any number of active users regardless of its code rate .
a drawback of the reed - muller mapped mc - cdma is that it could be employed only for a small spreading factor due to the demapping complexity at the receiver .
this paper has presented a coded mc - cdma system where the information data is encoded by a reed - muller subcode for the sake of papr reduction . in the system ,
the codeword is then fully - loaded to walsh - hadamard spreading sequences , where the spreading and the despreading processes are efficiently implemented by the walsh - hadamard transform ( wht ) .
we have established the polynomial representation of a coded mc - cdma signal for theoretical analysis of the papr .
we have then developed a recursive construction of the reed - muller subcodes which provide the transmitted mc - cdma signals with the bounded papr as well as the error correction capability .
we have also investigated a theoretical connection between the code rates and the maximum papr in the coded mc - cdma .
simulation results showed that the papr of the coded mc - cdma signal is not only theoretically bounded , but also statistically reduced by the reed - muller coding schemes .
in particular , it turned out that the coded mc - cdma could solve the papr problem of uncoded mc - cdma by dramatically reducing its papr for the small number of users .
finally , the theoretical and statistical studies exhibited that the reed - muller subcodes are effective coding schemes for peak power control in mc - cdma with small and moderate numbers of users , subcarriers , and spreading factors .
we believe this work gives us theoretical insights for papr reduction of mc - cdma and s - ofdm by means of an error correction coding .
ieee standard 802.11 - 2007 , ieee standard for information technology - local and metropolitan area networks - specific requirements , part 11 - wireless lan medium access control ( mac ) and physical layer ( phy ) specifications .
ieee802.16e-2005 , ieee standard for local and metropolitan area networks , part 16 - air interface for fixed and mobile broadband wireless access systems , amendment 2 : physical and medium access control layers for combined fixed and mobile operation in licensed bands .
k. fazel , s. kaiser , and m. schnell , `` a flexible and high performance celluar mobile communications systems based on orthogonal multicarrer ssma , '' _ wireless personal communications _ , vol . 2 , pp . 121 - 144 , 1995 .
a. e. jones , t. a. wilkinson , and s. k. barton , `` block coding scheme for reduction of peak to mean envelope power ratio of multicarrier communication schemes , '' _ electron .
_ , vol . 30 , pp . 2098 - 2099 , 1994 .
d. a. wiegandt , z. wu , and c. r. nassar , `` high - throughput , high - performance ofdm via pseudo - orthogonal carrier interferometry spreading codes , '' _ ieee trans .
_ , vol . 51 , no . 7 , pp . 1123 - 1134 , july 2003 . m. g. parker , `` close encounters with boolean functions of three different kinds , '' _
2nd international castle meeting on coding theory and applications _ , valladolid , september 2008 . also available at _ lecture notes in computer science ( lncs ) _ vol . 5228 , pp .
137 - 153 , 2008 .
m. g. parker and c. tellambura , `` golay - davis - jedwab complementary sequences and rudin - shapiro constructions , '' _ manuscript .
available in `` http://www.ii.uib.no/@xmath430matthew/constabent2.pdf '' .
j. h. conway and n. j. a. sloane , `` soft decoding techniques for codes and lattices , including the golay codea and the leech lattice , '' _ ieee trans . inform . theory _
it-32 , no . 1 ,
41 - 50 , jan . 1986 | reed - muller codes are studied for peak power control in multicarrier code - division multiple access ( mc - cdma ) communication systems . in a coded mc - cdma system ,
the information data multiplexed from users is encoded by a reed - muller subcode and the codeword is fully - loaded to walsh - hadamard spreading sequences .
the polynomial representation of a coded mc - cdma signal is established for theoretical analysis of the peak - to - average power ratio ( papr ) .
the reed - muller subcodes are defined in a recursive way by the boolean functions providing the transmitted mc - cdma signals with the bounded papr as well as the error correction capability . a connection between the code rates and the maximum papr is theoretically investigated in the coded mc - cdma .
simulation results present the statistical evidence that the papr of the coded mc - cdma signal is not only theoretically bounded , but also statistically reduced .
in particular , the coded mc - cdma solves the major papr problem of uncoded mc - cdma by dramatically reducing its papr for the small number of users .
finally , the theoretical and statistical studies show that the reed - muller subcodes are effective coding schemes for peak power control in mc - cdma with small and moderate numbers of users , subcarriers , and spreading factors .
boolean functions , multicarrier code - division multiple access ( mc - cdma ) , orthogonal frequency - division multiplexing ( ofdm ) , peak - to - average power ratio ( papr ) , reed - muller codes , spreading sequences , walsh - hadamard sequences , walsh - hadamard transform . |
type ia supernovae ( sne ia ) remain the most practical and well - exploited cosmological probe offering an immediate route to understanding ` dark energy '
. measures of distant events are being used to distinguish between einstein s cosmological constant , @xmath2 , and a scalar field whose equation of state parameter @xmath3@xmath4 @xcite . yet despite remarkable observational progress , there is no satisfactory theory explaining a sn ia event .
the mechanism by which a white dwarf accretes additional material is unclear as is the nature of the explosion itself @xcite . to facilitate progress ,
observers employ a variety of empirical correlations to reduce the intrinsic scatter of the sn ia hubble diagram .
sne ia were initially considered a one or two parameter family with the light curve width and rest - frame color as the key variables .
however improved data have revealed important correlations with the host galaxy .
events are not only more common in star - forming hosts per unit stellar mass but their light curve properties differ from those seen in quiescent galaxies @xcite , an effect that has direct consequences for their use over large look - back times @xcite .
these discoveries naturally raise the question of what further evolutionary changes might be present in the sn ia population .
a long - standing concern has been the unknown effect of an _ evolving progenitor composition _ , both in terms of a possible redshift - dependent bias , and in producing an intrinsic dispersion that could limit the effectiveness of large future surveys .
a keck study of high quality rest - frame near - uv sne ia spectra at intermediate redshift ( @xmath5 ) reveals a surprising diversity at short wavelengths where some models predict a sensitivity to metallicity ( * ? ? ?
* hereafter e08 , see also foley et al .
a large u - band dispersion had earlier been claimed in the local photometric survey of @xcite , although subsequent photometry has challenged the amount @xcite .
although some models predict a uv dispersion might arise from variations in the progenitor composition @xcite , the magnitude of the effect seen by e08 exceeds that predicted for reasonable compositional differences .
@xcite have shown a large fraction of the uv flux can be formed by reverse - fluorescence scattering which affects the dependence on composition .
if the observed dispersion is found to arise from some evolutionary trend , it could bias future @xmath6 sne campaigns that typically sample from this wavelength region .
* hereafter s09 ) compare mean sn ia spectra over a redshift path of @xmath7 and find no strong evolution , however , only three local uv spectra were available at the time , seriously limiting the comparison and giving no local measure of the uv dispersion .
efforts to understand the uv behavior of sne ia have subsequently intensified .
using the uv optical telescope on - board swift , @xcite and @xcite have confirmed the presence of a dispersion increase to shorter wavelengths in local sne ia .
comparisons between local and intermediate redshift data remain confused and so the question remains as to whether some component of the significant diversity of uv spectra seen in distant sne ia is an evolutionary phenomenon or represents some as yet unexplained diversity in the sn ia mechanism . as described in e08 ,
spectroscopic studies offer a major advantage over photometric investigations as they eliminate uncertainties arising from k - corrections and , with adequate data , the mean and dispersion can be investigated in the context of known metallic features . following the successful repair of the uv - capable space telescope imaging spectrograph ( stis ) aboard the hubble space telescope ( @xmath8 ) during the 2009 servicing mission 4 , it has become possible to make significant progress in addressing the above questions . using a non - disruptive target of opportunity campaign
( go 11721 , pi : ellis ) , we are securing maximum light stis spectra for 35 sne ia located in the hubble flow . here
we present the results from the first 12 events from this program , augmented by three events from earlier archival data , that provide comparable statistics to the survey of e08 at intermediate redshift for the analysis undertaken here .
lccl ptf-09dlc & @xmath9 & @xmath10 & 0.0675 + ptf-09dnl & @xmath11 & & 0.0231 + ptf-09dnp & @xmath12 & @xmath13 & 0.0373 + ptf-09fox & @xmath14 & @xmath15 & 0.0718 + ptf-09foz & @xmath16 & @xmath17 & 0.0543 + ptf-10bjs & @xmath18 & @xmath19 & 0.0300 + ptf-10fps & @xmath20 & @xmath21 & 0.0215 + ptf-10hdv & @xmath22 & @xmath23 & 0.0533 + ptf-10hmv & @xmath24 & @xmath25 & 0.0324 + ptf-10icb & @xmath26 & @xmath27 & 0.0086 + ptf-10mwb & @xmath28 & @xmath29 & 0.0313 + sn 2009le & @xmath30 & @xmath31 & 0.0178 + sn 1981b & @xmath32 & @xmath33 & 0.0060 + sn 1992a & @xmath34 & @xmath35 & 0.0061 + sn 2001ba & @xmath36 & @xmath37 & 0.0305
a significant challenge in delivering targets to hst for observations at maximum light is the need to detect and identify convincing sne ia candidates soon after explosion . for a _ non - disruptive _ target of opportunity ( too ) campaign , defined as one where phase ii observations are submitted for inclusion in the hst schedule built for the second week following the submission
, sn confirmation must occur @xmath38 calendar days prior to submission . as a result ,
spectroscopic sne ia target confirmations @xmath39 days before maximum light are necessary to acquire stis observations with phases @xmath40 days from maximum light and match a distribution of intermediate - redshift near uv spectra secured by the e08 campaign .
figure [ program ] illustrates the timeline of the data acquisition for the 12 sne of our hst program presented here from detection , to ground - based spectroscopic confirmation , to stis observation and shows the ability of ptf to detect sn ia outbursts as early as @xmath41 days prior to maximum light .
the identification of sne ia suitable for hst non - disruptive observations proceeded in two stages .
the early identification of candidate events was based on photometric survey data .
eleven of the 12 newly - discovered sne were identified between 2009 august 19 and 2010 june 29 by the palomar transient factory ( ptf ; * ? ? ?
* ; * ? ? ?
as with the cfht supernova legacy survey ( snls * ? ? ?
* ) utilized by the e08 intermediate redshift program , ptf is a rolling search for transient events unbiased to the nature of the host galaxy . the remaining event ( sn 2009le ) was triggered on the reported results of an independent search ( * ? ? ?
* chase ) during a period when the palomar observatory was under extensive ash clouds from the 2009 california forest fires .
[ 0.75 ] [ 0.90 ] reduced to rest - frame wavelengths .
each label gives the ptf identification ( with 09le indicating sn 2009le ) and phase of stis observations .
spectra in red represent those not included in the mean and dispersion analysis ( see text).,title="fig : " ] ground - based spectroscopic follow - up acquired within one to a few days after photometric detection represents the second stage of confirmation essential for determining the sn type and redshift .
spectroscopy was performed using regularly scheduled time and occasional too interrupts on the following telescopes / instruments : the low - resolution imaging spectrometer ( lris ; * ? ? ?
* ; * ? ? ?
* ) and deep imaging multi - object spectrograph ( deimos ; * ? ?
* ) on the keck telescopes , the gemini multi - object spectrograph ( gmos ; * ? ? ?
* ) on the gemini telescopes , fors @xcite and x - shooter @xcite on the eso very large telescopes , the double spectrograph ( dbsp ; * ? ? ?
* ) on the palomar hale telescope , and the intermediate dispersion spectrograph and imaging system ( isis ) on the william herschel telescope . by matching flux - calibrated spectra with selections from an extensive spectral database @xcite ,
improved phases typically accurate to @xmath42 days were obtained .
near - uv spectra were acquired using the space telescope imaging spectrograph ( stis ; * ? ? ?
* ; * ? ? ?
* ) between 2009 september 02 and 2010 july 13 ( fig .
[ program ] , red diamonds ) .
to best match our @xmath5 keck spectra ( e08 ) , we used the stis / ccd 430l observing mode with @xmath43 cr - split exposures totaling 1 orbit , ensuring good signal - to - noise ratio coverage down to a rest wavelength below @xmath44 . a montage of the stis spectra is shown in figure [ stis_spec ] .
associated near - simultaneous ground - based optical spectra were also taken using further regularly scheduled time and too interrupts using the facilities described above .
all data were reduced using standard iraf and idl data reduction routines .
pre- and post - maximum light ptf @xmath45-band photometric observations were used to determine the initial phase and stretch for each @xmath8 triggered event using the @xmath46 light curve fitter @xcite .
details for the sn ia sample here and the three archival sne ia ( discussed in s09 ) are listed in table [ details ] . in 4 cases ( 09dnl ,
09dnp , 09fox , 10fps ) , the ptf light curves are of marginal quality or too poorly sampled for an accurate determination of one or both of the phase and stretch .
one event ( sn 2009le ) has no ptf photometry and its phase was determined from the chase estimate @xcite which is consistent with our spectroscopic estimate . in our final analysis (
3 ) , this imprecision affects the discussion of two of our stis spectra ( 09dnl , 2009le ) .
[ 0.75 ] prior to hst cycle 13 , uv spectra were secured for 13 local sneia , but only 5 were studied close to maximum light and 2 of these were peculiar .
those suitable from earlier work are listed in table [ details ] and shown in figure [ stis_spec ] . in cycle 13
a too campaign began to increase the sample ( go 10182 , pi : filippenko ) but the failure of stis curtailed this program .
remarkably , prior to our study , more was known about the pre - maximum and maximum light uv spectra of sneia at @xmath47 than at @xmath48 .
our stis sample has now dramatically improved this situation . in order to compare the mean uv spectra of local sneia and their dispersion with the sample discussed by e08
, we must match the phase ( and ideally the stretch ) distributions of the two samples .
such a matched comparison was not possible in s09 due to the paucity of the local data . to maximize the utility of the new hst data ,
we adopt a phase range of @xmath49 to @xmath50d ( figure [ program ] ) bounded by the earliest phase of the hst data ( @xmath49 ) and a @xmath50d phase limit similar to that applied in e08 to minimize phase evolution effects on the mean and dispersion .
this criterion leads to 11 events from the hst+archival sample drawn from table [ details ] and 16 events from e08 . ignoring the two events for which accurate stretches can not be determined , the mean values are @xmath51 and @xmath52 which are consistent within the errors .
excluding the two supernovae for which reliable stretches are not available does not significantly affect our results in the following subsections .
s09 presented a comparison of the mean uv spectra determined by e08 at @xmath5 , those secured using the acs grism by @xcite at @xmath53 , and three local spectra from archival data listed in table [ details ] .
examining the @xmath5 uv spectra , s09 found some decrease with redshift in the strength of intermediate mass element features ( si ii , ca ii , and mg ii ) , but it was argued this could arise in part due to the natural drift to luminous , larger stretch events expected at high redshift . below @xmath54 3600
the mean local spectrum was highly uncertain . to facilitate a proper comparison to s09
, we construct our mean near - uv spectrum following the procedure discussed in e08 .
briefly , the spectra are normalized to have the same flux through a box filter defined between rest - frame @xmath55 and the variation in the mean spectrum is estimated via bootstrap - resampling .
use of other box filters , including the full wavelength range common to all spectra , does not significantly affect the results .
figure [ best ] shows the normalized mean spectrum and region containing @xmath56% of 100
bootstrap - resampled mean spectra for both the low- and intermediate - redshift matched samples .
although the low redshift sn ia mean spectrum closely resembles that at @xmath5 , as in s09 , we notice a marked decrease with increasing redshift in the depth of the si ii and ca ii blend near 3800 , mg ii near 4300 , and those of iron group elements below @xmath57 .
the utility of the comparison is more advantageous than that conducted by s09 for wavelengths below 3500 since our new local sample is much larger than that used by s09 and the phase and stretch distributions are better matched .
farther into the uv we note that the differences between the mean local and @xmath580.5 spectra become particularly significant .
since the stretch distributions of the two samples are similar , this could be a genuine effect rather than one arising from samples biased to more luminous and bluer events @xcite .
analysis of our eventual full sample will clarify this important point .
[ 0.75 ] e08 demonstrated a significant increase in the variance of their @xmath580.5 sne ia uv spectra for wavelengths below @xmath59 , both by comparing individual deviations of @xmath60 maximum light spectra from their mean in units of the dispersion @xmath61 , and via photometric colors measured directly from color - corrected keck spectra .
for the first time , our low - redshift sample of uv sn ia spectra is large enough to perform a similar analysis .
ideally one would color- correct the local spectra following the procedure discussed in e08 using the salt2 color law @xcite but this requires host- corrected multi - color data that must await late - time reference images in bands other than the ptf @xmath45-band .
we experimented with estimating the host contamination in @xmath62 and @xmath63 from the contemporary data but concluded the uncertainties are too great at this time .
accordingly , a comparison with data from e08 that is not color - corrected is more appropriate .
it is important to note that the color correction e08 applied did not significantly change the uv scatter and its wavelength dependent trend .
the key question we seek to address is whether the dispersion trend is generic to all sne ia , independent of redshift , or largely a feature _ of the intermediate redshift data only _ , possibly implying some evolutionary effect .
figure [ disp ] shows that the wavelength - dependent dispersion is indeed present in the local data , consistent with the photometric claims of @xcite and @xcite . in the region that contains 90% of the mean values from bootstrap resampling , the variation from one spectrum to another with respect to the mean increases below @xmath59 as in e08 .
clearly several of the features which vary between the local and e08 samples discussed earlier contribute to the dispersion suggesting a compositional origin .
even if we exclude the region dominated by strong features and consider the average dispersion from the mean spectrum in the regions uv : 2900 - 3500 and optical : 4100 - 5200 allowing for the spectrophotometric uncertainties , we find that the dispersion increases from the optical to the uv by a similar factor of @xmath642.3 in both the local and @xmath580.5 samples .
there is marginal evidence that the uv scatter may be larger in the @xmath580.5 data than in the local sample , but the spectrophotometric uncertainties for the individual spectra are naturally larger .
overall , we conclude that the uv spectral dispersion is most likely a feature of the sne ia population and not an evolutionary effect .
our initial results clarify and quantify indications from earlier work .
although the sn ia mean spectrum close to maximum light appears to have remained remarkably similar over the past 5 gyr , we find the decrease in the strength of the metallic features with increasing redshift noted by s09 also present in our more representative comparison .
given that the mean stretches of the local and [email protected] samples are similar , this may represent the expected decrease in metallicity over this epoch .
we can address this possibility in more detail with the completion of our survey .
equally important is that we observe a strong wavelength dependent scatter in the rest - frame uv spectra of our local sample , as noted in e08 .
independent of the calibration questions that have plagued recent photometric studies , the spectra demonstrate that the uv scatter is generic to sne ia over a wide range of cosmic time and is not likely an evolutionary effect .
much of this behavior can be attributed to the varying absorption line strengths of intermediate mass elements occupying the uv wavelength region , supporting the notion that the uv scatter arises from compositional differences between events .
in addition to strengthening these conclusions with a larger sample , it is now clear that further progress will follow more detailed multi - phase uv studies of selected local events .
hst has recently been awarded for such a program ( go 12298 , pi : ellis ) .
rse acknowledges support from doe grant de - sc0001101 , ms from the royal society , a.g . from the israeli science foundation and a european union marie curie fellowship .
support for program go 11721 was provided by nasa through a grant from the space telescope science institute , which is operated by aura inc . under nasa contract nas5 - 26555 .
amanullah , r. , lidman , c. , rubin , d. et al .
2010 , , 716 , 712 appenzeller , i. et al .
1998 , eso messenger , 94 , 1 .
astier , p. , et al .
2006 , , 447 , 31 balland , c. , et al . 2009 , , 507 , 85 branch , d. , lacy , c. h. , mccall , m. l. , sutherland , p. g. , uomoto , a. , wheeler , j. c. , & wills , b. j. 1983 , , 270 , 123 brown , p. , et al .
2010 , astro - ph/1007.4842 challis , p. , & berlind , p. 2009
, central bureau electronic telegrams , 2025 , 1 conley , a. , et al .
2008 , , 681 , 482 ellis , r. s. , et al .
2008 , , 674 , 51 faber , s. m. , et al .
2003 , , 4841 , 1657 foley , r. j. , filippenko , a. v. , & jha , s. w. 2008 , , 686 , 117 guy , j. , et al . 2007 , , 466 , 11 hflich , p. et al .
2000 , , 528 , 590 hook , i. m. , jrgensen , i. , allington - smith , j. r. , davies , r. l. , metcalfe , n. , murowinski , r. g. , & crampton , d. 2004 , , 116 , 425 howell , d. a. , et al .
2005 , , 634 , 1190 howell , d. a. , sullivan , m. , conley , a. , & carlberg , r. 2007 , , 667 , l37 jha , s. , kirshner , r.p . ,
challis , p. , et al .
2006 , 131 , 527 kessler , r. et al .
2009 , , 185 , 32 kimble , r. a. , et al .
1998 , , 492 , l83 kirshner , r. p. , et al .
1993 , , 415 , 589 law , n. m. , et al .
2009 , , 121 , 1395 lentz , e.j .
, et al . 2000 , 530 , 966 .
livio , m. 1999 astro - ph/9903264 mccarthy , j. k. , et al . 1998 , , 3355 , 81 miller , j. s. , & stone , r. p. s. 1993 , the kast double spectrograph ( lick obs . tech .
66 ; santa cruz : lick obs . ) milne , p.a .
2010 , astro - ph/1007.5279 oke , j. b. , & gunn , j. e. 1982 , , 94 , 586 oke , j. b. , cohen , j. g. , carr , m. , cromer , j. , dingizian , a. , harris , f. h. , labrecque , s. , lucinio , r. , schaal , w. , epps , h. , & miller , j. 1995 , , 107 , 375 pignata , g. , et al .
2009 , atel , 2022 rau , a. , et al .
2009 , , 121 , 1334 riess , a. g. , et al .
2004 , , 600 , l163 riess , a. g. , et al .
2007 , , 659 , 98 sauer , d.n . , et al .
2010 , , 391 , 1605 sullivan , m. , et al .
2006 , , 648 , 868 sullivan , m. , et al .
2006 , , 131 , 960 sullivan , m. , ellis , r. s. , howell , d. a. , riess , a. , nugent , p. e. , & gal - yam , a. 2009 , , 693 , l76 sullivan , m. , et al .
2010 , , submitted vernet , j. , et al .
2009 , eso messenger 138 , 4 woodgate , b. e. , et al .
1998 , , 110 , 1183 | we present the first results of an ongoing campaign using the stis spectrograph on - board the hubble space telescope ( hst ) whose primary goal is the study of near ultraviolet ( uv ) spectra of local type ia supernovae ( sne ia ) . using events identified by the palomar transient factory and subsequently verified by ground - based spectroscopy
, we demonstrate the ability to locate and classify sne ia as early as 16 days prior to maximum light .
this enables us to trigger hst in a non - disruptive mode to obtain near uv spectra within a few days of maximum light for comparison with earlier equivalent ground - based spectroscopic campaigns conducted at intermediate redshifts , @xmath0 .
we analyze the spectra of 12 type ia supernovae located in the hubble flow with @xmath1 .
although a fraction of our eventual sample , these data , together with archival data , already provide a substantial advance over that previously available . restricting samples to those of similar phase and stretch , the mean uv spectrum agrees reasonably closely with that at intermediate redshift , although some differences are found in the metallic absorption features .
a larger sample will determine whether these differences reflect possible biases or are a genuine evolutionary effect .
significantly , the wavelength - dependent dispersion , which is larger in the uv , follows similar trends to that observed at intermediate redshift and is driven , in part , by differences in the various metallic features . while the origin of the uv dispersion remains uncertain , our comparison suggests that it may reflect compositional variations amongst our sample rather than being predominantly an evolutionary effect . |
the interest in helical and chiral phases of frustrated low - dimensional quantum magnets has been triggered by recent experimental results . while many copper - oxide based materials predominantly realize antiferromagnetic exchange interactions , several candidate materials with magnetic properties believed to be described by frustrated ferromagnetic chains have been identified,@xcite including rb@xmath0cu@xmath0mo@xmath1o@xmath2 ( ref . ) , licuvo@xmath3 ( refs . ) , and li@xmath0zrcuo@xmath3 ( ref . ) .
the frustrated antiferromagnetic chain is well - studied,@xcite but the magnetic phase diagram of the model with ferromagnetic nearest - neighbor interactions remains a subject of active theoretical investigations.@xcite in this work we consider a parameter regime that is in particular relevant for the low - energy properties of licuvo@xmath3 , corresponding to a ratio of @xmath4 between the nearest neighbor interaction @xmath5 and the frustrating next - nearest neighbor interaction @xmath6 . as the interchain couplings for this material are an order of magnitude smaller than the intrachain
ones,@xcite we analyze a purely one - dimensional ( 1d ) model . apart from mean - field based predictions , @xcite the nature of the ground state in a magnetic field @xmath7
is not yet completely known .
therefore , combining the bosonization technique with a numerical analysis we determine ground - state properties and discuss the model s elementary excitations .
the hamiltonian for our 1d model reads : @xmath8 where @xmath9 represents a spin one - half operator at site @xmath10 .
bosonization has turned out to be the appropriate language for describing the regime @xmath11 of eq .
( [ eq:1.1 ] ) .
this result has been established by studying the magnetization process yielding a good agreement between field theory and numerical data.@xcite the derivation of the effective field theory is summarized in sec .
[ sec : field ] . here , we extend on such comparison of analytical and numerical results and further confirm the predictions of field theory by analyzing several correlation functions in sec .
[ sec : corrs ] .
then , in sec . [ sec : ex ]
, we numerically compute the one- and two - spin flip excitation gaps and compare them to field - theory predictions . finally , sec . [
sec : sum ] contains a summary and a discussion of our results .
we start from an effective field theory describing the long - wavelength fluctuations of eq .
( [ eq:1.1 ] ) . in the limit of strong next - nearest neighbor interactions
@xmath12 , the spin operators can be expressed as : @xmath13 @xmath14 is the fermi - wave vector and @xmath15 enumerates the two chains of the zig - zag ladder . in relation with eq .
( [ eq:1.1 ] ) , note that @xmath16 ( @xmath17 ) for @xmath10 odd ( even ) . @xmath18 and @xmath19 are compactified quantum fields describing the out - of - plane and in - plane angles of fluctuating spins obeying gaussian hamiltonians : @xmath20 with @xmath21 = i\theta ( y - x)$ ] , where @xmath22 is the heaviside function .
sub - leading terms are suppressed in eq .
( [ bosfor ] ) .
@xmath23 is the magnetization of decoupled chains , related to the real magnetization @xmath24 of the zig - zag system by : @xmath25 @xmath26 and @xmath27 are the luttinger liquid ( ll ) parameter and the spin - wave velocity of the decoupled chains , respectively .
the nonuniversal amplitude @xmath28 appearing in the bosonization formulas ( [ bosfor ] ) has been determined from density matrix renormalization group ( dmrg ) calculations.@xcite note that in our notation @xmath29 at saturation .
now we perturbatively add the interchain coupling term to two decoupled chains , each of which is described by an effective hamiltonian of the form eq .
( [ spinchainbosham ] ) and fields @xmath30 and @xmath31 , @xmath32 . for convenience ,
we transform to the symmetric and antisymmetric combinations of the bosonic fields @xmath33 and @xmath34 . in this basis and apart from terms @xmath35 of the form ( [ spinchainbosham ] ) , the effective hamiltonian describing low - energy properties of eq .
( [ eq:1.1 ] ) contains a single relevant interaction term with the bare coupling @xmath36 : @xmath37 and the renormalized ll parameters @xmath38 are , in the weak coupling limit : @xmath39 @xmath40 is the luttinger - liquid parameter of the soft mode of the zig - zag ladder .
the hamiltonian ( [ symantisym ] ) represents the minimal effective low - energy field theory describing the region @xmath41 of the frustrated fm spin-@xmath42 chain for @xmath43.@xcite the relevant interaction term @xmath44 opens a gap in the @xmath45 sector . since @xmath46 , relative fluctuations of the two chains are locked .
this implies that single - spin flips are gapped with a sine - gordon gap in the sector describing relative spin fluctuations of the two - chain system.@xcite gapless excitations come from the @xmath47 channel , _ i.e. _ only those excitations are soft where spins simultaneously flip on both chains .
dmrg results show that this picture applies to a large part of the magnetic phase diagram.@xcite
we now turn to the ground state properties of eq .
( [ eq:1.1 ] ) as a function of magnetization , concentrating on several correlation functions in order to identify the leading instabilities .
note that our analysis is only valid if @xmath48 .
apart from a term representing the magnetization @xmath24 induced by the external field , the longitudinal correlation function shows an algebraic decay with distance @xmath49 : @xmath50 the constants @xmath51 , @xmath52 , appearing here and in eq .
( [ nematic ] ) will be determined through a comparison with numerical results .
in contrast to eq .
( [ zzcor ] ) , the transverse @xmath53-correlation functions decay exponentially reflecting the gapped nature of the single spin - flip excitations .
here we do not restrict ourselves to the equal - time expression only , because we will need non - equal time correlation functions to extract the finite - size corrections to the gap later on .
we obtain : @xmath54 where @xmath55 stands for the euclidean time , @xmath56 is the @xmath57 gap , and @xmath58 in the weak coupling limit .
the kronecker delta strictly applies to the thermodynamic limit , while on the lattice an additional contribution for @xmath59 exists .
it is noteworthy that , different from eq .
( [ eq : sstrans ] ) , the in - plane correlation functions involving bilinear spin combinations decay algebraically .
this stems from the gapless nature of @xmath47 excitations .
in fact , these are the slowest decaying correlators close to the saturation magnetization : @xmath60 this result is reminiscent of a partially ordered state because the ordering tendencies in this correlation function are more pronounced than those of the corresponding single - spin correlation function eq .
( [ eq : sstrans ] ) .
therefore , we call the correlator ( [ nematic ] ) ` nematic ' .
furthermore , we will refer to a situation where eq .
( [ nematic ] ) is the slowest decaying one among _ all _ correlation functions as a ` nematic - like phase ' . by virtue of the exponential decay in ( [ eq : sstrans ] ) ,
the correlator ( [ nematic ] ) is proportional to : @xmath61 the term @xmath62 appearing in the case of the @xmath63 zig - zag ladder corresponds to the operator @xmath64 in the case of a @xmath65 chain .
one can think of an effective @xmath65 spin formed from two neighboring @xmath63 spins coupled by the ferromagnetic interaction .
a similar behavior of correlation functions , namely the exponential decay of in - plane spin components and the algebraic decay of their bilinear combinations , is encountered also in the @xmath66 phase of the anisotropic @xmath65 chain@xcite and in the spin-1 chain with biquadratic interactions , see , e.g. , ref . .
the algebraic decay of the nematic correlator as opposed to the exponential decay of ( [ eq : sstrans ] ) suggests that there are tendencies towards nematic ordering in this phase .
depending on the value of @xmath40 the dominant instabilities are either spin - density - wave ones for @xmath67 or nematic ones for @xmath68 . from the result for @xmath40 given in eq .
( [ ll ] ) one can perturbatively evaluate the crossover value of @xmath5 : @xmath69 the nematic correlator ( [ nematic ] ) is the slowest decaying one , _
i.e. _ one is in the nematic - like phase .
the behavior of the cross - over line can be read off from the behavior of @xmath26 : @xmath26 increases monotonically with @xmath23 , tends to @xmath70 for @xmath71 , and satisfies @xmath72 for @xmath73 ( see , e.g. , refs . ) .
therefore , we have @xmath74 for @xmath29 with increasing ferromagnetic @xmath75 for decreasing @xmath24 .
this means that for @xmath76 a regime opens at high @xmath24 where nematic correlations given by eq .
( [ nematic ] ) dominate over spin - density - wave correlations given by eq .
( [ zzcor ] ) , in agreement with chubukov s prediction.@xcite , and magnetization @xmath77 : ( a ) longitudinal component @xmath78 , ( b ) transverse component @xmath79 , ( c ) spin nematic @xmath80 .
@xmath10 is the distance in a single - chain notation .
ed results for periodic boundary conditions are shown by symbols , fits by lines .
note the logarithmic scale of the vertical axis in panel ( b ) . ]
now we check the correlation functions obtained within bosonization against exact diagonalization ( ed ) results .
numerical data obtained for @xmath81 and @xmath77 on finite systems with periodic boundary conditions are shown in fig .
[ fig : correl ] .
this parameter set allows for a clear test of the above predictions , but represents the generic behavior in the phase of two weakly coupled chains .
to take into account finite - size effects we use the observation that for a conformally invariant theory , any power law on a plane becomes a power law in the following variable defined on a cylinder of circumference @xmath82 : @xmath83 first we fit the nematic correlator given by eq .
( [ nematic ] ) , which from bosonization is expected to be the leading instability at high magnetizations .
using the part with @xmath84 of the @xmath85 data shown in fig .
[ fig : correl]c , we find @xmath86 , @xmath87 , and @xmath88 . fig .
[ fig : correl]c shows that all finite - size results for the nematic correlator are nicely described by this fit with the dependence on @xmath82 taken into account by substituting eq .
( [ eq : confmap ] ) for the power laws .
moreover , from @xmath68 , we see that the system is indeed in the region dominated by nematic correlations for @xmath77 and @xmath89 .
now we turn to the longitudinal correlation function which we fit to the bosonization result eq .
( [ zzcor ] ) .
since most numerical parameters have been determined by the previous fit , only one free parameter is left which we determine from the numerical results of fig .
[ fig : correl]a for @xmath85 and @xmath90 as @xmath91 .
predictions for other system sizes are again obtained by substituting eq .
( [ eq : confmap ] ) for the power laws .
the agreement in fig .
[ fig : correl]a is not as good as in fig .
[ fig : correl]c
. however , it improves at larger distances @xmath10 and system sizes @xmath82 , indicating that corrections omitted in eq .
( [ zzcor ] ) are still relevant on the length scales considered here .
finally , the @xmath53-correlation function is shown in fig . [
fig : correl]b with a logarithmic scale of the vertical axis of this panel . the exponential decay predicted by eq .
( [ eq : sstrans ] ) is verified .
one further observes that correlations between the in - plane spin - operators belonging to different chains ( odd @xmath10 ) are an order of magnitude smaller than on the same chain ( even @xmath10 ) .
this suppression of correlations between different chains corresponds to the @xmath92 symbol in ( [ eq : sstrans ] ) , which strictly applies only in the thermodynamic limit and for large distances .
we summarize the main result of this section : in - plane spin correlators are exponentially suppressed for any finite value of the magnetization in the parameter region @xmath93 .
the ground state crosses over from a spin - density - wave dominated to a nematic - like phase with increasing magnetic field , with the crossover line given by eq .
( [ crossover ] ) .
we next address the excitation spectrum .
since the gap to @xmath57 excitations should be directly accessible to microscopic experimental probes such as inelastic neutron scattering or nuclear magnetic resonance , we analyze its behavior as a function of magnetization . sufficiently below the fully polarized state
the gap can be calculated analytically using results from sine - gordon theory .
in addition , to leading order of the interchain coupling , one can get qualitative expressions using dimensional arguments for the perturbed conformally invariant model : @xmath94^{\frac{1}{\nu } } \ , , \ ] ] where @xmath95 .
@xmath96 and @xmath97 can be determined numerically from the bethe ansatz integral equations.@xcite with this information and eqs .
( [ mag ] ) and ( [ gap ] ) we determine the qualitative behavior of the single - spin gap @xmath56 as a function of @xmath24 : it increases from zero at zero magnetization , reaches a maximum at intermediate magnetization values , then shows a minimum and , upon approaching the saturation magnetization , it increases again . as our formulas do not strictly apply at @xmath98 , the notion of a vanishing gap at zero magnetization
may be a spurious result .
note that when the fully polarized state is approached , the magnetization increases in an unphysical fashion since in this limit bosonization becomes inapplicable . at the point where the magnetization saturates the exact value of the gap can be obtained from the following mapping to hard - core bosons : @xcite @xmath99 comparing eq .
( [ holstein ] ) with eq .
( [ bosfor ] ) one recognizes the leading terms in haldane s harmonic fluid transformation for bosons.@xcite using a ladder approximation which is exact in the two - magnon subspace we arrive at : @xmath100 in eq .
( [ binding ] ) we have represented the gap as a difference of two terms : the quantum and the classical instability fields emphasizing its _ quantum origin_. as a function of magnetization @xmath24 .
panel ( a ) shows the single - spin excitation gap ( [ eq : ssgap ] ) , panel ( b ) the finite - size gap ( [ eq : tsgap ] ) for two flipped spins multiplied by the chain length @xmath82 . ] in order to verify these field theory predictions , we perform complementary numerical computations using the dmrg method . @xcite open boundary conditions are imposed and we typically keep up to 400 dmrg states . from dmrg
we obtain the ground - state energies @xmath101 as a function of total @xmath102 . for those values of @xmath102 that emerge as a ground state in an external magnetic field
we compute the single - spin excitation gap from @xmath103 fig .
[ fig : gap]a shows numerical results for @xmath104 at a selected value of @xmath105 for the largest system sizes investigated .
we find that the finite - size behavior of the gap @xmath106 for system sizes @xmath107 is well described by a @xmath108 correction .
this will be further corroborated by field - theoretical arguments outlined below .
therefore , we extrapolate it to the thermodynamic limit using a fit to the form @xmath109 allowing for an additional @xmath110 correction for those values of @xmath24 where at least 4 different system sizes are available .
this extrapolation is represented by the full circles in fig .
[ fig : gap]a ; errors are estimated not to exceed the size of the symbols . our extrapolation for @xmath104 is consistent with a vanishing gap at @xmath111 in agreement with previous numerical studies @xcite although bosonization predicts a non - zero possibly very small gap .
@xcite the behavior of @xmath56 confirms the picture described above : the gap is non - zero for @xmath112 , goes first through a maximum and then a minimum and finally approaches @xmath113 given by eq .
( [ binding ] ) for @xmath114 .
we further wish to point out that for chains with periodic boundary conditions , the coefficient @xmath115 of the finite - size extrapolation eq .
( [ eq : extra ] ) is determined by the spin - wave velocity and the critical exponent of the soft mode from the @xmath47 channel . indeed , using eq .
( [ eq : sstrans ] ) where we can set @xmath116 , and use the conformal mapping ( [ eq : confmap ] ) to the cylinder , we see that the leading finite - size correction to the gap is : @xmath117 note that we have to replace sin with sinh in eq .
( [ eq : confmap ] ) in order to extract a gap , since we are dealing with euclidean time .
in addition we used the fact that in our approximation the effective hamiltonian ( [ symantisym ] ) is a direct sum of symmetric and antisymmetric sectors .
moreover , it is only the symmetric sector enjoying conformal invariance and consequently we perform the replacement @xmath118 only in the symmetric sector .
the antisymmetric sector has a spectral gap and its contribution to the finite - size corrections of the single - spin flip excitation energy are exponentially suppressed with system size.@xcite with this method one can not fix the amplitudes of the @xmath110 term and beyond .
note furthermore that there may be additional surface terms for open boundary conditions as employed in the numerical dmrg computations .
nevertheless there is a dominant @xmath108 correction in any case .
next , we briefly look at the @xmath47 excitations .
their finite - size gap is , in analogy to eq .
( [ eq : ssgap ] ) , computed with dmrg from @xmath119 fig .
[ fig : gap]b shows numerical results for @xmath120 again at the value @xmath105 .
one observes that the scaled finite - size gaps collapse onto a single curve which shows that @xmath121 scales linearly to zero with @xmath108 , exactly as expected for gapless excitations in 1d .
furthermore , we observe that the scaled quantity @xmath120 vanishes as one approaches saturation @xmath29 which indicates a vanishing of the velocity of the corresponding excitations at saturation . computed for @xmath122 and @xmath123 .
the wave vector is given relative to the ground state wave vector ( 0 for @xmath124 and @xmath125 for @xmath126 ) . ]
we proceed by discussing the wave - vector dependence of the @xmath127 excitation , while we remind the reader that the low - energy excitations are in the @xmath47 sector .
[ fig : dispersion ] shows representative ed results obtained for rings with @xmath122 and @xmath123 . for ground states with low @xmath102
, the @xmath127 excitation spectrum looks similar to the continuum of spinons . on the other hand , close to saturation one has single - magnon excitations with a minimum given by the classical value of the wave vector @xmath128.@xcite we read off from fig .
[ fig : dispersion ] that upon lowering the magnetic field , this minimum shifts from the classical incommensurate value towards @xmath129 ,
_ i.e. _ the value appropriate for two decoupled chains .
this renormalization of the minimum of the magnon excitations towards the value of decoupled chains can be interpreted in terms of quantum fluctuations , which are enhanced when the density of magnons increases .
a strong quantum renormalization of the pitch angle from its classical value at zero magnetization was previously observed by the coupled - cluster method and dmrg calculations.@xcite
we have combined numerical techniques with analytical approaches and mapped out the ground state phase diagram of the frustrated ferromagnetic spin chain in an external magnetic field .
we have established that with increasing magnetic field , the ground state crosses over from a spin - density - wave dominated to a nematic - like phase .
single spin flip excitations are gapped , giving rise to an exponential decay of in - plane spin correlation functions in both regimes .
we have studied the single- and two - spin flip excitation energy numerically . using tools from conformal field theory
we have further shown that the amplitude of the leading @xmath108 correction term to the single - spin flip gap is determined by the critical exponent and the spin - wave velocity of the soft mode . finally , in order to apply our findings to the material licuvo@xmath3
, one should take into account interchain interactions as well as anisotropies , which are expected to be present in this system.@xcite at low fields , a helical state has been observed experimentally.@xcite on the other hand , for the purely one dimensional case , we have shown that upon increasing the magnetic field there is a competition between spin - density - wave and nematic - like tendencies .
those are the two leading instabilities at high magnetizations and thus they are the natural candidates to become long - range ordered in higher dimensions .
the question whether there are true phase transitions at high fields in higher dimensions is beyond the scope of the current work .
we thank a. feiguin for providing us with his dmrg code used for large scale calculations .
most of t.v.s work was done during his visits to the institutes of theoretical physics at the universities of hannover and gttingen , supported by the deutsche forschungsgemeinschaft .
the hospitality of the host institutions is gratefully acknowledged .
. also acknowledges support from the georgian national science foundation under grant n @xmath130 .
lptms is a mixed research unit 8626 of cnrs and university paris - sud .
is supported by the deutsche forschungsgemeinschaft ( project no . ho 2325/4 - 1 ) , and f.h .- m .
is supported by nsf grant no .
dmr-0443144 .
m. enderle , c. mukherjee , b. fk , r. k. kremer , j .-
broto , h. rosner , s .- l
. drechsler , j. richter , j. malek , a. prokofiev , w. assmus , s. pujol , j .- l .
raggazzoni , h. rakoto , m. rheinstdter , and h. m. rnnow , europhys . lett . * 70 * , 237 ( 2005 ) . | magnetic field effects on the one - dimensional frustrated ferromagnetic chain are studied by means of effective field theory approaches in combination with numerical calculations utilizing lanczos diagonalization and the density matrix renormalization group method .
the nature of the ground state is shown to change from a spin - density - wave region to a nematic - like one upon approaching the saturation magnetization .
the excitation spectrum is analyzed and the behavior of the single spin - flip excitation gap is studied in detail , including the emergent finite - size corrections . |
within the last decades , model predictive control ( mpc ) has grown mature for both linear and nonlinear systems , see , e.g. , @xcite or @xcite .
although analytically and numerically challenging , the method itself is attractive due to its simplicity : in a first step , a new measurement of the current system state is obtained which is thereafter used to compute an optimal control over a finite optimization horizon . in the third and last step , a portion of this control is applied to the process and the entire problem is shifted forward in time rendering the scheme to be iteratively applicable .
stability of the mpc closed loop can be shown by imposing endpoint constraints , lyapunov type terminal costs or terminal regions , cf .
@xcite and @xcite . here , we study mpc schemes without these ingredients for which stability and , in addition , bounds on the required horizon length can be deduced , both for linear and nonlinear systems , cf .
@xcite and @xcite .
we follow the recent approach from @xcite extending @xcite to continuous time systems which not only guarantees stability but also reveals an estimate on the degree of suboptimality with respect to the optimal controller on an infinite horizon . in this work ,
we show how the essential assumption needed to apply the methodology proposed in @xcite can be practically verified . then , based on observations drawn from numerical computations , implementable mpc algorithms with variable control horizons are developed which allow for smaller optimization horizons while maintaining stability or a desired performance bound . to overcome the lack of robustness implied by prolonging the control horizon and , thus , staying in open loop for longer time intervals , conditions are presented which ensure that the control loop can be closed more often .
similar ideas were introduced in @xcite for a discrete time setting .
last , the computational effort is further reduced by introducing slack which allows to violate our main stability condition a relaxed lyapunov inequality temporarily , cf .
@xcite .
the paper is organized as follows : in section [ section : preliminaries ] the problem formulation is given . in the ensuing section [ section : stability and performance bounds ] ,
we summarize stability results from @xcite and propose a technique to verify the key assumption which is illustrated by an example .
thereafter , we present algorithms which allow for shortening the optimization horizon by using time varying control horizons . before drawing conclusions , section [ section : slack stability ] contains results on how stability
may be guaranteed by using weaker stability conditions .
let @xmath0 and @xmath1 denote the set of natural and real numbers respectively and @xmath2 the euclidean norm on @xmath3 , @xmath4 . a continuous function @xmath5 is called class @xmath6-function if it satisfies @xmath7 , is strictly increasing and unbounded .
a continuous function @xmath8 is said to be of class @xmath9 if for each @xmath10 we have that @xmath11 holds , and for each @xmath12 the condition @xmath13 is satisfied . within this work we consider nonlinear time invariant control systems @xmath14 where @xmath15 and @xmath16 denote the state and control at time @xmath12 .
constraints can be included via suitable subsets @xmath17 and @xmath18 of the state and control space , respectively .
we denote a state trajectory which emanates from the initial state @xmath19 and is subject to the control function @xmath20 by @xmath21 . in the presence of constraints ,
a control function @xmath22 is called admissible for @xmath23 on the interval @xmath24 if the corresponding solution @xmath25 exists and satisfies @xmath26 , \quad \ ; \text{and } \ ;
\quad \u(t ) \in \bu , t \in [ 0,t).\end{aligned}\ ] ] the set of these admissible control functions is denoted by @xmath27 . for an infinite time interval , @xmath28
is called admissible for @xmath23 if @xmath29 holds for each @xmath30 and the respective set is denoted by @xmath31 . for system
we assume an equilibrium @xmath32 to exist , i.e. @xmath33 holds .
our goal is to design a feedback control law @xmath34 such that the resulting closed loop is asymptotically stable with respect to @xmath35 , i.e. there exists @xmath36 such that @xmath37 , @xmath12 , holds for all @xmath38 where @xmath39 denotes the closed loop trajectory induced by @xmath40 .
the stabilization task is to be accomplished in an optimal fashion which is measured by a cost functional . to this end
, we introduce the continuous running cost @xmath41 satisfying @xmath42 then , for a given state @xmath43 , the cost of an admissible control @xmath44 is @xmath45 the computation of a corresponding minimizer is , in general , computationally hard due to the curse of dimensionality , cf .
hence , we use model predictive control ( mpc ) to approximately solve this task .
the central idea of mpc is to truncate the infinite horizon , i.e. to compute a minimizer @xmath46 of the cost functional @xmath47 which can be done efficiently using discretization methods and nonlinear optimization algorithms , see , e.g. , @xcite or ( * ? ? ?
* chapter 10 ) .
furthermore , we define the corresponding optimal value function @xmath48 , @xmath49 . to obtain an infinite horizon control ,
only the first portion of the computed minimizer is applied , i.e. we define the feedback law via @xmath50 for the so called control horizon @xmath51 .
last , the optimal control problem is shifted forward in time which renders algorithm [ preliminaries : alg : mpc ] to be iteratively applicable . given : @xmath52 1 .
measure the current state @xmath53 2 . compute a minimizer @xmath54 of and define the mpc feedback law via 3 .
implement @xmath55 , shift the horizon forward in time by @xmath56 and goto ( 1 ) the closed loop state trajectory emanating from the initial state @xmath57 subject to the mpc feedback law @xmath58 from algorithm [ preliminaries : alg : mpc ] is denoted by @xmath59 .
furthermore , @xmath60 denotes the control function obtained by concatenating the applied pieces of control functions , i.e. @xmath61 the resulting mpc closed loop cost are given by @xmath62 note that we tacitly assume that problem is solvable for all @xmath63 and the minimum is attained in each step of algorithm [ preliminaries : alg : mpc ] . for a detailed discussion of feasibility issues
we refer to ( * ? ? ?
* chapter 8) .
due to the truncation of the infinite horizon , stability and optimality properties of the optimal control may be lost . yet , stability can be shown if the optimization horizon is sufficiently long , cf .
additionally , an optimization horizon length @xmath64 can be determined for which both asymptotic stability as well as a performance bound on the mpc closed loop in comparison to the infinite horizon control law hold .
[ stability and performance bounds : thm : stability ] suppose a control horizon @xmath65 and a monotone bounded function @xmath66 satisfying @xmath67 for all @xmath43 to be given .
if @xmath68 is chosen such that @xmath69 holds for @xmath70 \left [ 1 - e^{-\int_{t-\delta}^t b(t)^{-1 } dt } \right ] } , % \label{\hspace*{-0.1cm}\alpha_{t , \delta } : = 1 - \frac{e^{-\int_\delta^t b(t)^{-1 } dt}}{\left [ 1- e^{- \int_\delta^t b(t)^{-1 } dt } \right ] \left [ e^{\int_{t-\delta}^t b(t)^{-1 } dt } -1 \right ] } , \end{aligned}\ ] ] then the relaxed lyapunov inequality @xmath71 as well as the performance estimate @xmath72 are satisfied for all @xmath43 .
if , additionally , there exist @xmath6 functions @xmath73 , @xmath74 such that @xmath75 hold for all @xmath43 , then the mpc closed loop is asymptotically stable for horizon length @xmath64 . a detailed proof of theorem [ stability and performance bounds : thm : stability ] is given in @xcite .
we point out that for a given control horizon @xmath65 and a desired performance specification @xmath76 on the mpc closed loop , there always exists an optimization horizon @xmath68 such that @xmath77 is satisfied , cf .
@xcite . interpreting exemplarily
, the choice @xmath78 corresponds to asymptotic stability of the mpc closed loop whereas @xmath79 limits the cost of the mpc control to double the cost of the infinite horizon control .
the crucial assumption which has to be verified in order to apply theorem [ stability and performance bounds : thm : stability ] is the growth condition . here
, we demonstrate this by means of the following example : [ stability and performance : ex : generator ] consider the system dynamics of a synchronous generator given by @xmath80 with constants @xmath81 , @xmath82 , @xmath83 , @xmath84 , @xmath85 and @xmath86 , cf .
@xcite . the equilibrium we wish to stabilize
is located at @xmath87 and the running costs are defined as @xmath88 where the parameter @xmath89 is used to penalize the taken control effort . due to physical considerations @xmath90 and @xmath91 are restricted to the interval @xmath92 $ ] and @xmath93 respectively . for example
[ stability and performance : ex : generator ] , we want to determine an optimization horizon length @xmath64 such that asymptotic stability of the resulting mpc closed loop is guaranteed , i.e. @xmath78 . to achieve this goal , the following methodology can be applied to compute a function @xmath94 such that holds : 1 . for each state @xmath43 derive a monotone function @xmath95 satisfying @xmath96 2 .
then , @xmath97 is defined pointwise as @xmath98 . to accomplish step ( 1 ) we exploit the fact that optimality of this bound is not needed .
hence , for each @xmath99 for some @xmath100 we solve @xmath101 over the class of sampled data systems with zero order hold and sampling period @xmath102 , i.e. control functions satisfying @xmath103 for @xmath104 , @xmath105 .
this yields monotone bounds @xmath106 , @xmath107 , satisfying by dividing by @xmath108 .
then , step ( 2 ) is carried out in order to compute @xmath97 .
since neither terminal constraints nor costs were imposed , @xmath109 is monotonically increasing in @xmath110 which implies @xmath111 , @xmath112 $ ] .
note that @xmath113 is only computed on a `` sufficiently '' large interval @xmath114 $ ] , @xmath115 , such that a desired performance estimate can be concluded by theorem [ stability and performance bounds : thm : stability ] . while the proposed procedure allows to verify step ( 1 ) rigorously ,
we carry out step ( 2 ) only approximately .
the reason is twofold : firstly , throughout this paper we assume feasibility for the set @xmath116 , i.e. @xmath116 is control invariant , cf .
@xcite for a definition of control invariance .
hence , a suitable subset has to be computed a priori if this assumption is violated , see ( * ? ? ?
* chapter 8) .
for example [ stability and performance : ex : generator ] , numerical computations show that an appropriately chosen level set @xmath117 of the value function @xmath118 is control invariant . secondly , this set has to be discretized . to this end
, a grid of initial values @xmath119 with discretization stepsize @xmath120 within each direction @xmath121 is used .
then , @xmath97 is approximately determined for the set @xmath117 by taking the supremum with respect to all states contained in the intersection @xmath122 . in conclusion
, the proposed approach allows to rigorously ensure the relaxed lyapunov inequality except for the state space discretization assuming that a control invariant set is given .
[ stability and performance : ex : bound ] regarding example [ stability and performance : ex : generator ] , let @xmath123 and the control horizon @xmath124 be given .
additionally , we define @xmath125 \times [ \xs_2 - a_2 , \xs_2 + a_2 ] \times [ \xs_3 - a_3 , \xs_3 + a_3 ] $ ] with @xmath126 , @xmath127 , @xmath128 and discretization stepsize @xmath129 , @xmath130 , and focus on the level set @xmath131 whose convex hull satisfies all physical constraints , see figure [ preliminaries : fig : set of initial values ] for an illustration of @xmath122 .
.,scaledwidth=42.0% ] then , formula enables us to determine an optimization horizon @xmath64 such that the relaxed lyapunov inequality holds with @xmath132 . for the computed function
@xmath113 this methodology yields asymptotic stability of the mpc closed loop for optimization horizon @xmath133 .
note that the presented method verifies condition for control functions @xmath134 , and allows to conclude asymptotic stability of the closed loop via theorem [ stability and performance bounds : thm : stability ] , cf .
* remark 2.7 ) .
in section [ section : stability and performance bounds ] we showed how to ensure asymptotic stability for the proposed mpc scheme for a given @xmath56 . in this section
, we investigate the impact of @xmath56 on the required optimization horizon length @xmath64 .
considering example [ stability and performance : ex : generator ] , we compute @xmath135 , @xmath136 , and @xmath137 for @xmath138 , @xmath139 , cf . figure [ increasing the control horizon : fig : horizons]a . here
, @xmath140 holds for @xmath141 whereas this stability criterion holds for significantly shorter optimization horizons if the control horizon is chosen equal to @xmath142 , that is @xmath143 for @xmath144 .
hence , enlarging the control horizon seems to induce an improved performance index @xmath145 .
\a ) for varying @xmath64 and different choices of @xmath56 .
b ) development of @xmath146 depending on the control horizon @xmath56.,title="fig:",scaledwidth=21.65% ] b ) for varying @xmath64 and different choices of @xmath56 .
b ) development of @xmath146 depending on the control horizon @xmath56.,title="fig:",scaledwidth=21.65% ] this numerical result motivates to investigate the influence of @xmath56 on @xmath145 . to this end
, we fix the optimization horizon @xmath64 and compute @xmath147 for @xmath148 , cf .
figure [ increasing the control horizon : fig : horizons]b , which leads to the following observations : firstly , a symmetry property seems to hold , i.e. @xmath149 .
secondly , the performance estimates appear to increase up to the symmetry axis @xmath142 . both properties have been shown for systems which are exponentially controllable in terms of their stage costs , i.e. @xmath150 for an overshoot constant @xmath151 and a decay rate @xmath152 , cf .
@xcite . using the computed function @xmath113 instead of exponential decay
we obtain @xmath153 and therefore better horizon estimates as shown in @xcite for a discrete time setting . despite the more general setting
, symmetry still follows directly from formula .
[ increasing the control horizon : cor : symmetry ] the performance estimate @xmath145 given by formula satisfies @xmath154 for @xmath51 , i.e. @xmath145 is symmetric with symmetry axis @xmath155 . unlike symmetry , we conjecture that there exists a counterexample negating monotonicity even if @xmath113 satisfies the additional condition ( * ? ? ?
* inequality ( 21 ) ) which is , however , violated in example [ stability and performance : ex : generator ] .
for such an example holds but @xmath146 is not monotone on @xmath156 $ ] , cf .
@xcite for a counterexample in the discrete time setting . yet , for example [ stability and performance : ex : generator ] numerical results indicate that the monotonicity property holds and allows to conclude asymptotic stability of the mpc closed loop for significantly shorter optimization horizons .
[ increasing the control horizon : ex : generator2 ] again consider example [ stability and performance : ex : generator ] .
if we impose the horizon @xmath157 , then from figure [ increasing the control horizon : fig : horizons]b we observe that stability of the closed loop can only be guaranteed if @xmath158 $ ] . if the horizon length is increased to @xmath159 , figure [ increasing the control horizon : fig : horizons]a shows that for any chosen @xmath160 $ ] stability can be concluded . in
the context of arbitrary monotone and bounded functions @xmath161 another interesting fact arises if we consider arbitrarily small @xmath56 : [ increasing the control horizon : thm : divergence ] for any optimization horizon @xmath162 we have that @xmath145 goes to @xmath163 for @xmath164 . in particular , our stability condition @xmath165 can not be maintained for arbitrarily small control horizon @xmath56 .
* proof * : follows directly from formula .
note that this assertion was solely shown for an exponentially controllable setting both for discrete and continuous time systems , cf .
* section 4 ) and ( * ? ? ?
* section 5.1 ) .
hence , the key contribution is the observation that this conclusion can also be drawn without the restriction to this particular class of systems .
motivated by example [ increasing the control horizon : ex : generator2 ] and theorem [ increasing the control horizon : thm : divergence ] , we propose an algorithm to obtain a control horizon length @xmath56 such that @xmath146 exceeds a predefined suboptimality bound @xmath166 . to this end
, we introduce a partition @xmath167 , @xmath168 , of @xmath169 $ ] with @xmath170 .
such a setting naturally arises in the context of digital control for sampled data systems with zero order hold .
yet , we like to note that algorithm [ increasing the control horizon : alg : control horizon ] is not limited to the digital control case .
additionally , we like to stress that monotonicity of @xmath146 in @xmath56 is the center of this algorithm . given : @xmath162 , @xmath167 with @xmath168 and @xmath171 1 .
measure the current state @xmath53 2 .
set @xmath172 and compute a minimizer @xmath173 of and @xmath174 .
+ do 1 . if @xmath175 : set @xmath56 according to exit strategy and goto ( 3 ) 2 .
set @xmath176 , @xmath177 and compute @xmath178 3 .
compute @xmath146 , i.e. @xmath179 + while @xmath180 3 . implement @xmath181 , shift the horizon forward in time by @xmath56 and goto ( 1 ) algorithm [ increasing the control horizon : alg : control horizon ] combines two aspects : the improved performance estimates obtained for larger control horizons @xmath56 , and the inherent robustness resulting from using a feedback control law which benefits from updating the control law as often as possible . in order to illustrate this claim
let us consider the numerical example [ stability and performance : ex : generator ] again . from figure
[ increasing the control horizon : fig : horizons]b we observed that asymptotic stability of the mpc closed loop can be shown for @xmath182 .
indeed , algorithm [ increasing the control horizon : alg : control horizon ] only uses @xmath183 in each step independent of the chosen initial condition .
hence , mpc with constant control horizon @xmath124 is performed safeguarded by our theoretically obtained estimates .
consequently , no exit strategy is needed since step ( 2a ) is excluded .
we compute @xmath145 by equation for all @xmath184 in order to test out the limits of algorithm [ increasing the control horizon : alg : control horizon ] . here
, @xmath185 is the smallest optimization horizon such that @xmath186 is satisfied for all @xmath184 .
however , using algorithm [ increasing the control horizon : alg : control horizon ] allows to ensure this conditions for @xmath187 for varying control horizon @xmath188 .
figure [ increasing the control horizon : fig : ex : n5 ] shows the sets of states @xmath184 for which @xmath146 computed by is less than zero for the cases @xmath189 and @xmath190 .
again , we see that if we allow for larger control horizons @xmath56 , then the performance bound @xmath146 increases , i.e. the set of state vectors for which stability can not be guaranteed shrinks . hence , for the considered example [ stability and performance : ex : generator ] algorithm [ increasing the control horizon : alg : control horizon ] allows to drastically reduce the horizon @xmath64 while maintaining asymptotic stability of the closed loop .
\a ) for which stability can not be guaranteed for @xmath185 , @xmath136 with ( a ) @xmath189 and ( b ) @xmath191.,title="fig:",scaledwidth=21.65% ] b ) for which stability can not be guaranteed for @xmath185 , @xmath136 with ( a ) @xmath189 and ( b ) @xmath191.,title="fig:",scaledwidth=21.65% ] the downside of considering potentially large control horizons @xmath56 is the possible lack of robustness in case of disturbances . utilizing the introduced partition @xmath167 , we can perform an update of the feedback law at time @xmath192 via @xmath193 where we extended the notation of the open loop optimal control to @xmath194 to indicate which initial state is considered .
such an update can be applied whenever the following lyapunov type update condition holds , see also @xcite for the discrete time setting . [ increasing the control horizon : thm : update ]
let @xmath171 , @xmath162 and a partition @xmath167 , @xmath105 , of @xmath169 $ ] with @xmath170 be given .
for @xmath195 suppose holds with @xmath196 for some @xmath43 .
let @xmath197 be a minimizer of for some @xmath198 .
if additionally @xmath199 holds , then the respective mpc feedback law @xmath200 can be modified by and the lower bound @xmath166 on the degree of suboptimality is locally maintained .
we like to stress that if the stabilizing partition index @xmath201 is known , then proposition [ increasing the control horizon : thm : update ] allows for iterative updates of the feedback until this index is reached .
hence , only step ( 3 ) of algorithm [ increasing the control horizon : alg : control horizon ] needs to be adapted , see algorithm [ increasing the control horizon : alg : update ] for a possible implementation . 1 .
for @xmath202 do 1 .
implement @xmath203 2 .
compute @xmath204 3 .
if holds : update @xmath205 via + shift the horizon forward in time by @xmath56 and goto ( 1 )
usually , the relaxed lyapunov inequality is tight for only a few points in the state space .
hence , if a closed loop trajectory @xmath206 visits a point for which is not an equality , then we can compute the occuring slack along the closed loop via @xmath207 this slack can be used to weaken the requirement by considering the closed loop instead of the open loop solution . for simplicity of exposition , we formulate the following result using constant @xmath56 , yet the conclusion also holds in the context of time varying control horizons as in algorithm [ increasing the control horizon : alg : control horizon ] .
[ slack stability : thm : convergence ] consider an admissible feedback law @xmath208 , an initial value @xmath209 and @xmath210 to be given . furthermore , suppose there exist @xmath211 functions @xmath73 , @xmath74 such that @xmath212 and @xmath213 hold for all @xmath43 .
if additionally @xmath214 from converges for @xmath110 tending to infinity , then the mpc closed loop trajectory with initial value @xmath57 behaves like an asymptotically stable solution .
furthermore , the following performance estimate holds @xmath215 * proof * : let the limit of @xmath216 for @xmath217 be denoted by @xmath218 .
then , for given @xmath219 , there exists a time instant @xmath220 such that @xmath221 holds for all @xmath222 . as a consequence ,
we obtain @xmath223 since @xmath224 as well as @xmath225 hold , this inequality implies @xmath226 hence , boundedness of the integral on the right hand side can be concluded .
now , due to positivity and continuity of @xmath227 we have @xmath228 . in turn
, the latter ensures @xmath229 and , thus , @xmath230 for @xmath110 approaching infinity .
inequality is shown by using @xmath231 in combination with to obtain @xmath232 then , since @xmath233 , the assertion follows .
note that within theorem [ slack stability : thm : convergence ] we did not assume semipositivity but convergence of @xmath214 to conclude stability . here , we like to stress that @xmath234 implies a suboptimality index @xmath235 . clearly
, both the stability and performance result shown in theorem [ slack stability : thm : convergence ] can be extended to assertions for all @xmath43 if @xmath236 converges for every @xmath43 or a uniform lower bound can be found , i.e. @xmath237 . apart from its theoretical impact , @xmath236 is also meaningful at runtime of the mpc algorithm .
for instance , the condition @xmath238 can be checked at each time instant @xmath239 , @xmath105 , instead of @xmath77 , cf . .
this is particularly useful since accumulated slack can be used in order to compensate local violations of @xmath240 , i.e. weakening the stability condition , as long as the overall performance is still satisfactory .
if @xmath241 occurs within the mpc algorithm , the slack can also be used to form an exit strategy . to this end
, we denote the performance of the mpc closed loop until time @xmath110 by @xmath242 now , if @xmath243 but @xmath244 , then stability is still maintained , yet the current performance index is worse than the desired bound @xmath166 .
again we consider example [ stability and performance : ex : generator ] and analyze the mpc closed loop for @xmath58 with @xmath187 and @xmath183 .
if we choose the initial value @xmath245 , then we observe from figure [ slack stability : fig : alpha ] that the local estimate @xmath146 drops below zero for @xmath246 , i.e. stability can not be guaranteed for @xmath189 .
yet , computing @xmath247 according to shows that the relaxed lyapunov inequality is satisfied after two steps of the mpc algorithm .
hence , using the slack information incorporated in @xmath247 allows to conclude asymptotic stability . and @xmath247 for @xmath187 and @xmath124 along a specific closed loop solution.,scaledwidth=42.0% ]
we have shown a methodology to verify the assumptions introduced in @xcite under which stability of the mpc closed loop without terminal constraints can be guaranteed .
additionally we presented an algorithmic approach for varying control horizons which allows to reduce the optimization horizon length .
last , we provided robustification methods via update and slack based rules .
this work was supported by the dfg priority research program 1305 `` control theory of digitally networked dynamical systems '' , grant .
gr1569/12 - 1 , and by the german federal ministry of education and research ( bmbf ) project `` snimored '' , grant no .
03ms633 g .
p. giselsson .
daptive nonlinear model predictive control with suboptimality and stability guarantees . in _ proceedings of the 49th conference on decision and control , atlanta , ga ,
usa _ , pages 36443649 , 2010 .
s. s. keerthi and e. g. gilbert .
optimal infinite horizon feedback laws for a general class of constrained discrete - time systems : stability and moving horizon approximations .
_ journal of optimization theory and applications _ , 57:0 265293 , 1988 . | : the stability analysis of model predictive control schemes without terminal constraints and/or costs has attracted considerable attention during the last years .
we pursue a recently proposed approach which can be used to determine a suitable optimization horizon length for nonlinear control systems governed by ordinary differential equations . in this context
, we firstly show how the essential growth assumption involved in this methodology can be derived and demonstrate this technique by means of a numerical example .
secondly , inspired by corresponding results , we develop an algorithm which allows to reduce the required optimization horizon length while maintaining asymptotic stability or a desired performance bound .
last , this basic algorithm is further elaborated in order to enhance its robustness properties .
lyapunov stability ; predictive control ; state feedback ; suboptimal control |
the ridge structure in two - particle correlation has been studied in nuclear collisions at the relativistic heavy - ion collider ( rhic ) for several years @xcite and is now also seen in @xmath5 collisions at the large hadron collider ( lhc ) @xcite .
the nature of that structure is that it is narrow in @xmath6 ( azimuthal angle @xmath7 relative to that of the trigger ) but broad in @xmath0 ( pseudorapidity @xmath1 relative to the trigger ) . in ref .
@xcite the range in @xmath0 is found to be as large as 4 .
so far there is no consensus on the origin of the ridge formation .
it has been pointed out that the wide @xmath0 distribution implies long - range correlation @xcite .
that is a view based mainly on theoretical ideas without any comparison between the @xmath1 ranges of single - particle distribution and two - particle correlation .
we make that comparison here using only experimental data from phobos @xcite .
it is found that the large @xmath0 ridge is related simply to a shift of the pseudorapidity and an integral over the trigger @xmath1 .
that is a phenomenological observation without any theoretical input .
any successful model of ridge formation should be able to explain that relationship .
there are subtleties about the single - particle for all charges , @xmath8 , that to our knowledge has not been satisfactorily explained in all its details . since it sums over all charges , hadrons of different types are included , making @xmath8 to be quite different from @xmath9 .
that difference can not be readily accounted for in any simple hadronization scheme .
fortunately , detailed examination of @xmath8 is not required before we find its relationship to the ridge @xmath10 , since both are for unidentified charged hadrons , and the empirical verification is based on the data from the same experimental group ( phobos ) . after the phenomenological relationship between @xmath10 and @xmath8
is established in sec . 2 , we give an interpretation of the phenomenon in sec .
it is not our objective to give a review of all other models that can reproduce the data on the ridge structure and assess their likelihood to explain the empirical observation made in sec . 2 .
we offer only to show the possibility , contrary to conventional wisdom , that there is no need for long - range longitudinal correlation .
our conclusion is given in sec .
our focus is on the phobos data on two - particle correlation measured with a trigger particle having transverse momentum @xmath11 gev / c in au + au collisions at @xmath12 gev @xcite .
the pseudorapidity acceptance of the trigger is @xmath13 .
the per - trigger ridge yield integrated over @xmath14 , denoted by @xmath15 , includes all charged hadrons with @xmath16 mev / c at @xmath17 and @xmath18 mev / c at @xmath19 , where the superscript @xmath20 stands for associated particle in the ridge .
for simplicity we use the notation @xmath21 , @xmath22 , @xmath23 , @xmath24 , @xmath25 , @xmath26 .
since all ridge particles are included in the range @xmath14 , no further consideration is given to the @xmath7 dependence .
our aim is to relate the ridge distribution in @xmath0 to the single - particle distribution in @xmath1 .
we first make a phenomenological observation using only phobos data for both distributions . after showing their relationship
, we then make an interpretation that does not involve extensive modeling . to do meaningful comparison
, it is important to use single - particle @xmath1 distribution , @xmath8 , that has the same kinematical constraints as the ridge distribution .
that is , it involves an integration over @xmath27 and a sum over all charged hadrons @xmath28 where @xmath29 , and the lower limit of the @xmath27 integration is @xmath30 mev / c in keeping with the acceptance window of @xmath31 @xcite .
the data on @xmath15 are for 0 - 30% centrality .
phobos has the appropriate @xmath8 for 0 - 6% , 6 - 15% , 15 - 25% and 25 - 35% centralities @xcite , as shown in fig .
thus we average them over those four bins . the result is shown in fig .
1(b ) by the small circles for 0 - 30% centrality .
those points are fitted by the three gaussian s , located at @xmath32 and @xmath33 , @xmath34 + a_1\exp[-(\eta-\hat\eta)^2/{2\sigma_1 ^ 2 } ] + a_1\exp[-(\eta+\hat\eta)^2/{2\sigma_1 ^ 2}]\ } \label{2}\end{aligned}\ ] ] shown by the solid ( red ) line in that figure with @xmath35 .
the dashed line shows the central gaussian , while the dash - dotted line shows the two side gaussians .
the purpose of the fit is only to give an analytic representation of @xmath8 to be used for comparison with the ridge distribution .
note that @xmath36 is large @xmath37 , and thus stretches the width of the @xmath1 .
gev for ( a ) various centrality bins and ( b ) 0 - 30% centrality .
data are from ref .
the ( red ) line in ( b ) is a fit using eq .
( [ 2 ] ) , whose first term is represented by the dashed line and the other two terms by the dash - dotted line . ]
we now propose the formula @xmath38 where @xmath39 is a parameter that summarizes all the experimental conditions that lead to the magnitude of the ridge distribution measured relative to the single - particle distribution .
in particular , @xmath39 does not depend on @xmath40 or @xmath41 ; otherwise , the equation is meaningless in comparing the @xmath1 dependencies . there is no theoretical input in eq .
( [ 3 ] ) , except for the question behind the proposal : how much of the @xmath0 distribution can be accounted for by just a mapping of @xmath42 with a shift due to the definition @xmath23 , and a smearing due to the trigger acceptance , @xmath43 ?
another way of asking the question is : if the experimental statistics were high enough so that the trigger s @xmath1 range can be very narrow around @xmath44 , how much difference would there be between the ridge distribution in @xmath0 and the pseudorapidity distribution in @xmath1 ? ) using @xmath1 from fig
. 1 @xcite.,scaledwidth=100.0% ] the proposed formula in eq .
( [ 3 ] ) is tested by substituting the fit of @xmath8 according to eq .
( [ 2 ] ) into the integrand on the right - hand side .
the result is shown in fig .
2 with @xmath39 being adjusted to fit the height of the ridge distribution ; its value is @xmath45 .
the peak in the data around @xmath46 is , of course , due to the jet component associated with the trigger jet and is not relevant to our comparison here .
it is evident that the large @xmath0 distribution is well reproduced by eq .
( [ 3 ] ) . in qualitative terms
the width of the ridge distribution is due partly to the width of @xmath8 and partly to the smearing of @xmath40 , which adds another 1.5 to the width .
no intrinsic dynamics of long - range longitudinal correlation has been put in .
note that the center of the plateau in @xmath0 is at @xmath47 , which is the average of the shift due to @xmath40 being integrated from 0 to 1.5 .
it suggests that if @xmath40 were fixed at @xmath48 when abundant data become available , then the width of @xmath49 would be only as wide as that of the single - particle @xmath8 .
no theoretical prejudice has influence these observations .
we now consider an interpretation of what eq .
( [ 3 ] ) implies , given the empirical support for its validity from fig . 2 .
since the observed ridge integrates over trigger @xmath1 , we write it as = _ 0 ^ 1.5 d_1_h_2 dp_2 p_2 .
r^h_2(_1,_2,p_2 ) |__2 = _ 1 + , [ 4 ] where we exhibit also explicitly the sum over the hadron type of the ridge particle @xmath50 and the integral over its momentum , denoted by @xmath51 . according to the definition of correlation @xmath52
, we can express the per - trigger ridge correlation as r^h_2(_1,_2,p_2 ) = _
h_1 dp_1 p_1 _ 2^h_1h_2(b+r)(_1,p_1,_2,p_2)_1^h_1(_1,p_1 ) -_1^h_2(b)(_2,p_2 ) , [ 5 ] where @xmath53 and @xmath54 in the superscript denote background and ridge , respectively .
the jet component in the associated - particle is excluded in eq .
( [ 5 ] ) . on the other hand , with eq . ( [ 1 ] ) substituted into eq . ( [ 3 ] ) we have , using @xmath41 and @xmath51 instead of @xmath1 and @xmath27 , = _ 0 ^ 1.5 d_1_h_2 dp_2 p_2 .
1^h_2(_2,p_2 ) |__2 = _ 1 + .
[ 6 ] comparing eq .
( [ 6 ] ) to ( [ 4 ] ) we see that the ridge @xmath55 is to be related to the phenomenological quantity @xmath56 .
thus the crux of the relationship between the ridge and inclusive s involves the interpretation of @xmath57 . to that end
let us first write @xmath58 in the form _
1^h_2(_2,p_2)=dn^h_2d_2p_2dp_2=h^h_2(_2,p_2 ) v(p_2 ) , [ 7 ] where @xmath59 is the transverse component that contains the explicit exponential behavior of @xmath51 .
although @xmath60 has some mild @xmath51 dependence due mainly to mass effects of @xmath50 , the average transverse momentum @xmath61 is determined primarily by the inverse slope @xmath62 and is not dependent on @xmath41 .
this is an approximate statement that is based on the brahms data @xcite , which show that @xmath63 is essentially independent of rapidity .
since @xmath39 serves as the phenomenological bridge between @xmath64 and @xmath58 , the key question to address is : which of the two components , the longitudinal @xmath60 or the transverse @xmath65 , does the two - particle correlation exert its most important influence in relating @xmath64 to @xmath58 ?
if there is longitudinal correlation from early times as in @xcite , then its effect must be to convert @xmath60 to @xmath55 . in that case
@xmath65 is relegated to the secondary role due to radial flow ( which is , nevertheless , essential in explaining the @xmath6 restriction as in refs .
@xcite ) . on the other hand ,
if there is no intrinsic long - range longitudinal correlation , then @xmath60 is unaffected , and the ridge can only arise from the change in the transverse component , @xmath65 .
without phenomenology one would think that the first option is more reasonable , when @xmath66 is regarded as large , and especially when there is an inclination based on theoretical ideas that there is long - range correlation . with the ridge phenomenology encapsuled in eq .
( [ 3 ] ) pointing to direct relevance of @xmath60 , the question becomes that of asking : @xmath67 is large compared to what ?
if it is now recognized that @xmath67 is not large compared to the @xmath1 range of @xmath68 after the widening due to @xmath40 smearing ( remarked at the end of the previous section ) is taken into account , then the need for a long - range dynamical correlation to account for the structure of @xmath55 is lost .
we describe below a possible explanation based on the second option of no long - range correlation .
the key is to accept the suggestion of the data that the unmodified longitudinal component @xmath60 is sufficient .
a series of articles have treated the subject of ridge formation in the recombination model @xcite , beginning with ( a ) the early observation of pedestal in jet correlation @xcite , to ( b ) its effects on azimuthal anisotropy of single - particle at mid - rapidity @xcite , and then to ( c ) the dependence on the azimuthal angle @xmath69 of the trigger relative to the reaction plane @xcite .
forward productions in d - au and au - au collisions have also been studied in @xcite .
our consideration here of ridge formation at @xmath70 is an extension of earlier studies with the common theme that ridges are formed as a consequence of energy loss by semihard or hard partons as they traverse the medium .
the details involve careful treatment of the hadronization process with attention given to both the longitudinal and transverse components .
the @xmath7 dependence has been studied thoroughly in @xcite , and the @xmath1 dependence should take into account of the experimental fact that the @xmath71 ratio can be large @xmath72 at large @xmath1 @xcite so that @xmath60 in eq .
( [ 7 ] ) can be properly reproduced , which is a subject to be reported elsewhere @xcite . for our purpose here we need not repeat all the details , but state only that the effect of jets on the medium is that the energy loss enhances the thermal fluctuations of the soft partons in the early stage ( before hubble - like expansion takes place ) so that even at @xmath70 the transverse component @xmath65 in eq .
( [ 7 ] ) is sensitive to the thermal enhancement .
the implementation of that basic physical idea on the formation of ridge particles ( called pedestal at the time ) was first carried out in ref .
@xcite in terms of thermal - shower recombination for the trigger , and of thermal - thermal recombination for the associated particles .
the result of the collective investigation @xcite can be summarized by r^h_2(_1,_2,p_2)=c h^h_2(_2,p_2 ) v_r(_1,_2,p_2 ) , [ 8 ] where v_r(_1,_2,p_2)= v_b+r(_1,_2,p_2)- v_b(p_2 ) , [ 9 ] which has the structure of eq . (
[ 5 ] ) , but only in the transverse part .
the constant @xmath73 characterizes the magnitude of the ridge , which can depend on many factors that include the fluctuations in the initial configuration , the details of correlation dynamics , the experimental cuts , the @xmath6 interval where the ridge is formed and the related scheme of background subtraction .
its value ( that was not calculated ) does not affect the relationship between the @xmath1 dependencies of the two sides of eq .
( [ 8 ] ) .
of more relevant concern to us are the transverse components , which may be written more explicitly as v_b(p_2)=(-p_2/t ) , v_b+r(_1,_2,p_2)=. [ 10 ] thus @xmath62 is the inverse slope of the inclusive thermal , regarded in eq .
( [ 9 ] ) as the background , while @xmath74 represents the enhancement effect due to the semihard or hard parton .
the parton s trajectory toward the near - side surface has a conical vicinity in which the medium has increased thermal activity because of the energy loss of the parton . in the initial configuration of spatial uncertainty in @xmath75
we allow the right - moving soft partons to start from the left side of the thermally enhanced cone , and similarly the left - moving soft partons to start from the right side , so that the transverse of the soft partons can change from @xmath62 to @xmath76 due to their passage through the enhanced cone .
this is transverse broadening as in the conventional interpretation of the cronin effect @xcite but only for associated particles on the near side within a restricted region of @xmath6 around the trigger @xmath77 .
furthermore , the polar - angular relationship between the soft partons and the semihard parton can fluctuate significantly during the broadening process that affect both the longitudinal and transverse momenta of the associated particles .
@xmath76 is a measure of the average effect on those particles .
it should not depend significantly on @xmath0 , just as the cronin effect does not have sensitive dependence on the longitudinal momentum .
thus we shall set @xmath78 to be roughly constant in @xmath0 to about the same degree of approximation as @xmath62 in @xmath65 is independent of @xmath41 in eq .
( [ 7 ] ) .
to let @xmath76 be independent of @xmath0 is not equivalent to putting in by hand long - range longitudinal correlation ; that is analogous to the proposition in the flux - tube model that assuming similar radial flow for all parts of a tube of partons ( so that they all gain the same average @xmath79 ) does not imply that long - range longitudinal correlation is inserted by hand
. there is , however , transverse correlation in the sense that the @xmath51 dependence of @xmath80 is the same at all @xmath41 .
that is similar to the familiar phenomenon that rising tide raises all boats , where the boat heights are transverse to the spatial separation that is longitudinal , the former being correlated , the latter not .
experimentally , for trigger momentum in the interval @xmath81 gev / c , it is found that @xmath82 is less than 20% of @xmath62 @xcite .
putting the various parts together , we have v_r(p_2 ) = e^-p_2/t(e^p_2/t-1 ) , = 1t-1t=ttt. [ 11 ] this expression was first obtained in refs .
@xcite as a description of the ridge without trigger .
it was noted there that @xmath83 as @xmath84 , and that @xmath85 sets the scale for @xmath86 for @xmath87 gev / c in agreement with the data on elliptic flow .
since the lower limit of the acceptance window for associated - particle momentum in the phobos experiment is low @xcite , the relevant values of @xmath62 and @xmath76 may not be the same as those measured by star @xcite .
for the purpose of quantitative orientation they may be set at @xmath88 gev and @xmath89 gev , so that @xmath90 gev .
the exact values are not important to our qualitative conclusion to be drawn below .
we now substitute eq .
( [ 8 ] ) in ( [ 4 ] ) and use ( [ 7 ] ) with @xmath65 identified as @xmath91 to obtain = _ 0 ^ 1.5 d_1_h_2 dp_2 p_2 .
cv_r(p_2)v_b(p_2 ) _
1^h_2(_2,p_2 ) |__2 = _ 1 + .
[ 12 ] comparing this equation with eq .
( [ 6 ] ) , we come to the conclusion that @xmath39 is a phenomenological approximation of @xmath92 in the region where it contributes most to the integral over @xmath51 . by itself
@xmath93 increases with @xmath51 as @xmath94 , but it is damped more severely by the exponential decrease of @xmath68 for @xmath95 gev / c , since @xmath96 .
thus @xmath92 may be approximated by a constant @xmath39 in the region where the integrand is maximum at around @xmath97 gev / c . in so doing ,
we obtain eq .
( [ 6 ] ) and therefore the phenomenological relation given by eq .
( [ 3 ] ) .
equation ( [ 12 ] ) implies that there is transverse correlation , but no explicit longitudinal correlation beyond what is implicitly contained in @xmath58 . the transverse correlation that we refer to is not what one usually associates with the correlation between hadrons in the fragments of a high-@xmath79 jet .
all of those fragments are in a small range of @xmath0 and have transverse - momentum fractions that are correlated .
they populate the peak in fig . 2 . in our problem about the ridge we have been concerned with the transverse momentum of a particle associated with a trigger outside that peak .
the former reveals the effect of the medium on the jet , while the latter reveals the effect of the jet on the medium .
that is the basic difference between the jet and ridge components of the associated particles . since semihard or hard scattering takes place early
, transverse broadening can take place for soft partons ( the medium ) moving through the interaction zone , leading to the ridge structure .
an issue that this study has brought up is the usage of the word `` large '' in referring to the range of @xmath0 in the ridge structure . our phenomenological observation in eq .
( [ 3 ] ) , substantiated by figs . 1 and 2 , does not reveal any quantitative definition of what large @xmath0 means . to say that @xmath70 is large is a figurative description until it is followed by a statement on dynamics . to be able to relate large @xmath0 to dynamical long - range correlation
is a worthy theoretical endeavor , but more can be added to its phenomenological relevance if it can also elucidate the empirical connection between the two sides of eq .
( [ 3 ] ) . since the pseudorapidity @xmath8 involves an integration over @xmath27 and sum over charged hadrons , as expressed in eq .
( [ 1 ] ) , a full understanding of its @xmath1 dependence must take into account the experimental fact from brahms that protons are produced more than twice as many as pions at @xmath98 and @xmath99 gev / c in central collisions @xcite .
in fact , it is seen in fig .
1(a ) that the phobos data on @xmath8 at 0 - 6% centrality show a bump at @xmath100 , which is undoubtedly related to proton production .
the widening of the @xmath1 is a hadronization problem that involves late - time physics .
long - range longitudinal correlation is early - time physics . to connect
the two is a theoretical problem that remains to be done .
most models emphasizes one or the other .
equation ( [ 3 ] ) that connects @xmath49 to @xmath8 is an empirical statement without any direct revelation on the nature of the dynamics throughout all times of the collision process .
it is therefore a worthy goal to explain that connection from all approaches .
the approach that we have taken provides a possible solution involving no long - range longitudinal correlation in addition to what accounts for the single - particle @xmath101 .
since @xmath101 is an integral over the two - particle @xmath102 , whatever intrinsic correlation that exists among the constituents must contribute to the properties of @xmath101 . in this paper
we have not delved into the details of @xmath101 ; we have only made the observation that , given @xmath101 , it is possible to get @xmath49 without additional longitudinal correlation of any range .
it gives a hint on how the ridge structure is to be understood .
but as with any model on particle production , the burden of proof is then shifted heavily toward an explanation of @xmath101 in all its details , only the integrated form of which is shown in fig . 1 .
although we have given a simple interpretation of the phenomenological quantity @xmath39 , the importance of this work leans more toward the finding of the relationship between @xmath49 and @xmath8 and of the substantiation it receives from the two pieces of data from phobos exhibited in figs . 1 and 2
. an extension of our interpretation would lead naturally to the prediction that similar phenomenon will be found at lhc .
this work was supported , in part , by the u. s. department of energy under grant no .
de - fg02 - 92er40972 . | a relationship between the ridge in @xmath0 and the single - particle in @xmath1 is proposed .
it is then verified by use of the data from phobos on both s. the implication seems to point to the possibility that there is no long - range longitudinal correlation .
an interpretation of the phenomenological observation along that possibility is developed .
.75 cm * charles b. chiu@xmath2 and rudolph c. hwa@xmath3 * .5 cm @xmath2center for particles and fields and department of physics , + university of texas at austin , austin , tx 78712 , usa + @xmath4institute of theoretical science and department of physics , + university of oregon , eugene , or 97403 - 5203 , usa .5 cm .5 cm |
colouring a graph is the problem of finding the minimal number of colours required to assign each vertex a colour such that no adjacent vertices have the same colour .
it is a classic problem in computer science and one of karp s 21 np - complete problems @xcite .
colouring remains np - complete if bounded to three colours @xcite .
also , if a 3-colourable graph is given , it remains np - hard to find a colouring , even if four colours are allowed @xcite .
dually chordal graphs are closely related to chordal graphs , hypertrees and @xmath2-acyclic hypergraphs , which also explains the name of this graph class ( see @xcite for more ) .
section [ sec : duallychordal ] will show , that colouring is np - complete for dually chordal graphs if bounded to four colours .
so the aim of this paper is to find a linear time algorithm for the 3-colouring problem .
next to the classic vertex colouring problem , there are the problems of colouring only edges or edges and vertices of a graph . for
dually chordal graphs they have been investigated by de figueiredo et al . @xcite .
let @xmath3 be a graph with the vertex set @xmath4 , the edge set @xmath5 . for this paper ,
any graph is finite , undirected , connected and without loops or multiple edges . for a set @xmath6 , @xmath7
$ ] denotes the induced subgraph of @xmath0 with the vertex set @xmath8 .
let @xmath9 be the complement of @xmath0 , such that @xmath10 .
a set of vertices @xmath11 is a _ clique _ if for each pair @xmath12 ( @xmath13 ) , @xmath14 and @xmath15 are adjacent .
a clique of size @xmath16 is denoted as @xmath17 .
the number of vertices in a maximum clique in @xmath0 is the _ clique number @xmath18 _ of @xmath0 .
diamond _ be a graph with four vertices and five edges .
the edge connecting the vertices with degree 3 is called _ mid - edge_. a _ chordless cycle _ @xmath19 has @xmath20 vertices @xmath21 and the edges @xmath22 ( index arithmetic modulo @xmath20 ) .
if a chordless cycle has at least five vertices , it is also called @xmath23 . a _ wheel _
@xmath24 is a @xmath19 plus a vertex @xmath25 and the edges @xmath26 ( @xmath27 ) . a chordless cycle @xmath19 or wheel @xmath24 is _ odd _
if @xmath20 is odd . let @xmath28 denote the _ open _ and @xmath29 : = n(v ) \cup \{v\}$ ] the _ closed neighbourhood _ of the vertex @xmath15 .
a graph is _ locally connected _ if for all vertices @xmath15 the open neighbourhood @xmath30 is connected .
a vertex @xmath31 $ ] is a _
maximum neighbour _ of @xmath15 , if for all @xmath32 $ ] , @xmath33 \subseteq n[u]$ ] holds .
note that @xmath34 is not excluded .
a vertex ordering @xmath35 is a _
maximum neighbourhood ordering _
if every @xmath36 ( @xmath37 ) has a maximum neighbour in @xmath38 $ ] .
a vertex @xmath15 is an articulation point of @xmath0 if @xmath39 $ ] is not connected . if a graph has no articulation point it is _ biconnected _ ( or _ 2-connected _ ) .
maximal biconnected subgraphs are called _
blocks_. any connected graph can decomposed into a block tree .
if two blocks intersect , they have exactly one common vertex which is an articulation point .
an _ independent set _ is a vertex set @xmath40 such that for all vertices @xmath41 , @xmath42 holds . a graph is _
@xmath20-colourable _ if @xmath4 can be partitioned into @xmath20 independent sets @xmath43 with @xmath44 and @xmath45 ( @xmath46 ) .
the _ chromatic number _
@xmath47 of a graph @xmath0 is the lowest @xmath20 such that @xmath0 is @xmath20-colourable .
the _ @xmath20-colourability _ problem asks if a graph is @xmath20-colourable .
accordingly the _ colourability _ problem asks for the chromatic number of a graph .
it is easy to see , that 2-colourability can be solved in linear time for each graph .
a graph @xmath0 is _ perfect _ if for each induced subgraph @xmath48 of @xmath0 the chromatic and the clique number are equal ( @xmath49 ) .
chudnovsky et al .
@xcite have shown that a graph is perfect if and only if it is @xmath50-free for all @xmath51 .
this is known as the strong perfect graph theorem .
perfect graphs can be recognised and coloured in polynomial time @xcite@xcite . for a graph @xmath0
, @xmath52 denotes the _ clique graph _ of @xmath0 , where each vertex in @xmath52 represents a maximal clique of @xmath0 and two vertices are connected if the corresponding cliques have a common vertex , i.e. @xmath53 with @xmath54 and @xmath55 .
a graph @xmath0 is _ chordal _ if it is @xmath19-free ( @xmath56 ) .
a graph is _ clique - chordal _ if its clique graph is chordal .
dually chordal graphs are originally defined by a maximum neighbourhood ordering .
a graph is _ dually chordal _ if and only if it has a maximum neighbourhood ordering .
brandstdt et al .
@xcite give an overview of characterisations for dually chordal graphs .
one of it allows to recognise graphs of this class in linear time ( with an algorithm by tarjan and yannakakis @xcite ) .
another characterisation was developed in @xcite : [ theo : dcdef ] let @xmath57 the set of vertices on the path from @xmath14 to @xmath15 in the tree @xmath1 , with @xmath58 . for a graph @xmath3 the following conditions are equivalent : 1 .
@xmath0 is dually chordal .
[ item : dcdefcliquetree ] there is a spanning tree @xmath1 of @xmath0 such that every maximal clique of @xmath0 induces a subtree in @xmath1 .
[ item : dcdefpathtree ] there is a spanning tree @xmath1 of @xmath0 such that for every edge @xmath59 the following holds : @xmath60 [ lem : cliquetreeequalpathtree ] the spanning trees in conditions [ item : dcdefcliquetree ] and [ item : dcdefpathtree ] of lemma [ theo : dcdef ] are equal . since @xmath57 does not include the vertices @xmath14 and @xmath15 , ( according to the notation for the open and closed neighbourhood ) let @xmath61 : = p_t(u , v ) \cup \{u , v\}$ ] . if @xmath59 then @xmath61 $ ] is a clique .
is is easy to see that an arbitrary graph becomes dually chordal if a vertex adjacent to all vertices is added .
this leads to a reduction for several np - complete problems , including colourability .
the idea for this method was already given by brandstdt et al .
@xcite .
[ theo:4coldc ] 4-colourability is np - complete for dually chordal graphs .
let @xmath3 be an arbitrary graph .
thus , @xmath62 with @xmath63 is dually chordal .
it follows that @xmath0 is 3-colourable if and only if @xmath64 is 4-colourable .
because 3-colourability is np - complete in general @xcite , 4-colourability is np - complete for dually chordal graphs .
obviously a graph has to be @xmath65-free to be 3-colourable .
so this section investigates @xmath65-free dually chordal graphs . for this section let @xmath3 be a @xmath65-free dually chordal graph .
the spanning tree @xmath66 of @xmath0 and the set @xmath61 $ ] are defined as in condition [ item : dcdefpathtree ] of theorem [ theo : dcdef ] and the paragraph after lemma [ lem : cliquetreeequalpathtree ] .
[ cor : k4freepedges ] if @xmath59 then @xmath67| \leq 3 $ ] , i.e. @xmath68 . for the next lemma the notion of a _ route _ is defined : a _ route _ on @xmath0 is a list of vertices @xmath69 such that @xmath70 or @xmath71 for all @xmath72 .
[ lem : circlewheel ] let @xmath73 be a @xmath19 in @xmath0 with @xmath56 .
there is a vertex @xmath74 such that @xmath73 and @xmath74 form a wheel .
also no edge of @xmath73 is in @xmath1 .
let @xmath73 be a chordless cycle with the vertices @xmath75 ( @xmath56 ) and the edges @xmath76 ( index arithmetic modulo @xmath20 ) . based on corollary [ cor : k4freepedges ] for every edge
there is the set of vertices @xmath77 = \
{ c_i , w_i , c_{i+1 } \}$ ] with @xmath78 if and only if @xmath79 .
thus , it is possible to build a route @xmath80 on @xmath1 ( see figure [ fig : circlewheel ] ) .
and the route @xmath81 ( waved ) . ] because all the edges of @xmath81 are in @xmath1 , @xmath81 also induces a tree .
therefore @xmath81 can be eliminated by removing leaves .
let @xmath82 be a subroute of @xmath81 and @xmath36 a leaf .
thus , if @xmath83 and @xmath84 then @xmath85 , so @xmath86 or @xmath87 can also removed from @xmath81 with @xmath36 . now assume that no vertex of @xmath73 is a leaf . in this case
a vertex @xmath88 is a leaf .
so it follows , that @xmath89 . therefore there must be a leaf @xmath90 and also @xmath91 ( otherwise @xmath92 )
this allows to remove @xmath90 and @xmath88 from @xmath81 . continuing this procedure
, it follows that @xmath93 and @xmath94 for all @xmath16 .
thus , no edge @xmath76 is an edge of @xmath1 and there is a vertex @xmath74 , such that @xmath73 and @xmath74 form a wheel .
[ theo:3coliffperfect ] a dually chordal graph @xmath0 is 3-colourable if and only if @xmath0 is perfect and @xmath65-free .
@xmath95 by definition of perfect graphs .
@xmath96 @xmath0 is 3-colourable , so it is @xmath65-free .
let @xmath97 .
it is easy to see , that no @xmath98 is 3-colourable .
therefore , @xmath0 has no @xmath98 . from the 3-colourability it also follows that there is no odd wheel in @xmath0 and by lemma [ lem : circlewheel ] no odd hole ] .
thus , by the strong perfect graph theorem , @xmath0 is perfect ( recall , @xmath99 ) .
theorem [ theo:3coliffperfect ] now leads to a polynomial time algorithm for 3-colourability of dually chordal graphs .
after showing that there is a polynomial time algorithm , this section presents a linear time algorithm for 3-colourability of dually chordal graphs based on the following theorem : [ theo : locconconstruction ] each connected locally connected graph can be constructed by starting with an arbitrary edge and then add only vertices that has at least two adjacent neighbours in the already constructed graph .
let @xmath100 be the ordering the vertices are added to construct a locally connected graph @xmath3 starting with an edge @xmath101 . also let @xmath102 with @xmath103 and @xmath104
. now let @xmath105 $ ] ( @xmath106 ) be a locally connected subgraph of @xmath0 .
assume , there is no vertex @xmath74 adjacent to two vertices in @xmath107 . then @xmath0 is not connected or @xmath74 has only one neighbour @xmath108 in @xmath107 . in second case , because @xmath0 is locally connected , there is a vertex set @xmath109 connecting @xmath74 with the neighbours of @xmath108 that are already in @xmath107 .
thus , there is a vertex @xmath110 adjacent to @xmath108 and a neighbour of @xmath108 in @xmath107 .
therefore , set @xmath111 .
theorem [ theo : locconconstruction ] allows to colour graphs whose blocks are locally connected by using the following strategy : select an uncoloured vertex with a minimal number of available colours and give it an available one . repeat this until all vertices are coloured or there is no available colour for a vertex .
algorithm [ algo:3colloccon ] describes this more detailed : [ algo:3colloccon ] + * input : * a graph @xmath3 whose blocks are locally connected . +
* output : * a 3-colouring for @xmath0 if and only if @xmath0 is 3-colourable . 1 .
give every vertex @xmath15 a set of available colours @xmath112 . also every vertex gets the possibility to get marked as coloured . at the beginning
no vertex is marked .
2 . * while * there are uncoloured vertices 1 . select an uncoloured vertex @xmath15 for which @xmath113 is minimal .
if @xmath114 , * stop : * @xmath0 is not 3-colourable .
3 . disable all but one colour in @xmath115 and mark @xmath15 as coloured .
4 . disable @xmath115 for all neighbours of @xmath15 ( @xmath116 ) .
[ theo:3collocconalog ] algorithm [ algo:3colloccon ] works correctly and can be implemented in linear time .
_ correctness_. after selecting the first two vertices , the algorithm only selects vertices with two adjacent neighbours , i.e. with @xmath117 , until all vertices of a locally connected block are coloured ( theorem [ theo : locconconstruction ] ) .
therefore , the colouring is unique for a locally connected block . also , if a vertex @xmath15 with @xmath118 and the coloured neighbour @xmath14 is selected , @xmath14 is an articulation point .
this allows to treat @xmath14 and @xmath15 as the first coloured vertices of a locally connected graph .
therefore , each available colour for @xmath15 leads to a correct colouring .
_ complexity_. it is easy to see , that line 1 runs in linear time and line 2(b ) in constant time such as line 2(c ) .
line 2(d ) is bounded by the number of neighbours . because @xmath119 , the full time needed ( overall iterations ) for line 2(d ) is in @xmath120 .
line 2(a ) can be implemented by using three doubly linked lists as queues .
the lists include the uncoloured vertices .
one is for vertices with @xmath121 , one for @xmath118 and one for @xmath117 .
if there is also a pointer from each vertex to its entry in the list , adding and removing a vertex can be done in constant time .
thus , line 2(a ) runs in constant time . another linear time algorithm for 3-colourability of locally connected graphs
was presented by kochol @xcite , but is more complicated .
dually chordal graphs are a subclass of clique - chordal graphs @xcite .
the next lemma will show , that a 3-colouring for clique - chordal graphs can be be found with algorithm [ algo:3colloccon ] .
each block of a clique - chordal graphs is locally connected .
let @xmath122 be a clique - chordal graph and @xmath123 its clique graph .
assume that @xmath0 has a block which is not locally connected .
then there is a vertex @xmath15 such that its neighbourhood is not connected and it is no articulation point .
therefore , there is a chordless cycle @xmath124 fulfilling the conditions that there is no path in @xmath125 $ ] connecting the two neighbours of @xmath15 in @xmath124 .
let @xmath126 be the smallest @xmath124 and using the following notation : @xmath14 , @xmath15 , @xmath74 and @xmath127 , @xmath128 , @xmath129 are vertices of @xmath126 with @xmath130 , @xmath131 , @xmath132 , @xmath133 , @xmath134 , @xmath135 and @xmath136 . if @xmath126 is a @xmath137 , then @xmath138 and @xmath139 .
figure [ fig : cycleproofcliquechordal ] illustrates this notation . and the vertex @xmath140 . ]
there is no clique in @xmath0 that contains two edges of @xmath126 , otherwise @xmath126 would not be chordless .
thus , each edge represents another clique of @xmath0 .
let @xmath141 , @xmath142 , @xmath143 and @xmath144 denote these cliques .
because the cliques have a common vertex if the corresponding edges have one , they are forming a cycle @xmath145 in @xmath146 .
@xmath146 is chordal .
therefore , @xmath145 has at least one chord . without loss of generality ,
let @xmath147 be such a chord .
thus , @xmath141 and @xmath144 have a common vertex @xmath140 that is adjacent to @xmath14 , @xmath15 , @xmath128 and @xmath129 . if @xmath148 , @xmath140 is adjacent to @xmath74 .
if @xmath149 , the sequence @xmath150 induces a chordless cycle smaller than @xmath126 . in both cases
@xmath126 is not the smallest cycle @xmath124 such that there is no path in @xmath125 $ ] connecting the two neighbours of @xmath15 in @xmath124 .
therefore , @xmath0 is nearly locally connected .
it follows : the 3-colourability problem can be solved in linear time for dually chordal and clique chordal graphs .
after introducing dually chordal graphs and investigating the connection between @xmath65-free dually chordal and perfect graphs , this paper presented a linear time algorithm to find a 3-colouring by using the structure of locally connected graphs .
this algorithm also computes a correct 3-colouring for each graph whose blocks are locally connected .
grtschel , m. , lovsz , l. , schrijver , a. : polynomial algorithms for perfect graphs . in : berge ,
c. , chvtal , v. ( eds . ) topics on perfect graphs , north - holland mathematics studies , vol .
north - holland ( 1984 ) kochol , m. : linear algorithm for 3-coloring of locally connected graphs . in : jansen , k. , margraf , m. , mastrolilli , m. , rolim , j. ( eds . ) experimental and efficient algorithms , pp .
2647 in lecture notes in computer science , springer berlin / heidelberg ( 2003 ) tarjan , r.e . , yannakakis , m. : simple linear - time algorithms to test chordality of graphs , test acyclicity of hypergraphs , and selectively reduce acyclic hypergraphs .
siam j. comput .
13 , 566579 ( 1984 ) | a graph @xmath0 is dually chordal if there is a spanning tree @xmath1 of @xmath0 such that any maximal clique of @xmath0 induces a subtree in @xmath1 .
this paper investigates the colourability problem on dually chordal graphs .
it will show that it is np - complete in case of four colours and solvable in linear time with a simple algorithm in case of three colours .
in addition , it will be shown that a dually chordal graph is 3-colourable if and only if it is perfect and has no clique of size four . |
geometric aspects of critical phenomena are characterized by random fractals such as conformally invariant fluctuations of local order parameters .
they have been extensively studied from various different points of view , especially in two dimensions ( 2d ) where the conformal invariance imposes strong constraints on the structure of critical phenomena . among them , the schramm - loewner evolutions ( sles ) @xcite , which directly describe geometric aspects of 2d critical phenomena through simple 1d brownian motions , have brought a renewed interest in the theory of random fractals ( see @xcite for reviews ) . the sle is a stochastic process defined in the upper half plane @xmath1 .
its evolution is described by the ordinary differential equation @xmath2 where @xmath3 is a brownian motion on @xmath4 , starting at the origin ( i.e. @xmath5 ) , and its expectation value and variance are given by @xmath6=0 $ ] and @xmath7=\kappa dt$ ] , respectively . here
@xmath8 is a diffusion coefficient which essentially characterizes the sle process .
the sle has a solution up to the explosion time @xmath9 , i.e. the first time when @xmath10 hits the singularity @xmath11 .
let @xmath12 be the hull at time @xmath13 ( see fig .
[ hull ] for a schematic view ) .
then @xmath14 ( @xmath15 ) is an increasing family of hulls : @xmath16 for @xmath17 .
moreover @xmath10 with @xmath18 at @xmath19 ( hydrodynamic normalization ) is the unique conformal map uniformizing the complement of the hull @xmath14 in the upper half plane @xmath1 : @xmath20 ( see fig . [ uniform ] ) .
the image @xmath11 by @xmath21 defines the tip @xmath22 of the growing random curve .
more precisely , it is expressed as @xmath23 .
denotes the tip of the trace ) ( a ) and its hull @xmath14 ( b).,scaledwidth=85.0% ] the connection between the sle and conformal field theory ( cft ) is well understood @xcite .
specifically , it can be accomplished by noticing that cft correlation functions @xmath24 are sle martingales : @xmath25 , where @xmath26 and @xmath27 are boundary condition changing ( bcc ) operators with conformal weights @xmath28 , which are inserted at the points @xmath29 and @xmath30 , respectively ( see @xcite or next section for details ) .
thus one can find the sle corresponds to the minimal conformal field theory @xmath31 ( @xmath32 , @xmath33 are coprime integers satisfying the condition @xmath34 ) where the central charge and the conformal weights of the primary fields are , respectively , given by @xcite @xmath35 then the diffusion constant @xmath36 and the conformal weight @xmath28 of the boundary field @xmath37 are , respectively , expressed as @xmath38 namely the boundary field @xmath37 is degenerate at level two , i.e. possesses a null field at level two .
an extension to the sle connecting with conformal field theories with lie algebra symmetries ( i.e. wzw models @xcite ) has been achieved by adding the extra brownian motion @xmath39 on a ( semisimple ) group manifold @xmath40 associated with a lie algebra @xmath41 , where @xmath42 s ( @xmath43 ) stand for any representation of the lie algebra generators @xcite ( see @xcite for a very different approach ) .
the evolution of this additional stochastic process @xmath44 is defined as @xmath45=0,\,\ , \mathbf{e}[d\vartheta_t^a d\vartheta_t^b]=\delta^{ab } dt \,\,\,(\vartheta_t^a\in \mathbb{r } ) .
\label{lie}\ ] ] the combination of and defines a fractal curve living on the lie group @xmath40 manifold .
the sle martingale @xmath25 , which should be satisfied by the cft correlation function ( note that the bcc operators and the operator @xmath46 take their values on @xmath40 ) , determines the relation of the sle defined by and with the corresponding wzw model .
some extensions to the sle with other additional symmetries have also been done in @xcite . : @xmath47 and its inverse .
, scaledwidth=85.0% ] in this paper we generalize the sle for the system containing multiple random interfaces which possess additional lie algebra symmetries , according to the theory developed in @xcite .
assuming that each sle interface grows under an independent martingale in the infinitesimal time interval , we extend together with to the case for the system with multiple interfaces . the evolution ( cf . for the single case ) characterizing the geometric aspect of the interfaces is described by a multiple brownian motion on the real line , while the evolution ( cf . for the single case ) expressing the algebraic aspect is described by a multiple brownian motion on the lie group @xmath40 . both the brownian motions , however , have drift terms describing the interaction among the sle traces . taking into account the sle martingale
, one finds that these drift terms are determined by the partition function in the corresponding wzw model .
moreover this partition function characterizes the configuration of the sle traces . as an example
, we apply our formula to the sle for the @xmath0-wzw models . utilizing the fusion rules for @xmath0
, we conjecture that there is a one - to - one correspondence between the partition functions and the topologically inequivalent configurations of the sle traces .
further , we exactly compute the probabilities of occurrence for certain configurations of traces ( that correspond to crossing probabilities ) for the triple sle , which can be obtained by solving the knizhnik - zamolodchikov ( kz ) equation @xcite .
the paper is organized as follows . in the subsequent section
, we describe some basic notions required in this paper . in section 3 ,
the multiple sle for conformal field theory with lie algebra symmetries is formulated .
the drift terms of the driving brownian motions are explicitly determined in section 4 .
a concrete application of the formula to the sle for the @xmath0-wzw model is given in section 5 .
in this section , we describe several theoretical foundations required in subsequent sections . in the former part of this section , we introduce sle martingales from the point of view of statistical mechanics ( see @xcite for details ) .
sle martingales are given by cft correlation functions involving bcc operators , and hence they become a key to decipher the relation between sle and cft . in the latter part , general properties of correlation functions
are explained in the case of the wzw model . to formulate the sles correctly describing the behavior of 2d interfaces in critical systems
, one must construct sle martingales in terms of the corresponding statistical systems defined on @xmath1 .
let @xmath48 be the thermal average of an observable @xmath46 defined in @xmath1 , and @xmath49 be the thermal average under a given shape of configuration of ( multiple ) interfaces , where @xmath50 denotes the shape of configuration with its occurrence probability @xmath51 $ ] .
then the thermal average @xmath48 must be given by @xmath52 = \sum_{\{\gamma_t\ } } \mathbf{p}[\{\gamma_t\}]\bra \mathcal{o}\ket|_{\{\gamma_t\}},\ ] ] where the average @xmath53 $ ] should be taken over all the possible configurations .
the conditional expectation value @xmath49 is thus time independent ( i.e. conserved in mean ) , and therefore it is a martingale . here and in what follows , we denote it by @xmath54 . at the critical point where the conformal invariance is expected in the system , the above observable
can be described in terms of the cft correlation functions .
for the situation where the number of the interfaces under consideration is @xmath55 , it reads @xmath56 where @xmath57 s denote the positions of the tips of the interfaces .
note that the cft correlation functions are defined on the domain @xmath1 removing the hull @xmath14 , i.e. @xmath58 ( see fig .
[ corr ] ) .
the operators @xmath59 s inserted at the positions of the tips denote primary bcc operators with conformal weights @xmath28 : under a local conformal map @xmath60 , a primary field @xmath61 transforms as @xmath62 : @xmath63 the denominator in stands for the cft partition function with a specific boundary condition fixed by the bcc operators @xmath59 . applying the conformal uniformizing map @xmath10 ( written with the same symbol as that used for the single sle ) , we obtain @xmath64 here @xmath65 , and @xmath66 is the image of @xmath46 by the map @xmath67 .
note that the jacobians coming from the conformal map on @xmath68 have been canceled in the numerators and the denominators holds for a global conformal map @xmath69 , where the field @xmath70 transforms as @xmath71 @xcite . ] .
now sle martingales @xmath54 are expressed as the cft correlation functions on @xmath72 , where the bcc operators @xmath73 are inserted at the points @xmath74 ( see fig .
[ corr ] ) . to proceed further ,
let us consider the case when the operator @xmath46 is a product of an arbitrary number of primary fields @xmath75 at positions @xmath76 and with conformal weights @xmath77 . by construction
, the uniformizing map @xmath10 can be analytically extended to the lower half plane : @xmath78 .
then , the doubling trick can be applied to the cft correlation functions .
the result reads @xmath79 where we denote that @xmath80 ; @xmath81 ; @xmath82 ; @xmath83 , @xmath84 for @xmath85 . here
@xmath86 ( @xmath87 ) stands for the holomorphic ( antiholomorphic ) part of the field @xmath88 . in this paper , we analyze the sle martingales for the system that possesses additional lie algebra symmetries . namely we construct the multiple sle for the wzw models which are one of the most fundamental cfts with extra lie algebra symmetries . in this case , the primary fields constructing the sle martingale possess internal degrees of freedom , such as spin " .
it can be transformed to the one defined on @xmath72 by the uniformizing map @xmath10 ( b).,scaledwidth=85.0% ] let us introduce several properties for correlation functions of wzw primary fields to analyze the sle martingales .
the wzw model is a cft described by a field @xmath89 taking values in a group manifold @xmath40 associated with a lie algebra @xmath41 @xcite .
the model is invariant under @xmath90 where @xmath91 and @xmath92 denote arbitrary matrices valued in @xmath40 .
this invariance gives rise to noether currents @xmath93 and @xmath94 , which can be written as @xmath95 where @xmath42 s stand for any matrix representation of the generator of @xmath41 , with commutation relations @xmath96=\sum_c
i f^{ab}_{\,\,\,\,\,\,c } t^c$ ] .
the parameter @xmath97 is a positive integer referred to as the level .
hereafter we only consider the holomorphic components , as if there were no boundary ( cf . )
let @xmath98 be a correlator of @xmath40-valued fields .
then the infinitesimal transformation @xmath99 leads to the ward identity @xmath100 on the other hand , conformal aspects of the wzw model are described by the stress energy tensor @xmath101 : @xmath102 where @xmath103 denotes the dual coxeter number of @xmath41 .
note that @xmath104 means the normal ordering " defined as @xmath105 the infinitesimal conformal transformation @xmath106 leads to the ward identity @xmath107 this geometric part of the ward identity together with the algebraic part are a key ingredient in analyzing the sle martingale .
the current @xmath108 and the stress energy tensor @xmath101 can be expanded in terms of the modes @xmath109 and @xmath110 , respectively .
namely @xmath111 where @xmath112 and @xmath110 denote , respectively , the generators . in that case ,
the structure constants can be written as @xmath113 , where @xmath114 is antisymmetric in all three indices .
] of the affine lie algebra @xmath115 with level @xmath97 , and the generators of the virasoro algebra : @xmath116=\sum_c i f^{ab}_{\,\,\,\,\,\,c } j^c_{n+m}+k n \delta^{a , b}\delta_{n+m,0 } , \nn \\ & [ l_n , j_m^a]=-m j_{n+m}^a , \nn \\ & [ l_n , l_m]=(n - m)l_{n+m}+\frac{c}{12}(n^3-n)\delta_{n+m,0}. \label{vir}\end{aligned}\ ] ] the central charge @xmath117 of the virasoro algebra is then given by @xmath118 by construction , the virasoro generators are not independent of the affine generators : @xmath119 where the normal ordering @xmath104 means that the operator with larger index @xmath120 is placed at the rightmost position .
the primary fields in the wzw model are defined as the fields transforming covariantly with respect to the @xmath121 transformation : @xmath122 .
together with the conformal covariance ( see ) , these properties can be expressed in terms of the operator product expansions ( ope ) via the ward identities and : @xmath123 where the field @xmath124 takes values in the representation specified by the highest weight @xmath125 , and @xmath126 is the generator @xmath42 in that representation . furthermore utilizing the field - state correspondence ,
i.e. @xmath127 , we can translate these properties into @xmath128 insertion of the relation into the l.h.s .
of the first equation in the above yields @xmath129 in the last equality we used the explicit form of the eigenvalue of the quadratic casimir . the quantity @xmath130 denotes the weyl vector , i.e. the sum of the fundamental weights @xmath131 , where @xmath132 is the rank of @xmath41 . comparing the r.h.s . in the first equation in with the above
, we find @xmath133 since the virasoro generators are expressed in terms of affine generators as , the other arbitrary states are of the form @xmath134 with @xmath135 positive integers .
let @xmath136 and @xmath137 be the descendent fields corresponding to the states @xmath138 and @xmath139 , respectively . combining the ope and the mode expansions , one finds that the correlation functions @xmath140 and @xmath141 , where @xmath142 , satisfy the following equations : @xmath143 , \nn \\ & \bra ( j_{-l}^a \psi_\lam)(z ) x\ket=\mathcal{j}_{-l}^a\bra \psi_\lam(z ) x \ket , \quad & & \mathcal{j}_{-l}^a=\sum_{j=1}^m \frac{t_{\lam_j}^a}{(z_j - z)^l}. \label{lj - op}\end{aligned}\ ] ] note that the global @xmath40-invariance requires @xmath144 . using the ward identity with constant @xmath145 and the ope
, we obtain a relation satisfied by the correlation function : @xmath146 this constraint together with the global conformal invariance ( or equivalently @xmath147-invariance ) fix the structure of the two- and three - points correlation functions .
to close this section , let us derive a crucial equation called knizhnik - zamolodchikov ( kz ) equation @xcite satisfied by the correlation functions of the wzw primary fields .
the constraint stems from the fact that the virasoro generators are not independent of the affine generators as in . for @xmath148 , we have @xmath149 .
then a null state @xmath150 is given by @xmath151 where we used the property in .
using the field - state correspondence and inserting the property into the correlation functions @xmath152 , one obtains the kz equation : @xmath153 here we used the translation invariance @xmath154 , which can be easily verified from and the ope by setting @xmath155 .
now we generalize the sle and for the system containing multiple random interfaces with additional lie algebra symmetries . in the infinitesimal time interval
, we expect that each sle interface grows under an independent martingale .
let us discuss the case where the number of the interfaces is @xmath55 .
then the uniformizing map @xmath10 with the hydrodynamic normalization may be of the form @xcite ( see also @xcite for other approaches ) .
@xmath156 where @xmath157 s mean infinitesimal time intervals satisfying the condition @xmath158 .
the random processes @xmath159 ( @xmath160 ) , which play a role as driving forces for the growth of interfaces , should be written as the it stochastic differential equations : @xmath161 where @xmath162 is an @xmath163-valued brownian motion whose expectation value and variance are , respectively , given by @xmath164=0 , \quad \mathbf{e}[d\xi_{\alpha t } d\xi_{\beta t}]=\delta_{\alpha\beta}dq_\alpha .
\label{br1}\ ] ] namely @xmath157 prescribes the growth rate of each interface . the quantity @xmath165 denotes a drift term proportional to @xmath166 , which comes from interactions among interfaces , and will be determined later by the sle martingale .
to extend to evolutions with lie algebra symmetries , we define a stochastic process @xmath167 $ ] ( @xmath168 ) living on a lie group manifold @xmath121 , where @xmath44 is written as @xmath169 with @xmath170 here @xmath171 is an @xmath172-valued brownian motion with @xmath173=0 , \quad \mathbf{e}[d\vartheta^a_{\alpha t } d\vartheta_{\beta t}^b]= \delta^{ab}\delta_{\alpha\beta}dq_\alpha , \label{br2}\ ] ] and @xmath174 stands for a drift term proportional to @xmath157 , which will also be determined later . for @xmath175 , we must define @xmath176 the growth of interfaces with affine lie algebra symmetries is described by both the geometric and the algebraic components .
the remaining problem is to determine the drift terms appearing in the driving forces and in the brownian motion . to achieve it
, we must evaluate the variation of the sle martingale @xmath54 involving wzw primary fields .
for the wzw models , the sle martingale is written as @xmath177 where @xmath80 and @xmath178 ( see section 2.1 for details ) .
note that the primary fields @xmath179 and @xmath180 constructing the correlation function take values in the representation specified by the highest weights @xmath181 and @xmath182 , respectively .
the drift terms are determined by the condition which makes @xmath54 to be a martingale . to simplify the notations
we sometimes omit the index @xmath13 , @xmath183 and @xmath184 ( e.g. @xmath185 , @xmath186 , @xmath187 , @xmath188 , etc . ) , if there is no ambiguity .
let @xmath189 , @xmath190 and @xmath191 be the denominator , numerator and jacobian factor of , respectively .
first we explicitly compute @xmath192 .
a simple manipulation leads to @xmath193 where in the last equality , we applied the geometric component of the sle .
thus we obtain the variation of the jacobian factor : @xmath194 next we shall calculate @xmath195 .
the it derivative of the bulk fields @xmath196 is given by @xmath197\phi(y).\end{aligned}\ ] ] here we applied to the first term , and to the second and third terms .
for the third term we also used the property .
similarly , for the boundary fields @xmath198 , we obtain @xmath199 \phi_{\alpha}(x_\alpha ) \nn \\ & = \left[dq_\alpha\frac{\kappa}{2}\der_{x_\alpha}^2 + d x_\alpha \der_{x_\alpha } + \sum_{a=1}^{\dim\g}\sum_{\substack{\beta=1\\ \beta\ne\alpha}}^m \left(\frac{dp_\beta^a t_{\alpha}^a}{x_\alpha - x_\beta}+ \frac{\tau}{2 } \frac{dq_\beta t_\alpha^at_\alpha^a}{(x_\alpha - x_\beta)^2}\right ) \right]\psi_{\alpha}(x_\alpha).\end{aligned}\ ] ] the above two relations together with give the form of @xmath200 .
explicitly it reads @xmath201 where we define @xmath202 and @xmath203 ( resp .
@xmath204 ) is given by subtracting the first sum depending on @xmath205 from the r.h.s .
of the first ( resp .
second ) equation in .
note that the operator @xmath206 ( resp .
@xmath207 ) characteristically appears in the correlation functions between a descendent field @xmath208 ( resp .
@xmath209 ) and some composite primary field ( see ) . to derive , we have also applied the identity @xmath210 , which comes from the translation invariance of the correlation functions ( see the explanation below ) . in completely the same manner , @xmath211
can also be evaluated : @xmath212 by construction , we must set @xmath213 ( @xmath214 ) for the conformal weights of the bcc operators .
furthermore , due to the constraint between the conformal weight @xmath215 and the representation @xmath125 , all the bcc operators @xmath216 except for @xmath217 must take values in the representation specified by @xmath183 , or its conjugate @xmath218 ( note that @xmath219 holds ) .
the conformal weight @xmath220 for the bcc operator inserted at @xmath221 is determined by fusion rules .
thanks to this restriction , we can determine , in principle , the relation of the parameters @xmath36 , @xmath222 and the weight @xmath223 of the bcc operators by considering the case that there is only a single interface ( @xmath224 ) as in @xcite . we see this in the next subsection . for the single sle ( set @xmath225 and @xmath224 ) , one sees @xmath226 due to the translation invariance of the two - point correlation functions ( it can also be obtained directly from by setting @xmath224 ) .
therefore the sle martingale gives a constraint to the correlation function @xmath190 , i.e. @xmath227=0 $ ] . from the relation and the definitions and ,
one easily see that this restriction leads to @xmath228 , @xmath229 and @xmath230 namely any drift terms do not show up in the driving forces and , as expected .
moreover the constraint indicates that the bcc operators must have null states at level 2 .
namely @xmath54 is a martingale if and only if the bcc operators have the following null states at level 2 : @xmath231 here @xmath110 and @xmath112 are , respectively , the virasoro and affine generators .
this is equivalent to the conditions @xmath232 and @xmath233 , which respectively lead to @xmath234|\psi_{\lam_1}\ket=0 , \quad ( \kappa + \tau h^{\vee}-4)j_0^b|\psi_{\lam_1}\ket=0 .
\label{ns - condition}\ ] ] this necessary and sufficient condition is rather involved .
for some simple cases ( e.g. @xmath0 ) , however , we can directly solve the above equations by acting the generator @xmath235 .
( note that more elegant procedure utilizing the kz - equation has been developed in @xcite . ) for later convenience , here we explicitly write down the results for the @xmath0-wzw model . the central charge @xmath117 and the conformal weight @xmath236 of the bcc operator @xmath237 in the spin-@xmath238 representation
are , respectively , written as @xmath239 where @xmath240 denotes the fundamental weight , and we used @xmath241 .
from the direct evaluation of for @xmath0 case , one finds that the conditions in are valid only for the case that the bcc operator carries spin-1/2 ( @xmath242 ) @xcite
. then @xcite @xmath243 for @xmath244 , @xmath36 and @xmath222 can not be specified , and only the relation @xmath245 is imposed .
this case , however , corresponds to a @xmath246 cft ( cf . ) , and therefore we shall set @xmath247 .
now we identified the bcc operators @xmath249 .
namely they have null states at level 2 and must satisfy the condition . utilizing the field - state correspondence , one finds that the first sums in and vanish .
thus the drift term of the it derivative of the cft correlation function @xmath250 is explicitly given by @xmath251=&j_t^\phi \sum_{\alpha=1}^m\left(df_\alpha - \kappa dq_\alpha \der_{x_\alpha}\log z_t -2\sum_{\substack{\beta=1\\\beta\ne \alpha}}^m \frac{dq_\beta } { x_\alpha - x_\beta } \right)\der_{x_\alpha}\left(\frac{z_t^\phi}{z_t}\right ) \nn \\
& \qquad+ \frac{j_t^\phi}{z_t } \sum_{\alpha=1}^m\sum_{a=1}^{\dim\g } \left [ \left(dg_\alpha^a- \frac{\tau dq_\alpha}{z_t}\widetilde{j}^{a}_{\alpha,-1}z_t \right)\left(j_{\alpha,-1}^a z_t^\phi- \frac{z_t^\phi\widetilde{\j}_{\alpha,-1}^a}{z_t}z_t\right ) \right],\end{aligned}\ ] ] where we substituted the relations and . by recalling that @xmath54 is the sle martingale
, the above quantity must be zero .
thus one finds the drift terms @xmath252 and @xmath253 are described by the partition function @xmath189 : @xmath254 to summarize , we have constructed the multiple sles for @xmath255-wzw models : @xmath256 the drift terms @xmath165 and @xmath174 are given by @xmath257 where @xmath162 ( resp .
@xmath171 ) is the @xmath163 ( resp .
@xmath172)-valued brownian motion whose expectation value and variance are given by @xmath258=0 , \quad & & \mathbf{e}[d\xi_{\alpha t } d\xi_{\beta t}]=\delta_{\alpha\beta}dq_\alpha , \nn \\ & \mathbf{e}[d\vartheta^a_{\alpha t } ] = 0 , \quad & & \mathbf{e}[d\vartheta^a_{\alpha t } d\vartheta_{\beta t}^b]= \delta^{ab}\delta_{\alpha\beta}dq_\alpha.\end{aligned}\ ] ] @xmath189 is a partition function involving the bcc operators : @xmath259 where all the bcc operators @xmath216 except for @xmath217 must take values in representations specified by @xmath183 , or its conjugate @xmath218 .
the conformal weight @xmath220 for the bcc operator inserted at @xmath221 is determined by fusion rules ( see next section for example ) .
the structure of the partition function is described by both the global @xmath40-invariance ( cf . ) , and the kz - equation ( cf . )
as a concrete application of our formula , let us consider multiple sles for the @xmath0-wzw model where the bcc operators carry spin-1/2 , and discuss topologies for the sle traces
. for the @xmath0 case , the fusion rules of the primary fields are similar to those in the minimal cfts .
therefore it is natural to extend the argument @xcite ( see also @xcite ) describing the geometric configurations of the sle traces in minimal cfts to the @xmath261-wzw model .
thus we make the following conjecture .
[ conj - config ] there exists a one - to - one correspondence between topologically inequivalent configurations of @xmath0 multiple sle traces and the independent solutions of the kz - equation satisfied by the partition functions .
we describe this conjecture more specifically .
let @xmath159 ( @xmath160 ) be the positions where the @xmath55 sle traces start to grow , and be ordered as @xmath262 .
consider the case that the traces eventually form @xmath263 disjoint curves in @xmath1 so that each point @xmath159 is an end point of exactly one curve and @xmath221 is an end point of exactly @xmath264 curves .
namely @xmath120 disjoint curves form arches ( more precisely @xmath120 pairs of growing curves hit each other s tips and consequently form @xmath120 arches ) and other @xmath265 curves converge toward the point at @xmath221 ( see fig . [ config ] for @xmath266 and @xmath267 ) .
then the number of topologically inequivalent configurations are given by @xmath268 this is nothing but a kostka number appearing as the coefficient of the irreducible decomposition for the @xmath55 tensor product of the @xmath269 fundamental representation @xmath270 into @xmath271 ( @xmath272 denotes the integral representation with the highest weight @xmath273 , where @xmath274 and @xmath240 is the fundamental weight of @xmath269 ) : @xmath275 where @xmath276 denotes the integer part of @xmath277 .
now we mention that the relation between the geometric configurations and the cft partition functions @xmath278 ( hereafter we omit the index @xmath13 to simplify the notation ) . to this end
, we consider the fusions of the @xmath0 primary fields : @xmath279 where @xmath280 stands for the @xmath0 bcc primary field taking values in the integral representation @xmath272 . for sufficiently large @xmath97 ,
the above rules reduce to the decomposition . applying the fusion procedures to the bcc operators @xmath180 ( @xmath160 ) recursively
, one can reduce @xmath278 to a two point function involving the fusion operator and the bcc operator at @xmath221 .
thus , for the non - vanishing partition functions , the conformal weight @xmath220 of the bcc operator @xmath217 must be equivalent to that of the fusion operator .
[ fusion - fig ] shows the possible conformal weight @xmath220 up to @xmath266 . due to the fusion procedures , for sufficiently large @xmath97
, one finds that the number of the paths from the left to @xmath281 corresponds to the number of topologically distinct configurations for the sle traces , i.e. @xmath282 . for generic @xmath97 , the number of paths is , in general , constrained by the fusion rules , which affects the structure of the partition functions , and therefore the realization of the geometric configurations . and @xmath267 .
there are two ( @xmath283 ) topologically inequivalent configurations.,scaledwidth=70.0% ] for instance , the configuration where all the curves eventually converge toward @xmath221 ( i.e. no arches ) corresponds to the path to the weight @xmath284 . by a standard argument for the cft correlation functions ,
one obtains the corresponding partition function @xmath278 : @xmath285 one can easily check that this partition function satisfies the kz - equation .
the factorized correlation function with the same exponents @xmath286 does not exist for @xmath287 , indicating that the no - arch configuration is allowed only when @xmath288 . according to the argument in @xcite
, we confirm our conjecture by considering the simplest case ( @xmath289 ) , where only two sle traces exist in the upper half plane @xmath1 .
to analyze specifically , let us write down the partition function @xmath278 .
up to a constant factor , it is given by @xmath290 where the exponent is @xmath291 .
the fusion rules ( see also fig . [ fusion - fig ] ) indicate that @xmath292 ( @xmath293 ) or @xmath294 ( @xmath295 ) .
then using the relation , one arrives at @xmath296 ( @xmath293 ) or @xmath297 ( @xmath295 ) .
correspondingly the partition functions are @xmath298 by inserting them into and , the driving processes @xmath299 and @xmath300 characterizing the geometric aspects of the sle traces become : @xmath301 where we normalized the variances by @xmath302 so that @xmath303 with two independent standard brownian motions : @xmath304=0 $ ] and @xmath305=\delta_{\alpha \beta } dt$ ] .
these driving processes with the sle describe two curves emerging from two points @xmath306 and @xmath307 .
setting @xmath308 and rescaling the time by @xmath309 , we reduce the processes to the following bessel process : @xmath310 with the effective dimension @xmath311
. substitution of the exponent @xmath312 and @xmath36 yields @xmath313 recalling that the bessel process is recurrent ( resp . not recurrent ) if @xmath314 ( resp .
@xmath315 ) ( see @xcite , for example ) , we conclude that the driving processes @xmath316 hit each other with probability 1 for @xmath294 and never hit for @xmath292 .
the hit of the driving processes means the hit of the tips of the sle traces , and hence the case for @xmath294 describes a single curve ( i.e. single arch ) whose end points are @xmath317 and @xmath318 , while the case for @xmath319 describes two curves converging to the point at @xmath221 .
the above observation agrees with conjecture [ conj - config ] . . the primary fields with weight @xmath320 for @xmath321 do not exist due to the fusion rules . for instance
, for @xmath322 , the primary fields in the shadow area vanish.,scaledwidth=40.0% ] let us discuss a more non - trivial example : the triple sles ( @xmath323 ) .
the geometric properties are characterized by the partition function : @xmath324 where @xmath325 and by construction the bcc operators @xmath326 ( @xmath327 ) take values in the fundamental representation @xmath328 . from the fusion rules
( see also fig .
[ fusion - fig ] ) , the conformal weight @xmath329 of the bcc operator @xmath330 inserted at @xmath221 is @xmath331 ( @xmath332 ) or @xmath333 ( @xmath295 ) .
when the level takes its value in the range @xmath332 and @xmath334 , there exists the factorized partition function with the same exponents @xmath286 : @xmath335^{\frac{1}{2(k+2 ) } } \quad ( k \ge 3).\ ] ] according to conjecture [ conj - config ] , we expect that this partition function describes three curves converging toward @xmath221 . and
there are two ( @xmath337 ) topologically inequivalent configurations .
these configurations are described by the partition function involving the 4 bcc operators with the same conformal weights @xmath338 .
the configuration in the left ( resp .
right ) panel is characterized by the partition function @xmath339 ( resp .
@xmath340 ) . ,
scaledwidth=70.0% ] the case for @xmath333 is more interesting , because two topologically inequivalent configurations do exist ( see fig .
[ config2 ] ) . by conformal mapping @xmath341 ,
the four points @xmath342 are transformed to the points @xmath343 , @xmath344 , @xmath345 and @xmath346 , where @xmath347 .
thus the partition function @xmath278 can be expressed as @xmath348 the correlation function @xmath349 can be calculated by solving the kz - equation @xcite .
its explicit form reads @xmath350 with @xmath351 where @xmath352 s are expressed by the hypergeometric function @xmath353 : @xmath354 and the coefficients @xmath355 are respectively , given by @xmath356 note that @xmath357 and @xmath358 satisfy the relation @xmath359 .
for @xmath244 , @xmath360 holds do not exist for @xmath244 .
] , and then the partition functions reduce to simple forms @xmath361 for generic @xmath97 , the behaviors of the partition functions close to the position @xmath362 and @xmath363 are described as @xmath364 thus the fusion rules and conjecture [ conj - config ] imply that @xmath339 describes two curves connecting the points @xmath365 $ ] and @xmath366 $ ] ( configuration @xmath367 ; see the left panel in fig [ config2 ] ) while @xmath368 corresponds to the configuration @xmath369 $ ] and @xmath370 $ ] ( configuration @xmath371 ; see the right panel in fig [ config2 ] ) .
obtained by applying a conformal transformations @xmath372 to the configurations in fig .
[ config2 ] ( @xmath373 denotes @xmath374 ) .
the probability of the occurrence of the configuration ( a ) ( resp .
( b ) ) is exactly the same as that of the occurrence of the configuration ( a ) ( resp .
( b ) ) in fig .
[ config2].,scaledwidth=85.0% ] utilizing the partition functions , we can exactly compute the probability @xmath375 $ ] ( resp .
$ ] ) of the occurrence of the configuration @xmath367 ( resp .
@xmath371 ) with the initial condition that the three curves emerging from the point @xmath377 , @xmath378 , @xmath379 by @xmath380=\frac{z_{\rm c_1}(x)}{z_{\rm c_1}(x)+z_{\rm c_2}(x ) } , \quad \mathbf{p}[{\rm c_2}]=\frac{z_{\rm c_2}(x)}{z_{\rm c_1}(x)+z_{\rm c_2}(x)}.\ ] ] ( see @xcite for similar formulas for minimal cfts . ) since the measure is conformally invariant , the probabilities in arbitrary ( simply connected ) domain @xmath381 can be easily reproduced by use of @xmath349 and the conformal map @xmath372 mapping @xmath1 to @xmath381 . more explicitly ,
the probability @xmath382 $ ] ( resp .
@xmath383 $ ] ) is exactly the same as that of the occurrence of the configuration with two curves joining the boundary points @xmath384 $ ] and @xmath385 $ ] ( resp .
@xmath386 $ ] and @xmath387 $ ] ) on @xmath388 ( see fig .
[ crossing - fig ] ) ( note that @xmath373 stands for @xmath374 @xmath389 ) ) . in this sense
, the arch probabilities can be interpreted as crossing probabilities of the interfaces . for @xmath244 ,
the formula becomes very simple : @xmath380=1-x \quad ( k=1).\ ] ] in fig .
[ arch - fig ] , the arch probabilities @xmath375 $ ] are depicted for several values of @xmath97 . as shown in fig .
[ arch - fig ] , the @xmath277-dependence of the arch probabilities close to @xmath390 in the wide range of @xmath277 , as increasing @xmath97 .
$ ] for various level @xmath97.,scaledwidth=70.0% ]
the present work was partially supported by grants - in - aid for young scientists ( b ) no .
21740285 and for scientific research ( c ) no .
24540393 from japan society for the promotion of science .
o. schramm , isr .
* 118 * ( 2000 ) 221 .
[ arxiv : math/9904022v2 ] .
s. rohde and o. schramm , ann . math .
* 161 * ( 2005 ) 879 .
[ arxiv : math/0106036v4 ] .
w. werner , _ random planar curves and schramm - loewner evolutions _ , in _ lectures on probability theory and statistics _ , lecture notes in mathematics vol .
1840 , pp 107 - 195 , springer , berlin , 2004 .
[ arxiv : math/0303354v1 ] .
w. kager and b. nienhuis , j. stat . phys .
* 115 * ( 2004 ) 1149 .
[ arxiv : math - ph/0312056v3 ] .
j. cardy , ann . phys .
* 318 * ( 2005 ) 81 .
[ arxiv : cond - mat/0503313v2 ] . g.f .
lawler,_conformally invariant processes in the plane _ , mathematical surveys and monographs , vol .
114 , american mathematical society , providence , ri , 2005 .
m. bauer and d. bernard , phys .
* 432 * ( 2006 ) 115 .
[ arxiv : math - ph/0602049v1 ] .
gruzberg , j. phys . a : math
* 39 * ( 2006 ) 12601 .
[ arxiv : math - ph/0607046v2 ] .
s. smirnov , proc .
congr . math .
* 2 * ( 2006 ) 1421 .
[ arxiv:0708.0032v1 ] .
m. bauer and d. bernard , phys .
b * 543 * ( 2002 ) 135 .
[ arxiv : math - ph/0206028v3 ] .
m. bauer and d. bernard , comm .
* 239 * ( 2003 ) 493 .
[ arxiv : hep - th/0210015v3 ] .
r. friedrich and w. werner , c.r .
paris , ser .
* 335 * ( 2002 ) 947 .
[ arxiv : math/0209382v2 ] .
r. friedrich and w. werner , comm .
* 243 * ( 2003 ) 105 .
[ arxiv : math - ph/0301018v3 ] .
m. bauer and d. bernard , phys .
b * 557 * ( 2003 ) 309 .
[ arxiv : hep - th/0301064v2 ] .
m. bauer and d. bernard , ann .
poincar * 5 * ( 2004 ) 289 .
[ arxiv : math - ph/0305061v1 ] .
m. bauer and d. bernard , phys .
b * 583 * ( 2004 ) 324 .
[ arxiv : math - ph/0310032v1 ] .
m. bauer , d. bernard , k. kytl , j. stat .
* 120 * ( 2005 ) 1125 .
[ arxiv : math - ph/0503024v2 ] .
belavin , a.m. polyakov , a.b .
zamolodchikov , nucl .
b * 241 * ( 1984 ) 333 .
di francesco , p. mathieu , and d. snchal , _ conformal field theory _ , springer , new york , 1997 .
novikov , usp . mat .
nauk * 37 * ( 1982 ) 3 .
e. witten , comm .
* 92 * ( 1984 ) 455 .
knizhnik and a.b .
zamolodchikov , nucl .
b * 247 * ( 1984 ) 83 .
e. bettelheim , i.a .
gruzberg , a.w.w .
ludwig and p. wiegmann , phys .
lett * 95 * ( 2005 ) 251601 .
[ arxiv : hep - th/0503013v3 ] .
a. alekseev , a. bytsko and k. izyurov , lett . math
* 97 * ( 2011 ) 243 .
[ arxiv:1012.3113v2 ] .
j. rasmussen , afr .
. phys . * 4 * ( 2007 ) 1 .
[ arxiv : hep - th/0409026v2 ] .
j. rasmussen , lett .
* 68 * ( 2004 ) 41 .
[ arxiv : math - ph/0312010v1 ] .
j. rasmussen , j. stat .
( 2004 ) p09007 .
[ arxiv : math - ph/0408011v2 ] .
j. nagi and j. rasmussen , nucl .
b * 704 * ( 2005 ) 475 .
[ arxiv : math - ph/0407049v1 ] .
r. santachiara , nucl .
b * 793 * ( 2008 ) 396 .
[ arxiv:0705.2749v2 ] .
a. nazarov , jetp lett .
* 96 * ( 2012 ) 93 - 96 .
[ arxiv:1205.6104v1 ] .
a. nazarov , j. phys : conf . ser . * 343 * ( 2012 ) 012085 .
[ arxiv:1112.4354v1 ] .
j. cardy , nucl .
b * 240 * ( 1984 ) 514 .
j. cardy , j. phys .
a : math . gen .
* 36 * ( 2003 ) l379 .
[ arxiv : math - ph/0301039v49 ] .
j. dubdat , comm .
pure appl . math .
* 60 * ( 2007 ) 1792 .
[ arxiv : math/0411299v4 ] .
k. graham , j. stat .
( 2007 ) p03008 .
[ arxiv : math - ph/0511060v1 ] .
kozdron and g.f .
lawler , fields inst .
* 50 * ( 2007 ) 199 .
[ arxiv : math/0605159v2 ] .
kozdron , j. phys . a : math
* 42 * ( 2009 ) 265003 .
[ arxiv:0905.2430v1 ] . | we provide multiple schramm - loewner evolutions ( sles ) to describe the scaling limit of multiple interfaces in critical lattice models possessing lie algebra symmetries .
the critical behavior of the models is described by wess - zumino - witten ( wzw ) models . introducing a multiple brownian motion on a lie group as well as that on the real line
, we construct the multiple sle with additional lie algebra symmetries .
the connection between the resultant sle and the wzw model can be understood via sle martingales satisfied by the correlation functions in the wzw model . due to interactions among sle traces ,
these brownian motions have drift terms which are determined by partition functions for the corresponding wzw model . as a concrete example
, we apply the formula to the @xmath0-wzw model . utilizing the fusion rules in the model
, we conjecture that there exists a one - to - one correspondence between the partition functions and the topologically inequivalent configurations of the sle traces .
furthermore , solving the knizhnik - zamolodchikov equation , we exactly compute the probabilities of occurrence for certain configurations ( i.e. crossing probabilities ) of traces for the triple sle . |
spin - orbit couplings open a route to control quantum electron spin by electric means .
one of the efficient methods to inject or generate electron spin in non - magnetic semiconductors is the spin hall effect , in which an electric current or an electric field may induce a transverse spin current in the systems with strong spin - orbit couplings .
early theories @xcite proposed that the spin current is caused by asymmetric scattering of electrons with spin - up and -down in impurity potentials , named as extrinsic spin hall effect . in recent years
it was demonstrated that the spin - orbit coupling in the electron bands can also lead to an intrinsic spin hall effect in either p - doped or n - doped semiconductors.@xcite both extrinsic and intrinsic spin hall effects were confirmed experimentally in various systems.@xcite a two - dimensional electron gas ( 2deg ) with a rashba coupling was proposed to exhibit an intrinsic spin hall effect.@xcite the spin - orbit coupling in 2deg modifies the electron band structure , and may lead to interesting magnetotransport properties , such as the beating phenomenon in the shubnikov - de haas ( sdh ) oscillation.@xcite when the system is subjected to an external magnetic field , the zeeman splitting will also change the spin - dependent electron bands .
the interplay of the spin - orbit coupling and the zeeman coupling produces the crossing of electron energy levels . based on this property
, it was proposed that a tiny electric field may remove the additional degeneracy of energy levels and produces a finite spin current if the fermi surface sweeps across the crossing point of energy levels . as a result
, there exhibits a divergent spin hall conductance.@xcite this resonant spin hall effect was also discussed in p - doped systems in a magnetic field.@xcite however , impurities in the system make the issue more subtle.@xcite the vertex correction in the self energy turns out to cancel the spin hall conductance even in a weak disorder limit in 2deg with linear rashba coupling , while the spin hall conductance survives in p - doped luttinger model and the systems with cubic spin - orbit couplings.@xcite the disorder effect strongly depends on the symmetry of the spin - orbital coupling and the dispersion .
now whether the resonant spin hall effect can survive in a finite density of impurities becomes an issue to be answered .
this is the motivation of the present work . here
we present a full investigation on the disorder effect of resonant spin hall effect in 2deg with the rashba coupling in a tilted magnetic field .
the impurity effect is considered by the self - consistent born approximation ( scba ) and the vertex correction in the ladder approximation .
we found the impurity effect will suppress the resonant spin hall conductance at the resonant point , and produce a double peak structure of the spin hall conductance around the point .
a tilted magnetic field is applied to enhance the effective zeeman splitting,@xcite and to recover the effect when the energy level splitting is larger than the energy broadening by impurities
we consider a 2deg in the x - y plane with the rashba spin - orbit interaction in a tilted magnetic field .
the perpendicular component of the tilted field is @xmath0 and the in - plane component is chosen to be along the x - direction @xmath1 , where @xmath2 is the angle between the field and the z - direction .
we take the laudau gauge for the vector potential of the field @xmath3 .
the total hamiltonian including the zeeman energy is given by@xmath4+\frac{\lambda}{\hbar}[(p_{x}+eb_{\bot}y)\sigma_{y}-p_{y}\sigma_{x}]\nonumber\\ & -\frac{1}{2}g_{s}\mu_{b}b_{\bot}\sigma_{z}+\frac{1}{2}g_{s}\mu_{b}b_{\bot } \tan\theta\sigma_{x } , \label{h_soc}\ ] ] where @xmath5 @xmath6 are the electron s effective mass , charge , and lande @xmath7 factor , respectively .
@xmath8 is the bohr magneton , @xmath9 is the strength of rashba spin - orbit coupling , and @xmath10 are the pauli matrices .
we take a periodic boundary condition along the x direction , hence the momentum @xmath11 is a good quantum number .
an analytical solution can be obtained in the case of @xmath12.@xcite there were some studies on spin transport based on the solution.@xcite inclusion of the tilted field makes the problem much more complicated , and an analytical solution is not available at present . in the following approach ,
we choose the energy eigenstates for the system without the spin - orbit coupling ( @xmath13 ) and @xmath12 as a set of basis , @xmath14 where the magnetic length @xmath15 , the spin index @xmath16 and @xmath17 is the length of the 2deg , @xmath18 is the eigenstate of the @xmath19 energy level of a linear oscillator with the frequency @xmath20,@xcite and @xmath21 is the eigenstate of spin @xmath22 . when the tilt angle @xmath12 , the system can be solved exactly.@xcite the eigenvalues of @xmath23 are given by@xmath24 where @xmath25 and @xmath26 with @xmath27 the mass of a free electron , @xmath28 for @xmath29 , and @xmath30 for @xmath31 .
the states @xmath32 have a degeneracy @xmath33 , corresponding to @xmath34 values of @xmath35 .
the eigenstate has the form,@xmath36 where @xmath37 , and for @xmath31 , @xmath38 with @xmath39 .
one of the features of the solution is the crossing of the energy levels as functions of the magnetic field , which is caused by the competition between the spin - orbit coupling and zeeman energy splitting .
for the two levels @xmath40 and @xmath41 , the condition for the crossing is determined by@xcite@xmath42 this point is called the resonant point for resonant spin hall effect .
this additional degeneracy due to the competition between the spin - orbit coupling and the zeeman energy of the perpendicular field can be removed by a tilted field . in the case of @xmath43
the energy levels can be calculated numerically . using the expression in eq .
( 4 ) , we may make a truncation approximation by keeping the landau levels with @xmath44 such that the dimensionality of matrix is reduced to @xmath45 numerical diagonalization of the matrix can give us the energy eigenvalues .
alternatively , the gap can also be calculated approximately by the degenerate perturbation theory .
we take the partial hamiltonian @xmath46 as a perturbation , and express it in the subspace spanned by the two states @xmath47 and @xmath48 near the resonant point,@xmath49{cc}0 & i\delta/2\\ -i\delta/2 & 0 \end{array } \right ] , \label{h'}\ ] ] where the gap @xmath50 is@xmath51 in fig .
[ gap ] , we present the energy levels of @xmath12 as a function of @xmath52 the parameters used are @xmath53 , @xmath54 , and @xmath55.@xcite the red arrow denotes a level crossing at @xmath56 .
the insert shows the energy gap as a function of the tilt angle @xmath2 with @xmath57 .
we notice that the numerical and analytical results are in good agreements .
, the arrow denotes a level - crossing point , which develops into a gap when the tilt angle increases .
the inset shows the gap as a function of the tilt angle , reflects the accordance of the numerical calculation and the analytic expression eq .
( [ delta ] ) . the energy has been scaled by @xmath58 , scaledwidth=45.0% ] in this section , we briefly review the general formalism of linear response theory of self - consistent born approximation ( scba ) for electron transport .
we shall use this technique to investigate the transport properties of 2deg with a rashba coupling in a tilted magnetic field , especially near the resonant point .
the effect of impurities will be taken into account in this formalism .
we consider a random configuration of impurities with short - range potentials @xmath59 , where @xmath60 is the position of the @xmath61 impurity .
the density of the impurities is @xmath62 . generally speaking
, the green s functions for a specific configuration of random potential can be written as @xmath63^{-1}$ ] , where @xmath64 and @xmath65 correspond to the retarded and advanced green s function , respectively .
all transport quantities can be expressed in terms of the green s function after averaging all possible configurations of the impurities . using the conventional perturbation expansion with respect to @xmath66
, we can obtain the dyson equation for the averaged green s function @xmath67 .
the impurity effect is absorbed by a self - energy function @xmath68 as follows , @xmath69,\ ] ] where @xmath70 means the average over all the impurity configurations . in the scba
, the self - energy operator can be expressed by @xmath71.@xcite in the representation of the landau levels , @xmath72 , @xmath73 , and @xmath74 are expressed as matrices .
for such a spin - independent impurity potential , previous works@xcite proved that the self - energy is independent of @xmath75 and @xmath35 for a spin - independent landau system .
we find that the self - energies for a spin - orbit coupling system are independent of @xmath75 and @xmath35 , @xmath76 here we dropped the index @xmath35 in @xmath67 because the averaged green s functions are k - independent . with the averaged green s function in mind
, we can use the kubo formula to calculate the linear response of any physical quantity @xmath77 to an external electric field @xmath78 in the @xmath79 direction , @xmath80 as the single - particle version of the kubo formula , the streda s formula is a conventional and powerful tool to study the transport property of 2deg system under a magnetic field.@xcite at the zero temperature , the formula is given by@xmath81\right\rangle _ { c } , \label{o}\ ] ] where @xmath82 is the spectral function and @xmath83 $ ] is the velocity operator , . because there are products of two green s functions in the impurity average @xmath70
, the vertex correction has to be included . for a specific density of charge carriers ,
the fermi energy as a function of the magnetic field is determined by @xmath84\right\rangle _ { c}.\ ] ] usually the streda s formula is applied to calculate the electric conductance by replacing @xmath77 by an electric current operator , @xmath85 . in the present work , we intend to explore the spin transport in the system .
the spin current is defined as @xmath86 , which is a tensor determined by both the motion direction of an electron and its polarization . in the framework of linear response theory ,
the spin hall conductivity @xmath87 , the ratio of the spin hall current to an external field , can be calculated by substituting @xmath88 ( @xmath89 ) in eq .
( [ o ] ) .
the spin hall conductivity comes from the contribution of all the electrons below the fermi level .
opposite to those for the conductivity and hall conductivity , it can not be reduced to a fermi edge quantity@xcite because the spin current is not a commutator of any operator and the hamiltonian.@xcite for the purpose of our numerical calculation , we transform the streda s formula into the following form,@xmath90 , \label{sigma_alpha_munu}\ ] ] where @xmath91 ^{+}. \label{k2}\ ] ] @xmath92 are determined in a set of bethe - salpeter - like equations , @xmath93 ^{-1}f_{\nu}^{+-}+f_{\nu}^{+-}\left [ g^{-}\right ] ^{-1}\frac{dg^{-}}{de}\nonumber\\ & + g^{+}\left\langle vk_{\nu}^{+-}v\right\rangle _
{ c}g^{-},\label{k+-}\\ k_{\nu}^{++ } & = \frac{dg^{+}}{de}\left [ g^{+}\right ] ^{-1}f_{\nu}^{++}+g^{+}\left\langle vk_{\nu}^{++}v\right\rangle _ { c}g^{+}.\end{aligned}\ ] ] where @xmath94 ( @xmath95 ) is the vertex operator , which satisfy the bethe - salpeter equation in the ladder approximation,@xmath96 these equations can be solved self - consistently and eq .
( [ k+- ] ) and ( [ bs ] ) have multi - solutions . from the continuity equation for charge current in equilibrium
, we can derive an auxiliary equation,@xmath97=0 . \label{con1}\ ] ] differentiating eq .
( [ con1 ] ) with respect to @xmath98 leads to another auxiliary equation,@xmath99=-tr[f_{\nu}^{+-}\frac{d\left ( \sigma^{-}-\sigma^{+}\right ) } { de}].\ ] ] @xmath100 is determined by an another self - consistent equation , @xmath101 since the density of states for each landau level in scba has a semi - elliptic form,@xcite it approaches zero and @xmath100 becomes infinity at the edge of each level .
we find the integrand in eq .
( [ sigma_alpha_munu ] ) is always convergent in numerical calculation , because the concurrence of @xmath102 and @xmath103 in eq .
( [ k1 ] ) , ( [ k2 ] ) .
now we are ready to calculate the spin hall conductivities numerically . in this paper ,
the electron density is fixed at @xmath104 .
this value of @xmath105 promises that the fermi level is located near the resonant point with the filling factor @xmath106 while the magnetic field sweeps over the point , i.e. , @xmath107 as indicated in fig .
the other parameters are as the same as those used in fig .
[ gap ] . the density of impurities @xmath108 and the strength of the impurity potential @xmath74 are combined in one parameter scattering strength @xmath109 .
we assign @xmath110 various values to investigate the impurity effect . , for various scattering strength @xmath110 .
, scaledwidth=45.0% ] we first discuss the disorder effect of spin hall conductance , especially near the crossing point .
we apply the formula in eq.([sigma_alpha_munu ] ) to calculate the spin hall conductivity @xmath111 around the resonant point @xmath112 numerically for various strengths of disorder .
numerical results are plotted in fig .
[ sigmazxy_impurity ] .
the dashed curve for @xmath113 is from the solution in ref . [ ] .
the key feature of the disorder effect is the suppression of the spin hall conductivity at the resonant point .
the large spin hall conductance exhibits when the field deviates from the resonant point and forms a double peak structure .
the weight of the spin hall conductivity increases as the impurity strength decreases , which reflects the intrinsic properties of the resonance . to understand the suppression of resonant spin hall conductance , we plotted in fig
. [ cross ] the distribution of the spin hall conductivity on the electron s energy , @xmath114 , around the crossing point .
the total spin hall conductivity @xmath115 are contributed by all the electron states under the fermi level .
it was observed that the energy levels are broadened due to the impurity scattering , and the distribution of the spin hall conductivity is inhomogeneous .
so the magnitude and even the sign of the spin hall conductivity can be varied by the fermi level or the electron density . in the clean limit ,
a tiny external electric field can open a gap between the crossing levels , which leads to spin hall conductance divergent .
however , after the impurity scattering is taken into account , a tiny external field can not open an energy gap any more because of the level broadening . as a result , the spin hall conductance in a weak field limit will be suppressed . however , once the external field becomes stronger than the level broadening , a large spin hall conductance will appear .
this can be seen from the case that the magnetic field deviates from the crossing point , i.e. , the additional degeneracy of the two levels will be lifted , and a strong spin hall conductance recovers . this is the physical origin of the double peak structure of the resonant spin hall conductance .
it is worth stessing that this suppression of resonant spin hall conductance is different from the case in the absence of the zeeman term .
the zeeman splitting may produce a non - zero spin hall conductance in the rashba system.@xcite to further illustrate the formation of the resonant spin hall effect , we investigate the effect of the tilted magnetic field .
[ sigmazxy_theta ] shows the dependence of the spin hall conductivity on the tilt angle near the resonant point . for the purpose of numerical calculation
, we take the scattering strength @xmath116 . as the tilted angle increases , the spin hall conductivity at the resonant point will increase very quickly , and the two peaks finally integrate into one . after that point
, the spin hall conductivity begins to decrease .
these behaviors can be understood as the competition between the disorder broadening of the energy levels and the degeneracy lifting by the tilted field . in the clean limit
, the tilted field will remove the degeneracy of the energy crossing levels as shown in fig .
we can estimate the peak height of the spin hall conductivity as a function of the the tilt angle by a perturbation calculation adopted in sec .
[ sec_hamiltonian ] . diagonalizing the truncated two - level hamiltonian in eq .
( [ h ] ) , we get the modified eigenstates @xmath117 , and the energy correction @xmath118 with eq .
( [ delta ] ) .
if the fermi level just lies between the energy levels , the spin hall conductivity is mainly attributed to @xmath119 , which can be calculated by the kubo formula,@xcite @xmath120 , \label{sigmazxy_analytic}\ ] ] where @xmath121 .
the impurity scattering will cause the level broadening . if the level broadening is larger than the gap caused by the tilted field , the impurity effect is dominant
. otherwise the tilted field effect will be dominant .
the level broadening is characterized by the half - width of the semi - elliptic density of states of the landau levels , which can be estimated approximately by@xcite @xmath122 . when @xmath123 , the effect of the tilted field turns out to be dominant .
the insert of fig .
[ sigmazxy_theta ] presents the data of the peak height of the spin hall conductivity as a function of @xmath2 from the numerical calculation ( diamond ) and from the analytic formula ( solid line ) in eq .
( [ sigmazxy_analytic ] ) , they are in a good agreement when @xmath124 , which is very close to the estimated value @xmath125 from @xmath126 . .
the insert compares the peak height of the spin hall conductivity as function of @xmath2 from the numerical calculation ( diamond ) and from the analytical formula eq .
( [ sigmazxy_analytic ] ) ( solid line ) . the scattering strength @xmath127 .
, scaledwidth=45.0% ]
in summary , we applied the streda s formula to study the disorder effect of the resonant spin hall effect in the 2deg system with the rashba interaction in a tilted magnetic field .
considering the vertex corrections in the self energy , we find that the main effect of the impurity scattering is to broaden the landau levels . in the framework of linear response ,
the electric field is taken to approach zero , and the energy splitting caused by the electric field is always less than the broadening of the landau levels .
thus a tiny external field can not remove the additional degeneracy of the energy levels at the resonant point . as a result
, the spin hall conductance will be suppressed at the point . when the magnetic field slightly deviates the resonant point or a tilted field
is applied , the degeneracy will be removed , a large spin hall conductance will be recovered .
the spin hall conductance exhibits a double peak around the resonant point . from the effect of a tilted field ,
we believe that a finite electric field , if it is strong enough to overcome the energy level broadening , will recover the spin hall effect even at the resonant point .
this is quite different from the disorder effect of spin hall effect in the rashba system in the absence of magnetic field.@xcite the work is partly motivated with the tilted magnetic field experiments in r. r. du s group , to whom we wish to thank for stimulating discussions .
this work was supported by the research grant council of hong kong under grant no .
: hku 7041/07p and hku 10/crf/08 . | we study the disorder effect of resonant spin hall effect in a two - dimension electron system with rashba coupling in the presence of a tilted magnetic field .
the competition between the rashba coupling and the zeeman coupling leads to the energy crossing of the landau levels , which gives rise to the resonant spin hall effect . utilizing the streda s formula within the self - consistent born approximation
, we find that the impurity scattering broadens the energy levels , and the resonant spin hall conductance exhibits a double peak around the resonant point , which is recovered in an applied titled magnetic field . |
classically , a particle with mass @xmath0 submitted to a time- and momentum - independent smooth potential @xmath1 in @xmath2 is constrained to move following the phase space trajectory given by the hamilton equations @xmath3 smoothness in @xmath1 guarantees that the equations above have a unique solution @xmath4 around all initial condition @xmath5 ( _ i.e. _ , for @xmath6 sufficiently small ) , thus we can define the classical hamiltonian flow @xmath7 by setting @xmath8 .
further conditions on the regularity and growth rate of @xmath1 imply more good properties , so if @xmath9 is bounded , one can extend @xmath10 for all @xmath11 , for any @xmath5 in the phase space@xcite . in quantum mechanics ,
the state evolution of a similar system is described by a function @xmath12 obeying to the schrdinger equation with initial data @xmath13 where the initial @xmath14 data satisfy @xmath15 and @xmath16 is a parameter generally reminiscent from some rescaling procedure , but that can also be seen as the planck s constant in a system with mass of order @xmath17 , in which case the wigner measures can be viewed as the classical limits of the system s mass distribution , as we will explain next .
if @xmath1 satisfies the kato - rellich conditions ( @xmath1 continuous and @xmath18 ) , then the hamiltonian operator @xmath19 with domain in the sobolev space @xmath20 is essentially self - adjoint ( @xcite,@xcite ) and has a unique solution for all @xmath11 , which is given by @xmath21 now , for @xmath11 , we define the _ wigner transform _ associated to the solution of , @xmath22 which can be interpreted as the quantum version of the mass probability density on phase space in classical statistical physics and provides a link between these two theories@xcite .
when @xmath23 , the wigner transform converges to a finite and positive measure @xmath24 on @xmath25 ( @xcite , @xcite , @xcite ) in the sense that , given a sequence @xmath26 converging to @xmath27 , we can extract a subsequence may depend on the subsequence , this is why we refer to it as _ a _ semiclassical limit , not necessarily _ the _ classical one ( examples of non - unicity in @xcite ) . ]
@xmath28 such that , for all @xmath29 , @xmath30 again , more regularity on @xmath1 implies more good properties@xcite .
the wigner measure is always absolutely continuous with respect to the lebesgue measure @xmath31 , so @xmath32 , where @xmath33 is a @xmath34 function .
however , if for instance @xmath9 is bounded , one can show by the ascoli - arzel theorem that in this case , given @xmath35 , @xmath36 \ni t \longmapsto \mu_t \in \mathcal{d}'(\r^{2d})$ ] is continuous and that there exists another subsequence @xmath37 such that , for each @xmath38 $ ] and @xmath39 , @xmath40 furthermore , for such a regular @xmath1 , the semiclassical measures satisfy a liouville equation @xmath41 or equivalently @xmath42 where @xmath43 is the correspondent semiclassical limit of @xmath44 .
this last equation is interpreted as a transport phenomenon along the classical flow @xmath7 , which can be easily seen by picking up the test functions @xmath39 and verifying that @xmath45 due to and .
moreover , the test functions @xmath39 are called _ classical observables _ and are related with the _ quantum observables _ @xmath46 by the weyl quantization formula , @xmath47 which provides self - adjoint operators for real - valued symbols@xcite .
these objects , called _ semiclassical pseudodifferential operators _ , relate to the wigner transform by means of the formula @xmath48 the quadratic form @xmath49 gives in quantum mechanics the average value of the observable @xmath50 for a system in the quantum state @xmath51 , exactly as does the integral @xmath52 in classical statistical mechanics for the observable @xmath53 in a system with mass probability density @xmath54 over the phase space .
besides , last equation carries @xmath55 which allows us to understand the wigner measures as mass distributions , linking the quantum evolution of @xmath56 given by the schrdinger equation with that of its semiclassical limits that gives a dirac mass on a certain point of the phase space , corresponding to a singular particle of mass @xmath57 ( classically localized on that point ) whose quantum evolution is described by .
although it is always possible to find a sequence of quantum states concentrating to a dirac mass@xcite , this is not always the case , since one may have @xmath24 constinuously spread over the phase space . in any case , the quantum - classical correspondence is better understood statistically ( @xcite , @xcite ) , in which frame the case of the dirac mass ( or a sum of puntucal dirac deltas with total mass @xmath57 ) should be seen as a special case of statistical distribution . ] .
it happens that a very large class of relevant problems do not present potentials with all such regularity .
for instance , _ conical potentials _ , which are of the form @xmath58 with @xmath59 not growing faster than some non - decreasing positive polynomial @xmath60 ( which also bounds @xmath61 ) , @xmath62 , @xmath63 , such that @xmath64 is full rank , @xmath65 , and with @xmath1 and @xmath66 satisfying the kato - rellich conditions . as we shall see
, they raise interesting mathematical questions .
similar problems have been being treated in works like @xcite , @xcite and @xcite in a probabilistic way . in other works
authors have been analysing the deterministic behaviour of the wigner measures under the conical potentials defined above , more noticeably in @xcite , where they found a non - homogeneous version of whose inhomogeneity is an unknown measure supported on @xmath67 a set onto which it is not generally possible to extend the classical flow in a unique manner , although it is possible everywhere else .
[ rem : cotangent ] the set @xmath68 corresponds exactly to the tangent bundle to @xmath69 , since any curve @xmath70 over @xmath69 ( _ i.e. _ , such that @xmath71 ) passing on @xmath72 at @xmath73 must satisfy @xmath74 . in our work , however , we stay within a structure of phase space , so it will be natural to identify @xmath68 as the cotangent bundle @xmath75 .
this suggests an intriguing possibility involving irregular potentials : what happens to the wigner measures in a system where the potential allows a complete quantum treatment , but causes the classical flow to be ill defined ?
is there some selection principle from the quantum - classical correspondence that could provide information enough for describing the transport of the measure where the classical flow fails ? to fix some ideas , forget for a moment about the measures and think of a classical particle submitted to conical potentials like @xmath76 . naturally , the trajectories are well defined everywhere away from the line @xmath77 , and it turns out that they can be continuously extended onto @xmath78 in a unique manner provided that @xmath79 , as in figure [ fig : flow_classical ] . 2 near the origin .
the arrows indicate their orientation.,title="fig:",scaledwidth=40.0% ] + near the origin .
the arrows indicate their orientation.,title="fig:",scaledwidth=26.0% ] + in the case @xmath80 , the flow can be uniquely defined even over @xmath81 by setting it constant at the origin , as shown in figure [ fig : flow_classical](a ) ; more generaly , in higher dimensional cases , like for @xmath82 , @xmath83 with @xmath84 and @xmath85 , there is still room for the particle to move inside the singular set @xmath86 and it must have some non ambiguous behviour therein , induced by the unique quantum evolution of @xmath87 and their well defined concentration to @xmath88 . however , the dynamics in @xmath68 is classically unknown , for @xmath89 makes no sense for @xmath90 , so we can not rely on the sole hamiltonian trajectories to characterize the transport phenomenon that the semiclassical measure undergoes over the singularities . furthermore , there are other kinds of difficulties . in the case
@xmath91 , even in dimension @xmath57 there is no unique extension for the flow all over the phase space .
as we can see in figure [ fig : flow_classical](b ) , when coming from the right - hand side below , there are different alternative trajectories after reaching @xmath77 with zero momentum : going back to the right upwards , crossing to the left downwards , staying at @xmath92 , or staying there for a moment and then resume moving to one or to the other side .
let us treat this problem in three different steps . in @xcite
, the authors proved that the hamiltonian flow can always be continuously extended in a unique manner to @xmath93 and that , whenever the wigner measure does not charge the singulities in the phase space , _ i.e. _ , while @xmath94 , then @xmath24 follows these unique continuous extensions .
this result is grounded on the facts that @xmath88 does not charge the set @xmath93 for more than a negligible time , more precisely that @xmath95 almost everywhere in @xmath96 with respect to @xmath31 ( for a matter of completeness , we re - obtain this result in lemma [ lem : m_is_zero_and_etc ] ) , and that the measures obey to the standard liouville equation with @xmath1 away from @xmath69 , where the potential is regular .
in this paper we will obtain in section [ sec : establishing_liouville ] a complete description of the dynamics to which the semiclassical measures ought to obey , including near and inside the singularities , by driving an approach similar to that of @xcite , which makes an extensive use of _ symbolic calculus _ ( section [ sec : symbolic_calculus ] ) and _ two - microlocal measures _ ( section [ sec : two_microlocal_analysis ] ) : [ th : main_result_1 ] let @xmath56 be the solution to the system with a conical potential of the form , and denote @xmath97 .
then the correspondent wigner measures @xmath24 obey to the @xmath98 equation : @xmath99 where @xmath100 is the indicatrix of @xmath101 ( and @xmath102 of it complementar inside @xmath103 ) and for each @xmath104 , @xmath105 is the orthogonal projector over @xmath106 inside @xmath2 .
furthermore , decomposing @xmath103 in a neighbourhood of @xmath69 as the bundle @xmath107 with fibers @xmath108 ( and elements @xmath109 ) , there exists a measure @xmath110 over @xmath111 , where @xmath112 , satisfying the asymmetry condition @xmath113 and such that @xmath114 [ rem : remark_in_proof ] observe that equation is the sum of two liouville terms , one for the potential @xmath1 outside @xmath68 and another for @xmath115 over it .
usual transport under @xmath1 is assured away from @xmath68 by @xcite , thus , if no trajectories outside @xmath68 lead to or from it , these both terms are shown to cancel on their own against functions in @xmath116 and consider the local conservation of mass . ]
and the equation decouples into two independent transport phenomena , one inside and the other outside @xmath68 , the regularity of the insider flow being guaranteed by the picard - lindelf theorem , as @xmath117 is smooth in the topological space @xmath68 .
[ rem : static_equilibrium ] the second part of the theorem , the asymmetry formula , will be discussed ahead in section [ sec : second_question ] and turns out to indicate that any mass that stay on the singularity will be in static equilibrium over it .
remark [ rem : remark_in_proof ] immediately gives : [ th : main_result_2 ] in the same conditions of theorem [ th : main_result_1 ] , suppose that no hamiltonian trajectories lead into @xmath68 .
call @xmath7 the hamiltonian flow defined by the trajectories induced by @xmath1 for @xmath118 and by @xmath119 for @xmath120 . then , writing @xmath121 , one has @xmath122 for all @xmath11 , where @xmath43 is the wigner measure of the family @xmath123 .
more precisely , suppose that the hamiltonian flow @xmath7 can be extended in a unique way everywhere in a region @xmath124 .
it is sufficient in theorem [ th : main_result_2 ] that we choose a time interval @xmath125 such that @xmath126 to assure the transport for any @xmath127 ; this recovers the result in @xcite for a particular choice where @xmath128 . in words , it is sufficient that our measure transported by the flow does not hit any point where the trajectories split , the fact that it may charge the singularities being irrelevant .
now , what happens if some trajectories hit @xmath68 ?
first , realize that in this case there is never uniqueness , since there are necessarily the outgoing trajectory ( which is the reverse of the incoming one ) and the one whose projection on @xmath75 evolves freely and , more importantly , whose projection on @xmath129 remains static , what is always kinectically admissible ( with @xmath130 , the velocity normal to @xmath69 is @xmath27 , for @xmath131 is actually @xmath132 , as informs theorem [ th : main_result_1 ] ) .
the equation in theorem [ th : main_result_1 ] says that either a transfer of mass between these two regimes , the insider and the outsider , and the continuation of the exterior transport are possible to happen , but gives no more information .
the second part of the theorem , however , may solve the question .
it has got a rich geometric interpretation , saying that the mass distribution on @xmath68 has some asymmetry around @xmath69 due to the `` shape '' @xmath133 of the conical singularity , and that it is deformed by an exterior normal force @xmath134 , in such a manner that portion of mass that remains over the singularity wil be in static equilibrium .
indeed , the measure @xmath110 gives the mass distribution in a sphere bundle with fibers @xmath135 around the singularity , _
i.e. _ , the directions @xmath136 in the exterior space from where the solutions @xmath56 concentrate more or less intensively to the singular points . if @xmath137 , then of course the measure does not stay at the singularity , so it necessarily continues under the exterior regime ( regardless of whether it is unambiguously defined or not ) and we got all the information @xmath110 may give ; let us then suppose @xmath138 .
loosely , let us also consider @xmath110 as a function of @xmath139 ( as if it was absolutely continuous with respect to @xmath140 ) and let us say that @xmath141 gives the total mass @xmath142 on the point @xmath139 ( though it actually gives a mass density over the phase space ) at the instant @xmath6 , supposed not to be @xmath27 .
naturally , @xmath143 gives the average vector of concentration to this point , whose normalization by @xmath142 we will call @xmath144 .
realize that the speed the mass may have tangentially to the singular space @xmath69 , that we call @xmath145 , plays no role in dictating how the quantum concentration will happen thereon ; with simple hypotheses on the family @xmath123 ( like @xmath146-oscillation , see @xcite ) , one has @xmath147 , and the same for @xmath110 since @xmath148 , so we could be working directly with @xmath149 .
well , the derivative of a potential is a force , so let us call @xmath150 the force normal to the singular manifold @xmath69 at the point @xmath151 .
consequently , the asymmetry formula in theorem [ th : main_result_1 ] is imposing a simple condition on the mass concentration : @xmath152 which is to say that , in average , the mass should concentrate alongside the exterior normal force @xmath153 .
how strong the concentration will be there with respect to the other points , or whether it is going to be attractive or repulsive , will depend on the shape of the conical singularity at the different points of @xmath69 , described ( so as to say ) by @xmath133 .
an example is illustrated in figure [ fig : asymetry ] below .
encircled by its normal bundle in sphere @xmath154 ( with fibres @xmath155 ) . the mass concentration given by @xmath110
is represented by the more or less strongly shadowed regions , and is directed by the normal force @xmath153 which is spinning around @xmath69 in the example . here
, we took @xmath156 and @xmath157.,scaledwidth=90.0% ] the reason why we have said that equation is a condition of equilibrium is that the expression inside the integral therein is similar to what would be the total force normal to @xmath69 , _
i.e. _ @xmath158 , when calculated on @xmath69 ( where @xmath159 ) and making sense of @xmath136 as some limit of @xmath160 ( a vector in @xmath161 ) when @xmath72 approaches the singularity from a particular direction .
so we are also tempted to interpret as saying that the total force normal to @xmath69 on some mass staying on the singularity , to be given by the integral , is @xmath27 : this mass is in static equilibrium . besides , in formula , we have @xmath162 , since @xmath144 is an average of norm @xmath57 vectors in @xmath161 .
consequently , we must have @xmath163 not to be led to an absurd .
if this is not the case , then we must have @xmath137 in order to satisfy the asymmetry condition trivially , which means that the wigner measure will not stay on the singularity .
this reasoning will be made rigorous in section [ sec : asymmetry_condition_in_use ] , where we will prove : [ th : main_result_4 ] if for some @xmath164 one has @xmath165 , then there exists a neighbourhood @xmath166 of @xmath151 such that @xmath137 over @xmath167 , where @xmath168 .
once we know some cases where the mass is forbidden to stay over the singularity , it is worth studying deeperly the ways it can get in and out @xmath68 : [ th : accessory_result_1 ] supposing @xmath169 , for any trajectory @xmath4 leading out from or into @xmath68 in @xmath164 at @xmath170 , set the @xmath171 vector @xmath172 ; then , for any sequence @xmath173 converging to @xmath170 , @xmath174 has points of adherence @xmath175 , which are non - zero roots of @xmath176 conversely , for any @xmath177 and @xmath178 satisfying , there exists a unique continuous extension of the classical flow which passes by @xmath151 at @xmath179 without staying on @xmath151 and whose correspondent limits @xmath180 exist and are equal to @xmath175 .
if has no non - zero roots , then no trajectory leads in or out @xmath68 in @xmath151 .
if @xmath181 , then @xmath182 may have @xmath27 as a point of adherence , but even in this case @xmath183 will have non - zero adherence points that we denote @xmath184 .
we will abusively call @xmath185 `` zero roots '' of and say that trajectories reach @xmath68 in @xmath151 following directions @xmath175 if @xmath186 both exist and are equal to @xmath187 .
sometimes it may be that an incoming trajectory only approaches this limit asymptotically at @xmath188 .
[ th : accessory_result_2 ] if @xmath181 , then at least one of the following affirmations holds : * equation has non - zero roots @xmath175 and there are unique trajectories leaving and arriving on @xmath68 in @xmath151 through the directions @xmath175 ; * equation has `` zero roots '' , in the sense that there are @xmath189 such that @xmath190 and either there is no trajectories reaching @xmath68 in @xmath151 through @xmath185 , or they exist but do not arrive onto @xmath68 within any finite time ; * the classical flow does not touch @xmath68 in @xmath151 . in any case , if equation has no roots ( `` zero '' or non - zero ) , then no classical trajectory passes by @xmath68 in @xmath164 . in @xcite , we will endeavour a more precise study of the link between @xmath110 and the classical flow , generalizating the link between theorem [ th : main_result_4 ] and theorems [ th : accessory_result_1 ] and [ th : accessory_result_2 ] . in section
[ sec : asymmetry_condition_in_use ] we will work out the proof of theorems [ th : main_result_4 ] , [ th : accessory_result_1 ] and [ th : accessory_result_2 ] in coordinates that are more suitable to understand @xmath191 as an approaching direction .
besides , we will see in section [ sec : examples ] , by means of a number of examples , that these results allow a full classification of the types of trajectories that may reach or stay on the singularity . in some cases , like for @xmath192 or @xmath193 in @xmath2 , they allow a full resolution of the problem when the inequality in theorem [ th : main_result_4 ] holds , since by solving explicitly one can verify that there is only a unique trajectory leading in and out the singularity without staying thereon , and necessarily the measures will follow it and not charge @xmath68 . in short , so far we have seen that whenever we have a well defined flow , we know what the semiclassical measures do : they are transported thereby . if the flow presents trajectory splits , they necessarily happen on @xmath68 , where there is always the possibility of regime change between outsider and insider flows .
then , thanks to the measure @xmath110 , we may be able to obtain enough information to decide whether the measures stay or not on the singularity , and in case they stay , we know that they will be carried by the flow generated by @xmath115
. finally , a last problem is : if a measure does not stay on @xmath68 and continues in the exterior flow , but even though there are different trajectories to take , can we derive from the well posed quantum evolution some general criterion for choosing the actual trajectories that the measure will follow ? is there any selection principle for the classical movement of a particle under such conical potentials ? as we will see in section [ sec : splitting ] , the answer is negative .
the path a wigner measure ( or a particle ) takes after its trajectory splits depends crucially on its quantum state concentration , so any selection principle making appeal only to purely classical or semiclassical information is to be dismissed .
this can be justified by : [ th : main_result_3 ] let be @xmath194 in @xmath96 and @xmath195 and @xmath196 the wigner measures associated to the solutions of with initial data @xmath197 with @xmath198 , @xmath199 and @xmath200 supported on @xmath201 .
then , for @xmath202 , @xmath203 nevertheless , for @xmath204 , @xmath205 whereas @xmath206 in pictures , the particle @xmath195 follows the path in figure [ fig : paths](a ) , and the particle @xmath196 moves as in figure [ fig : paths](b ) . 2 , but then diverging for @xmath204 , which dismisses any selection principle based only on classical information about the problem.,title="fig:",scaledwidth=30.0% ] + , but then diverging for @xmath204 , which dismisses any selection principle based only on classical information about the problem.,title="fig:",scaledwidth=30.0% ] + this result will be obtained with the help of approximative solutions of called _ wave packets _ , which are @xmath207 functions generally of the form @xmath208}\ ] ] ( @xmath209 is the classical action ) , to be properly introduced in section [ sec : the_wave_packets ] .
the standard methods using wave packet presented in that section , however , only apply for smooth flows , and in both cases the trajectories in figure [ fig : paths ] have problems over the axis @xmath210 , not to speak about the lack of regularity of @xmath1 . in the case of the returning particle
, the problem will be solved by decomposing the initial data into two pieces , for @xmath201 and @xmath211 , and treating each one with a standart wave packet set to follow a different parabola .
we will see in proposition [ prop : psi_follows_the_path_1 ] that the wigner measure initially set on the singularity will break out into two pieces @xmath212 and @xmath213 with weights given by the total mass of the initial data over @xmath214 and @xmath215 respectively , @xmath216 and @xmath217 , each piece gliding to a different side as in figure [ fig : rebounding_particle ] . yet
, this does not give a full example of non - unicty as in theorem [ th : main_result_3 ] , since , if we evolve the pieces @xmath212 and @xmath213 to the past , we realize that they do not come from the same side , and this fact could indicate some kind of selection principle .
( right ) and @xmath213 ( left ) .
the full line indicates the path for @xmath218 , whereas the dashed line indicates the past trajectories that the measures ought to have followed in @xmath219 to reach the singularity.,scaledwidth=30.0% ] constructing a quantum solution whose semiclassical measure behaves like in figure [ fig : paths](b ) will be more difficult and will require us to consider a family of wave packets following different trajectories with smaller and smaller initial momenta @xmath220 , that in some sense converge to the aimed path with @xmath221 , as illustrated in figure [ fig : crossing_particle ] .
we will then study the concentration of the wave packets with @xmath146 going to @xmath27 at the same time as the trajectories concentrate , by making @xmath220 go to @xmath27 with a suitable power of @xmath146 . for @xmath222 , approaching the aimed one with @xmath221 .
[ fig : crossing_particle ] in section [ sec : preliminaries ] , we will introduce the fondamentals of our analysis : wave packet approximations ( sec .
[ sec : the_wave_packets ] ) , symbolic calculus ( sec .
[ sec : symbolic_calculus ] ) and two - microlocal measures ( sec .
[ sec : two_microlocal_analysis ] ) . in section
[ sec : splitting ] , we will construct the solutions of the schrdinger equation that lead to theorem [ th : main_result_3 ] ; the case keeping on the same parabola is treated in sec .
[ sec : same_parabola_1 ] , the other one in sec .
[ sec : other_parabola ] .
finally , in section [ sec : establishing_liouville ] we will prove theorem [ th : main_result_1 ] , firstly in a particular version for subspaces , what will be done step by step from sec . [
sec : kinetic_term ] to [ sec : absolute_continuity ] ( the part where we effectly establish the dynamical equation and the asymmetry condition being section [ sec : establishing_the_equation ] ) . then this version will be immediately extended to the general case thanks to the coordinate change that we will have set in section [ sec : reducing ] . in section [ sec : asymmetry_condition_in_use ]
we will use the asymmetry condition to prove theorem [ th : main_result_4 ] .
theorems [ th : accessory_result_1 ] and [ th : accessory_result_2 ] are also proven in this section , and in sec .
[ sec : examples ] we conclude by showing with examples [ ex:1 ] to [ ex : last ] how the results in this article allow a full classification of behaviours the wigner measures present and , sometimes , give a full description of the transport phenomenon .
in this section , we will present the basics of the main tools that we use in this work .
first the wave packet method for approximating solutions of the schrdinger equation ( see for example @xcite and @xcite , or @xcite for a generalized notion of wave packet ) that we will adapt later in section [ sec : splitting ] , then some simple results in standard symbolic calculus ( @xcite , @xcite ) which will provide a guideline for proving theorem [ th : main_result_1 ] , and last some notions about the two - microlocal measures ( @xcite , @xcite , @xcite ) , that we will deploy in order to accomplish the necessary refined analysis for obtaining either the dynamical equation for the wigner measures and the asymmetry condition on the mass concentration around the singular manifold .
for the sake of simplicity , we will use these measures in a specialized version for @xmath223-codimensional subspaces of @xmath2 .
this does not carry any loss of generality , since in section [ sec : reducing ] we will be able to perform a change of coordinates in the problem that will lead us to analyse such a subspace . the geometric nature of these measures , as well as that of the semiclassical ones and of liouville like equations , makes it possible to move between different coordinate systems respecting the structures of cotangents , normal and conormal bundles over the singular set . for a @xmath224 potential @xmath1 and one of its hamiltonian trajectories @xmath225
, we define the wave packet with initial profile @xmath226 following @xmath225 as @xmath227},\ ] ] where @xmath209 is the classical action @xmath228 and @xmath229 satisfies the @xmath146-independent differential system @xmath230 [ lem : path_wave_packet ] any semiclassical measure associated to the family @xmath231 is @xmath232 a straightforward calculation .
another virtue of the wave packets is that they provide approximative solutions to the schrdinger equation with convenient initial data , as stated in : [ prop : evolution_norm ] let be @xmath2330,\tau\right [ \subset \r$ ] and @xmath234 \ , 0 , \tau \right [ \right\rbrace$ ] . if @xmath9 exists and is integrable in @xmath235 , and if @xmath56 is the solution of the schrdinger equation with potential @xmath1 and initial data @xmath236 , then letting be @xmath237 the wave packet initially centred in @xmath5 with profile @xmath226 , we have @xmath238 0,\tau \right [ } \ , \left\| \frac{1}{\eps } r_{s}^{\eps } v_{s } \right\|_{l^2(\r^d ) } ds,\ ] ] where @xmath239 after a direct calculation , one obtains the following differential system for @xmath237 : @xmath240 where @xmath241 is explicitly given by equation .
now , we compare @xmath242 and @xmath243 by evaluating @xmath244 which gives , in view of the equations for @xmath56 , @xmath237 and the self - adjointness of @xmath245 , @xmath246 thus , for any @xmath247 0 , \tau\right[$ ] , we have @xmath248 naturally , the function @xmath249 is continuous and , at @xmath73 , @xmath250 .
hence , by choosing sequences @xmath251 and @xmath252 in @xmath2330,\tau\right[$ ] such that @xmath253 and @xmath254 , the proposition follows .
call @xmath24 the semiclassical measure linked to the exact family of solutions @xmath255 with initial data as in the theorem above .
if @xmath256 exists and is integrable , then @xmath257 and , consequently , given @xmath35 , for any @xmath38 $ ] , @xmath258 @xmath1 being at least of class @xmath259 , one can verify from a taylor formula that @xmath260 and , moreover , that @xmath241 introduces in the schrdinger equation a @xmath14 error of order @xmath261 .
thus , from proposition [ prop : evolution_norm ] , it is clear that @xmath262 in @xmath14 ; for any @xmath263 $ ] , this gives , when @xmath264 , that the wigner measure of @xmath56 shall coincide of that of the wave packets .
the conclusion comes from lemma [ lem : path_wave_packet ] .
actually , the approximation in the corollary remains good for @xmath6 smaller than the ehrenfest time @xmath265 , as @xmath266 when @xmath23 ; more details in @xcite .
estimates beyond the ehrenfest time are given in @xcite .
observe that even if @xmath1 is not as regular as we required , we can still write @xmath241 as in for any @xmath6 such that @xmath267 and @xmath268 make sense , although in this case it is not clear which is the order of the approximation the wave packet furnishes , nor even whether it is negligible in the semiclassical limit . finally , observe that it is also possible to write the actual solution @xmath242 with initial state @xmath269 under the wave packet form : one defines a @xmath270 such that @xmath271 } , \ ] ] which consequently obeys to @xmath272 which is nothing else than the exact schrdinger equation written in a different form .
let us consider the @xmath146-pseudodifferential operators @xmath273 of symbols @xmath274 given in formula . of central importance
is the fact that they are uniformly bounded in @xmath275 with respect to @xmath146 ( @xcite , @xcite ) : there exist constants @xmath276 such that @xmath277 or else @xmath278 [ rem : schur_upper_bound ] inequalities and give upper bounds for the schur estimate of the norm of @xmath50 .
besides , these formul remain valid when one uses different scales of @xmath146 in different variables , like for an operator @xmath279 , with @xmath280 and @xmath281 . nonetheless , formula can be used for more general symbols , although we may loose boundedness , good properties for symbolic calculation , and be forced to restrict their domains .
in particular , for @xmath1 satisfying the kato - rellich conditions , @xmath282 doted with domain @xmath20 , is unbounded , although it is still self - adjoint .
thus , taking a test function @xmath283 and @xmath242 solution of , the semiclassical measure @xmath24 linked to the family @xmath284 may be given as @xmath285 ( here @xmath286 stands for _ semiclassical limit _ ,
abstracting which particular subsequence @xmath287 is to be taken ) . from this expression
we can evaluate the distribution @xmath288 : @xmath289 moreover , in view of the schrdinger equation , @xmath290 \psi_{t}^{\eps } \ , , \psi_{t}^{\eps } \right>. \label{eq : d_op}\ ] ] by standart symbolic calculus ( many of whose formul may be found in @xcite , for instance ) , in the smooth case we would get @xmath291 = \op_\eps \left ( ( \xi \centerdot \partial_x - \nabla v(x ) \centerdot \partial_\xi ) a \right ) + \mathcal{o}(\eps),\ ] ] which ultimately induces equation for @xmath24 in the sense of the distributions .
the conical singularities that @xmath1 presents , however , will require a specific treatment .
roughly , we will have to re - derive `` by hand '' adapted formul for a correct symbolic calculus with such potentials , which we will do progressively in sections [ sec : the_inner_part ] , [ sec : the_outer_part ] and [ sec : the_middle_part ] .
now , let us define a new symbol class @xmath292 composed by symbols @xmath293 such that * for each @xmath294 , @xmath295 is compactly supported on @xmath296 .
* there exists some @xmath297 and a function @xmath298 such that , for @xmath299 , one has @xmath300 .
these symbols will be quantized as @xmath301 observe that the right hand term above is just the banal quantization of a @xmath146-dependent @xmath103 function as in .
[ prop : two_microlocal_measures ] there exists a measure @xmath302 on @xmath303 and a trace class operator valued measure @xmath304 on @xmath305 , both positive , such that , for @xmath306 and @xmath283 , @xmath307 where @xmath308 is the weyl quantization of the symbol @xmath309 with @xmath310 and @xmath24 is the usual wigner measure related to @xmath56 .
furthermore , for a function @xmath311 taking values on the set of compact operators on @xmath312 , one has @xmath313 where @xmath314 is the trace class operator with kernel @xmath315 finally , the terms in are obtained respectively from those in the decomposition @xmath316 in the limit when @xmath23 , then @xmath317 , and last @xmath318 , where @xmath319 is a cut - off function such that @xmath320 , @xmath321 for @xmath322 and @xmath323 for @xmath324 .
[ rem : p_equals_d ] if @xmath325 , then @xmath326 is just the adjoint of the projector over @xmath327 , with kernel @xmath328 .
it follows that @xmath329 then , because @xmath330 is compact , in the limit @xmath23 one has that it is simply @xmath331 , where @xmath332 is some weak limit of @xmath333 .
a very general treatment of this result can be found in @xcite .
the introduction of @xmath334 is a trivial addition of ours in order to enlighten the calculations to come .
[ rem : definition_m ] observe that @xmath304 induces a measure @xmath335 on @xmath336 by means of the formula @xmath337 above , @xmath53 has no need to be in @xmath292 ; it is sufficient that it be compact supported in all variables .
[ lem : m_0_gives_m_0 ] @xmath338 if and only if @xmath339 .
that @xmath340 implies @xmath339 , it is obvious . for the converse ,
it is necessary to show that @xmath339 implies @xmath341 for all compact supported functions @xmath342 taking values in the set of compact operators , since this is the set whose dual are the trace class operators .
observe that it is sufficient to consider @xmath343 hilbert - schmidt , given that these operators are dense in the set of the compact ones .
so , we can consider that it has a kernel @xmath344 and , defining @xmath345 with some @xmath346 , it follows that @xmath347 and we are done .
[ lem : m_absolutely_continuous ] the measure @xmath335 is absolutely continuous with respect to @xmath348 . indeed , if @xmath349 unless for a null measure set with respect to @xmath350 , then @xmath351 and the result follows from the definition in remark [ rem : definition_m ] .
[ lem : two_microlocal_decomposition ] the semiclassical measure @xmath24 decomposes as @xmath352 where @xmath302 is as in proposition [ prop : two_microlocal_measures ] and @xmath353 . if @xmath354 , define @xmath355 for all @xmath294 .
thus , @xmath356 and @xmath357 , and we can use proposition [ prop : two_microlocal_measures ] . with the two - microlocal measures , we are equipped to tackle the analysis of the singular term of the commutator in .
in the introduction we pointed out that the non - uniqueness of the classical flow for the present case only plays a relevant role when the initial data concentrate to a point belonging to a trajectory that leads to the singularity .
the behaviour of the measure will depend on the concentration rate and oscillations of the quantum states @xmath56 as well as on other characteristics of this family over the crossings , such as the region where these states are supported .
below , we will prove results that altogether are slightly more general than theorem [ th : main_result_3 ] .
we will present concrete cases of solutions to the schrdinger equation with the conical potential @xmath194 , @xmath358 , that concentrate to a branch of one of the parabol leading to the singularity in figure [ fig : flow_classical](b ) , and thereafter either swap to the other parabola ( section [ sec : other_parabola ] , proposition [ prop : crossing_psi ] ) or keep on the same one ( sections [ sec : same_parabola_1 ] , proposition [ prop : psi_follows_the_path_1 ] ) .
these examples refute any possibility of a classical selection principle allowing one to predict the evolution of a particle ( _ i.e. _ , a wigner measure concentrated to a single point ) after the singularity , since they show two particles subjected to the same potential and following the same path for any @xmath219 , but then going each to a different side for @xmath204 .
let us consider the trajectories @xmath359 in this section we will prove : [ prop : psi_follows_the_path_1 ] let be @xmath56 the solution to the schrdinger equation with @xmath194 in @xmath96 with initial datum @xmath360 with @xmath361 . the semiclassical measure associated to the family @xmath362
is given by @xmath363 where the weights @xmath364 are given by @xmath365 given an arbitray @xmath366 , let us cut the the evolved state @xmath333 in three parts , @xmath367 where @xmath368 , @xmath369 and @xmath370 solve the schrdinger equation with initial data @xmath371 chosen in such a way that @xmath372 , all these three functions smooth and taking values in @xmath373 $ ] .
the middle term s semiclassical measure has total mass of order @xmath374 ; as a consequence , the full wigner measure of @xmath56 will be , for any @xmath366 : @xmath375 where @xmath376 are the measures associated to @xmath377 , @xmath378 is a measure with total mass of order @xmath379 issued from the middle term and its interferences with the other terms , and @xmath380 is the interference measure between @xmath381 and @xmath382 , which satisfies , for any test function @xmath383 , the estimate @xmath384 as widely known in semiclassical calculus . for the study of @xmath376 ,
let us introduce @xmath385 , the wave packets defined in for @xmath11 , having profiles @xmath386 that obey to @xmath387 which is nothing more than the profile equation _ with the smooth potentials _
@xmath388 ( which admit the trajectories ) .
[ lem : psi_follows_varphi_other ] for any @xmath366 , @xmath389,l^2(\r)\right ) } = 0.\ ] ] ( the proof is postponed . )
so , the wigner measures for the components @xmath390 are the same as those for @xmath391 , which one computes explicitly : @xmath392 observe that equation implies that @xmath393 is supported on the intersection of the supports of @xmath394 and @xmath395 , but from the last formula the support of @xmath110 turns out to be different from @xmath396 only for @xmath73 .
even for @xmath73 we can consider @xmath397 , since this measure must be absolutely continuous with respect to @xmath376 , and these are absolutely continuous with respect to the lebesgue measure @xmath31 .
finally , as @xmath379 is arbitrary , we take the limit @xmath318 and it follows from that @xmath398 , where : @xmath399 as we had in the proposition s statement . to begin with
, since for @xmath400 we have @xmath401 for any @xmath402 , @xmath403 obeys to with error @xmath404 defined according to ; let us calculate it for @xmath400 : @xmath405 which gives @xmath406 additionally , one can solve equation for the profile @xmath386 of @xmath407 explicitly ; writing down its solution , @xmath408 it remains clear that @xmath386 admits a finite development like @xmath409 where the first term in the right hand side has support on @xmath410 .
[ rem : boundedness ] from and the fact that its initial datum is @xmath411 , it follows that , for any @xmath35 and @xmath412 , one has @xmath413,l^2(\r)\right)$ ] , which natually implies that @xmath414,l^2(\r)\right)$ ] . therefore , from expressions and : @xmath415 which results in @xmath416 } \left\| \psi_{\pm , t}^{\eps,\delta } - \varphi_{\pm , t}^{\eps,\delta } \right\|_{l^2(\r ) } \leqslant \pm \frac{2}{\sqrt{\eps } } \int_{0}^{\pm t } s^{2 } \left\| y \ , \tilde{v}_{s}^{\delta,\pm}\left(y\right ) 1\!\!1_{\pm y \leqslant -\frac{s^{2}}{2\sqrt{\eps } } } \right\|_{l^2(\r ) } ds\ ] ] after applying proposition [ prop : evolution_norm ]
. to evaluate this integral , fix @xmath417 and denote @xmath418 : 1 . for @xmath419
, gives @xmath420 with @xmath421 constant .
the fact that @xmath422 is bounded comes from remark [ rem : boundedness ] .
2 . for the estimate for @xmath423 \cup
\left [ \tau_{\eps } , t \right]$ ] , remark that @xmath424 implies @xmath425 , then : @xmath426 boundedness for the norm in last equation s right hand side comes from previous remark [ rem : boundedness ] .
+ finally , form , @xmath427 for a constant @xmath428 .
because of our choice of @xmath417 , both estimates and go to @xmath27 with @xmath23 , and we prove the proposition .
so , in this section we saw the example of a case where the initial measure splits in two pieces , each one gliding to its side as in figure [ fig : rebounding_particle ] according to the quantum distribution of mass along the @xmath72-axis .
remark that there is no crossings at all .
the part @xmath213 that goes to the left downwards did not come from the same side as @xmath212 , but from the very left .
the portion that was in the right for @xmath219 stays at the right hand side . now , consider for @xmath429 the hamiltonian trajectories of @xmath194 : @xmath430 in this section , we will prove : [ prop : crossing_psi ] if @xmath431 is solution to the schrdinger equation with @xmath194 in @xmath96 , @xmath432 and initial data @xmath433 with @xmath434 , then the associated semiclassical measures are @xmath435 besides , if we take @xmath436 with @xmath198 , then the corresponding semiclassical measure will be @xmath437 for the case with @xmath432 constant , just apply lemmata [ lem : psi_follows_varphi ] and [ lem : psi_follows_the_path ] ahead chosing @xmath438 . for the case @xmath436 ,
same thing , but of course taking @xmath198 .
before we proceed to the lemmata , let us define @xmath439 after using the trajectories in , so as @xmath439 satisfies the following system : @xmath440 and we have : above , for @xmath400 we have @xmath441 and @xmath442 write down @xmath443 and observe that @xmath444 for the trajectory in .
now , define @xmath445 and , given some small @xmath446 , consider the following wave packet profile equation linked to the trajectory : @xmath447 \cup \left[\tau_\eps , t\right ] .
\right.\ ] ] we will call @xmath448 the wave packet defined as in for trajectory , having profile @xmath449 for @xmath450 $ ] and profile @xmath451 otherwise .
[ lem : psi_follows_varphi ] for @xmath452 with @xmath432 and @xmath453 , one has @xmath454 , l^2\left(\r\right)\right ) } = 0.\ ] ] let us treat the problem partitioning it in zones by choosing @xmath455 with @xmath456 : 1 .
@xmath450 $ ] : + denote depends on @xmath146 , we will drop down the dependencies on @xmath220 in order not to overcharge the notation .
we will also let the dependency of the trajectories on @xmath146 implicit until it be crucial to take it into account . ] @xmath457 ; then @xmath458 , @xmath459 and consequently @xmath460 given that @xmath461 satisfies @xmath462 one can estimate @xmath463 the last line coming from the fact that @xmath464 is constant ( @xmath465 and quantum normalization ) and that @xmath466 + besides , @xmath467 + taking into account the domain restrictions of @xmath468 ( see ) , for @xmath469 one must have satisfied the inequalities @xmath470 and @xmath471 in order not to have @xmath468 null , which means that , for fixed @xmath472 , @xmath6 is comprised in @xmath473 $ ] , where @xmath474 is one of the roots of @xmath475 .
further , remark that @xmath476 ; as a consequence , @xmath477 which implies @xmath478 + now , considering that @xmath479 and that we are compactly supported in @xmath472 ( @xmath480 still has factor @xmath53 and @xmath481 ) , it can be made the estimative @xmath482 in @xmath96 , which in any case gives @xmath483 for some constant @xmath484 , hence @xmath485 + finally , this results in a superior bound for @xmath486 in @xmath487 $ ] ; for a constant @xmath488 : @xmath489,l^2(\r)\right)}^2 \lesssim c_2 \frac{\tau_\eps}{|\eta_\eps| } + c_1 \frac{\tau_\eps^2}{\sqrt{\eps } |\eta_\eps|}\ ] ] as @xmath490 by assumption , the only additional constraint we need in order to the bound above be small when @xmath23 is @xmath491 2
. @xmath492 \cup \left[\tau_\eps , t\right]$ ] : + hereafter denote @xmath493 .
now @xmath494 obeys to the equation @xmath495 and , therefore , @xmath496 according to @xmath6 being positive or negative . + recalling
, one has @xmath468 non - zero only in the region @xmath497 , so @xmath498 ; this gives @xmath499 + [ lem : norms ] for @xmath500 , @xmath501 and @xmath502 , there exists @xmath503 constant such that @xmath504 .
+ ( the proof is postponed . ) + as a conclusion , for @xmath146 small enough we have @xmath505 which carries the new constraint : @xmath506 for some @xmath507 .
the proposition is proven once we remark that for any @xmath453 , one can find a positive integer @xmath508 such that both and will be satisfied for @xmath456 . to evaluate @xmath509 ,
observe that by recurrence one can show that , for @xmath510 ( formally considering @xmath511 ) , @xmath512 and , since @xmath513 satisfies the same equation as @xmath514 , @xmath515 this implies @xmath516 where @xmath517 and @xmath518 essentially , what we have got is @xmath519 @xmath520 and @xmath521 appropriate coefficients .
remark that we are being able to diminish the coefficient of @xmath522 by augmenting the order of the derivative @xmath513 .
roughly , by exhausting the powers @xmath522 in the developments above , we obtain @xmath523 with @xmath146-independent coefficients @xmath524 . moreover , from equation [ eq : v_dv_afterwards ] , @xmath525 is constant in time , and @xmath526 is compactly supported in @xmath472 , so we can simplify even more the last estimation : @xmath527 making use of and the initial condition , let us calculate the remaining quantities : @xmath528 where @xmath529 are complex coefficients . the way for calculating the expression above is the following : if conditions and are fulfilled , then we have @xmath530 , which causes @xmath531 to be always greater than @xmath532 . then , using , we get @xmath533 , and using , @xmath534 whenever @xmath535 and @xmath502 .
additionally , @xmath536 for @xmath537 , as we can verify explicitly .
it follows that , for @xmath151 such that @xmath538 , @xmath539 so @xmath540 will be dominated by a term of order @xmath541 and , finally , inequality will be bounded by a term of order @xmath542 , what we wanted to show .
this completes the proposition s proof .
[ lem : psi_follows_the_path ] with @xmath543 , @xmath432 , the semiclassical measure associated with the family @xmath544 is transported by a trajectory of shape : @xmath545 if @xmath198 , and @xmath546 if @xmath438 .
the fact that @xmath547 concentrates to a measure that follows the aimed path is not guaranteed by lemma [ lem : path_wave_packet ] since the initial data we inserted in the wave packet equation is not @xmath146-independent , as we required in section [ sec : the_wave_packets ] .
let us then calculate the concentration of @xmath547 indirectly . to begin with , if conditions and are fulfilled , then @xmath548 and consequently , from , and : @xmath549 thus , setting @xmath550 , one has @xmath551 where @xmath552 and @xmath553 and @xmath554 exist and are locally bounded almost everywhere in @xmath96 .
these facts and standard symbolic calculus allow a straighforward calculation showing that @xmath555 concentrates to the measure @xmath556 .
now , define @xmath557 the functions in @xmath558 that satisfy the systems @xmath559 it is possible to affirm that the semiclassical measures of @xmath560 will by carried by the flow @xmath561 , since by standard results ( see the introduction ) they should obey to the usual liouville equation with a null potential .
but because initially they are concentrated to the point @xmath562 in the phase space , we get @xmath563 for all @xmath564 .
well , for @xmath565 , @xmath566 , so observe that , for @xmath567 $ ] , @xmath568};\ ] ] consequently , by picking up a @xmath569 , one gets @xmath570 with the error coming from @xmath571 where @xmath572 is constant and we used lemma [ lem : norms ] . if @xmath438 , then @xmath573 is constant and we get , for @xmath574 \setminus \lbrace 0 \rbrace$ ] , @xmath575 which also holds for @xmath576 due to the initial condition for @xmath448 , implying that @xmath577 for all @xmath578 $ ] .
if @xmath198 , we can still have @xmath579 and now the error @xmath580 comes from the difference between calculating @xmath581 with the trajectories @xmath582 or the with `` limit '' path @xmath583 , which must be negligible in compact times for @xmath146 small enough , given that @xmath581 is smooth and the flow that defines the trajectories is stable in the region where we are . since it is possible to choose @xmath584 within conditions and , an argument similar to the previous one gives @xmath585 for all @xmath578 $ ] when @xmath198 .
[ rem : eta_positive ] all results in this section also work taking @xmath586 and swapping @xmath6 negative for positive and conversely in the definition of the trajectory . hence , we have found that it is possible that a particle arrive into the singularity from the up left or from the down right and that it continue to the other side down or up , as partially indicated in figure [ fig : paths](b ) .
moreover , we also proved that the wave packet approximation is valid for the non - smooth trajectories indicated in figure [ fig : crossing_particle ] ( and for the revese ones not indicated in the picture ) .
in view of the developments in section [ sec : symbolic_calculus ] , from equation we are left with the analysis of the commutator @xmath587 = \frac{i}{\eps}\left [ -\frac{\eps^{2}}{2}\delta \ , , \op_{\eps}\left(a\right ) \right ] + \frac{i}{\eps}\left [ \ , v_s \ , , \op_{\eps}\left(a\right ) \right ] + \frac{i}{\eps}\left [ \ , \|g(x)\|f(x ) \ , , \op_{\eps}\left(a\right ) \right].\ ] ] we will look separately into each term .
the first one is kinetic , the other two dynamical .
the first and the second are regular enough so one can use standard symbolic calculus ; this presents no difficulties and we will treat them explicitly in sections [ sec : kinetic_term ] and [ sec : dynamical_smooth ] only for a matter of completeness .
the third term is complicate because of the conical singularities it presents , which will require us to employ the two - microlocal analysis in section [ sec : two_microlocal_analysis ] .
this strategy was followed in @xcite , but here we will describe the two - microlocal measures in more details . prior to proceeding to this kind of analysis , however , we will need to restrict ourselves to the case where @xmath588 , with @xmath83 , @xmath84 and @xmath63 , in other words , to the case where the manifold @xmath69 formed by the singularities is actually a subspace @xmath312 .
it is in this context that we will be able to prove proposition [ prop : main_result ] , which is a particular version of theorem [ th : main_result_1 ] for @xmath589 .
reducing the general case to this one is the subject of next section . for a general conical potential ,
thanks to @xmath591 having maximal rank we can define a local change of coordinates @xmath592 in neighbourhoods of @xmath2 where @xmath593 for some function @xmath594 locally depending on @xmath595 in such a manner that @xmath596 has maximal rank and , if @xmath104 , then @xmath597 is orthogonal to @xmath131 may be contructed as follows : let be @xmath598 an open neighbouhood of @xmath69 ; choose @xmath599 a local diffeomorphism ; take @xmath600 a cylindrical neighbourhood of @xmath69 such that @xmath601 .
pick up an open @xmath602 , so @xmath603 is given in geodesic coordinates by @xmath604 for some @xmath605 with coordinates @xmath606 and @xmath607 .
define @xmath608 as @xmath609 .
it follows that @xmath596 is diffeomorphic over @xmath610 and null over @xmath161 , but since @xmath611 ( see remark [ rem : cotangent ] ) , @xmath612 and we are done . ] .
now , for the sake of clarity let us consider the coordinate change in tangent space induced by @xmath592 : @xmath613 \xi.\end{aligned}\ ] ] writing @xmath614 , we have the decomposition @xmath615 where @xmath616 and @xmath617 are suitable projectors inside @xmath2 over the kernels of @xmath591 and @xmath596 ; if @xmath104 , they are orthogonal . realize that @xmath618 and @xmath619 are invertible ( due to the maximality of their ranks ) ; let us denote their inverses simply by @xmath620 and @xmath621 .
thus one has : @xmath622 it follows that @xmath623 can be inverted in terms of @xmath620 and @xmath621 ; its inverse is @xmath624.\ ] ] analogously , @xmath625 and @xmath626 are invertible as soon as their counter - domains are restricted to @xmath627 and @xmath628 ( in which case the transpose of the relations in holds ) , allowing us to write the coordinate transformation in cotangent space as : @xmath629 \xi.\end{aligned}\ ] ] geometrically , let be the manifold @xmath97 , parametrized locally by the variable @xmath630 . for a @xmath104 ,
the cotangent space can be described by @xmath631 ( see remark [ rem : cotangent ] ) and its elements are parametrized by @xmath632 .
the variables @xmath633 and @xmath634 in @xmath312 will be associated with the normal and conormal spaces @xmath635 and @xmath636 .
we will also be using variables @xmath637 for @xmath638 , which will be identified as elements of the normal space in sphere of @xmath69 , @xmath639 .
[ rem : coordinate_global ] due to the fact that the coordinate @xmath633 is defined equally in any local charts by @xmath640 , one can define functions on @xmath312 , more specifically on @xmath641 , simply by its explicit formulation in @xmath633 ; further , @xmath69 not being empty allows us to always calculate a function at @xmath642 .
these facts are implicitly used in the calculations to come . at this point
we shall state a central result in semiclassical analysis ( see @xcite or @xcite ) : [ prop : change_variables ] let @xmath592 diffeomorphism on @xmath2 in the sense of manifolds , and @xmath643 the correspondent cotangent bundle transformation , @xmath644 .
let @xmath645 be the operator such that @xmath646 , where @xmath647 is the jacobian of @xmath592 .
then , @xmath648 and @xmath649 where @xmath650 is a seminorm for @xmath53 like or .
besides , denoting @xmath651 and @xmath652 , the local expression for @xmath653 satisfies the equation @xmath654 as a consequence , if @xmath655 is the measure associated to the family @xmath656 , for any symbol @xmath39 , @xmath657 one has @xmath658 _ i.e _ , any result got for the new measure @xmath655 can be immediately transferred to the original @xmath24 by simply changing the coordinate system ; analogous arguments work for the two - microlocal measures . noting @xmath659 the set of measures over the locally convex space @xmath1
, we get : [ cor : easy ] geometrically , the wigner and the two - microlocal measures to be built ahead will be @xmath660 , @xmath661 and @xmath662 , where the bundles @xmath107 and @xmath663 have fibers @xmath664 and @xmath665 therefore , theorem [ th : main_result_1 ] will follow directly from proposition [ prop : main_result ] and a trivial application of corollary [ cor : easy ] . since in the rest of this work the variables that we will write
are going to be dummy , we will not care about marking the differences between @xmath666 and @xmath667 , nor about keeping the notations @xmath668 , @xmath655 and @xmath669 in contrast to @xmath51 , @xmath24 and @xmath1 .
in short , now we can fairly relay on the study of the concentration of a family @xmath56 satisfying equation with a potential @xmath670 with @xmath84 , _
i.e. _ , satisfying the schrdinger with a modified hamiltonian operator @xmath671 so as that we become interested in the commutators @xmath672 = \frac{i}{\eps}\left [ -\frac{\eps^{2}}{2}t_\phi\delta t_\phi^ * \ , , \op_{\eps}\left(a\right ) \right ] + \frac{i}{\eps}\left [ \ , v_s \ , , \op_{\eps}\left(a\right ) \right ] + \frac{i}{\eps}\left [ \ , \|x'\|f(x ) \ , ,
\op_{\eps}\left(a\right ) \right].\ ] ] in next sections we will analyse separately each one of these pieces . let us start the computation of the first term in the right - hand side of by the following exact calculation , with arbitrary @xmath673 : @xmath674 \psi(x ) \\ & = & \frac{i\eps}{2(2\pi\eps)^d}\int_{\r^{2d } } \left [ e^{\frac{i}{\eps}\xi \centerdot ( x - y ) } a\left(\frac{x+y}{2},\xi\right ) \delta_y \psi(y ) - \delta_x \left ( e^{\frac{i}{\eps}\xi \centerdot ( x - y ) } a\left(\frac{x+y}{2},\xi\right ) \right ) \psi(y ) \right ] d\xi dy \\ & = & \frac{i\eps}{2(2\pi\eps)^d}\int_{\r^{2d } } \left ( \delta_y - \delta_x \right ) \left ( e^{\frac{i}{\eps}\xi \centerdot ( x - y ) } a\left(\frac{x+y}{2},\xi\right ) \right ) \psi(y ) \ , d\xi dy \\ & = & \frac{1}{(2\pi\eps)^d}\int_{\r^{2d } } e^{\frac{i}{\eps}\xi \centerdot ( x - y ) } \xi \centerdot \partial_x a\left(\frac{x+y}{2},\xi\right ) \psi(y ) \ , d\xi dy \\
& = & \op_\eps \left(\xi \centerdot \partial_x a(x,\xi)\right ) \psi(x).\end{aligned}\ ] ] observe that @xmath675 , thus the pseudodifferential operator above can be extended to @xmath14 , where it will be uniformly bounded with respect to @xmath146 .
moreover , using last identity , @xmath676 & = & t_\phi \frac{i}{\eps } \left [ -\frac{\eps^2}{2 } \delta \ , , \ , t_\phi^ * \ , \op_\eps(a ) \ , t_\phi \ , \right ] t_\phi^ * \\ & = & t_\phi \frac{i}{\eps } \left [ -\frac{\eps^2}{2 } \delta \ , , \ , \op_\eps\left(a\circ \tilde{\phi}\right ) \ , \right ] t_\phi^ * + o\left(\eps\right ) \\ & = & t_\phi \ , \op_\eps \left ( \xi \centerdot \partial_x \left(a \circ \tilde{\phi } \right ) \right ) t_\phi^ * + o\left(\eps\right ) \\ & = & \op_\eps\left ( \left ( \xi \centerdot \partial_x \left(a \circ \tilde{\phi } \right ) \right)\circ \tilde{\phi}^{-1 } \right ) + o\left(\eps\right ) \\ & = & \op_\eps\left(d(x)\xi \centerdot \partial_x a \right ) + o\left(\eps\right),\end{aligned}\ ] ] where @xmath677 reads : @xmath678 \left [ \ : ^t\nabla g\left(\phi^{-1}(x)\right ) \;\ ; ^t\nabla f\left(\phi^{-1}(x)\right ) \ : \right].\ ] ] the result is : @xmath679 \psi^\eps_t \ , , \psi^\eps_t \right > dt \nonumber \hspace{-2 cm } & & \\ & \tend{\eps}{0 } & - \left < d(x ) \xi \centerdot \partial_x \mu(t , x,\xi ) \
, , \xi(t ) \ , a(x,\xi ) \ , \right>_{\r\times\r^{2d}}. \label{eq : kinetic_part}\end{aligned}\ ] ] [ rem : convenience ] observe that @xmath680 and @xmath681 . back to the original coordinates ,
this gives @xmath682 and @xmath683 .
besides , from we have @xmath684 and @xmath685 , which implies @xmath686 as we chose @xmath687 so as to have @xmath616 and @xmath617 orthogonal when @xmath104 , we are left with @xmath688 , where @xmath689 is invertible for @xmath690 and the same for @xmath691 analogously defined . consider the taylor development @xmath692 and the fact that @xmath693 $ ] has kernel @xmath694 as in standard symbolic calculus , we use formula @xmath695 to exchange the factors @xmath696 in by factors @xmath697 in . because both @xmath698 and @xmath699 do not grow faster than some polynomial and @xmath53 is compactly supported , it is a direct computation to get @xmath700 = \op_\eps \left ( -\nabla v_s \centerdot \partial_\xi a \right ) + \mathcal{o}\left(\eps\right),\ ] ] where @xmath701 tends to @xmath27 in @xmath275 and @xmath702 is uniformly bounded with respect to @xmath146 .
this naturally gives @xmath703 \psi^\eps_t \ , , \psi^\eps_t \right > dt \nonumber \hspace{-1 cm } & & \\ & \tend{\eps}{0 } & \left < \ , \nabla v_s ( x ) \centerdot \partial_\xi \mu ( t , x,\xi ) \ , , \xi(t ) \ , a(x,\xi ) \
, \right>_{\r \times \r^{2d}}.\end{aligned}\ ] ] in order to analyse the commutator with @xmath704 in , we will introduce a cut - off @xmath705 , @xmath320 , @xmath706 for @xmath707 and @xmath708 for @xmath709 .
let us cut the symbol @xmath53 in three parts using parameters @xmath710 and @xmath711 , as follows : @xmath712 in the context of two - microlocal analysis , each of these pieces is related to a different two - microlocal measure , and that is what we will be talking about in the next sections .
defining @xmath713 , one calculates : @xmath714 \psi_t^\eps \ , , \psi_t^\eps \right > \nonumber \\ & = & \frac{i}{\eps\left(2\pi\eps \right)^{d } } \int_{\r^{3d } } e^{\frac{i}{\eps}\xi \centerdot \left ( x - y \right ) } \ , a\left(\frac{x+y}{2},\xi\right ) \chi\left(\frac{x'+y'}{2\eps r}\right ) \left ( \|x'\| f(x ) -
\|y'\| f(y ) \right ) \nonumber \\ & & \hspace{9 cm } \psi_t^\eps\left(y\right ) \overline{\psi_t^\eps\left(x\right ) } \ ; dxd\xi dy \nonumber \\ & = & \int_{\r^{3d } } \frac{ie^{i \xi \centerdot \left ( x - y \right)}}{\left(2\pi \right)^{d } } \ , a\left(\eps\frac{x'+y'}{2},\frac{x''+y''}{2},\xi',\eps\xi''\right ) \chi\left(\frac{x'+y'}{2r}\right ) \nonumber \\ & & \hspace{4.2 cm } \left ( \|x'\| f(\eps x',x '' ) - \|y'\| f(\eps y ' , y '' ) \right ) \tilde{\psi}_t^\eps\left(y\right ) \overline{\tilde{\psi}_t^\eps\left(x\right ) } \ ; dxd\xi dy \nonumber \\ & = & \int_{\r^{3d } } \frac{ie^{i \xi \centerdot \left ( x - y \right)}}{\left(2\pi \right)^{d } } \ , a\left(0,\frac{x''+y''}{2},\xi',\eps\xi''\right)\chi\left(\frac{x'+y'}{2r}\right ) f\left(0,\frac{x''+y''}{2}\right ) \left ( \|x'\| - \|y'\| \right ) \nonumber \\ & & \hspace{8 cm } \tilde{\psi}_t^\eps\left(y\right ) \overline{\tilde{\psi}_t^\eps\left(x\right ) } \ ; dxd\xi dy + r^\eps,\end{aligned}\ ] ] with @xmath468 an error of order @xmath146 in @xmath96 whose analysis will be postponed .
now , for each @xmath715 , denote by @xmath716 the integral kernel of the @xmath717 operator @xmath718 $ ] , where @xmath719 is the weyl quantization of the symbol @xmath720 , and , as in , by @xmath721 the kernel of the correspondent bounded @xmath722 operator @xmath723 ( which is an operator - valued generalization of the wigner transform @xmath724 ) introduced in proposition [ prop : two_microlocal_measures ] .
then , the object in the previous calculation reads @xmath725 which gives @xmath726 \psi_t^\eps \ , , \psi_t^\eps \right > = { \rm tr } \int_{\r^{2(d - p ) } } a^r_{\left(x'',\xi''\right ) } u^\eps_{\left(t,\ , x'',\xi''\right)}dx''d\xi '' + r^\eps.\ ] ] regarding the error : @xmath727 where @xmath728 and @xmath729 are the integral operators with the respective kernels : @xmath730 and @xmath731 @xmath732 @xmath733 ds.\ ] ] [ lem : b_c_bounded ] the operators @xmath728 and @xmath729 are uniformly bounded with respect to @xmath146 .
we will use the conventional schur test and the fact that both @xmath734 and @xmath735 are rapidly decreasing functions : @xmath736 with @xmath737 ( recall that @xmath738 is non - decreasing , @xmath60 the bounding polynomial in the definition of conical potentials given in the beginning of section [ sec : statement ] ) , and exactly the same estimate holds for @xmath739 . again the same argument for the other operator : @xmath740 the same holding for @xmath741 .
as one sees , these bounds are independent of @xmath146 , so the lemma holds .
now we only need to focus on the lasting term ; from what we have seen in section [ sec : two_microlocal_analysis ] , in the limit where @xmath23 and then @xmath317 , it gives @xmath742 \psi_t^\eps \ , , \psi_t^\eps \right > dt \nonumber \hspace{-8 cm } & & \\ & \longrightarrow & { \rm tr } \left < \ , m\left(t , x'',\xi''\right ) \
, , \ , i \left [ \ , \|y\|f\left(0,x''\right ) , \xi(t ) \ , a^{w}\left(0,x'',\partial_y,\xi''\right ) \ , \right ] \
, \right>_{\r \times \r^{2(d - p ) } } , \label{eq : inner_part}\end{aligned}\ ] ] where @xmath304 is the two - microlocal operator - valued measure in proposition [ prop : two_microlocal_measures ] . in section [ sec : establishing_the_equation ] , @xmath304 will be shown to be zero ; for now , let us prove : [ lem : estimate_for_m ] one has got the estimate @xmath743 \ , \right>_{\r \times \r^{2(d - p ) } } \hspace{2 cm } \nonumber \\ \leqslant tk \sup_{\underset{|\alpha| \leqslant d+1}{\alpha\in \mathbb{n}_0^d } } \ , \sup_{x '' \in \r^{d - p } } \int_{\r^d } \left\| \partial_\xi^\alpha \partial_{\xi ' } a(0,x'',\xi ) f(0,x '' ) \right\|_{\r^p } d\xi,\end{aligned}\ ] ] where @xmath744 are constants , @xmath745)$ ] and @xmath39 . from a calculation similar to that we made in and similar estimates , it follows that @xmath726 \psi_t^\eps \ , , \psi_t^\eps \right > = \left < p^\eps \ , \tilde{\psi}^\eps_t \ , , \ , \tilde{\psi}^\eps_t \right > + \mathcal{o}(\eps),\ ] ] where @xmath746 is the operator with kernel @xmath747 ( recall : @xmath748 ) .
further , let be @xmath749 the kernel of the operator @xmath750 , with @xmath751 where @xmath752 .
observe that @xmath753 , which causes the schur estimate for the norm of @xmath754 to be greater than that for @xmath746 .
besides , the schur estimate for @xmath754 is upper bounded by an estimate of type ( see remark [ rem : schur_upper_bound ] ) , which must , consequently , be an upper bound for the norm of @xmath746 as well .
this estimate is the one we stated in the lemma . in @xcite , an estimate that turns up to be equivalent to last lemma
was obtained by noticing directly that @xmath755 $ ] is bounded uniformly with respect to @xmath146 , which was used in proving that @xmath756 $ ] is itself bounded .
we start by proving with standard symbolic calculus the technical result below : [ lem : better_commutator ] for @xmath366 , @xmath757 , one has the following estimation in @xmath275 : @xmath758 = \frac{i}{\eps}\left [ \| x ' \| f(x ) \left(1-\chi\left(\frac{x'}{\delta}\right)\right ) \ , , \op_\eps \left ( a(x,\xi ) \right ) \right ] \\ - \op_\eps \left(\frac{1}{\delta } \|x'\|f(x ) \ , \chi ' \left ( \frac{x'}{\delta } \right)\centerdot\partial_{\xi'}a(x,\xi)\right ) + \mathcal{o}\left(\frac{\eps}{\delta}\right).\end{aligned}\ ] ] in view of the identities @xmath759 where @xmath760 and @xmath761 have integral kernels @xmath762 and @xmath763 we have @xmath758 = \frac{i}{\eps}\left [ \| x ' \| f(x ) \left(1-\chi\left(\frac{x'}{\delta}\right)\right ) \ , , \op_\eps \left ( a(x,\xi ) \right ) \right ] \\ + \underbrace{\frac{i}{\eps}\left ( \|x'\| f(x ) r_{\eps,\delta}^l - r_{\eps,\delta}^r \|x'\| f(x ) \right)}_{r_{\eps,\delta}},\end{aligned}\ ] ] where @xmath764 has kernel @xmath765 using taylor developments for @xmath319
, this kernel can be written in the form @xmath766\end{aligned}\ ] ] and still , with developments for @xmath767 , @xmath768\end{aligned}\ ] ] where @xmath769 @xmath770 and @xmath771 observe now that @xmath772 , @xmath773 and @xmath774 are bounded and , furthermore , @xmath773 is supported on @xmath775 ( given that @xmath776 is null for @xmath777 ) .
this , along with the identity @xmath778 , allows us to write @xmath764 as @xmath779 where @xmath780 has an integral kernel such that @xmath781 where @xmath782 is some constant big enough .
the lemma will follow as one uses the estimation above for the schur test to show that @xmath783 .
[ rem : useful_in_middle_part ] replacing @xmath379 by @xmath784 and @xmath785 by @xmath786 , lemma [ lem : better_commutator ] gives @xmath787 = \frac{i}{\eps}\left [ \| x ' \| f(x ) \left(1-\chi\left(\frac{x'}{\eps r}\right)\right ) \
, , \op_\eps \left ( a(x,\xi)\ , \chi\left(\frac{x'}{\delta}\right ) \right ) \right ] \\ - \op_\eps \left(\frac{1}{\eps r } \|x'\|f(x ) \ , \chi ' \left ( \frac{x}{\eps r } \right)\centerdot\partial_{\xi'}a(x,\xi)\ , \chi\left(\frac{x'}{\delta}\right)\right ) + \mathcal{o}\left(\frac{1}{r}\right),\end{aligned}\ ] ] which is going to be remarkably useful in section [ sec : the_middle_part ] .
[ lem : commutator_smooth ] for @xmath366 , @xmath757 , one has @xmath788 = \op_\eps \left ( - \nabla\left(\|x'\|f(x)\right ) \centerdot \partial_{\xi } a(x,\xi )
\left(1-\chi\left(\frac{x'}{\delta}\right ) \right ) \right ) \\ + \op_\eps \left ( \frac{1}{\delta } \|x'\|f(x ) \ , \chi ' \left ( \frac{x'}{\delta } \right)\centerdot\partial_{\xi'}a(x,\xi ) \right ) + \mathcal{o}\left(\eps\right ) + \mathcal{o}\left(\frac{\eps}{\delta}\right)\end{aligned}\ ] ] in @xmath275 .
because @xmath789 is everywhere smooth , we can apply usual symbolic calculus as in [ eq : taylor_smooth ] and [ eq : smooth_piece ] to get @xmath790 = \op_\eps \left ( - \partial_x \left ( \|x'\|f(x ) \right ) \centerdot \partial_{\xi } a(x,\xi ) \left(1-\chi\left(\frac{x'}{\delta } \right ) \right ) \right ) + r^{\eps,\delta}\ ] ] where @xmath791 has kernel @xmath792 with @xmath773 the matrix @xmath793 it is easy to see that @xmath794 , with @xmath795 constants , and therefore @xmath796 the same estimation holding for @xmath797 , it follows by the schur test that @xmath798 .
the rest of the proof consists on the basic derivation @xmath799 and on the linearity of the pseudodifferential operators with respect to their symbols .
[ rem : useful_in_middle_part_2 ] analogously to remark [ rem : useful_in_middle_part ] , in section [ sec : the_middle_part ] we will use that @xmath800 = \op_\eps \left ( - \nabla\left(\|x'\|f(x)\right ) \centerdot \partial_{\xi } a(x,\xi ) \left(1-\chi\left(\frac{x'}{\eps
r}\right ) \right ) \chi\left(\frac{x'}{\delta}\right ) \right ) \\ + \op_\eps \left ( \frac{1}{\eps r } \|x'\|f(x ) \ ,
\chi ' \left ( \frac{x'}{\eps r } \right)\centerdot \partial_{\xi'}a(x,\xi )
\chi\left(\frac{x'}{\delta}\right ) \right ) + \mathcal{o}\left(\eps\right ) + \mathcal{o}\left(\frac{1}{r}\right).\end{aligned}\ ] ] there , however , there will be a tiny question about the limit in @xmath146 . combining lemmata [ lem : better_commutator ] and [ lem : commutator_smooth ] ,
we obtain @xmath758 = \op_\eps \left ( - \nabla\left(\|x'\|f(x)\right ) \centerdot \partial_{\xi } a(x,\xi ) \left(1-\chi\left(\frac{x'}{\delta}\right ) \right ) \right ) \\
+ \mathcal{o}(\eps ) + \mathcal{o}\left(\frac{\eps}{\delta}\right),\end{aligned}\ ] ] which gives , in the limit where @xmath23 and then @xmath318 , @xmath801 \psi^\eps_t \ , , \psi^\eps_t \right > dt \nonumber \hspace{-7 cm } & & \\ & \longrightarrow & \left < \ , \nabla\left ( \|x'\|f(x ) \right ) \centerdot \partial_\xi \mu(t , x,\xi ) 1\!\!1_{\left\lbrace x ' \neq 0 \right\rbrace } \ , , \xi(t ) \ , a(x,\xi ) \ , \right>_{\r\times\r^{2d}}. \label{eq : outer_part}\end{aligned}\ ] ]
as seen in remarks [ rem : useful_in_middle_part ] and [ rem : useful_in_middle_part_2 ] , the calculations in section [ sec : the_inner_part ] lead to @xmath802 \hspace{-5 cm } & & \\ & = & \op_\eps \left ( - \nabla\left(\|x'\|f(x)\right ) \centerdot \partial_{\xi } a(x,\xi ) \left(1-\chi\left(\frac{x'}{\eps r}\right ) \right ) \chi\left(\frac{x'}{\delta}\right ) \right ) + \mathcal{o}\left(\eps\right ) + \mathcal{o}\left(\frac{1}{r}\right),\end{aligned}\ ] ] which implies in the two - microlocal limit when @xmath23 , then @xmath317 and finally @xmath318 : @xmath803 \psi^\eps_t \ , , \psi^\eps_t \right > dt \nonumber \hspace{-9.5 cm } & & \\ & \longrightarrow & \left < \delta(x ' ) \otimes \int_{\mathcal{s}^{p-1 } } f(x ) \ , \omega \centerdot \partial_{\xi ' } \nu_\infty(t , x'',\xi , d\omega ) \ , , \xi(t ) \ , a(x,\xi ) \right>_{\r\times\r^{2d } } , \label{eq : middle_part}\end{aligned}\ ] ] where @xmath302 is the two - microlocal measure on sphere introduced in proposition [ prop : two_microlocal_measures ] , and @xmath379 is the usual counting measure ( or dirac mass ) . from equations , and , more the results in the last sections , equations , , , and
, we obtain the equation @xmath804 , a^w(t,0,x'',\partial_y,\xi '' ) \ , \right>_{\r\times\r^{2(d - p)}}\end{aligned}\ ] ] for the full wigner measure @xmath24 tested against a function @xmath805 .
we will now work out this expression until we prove proposition[prop : main_result ] ahead .
[ lem : m_is_zero_and_etc ] in the two - microlocal decomposition given in lemma [ lem : two_microlocal_decomposition ] , the operator valued measure @xmath304 is zero and @xmath806 for some measure @xmath110 on @xmath807 .
consequently , equation reads @xmath808 for all test functions @xmath805 . to begin with , re - write equation as @xmath809 , a^w(t,0,x'',\partial_y,\xi '' ) \ , \right>.\end{aligned}\ ] ] now , recall that the term in the trace obeys to estimate given in lemma [ lem : estimate_for_m ] ; besides , since @xmath24 is a measure in @xmath810 , we have @xmath811 where @xmath812 since @xmath24 is finite and @xmath813 , the support of @xmath53 , is compact . obviously the very same estimate ( with @xmath814 in the right - hand side ! ) is valid for @xmath815 and for @xmath816 .
so , for test functions of the form @xmath817 , equation gives ( recalling also remark [ rem : convenience ] ) : @xmath818 that at the limit where @xmath318 results in @xmath819 which means that @xmath820 is supported on @xmath821 , and , by positivity , @xmath822 is necessarily supported on @xmath823 and the measure @xmath335 introduced in remark [ rem : definition_m ] in @xmath824 .
regarding @xmath302 , any measure supported on @xmath825 can only be a dirac mass thereon , where @xmath826 and @xmath827 is the @xmath828-th distribution derivative of the dirac @xmath379 . as our distribution must be a positive measure as well ,
the only allowed term in this development is the one with @xmath829 .
] , whence there must be @xmath110 as stated in the lemma . however , for @xmath335 we already knew that it was absolutely continuous with respect to @xmath830 ( lemma [ lem : m_absolutely_continuous ] ) , so being zero almost everywhere in @xmath831 is , in other words , saying that @xmath339 . from lemma [ lem : m_0_gives_m_0 ]
, this implies that @xmath338 and , finally , that so is the term in the trace in equation . to conclude , just re - write attaching all the information we have just got and verify that it simplifies to .
[ rem : measure_over_the_singularity ] as a scholium of the last proof , one has that @xmath24 is not supported on the region of the @xmath831 axis away from the origin , @xmath832 .
this implies that @xmath833 , which is a result that @xcite had already obtained with a similar argument .
[ lem : technical ] one has the identity @xmath834 .
recall estimate , which holds for @xmath815 and for @xmath835 as well .
thus , for test functions of the form @xmath836 and proceeding in the same manner as in the proof of lemma [ lem : m_is_zero_and_etc ] , equation gives , in the limit where @xmath318 , @xmath837 which means that the distribution @xmath838 is supported on @xmath839 .
but of course this carries that it is null .
[ lem : symmetry_condition ] the measure @xmath110 introduced in lemma [ lem : m_is_zero_and_etc ] obeys to the following identity in the sense of the distributions on @xmath840 : @xmath841 where @xmath66 and @xmath842 are as in .
for test functions of the form @xmath843 , using estimate and lemma [ lem : technical ] , the present lemma follows from at the limit @xmath318 .
to finish establishing a liouville equation for @xmath24 , let us put all lemmata [ lem : m_is_zero_and_etc ] , [ lem : technical ] , [ lem : symmetry_condition ] and remark [ rem : convenience ] together and write equation in a distributional and clearer way : @xmath844 or , more explicitly , in view of remark [ rem : measure_over_the_singularity ] : [ prop : main_result ] let be @xmath845 a conical potential with @xmath84 , @xmath63 , @xmath846 .
let be @xmath242 the solution to the schrdinger equation with potential @xmath1 .
then , the wigner measure associated to the concentration of @xmath56 in the limit @xmath23 can be decomposed as @xmath847 where the measure @xmath110 satisfies the asymmetry condition @xmath848 besides , it obeys to the following distributional equation in @xmath849 : @xmath850 where@xmath851 .
[ sec : absolute_continuity ] from equation , it is obvious that @xmath24 is absolutely continuous with respect to the lebsgue measure @xmath31 , which implies the existence of a function @xmath852 taking values in the set of positive measures on @xmath103 such that @xmath121 .
the same holds for the two - microlocal measure @xmath110 , since it is positive and absolutely continuous with respect to @xmath24 , carrying the existence of a function @xmath853 such that @xmath854 .
nevertheless , the same is not true for continuity .
in fact , in @xcite it was shown that @xmath33 is continuous inside a compact @xmath36 $ ] by verifying that , for @xmath39 , the commutator @xmath855 $ ] is uniformly bounded with respect to @xmath146 ( which we have also done indirectly during the calculations in this section ) , so , from , one sees that the family @xmath856 is equicontinuous in @xmath36 $ ] in addition to being equibounded , which implies the continuity of @xmath36 \ni t \longmapsto \left <
\mu_t , a \right>_{\r^{2d}}$ ] by the ascoli - arzel theorem .
this is not true for @xmath857 in general , as the examples in theorem [ th : main_result_3 ] show . in both cases there
, we have @xmath858 for any @xmath400 and @xmath859 , so @xmath860 if and only if @xmath73 , which shows lack of continuity for @xmath857 in spite of @xmath88 .
consequently , it makes no sense to think of @xmath110 on a @xmath6 by @xmath6 basis unless directly linked to @xmath24 , as in lemma [ lem : two_microlocal_decomposition ] .
in particular , observe that there may be violations of the asymmetry condition for a particular instant @xmath6 or , more generally , for time sets of null lebesgue measure .
this is not at all a contradiction , since this condition is only valid _ in average _ in time . on the other hand
, one could use an argument of continuity for @xmath88 to see , in the one - dimensional case @xmath861 , that a family @xmath862 whose initial data concentrate to @xmath863 will remain concentrating to this same point , @xmath864 . in this case
, condition allows a complete description of @xmath110 : @xmath865
in last section we presented a trivial application of the asymmetry condition . in this section
we will apply it to obtain theorem [ th : main_result_4 ] . moreover
, we will prove theorems [ th : accessory_result_1 ] and [ th : accessory_result_2 ] .
in next section we will give examples of applications of these results .
so , to start with : since @xmath66 and @xmath866 are continuous functions all over @xmath2 , there exists a neigbourhood @xmath867 of @xmath151 such that the inequality @xmath868 holds strictly for every @xmath869 .
let be @xmath870 , and let us test expression against this function : @xmath871 further , realize that the testing term is always non - zero whithin the support of @xmath53 : @xmath872 as a consequence , since @xmath110 is always positive , one must have @xmath137 over the support of @xmath53 , more precisely , over @xmath873 .
now , in order to study the classical flow in more detail , remark that in the transformed coordinates introduced in section [ sec : reducing ] , the equation of motion for the component of @xmath72 in @xmath69 , @xmath874 , reads : @xmath875 let us suppose that @xmath4 is a trajectory that reaches the phase space singularity @xmath68 on the point @xmath164 at a time @xmath73 .
so we have @xmath876 , @xmath877 , and , for @xmath400 , we can define @xmath878 , or else : @xmath879 let be @xmath173 such that @xmath880 . necessarily the sequence @xmath881 will have convergent subsequences .
let us call a _ point of adherence _ of the function @xmath882 the limits of convergent subsequences of @xmath881 , where @xmath883 is some sequence converging to @xmath27 .
let us define the points of adherence of @xmath884 in an analogous manner , when they exist .
[ lem : conditions_on_the_flow ] let be @xmath885 ; then a vector @xmath886 is a point of adherence of @xmath887 if and only if @xmath888 is a point of adherence of @xmath882 . in this case
, @xmath889 is a non - zero solution to the equation @xmath890 geometrically , this equation reads ( remembering that @xmath891 is the derivative normal to @xmath69 ) : @xmath892 where @xmath185 is a point of adherence of the @xmath171 vector function @xmath893 .
first , let us suppose that all points of adherence of @xmath887 are non - zero . at the end of the proof , we will verify this fact for the case where @xmath894 .
so , observe that @xmath895 for any @xmath400 , so if @xmath889 is non - zero and point of adherence of @xmath887 , then necessarily @xmath888 is point of adherence of @xmath882 .
the converse comes from the definition of @xmath887 in : since at the limit @xmath896 one has @xmath897 , @xmath898 , @xmath899 and @xmath900 .
thus , whenever a sequence @xmath883 is such that @xmath881 converges to some limit that one can write @xmath901 , @xmath902 also converges to a vector @xmath889 , that we suppose non - zero . by the previous identity @xmath895
, one must have @xmath903 .
equation comes from taking the limits @xmath880 in equation .
multiplying its both sides by @xmath904 , one gets @xmath905 two things remain before completing the proof : recognizing equation from , and proving that any point of adherence of @xmath887 is non zero under the lemma s hypotheses . the latter is done by remarking that , if we lift the hypothesis that all points of adherence of @xmath887 are non zero , then in the same way that we obtained , we would get , for any point of adherece @xmath906 of @xmath887 : @xmath907 where @xmath889 is some point of adherence of @xmath882 not necessarily a function of @xmath906 . however , if @xmath908 , surely @xmath909 ( as we saw in the proof of theorem [ th : main_result_4 ] ) , so @xmath906 is non - zero . finally , recall that @xmath874 is the coordinate of the variable in the normal bundle @xmath171 ; calling it @xmath54 in the original coordinates , we have @xmath910 , so in we have just a point of adherence of @xmath911 , _
i.e. _ , @xmath912 .
[ rem : directions ] in particular , the previous lemma implies that whenever a trajectory reaches @xmath68 , then @xmath887 has some points of adherence and they are all non - null . hence their interpretation as the directions in space through which the trajectories get in or out the singularities
as we have seen in the remark above , for any trajectory arriving on @xmath68 one can talk about its approaching directions ( not necessarily unique ) ; if @xmath913 , then these directions are submitted to satisfy equation , which implies , in particular , that if this equation has no non - zero roots , then by absurd no trajectory at all can reach @xmath68 .
conversely , lemma [ lem : conditions_on_the_flow ] does not say whether there are actual trajectories approaching @xmath68 in all the directions satisfying . below we will verify that indeed any @xmath185 satisfying is realized by some trajectory .
[ lem : part_proof_2 ] if @xmath914 satisfy , then there exists a unique trajectory @xmath4 reaching @xmath68 at @xmath73 in a point @xmath164 such that @xmath915 .
first , let us choose @xmath916 .
for @xmath917 and @xmath918 such that @xmath919 and @xmath917 is the coordinate of @xmath151 , take @xmath920 , @xmath921 and @xmath922 sufficiently small so as the set @xmath923 } = \left\lbrace ( x,\xi,\vartheta ) \in \left(c^0 \left([-\tau,\tau]\right)\right)^3 : ( x,\xi,\vartheta)(0 ) = \left(x_0,\xi_0,\lambda \theta_0\right ) \ ; { \rm and } \hspace{3 cm } \right .
\sup_{t \in [ -\tau,\tau ] } \left ( \| x(t ) - x_0 \| + \| \xi(t ) - \xi_0 \| + \| \vartheta(t ) - \lambda \theta_0 \| \right ) \leqslant \delta \right\rbrace\end{aligned}\ ] ] fits in a proper definition of the application @xmath924 } \longrightarrow \mathcal{b}^{[-\tau,\tau]}$ ] given by : @xmath925 @xmath926 and last @xmath927 observe that taking @xmath928 and @xmath929 as small as necessary , @xmath930 becomes a contraction on @xmath931}$ ] , a set which is complete with respect to its natural supremum norm .
consequently , by banach s fixed point theorem , there exists a unique triple @xmath932}$ ] such that @xmath933 ; this is equivalent to saying that the system @xmath934 with initial data @xmath935 admits a unique solution , which must be such that @xmath936 @xmath937 .
therefore , @xmath4 must be a trajectory of our conical problem with the properties listed in the lemma for @xmath938 . for @xmath939 ,
define the sets @xmath940}$ ] and @xmath941}$ ] and procceed as above in order to show the existence for @xmath942 $ ] of trajectories @xmath943 such that @xmath944 . besides , as @xmath945 and @xmath946 , we can build a continuous path @xmath4 by setting @xmath947 and @xmath948 for @xmath565 , and this new trajectory will be hamiltonian and the unique one to meet the properties stated in the lemma .
bringing together lemmata [ lem : conditions_on_the_flow ] and [ lem : part_proof_2 ] , one proves theorem [ th : accessory_result_1 ] . in order to obtain theorem [ th : accessory_result_2 ] , we observe the following facts : 1 .
the hypothesis @xmath908 played no role in the demonstration of lemma [ lem : part_proof_2 ] , for we had supposed @xmath949 at first .
2 . in the proof of lemma [ lem : conditions_on_the_flow ] , this hypothesis was needed only in showing that @xmath949 ; it was not necessary in getting equation [ eq : blabla ] , where we could still have adherence points @xmath949 and @xmath950 , giving equation ( a `` zero root '' of ) .
in any case , if some trajectory is to arrive onto @xmath68 in @xmath151 , then either or must be satisfied .
3 . if @xmath889 is a `` zero root '' , then we will see in example [ ex : equal_no ] that it may happens that no trajectory following the direction @xmath889 ever touches @xmath68 in @xmath151 .
on the other hand , in example [ ex : equal_yes ] we will see a case where there is a trajectory approaching the singularity , but only asymptotically .
the proof of theorem [ th : accessory_result_2 ] will be complete after showing that any such trajectories only reach @xmath68 after an infinite time . letting be @xmath4 a trajectory that is not on @xmath68 at @xmath73 ,
define @xmath951 , ( x(s),\xi(s ) ) \notin \omega \right\rbrace$ ] and @xmath952 .
if @xmath949 obeys to with @xmath950 and we suppose that there is a trajectory @xmath4 such that @xmath953 , that for any @xmath954 there is @xmath955 for which @xmath956 and such that @xmath957 , then @xmath958 .
consider @xmath669 a smooth extension of @xmath1 outside the closure of @xmath867 such that @xmath959 and that @xmath960 for any @xmath961 .
such a @xmath669 is possible to be found since from the hypotheses @xmath949 and @xmath950 in one knows that @xmath962
. then @xmath963 is at the same time a hamiltonian trajectory of a conical potential and of a standard problem with smooth potential , for which case it is widely known that no trajectory can arrive at an extremum of @xmath669 with null speed @xmath831 within a finite time ( since this would break the unicity of the constant solution @xmath964 ) . in this section
we will give examples of how theorems [ th : main_result_4 ] , [ th : accessory_result_1 ] and [ th : accessory_result_2 ] can be used in order to classify the trajectories that arrive on a conical singularity and , sometimes , completely describe the transport phenomenon to which the semiclassical measures are submitted .
in particular , examples [ ex : equal_no ] and [ ex : equal_yes ] are part of the reasoning that led to the obtaintion of the second assertive in theorem [ th : accessory_result_2 ] .
to begin with , let us consider @xmath192 and @xmath965 , so @xmath966 .
[ ex:1 ] take @xmath967 and @xmath968 . 2 +
+ then admits no non - zero solutions , which is consistent with the fact that the classical flow @xmath7 presents no trajectories hitting the singularity . in this case
, the asymmetry condition gives that @xmath969 with @xmath970 the total mass over the singularity and does not depend on @xmath6 .
the semiclassical transport here is thus completely solvable starting from some initial measure .
[ ex:2 ] consider @xmath967 and take @xmath971 .
2 + + then admits two solutions : @xmath972 and @xmath973 , which is consistent with the fact that the classical flow @xmath7 does present trajectories hitting the singularity from both directions @xmath201 and @xmath211 . in this case , we will have @xmath974 with @xmath975 the total mass over the singularity , which may vary with @xmath6 .
[ ex : equal_no ] take @xmath976 and @xmath968 . 2 +
+ then equation has no non - zero roots , but equation admits any solution @xmath977 . in this case , there are no trajectories hitting the singularity from @xmath215 .
the semiclassical measure in sphere will be @xmath978 where the total mass over the singularity @xmath970 is constant .
again , this is a completely solvable case .
[ ex : equal_yes ] pick up @xmath976 and @xmath979 . 2 +
+ then equation admits one solution : @xmath980 , and equation also admits solutions : any @xmath981 .
this is consistent with the fact that the classical flow @xmath7 presents atrajecotry hitting the singularity from the direction @xmath211 , and in this case also from the direction @xmath201 .
however , the trajectory from the positive side takes an infinite time to get close to the singularity .
one has : @xmath982 the total mass on the singularity possibly changin with time .
[ ex:5 ] take @xmath983 and choose @xmath968 .
2 + + then equation admits a unique solution : @xmath984 , which is consistent with the fact that the classical flow @xmath7 only presents a trajectory hitting the singularity from the direction @xmath211 . in this case , since @xmath985 , we will have @xmath137 , so the semiclassical measure will necessarily follow the flow extended so as the singularity will be part of the parabola in @xmath211 . the problem is hence completely solvable , even though there are trajectories leading to the singular set
now let be @xmath986 and denote @xmath987 .
choose an exterior potential @xmath988 , @xmath971 and @xmath989 .
then @xmath990 and no @xmath185 satisfies equation . as a conclusion ,
no trajectory in this case can hit the singularity at @xmath77 .
[ ex : last ] last , we will take the same @xmath988 , @xmath971 , but @xmath991 .
then @xmath992 but now equation admits any solution @xmath993 with @xmath994 .
this is a case where the exterior force polarizes the flow in its direction , but leaves it free to turn around a circle of radius @xmath995 in the orthogonal plane .
we would like to thank dr .
clotilde fermanian for his very helpful support and fruitful advises all over the writing of this paper .
we also thank dr .
fabricio maci for the help in sections [ sec : the_outer_part ] and [ sec : establishing_the_equation ] .
l. ambrosio , a. figalli , _ almost everywhere well - posedness of continuity equations with measure initial data _ , c. r. math .
paris 1 , * 348 * , 249 - 252 ( 2010 ) .
l. ambrosio , a. figalli , g. friesecke , j. giannoulis , t. paul , _
semiclassical limit of quantum dynamics with rough potentials and well - posedness of transport equations with measure initial data _
, * 64 * , 1199 - 1242 ( 2011 ) .
a. athanassoulis , _ generalized wave packets for the semiclassical schrdinger equation with power non - linearitis _ , arxiv:1505.04707[math.ap ] ( 2015 ) .
a. athanassoulis , t. paul , _
strong and weak semiclassical limits for some rough hamiltonians _ , math .
models methods appl .
, * 22 * , 1250038 ( 2012 ) .
a. athanassoulis , t. paul , _ on the selection of the classical limit for potentials with bv derivatives _ , j. dyn .
equat . , * 25 * , 33 - 47 ( 2013 ) n. anantharaman , f. faure , c. fermanian , _ chaos en mcanique quantique _ , ed .
p. harinck , a. plagne , c. sabbah , cole polytechnique , palaiseau , 2014 .
v. chabu , _ work in progress_. m. combescure , d. robert , _ coherent states and applications in mathematical physics _ , springer science & business media , 2012 .
m. dimassi , j. sjstrand , _ spectral asymptotics in the semiclassical limit _ , cambridge university press , 1999 . c. fermanian , _ mesures semiclassiques 2-microlocales _ , c.r .
paris 1 , * 331 * , 515 - 518 ( 2000 ) . c. fermanian , p. grard , c. lasser , _
wigner measure propagation and conical singularity for general initial data _ , arch .
, * 209 * , 209 - 236 ( 2012 ) . c. fermanian , c. lasser , _ wigner measures and codimension two crossings _ , j. math .
* 44 * , 507 - 527 ( 2003 ) .
a. figalli , m. ligab , t. paul , _
semiclassical limit for mixed states with singular and rough potentials _ , to appear in indiana univ .
journal , * 61 * , 193 - 222 ( 2012 ) .
p. grard , _ mesures semiclassiques et ondes de bloch _ , sminaire sur les quations aux drives partielles 1990 - 91 , xvi , cole polytechnique , palaiseau , 1991 .
p. grard , p. markowich , n. mauser , f. poupaud , _ homogeneization limits and wigner transforms _ , comm .
, * 50 * , 323 - 379 ( 1997 ) ; + erratum : comm . pure . appl . math , * 53 * , 280 - 281 ( 2000 ) .
g. hagedorn , _ semiclassical quantum mechanics _ ,
* 71 * , 77 - 93 ( 1980 ) .
l. harris , j. lukkarinen , s. teufel , f. theil , _ energy transport by acoustic modes of harmonic lattices _
, siam j. math .
, * 40 * , 1392 - 1418 ( 2008 ) . t. kato , _ perturbation theory for linear operators _ , springer verlag berlin heidelberg , 1995 .
p. lions , t. paul , _
sur les mesures de wigner _ ,
rev . mat .
iberoamericana , * 9 * , 553 - 618 ( 1993 ) . l. miller , _
propagation dondes semiclassiques travers une interface et mesures 2-microlocales _ , thse de lcole polytechnique , palaiseau , 1996 .
a. mielke , _ macroscopic behavior of microscopic oscillations in harmonic lattices via wigner - husime transforms _ , arch .
, * 181 * , 401 - 448 ( 2006 ) .
f. nier , _ a semiclassical picture of quantum scatering _ , ann .
sci . ecole norm . sup . 4 , * 29 * , 149 - 183 ( 1996 ) . m. reed , b. simon , _ methods of modern mathematical physics ii : fourier analysis , self - adjointness _ , academic press , 1975 .
r. schubert , r. vallejos , f. toscano , _ how do wave packets spread ?
time evolution on ehrenfest time scales _ , j. phys .
a , * 45 * , 215307 ( 2012 ) . l. tartar , _ h - measures , a new approach for studying homogeneization , oscillation and concentration effects in partial differential equations _ , proc .
edinb . a , * 115 * , 193 - 230 ( 1990 ) .
m. zworski , _ semiclassical analysis _ , american mathematical society , 2012 . | in this article we study the propagation of wigner measures linked to solutions of the schrdinger equation with potentials presenting conical singularities and show that they are transported by two different hamiltonian flows , one over the bundle cotangent to the singular set and the other elsewhere in the phase space , up to a transference phenomenon between these two regimes that may arise whenever trajectories in the outsider flow lead in or out the bundle .
we describe in detail either the flow and the mass concentration around and on the singular set and illustrate with examples some issues raised by the lack of unicity for the classical trajectories at the singularities despite the unicity for the quantum solutions , dismissing any classical selection principle , but in some cases being able to fully solve the propagation problem . + * keywords * : schrdinger equation , wigner measures , two - microlocal measures , symbolic calculus , quantum - classical correspondence , wave packet approximation . |
`` resource / service as a utility '' once a dream , is now a reality we are living with .
utility computing is not a new concept , but rather it has quite a long history . among the earliest references
is by john mccarthy .
last two decades of information technology ( it ) development has witnessed the specific efforts done to make this statement of john mccarthy a reality .
utility computing is providing basics for the current day resource utilization .
cluster , grids and now cloud computing have made this vision a reality .
e - collaborations also called _ virtual organizations _ have been evolved with the technological and paradigm shift .
cluster computing offered more centralized resource pool , while grid computing remaind in need via hardware and computation cycles offerings to the scientific community .
grid computing models observed a deadlock after the introduction of cloud computing concepts .
this situation was a result of missing economic business models , although some work was done on this issue even by the authors @xcite .
based on pay - as - you - use criteria , cloud computing is still in early stage . however , the economic component of cloud computing is a central focus of actual research @xcite .
research efforts are going on to establish the basis of cloud computing as every - thing - as - a - service paradigm .
infrastructure providing resources as a utility must be dynamic , scalable and reliable .
orchestration of resources across the globe , named as virtual organization ( vo)/virtual enterprize ( ve ) has been extensively deployed to achieve this target .
change in the hardware and software technology , computing paradigms algorithm and procedures , incorporation of knowledge rather information and data , made the concepts of vo vague .
though vo had been created utilizing the best technology known to that time , but the success was short lived .
there are three main issues , which has to be considered in order to understand : * advancement in hardware / software technology .
* birth of new computing paradigms . * changed nature of resources and requirements from end user .
we are living in the age of transformation .
a paradigm shift is one that effects the society as a whole . according to peter drucker such transformation place over fifty to sixty - year periods @xcite . in his book `` post - capitalist society '' , he outlines three earlier periods of dramatic changes in the western world . *
the rise of medieval craft guilds and urban centuries . long distance trade ( thirteenth - century europe ) @xcite . * the renaissance period of gutenburg s printing press and lutheran reformation ( 1455 - 1519 ) @xcite . * the industrial revolution , starting with watt s steam engine ( 1776 - 1815 ) @xcite .
existing technologies and paradigms do not vanish with the birth of new concepts rather they adopt what is positive and remove what is not required .
technology and paradigm used to form vo have also faced this transformation .
for example networking , distributed computing , cluster computing , grid computing , utility computing and now cloud computing , all are related and are improvements of the existing concepts .
when technology changes or improves , paradigm needs an upgrade too .
new methods and algorithms are created to support the hardware .
another main factor is the requirements from the user community .
the user community puts a demand on the technology and computing paradigm and they evolve accordingly .
`` resources / services as a utility '' is main theme of collaboration . to achieve the goal(s ) , organizations and individuals gather all the resources available .
the spectrum of availability has covered the whole globe . today , time and space are not a limit due to information and communication technology ( ict ) advancements .
this revolution has an impact on the resources types .
initial collaborations offered only storage and downloading ( p2p networks ) , computing cycles and storage space ( grid computing and cluster computing ) .
main focus remained at hardware and software sharing , but vos for scientific research initiated another requirement , i.e. need of a human expert to guide the beginners in the said domain .
expert becomes an integral part of collaborations . also , the two way contribution ( duplex ) motivated us to review and categorize the resource in the vicinity of vo .
the categorization we presented is also vigilant to depict the general pattern of resources in any domain .
vo is the right place for both technology and computing paradigm to merge and achieve the objectives . in the past two decades
collaborative computing has remained main concern of technology produced .
optimization of time and heterogeneous resources by building vo is the key point of today s research directions .
vision of a vo has evolved with the networking and distributed computing concepts .
research community recognizes vo with different names , e.g. collaboratories @xcite @xcite , e - science or e - research @xcite @xcite , distributed work groups or virtual teams @xcite @xcite , virtual environments , virtual enterprize @xcite and online communities @xcite @xcite .
initially , focus was to improve business by utilizing the ability to gather resources which are scattered across the dimensions of time , space and structure . with the advent of modern technology
, vo has encompassed almost all fields of life .
we can say that every human will be soon part of a vo .
vo s concepts need to be revisited with this evolution in general .
vos have been visioned from the business perspective in early 1990s .
pervasiveness of technology and improvements in computing paradigms has extended the domain of vo to cover all the areas where individuals and organizations meet to achieve some goal ( formal virtual organization ) or without any specific common objective , e.g. social networks ( informal vo ) . to the best of our knowledge ,
till now there are no standard procedures or patterns for how vo should be created and evolve to accommodate the changes in its integral parts or entities .
lack of standards for vo motivated us to provide a standard vision of e - collaboration incorporating both paradigm and technology shift as a reference architecture ( ra ) to achieve common objective(s ) in any domain .
our research efforts also introduced new concepts regarding resources and stakeholder of a vo . to provide a standard for vo , we consider the existing technologies and paradigms .
service oriented architecture ( soa ) , web 2.0 and web 3.0 are the underlying technological platform , and computing paradigms include utility computing and cloud computing . during our research process
we studied the existing infrastructures available for vo . utilizing electronic collaborations for achieving common goals
is a tradition rather a requirement .
distributed resources are gathered using an infrastructure and are exploited to obtain the said results . in it world such collaboration
is known as vo .
idea is to provide resources as a utility to the end user .
service - oriented infrastructures need to act dynamically to fulfil the demands from organizations and businesses .
we encountered the following addressable issues : * does existing electronic collaboration approaches follow a standard ? * can we define patterns without predefined standards ? * does existing infrastructures fulfil the requirements of participating entities ? *
are the existing infrastructures dynamic and adaptable to the rapidly updating it and business world ? * can we design a generic platform to integrate resources from multiple domains using essential and optional parts ?
our research aims to answer these questions .
vo s creation process lacks standards / patterns / methods @xcite .
we analyzed existing vos and the process of their creation through available documentation .
we found the following answers to the above questions : * currently , there exist no specific standards for building vo or e - collaboration . *
existing infrastructures are modified for specific domain needs and can not be generalized to all the domains . since
, existing technology is used without following any standard for creating a vo , it is hard to foresee incoming demands from the participants .
* we require a generic platform to integrate resources from a single domain or multiple domains . * defining a generic platform on the basis of everything - as - a - service ( xaas)@xcite concept is a solution .
definition of participating components as : * * essential parts * * optional parts we present the following solutions for these obstacles : * generalized patterns for building a vo . * defining components of a vo . * providing new definitions and examples of _ resources _ and _ stakeholder _ in different domains and justifying them in real world . * presenting a reference architecture for virtual organization ( ravo ) which can be applied as a starting point for any community ( belonging to a single or multiple domains ) to collaborate . the layout of the paper is as follows : in the next section [ sec : vo ] we introduce the notion of virtual organization . in section [ sec : creatingvo ]
we present the principal steps for building a virtual organization , which we map to the process of our blueprint design .
the generic reference architecture for virtual organizations , which we call ravo , is laid out in section [ sec : ravo ] . for justification of our approach a proof - of - concept implementation based on ravo is presented in section [ sec : n2sky ] .
finally the paper is closed with a conclusion of the findings .
a virtual organization is a non - physical communication model with the purpose is of a common goal .
it is built up from people and organizations with respect to geographical limits and nature .
additionally , a virtual organization provides typically a collection of logical and physical resources distributed across the globe . from a conceptual point of view a _ virtual organization _ resembles a detailed non - physical problems solving environment .
many definitions have been presented and many different terms arose , e.g. collaboratories @xcite , e - science or e - research @xcite , distributed work groups or virtual teams @xcite , virtual environments , and online communities @xcite .
a virtual organization can comprise a group of individuals whose members and resources may be dispersed geographically and institutionally , yet who function as a coherent unit through the use of a cyber infrastructure @xcite .
virtual organizations are typically enabled by and provide shared and , in most cases , real - time access to centralized or distributed resources , such as community specific tools , applications , data , instruments , and experimental operations , which was a key research areas of the authors @xcite in the past .
the different types of virtual organizations depend upon mode of operation , goals they achieve and life span for which they exists .
regardless of the objectives , virtual organizations possess some common characteristics .
virtual organizations provide distributed access across space and time .
structures and processes running a virtual organization are dynamic .
email , video conferencing , telepresence , awareness , social computing and group management tools are used to enable collaboration among the participants @xcite .
operational organizations are supported by simulation , database and analytical services . in daily life
we come across many virtual organizations in terms of social networks as well ( e.g. facebook , myspace ) .
it can be phrased that soon every human on this earth will considered to be a part of some virtual organization serving its purpose in the said organization .
generally , a virtual organization provides a global problem solving platform .
it is difficult to specify or restrict the domain for which they are serving .
some advantageous roles played by virtual organizations are facilitator of access ( birn @xcite , lead @xcite , nanohub @xcite ) , cabig@xcite , lhc computing grid , enhancer of problem solving processes ( teragrid @xcite ) and key to competitiveness ( geon @xcite ) .
virtual organizations have served in the field of earthquake engineering ( southern california earthquake center ( scec ) ) , cancer research ( cancer biomedical informatics grid ( cabig ) @xcite ) , climate research ( earth system grid @xcite ) , high - energy physics ( large hadron collider ) , and computer science .
other communities are now forming virtual organizations to study system - level science .
these virtual organizations are addressing problems that are too large and complex for any individual or institution to tackle alone .
it simply is not possible to assemble at a single location all of the expertise required to design a modern accelerator , to understand cancer , or to predict the likelihood of future earthquakes .
virtual organizations allow humanity to tackle previously intractable problems @xcite .
the creation of a virtual organization is time consuming and should be a well planned activity . in this section
we will discuss virtual organization and technology from different perspectives .
both aspects are required to support each other .
technology provides the basic infrastructure for a virtual organization to exist . a virtual organization in turn places demands on information technology and shapes the evolution of technology . for the last decade
the virtual organization is one of the most discussed collaboration environment ; but still no standards exist . from this discussions we assume a step wise approach which is helpful in the creation of virtual organization .
it can be separated in two phases which are detailed below .
the definition of a virtual organization starts with a series of questions , which are very critical in order to proceed .
these questions ( qx ) are listed in the following : * q1 : why to form a virtual organization ?
what are the reasons of an organization to create a virtual organization ?
* q2 : what is the motivation behind participation ? why should other persons , institutes , service providers , etc
. want to participate in a virtual organization ?
* q3 : what services are offered by a virtual organization ? * q4 : how are these services fared ? what is the type of the resources / business model ?
* q5 : who are the intended users ? who will eventually use and get benefited from this virtual organization ? * q6 : what is the life of ( membership of ) a virtual organization ?
are temporal alliance or permanent participation expected ?
based on these q&a activity it is necessary to identify the building blocks of a virtual organization .
gannon @xcite has identified main components of a virtual organization .
these components are * _ common interest . _ the reason to form a virtual organization , * _ users . _ the participants of a virtual organization , * _ tools and services .
_ this is a crucial part of a virtual organization , which maintains the overall working environment and saves the existing patterns to be reused in order to reduce time to solve similar problems .
a virtual organization requires a collection of shared analysis tools ( e.g. visualization tools and provenance tools ) .
tools can be integrated into specific virtual organization work flows and can be shared and reused .
they are used to collect data and publish results . * _ data . _
a virtual organization contains two types of data , generally categorized as meta data and operational data that is being operated by tools .
the component identification process provides the basic building blocks .
the designer of a virtual organization can decide what to be improved and further included in the design process .
also , each component can be given a unique definition by the designer in context of a virtual organization being created .
detailed information about creation , management and application area of virtual organizations is available in @xcite .
according to gerrit muller @xcite there are two simultaneous trends , * increasing complexity , scope and size of the system of interest , its context and the organizations creating system @xcite . *
increasing dynamics and integration : shorter time market , more interoperability , rapid changes and adaption in the field , in a highly competitive market , for example cost and performance pressure @xcite .
these trends form basis for our proposed ra as well .
vos are developed as distributed system at multiple locations , by multiple entities , consisting of multiple applications by multiple vendors , merging multiple domains for providing solutions to multiple problems .
ra comes in scene where the multiplicity reaches a critical mass triggering a need to facilitate product creation and life cycle support in this distributed open world @xcite .
we detail the ravo in the subsequent sections .
we define ravo as `` an open source template that does not only depict the architectural patterns and terminology , but also defines the boundaries where heterogeneous resources from different domains merge collaboratively into a common framework '' .
a ra has a life span and is dependent on the target architecture and possibly other ras . as guideline for our effort
we closely analyzed the ra presented by shaman ( european commission , ict-216736 ) , gerrit muller @xcite and nexof@xcite .
ravo provides * a common lexicon and taxonomy . * a common ( architectural ) vision .
* modularization and complementary context . * a layered approach(bottom - up ) .
a common vision facilitates the participating entities to work as a team to achieve their decided goals .
modularization helps to integrate different domains thereby decreasing the efforts and context information make the dynamic nature of the architecture consistent .
we aim for developing a ra which allows for new forms of it infrastructure coping with new collaborative processing paradigms , as grid computing and cloud computing .
thus we have to deliver an environment to allow for the new _ internet of services and things _ accommodating the novel service stack , as iaas , paas and saas .
architecture is classified into different layers according to the service each layer provides .
layered architecture is chosen because it helps to group different components ( logical and physical ) according to the degree of relatedness and required functionality .
ravo is based on spi model .
layered approach is used to achieve the goal of providing all the resources as a service .
layers are distributed into 3 broad categories , iaas , paas , saas figure [ fig : csvo ] presents the framework for vo using the spi model .
the layers are distributed into 3 broad categories , iaas , paas and saas thus resulting in xaas .
[ [ section ] ] in context of ravo , saas is composed of a service layer .
it contains domain specific applications ( dsa ) accessible by all users .
dsas are the combination of several user interfaces and business models found in the vo layer .
users , who only use the platform to solve their domain specific problems and do not contribute to the vo , find an entry point at this layer .
* _ service layer _ : it has open source , downloadable software , categorized in domains .
the service layer packages several services provided by the vo layer to be subscribable entities .
these entities include generic functionality to query information from the problem domain as well as the means to perform data mining on the compound data created or provided by the combination of the services .
[ [ section-1 ] ] in ravo two layers , namely vo layer and abstract layer , cover paas . * _ abstract layer _ : this layer is composed of essential tools which enable the whole framework to be exploited in a dynamic manner .
the set of tools consist of provenance , workflow , graphical tools and any other domain specific tools which are used to enhance the reuse of the resources for a diverse set of problem solving activity .
each tool provides its own functionality , its own user interface description @xcite , as well as an abstract api ( identical for each tool ) to access the resource in factory layer . *
_ vo layer _ : this layer is the entry point for user .
it provides the realization of the user interface description and defines a business model on top of the abstract layer to set usage cost according to usage statistics .
participating entities can agree on a usage model and build a cost trust for selling their resources . in context of vo ,
contributor / subject users ( who not only use the resources offered by a vo but also contribute to the vo ) are authorized to access the system on this layer .
all have access to the system on paas layer .
[ [ section-2 ] ] in ravo logical and physical resources are considered to be the part of iaas .
this part consists of two sub - layers in ravo : factory layer and infrastructure enabler layer .
only users with administrative rights have access to this layer . * _ factory layer _ :
belongs to the iaas category and contains resources for ravo .
resources are described as physical and logical resources .
physical resources comprise of hardware devices for storage and computation cycles in a distributed manner .
logical resources contain expert s knowledge that supports the problem solution activity thereby reducing time to reach the specified goal . * _ infrastructure enabler layer _ : allows access to the resources provided by the factory layer .
it consists of protocols , procedures and methods to contact the desired resources for a problem solving activity .
it acts as a glue or medium to reach the desired resources based on user request .
[ [ section-3 ] ] all layers are providing their functionality in a pure service - oriented manner so we can say that ravo is xaas .
vos is a broad category of distributed systems .
it is envisioned as a combined effort of multiple entities ( organizations , people , hw , sw ) for achieving a goal .
building ra for vo is effective in many ways .
ravo forms basis for vos belonging to any domain .
it improves the effectiveness by managing synergy , providing guidance for collaboration , generic framework , managing and sharing the architectural patterns .
interoperability is the most critical aspect of collaborative computing and vos main feature .
it determines the usability , performance and dependability of user level applications @xcite .
integration cost and time are also important factors in context of interoperability .
ravo supports interoperability by defining a negotiation model / trust for the participating entities thereby supporting the effective re - use of patterns .
many ra focuses the technical architecture only .
according to saf meeting conclusion , a ra should address @xcite , * technical architecture .
* business architecture .
* customer context .
ravo well addresses these three aspects .
it presents a technical architecture specifying the must participating modules , apis , protocols and platform to support vo .
ravo offers a business model which is open according to the participating entities conditions for resource sharing .
business model and customer context overlap .
ravo explicitly defines roles of participating entities as _ subject _ , _ consumer _ , _ producer _ and _ administrators_. elaboration of roles makes it easy to dynamically update the business model as an entity changes the role .
ravo supports feedback from the participating entities which is helpful in improving and maintaining the existing ra .
these concepts are already detailed in ravo section .
ra is a perceived image of existing technologies .
designing ra is a challenging job because it needs sufficient proof to justify its need in the said context .
ravo focuses on vos . to the best of our knowledge
, there is no standard pattern or framework which can be used to create a vo from scratch .
our vision is to provide the vo community a complete framework for identifying main components and abstract a life cycle to create vo from scratch .
it grasps knowledge from existing structures such as nexof @xcite and shamans ( european commission , ict-216736 ) .
guidelines are used to modify the requirements into an ra which supports creation , dynamic evolution and maintenance of a vo .
viewpoint is defined as a specification of the conventions for constructing and using a view . a pattern or template from which to develop individual views by establishing the purposes and audience for a view and the techniques for its creation and analysis @xcite .
view is a representation or description of the entire system from a single perspective .
stakeholder is the viewer , who perceives the system according to her role .
viewpoint has a name , stakeholders addressed by it and concerns to be addressed by the viewpoint , and the language , modeling techniques or analytical methods to be used in constructing a view based on the viewpoint @xcite . according to these definitions , viewpoints extracted from the concerns of the stakeholders are shown in figure [ fig : vp ] .
these viewpoints are detailed in the following sections .
stakeholders collaborate to form a vo .
all participants of a vo have an objective ( personal or organizational ) to achieve via this collaboration .
sub - viewpoints are , * domain definition : depends on the type of problem solution , target domain can be one or multiple .
thus , stakeholders can be from one or multiple domains .
* participation level : participation can be individual or at organizational . * duration : stakeholder remain part of the vo according to the membership duration agreed upon among the collaborative entities .
it can vary depending on the type of vo , partial or permanent ( either participation is required for a specific part or throughout ) and a business model in case of profitable organizations . * types of contribution : it is decided by the role assigned to a specific stakeholder in the context of a vo . this viewpoint formulates the boundaries of a vo and its participants .
all the participants must clearly put their requirement and goals while building collaboration .
these requirements should reflect any assumption made on the architecture and the respective requirements stemming from the assumptions .
once requirements are defined ( in form of a list , catalogue ) , vo has a vision to achieve and targets are set accordingly .
trust governance viewpoint is very important to any collaboration specially for vos .
keeping stakeholders and resources glued together to achieve a target is only achievable via strong trust .
following sub - viewpoints are defined in this context : * trust / policy formation : experts and planners from participating entities prepare an agreed upon policy / model / contract .
this policy defines the rules for participating and leaving vo , contributing and consuming resources , penalties for violation and measures to keep consistent and just to all the stakeholders .
* objective catalogue : this viewpoint provides the list of all the contracts and agreements in a documented form necessary for authentication , authorization and stakeholder management . * reviewing policy : due to dynamic nature of vo , the policies and contracts are reviewed to be in the accordance of change in requirements , technological updates , removal and entrance of participants .
* business perspective : this viewpoint is optional depending upon the type of vo .
profitable vo have a business model for metering , billing in addition to authentication and user management .
this view point provides details of participating components for realizing the vo on technological and system level .
it is further divided into 4 sub - viewpoints which are briefed here as , * data : this viewpoint aims to depict the types of data utilized in collaboration .
two broad identifications are found as _ meta data _ and _ operational data_. problem nature , domain and participating entities decide on the data source and security in collaborative efforts .
data and relationship between different components can be represented using table , mat , uml class diagrams , activity diagrams and component diagrams .
* applications and tools : this viewpoint describes the list of running applications and tools utilized in a problem solving activity .
this view can be further divided according to requirement .
roles of stakeholders also decide the access to different available tools and applications at multiple levels ( interface , infrastructure , platform and so on ) .
distribution and relationship among applications , tools and components can be shown using uml component diagram .
* resources : this viewpoint explains the list of resources ( table , list ) , their owners , availability , usage cost ( in case of profitable organization ) and access rights .
we have to sub - viewpoints : * * subject : an important viewpoint which defines stakeholder which consume and contribute to the resources simultaneously . * * enabler : this viewpoint details the stakholders which are related to deployment , configuration , monitoring and lifestyle management .
roles assigned in this viewpoint are developers , administrators , business providers , planners and experts . * log catalogs : this viewpoint keep track of activities which are carried out during problem solving activities .
dynamic collaboration environments need to this record for the feedback and improvement .
this viewpoint lists the best available technology currently deployed .
if new technology is employed which is not listed then it should be added to the list later .
it is very helpful keeping vo consistent with the upcoming demands from business and user requirements and advancement in new computing paradigms and methods .
platforms used for collaboration have remained in a constant up - gradation .
choice must be made on technology by giving weighage to qos , security , cost effective and timely solution to the end user .
an important sub - viewpoint of technological aspect is virtualization .
it provides the way to reuse hardware cost , respond dynamically and maximize resource utilization and easy relocation .
virtualization viewpoint deals with logical resources rather than physical resources .
all these viewpoints are shown in the diagram figure [ fig : vp ] these viewpoints can be represented using lists , tables , uml tools , and other requirement specification tools available .
they are also extendable and organizations can add any further categories according to their goals .
ravo is composed of multiple layers and each layer provides a set of components which are the building block of a vo .
selection of these building blocks is subjected to various aspects ( i.e. life span , nature ( dynamic or static ) , type , formal , informal and so on ) .
we define interfaces for these components by specifying parameters ( mandatory and optional ) , methods and necessary conditions for their executions .
vo needs to keep specific information in general , when created .
it possesses some characteristics , ( e.g. unique i d , date created , description about purpose domain etc ) .
it also requires to maintain information about participating organizations and individuals .
it is a must to maintain and update the information about the resource providers in a vo .
organization offering resources , time period for which resources are made available and access rights are potential characteristics .
[ [ query - interface . ] ] query interface .
+ + + + + + + + + + + + + + + + ravo proposes query interface as a mandatory component at service layer .
user is facilitated with remote or desktop access .
query interface enables user to search for their problem solution in knowledge base .
knowledge base contains history of problems solved previously . on successful query user
is provided with appropriate output . in case of no matching solution ,
query is processed and problem solutions is provided to user and knowledge base is updated .
query interface must provide login facility , identify the query type , check for existing solutions and must maintain a tolerable response time .
[ [ domain - specific - application - dsa . ] ] domain specific application ( dsa ) .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + dsa is a mandatory component dsa provide user with the ability to either download the applications and run on their own systems or use them at vo platform for problem solution .
the range of applications depends on the domain and type of vo .
stakeholder can share their applications paid or non - paid basis .
sharing of application can be conditional ( e.g. fully or partially paid in case of profitable organization ) .
information maintained about dsa must include how it is accessible ( online or offline ) , access rights and cost ( as defined in the contract / business model ) . [ [ data - mining - tools . ] ] data mining tools .
+ + + + + + + + + + + + + + + + + + data mining tools are an optional component of ravo .
they are a must for analytical and scientific research based vos .
interface for data mining tools include tool description , access rights and manul / help .
paas layer is composed of two layers , namely 3-vo layer and 2-abstract layer .
component specification is detailed below .
[ [ vo - trust . ] ] vo trust .
+ + + + + + + + + vo layer consists of two main mandatory components .
vo trust is the most important of all components .
it is formed by combining different modules and performs multiple tasks .
it is responsible for _ authentication _ and _ authorization _ of vo members .
authorization is done on the basis of _ roles _ defined in the _ contract / business model_. vo trust have a mandatory emphcontract which consists of policies to achieve the goals of vo . in profitable or partially profitable vos business model is also mandatory component of vo trust . in ravo business model
is optional and depends on the type of vo , however contract is mandatory .
access rights are defined in contract or business model .
different methods are available to define the access rights . organization models and access rights are comprehended in @xcite . according to
the authors access rights might be subjected to _ organizational _ and _ direct _ change @xcite .
all components of vo trust are synchronized to maintain the vo .
each component is assigned a specific task and output of one component provides input to the other component .
vo trust has a _ resource information _ component that acts like a _
registry_. it keeps necessary details about all the resources available in vo . [ [ user - interface . ] ] user interface .
+ + + + + + + + + + + + + + + user interface is a mandatory component of vo layer .
it provides access to the platform services offered by vo .
user is authenticated and authorized using login option . after authorization
, user can formulate different queries and perform actions .
these facilities are realized using a web portal .
[ [ workflow - tools . ] ] workflow tools .
+ + + + + + + + + + + + + + + abstract layer is a sub layer of paas layer .
it includes different components .
workflow tools is a mandatory component of this layer .
workflow management is a critical aspect of a vo in any domain .
it supports provenance management which plays vital role in monitoring and maintaining a vo .
workflow can be interpreted in different forms ( e.g. graphical , textual , source code ) .
interpretation mode is chosen on the level of audience a vo possess .
workflow tools keep track of all the processes active in vo .
process management can be included as a sub component of a workflow tools .
dynamic adaption of in - process workflow is an essential part of any workflow management system .
classification of approaches along their strength and limitations used for dynamic adaption in workflow systems are detailed in @xcite .
flexibility criteria in process management to handle the foreseen and unforeseen behaviors are categorized in @xcite .
workflow tools allow user to define workflows for a problem solving activity .
the participants responsible at each stage of this activity are notified and are responsible for delivering the promised results .
workflows are reusable and reduce redundancy and time for similar problems .
information maintained consist of workflow i d , type , status , access rights , how it interprets the results and process management .
workflows are used by provenance management to track the problem solving activity on user demand .
[ [ provenance - tools . ] ] provenance tools .
+ + + + + + + + + + + + + + + + + with the advent of financial computing systems , as well as of data - intensive scientific collaborations , the source of data items , and the computations performed during the incident processing workflows have gained increasing importance @xcite .
provenance of a resource is a record that describes entities and processes involved in producing and delivering or otherwise influencing that resource . in a vo ,
provenance forms a critical foundation for enabling trust , reproduction and autentication .
provenance assertions are a form of contextual metadata and can themselves become important records with their own provenance .
provenance tools are mandatory and included in abstract layer of ravo .
provenance management is dependent on authorization , query management and workflow management .
[ [ graphical - interface . ] ] graphical interface .
+ + + + + + + + + + + + + + + + + + + + graphical interface is a mandatory component of abstract layer .
it facilitates users to perform different task in vo web portal .
it provides an understandable interface to interact with the vo .
[ [ resource - management . ] ] resource management .
+ + + + + + + + + + + + + + + + + + + + resource management is a mandatory component of abstract layer .
it provides a mechanism to select and aggregate resources for a problem solving activity . depending upon the underlying technology ,
vo developers can deploy different resource management tools .
necessary information maintained depends on the resource type and interest of participating entities .
basic information includes resource s unique identification , categorization as logical or physical , owner information , access rights and costs etc .
iaas layer is composed of infrastructure enabler layer and factory layer . this layer from the fabric of ravo .
all the resources are avaialble in factory layer and are exploited through infrastructure enabler layer . [
[ infrastructure - enabler . ] ] infrastructure enabler .
+ + + + + + + + + + + + + + + + + + + + + + + this module is depending on the underlying technology .
qos , service level agreement ( sla ) , security , fault tolerance and disaster management are most important issues specifically in clouds @xcite .
these aspects have to be implemented on the bases of terms and conditions presented by participating entities .
financial aspect is another limitation for the implementation of these modules .
any other desired aspects can be added to extend the infrastructure enabler layer .
components are dependent on the decision of the developers .
ravo identifies least basic and gives developers an open end to use them as mandatory or optional in their target vo .
[ [ resource - catalogue . ] ] resource catalogue .
+ + + + + + + + + + + + + + + + + + + this module is part of factory layer but not explicitly shown in ravo .
it acts like a database for the resources .
vo developers can include it at any layer according to their needs .
ravo keeps it at the factory layer as a mandatory component .
it contains information about resource management .
[ [ expert . ] ] expert .
+ + + + + + + expert represents the logical resource in ravo factory layer .
an expert plays important role in problem solving activity .
expert can be contacted online during the problem solving process or she can be accessed offline .
vo must maintain detailed information about expert so that this feature can be fully exploited .
[ [ data - service . ] ] data service .
+ + + + + + + + + + + + + data services is a mandatory component of factory layer .
it represents the physical resource in ravo .
data stores are important scientific and research based vos .
[ [ computational - services . ] ] computational services .
+ + + + + + + + + + + + + + + + + + + + + + + computational services are mandatory component of factory layer .
they also form the physical resources offered by a vo .
as proof - of - concept of our approach we used ravo as a design blueprint for implementing a cloud based vo for neural network research , namely , n2sky .
we based the development of n2sky on the blueprint provided by ravo and produced a concrete instance out of our proposed standard @xcite .
this section compares n2sky with ravo to reveal the process of creation of n2sky .
the comparison justifies and proves how ravo supported different development phases of n2sky .
we explain n2sky as an instantiation of ravo but with concrete components .
we divide this comparison in 3 levels .
first , _ requirement analysis phase _ that defined boundaries of n2sky .
second , _ component identification phase _ which made it easy to identify the components of n2sky and also choose between optional and mandatory components .
third , _ implementation phase _ that reveals how technology independence , xaas and layered distribution of components made it helpful to implement the system .
the stakeholders envisioned in ravo are also implemented as part of n2sky . in the previous section we detailed a series of questions which must be answered by the responsible authorities for creating a vo .
n2sky utilizes this pattern for defining the requirements boundary of the system .
these questions are answered in detail in an interview by engaged software engineering experts , whic are presented in section [ sec : evaluation ] .
n2sky is a layered architecture instantiated from ravo .
the n2sky is shown in figure [ fig : n2s ] .
n2sky is also presented as an xaas , based on cloud spi model .
it consists of 3 layers , namely saas , paas and iaas .
these layers have sublayers similar to ravo .
each layer has some components which are either mandatory or optional depending upon their participation in vo .
figure [ fig : csvo ] shows ravo framework .
a detailed , tabular comparison of ravo and n2sky components is given in figure [ fig : comp ] .
section [ sec : ravo ] presented interface specification for components . here , we analyze how these interface specifications were used in n2sky .
we compare the underlying framework ravo with its instantiation as n2sky , in a top - down fashion .
we start with saas layer .
saas layer of ravo consists of optional and mandatory components .
choice of components and decision on their status ( mandatory and optional ) is open for the developers .
the inclusion of components is dependent on the requirement definition by the stakeholders .
saas layer has one layer , named service layer .
here tables are included for the sake of comparison . * query interface : ravo proposes query interface as a mandatory component at service layer . in n2sky , query interface is also included as a mandatory component . *
domain specific application ( dsa ) : dsa is a mandatory component .
n2sky has a simulation service but at neural network layer ( sub layer of paas ) .
n2sky includes dsa as nn specific applications .
n2sky is planned to include nn specific applications .
the simulation service provides the creation , training and simulation of neural objects which in turn are instances of nn paradigms .
currently , simulation services are provided at nn layer of n2sky . * data mining tools : data mining tools are an optional component of ravo .
n2sky has not included this option .
n2sky also has one layer , named service layer ( similar to ravo ) .
extended components included at service layer in n2sky are : * web portal : n2sky web portal is a mandatory component .
* smaprtphone app .
* hosted ui .
paas layer is composed of two layers , namely vo layer and 2-abstract layer .
component specification is detailed below . in n2sky paas
consists of 3-neural network layer and 2-abstract layer .
[ [ vo - layer - comparison - with-3-neural - network - layer . ] ] vo layer comparison with 3-neural network layer .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + in ravo vo layer has the following components : * vo trust : mandatory component of vo , which is responsible for enabling resources , defining policies to achieve a goal .
it has several components and is extendable according to the need and requirement of stakeholders .
n2sky has distributed trust component in to different modules . in n2sky ,
neural network layer has a management service component to serve the purpose .
other components are available at abstract layer namely , business model with slas and accounting . * user interface : user interface is a mandatory component for solving problem utilizing vo paas utility .
it provides an interface to interact with the vo .
n2sky also realizes this component as a part of web portal .
extended component of n2sky : * hosted component : provides and interface for components hosting platform .
* simulation service : already described in service layer comparison .
it is a mandatory component that is part of neural network layer of n2sky .
[ [ sub : alcompared ] ] abstract layer comparison .
+ + + + + + + + + + + + + + + + + + + + + + + + + + ravo and n2sky both have this sub layer named 2-abstract layer
. components of these layer in ravo and n2sky are compared . * resource management : resource management is a mandatory component of abstract layer .
it provides a mechanism to select and aggregate resources for a problem solving activity . depending upon the underlying technology ,
vo developers can deploy different resource management tools . in n2sky resource management
is achieved via mandatory registry component .
* workflow tools : n2sky also has a workflow system under development . * provenance tools : provenance tools are proposed in ravo but they are not included in n2sky . * graphical interface : a mandatory components which facilitates interaction with vo easier and helps user to get results in an understandable format
. it also assists user in formulating queries and browsing in vo environment . in n2sky graphical interface
is implemented as a web portal described earlier .
extended components supporting vo trust ( as proposed in ravo ) functionality :
* controlling and accounting : this component along with sla component serves as a business model . in ravo business model
is optional .
* usermanagement .
* access control .
* annotation service .
* knowledge management : it refers to expert s knowledge of ravo defined at factory level .
* component hosting platform .
iaas layer is composed of 1-infrastructure enabler layer and 0-factory layer . this layer froms the fabric of ravo .
all the resources are available in factory layer and are exploited through infrastructure enabler layer .
infrastructure enabler layer in ravo brings an open choice for the developers for underlying technology .
qos , service level agreement ( sla ) , security , fault tolerance and disaster management are aspects to be considered in particular .
further extension can be done by developers .
n2sky also have an infrastructure enabler layer .
it contains following components .
* data archive : implemented as a mandatory component of n2sky .
* component replication service .
factory level of ravo is also instantiated in n2sky .
it has following components in ravo * resource catalogue : resource catalogue module is an extension of resource management component .
it is a mandatory components .
it keeps information about resources which is of interest to vo . in n2sky
this task is achieved by registry component . *
computational services : ravo offers computational services as a mandatory component . in n2sky this component is realized by component replication service .
it is a mandatory component which act as n2sky paradigm archive service . * data services : this component of ravo
is realized by n2sky as a part of infrastructure enabler layer . *
expert s knowledge : n2sky implements this component of ravo as knowledge management as a subcomponent of abstract layer . based on the experiences drawn from development of n2sky
we could derive the following list of findings : * ravo provides strong theoretical grounds to clear the vision of vo developers and participants before they start building a community .
* requirement analysis and component identification phases enable developers to list mandatory and optional components .
the purpose is twofold .
first , must parts of the vo are confirmed .
second , optional parts leave room for future requirements and upgrades .
* ravo framework is flexible and generic .
components at different layers are moved or integrated with other parts as it eases the developing process .
* ravo is technology independent and it gives freedom of choosing any suitable tools and programming languages . *
ravo emphasis on providing graphical interface to ease the end user so that they can communicate and formulate their queries easily .
the interface should not be complicated that only professionals can interact .
* stakeholders and their roles are important to understand . pattern developed for ravo are used here extensively to design a business model for n2sky . for further justification of our approach also an informal evaluation was conducted .
hereby two software developers ( which we name a and b ) , who were involved in the final development of voci , were interviewed .
interview partner a is a senior researcher at our research group .
his expertise include soas ( including work on mobile devices ) , with a specific focus on process aware information systems .
he developed a light - weight modular process engine to fully support external monitoring and intervention .
he further published in the field of restful service description , composition and evolution .
he was interviewed regarding ravo as a staring point for developers in nn domain .
interview partner b was a master student at university of vienna , chose ravo as the template to develop a cloud - based virtual organization for nns called n2sky .
erwin mann has experience in implementing service - oriented architectures ( soas ) , service orchestration , to create workflows and in porting such systems to cloud - based environment .
n2sky brings together both nn paradigm developers and users who deal with problems that are beyond conventional computing possibilities .
n2sky provides a standardized description language for describing neural objects ( instances of neural paradigms ) called vinnsl .
furthermore n2sky provides a business model for researchers and students but also for any interested customer .
s core process is the simulation service including creation , training and evaluating of neural objects in a distributed manner in the cloud .
both researcher gave their opinion after critical analysis of ravo .
researcher a s abstraction of ravo in terms of `` q&a '' is helpful for the developers of vo in any domain .
researcher b has applied ravo and developing an instance in the domain of nn .
we deduce the following statements from this evaluation .
* ravo best fits needs of community for developing a vo from scratch . * ravo supports evolution of existing systems in to a vo .
n2sky is an example of such evolution .
* ravo is presented in a layered fashion , with a choice of mandatory and optional components .
layered approach make it easy to distribute the components in different layers and also developers are not bound to choose the exact distribution .
ravo is a flexible and extendable framework .
developer can change the components and move them to any desired layer .
for example in n2sky , components have been moved to different layer as compared to ravo . * ravo is not technology dependent .
both the researchers described their alternative choices which establishes the technological independence of ravo . * categorization of resources into logical and physical is a new dimension for vo developers .
inclusion of human expertise as a resource supports the demanding nature of problem solving ability , thereby increasing the level of trust in users .
* ravo presented a new concept of stakeholder , _
subject_. a unique idea of how a stakeholder can become a resource in a vo .
being consumer and producer at the same time is difficult to implement .
ravo make it easier by introducing the stakeholder categorization .
* ravo foresees a business model which is introduced in n2sky as a mandatory component .
stakeholder s roles are integrated in business model to set the usage and cost policy . the interviews and the derived results are detailed in @xcite .
the strong response from research community , motivated us to design a reference architecture for virtual organization ( ravo ) .
reference architectures are system specific and provide a level of detail required to translate the required capabilities as derived from missions , operational concepts , and operational architecture views , into projects within a capable package , which will increase system quality and decrease system development costs .
we developed a service - stack pattern to start a virtual organization from scratch and also presented a design blueprint for collaborative environments .
emphasis is on flexible and simple interface for interaction from the user perspective .
it supports addition of tools and application to the virtual organization environment .
we give guidelines for building a domain specific virtual organization for computational intelligence community and can extend it according to our requirements .
t. weishupl , f. donno , e. schikuta , h. stockinger , and h. wanek , `` business in the grid : big project , '' in _ grid economics & business models ( gecon 2005 ) of global grid forum _ , vol . 13 , ( seoul , korea ) , ggf , 2005 . t. weishupl and e. schikuta , `` towards the merger of grid and economy , '' in _ international workshop on agents and autonomic computing and grid enabled virtual organizations ( aac - gevo04 ) at the 3rd international conference on grid and cooperative computing ( gcc04 ) _ , vol . 3252/2004 of _ lecture notes in computer science _ , ( wuhan , china ) , p. 563570
, springer berlin / heidelberg , 2004 .
t. weishupl , c. witzany , and e. schikuta , `` gset : trust management and secure accounting for business in the grid , '' in _
6th ieee international symposium on cluster computing and the grid ( ccgrid06 ) _ , ( singapore ) , p. 349356
, ieee computer society , 2006 .
w. mach , b. pittl , and e. schikuta , `` a business rules driven framework for consumer provider contracting of web services , '' in _
15th international conference on information integration and web - based applications & services ( iiwas2013 ) _ , december 2013 .
w. mach and e. schikuta , `` a generic negotiation and re - negotiation framework for consumer - provider contracting of web services , '' in _
14th international conference on information integration and web - based applications & services ( iiwas2012 ) _ , ( bali , indonesia ) , acm , dec . 2012 .
c. kesselman , i. foster , j. cummings , k. a. lawrence , and t. finholt , `` beyond being there : a blueprint for advancing the design , development , and evaluation of virtual organizations . , '' tech .
may 2008 .
e. schikuta and t. frle , `` vipios islands : utilizing i / o resources on distributed clusters , '' in _
15th international conference on parallel and distributed computing systems ( pdcs02 ) _ , ( louisville , ky , usa ) , isca , 2002 .
e. schikuta , t. frle , and h. wanek , `` vipios : the vienna parallel input / output system , '' in _ 4th international euro - par conference _ ,
1470/1998 of _ lecture notes in computer science _ , ( southampton , uk ) , p. 953958
, springer berlin / heidelberg , 1998 .
p. brezany , t. mck , and e. schikuta , `` a software architecture for massively parallel input - output , '' in _ third international workshop applied parallel computing industrial computation and optimization ( para96 ) _ ( j. wasniewski , j. dongarra , k. madsen , and d. olesen , eds . ) , vol .
1184 of _ lecture notes in computer science _ , ( lyngby , denmark ) , p. 8596
, springer berlin / heidelberg , 1996 .
10.1007/3 - 540 - 62095 - 8_10 .
g. klimeck , m. mclennan , m. mannino , m. korkusinski , c. heitzinger , r. kennell , and s. clark , `` nemo 3-d and nanohub : bridging research and education , '' in _ nanotechnology , 2006 .
ieee - nano 2006 .
sixth ieee conference on _ , vol . 2 ,
pp . 441444 , ieee , 2006 .
c. catlett , w. allcock , p. andrews , r. aydt , r. bair , n. balac , b. banister , t. barker , m. bartelt , p. beckman , _
_ , `` teragrid : analysis of organization , system architecture , and middleware enabling new types of applications , '' _ hpc and grids in action , amsterdam _ , 2007 .
u. nambiar , b. ludaescher , k. lin , and c. baru , `` the geon portal : accelerating knowledge discovery in the geosciences , '' in _ proceedings of the 8th annual acm international workshop on web information and data management _ , pp . 8390 , acm , 2006 .
i. foster , `` service - oriented science : scaling escience impact , '' in _
iat 06 : proceedings of the ieee / wic / acm international conference on intelligent agent technology _ , ( washington , dc , usa ) , pp . 910 , ieee computer society , 2006 .
a. a. huqqani , p. beran , x. li , and e. s. .
, `` n2cloud : cloud based neural network simulation application , '' in _ in proceedings of the international joint conference on neural networks 2010 ( wcci 2010 ) , barcelona , spain _ , 2010 . s. rinderle - ma and m. reichert , `` managing the life cycle of access rules in ceosis , '' in _
enterprise distributed object computing conference , 2008 .
12th international ieee _ , pp . 257266 , ieee , 2008 .
r. hasan , r. sion , and m. winslett , `` introducing secure provenance : problems and challenges , '' in _ proceedings of the 2007 acm workshop on storage security and survivability _ , storagess 07 , ( new york , ny , usa ) , pp . 1318 , acm , 2007 .
i. u. haq , e. schikuta , i. brandic , a. paschke , and h. boley , `` sla validation of service value chains , '' in _
9th international conference on grid and cloud computing ( gcc10 ) _ , ( nanjing , jiangsu , china ) , p. 308313
, ieee computer society , 2010 .
i. u. haq , i. brandic , and e. schikuta , `` sla validation in layered cloud infrastructures , '' in _ economics of grids , clouds , systems , and services , 7th international workshop , gecon10 _ , vol .
6296 of _ lecture notes in computer science _ , ( ischia , italy ) , p. 153164
, springer berlin / heidelberg , 2010 .
i. u. haq , a. a. huqqani , and e. schikuta , `` a conceptual model for aggregation and validation of slas in business value networks , '' in _ 3rd international conference on adaptive business information systems ( abis09 ) _ , ( leipzig , germany ) , mar .
2009 . i. u. u. haq , a. paschke , e. schikuta , and h. boley , `` rule - based workflow validation of hierarchical service level agreements , '' in _ workshops at the grid and pervasive computing conference ( gpc09 ) _ , p. 96103
, ieee computer society , 2009 .
i. u. haq , a. huqqani , and e. schikuta , `` aggregating hierarchical service level agreements in business value networks , '' in _
7th international conference on business process management ( bpm09 ) _ , ( ulm , germany ) , p. 176192
, springer - verlag , 2009 . | `` united we stand , divided we fall '' is a well known saying .
we are living in the era of virtual collaborations .
advancement on conceptual and technological level has enhanced the way people communicate . everything - as - a - service once a dream , now becoming a reality .
problem nature has also been changed over the time . today , e
- collaborations are applied to all the domains possible .
extensive data and computing resources are in need and assistance from human experts is also becoming essential .
this puts a great responsibility on information technology ( it ) researchers and developers to provide generic platforms where user can easily communicate and solve their problems . to realize this concept , distributed computing
has offered many paradigms , e.g. cluster , grid , cloud computing . virtual organization ( vo )
is a logical orchestration of globally dispersed resources to achieve common goals .
existing paradigms and technology are used to form virtual organization , but lack of standards remained a critical issue for last two decades .
our research endeavor focuses on developing a design blueprint for virtual organization building process .
the proposed standardization process is a two phase activity .
first phase provides requirement analysis and the second phase presents a reference architecture for virtual organization ( ravo ) .
this form of standardization is chosen to accommodate both technological and paradigm shift .
we categorize our efforts in two parts .
first part consists of a pattern to identify the requirements and components of a virtual organization .
second part details a generic framework based on the concept of everything - as - a - service . |
future neutrino experiments will operate in the few - gev neutrino energy range , where quasielastic scattering ( ccqe ) and pion production channels are dominant interactions .
the charged current pion production interactions account for the biggest background for the neutrino oscillation disappearance measurements .
recently , three experiments ( k2k @xcite , miniboone @xcite and sciboone @xcite ) have presented results for cc pion production cross section in the few - gev neutrino energy region on nuclear targets . with high statistic samples it is possible to determine neutrino cross sections in this region with better precision than data from previous experiments@xcite . here , we present the ratio of cross sections of the charged current single pion production ( @xmath0 ) and quasielastic scattering ( @xmath1 ) as obtained from the miniboone experiment . in this measurement
the largest sources of uncertainty coming from the neutrino flux determination largely cancels out .
the booster neutrino beam at fermilab provides a neutrino source which is particularly well - suited to make this measurement ; about 40% of @xmath2 neutrino interactions in miniboone are expected to be ccqe and 24% @xmath3 .
the beam itself is composed of 93.6% @xmath2 with a mean energy of about 800 mev and 5.9% ( 0.5% ) @xmath4 ( @xmath5 ) contamination @xcite .
the neutrinos are detected in the miniboone detector @xcite , a 12.2 m diameter spherical tank filled with 818 tons of undoped mineral oil located 541 m downstream of the beryllium target . at the energies relevant to this analysis
, the products of the interactions produce primarily cerenkov light with a small fraction of scintillation light @xcite .
the light is detected by 1280 8-inch photomultiplier tubes ( pmts ) which line the miniboone inner tank .
this inner tank region is optically isolated from a surrounding veto region , instrumented with 240 pmts , that serves to reject incoming cosmic rays and partially contained neutrino interactions .
neutrino interactions within the detector are simulated with the v3 nuance event generator @xcite .
ccqe interactions on carbon are generated using the relativistic fermi gas model @xcite tuned to better describe the observed distribution of @xmath6 interactions in miniboone @xcite .
resonant @xmath7 events are simulated using the rein and sehgal ( r - s ) model @xcite , as implemented in nuance with an axial mass @xmath8 .
the angular distribution of the decaying pions in the center of mass of the recoiling resonance follows the helicity amplitudes of @xcite . in miniboone ,
@xmath9 of @xmath7 production is predicted to occur via the @xmath10 resonance , but 17 higher mass resonances and their interferences , as well as a nonresonant background that accounts for roughly 6% of @xmath7 events , are also included in the model .
coherently produced @xmath7 events are generated using the r - s model @xcite with the r - s absorptive factor replaced by nuance s pion absorption model and the overall cross section rescaled to reproduce miniboone s recent measurement of neutral current coherent @xmath11 production @xcite .
coherent @xmath12 production is predicted to compose less than 6% of the miniboone @xmath7 sample due to the small coherent cross section @xcite and the dominance of the @xmath13 resonance .
a geant3-based detector model simulates the response of the detector to particles produced in these neutrino interactions .
the @xmath6 and @xmath7 reconstruction requires a detailed model of light production and propagation in the tank to predict the charge distribution for a given vertex and muon angle .
the muon vertex , track angle , and energy , are found with a maximal likelihood fit , with the energy being determined from the total tank charge @xcite . in the case of @xmath7 events , only one track is reconstructed and assumed to be a @xmath14 .
the neutrino energy for both samples is reconstructed from the observed muon kinematics , treating the interaction as a 2-body collision and assuming that the target nucleon is at rest inside the nucleus : @xmath15 here @xmath16 is the mass of the proton , @xmath17 is the mass of the muon , @xmath18 is the mass of the neutron in @xmath6 events and of the @xmath19 in @xmath7 , @xmath20 is the reconstructed angle of the muon with respect to the beam axis ( in the lab frame ) , and @xmath21 is the reconstructed muon energy .
the cross - section for a neutrino interaction is defined as : @xmath22 where n is the true number of events of type x(@xmath7 or @xmath6 ) , @xmath23 is the number of nuclear targets in the detector , and @xmath24 is the flux of incident neutrinos .
because the incident neutrino flux and the number of nuclear targets are fixed properties of the beam and the detector ( i.e. independent of the process we are considering ) , @xmath23 and @xmath25 must be the same for both processes .
therefore we can write @xmath26 with all the correction terms put together , the cross section ratio in each energy bin @xmath27 is : @xmath28 where the subscript @xmath27 runs over bins in true neutrino energy , subscript @xmath29 indexes bins in reconstructed neutrino energy , @xmath30 denotes the number of events passing cuts for @xmath31 , @xmath6 , @xmath32 denotes the signal fraction , @xmath33 denotes the cut efficiency , and @xmath34 is a neutrino energy unsmearing matrix that acts on a reconstructed distribution to return the true distribution .
to map reconstructed to true energy , we form a migration matrix @xmath35 representing the number of mc events in bin @xmath27 of reconstructed energy and bin @xmath29 of true energy .
we then normalize each reconstructed energy bin to unity to obtain an unsmearing matrix .
this is equivalent to a bayesian approach discussed in @xcite ; it differs from the standard matrix inversion method in that the resulting unsmearing matrix is biased by the mc distribution used to generate it .
we account for this in our uncertainties by including a variation in the mc distribution used to generate the matrix . because we have good data / mc agreement
, this effect is small .
the advantage of this method is that it avoids the problems of numerical instability and the magnification of statistical errors which occur in matrix inversion .
this unsmearing procedure also proved insensitive to variations in neutrino energy reconstruction , confirming that it performs as intended .
for our primary measurement , we define @xmath7 signal as events with exactly one @xmath14 and one @xmath12 escaping the struck nucleus ( which we call @xmath7-like events ) and ccqe signal as those with exactly one @xmath14 and no pions ( @xmath6-like events ) .
both event classes may include any number of protons or neutrons , but no other types of hadrons .
the observed cross section ratio is then defined as the ratio of @xmath7-like to @xmath6-like events and thus has not been corrected for re - interactions in the struck nucleus .
the systematic uncertainties on the cross section ratio arise from five main sources : the neutrino flux ( which largely cancels in the ratio ) , the neutrino interaction cross sections ( which affect the background predictions ) , the target nucleon momentum distribution ( which accounts for the model dependence of our unfolded neutrino energy ) , hadron re - interactions in the detector , and the detector simulation ( which describes light propagation in the oil ) . in the region of highest statistics ( about 1 gev ) , there is roughly an 8% fractional error on the ratio resulting from hadron re - scattering in the detector , 6% from neutrino cross sections , 4% from the detector simulation , 2% from the nucleon momentum distribution , 2% from the neutrino flux , and 2% from the statistics of the two samples . the cross section ratio reported by all prior experimental measurements @xcite has been one in which the effects of final state interactions ( fsi ) in the target nucleus have been removed using mc or deuterium was used as a target . solely for the purpose of comparison ,
we now extract a similarly corrected value .
the fsi - corrected ratio is defined as the ratio of @xmath7 to @xmath6 events at the initial vertex and before any hadronic re - interactions .
thus , the signal fractions and cut efficiencies for the fsi - corrected ratio include corrections for intra - nuclear hadron re - scattering based on the mc s model for nuclear effects @xcite .
the measurement proceeds exactly as for the observed ratio , except that now we define @xmath7 and @xmath6 , rather than @xmath7-like and @xmath6-like , events as signal for the respective samples . here
we limit our comparison to those experiments which reported both @xmath6 and @xmath7 cross sections , using the same energy bins for each of these interactions , so as to facilitate comparison with our measured @xmath36 ratio .
our result agrees with both anl , which used a deuterium target , and k2k , which used @xmath37 . in order to make this comparison ,
the miniboone and k2k results have been re - scaled to an isoscalar target . to perform this correction
, we rescale the ratio by a factor of @xmath38 , where @xmath39 is the ratio of neutrons to protons in the target and @xmath40 is the fraction of @xmath12 production that is predicted ( by mc ) to occur on protons .
the resulting scaling factor is 0.80 for miniboone ; for k2k we use the factor of 0.89 provided in @xcite .
the results have not been corrected for their differing nuclear targets nor for the application of explicit invariant mass requirements ( although the latter are similar ) .
anl used an explicit cut on invariant mass w < 1.4 gev @xcite .
while no invariant mass cut is used in this analysis , the miniboone spectrum is such that @xmath7 events occur only in the region @xmath41 ; similarly , k2k s measurement covers the region w < 2 gev @xcite .
the dominant reason for the difference between the ratios presented in figure 1 is intra - nuclear pion absorption in @xmath7 events , which cause these events to look ccqe - like . as a result of @xmath12 absorption ,
a significant number of @xmath7 events migrate from @xmath3 sample to @xmath6-like one .
thus , the fsi - corrected ratio , is 15% to 30% higher than the observed ratio in our energy range . in summary
, miniboone has measured the ratio of @xmath7-like to @xmath6-like events for neutrinos with energy @xmath42 incident on @xmath43 .
this is the first time such a ratio has been reported .
additionally , the ratio of the @xmath7 and @xmath6 cross sections at the vertex has been extracted using mc to remove the effects of final state interactions , in order to facilitate comparison with previous experimental measurements .
the results are summarized in table i. the measured ratios agree with prediction @xcite and previous data @xcite . | using high statistics samples of charged current interactions , miniboone reports a model independent measurement of the single charged pion production to quasi - elastic cross section ratio on mineral oil without corrections for pion re - interactions in the target nucleus @xcite .
the result is provided as a function of neutrino energy in the range 0.4 gev < e < 2.4 gev with 11% precision in the region of highest statistics .
address = department of physics and astronomy + louisiana state university + baton rouge , la 70803 |
suppose we wish to model a finite system in which objects have various states and update their states in discrete time steps .
moreover , assume that the state to which an object updates depends only on the current state of that object along with the states of other nearby or connected objects .
we can capture such a system s behavior with a _ graph dynamical system_. a graph dynamical system contains a graph representing the connections between objects , a set of states that the objects can adopt , a collection of functions that model how each individual object updates its state in reaction to its neighbor s states , and a rule determining the scheme by which the objects update their states . in a series of papers published between 1999 and 2001 , barrett , mortveit , and
reidys introduced the notion of a sequential dynamical system ( sds ) , a graph dynamical system in which vertices update their states sequentially @xcite .
we draw most of our terminology and background information concerning sds from @xcite , a valuable reference for anyone interested in exploring this field . if @xmath14 is a vertex of a graph @xmath0 , we let @xmath15 denote the degree of @xmath14 .
we will often work in the finite field @xmath16 .
in doing so , we will convene to let @xmath17 for any @xmath18 . for any vector @xmath19 , we let @xmath20 .
furthermore , @xmath21 will denote the identity permutation @xmath22 ( the length @xmath23 of the permutation @xmath21 will always be clear from context ) .
an sds is built from the following parts : * an undirected simple graph @xmath0 with vertices @xmath1 . * a set of states @xmath4 .
we will typically use the set of states @xmath24 . * a collection of _ vertex functions
_ @xmath2 .
each vertex @xmath25 of @xmath0 is endowed with its own vertex function @xmath26 . *
a permutation @xmath3 .
the permutation @xmath5 is known as the _ update order_. let @xmath27 denote the state of a vertex @xmath14 .
suppose a vertex @xmath25 has neighbors @xmath28 , where @xmath29 .
we let @xmath30 for example , if the vertex @xmath31 has neighbors @xmath32 , @xmath33 , and @xmath34 , we let @xmath35 .
the vector @xmath36 , which lists all of the states of the vertices of @xmath0 in the order corresponding to the order of the vertex indices , is known as the _ system state _ of the sds .
note that if @xmath0 is a complete graph and @xmath25 is any vertex of @xmath0 , then @xmath37 is equal to the system state of the sds . from each vertex function @xmath38
, we can define the _ local update function _
@xmath39 by @xmath40 combining these local update functions with the update order + @xmath41 ( we have written the permutation @xmath5 as a word ) , we obtain the sds - map @xmath42 given by @xmath43 we will find it useful to introduce an intermediate " sds - map @xmath44 for each @xmath45 , which we define by @xmath46 thus , @xmath47 .
we convene to let @xmath48 denote the identity map from @xmath49 to @xmath49 .
the vector @xmath50 represents the system state of the sds obtained by starting with a system state @xmath51 and updating only the first @xmath52 vertices in the update order @xmath5 .
once the system updates all @xmath23 vertices ( known as a system update ) , the new system state is @xmath53 . given any sds on a graph @xmath0 with vertex functions @xmath2 and update order @xmath5 , we denote its sds - map @xmath54 by the triple @xmath55 $ ] .
if all of the vertex functions @xmath38 are equal to the same function @xmath7 , we will simply write @xmath8 $ ] for the corresponding sds - map . [ example1 ] consider the graph @xmath0 shown in block a of figure [ fig1 ] .
we define an sds over @xmath0 using the update order @xmath56 and the vertex functions @xmath38 given by @xmath57 @xmath58 @xmath59 and @xmath60 the initial system state of this sds is @xmath61 , as shown by the blue labels in block a of figure [ fig1 ] . because @xmath62 , we first update the vertex @xmath63 using the vertex function @xmath64 .
we have @xmath65 so the vertex @xmath63 updates to the new state @xmath66 .
this intermediate update is shown in block b of figure [ fig1 ] .
another way to understand the transition from block a to block b in the figure is to see that we have changed the system state of the sds by applying the local update function @xmath67 .
indeed , @xmath68 . in a similar fashion ,
we next update vertices @xmath33 , @xmath32 , and @xmath31 . letting @xmath69 $ ] , we find that @xmath70 , as shown in block e of the figure . in other words , through a sequence of local updates , the system update transformed the system state @xmath61 into the new system state @xmath71 .
the sds - map @xmath54 tells us how the states of the vertices of the graph @xmath0 change when we update the graph in a sequential manner .
a useful tool for visualizing how @xmath54 acts on the system s states is the _ phase space _ of the sds .
the phase space , denoted @xmath72 , is the directed graph with vertex set @xmath73 and edge set @xmath74 in other words , we draw a directed edge from @xmath51 to @xmath53 for each @xmath75 . as an example , the phase space of the sds given in example [ example1 ] is shown in figure [ fig2 ] .
notice that the phase space shown in figure [ fig2 ] has a single @xmath10-cycle ( formed from the vertices @xmath76 and @xmath77 ) . in general
, a phase space of an sds can have many cycles of various lengths .
this leads us to some interesting questions .
for example , when is it possible to have a phase space composed entirely of @xmath10-cycles ? can we find an upper bound on the number of @xmath10-cycles that can appear in the phase spaces of certain sds ?
if we can show , for instance , that certain sds defined over a graph with @xmath23 vertices can not have phase spaces consisting entirely of @xmath10-cycles , then it will follow that the very natural function @xmath78 can not be the sds - map of any of those sds . in this paper , we study the number of @xmath10-cycles that can appear in the phase spaces of sds defined over a complete graph @xmath79 in which all vertex functions are the same .
more formally , we give the following definition .
[ def3 ] let @xmath80 .
let @xmath81 be the set of all functions @xmath82 . for each @xmath83 and @xmath3
, let @xmath84 denote the number of @xmath10-cycles in the phase space @xmath85)$ ] . define @xmath86 our decision to use the identity update order @xmath21 in the definition of @xmath9 in definition [ def3 ] stems from a desire for convenience , but we lose no generality in making such a decision because we are working over complete graphs . to be precise ,
suppose we choose a permutation @xmath3 and a function @xmath87 .
define @xmath88 and @xmath89 by @xmath90 and @xmath91 .
we have @xmath92 .
one may use this fact to show that @xmath93(\gamma(\vec x))=\gamma([k_n , g,\pi](\vec x))\ ] ] for all @xmath94 .
this shows that @xmath95 is a graph isomorphism from the directed graph @xmath85)$ ] to the directed graph @xmath96)$ ] .
in other words , if we start with any @xmath3 and any @xmath87 , we can find a function @xmath88 such that the sds - maps @xmath97 $ ] and @xmath98 $ ] have isomorphic phase spaces .
consequently , @xmath99 in the next section , we reformulate the problem of determining @xmath9 for a given @xmath23 in terms of finding the clique number of a certain graph .
we then show that @xmath9 is equal to the maximum number of codewords in a binary code of length @xmath12 with minimum distance at least @xmath13 .
if @xmath100 is a set and @xmath101 is a function , we say an element @xmath51 of @xmath100 is a periodic point of period @xmath10 of @xmath54 if @xmath102 and @xmath103 .
typically , we use the field @xmath104 as our set of states . however
, in the following lemma , we may use any set of states @xmath4 so long as @xmath105 .
[ lem8 ] let @xmath106 be an integer .
let @xmath107 be a function , and let @xmath108 be a permutation .
suppose @xmath109 is a periodic point of period @xmath10 of the sds - map @xmath110 $ ] .
write @xmath111 . for each @xmath112 , we have @xmath113 . in particular , if @xmath114 , then @xmath115 . for any @xmath116 , recall that @xmath117 is the system state that results from starting with the initial system state @xmath51 and then updating the vertices @xmath118 ( in this order ) .
when we update the state of the vertex @xmath119 , the state to which @xmath119 updates is given by @xmath120 .
on the other hand , the state to which @xmath119 updates must be @xmath121 , the state of @xmath119 in the system state @xmath122 .
it follows that @xmath123 .
note that @xmath124 these same arguments show that @xmath125 for any @xmath116 and that @xmath126 .
if @xmath127 , then the preceding paragraph tells us that @xmath128 @xmath129 we also have @xmath130 @xmath131 notice that if @xmath132 , then updating the vertex @xmath133 does nt `` change '' anything ; that is , @xmath134 and @xmath135 with the help of the equations in and , this shows that if @xmath132 for some @xmath45 , then @xmath136 . as a consequence
, we see that if @xmath132 for some @xmath45 , then @xmath137 .
however , we are assuming that @xmath51 is a periodic point of @xmath54 of period @xmath10 , so it follows by definition that @xmath103 .
thus , @xmath138 for all @xmath45 , which proves the lemma .
let @xmath139 be a finite word over the alphabet @xmath104 .
we say that a vector @xmath140 _ contains the subsequence @xmath141 _ if there exist @xmath142 such that @xmath143 and @xmath144 for all @xmath145 .
for example , the vector @xmath146 contains the subsequence @xmath147 because @xmath148 , @xmath149 , and @xmath150 .
however , the vector @xmath151 does not contain the subsequence @xmath152 .
this leads us to the following definition .
[ def1 ] let @xmath141 be a finite word over the alphabet @xmath104 .
we define @xmath153 to be the set of vectors @xmath94 such that @xmath51 contains the subsequence @xmath141 .
[ def2 ] let @xmath106 be an integer .
let @xmath154 be the set of all elements of @xmath155 whose first coordinate is @xmath156 .
we define @xmath157 to be the undirected simple graph with vertex set @xmath158 and edge set @xmath159 the following two lemmas link the graphs @xmath157 to our study of @xmath10-cycles in the phase spaces of sds defined over complete graphs .
[ lem2 ] let @xmath106 be an integer , and let @xmath160 be a clique of order @xmath161 of @xmath157 .
there exists a map @xmath162 such that each @xmath163 is in a @xmath10-cycle of @xmath164)$ ] .
moreover , no two distinct @xmath165 are contained in the same 2-cycle of @xmath164)$ ] .
take any clique @xmath160 from the graph @xmath157 .
for each @xmath166 , let @xmath167 , where @xmath168 by the definition of @xmath157 .
define the map @xmath162 as follows : * if there are some @xmath169 such that @xmath170 then let @xmath171 .
* if there are some @xmath169 such that @xmath172 then let @xmath173 .
* otherwise , let @xmath174 .
we first need to show that @xmath7 is well - defined .
to do so , we show that for any @xmath175 and any @xmath176 with @xmath177 or @xmath178 , we have @xmath179 @xmath180 and @xmath181 this will show that we have not accidentally defined @xmath174 and @xmath182 for the same vector @xmath183 . applying the function @xmath184 to each side of yields , and applying @xmath184 to each side of yields ( we are using the injectivity of @xmath184 ) .
this shows that and are equivalent , so we only need to prove and to show that @xmath7 is well - defined .
choose such @xmath185 , and assume without loss of generality that @xmath186 .
first , suppose @xmath187 .
because we are assuming either @xmath177 or @xmath178 , this implies that @xmath188 .
the two vectors in can not be the same because their @xmath189 coordinates are different .
similarly , the two vectors in can not be the same because their @xmath190 coordinates are different .
this proves and in the case in which @xmath187 .
we now assume @xmath177 .
because @xmath191 and @xmath192 are adjacent in @xmath157 , the vector @xmath193 contains the subsequence @xmath194 .
this implies that there exist @xmath195 such that @xmath196 , @xmath197 , @xmath198 , and @xmath199 . if @xmath200 , then the two vectors in differ in their @xmath201 coordinates . if @xmath202 , then the two vectors in differ in their @xmath203 coordinates . if @xmath204 and @xmath205 , then @xmath206 , so the two vectors in differ in their @xmath207 coordinates
. this proves in all cases .
we next prove .
note that @xmath208 . since @xmath209 , the first coordinate of the vector on the right - hand side of is @xmath156 .
if @xmath210 , then the first coordinate of the vector on the left - hand side of is @xmath66 .
this shows that holds whenever @xmath210 , so assume @xmath211 .
if @xmath212 , then the two vectors in differ in their @xmath207 coordinates .
hence , we may also assume @xmath213 . since @xmath211 , the @xmath201 coordinate of the vector on the left - hand side of is @xmath214 .
since @xmath215 , the @xmath201 coordinate of the vector on the right - hand side of is @xmath216 . because @xmath197 , the two vectors in differ in their @xmath201 coordinates .
this proves , showing that @xmath7 is indeed well - defined .
let @xmath217 $ ] .
we now show that for each @xmath166 , the vector @xmath191 is in a @xmath10-cycle of @xmath72 .
choose some @xmath166 .
it follows from the definition of @xmath7 that for each @xmath218 @xmath219 and @xmath220 in particular , @xmath221 and @xmath222 .
in other words , @xmath191 is in a @xmath10-cycle of @xmath72 .
finally , choose some distinct @xmath175 .
it is easy to see that the vectors @xmath191 and @xmath192 are in distinct @xmath10-cycles of @xmath72 . since @xmath223 ,
the only way the vectors @xmath191 and @xmath192 could be in the same @xmath10-cycle of @xmath72 is if @xmath224 .
we have just shown that @xmath225 , so @xmath191 could only be in the same @xmath10-cycle as @xmath192 if @xmath226 . however , this is impossible because @xmath191 and @xmath192 have the same first coordinate ( namely , @xmath156 ) .
[ lem3 ] let @xmath106 be an integer , and let @xmath51 and @xmath227 be distinct nonadjacent vertices of @xmath157 .
let @xmath162 be a function .
if @xmath51 is in a @xmath10-cycle of the phase space @xmath164)$ ] , then @xmath227 is not in a @xmath10-cycle of @xmath164)$ ] .
suppose , by way of contradiction , that @xmath51 and @xmath227 are both in @xmath10-cycles of @xmath228)$ ] .
let @xmath217 $ ] .
let @xmath229 and @xmath230 . because @xmath231 , we may let @xmath232 be the smallest element of @xmath233 such that @xmath234 . similarly , we may let @xmath235 be the largest element of @xmath233 such that @xmath236 . note that @xmath237 because @xmath238 . because @xmath51 and @xmath227 are not adjacent in @xmath157 , the vector @xmath239 does not contain the subsequence @xmath194 .
this implies that @xmath240 in other words , @xmath241 . for the sake of convenience , let @xmath242 . because we are assuming that each of the vectors @xmath51 and @xmath227 is in a @xmath10-cycle of @xmath72 , we know from lemma [ lem8 ] that @xmath115 and @xmath243 .
hence , for each @xmath244 , @xmath245 and @xmath246 it follows that @xmath247 but also @xmath248
@xmath249 therefore , @xmath250 .
similarly , we have @xmath251 and @xmath252 @xmath253 this implies that @xmath254 , which is a contradiction .
we are now in a position to prove one of our crucial theorems .
let @xmath255 denote the clique number of a graph @xmath256 .
that is , @xmath256 contains a clique of order @xmath161 if and only if @xmath257 . [ thm1 ] for any integer @xmath106 , we have @xmath258 choose an integer @xmath106 , and let @xmath259 .
let @xmath260 be a clique of order @xmath161 of @xmath157 .
by lemma [ lem2 ] , there exists a map @xmath162 such that any distinct vectors @xmath261 are in distinct @xmath10-cycles of @xmath228)$ ] .
in particular , @xmath228)$ ] contains at least @xmath161 @xmath10-cycles . in the notation of definition [ def3 ] , @xmath262 , and @xmath263 .
thus , @xmath264 we now show that @xmath265 . by definition [ def3 ] ,
there exists a function @xmath87 such that @xmath266 . in other words , there are @xmath9 @xmath10-cycles in the phase space @xmath267)$ ] .
it follows from lemma [ lem8 ] that each @xmath10-cycle of @xmath267)$ ] contains exactly one vector whose first coordinate is @xmath156 . in the notation of definition [ def2
] , each of the @xmath9 @xmath10-cycles of @xmath267)$ ] contains exactly one vector that is a vertex of @xmath157 .
let @xmath268 be these vertices of @xmath157 .
choose some distinct @xmath269 . because each of the vectors @xmath270 is in a @xmath10-cycle of @xmath267)$ ]
, it follows from lemma [ lem3 ] that @xmath191 and @xmath192 must be adjacent in @xmath157 .
because @xmath52 and @xmath271 were arbitrary , this shows that @xmath272 is a clique of @xmath157 .
consequently , @xmath265 . given an integer @xmath106
, we may define a map @xmath273 by @xmath274 because the first coordinate of each vector in @xmath275 is @xmath156 , it should be clear that @xmath276 is a bijection ( in fact , it is an isomorphism of vector spaces ) .
furthermore , it is easy to see that if @xmath277 , then @xmath278 if and only if @xmath279 .
this motivates the following definition .
[ def4 ] given an integer @xmath280 , define @xmath281 to be the undirected simple graph with vertex set @xmath282 and edge set @xmath283 it follows from the preceding paragraph that @xmath284 is a graph isomorphism from the graph @xmath285 to the graph @xmath286 .
therefore , @xmath287 .
in the following section , we relate the graphs @xmath286 to error - correcting binary codes .
in coding theory , a binary code of length @xmath23 is a subset of @xmath155 .
the elements of a code are known as _
codewords_. the _ hamming distance _ between two codewords @xmath288 and @xmath289 , which we shall denote by @xmath290 , is simply the number of positions in which the vectors @xmath51 and @xmath227 have different coordinates .
that is , @xmath291 if @xmath292 is a nonempty binary code of length @xmath23 , then the _ minimum distance _ of @xmath292 , denoted @xmath293 , is the quantity given by @xmath294 if a code @xmath292 consists of a single codeword , we convene to let @xmath295 . of particular interest in coding theory
are binary codes with minimum distance at least @xmath13 .
such codes are known as one - bit error - correcting codes because any error formed by flipping a single bit ( that is , changing a single coordinate ) in a codeword can be detected and corrected ( this is not the case for binary codes with minimum distance less than @xmath13 ) .
let @xmath296 denote the maximum number of codewords that can appear in a binary code of length @xmath23 that has minimum distance at least @xmath13 .
an important problem in coding theory is the determination of the values of @xmath296 .
it is known that @xmath297 if @xmath298 is a power of @xmath10 , but most values of @xmath296 are not known .
given an integer @xmath280 , define @xmath299 to be the undirected simple graph with vertex set @xmath300 and edge set @xmath301 observe that two vectors @xmath302 are adjacent in @xmath299 if and only if the hamming distance @xmath290 between @xmath51 and @xmath227 is at least @xmath13 .
therefore , @xmath296 is equal to the clique number @xmath303 of the graph @xmath304 . from this , we obtain the following theorem .
[ thm2 ] for any positive integer @xmath23 , @xmath305 we saw at the end of the preceding section that @xmath306 .
therefore , in light of the preceding paragraph , we see that it suffices to show that @xmath307 .
consider the linear transformation @xmath308 given by @xmath309 the linear tranformation @xmath310 is an endomorphism with a trivial kernel , so it is an isomorphism .
furthermore , a vector @xmath311 contains the subsequence @xmath312 if and only if @xmath313 contains the subsequence @xmath194 .
this shows that @xmath310 is in fact a graph isomorphism from @xmath304 to @xmath286 , so @xmath307 .
we have defined @xmath9 to be the maximum number of @xmath10-cycles that can appear in a phase space of the form @xmath228)$ ] .
one could easily generalize these numbers by defining @xmath314 to be the maximum number of @xmath315-cycles that can appear in such a phase space .
one may show that @xmath316 for all positive integers @xmath23 ( the only possible fixed points of an sds - map @xmath317 $ ] are the all-@xmath156 s and all-@xmath66 s vectors of length @xmath23 ) .
is it possible to obtain general bounds for the numbers @xmath314 ?
could we perhaps relate the numbers @xmath314 to codes as we have done for the numbers @xmath318 ? there are , of course , many other natural ways to generalize the problems considered here
. one might wish to replace complete graphs with graphs that are , in some sense , almost " complete ( such as complements of cycle graphs ) .
we could also choose to ask similar questions about sds - maps of the form @xmath317 $ ] defined using a set of states @xmath4 with @xmath319 .
the author would like to thank padraic bartlett for introducing the author to sequential dynamical systems and giving excellent suggestions for the improvement of this paper . | a sequential dynamical system ( sds ) consists of an undirected simple graph @xmath0 with vertices @xmath1 , a collection of vertex functions @xmath2 , a permutation @xmath3 , and a collection of states @xmath4 .
the vertices update their states in a sequential order specified by @xmath5 using the vertex functions @xmath2 .
any sequential dynamical system gives rise to an sds - map @xmath6 $ ] , which maps each initial configuration of vertex states to the new configuration of states obtained after each vertex updates .
when all vertex functions are equal to the same function @xmath7 , we write the sds - map as @xmath8 $ ] .
let @xmath9 denote the maximum number of periodic orbits of period @xmath10 that an sds - map of the form @xmath11 $ ] can have .
we show that @xmath9 is equal to the maximum number of codewords in a binary code of length @xmath12 with minimum distance at least @xmath13 .
.2 in 2010 _ mathematics subject classification _ : primary 37e15 ; secondary 05c69 .
_ keywords : sequential dynamical system ; periodic orbit ; complete graph ; binary code . _ |
the evolution of the parton densities at small @xmath0 is a very rich but complicated issue .
the steep rise of the structure function @xmath2 at small @xmath0 is explained by the presence of a huge gluon number density .
the pure dglap evolution equations , meant to describe the evolution of the parton densities as a function of @xmath3 , are able to reproduce the rise of @xmath2 provided the input starting distributions are chosen properly . figure [ ccfm_variables ] shows the pattern of qcd initial - state radiation in a small-@xmath0 dis event , together with labels for the kinematics .
the gluon splitting function @xmath4 is given by : @xmath5 where @xmath6 and @xmath7 ( see fig . [ ccfm_variables ] ) is the ratio of the energy fractions of successive branchings in the gluon chain and @xmath8 is virtuality of the @xmath9-channel gluon with @xmath10 . in dglap the @xmath8 dependence of all emissions in the gluon chain are simplified by the observation that at not too small @xmath0 the dominant part to the cross section comes from the region of phase space where @xmath8 is very small .
however if @xmath0 or @xmath11 becomes very small , the collinear ( small @xmath12 ) approximation of dglap may be inadequate .
this region is treated by the bfkl evolution equation , which keeps the full @xmath12 integration but approximates the gluon splitting function with the asymptotic form @xmath13 from a detailed analysis of interference effects in a gluon chain , it was found that the proper evolution variable is the angle of the emitted gluon , and not the virtuality @xmath8 as in the dglap approximation , nor @xmath11 as in the bfkl approximation .
this angular ordering resulted in the new , and more complicated , ccfm evolution , which reproduces the bfkl and dglap approximations in the small and large @xmath0 limits respectively .
the ccfm equation naturally interpolates between the two extremes .
however , in ccfm the gluon splitting function contains only the singular terms in @xmath11 : @xmath14 with @xmath15 being the non - sudakov form factor to regulate the @xmath16 singularity .
the non singular terms of the splitting function ( see eq.([pgg ] ) ) are not obtained within the ccfm approximation . whereas at hera energies ( @xmath17 gev ) the total cross section of deep inelastic scattering can be reasonably well described with the dglap evolution equations ,
measurements of specific features of the hadronic final state indicate clear deviations from a pure dglap scenario . the cross section at low @xmath0 and large @xmath3 for a high @xmath18 jet in the proton direction ( a forward jet ) has been advocated as a particularly sensitive measure of small @xmath0 parton dynamics @xcite .
if the forward jet has large energy ( @xmath19 ) the evolution from @xmath20 to small @xmath0 can be studied .
when @xmath21 there is no room for @xmath3 evolution left and the dglap formalism predicts a rather small cross section in contrast to the bfkl / ccfm formalisms , which describe the evolution also in @xmath0 .
measurements performed at hera @xcite show that the prediction from the naive dglap formalism lies a factor @xmath22 below the data , whereas the data can be described by ccfm evolution equations @xcite .
the effect of new small @xmath0 parton dynamics is most clearly seen if the contribution from typical dglap dynamics is suppressed .
the forward jet production in deep inelastic scattering at small values of @xmath0 is one example of such a process .
however , the kinematics need to be investigated further .
typical event selection criteria at hera are : [ cols= " < " , ] the range in @xmath0 is typically @xmath23 .
the evolution takes place from the large @xmath20 down to the small @xmath0 with a typical range at hera energies of @xmath24 . in order to justify the use of an evolution equation ( instead of a fixed order calculation ) one would require at least 2 or more gluon emissions during the evolution . to roughly estimate the energy fractions @xmath25 of 3 gluon emissions between @xmath26 , one can assume that each gluon carries the same energy .
then the range of @xmath27 results in @xmath28 , which is far from being in the very small @xmath11 region , where the bfkl or ccfm approximations ( treating only the @xmath16 terms in the gluon splitting function ) are expected to be appropriate . in fig .
[ zval_fjet_hera ] we show the values of the splitting variable @xmath11 in events satisfying the forward jet criteria at hera energies obtained from the monte carlo generator @xcite . since the values of the splitting variable @xmath11 are indeed in the large @xmath11 region ( the majority has @xmath29 ) , it is questionable , whether the bfkl or ccfm evolution equations , including only the @xmath16 terms of the gluon splitting function , are already applicable . whereas the measurement at hera stops at @xmath30 , the available phase space at thera is enlarged by a factor of @xmath31 .
therefore the gluons along the chain will presumably have smaller @xmath11 values and the usage of the small @xmath0 evolution equations might be more justified . in fig .
[ zval_fjet_thera ] we show the @xmath11 values obtained from events satisfying the forward jet selection criteria at thera energies ( @xmath32 gev ) compared with the ones at hera energies ( @xmath33 gev ) .
the four plots correspond to different cuts on the minimum jet angle ( @xmath34 ) .
one clearly can observe , that the distribution of @xmath11 values becomes flat at small @xmath11 at thera energies and that the small @xmath11 values are no longer suppressed as in the hera kinematic region .
this clearly shows , that the application of small @xmath0 evolution equations at thera energies are justified and unavoidable , although the sensitivity to the treatment of large @xmath11 splittings does not completely go away . at hera
the forward jet cross section could be reasonably well described by including a resolved virtual photon contribution , because the phase space for small @xmath11 emissions at hera energies was relatively small .
the situation changes at thera energies : the phase space for small @xmath11 emissions is larger and effects from small @xmath0 evolution become more visible . in fig .
[ fig - fwdjet1 ] the cross section for forward jet production is shown as a function of @xmath0 for standard dglap prediction ( dotted ) , including in addition a contribution from resolved virtual photons ( dashed - dotted ) and the ccfm prediction ( dashed ) .
for comparison the prediction from ariadne@xcite is shown , which implements a semi - classical soft radiation model in a dipole cascade , and which currently gives the best overall description of small-@xmath0 hera data .
since the the available phase space at thera is enlarged by a factor of @xmath31 , the difference between the standard dglap - based calculation and the ccfm calculation is increased . moreover , at thera the ccfm approach predicts a larger cross section than the model with resolved virtual photon contributions added , and smaller cross section than ariadne , while all three models give comparable results at hera .
this gives a unique opportunity , not only to distinguish between the different approaches , but also to study details of the qcd cascade in a regime , where the new small-@xmath0 evolution equations should be appropriate . in fig .
[ fig - fwdjet2 ] the forward jet cross section is shown for different minimal jet angle cuts .
on the experimental side this requires complete acceptance both in the electron and proton direction down to the lowest possible angles . from the size of the cross section
@xmath35 one would like to reach at least @xmath36 for the forward jet measurement , but there are other reasons to ask even for @xmath37 . on the other hand ,
the luminosities needed for such measurements are moderate so that this question can be settled within one year of running at thera .
hera has conclusively shown the need to go beyond the standard dglap evolution equations in order to explain the data on small-@xmath0 final states .
the reach in @xmath0 is , however , not quite enough to really study details of the small-@xmath0 evolution to be able to distinguish between different approaches , such as the ccfm evolution , resolved virtual photons and the dipole cascade model . with the increased kinematical region available at thera
, this will become possible . the steeply rising cross section as @xmath0 gets smaller means that such measurements can be done even with moderate luminosity . on the other hand the demands on the forward coverage of the detector
is more critical ideally it should be possible to measure jets down to an angle of @xmath38 .
once a good understanding of the small-@xmath0 evolution is obtained , it should be possible to use the underlying @xmath39-factorization theorem in bfkl / ccfm , where observables are described in terms of process dependent off - shell matrix element and universal un - integrated parton densities , to make firm predictions of any other small-@xmath0 measurement , just as normal dglap parton densities and matrix elements are used today at large @xmath3 .
thera will be the only place where these un - integrated parton densities can be measured .
n. b. et al . , a comparison of deep - inelastic monte carlo event generators , in _ proceedings of the workshop on monte carlo generators for hera physics _ , edited by a. doyle , g. grindhammer , g. ingelman , h. jung ( , hamburg , 1999 ) , p. 10 . | we discuss some aspects of forward jet production as a signature for small @xmath0 physics at thera energies .
@xmath1 department of theoretical physics , lund university , sweden , [email protected] |
recently , atlas and cms collaborations have discovered the higgs boson with a mass around 126 gev @xcite , which would be a triumph of the standard model ( sm ) .
thus , the status of the sm as the basic theory describing the nature becomes further stabilized .
since the models based on low energy supersymmetry ( susy ) predicted a relatively light higgs mass , one might say that the 126 gev higgs boson supports also susy .
unfortunately , however , any evidence of new physics beyond the sm including susy has not been observed yet at the large hadron collider ( lhc ) .
it implies that the theoretical puzzles raised in the sm such as the gauge hierarchy problem , which have provided motivations of new physics for last four decades , still remain unsettled . in fact
, 126 gev is too heavy for a mass of the higgs appearing in the minimal supersymmetric sm ( mssm ) .
the basic reason of it is that the tree - level quartic coupling of the mssm higgs potential is given by the small sm gauge couplings unlike the sm . as a consequence ,
the predicted tree - level higgs mass in the mssm is lighter even than the @xmath14 boson mass @xmath15 @xcite .
thus , a large radiative correction for lifting the higgs mass is very essential in the mssm to account for the observed higgs mass .
the dominant radiative correction to the higgs mass in the mssm comes from the top quark yukawa coupling in the superpotential ( @xmath16 ) , because the relevant yukawa coupling constant is large enough ( @xmath17 ) . through the top quark yukawa coupling ,
the top quark and its super - partner , `` stop '' contribute to the radiative higgs mass ( @xmath18 ) and also the renormalization of the soft mass parameter of the higgs @xmath19 ( @xmath20 ) @xcite : @xmath21 , \\ & \quad \delta \widetilde{m}_2 ^ 2|_{\rm top}\approx \frac{3|y_t|^2}{8\pi^2}\widetilde{m}_t^2 ~{\rm log}\left ( \frac{\widetilde{m}_t^2}{m_g^2}\right ) \left[1 + \frac12\frac{a_t^2}{\widetilde{m}_t^2}\right ] , \end{split}\ ] ] where @xmath22 ( @xmath23 ) denotes the top quark ( stop ) mass , and @xmath24 is the vacuum expectation value ( vev ) of the higgs , @xmath25 with @xmath26 .
@xmath27 means the grand unified theory ( gut ) scale ( @xmath28 ) adopted as a cut - off of the model . for simple expressions , here we assumed that the `` @xmath29-term '' coefficient corresponding to the top quark yukawa coupling , @xmath30 dominates over @xmath31 , where @xmath32 stands for the `` @xmath32-term '' coefficient in the mssm superpotential . )
are obtained , when the su(2)@xmath33-doublet and -singlet stops are degenerate .
however , they would be good approximations , unless the stops are too hierarchical to be realized in supergravity ( sugra ) models . for the full expressions ,
refer to e.g. ref .
] as seen in eq .
( [ topstop ] ) , only useful parameters for enhancing the radiative higgs mass , @xmath34 are @xmath35 and @xmath36 in the mssm .
for the 126 gev higgs mass , it is known that a stop mass should be heavier than at least a few tev with the help of the @xmath36 term , if two - loop effects are also included @xcite . note that the radiative higgs mass squared , @xmath34 is maximized when @xmath36 fulfills @xmath37 @xcite .
the renormalization contribution in eq .
( [ topstop ] ) , @xmath38 is associated with the fine - tuning problem in the higgs sector , since the higgs vev and the resulting @xmath14 boson mass , which define the electroweak ( ew ) scale , are determined by @xmath39 ( @xmath40 ) [ and other ( soft ) mass parameters ] .
if @xmath35 is of order tev for the 126 gev higgs mass , thus , a tev scale fine - tuning among the soft parameters is needed to obtain the @xmath14 boson mass of @xmath41 ( `` little hierarchy problem '' ) .
particularly , the `` maximal mixing '' case , @xmath37 makes @xmath42 four times larger than the case of @xmath43 , which aggravates the fine - tuning problem . in order to mitigate the fine - tuning in the higgs sector , thus ,
the stop mass squared and @xmath30 should be as small as possible . at the moment , the stop mass bound is @xmath44 @xcite . ,
the stop mass is not constrained at all . if @xmath45 parity is broken , the lower bounds on stop and gluino masses can be relieved to @xmath46 @xcite .
] however , the radiative correction by a heavy gluino ( @xmath47 @xcite ) pulls up @xmath35 too much at the ew scale .
thus , the heavy gluino effect should be adequately suppressed for a light stop mass at low energy .
actually it could be compensated e.g. by very heavy squarks of the first and second generations ( @xmath48 ) @xcite through two loop effects @xcite , or by the wino but giving up the assumption of the universal gaugino mass at the gut scale . at the ew scale @xcite . ]
otherwise , one should assume a low messenger scale to minimize the renormalization group ( rg ) effect of a heavy gluino . in this paper
, we do nt specify a scenario for a small enough stop mass .
although one somehow acquires a small stop mass and successfully reduces the fine - tuning , however , one can not explain yet the observed higgs mass just with a small stop mass . with @xmath49 , we have just @xmath50 . for the 126 gev higgs mass , hence , the mass gap , @xmath51 for @xmath52 should somehow be filled up . for raising the higgs mass , thus ,
we need other ingredients rather than the stop mass .
as pointed out above , a too light tree - level higgs mass could be a crucial cause of the higgs mass problem in the mssm . in the next - to mssm ( nmssm ) , the `` @xmath32 term '' of the mssm ( @xmath53 )
is promoted to the renormalizable coupling with a new singlet superfield @xmath54 @xcite : @xmath55 it provides an additional quartic term in the higgs potential , @xmath56 , which is absent in the mssm , and so the higgs mass in the nmssm receives a _ tree - level _ correction resulting from that term : @xmath57 accordingly , the higgs mass could be significantly increased even with a relatively light stop , if the dimensionless coupling @xmath4 was of order unity . for @xmath58 , @xmath4 should be larger than @xmath59 with @xmath60 for explaining the observed higgs mass @xcite
. however , @xmath4 greater than @xmath61 at the ew scale turns out to bring a landau - pole ( lp ) below the gut scale through its rg evolution @xcite .
thus , only the quite narrow bands of the parameter space are left in the nmssm : @xmath62 note that @xmath63 is the upper bound for perturbativity of the model up to the gut scale , and @xmath64 maximizes eq .
( [ nmssmhiggs ] ) .
they imply that the nmssm accounts for the higgs mass just _ around the boundary of the theoretically valid parameter space_. for naturalness of the higgs mass in the nmssm , therefore , the permitted parameter space should be somehow enlarged . one way to relieve
the lp constraint on @xmath4 is to introduce a new gauge symmetry , under which @xmath54 and @xmath65 are charged : a strong enough new gauge interaction could hold @xmath4 within the perturbative regime up to the gut scale , and so the upper bound of @xmath4 at low energy could be relaxed . in ref .
@xcite , a u(1 ) gauge symmetry is considered for ameliorating the lp problem . since new u(1 )
gauge charges are assigned to @xmath65 , however , ordinary mssm matter fields should also carry proper u(1 ) charges for the desired yukawa couplings . as a result ,
the beta function coefficient of the new u(1 ) gauge coupling becomes much larger , which makes the u(1 ) gauge coupling quite smaller at low energies , and so the relaxation mechanism of the lp constraint becomes inefficient . in ref .
@xcite , the yukawa couplings between the newly introduced vector - like leptons and the ordinary mssm higgs were considered for raising the _ radiative _ higgs mass : @xmath66 where @xmath67 ( @xmath1 ) denote extra lepton doublets ( neutral singlets ) , and the last two terms are the mass terms for them @xcite . only if @xmath67 are heavier than @xmath1 , the former can immediately decay to the latter and sm fermions ( plus lsp ) .
so they do nt leave experimentally unacceptable signals .
since @xmath1 remain absolutely stable , they can be a promising dark matter candidate @xcite .
the higgs vevs are aligned in the directions of the neutral components of @xmath67 , and so there is no extra two photon enhancement in this model .
note that the gauge quantum numbers of @xmath67 are the same as those of the mssm higgs @xmath68 , and so the structures of the @xmath6 and @xmath69 terms in eq .
( [ vlep ] ) are the same as the @xmath4 term in eq .
( [ nmssmhiggs ] ) .
accordingly , the same lp constraint on @xmath4 ( @xmath70 ) could be imposed also to @xmath6 and @xmath69 . since one mssm superfield , @xmath19 ( @xmath71 ) couples to two extra superfields @xmath72 ( @xmath73 ) in the @xmath6 ( @xmath69 ) term of eq .
( [ vlep ] ) , an extra gauge symmetry can be introduced , under which _ all the mssm superfields remain neutral _ , and only the extra vector - like leptons , @xmath74 carry non - trivial extra gauge charges .
then , the mixings between the extra vector - like leptons and the ordinary mssm leptons are forbidden , and so such extra lepton doublets do nt raise phenomenological problems associated with the flavor changing neutral current ( fcnc ) . in ref .
@xcite , the smallest non - abelian gauge group , su(2)@xmath3 was considered , which is still _
asymptotically free_. so at least two pairs of vector - like leptons are necessary for composing su(2)@xmath3 doublets . in this case , the maximally allowed value of @xmath6 at low energy is significantly lifted up , @xmath75 .. in total , three pairs of @xmath76 are introduced in ref .
@xcite , where @xmath77 and one pair of @xmath67 are su(2)@xmath3 singlets . ]
only if @xmath78 is smaller than @xmath79 at the ew scale , thus , it does not blow up below the gut scale .
like the top and stop in the mssm , @xmath72 make contributions to the radiative higgs mass ( @xmath80 ) as well as the renormalization of the soft mass squared of @xmath19 ( @xmath81 ) @xcite : @xmath82_{q = m_g } , \end{split}\ ] ] where @xmath83 for the @xmath84 doublets , and @xmath85 is defined as @xmath86 .
@xmath87 denotes the mass squared of the fermionic component , @xmath88 . for simplicity ,
all the soft mass squareds of @xmath89 are set equal to @xmath90 . here
@xmath91 is assumed .
note that @xmath92 is proportional to @xmath93 , while @xmath94 is to @xmath95 . as in the mssm , @xmath94 is associated with the fine - tuning of the ew scale . for raising the radiative higgs mass @xmath92 , but holding @xmath94 , therefore , a larger @xmath6 and smaller masses of @xmath96 are preferred . in ref .
@xcite , it was assumed that @xmath97 and the soft parameter @xmath90 are smaller than @xmath35 .
actually it is possible , because experimental bounds on the leptonic particles are not severe yet .
the _ quartic _
power of @xmath6 in @xmath92 lifts the radiative higgs mass very efficiently , if @xmath78 is larger than unity . even for the stop mass squared of @xmath98 ,
thus , 126 gev higgs mass can be easily explained , only if the following relation is fulfilled : @xmath99 for @xmath100 , respectively , .
here we list the numbers estimated with @xmath98 . ] even without @xmath29-term contributions .
we note that @xmath101 ( or @xmath102 ) makes eq .
( [ requirement ] ) trivial . in this paper
, we will introduce an abelian gauge symmetry u(1)@xmath3 instead of su(2)@xmath3 for relaxing the lp problem associated with @xmath6 .
hence , only one pair of vector - like leptons , @xmath89 would be enough , because they do nt have to compose non - trivial multi - plets as in the case of a non - abelian gauge group .
consequently , the model could be much simplified .
however , since the u(1)@xmath3 gauge coupling , @xmath103 monotonically increases with energy unlike the non - abelian case , it is hard to get a relatively large @xmath103 at low energies .
it means that the relaxation mechanism of the lp problem using an abelian gauge symmetry would not be much efficient . due to the reason
, @xmath6 can not be large enough to explain 126 gev higgs mass with @xmath104 .
so we will consider also the susy breaking @xmath29-term corresponding to eq .
( [ vlep ] ) as well as the @xmath4 coupling eq .
( [ nmssmsuperpot ] ) of the nmssm .
even with a relatively smaller value of @xmath6 ( @xmath105 ) , thus , the radiative higgs mass could be sufficiently raised .
most of all , we will attempt to investigate how much the parameter space of @xmath4 and @xmath8 can be enlarged in this setup , compared to the case of the original nmssm .
in fact , the sm gauge quantum numbers of @xmath67 are the same as the mssm higgs doublets , @xmath68 , as mentioned above .
thus , introduction of such extra vector - like leptons is equivalent to introduction of new `` inert higgs doublets '' @xmath0 , only if the new higgs doublets do nt develop vevs .
hence , the @xmath6 and @xmath69 terms in eq .
( [ vlep ] ) can be regarded as extensions of the @xmath4 term in eq .
( [ nmssmsuperpot ] ) of the nmssm except for the fact that @xmath0 carry extra u(1)@xmath3 charges unlike the ordinary mssm higgs @xmath68 .
as pointed out above , the u(1)@xmath3 is necessary also for avoiding unwanted fcnc . in this paper
, we will look upon the extra leptons of ref .
@xcite as extra inert higgs doublets .
this paper is organized as follows . in section [ sec :
model ] , we will discuss the radiative higgs mass , and the fine - tuning in this model .
particularly , we will investigate the @xmath29-term effects in section [ sec : model ] . in section [ sec : lp ] , we will analyze the lp constraint on the coupling constants , @xmath6 and @xmath4 , and explore the allowed parameter space .
we will also compare our results with the cases in the absence of @xmath6 or @xmath4 couplings , or the gauged u(1)@xmath3 in section [ sec : lp ] .
section [ sec : conclusion ] is a conclusion .
by introducing one extra pair of the higgs doublets @xmath0 and singlets @xmath1 , we extend the nmssm higgs sector eq .
( [ nmssmsuperpot ] ) as follows : @xmath106 from the last two terms , @xmath0 and @xmath1 acquire the masses . @xmath107 and @xmath108 of order
ew scale can be naturally induced through the pecci - quinn symmetry breaking @xcite or susy breaking mechanism @xcite .
we assume @xmath109 such that @xmath0 immediately decay into @xmath1 plus sm fermions ( and lsp ) as in eq .
( [ vlep ] ) .
in fact , @xmath108 is not essential , because @xmath1 can get the masses also from the @xmath6 and @xmath69 terms , when the higgs vevs , @xmath110 and @xmath111 are developed .
@xmath108 can be even smaller than @xmath112 .
due to the relatively heavy masses of @xmath0 , we assume that @xmath0 do nt get non - zero vevs ( `` inert higgs doublet '' ) .
@xmath54 also get a mass and a vev by including its self - couplings in the superpotential , and also their soft terms that we do nt specify here @xcite .
actually , the @xmath69 term ( @xmath71 term in general ) is less helpful for raising the higgs mass , since its contribution to the radiative higgs mass would be @xmath8 suppressed for @xmath113 . for a simple analysis , thus , we will neglect @xmath69 coupling as in eq .
( [ vlep ] ) , assuming @xmath114 .
accordingly , in this paper we will consider only the following terms among the holomorphic soft terms : @xmath115 since the sm gauge quantum numbers of @xmath0 are the same as @xmath67 , the @xmath6 term of eq .
( [ superpot ] ) would be the same as the @xmath6 term of eq .
( [ vlep ] ) , if other quantum numbers are ignored .
unlike the model in eq .
( [ vlep ] ) , we introduce an extra u(1)@xmath3 gauge symmetry , whose charge assignment is presented in table [ tab : u(1)charge ] . as seen above , of course , introduction of an extra non - abelian gauge symmetry is very helpful for raising the radiative higgs mass , avoiding the lp problem . in this paper , however , we attempt to raise it just with an abelian gauge symmetry , considering also the helps coming from the @xmath4 term in eq .
( [ superpot ] ) and soft terms . as a result
, only one pair of vector - like superfields is introduced . if necessary
, one can extend the extra abelian gauge symmetry u(1)@xmath3 to u(1)@xmath116u(1)@xmath117u(1)@xmath118 , which could much enhance the mechanism for evading lp without introducing more matter .
nonetheless , we will focus on the case only with one extra u(1 ) in this paper .
c||c|c|c superfields & @xmath119 , @xmath120 , ( @xmath121 , @xmath122 ) & @xmath123 , @xmath124 , ( @xmath125 , @xmath126 ) & mssm superfields , @xmath54 + u(1)@xmath127 charge & @xmath128 & @xmath129 & @xmath130 since the u(1)@xmath3 assigns the non - zero charges only to the extra vector - like superfields as shown in table [ tab : u(1)charge ] , it forbids mixing between the ordinary mssm superfields and the newly introduced vector - like superfields in the bare superpotential .
accordingly , the new su(2)@xmath33 doublets @xmath0 do nt induce unacceptable fcnc phenomena .
@xmath1 can play the role of the higgs for spontaneous breaking of u(1)@xmath3 , if they got vevs .
just for simplicity of the mass spectrum , one can introduce additional singlets , @xmath131 for breaking u(1)@xmath3 .
they can acquire vevs just through the same mechanism with that for the mssm higgs .
we do nt discuss it here in details . in order to maintain the sm gauge coupling unification in the ( n)mssm , @xmath0 need to be supplemented with relatively heavier colored particles , @xmath77 to compose @xmath132 of su(5 ) or @xmath133 of flipped su(5 ) @xcite .
they could eventually decay to sm fermions and neutral particles via , e.g. @xmath134 , where @xmath135 is an mssm quark doublet .
in fact , relatively heavier extra colored particles @xmath77 can cure the small deviation of the gauge coupling unification appearing at the two - loop level in the mssm . in this paper , however , we do nt discuss also it .
although we introduced @xmath136 for completeness of the model , they do nt play an essential role for raising the radiative higgs mass : @xmath77 could make just indirect contributions for improving the rg behaviors of various relevant yukawa couplings , which will be discussed later . except for it , @xmath136 are almost spectators , concerning the radiative corrections of the higgs which will be discussed below . nonetheless , we note that many string models provide extra vector - like pairs of @xmath137 as well as singlets and extra u(1)s @xcite . in this sense , introducing one pair of them and studying their phenomenological implications , in particular , the effects on the higgs mass would be important , because the little hierarchy problem threatens the traditional status of the mssm at the moment . in the bases of @xmath138 , the squared mass matrix for the neutral scalar fields in this model takes the following form : @xmath139 + \left [ \begin{array}{c } { \rm nonholomorphic } \\
{ \rm soft~mass~matrix } \end{array } \right ] , ~ \end{aligned}\ ] ] where we ignored the contributions coming from @xmath69 , the mssm @xmath32-term ( or vev of @xmath54 ) , and `` @xmath125-term '' owing to their relative smallness .
one can suppose that the `` nonholomorphic soft mass matrix '' takes a simple diagonal form : @xmath140 . actually , the expressions of the mass eigenvalues for eq .
( [ mmatrix ] ) are too complicate .
thus , we set @xmath141 , assuming @xmath142 , i.e. su(2)@xmath33 doublets and singlets are almost degenerate as in the ( s)top sector , and will study the following two limited cases for relatively simple analytic expressions : @xmath143 for case i , then , the four eigenvalues of the squared mass matrix eq .
( [ mmatrix ] ) are presented as @xmath144 where we expand the eigenvalues in powers of @xmath145 up to its quartic terms just for future convenience . in eq .
( [ evaluesi ] ) , @xmath146 is defined as @xmath147 , and the coefficients , @xmath148 and @xmath149 are given by @xmath150}{2\sqrt{a_i^2(4+a_i^2 ) } } .
\end{split}\ ] ] by setting @xmath151 in eq .
( [ evaluesi ] ) , we can obtain also the eigenvalues of the squared mass matrix for the fermionic fields , @xmath152 : @xmath153 for case ii , the eigenvalues of @xmath154 are expressed as follows : @xmath155 where @xmath156 , and the coefficients , @xmath157 and @xmath158 are @xmath159 unless we assume a non - vanishing vev for the new higgs doublets , @xmath160 should be greater than @xmath161 such that @xmath162 . of course , the eigenvalues of @xmath152 in case ii are still given by eq .
( [ mf ] ) . with the mass spectra of eqs .
( [ evaluesi ] ) , ( [ mf ] ) , and ( [ evaluesii ] ) , one can calculate the radiative corrections by @xmath163 . concerning the radiative higgs mass and its renormalization , it is convenient to read them from the coleman - weinberg potential @xcite : @xmath164 , \end{split}\ ] ] where @xmath165 denotes the renormalization scale .
@xmath166 is expanded in powers of @xmath167 : ( @xmath168 ) : @xmath169 for case i , the coefficients of the quadratic and quartic terms of @xmath170 in eq .
( [ expansion ] ) are estimated as @xmath171 , \nonumber \\ & & \left(\partial^2_{h_u}\partial^2_{h_u^*}\delta v\right)_0\approx\frac{|y_n|^4}{4\pi^2}\bigg [ { \rm log}\left(1+z^2\right ) -\frac{a_{i}^4}{2}{\rm log}\left(\frac{1+z^2}{z^2}\right ) \label{quarti } \\ & & \qquad\qquad \qquad\quad + a_{i}^2(2+a_{i}^2)\left\{1-z^2{\rm log}\left(\frac{1+z^2}{z^2}\right)\right\}\bigg ] , \nonumber \end{aligned}\ ] ] where the function @xmath172 is given again by @xmath173 as in eq .
( [ vlepradi ] ) , and the parameter @xmath174 is defined as @xmath175 .
here we neglected @xmath176 in @xmath172s and the logarithmic functions in eqs .
( [ quadi ] ) and ( [ quarti ] ) , because it is supposed to be quite smaller than @xmath177 . for case
ii , the coefficients of eq .
( [ expansion ] ) are given by @xmath178 \qquad \\ & & \qquad\qquad\qquad\quad + a_{ii}^2 \left\ { f_q(\widetilde{m}^2+|m_b|^2)-f_q(\widetilde{m}^2-|m_b|^2)\right\}\bigg ] , \nonumber \\ & & \left(\partial^2_{h_u}\partial^2_{h_u^*}\delta v\right)_0\approx\frac{|y_n|^4}{32\pi^2}\bigg [ \left\{2 + 6a_{ii}^2+(1 + 4a_{ii}^2-a_{ii}^4)\frac{z^2}{\zeta^2}\right\}{\rm log}\left(z^2+\zeta^2\right ) \label{quartii } \\ & & \qquad\qquad\qquad\quad + \left\{2 - 6a_{ii}^2+(-1 + 4a_{ii}^2+a_{ii}^4)\frac{z^2}{\zeta^2}\right\}{\rm log}\left(z^2-\zeta^2\right ) \nonumber \\ & & \qquad\qquad\qquad\quad + \left(4 - 8a_{ii}^2\frac{z^2}{\zeta^2}\right){\rm log}z^2 -2\left(1-a_{ii}^4\right ) \bigg ] , \nonumber \end{aligned}\ ] ] where @xmath179 .
the quadratic term in eq .
( [ expansion ] ) ( @xmath180 ) with the coefficient of eq .
( [ quadi ] ) or ( [ quadii ] ) depends on the renormalization scale @xmath165 .
it renormalizes the tree - level soft mass parameter of @xmath19 appearing in the mssm lagrangian , @xmath181 together with the ( s)top contribution @xmath182 : @xmath183 inserting the rg solution of @xmath184 into eq .
( [ renorm ] ) yields the low energy value of @xmath39 ( @xmath185 ) , replacing @xmath165 in eq .
( [ renorm ] ) by a cut - off scale @xmath186 @xcite .
the soft terms are regarded as being generated at the messenger scale of susy breaking , since the soft terms would become non - local operators above the messenger scale .
thus , the messenger scale is adopted as the cut - off scale , and so we have @xmath187 where @xmath188 stands for the value of @xmath39 at the scale that it is generated , namely @xmath186 .
the ( s)top contribution , @xmath38 is presented as @xcite @xmath189_{q=\lambda } \\
\nonumber & & + \frac{3|y_t|^2}{16\pi^2}\frac{|a_t|^2}{\widetilde{m}_{tl}^2-\widetilde{m}_{tr}^2 } \bigg[f_q(\widetilde{m}_{tl}^2)-f_q(\widetilde{m}_{tr}^2)\bigg]_{q=\lambda } , \end{aligned}\ ] ] where @xmath190 ( @xmath191 ) denotes the soft mass squared of su(2)@xmath33 doublet ( singlet ) stop .
, under which eq .
( [ mt ] ) is a good approximation . if @xmath30 is comparable to the stop masses , however , @xmath192 in eq .
( [ mt ] ) should be replaced by the mass eigenvalues , @xmath193 after mass matrix diagonalization for a precise expression . ]
this is the dominant radiative correction to @xmath194 in the mssm .
note that @xmath38 is regular at @xmath195 .
moreover , it is quite insensitive to @xmath196 . in the limit of @xmath197 , eq .
( [ mt ] ) approaches to a much simple form : @xmath198 for case i , @xmath199 in eq .
( [ ewm2 ] ) is given by @xmath200_{q=\lambda } \label{fi } \\
\nonumber & \longrightarrow & \frac{|y_n|^2}{8\pi^2 } \widetilde{m}^2 { \rm log}\frac{\widetilde{m}^2}{\lambda^2}\times\left(1+\frac{a_{i}^2}{2z^2}\right ) \qquad { \rm for}~~\lambda^2~\gg~\widetilde{m}^2~\gg~ |\mu_h|^2 , \ ] ] while @xmath199 for case ii is @xmath201_{q=\lambda } \label{fii } \\
\nonumber & \longrightarrow & \frac{|y_n|^2}{8\pi^2 } \widetilde{m}^2 { \rm log}\frac{\widetilde{m}^2}{\lambda^2}\times\left(1+\frac{a_{ii}^2\zeta^2}{2z^2}\right ) \qquad { \rm for}~~\lambda^2~\gg~\widetilde{m}^2,~ |m_b|^2 ~\gg~|\mu_h|^2 . ~~ \ ] ] in the last lines of eqs .
( [ fi ] ) and ( [ fii ] ) , we took the large @xmath186 limit . as mentioned in introduction , the ew scale or the higgs vev
is determined by @xmath194 of eq .
( [ ewm2 ] ) and other ( soft ) mass parameters : @xmath194 eventually participates in the one of the extremum conditions for the higgs potential ( @xmath202 ) @xcite : @xmath203 which @xmath204 denotes the `` @xmath205 term '' coefficient .
it should , of course , be fulfilled around the vacuum state .
if @xmath194 is too large , it gives rise to a serious fine - tuning problem , because @xmath194 should be matched to the @xmath14 boson mass squared , @xmath206 [ @xmath207 in eq .
( [ extremum ] ) . for naturalness of the ew scale and its perturbative stability , thus , the dimensionful parameters in eqs .
( [ m2 ^ 2stop ] ) , ( [ fi ] ) , and ( [ fii ] ) should be small enough . also , a lower mediation scale of susy breaking is very helpful for relaxing the fine - tuning . in the mssm , eq .
( [ m2 ^ 2stop ] ) with stop mass heavier than @xmath208 would make the biggest contributions to @xmath194 .
moreover , @xmath209 in eq .
( [ m2 ^ 2stop ] ) deteriorates the fine - tuning problem . in order to minimize the fine - tuning , thus ,
the stop mass needs to be as light as possible , and the @xmath210 should also be suppressed . with a stop mass much heavier than @xmath208 and an @xmath210 comparable to it ,
however , the observed higgs mass would be more easily explained , as will be seen later , even if the fine - tuning problem becomes worse . in this paper
, we will take the experimental lower bound , @xmath211 , assuming a quite small @xmath30 term . under this condition ,
we will attempt to account for the observed 126 gev higgs mass , utilizing other ingredients contained in this model .
concerning the fine - tuning problem , smaller susy breaking @xmath29-term and susy(-breaking ) masses would be required also for the extra vector - like fields @xmath212 .
actually , light extra leptonic particles are still experimentally acceptable , only if they can immediately decay to the neutral particles . on the contrary ,
masses of extra colored particles are severely constrained from lhc data , and heavy enough extra colored particles coupled to the mssm higgs would cause a fine - tuning in the higgs sector .
it is the reason why we are particularly interested in the extra _ colorless _ particles . in the limit of @xmath213 and @xmath214 , however , @xmath215 and @xmath216 in eqs .
( [ fi ] ) and ( [ fii ] ) vanish , respectively . in this limit , therefore , larger values of @xmath217 can be taken without making the fine - tuning worse . instead
, @xmath218 should be small enough , because the other terms of eqs .
( [ fi ] ) and ( [ fii ] ) increase in this case . in the limit of @xmath219 , in contrast , @xmath215 in eq .
( [ fi ] ) becomes larger , while the first two terms of eq .
( [ fi ] ) cancel each other . in this case , thus , @xmath220 should be small enough for avoiding a too serious fine - tuning .
note that @xmath160 can not be smaller than @xmath161 in case ii , if the extra higgs do nt get a non - zero vev . in the case of eq .
( [ vlep ] ) , the @xmath221-term was not considered . instead , a large value of @xmath6 ( @xmath222 ) was possible @xcite .
so @xmath160 could not be much greater than @xmath97 . in this model , however , we can have a relatively large @xmath221-term .
so @xmath223 is permitted , only if @xmath224 .
for instance , if the u(1)@xmath3 gauge sector plays also the role of the susy breaking messenger @xcite , or the extra vector - like superfields @xmath212 carry also other gauge charges associated with a u(1)@xmath127 mediation of susy breaking apart from u(1)@xmath3 , a relatively large @xmath221 ( and also @xmath225 ) term can be generated , leaving intact the @xmath30 term . the quartic term in eq .
( [ expansion ] ) with the coefficient of eq .
( [ quarti ] ) or ( [ quartii ] ) , which is independent of the renormalization scale @xmath165 @xcite , makes contribution to the radiative correction to the _ physical higgs mass _ together with the ( s)top .
thus , the summation of all the tree - level and the radiative squared masses should yield the experimental value of the higgs squared mass : @xmath226 here the first and second terms are the tree - level higgs mass of the mssm and nmssm , while the last two terms correspond to the radiative corrections to it .
the ( s)top contribution @xmath34 in eq .
( [ fullmass ] ) is presented as @xcite @xmath227 , \end{aligned}\ ] ] which is regular also at @xmath195 . in the limit of @xmath195 ( @xmath228 )
, it approaches to @xmath229 .
\end{split}\ ] ] note that @xmath230 in eq .
( [ mssmradi ] ) is simply written as @xmath231 as seen in eq .
( [ topstop ] ) . even if eq .
( [ mssmradi ] ) is derived under the limit of @xmath195 , it is approximately valid over the large parameter space of @xmath232 , unless they are extremely hierarchical . particularly , if @xmath233 , eq .
( [ mssmradi ] ) is almost the maximum that eq .
( [ mssmradimass ] ) is able to reach using @xmath190 and @xmath191 for a given @xmath234 .
were it not for the last two terms in eq .
( [ mssmradi ] ) , thus , @xmath35 was only a useful parameter for raising the higgs mass in the mssm .
since the radiative higgs mass is a logarithmic function of @xmath35 in such a case , raising the higgs mass using @xmath35 would be a quite inefficient way .
the last two terms in eq .
( [ mssmradi ] ) are maximized when @xmath235 .
as mentioned before , however , they make the fine - tuning problem more serious . in this paper , thus , we will discuss the radiative higgs mass without considering the @xmath30 terms in eq .
( [ mssmradi ] ) , as mentioned above .
the contribution to the radiative higgs mass by @xmath163 in eq .
( [ fullmass ] ) , @xmath236 is computed with the quartic coefficient of eq .
( [ expansion ] ) , i.e. eq . ( [ quarti ] ) or ( [ quartii ] ) : @xmath237 \equiv \frac{v_h^2{\rm sin}^4\beta}{4\pi^2}|y_n|^4 \times z_i \\ & & \left(\longrightarrow\frac{v_h^2{\rm sin}^4\beta}{4\pi^2}|y_n|^4\left[{\rm log}z^2+\frac{a_i^2}{z^2}\left(1-\frac{1}{12}\frac{a_i^2}{z^2}\right)\right]\quad { \rm for}~z^2,~a_i^2\gg1\right ) \nonumber \end{aligned}\ ] ] and @xmath238 \equiv \frac{v_h^2{\rm sin}^4\beta}{4\pi^2}|y_n|^4 \times z_{ii } \nonumber \\ & & \left(\longrightarrow\frac{v_h^2{\rm sin}^4\beta}{4\pi^2}|y_n|^4\left[{\rm log}z^2+a_{ii}^2\frac{\zeta^2}{z^2}\left(1-\frac{a_{ii}^2}{12}\frac{\zeta^2}{z^2}\right)\right]\quad { \rm for}~z^2/\zeta^2 , ~a_{ii}^2\gg1\right ) \nonumber \end{aligned}\ ] ] for case i and ii , respectively .
although we do nt necessarily require @xmath239 or @xmath240 , we presented the asymptotic expressions of @xmath241 in eqs .
( [ massi ] ) and ( [ massii ] ) just for comparison with eq .
( [ mssmradi ] ) .
the shapes of @xmath242 and @xmath243 defined in eqs .
( [ massi ] ) and ( [ massii ] ) are displayed in fig . 1 and 2 in terms of @xmath244 and @xmath245 , respectively .
as seen in fig .
1-(a ) and 2-(a ) , the @xmath221-term in eq .
( [ soft ] ) is quite helpful for raising the radiative higgs mass : @xmath246 and @xmath243 , which are proportional to the radiative higgs mass , rapidly grow along the @xmath174 ( @xmath247 and @xmath248 ) direction(s ) for relatively larger @xmath247 and @xmath248 ( a smaller @xmath174 ) .
we note that @xmath246 and @xmath243 are in the range of @xmath249 for @xmath250 and @xmath251 in case i , and for @xmath252 , @xmath253 , and @xmath254 in case ii .
these parameter ranges are translated into the following scopes in terms of the lagrangian parameters : @xmath255 and @xmath256 they easily pass the lep constraint on extra leptons @xcite . moreover
, the extra charged leptons rapidly decay to the neutral ones and sm fermions in this model .
the above ranges of parameters do nt much affect the oblique parameters , @xmath257 and @xmath54 .
if @xmath258 and @xmath259 , @xmath260 and @xmath261 by the new vector - like fermions are estimated as @xcite @xmath262 the experimental best fit requires that @xmath263 ( @xmath264 ) for @xmath265 @xcite .
it can be satisfied only if @xmath266 , @xmath267 , @xmath268 for @xmath269 , respectively , and @xmath270 .
even if @xmath108 is much smaller than @xmath271 , the results are not much different : it can be shown that this case yields a bit smaller @xmath272 than those for @xmath258 using the formulas of refs .
the contributions of the new scalars to @xmath272 would be more suppressed due to their heavy masses .
the higgs mass in this model is determined with eqs .
( [ fullmass ] ) , ( [ mssmradi ] ) , ( [ massi ] ) , and ( [ massii ] ) . if the @xmath30 terms are ignored in eq .
( [ mssmradi ] ) , they are recast into @xmath273 \end{split}\ ] ] where @xmath274 and @xmath275 .
@xmath242 and @xmath243 are defined in eqs .
( [ massi ] ) and ( [ massii ] ) , respectively . only if the above equation is satisfied , thus
, the observed 126 gev higgs mass is explained .
however , we will constrain the parameters in eq . (
[ 126higgs ] ) such that the newly introduced superfields @xmath163 do nt raise a more serious fine - tuning problem in the higgs sector than that already resulted from the ( s)top . from eqs . (
[ m2 ^ 2stop ] ) , ( [ fi ] ) , and ( [ fii ] ) , thus , we have @xmath276 for the least tuning . here
we assumed @xmath277 . only with the experimental lower bound of the stop , @xmath278
, we will show that the higgs boson mass of 126 gev can still be explained with our extension of the nmssm , eq .
( [ superpot ] ) in the parameter space eq .
( [ ft ] ) , as mentioned above . in order to meet eq .
( [ 126higgs ] ) , a larger value of @xmath6 is favored : were it not for the @xmath221- and @xmath4 terms , @xmath78 should be greater than unity . for a larger parameter space of eq .
( [ ft ] ) , on the other hand , a smaller @xmath6 is preferred .
note that the _ quartic _ power of @xmath78 appears in eq .
( [ 126higgs ] ) , while just its _ quadratic _ power dependence arises in eq .
( [ ft ] ) . in ref .
@xcite , thus , @xmath279 was taken . instead ,
a relatively smaller @xmath160 and @xmath177 compensate it . in this case
, a unwanted lp would appear below the gut scale .
because of that , an extra non - abelian gauge symmetry , su(2)@xmath3 was introduced in ref .
@xcite . in this paper
, we will consider the possibility of @xmath224 and an extra u(1)@xmath3 gauge symmetry .
instead , we take into account of the @xmath4 and @xmath221 terms in eqs .
( [ superpot ] ) and ( [ soft ] ) . as seen in fig . 1 and 2
, a relatively larger value of @xmath221 can efficiently raise @xmath280 or the radiative higgs mass . for @xmath281 , the entire parameter spaces of @xmath244 and @xmath245 in fig .
1 and 2 are well - inside the upper bounds of eq .
( [ ft ] ) . for @xmath282 [ @xmath283 in case i [ ii ] , however , only the space inside the dotted line in fig .
1-(b ) [ 2-(b ) ] fulfills eq .
( [ ft ] ) . in both cases ,
we set @xmath284 .
as seen in eqs .
( [ massi ] ) , ( [ massii ] ) , and ( [ 126higgs ] ) , larger values of @xmath4 and @xmath6 are preferred for raising the tree - level and radiative higgs masses .
if such yukawa coupling constants are too large , however , they would blow up below the gut scale . in this section ,
we investigate the maximally allowed values for them at low energy scale ( @xmath285 ) for avoiding the lp constraints . from eq .
( [ superpot ] ) and the ordinary superpotential of the nmssm , the anomalous dimensions of the relevant superfields can read as follows : @xmath286 here we considered only the third generation of the mssm matter , @xmath287 , concerning the mssm yukawa couplings : @xmath288 , @xmath289 , and @xmath290 denote the top quark , bottom quark , and tau s yukawa couplings , respectively .
the @xmath291 in eqs .
( [ gammaextra ] ) , ( [ gammahiggs ] ) , and ( [ gammamatt ] ) stand for the three mssm gauge couplings associated with su(3)@xmath292 , su(2)@xmath33 , and u(1)@xmath293 gauge interactions . as discussed before , we introduced an extra u(1)@xmath3 gauge symmetry in order to resolve the lp problem associated with @xmath6 .
@xmath294 terms appear in eq .
( [ gammaextra ] ) due to the u(1)@xmath3 gauge interactions with the charge assignment in table [ tab : u(1)charge ] .
such mssm and extra gauge interactions make the _ negative _ contributions to the anomalous dimensions . considering the relevant superpotential , one can readily write down the rg equations for the yukawa coupling constants : @xmath295 , \vspace{0.3cm}\\ \frac{d|\lambda|^2}{dt}= \frac{|\lambda|^2}{8\pi^2}\bigg[|y_n|^2 + 4|\lambda|^2 + 3|y_t|^2 + 3|y_b|^2+|y_\tau|^2 - 3g_2 ^ 2-\frac35g_1 ^ 2 \bigg ] , \vspace{0.3 cm } \\ \frac{d|y_t|^2}{dt}=\frac{|y_t|^2}{8\pi^2}\bigg[|y_n|^2+|\lambda|^2 + 6|y_t|^2+|y_b|^2-\frac{16}{3}g_3 ^ 2 - 3g_2 ^ 2-\frac{13}{15}g_1 ^ 2 \bigg ] , \vspace{0.3 cm } \\ \frac{d|y_b|^2}{dt}=\frac{|y_b|^2}{8\pi^2}\bigg[|\lambda|^2+|y_t|^2 + 6|y_b|^2+|y_\tau|^2-\frac{16}{3}g_3 ^ 2 - 3g_2 ^ 2-\frac{7}{15}g_1 ^ 2 \bigg ] , \vspace{0.3 cm } \\ \frac{d|y_\tau|^2}{dt}=\frac{|y_\tau|^2}{8\pi^2}\bigg[|\lambda|^2 + 3|y_b|^2 + 4|y_\tau|^2 - 3g_2 ^ 2-\frac{9}{5}g_1 ^ 2 \bigg ] , \end{array } \right
. \end{aligned}\ ] ] where @xmath296 parametrizes the renormalization scale , @xmath297 .
the one - loop rg equations for the three mssm gauge couplings are integrable .
the rg solutions to them take the following form : @xmath298 where @xmath299 ( @xmath300 ) denotes the beta function coefficients for the gauge couplings of su(3)@xmath292 , su(2)@xmath33 and u(1)@xmath293 [ with the su(5 ) normalization ] . in the presence of the one pair of @xmath132 as in our case ,
they are given by @xmath301 , and the unified gauge coupling @xmath302 at the gut scale is estimated as @xmath303 .
the solution to the rg equation of the extra u(1)@xmath3 gauge coupling has also a similar form to eq .
( [ rgmssm ] ) : @xmath304 where @xmath305 parametrizes the u(1)@xmath3 breaking scale @xmath306 [ @xmath307 .
u(1)@xmath3 can be spontaneously broken at tev scale in the same manner of su(2)@xmath308u(1)@xmath293 breaking mechanism in the mssm .
@xmath309 in eq .
( [ zgauge ] ) denotes the beta function coefficient of @xmath310 , which is model - dependent . if the extra colored particles @xmath311 and the u(1)@xmath3 breaking higgs @xmath131 carrying u(1)@xmath3 of @xmath312 are also included , the charge assignment in table [ tab : u(1)charge ] yields @xmath313 .
we assume that all the gauge couplings are unified , @xmath314 at the @xmath27 scale . from the first equation in eq .
( [ rgeq ] ) , we can expect that the lp constraint can be relaxed by the additional negative contributions coming from the @xmath315 terms . as a result ,
the allowed maximal values for @xmath316 at low energies can be lifted up , compared to the case that the u(1)@xmath3 gauge symmetry is absent , and so the radiative higgs mass can be raised with a larger @xmath6 , particularly for a large @xmath8 as seen in eqs .
( [ massi ] ) and ( [ massii ] ) .
consequently , even relatively smaller value of @xmath4 can be consistent with the observed 126 gev higgs mass : the lower bound of @xmath4 and the upper bound of @xmath8 could be relaxed , compared to the case of nmssm , eq .
( [ nmssmpara ] ) . now let us discuss the cases of * a. * @xmath317 and @xmath318 , * b. * @xmath319 and @xmath320 , * c. * @xmath319 and @xmath321 without the u(1)@xmath3 gauge symmetry , and * d. * @xmath319 and @xmath321 with the u(1)@xmath3 gauge symmetry in order .
if the @xmath6 coupling is not introduced in the superpotential eq .
( [ superpot ] ) , u(1)@xmath3 can not affect the radiative higgs potential , because only the extra vector - like superfields have the non - zero charge of it , as seen in table [ tab : u(1)charge ] .
nonetheless , we have introduced extra one pair of @xmath132 , and so the mssm gauge couplings become a bit larger by them at higher energy scales , compared to those of the original ( n)mssm .
hence , the lp constraint of @xmath4 can slightly be relaxed by the stronger mssm gauge interactions via the rg equation of @xmath4 of eq .
( [ rgeq ] ) @xcite , even if no @xmath315 contribution is there .
since yukawa couplings monotonically increase with energy , throughout this paper we require that all the squared yukawa coupling constants discussed here should be smaller than the perturbativity bound , @xmath322 at the gut scale .
c|c|c|c|c|c & @xmath323 & @xmath324 & @xmath325 & @xmath326 & @xmath327 + @xmath328 & @xmath329 & @xmath330 & @xmath331 & @xmath332 & @xmath332 + @xmath333 & @xmath334 & @xmath335 & @xmath336 & @xmath337 & @xmath338 table [ tab : y_n=0 ] lists the maximally allowed low energy values of @xmath4 ( @xmath339 ) , avoiding the lp below the gut scale , and the needed values of @xmath4 for explaining 126 gev higgs mass ( @xmath340 ) with @xmath341 , depending on @xmath8 , _ when the @xmath6 coupling is absent , i.e. for @xmath317_. @xmath328 can be estimated using the rg equations eq .
( [ rgeq ] ) , while @xmath342 is determined by eq .
( [ 126higgs ] ) setting @xmath343 . since @xmath344 can not exceed @xmath328 to avoid the lp below the gut scale , @xmath8 should be smaller than 5.5 in this case .
we see that even if there is no @xmath6 coupling in the superpotential , the allowed ranges of the values for @xmath8 and @xmath4 become quite wider in the presence of one extra pair of @xmath132 : @xmath345 in this case , still only relatively small values of @xmath8 are consistent with the observed higgs mass and the perturbativity of the model up to the gut scale .
now , we discuss the lp constraint of @xmath6 , when @xmath320 in the superpotential eq .
( [ superpot ] ) . since
@xmath123 and @xmath120 are charged under u(1)@xmath3 , the u(1)@xmath3 gauge interaction is helpful for relaxing the lp problem for @xmath6 . if the u(1)@xmath3 gauge symmetry is not introduced , however , there does not appear the @xmath315 term in the rg equation of @xmath6 in eq .
( [ rgeq ] ) . as a result ,
elusion of a lp for @xmath6 is not efficient .
for comparison , table [ tab : lambda=0 ] displays the results of the both cases , @xmath346 and @xmath347 .
@xmath348 means the maximally allowed low energy value of @xmath6 needed for evading lp , which is obtained by performing the analysis of the rg equations eq .
( [ rgeq ] ) .
@xmath349 is the value of @xmath246 or @xmath243 required to account for the 126 gev higgs mass for each @xmath348 .
it can be estimated with eq .
( [ 126higgs ] ) .
@xmath350 indicates the value of @xmath6 needed for explaining the 126 gev higgs mass when @xmath280 takes the maximum value , @xmath351 .
for perturbativity up to the gut scale and naturalness of the model , @xmath6 should be located between @xmath348 and @xmath350 .
c||c|c|c|c|c|c|c|c @xmath352 & @xmath325 & @xmath353 & @xmath354 & @xmath355 & @xmath356 & @xmath357 & @xmath358 & @xmath270 + @xmath348 & @xmath331 & @xmath332 & @xmath338 & @xmath338 & @xmath338 & @xmath332 & @xmath332 & @xmath331 + @xmath349 & @xmath359 & @xmath360 & @xmath361 & @xmath362 & @xmath363 & @xmath364 & @xmath365 & @xmath366 + @xmath367 & @xmath325 & @xmath353 & @xmath354 & @xmath355 & @xmath356 & @xmath357 & @xmath358 & @xmath270 + @xmath348 & @xmath7 & @xmath368 & @xmath368 & @xmath368 & @xmath368 & @xmath368 & @xmath7 & @xmath369 + @xmath350 & @xmath370 & @xmath370 & @xmath369 & @xmath371 & @xmath372 & @xmath338 & @xmath338 & @xmath338 + @xmath349 & @xmath373 & @xmath374 & @xmath375 & @xmath376 & @xmath377 & @xmath378 & @xmath379 & @xmath380 in table [ tab : lambda=0 ] , the first two lines of @xmath381 correspond to the results of @xmath346 case , while the last three lines of @xmath382 to the case of @xmath347 . in the absence of the gauged u(1)@xmath3
i.e. in the @xmath346 case , @xmath349s exceed 4.5 throughout the range of @xmath13 .
hence , if one takes @xmath383 to ameliorate the fine - tuning problem in the higgs sector , @xmath6 should be greater than @xmath348 for the observed 126 gev higgs mass , and so it diverges below the gut scale . on the contrary , if the u(1)@xmath3 gauge interaction is turned on , it is possible to elude the fine - tuning and lp problems , explaining the observed higgs mass in the large @xmath8 range , @xmath384 and @xmath385 depending on @xmath8 . for @xmath386
, however , @xmath350 exceeds @xmath348 .
table [ tab : globalu(1 ) ] shows the results of the case , in which both the @xmath4 and @xmath6 couplings in eq .
( [ superpot ] ) are turned on , but the u(1)@xmath3 is not gauged [ or @xmath346 in eq .
( [ rgeq ] ) ] .
@xmath348 is defined as the maximal value of @xmath6 at low energy ( @xmath285 ) such that _ all _ the yukawa couplings considered here , @xmath387 do not blow up below the gut scale for a given low energy value of @xmath4 .
because of the lp constraint , @xmath6 could not be large enough in this case .
accordingly , @xmath246 and @xmath243 should be excessively larger than 4.5 in most parameter space as shown in table [ tab : globalu(1 ) ] in order to account for the 126 gev higgs mass .
it means that @xmath388 ( or @xmath160 ) and @xmath389 ( or @xmath221 ) need to be quite large , violating eq .
( [ ft ] ) .
c||ccc||ccc||ccc||ccc & & @xmath323 & & & @xmath325 & & & @xmath353 & & & @xmath355 & + @xmath4 & @xmath390 & @xmath59 & @xmath61 & @xmath390 & * @xmath391{0.6}$ ] * & @xmath61 & @xmath390 & @xmath59 & @xmath61 & @xmath390 & @xmath59 & @xmath61 + @xmath348 & @xmath392 & @xmath393 & @xmath394 & @xmath395 & @xmath329 & @xmath396 & @xmath395 & @xmath397 & @xmath398 & @xmath337 & @xmath399 & @xmath335 + @xmath349 & @xmath400 & @xmath401 & @xmath394 & @xmath402 & @xmath403 & @xmath404 & @xmath405 & @xmath406 & @xmath407 & @xmath408 & @xmath409 & @xmath410 + & & @xmath356 & & & @xmath357 & & & @xmath358 & & & @xmath270 & + @xmath4 & @xmath411 & @xmath412 & @xmath413 & @xmath411 & @xmath412 & @xmath413 & @xmath411 & @xmath412 & @xmath413 & @xmath411 & @xmath412 & @xmath413 + @xmath348 & @xmath338 & @xmath332 & @xmath331 & @xmath332 & @xmath332 & @xmath330 & @xmath332 & @xmath331 & @xmath337 & @xmath331 & @xmath330 & @xmath399 + @xmath349 & @xmath363 & @xmath414 & @xmath415 & @xmath364 & @xmath416 & @xmath417 & @xmath365 & @xmath414 & @xmath418 & @xmath366 & @xmath406 & @xmath419 although we present only the results for @xmath420 , @xmath59 , @xmath61 for @xmath421 in table [ tab : globalu(1 ) ] , @xmath280 turn out to be much larger than @xmath422 for @xmath423 and @xmath424 .
for @xmath425 , @xmath349 monotonically increases with @xmath4 .
however , @xmath349 already exceeds 4.5 even for @xmath320 .
note that the results for @xmath320 when @xmath425 reproduce the corresponding results of @xmath346 in table [ tab : lambda=0 ] . however , the @xmath6 coupling is definitely helpful for raising the radiative higgs mass , and so the consistent parameter space in this case should become broader .
we note that the sign of @xmath280 should flip between @xmath390 ( @xmath59 ) and @xmath59 ( @xmath61 ) for @xmath323 ( @xmath426 ) .
hence @xmath427 at some point between them .
@xmath427 is effectively equivalent to the case of @xmath317 ( @xmath428 , of course ) discussed in table [ tab : y_n=0 ] at low energy in explaining the higgs mass , because the left hand side of eq .
( [ 126higgs ] ) vanish anyway in the both cases , @xmath429 . comparing with @xmath344s of table [ tab : y_n=0 ] , thus
, we can expect that @xmath430 at @xmath431 ( 0.67 ) for @xmath323 ( @xmath426 ) . since now we have the @xmath6 coupling and @xmath221 term , which are helpful for explaining the observed higgs mass , even @xmath4 smaller than @xmath344 can be permitted .
actually , the following parameter ranges meet the perturbative constraint up to the gut scale , explaining the observed higgs mass : @xmath432 which means that the consistent _ points _ in table [ tab : y_n=0 ] become narrow _ bands _ in the parameter space , when the @xmath6 coupling and @xmath221 term are introduced . around the lower bounds of @xmath4 in eq .
( [ globalu(1)_lambda ] ) , it turns out that @xmath6 should be restricted to @xmath335 , @xmath433 , @xmath329 , @xmath397 for @xmath434 , respectively , since the upper bound coming from the lp constraint and the lower bound for the 126 gev higgs mass get to merge together . on the contrary , around the upper bounds of @xmath4 in eq .
( [ globalu(1)_lambda ] ) , @xmath6 is constrained only by lp , because the 126 gev higgs mass has been already explained with the the maximal @xmath4s and so the radiative higgs mass correction by @xmath435 proportional to @xmath436 should vanish at low energy .
thus , we have just the ( trivial ) constraints , @xmath437 for @xmath434 , respectively , around the upper bound of @xmath4 .
table [ tab : localu(1 ) ] presents the results when not only the @xmath4 , @xmath6 terms , but also the u(1)@xmath3 gauge symmetry are introduced particularly with @xmath347 . since the u(1)@xmath3 gauge coupling , @xmath103 monotonically increase with energy and @xmath438 , the perturbativity of u(1)@xmath3 gauge interaction is guaranteed throughout the energy range from the ew to the gut scale .
then , the negative contribution by u(1)@xmath3 gauge interactions could make the lp constraint on @xmath6 remarkably relaxed , as mentioned above .
although one takes a more larger value of @xmath128 , e.g. @xmath439 , which would be almost the maximal value of @xmath128 to maintain the perturbativity of u(1)@xmath3 gauge interaction at the gut scale , it turns out that a conspicuous improvement of the allowed parameter space is not achieved . in table
[ tab : localu(1 ) ] , @xmath348 again indicates the maximally allowed value of @xmath6 for a given @xmath4 at low energy : only if @xmath6 is smaller than @xmath348 around 1 tev energy scale , any yukawa couplings considered here do not reach the perturbativity bound ( @xmath440 ) below the gut scale . as in the previous tables ,
@xmath349 stands for the value of @xmath246 or @xmath243 required for explaining the 126 gev higgs mass , when the corresponding @xmath348 is taken .
@xmath350 means the value of @xmath316 needed for explaining the observed higgs mass when @xmath351 . in table
[ tab : localu(1 ) ] , the @xmath4s satisfying both the lp and higgs mass constraints are written inside the boxes .
we note that the allowed range of @xmath4 is remarkably enlarged particularly for larger values of @xmath441 .
[ actually , the results of @xmath354 show a similar pattern to the case of @xmath355 , even if they are not displayed in table [ tab : localu(1 ) ] . ]
thus , we have @xmath442 respectively , and roughly @xmath443 depending on @xmath8 .
note that the lists of @xmath320 are coincident with the results of @xmath347 in table [ tab : lambda=0 ] .
comparing with the parameter range of the nmssm , eq .
( [ nmssmpara ] ) , much larger values of @xmath8 are also allowed , and the lower bound of @xmath4 is remarkably relieved .
in particular , the lower bound of @xmath4 disappears for @xmath444 .
it is because the @xmath6 and @xmath221 terms of @xmath435 significantly raise the radiative higgs mass particularly for large @xmath8 .
moreover , the u(1)@xmath3 gauge interaction makes it possible that their contributions are further enhanced .
we note that the sign of @xmath349 is flipped between @xmath445 and @xmath61 for @xmath325 in table [ tab : localu(1 ) ] .
one can expect that @xmath349 vanishes at a point between them , which is effectively equivalent to @xmath317 of table [ tab : y_n=0 ] in explaining the higgs mass , because @xmath430 in the both cases . by performing a similar analysis to eq .
( [ globalu(1)_lambda ] ) , we get the following results : @xmath446 which are wider than the former results in eq .
( [ globalu(1)_lambda ] ) , because of the gauged u(1)@xmath3 . around the lower bounds of @xmath4 in eq .
( [ localu(1)_lambda ] ) , @xmath6 should be restricted to @xmath397 , @xmath372 , @xmath447 , @xmath371 for @xmath434 , respectively , and @xmath448 , while @xmath317 around the upper bounds .
the upper bounds of eq .
( [ localu(1)_lambda ] ) should coincide with those of eq .
( [ globalu(1)_lambda ] ) and also the results of table [ tab : y_n=0 ] . between the lower and upper bounds of @xmath4 ,
sizable intervals of @xmath6 can be allowed : e.g. for @xmath325 and @xmath445 ( @xmath449 ) , the permitted range of @xmath6 is given by @xmath450 @xmath451 ( @xmath452 ) , as seen from @xmath348 and @xmath350 of @xmath445 in table [ tab : localu(1 ) ] . for @xmath326 and @xmath453 ( @xmath449 ) ,
the allowed range of @xmath6 turns out to be @xmath454 @xmath455 ( @xmath452 ) . for the smaller @xmath8 ,
the intervals are relatively narrower .
however , they all should rapidly shrink to @xmath317 around the upper bounds of @xmath4 in eq .
( [ localu(1)_lambda ] ) , which correspond to the original nmssm limit at low energy .
c||ccccccccc & & & & & @xmath325 & & & & + @xmath4 & @xmath411 & @xmath456 & @xmath412 & @xmath457 & @xmath413 & * @xmath391{0.5}$ ] * & * @xmath391{0.6}$ ] * & @xmath61 & @xmath331 + @xmath348 & @xmath7 & @xmath7 & @xmath369 & @xmath458 & @xmath371 & @xmath447 & @xmath372 & @xmath335 & @xmath459 + @xmath350 & @xmath370 & @xmath370 & @xmath370 & @xmath370 & @xmath370 & @xmath447 & @xmath450 & @xmath370 & @xmath370 + @xmath349 & @xmath373 & @xmath460 & @xmath461 & @xmath462 & @xmath463 & @xmath464 & @xmath465 & @xmath466 & @xmath394 + & & & & & @xmath353 & & & & + @xmath4 & @xmath411 & @xmath456 & @xmath412 & @xmath457 & * @xmath391{0.4}$ ] * & * @xmath391{0.5}$ ] * & * @xmath391{0.6}$ ] * & * @xmath391{0.7}$ ] * & @xmath331 + @xmath348 & @xmath368 & @xmath7 & @xmath7 & @xmath369 & @xmath458 & @xmath371 & @xmath467 & @xmath468 & @xmath469 + @xmath350 & @xmath370 & @xmath370 & @xmath370 & @xmath370 & @xmath458 & @xmath467 & @xmath337 & @xmath303 & @xmath370 + @xmath349 & @xmath374 & @xmath470 & @xmath471 & @xmath472 & @xmath473 & @xmath474 & @xmath475 & @xmath476 & @xmath477 + & & & & & @xmath355 & & & & + @xmath4 & * @xmath391{0.0}$ ] * & * @xmath391{0.1}$ ] * & * @xmath391{0.2}$ ] * & * @xmath391{0.3}$ ] * & * @xmath391{0.4}$ ] * & * @xmath391{0.5}$ ] * & * @xmath391{0.6}$ ] * & @xmath61 & @xmath331 + @xmath348 & @xmath368 & @xmath368 & @xmath368 & @xmath7 & @xmath369 & @xmath371 & @xmath452 & @xmath450 & @xmath478 + @xmath350 & @xmath371 & @xmath447 & @xmath447 & @xmath452 & @xmath467 & @xmath338 & @xmath331 & @xmath370 & @xmath370 + @xmath349 & @xmath376 & @xmath479 & @xmath480 & @xmath481 & @xmath482 & @xmath482 & @xmath483 & @xmath484 & @xmath485 + & & & & & @xmath356 & & & & + @xmath4 & * @xmath391{0.0}$ ] * & * @xmath391{0.1}$ ] * & * @xmath391{0.2}$ ] * & * @xmath391{0.3}$ ] * & * @xmath391{0.4}$ ] * & * @xmath391{0.5}$ ] * & * @xmath391{0.6}$ ] * & @xmath61 & @xmath331 + @xmath348 & @xmath368 & @xmath368 & @xmath7 & @xmath7 & @xmath369 & @xmath371 & @xmath467 & @xmath398 & @xmath394 + @xmath350 & @xmath372 & @xmath372 & @xmath372 & @xmath338 & @xmath338 & @xmath338 & @xmath332 & @xmath370 & @xmath394 + @xmath349 & @xmath377 & @xmath486 & @xmath487 & @xmath475 & @xmath488 & @xmath489 & @xmath490 & @xmath491 & @xmath394 + & & & & & @xmath357 & & & & + @xmath4 & * @xmath391{0.0}$ ] * & * @xmath391{0.1}$ ] * & * @xmath391{0.2}$ ] * & * @xmath391{0.3}$ ] * & * @xmath391{0.4}$ ] * & * @xmath391{0.5}$ ] * & * @xmath391{0.6}$ ] * & @xmath61 & @xmath331 + @xmath348 & @xmath368 & @xmath368 & @xmath7 & @xmath369 & @xmath458 & @xmath447 & @xmath372 & @xmath492 & @xmath394 + @xmath350 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath332 & @xmath370 & @xmath394 + @xmath349 & @xmath378 & @xmath493 & @xmath494 & @xmath495 & @xmath496 & @xmath497 & @xmath498 & @xmath499 & @xmath394 + & & & & & @xmath358 & & & & + @xmath4 & * @xmath391{0.0}$ ] * & * @xmath391{0.1}$ ] * & * @xmath391{0.2}$ ] * & * @xmath391{0.3}$ ] * & * @xmath391{0.4}$ ] * & * @xmath391{0.5}$ ] * & @xmath59 & @xmath61 & @xmath331 + @xmath348 & @xmath7 & @xmath7 & @xmath7 & @xmath369 & @xmath371 & @xmath452 & @xmath450 & @xmath394 & @xmath394 + @xmath350 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath332 & @xmath370 & @xmath394 & @xmath394 + @xmath349 & @xmath379 & @xmath494 & @xmath500 & @xmath501 & @xmath502 & @xmath476 & @xmath503 & @xmath394 & @xmath394 + & & & & & @xmath270 & & & & + @xmath4 & * @xmath391{0.0}$ ] * & * @xmath391{0.1}$ ] * & * @xmath391{0.2}$ ] * & * @xmath391{0.3}$ ] * & * @xmath391{0.4}$ ] * & @xmath390 & @xmath59 & @xmath61 & @xmath331 + @xmath348 & @xmath369 & @xmath369 & @xmath458 & @xmath371 & @xmath452 & @xmath303 & @xmath394 & @xmath394 & @xmath394 + @xmath350 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath338 & @xmath370 & @xmath394 & @xmath394 & @xmath394 + @xmath349 & @xmath380 & @xmath504 & @xmath488 & @xmath505 & @xmath506 & @xmath507 & @xmath394 & @xmath394 & @xmath394
the observed higgs mass and the lp constraint seriously restrict the valid ranges of @xmath4 and @xmath8 in the nmssm , leaving only the narrow bands , @xmath9 and @xmath10 . here
the lower bound of @xmath4 and the upper bound of @xmath8 result from the 126 gev higgs mass . in order to relieve such severe bounds , we extended the nmssm with the vector - like superfields @xmath212 , and studied their coupling with the mssm higgs doublet , @xmath508 .
we introduced also a u(1 ) gauge symmetry , under which only the extra vector - like superfields are charged , but all the ordinary nmssm superfields remain neutral . with the help of such a u(1 )
gauge symmetry , the allowed value of @xmath6 at low energy can be lifted up to @xmath7 , evading a lp below the gut scale .
the @xmath6 term and the holomorphic soft terms can remarkably raise the radiative higgs mass particularly for large values of @xmath8 .
consequently , they invalidate the previous lower bound of @xmath4 and the upper bound of @xmath8 , significantly enlarging the valid parameter space . in particular ,
the lower bound of @xmath4 is completely removed for @xmath444 .
thus , we have @xmath509-@xmath59 for @xmath510 as a consistent parameter space , while @xmath413-@xmath511-@xmath61 for @xmath512 , and roughly @xmath513 , depending on @xmath8 . for @xmath514 ,
the effects coming from the extra matter become weaker , and so we have just a limited parameter range , @xmath515 .
however , the original nmssm parameter space should be contained in our case , and so relatively smaller @xmath6s in @xmath516 are also possible in small @xmath8 cases .
i thank jihn e. kim for valuable discussion .
this research is supported by basic science research program through the national research foundation of korea ( nrf ) funded by the ministry of education , grant no . 2013r1a1a2006904 .
for a review , for instance , see m. drees , r. godbole and p. roy , `` theory and phenomenology of sparticles : an account of four - dimensional n=1 supersymmetry in high energy physics , '' hackensack , usa : world scientific ( 2004 ) 555 p. references are therein .
m. s. carena and h. e. haber , prog .
* 50 * ( 2003 ) 63 [ hep - ph/0208209 ] .
see also a. djouadi , phys .
* 459 * ( 2008 ) 1 [ hep - ph/0503173 ] .
atlas collaboration , atlas - conf-2013 - 024 ; s. chatrchyan _ et al . _
[ cms collaboration ] , arxiv:1308.1586 [ hep - ex ] .
h. abe , t. kobayashi and y. omura , phys .
d * 76 * ( 2007 ) 015002 [ hep - ph/0703044 [ hep - ph ] ] ; d. horton and g. g. ross , nucl .
b * 830 * ( 2010 ) 221 [ arxiv:0908.0857 [ hep - ph ] ] ; j. e. younkin and s. p. martin , phys . rev .
d * 85 * ( 2012 ) 055028 [ arxiv:1201.2989 [ hep - ph ] ] ; h. abe , j. kawamura and h. otsuka , ptep * 2013 * ( 2013 ) 013b02 [ arxiv:1208.5328 [ hep - ph ] ] ; i. gogoladze , f. .nasir and q. .shafi , int .
j. mod .
a * 28 * ( 2013 ) 1350046 [ arxiv:1212.2593 [ hep - ph ] ] . for other types of singlet extensions of the mssm , see , for instance , a. delgado , c. kolda , j. p. olson and a. de la puente , phys .
lett . * 105 * , 091802 ( 2010 ) [ arxiv:1005.1282 [ hep - ph ] ] ; g. g. ross and k. schmidt - hoberg , nucl .
b * 862 * , 710 ( 2012 ) [ arxiv:1108.1284 [ hep - ph ] ] ; b. kyae and j. -c . park , phys .
d * 86 * ( 2012 ) 031701 [ arxiv:1203.1656 [ hep - ph ] ] ; b. kyae and j. -c . park , phys . rev .
d * 87 * ( 2013 ) 075021 [ arxiv:1207.3126 [ hep - ph ] ] .
m. masip , r. munoz - tapia and a. pomarol , phys .
d * 57 * ( 1998 ) r5340 [ hep - ph/9801437 ] . for more recent discussions ,
see also r. barbieri , l. j. hall , a. y. papaioannou , d. pappadopulo and v. s. rychkov , jhep * 0803 * ( 2008 ) 005 [ arxiv:0712.2903 [ hep - ph ] ]
. for early studies on vector - like matter , see t. moroi and y. okada , mod .
lett . a * 7 * ( 1992 ) 187 ; t. moroi and y. okada , phys .
b * 295 * ( 1992 ) 73 ; k. s. babu , i. gogoladze and c. kolda , hep - ph/0410085 ; k. s. babu , i. gogoladze , m. u. rehman and q. shafi , phys .
d * 78 * ( 2008 ) 055017 [ arxiv:0807.3055 [ hep - ph ] ] .
m. endo , k. hamaguchi , s. iwamoto , n. yokozaki and , phys .
d * 84 * ( 2011 ) 075017 ; [ arxiv:1108.3071 [ hep - ph ] ] .
t. moroi , r. sato and t. t. yanagida , phys .
b * 709 * ( 2012 ) 218 .
[ arxiv:1112.3142 [ hep - ph ] ] ; k. j. bae , t. h. jung and h. d. kim , phys .
d * 87 * ( 2013 ) 015014 [ arxiv:1208.3748 [ hep - ph ] ] ; w. -z .
feng and p. nath , phys .
d * 87 * ( 2013 ) 075018 [ arxiv:1303.0289 [ hep - ph ] ] .
s. m. barr , phys . lett .
b * 112 * ( 1982 ) 219 ; j. p. derendinger , j. e. kim and d. v. nanopoulos , phys .
b * 139 * ( 1984 ) 170 ; i. antoniadis , j. r. ellis , j. s. hagelin and d. v. nanopoulos , phys .
b * 194 * ( 1987 ) 231 .
for instance , see j. -h .
huh , j. e. kim and b. kyae , phys .
d * 80 * ( 2009 ) 115012 [ arxiv:0904.1108 [ hep - ph ] ] ; j. e. kim , j. -h .
kim and b. kyae , jhep * 0706 * ( 2007 ) 034 [ hep - ph/0702278 [ hep - ph ] ] .
see also j. e. kim and b. kyae , nucl .
b * 770 * ( 2007 ) 47 [ hep - th/0608086 ] ; j. e. kim and b. kyae , phys .
d * 77 * ( 2008 ) 106008 [ arxiv:0712.1596 [ hep - th ] ] ; k. -s .
choi and b. kyae , nucl .
b * 855 * ( 2012 ) 1 [ arxiv:1102.0591 [ hep - th ] ] .
m. s. carena , m. quiros and c. e. m. wagner , nucl .
b * 461 * ( 1996 ) 407 .
p. langacker , g. paz , l. -t .
wang and i. yavin , phys .
* 100 * ( 2008 ) 041802 [ arxiv:0710.1632 [ hep - ph ] ] ; p. langacker , g. paz , l. -t .
wang and i. yavin , phys .
d * 77 * ( 2008 ) 085033 [ arxiv:0801.3693 [ hep - ph ] ] . | we introduce one pair of inert higgs doublets @xmath0 and singlets @xmath1 , and consider their couplings with the higgs doublets of the minimal supersymmetric standard model ( mssm ) , @xmath2 .
we assign extra u(1)@xmath3 gauge charges only to the extra vector - like superfields , and so all the mssm superfields remain neutral under the new u(1)@xmath3 .
they can be an extension of the `` @xmath4 term , '' @xmath5 in the next - to - mssm ( nmssm ) . due to the u(1)@xmath3 ,
the maximally allowed low energy value of @xmath6 can be lifted up to @xmath7 , avoiding a landau - pole ( lp ) below the grand unification scale .
such colorless vector - like superfields remarkably enhance the radiative mssm higgs mass particularly for large @xmath8 through the @xmath6 term and the corresponding holomorphic soft term . as a result
, the lower bound of @xmath4 and the upper bound of @xmath8 can be relaxed to disappear from the restricted parameter space of the original nmssm , @xmath9 and @xmath10 .
thus , the valid parameter space significantly expands up to @xmath11 , @xmath12 , and @xmath13 , evading the lp problem and also explaining the 126 gev higgs mass naturally . |
metastability is a broad term describing the existence of a very sensitive equilibrium , possessing a weak form of stability / instability .
usually , such behavior is related to the presence of a small first eigenvalue for the linearized operator at the given equilibrium state , revealed at dynamical level by the appearance of slowly moving structures .
such circumstance comes into view in the analysis of different classes of evolutive pdes , and it has been object of a wide amount of studies , covering many different areas . among others ,
we emphasize the explorations on the allen cahn equation , started in @xcite , and the investigations on the cahn hilliard equation , with the fundamental contributions @xcite .
the analysis has been continued by many other scholars by means of a broad spectrum of techniques , and extended to a number of different models such as the gierer meinhardt and gray
scott systems ( see @xcite ) , keller
segel chemotaxis system ( see @xcite ) , general gradient flows ( see @xcite ) and many others .
the number of references is so vast that it would be impossible to mention all the contributions given in the area .
a pionereeing article in the analysis of slow dynamics for parabolic equations has been authored by g. kreiss and h .- o .
kreiss @xcite and concerns with the scalar viscous conservation law @xmath6 with the space variable @xmath7 belonging to a one - dimensional interval @xmath8 , @xmath9 .
the primary prototype for the flux function @xmath10 is given by the classical quadratic formula @xmath11 , so that partial differential equation in becomes the so - called _ ( viscous ) burgers equation_. the parameter @xmath12 is small .
problem is complemented with dirichlet boundary conditions @xmath13 for given data @xmath14 to be discussed in details .
burgers equation is considered as a ( simplified ) archetype of more complicate systems of partial differential equations arising in different fields of applied mathematics .
inspired by the equations of fluid - dynamics , the parameter @xmath15 is interpreted as a _ viscosity coefficient _ and the main problem is to identify and quantify its rle in the emergence and/or disappearance of structures .
formally , in the limit @xmath16 , the initial value problem reduces to a first - order quasi - linear equation of hyperbolic type @xmath17 whose standard setting is given by the _ entropy formulation_. hence solutions may have discontinuities , which propagate with speed @xmath18 such that @xmath19\!]}={[\![}f(u){]\!]}\qquad\qquad \textrm{({\sf rankine -- hugoniot relation})}\ ] ] and satisfy appropriate entropy conditions ( here @xmath20\!]}$ ] denotes the jump ) .
in addition , the treatment of the boundary conditions is more delicate than the parabolic case , because of the eventual appearance of boundary layers , @xcite .
concerning the flux function @xmath10 , let us assume that , for some @xmath21 , @xmath22 where @xmath23 are the boundary data prescribed in .
the last two assumptions guarantee that a jump with left value @xmath24 and right value @xmath25 satisfy the entropy condition and has speed of propagation equal to zero , as dictated by the rankine hugoniot relation . therefore
, the one - parameter family of functions @xmath26 defined by @xmath27 ( where @xmath28 denotes the characteristic function of the set @xmath29 ) is composed by stationary solutions of the equation in satisfying the boundary conditions .
the dynamics determined by boundary - initial value problem - is simple : for any datum @xmath30 with bounded variation , the solution converges _ in finite time _ to an element of @xmath26 ( see section [ sect : appl ] ) .
hence , at the level @xmath31 , there are infinitely many stationary solutions , generating a `` finite - time '' attracting manifold for the dynamics .
.15 cm for @xmath12 , the situation is different .
apart from the well - known smoothing effect , the presence of the laplace operator in has the effect of a drastic reduction of the number of stationary solutions satisfying : from infinitely many to a single stationary state ( see section [ sect : appl ] ) . such solution , denoted here by @xmath32 , converges in the limit @xmath16 to a specific element @xmath33 of the family @xmath26 .
the dynamical properties of for initial data close to the equilibrium configuration @xmath34 can be analyzed linearizing at the state @xmath34 @xmath35 in @xcite it shown that , in the case of burgers flux @xmath11 , the eigenvalues of @xmath36 with homogeneous dirichlet boundary conditions , are real and negative . moreover , as a consequence of the requiremente @xmath37 , there holds as @xmath38 @xmath39 for some @xmath21 independent on @xmath15 .
negativity of the eigenvalues implies that the steady state @xmath34 is asymptotically stable with exponential rate ; the precise description of the eigenvalue distribution shows that the large time behavior is described by term of the order @xmath40 and thus the convergence is very slow when @xmath15 is small to quantify the reduction order of the mapping @xmath41 , note that @xmath42 has order @xmath43 for @xmath44 and order @xmath45 for @xmath46 .
such is the picture relative to the behavior determined by an initial data close to the equilibrium solution @xmath34 .
the next question concerns with the dynamics generated by initial data presenting a sharp transition from @xmath47 to @xmath48 localized far from the position of the steady state @xmath34 .
[ fig1 ] , @xmath49 and @xmath50.,width=340 ] figure [ fig1 ] represents a numerical simulation of the solution to the initial value problem with boundary conditions , relative to the initial condition @xmath50 . starting with a decreasing initial datum
, a shock layer is formed in a short time scale , so that the solution is approximately given by a translation of the ( unique ) stationary solution of the problem .
once such a layer is formed , on a longer time scale , it moves towards the location corresponding to the equilibrium solution .
this article deals with the dynamics after the shock layer formation for @xmath15 small . in order to provide a detailed description of such regime , with special attention to the relation between the unviscous and the low - viscosity behavior , it is rational : +
to build up a one - parameter family of functions @xmath51 such that @xmath52 as @xmath38 , in an appropriate sense ; + to describe the dynamics of the solution to the initial - boundary value problem in a tubular neighborhood of the family @xmath51 .
a specific element @xmath53 of the manifold @xmath54 corresponds to the steady state @xmath34 of and the dynamics will asymptotically lead to such configuration .
.15 cm before describing in details the contribution of the paper , let us recast the state of the art on the topic . among others , the problem of slow dynamics for the burgers equation has been examined in @xcite and in @xcite , where different approaches have been considered .
the former is based either on _ projection method _ or on _ wkb expansions _ ; the latter stands on an adapted version of the _ method of matched asymptotics expansion_. the common aim is to determine an expression and/or an equation for the parameter @xmath55 , considered as a function of time , describing the movement of the transition from a generic point of the interval @xmath56 toward the equilibrium location @xmath57 . in both the contributions , the analysis is conducted at a formal level and validated numerically by means of comparision with significant computations . a rigorous analysis has been performed in @xcite ( and generalized to the case of nonconvex flux in @xcite ) , where one - parameter family of reference functions is chosen as a family of traveling wave solutions to the viscous equation satisfying the boundary conditions and with non - zero ( but small ) velocity .
the approach is based on the use of such traveling waves to obtain upper and lower estimates by the maximum principle , from which rigorous asymptotic formulae for the slow velocity are obtained .
slow motion for the viscous burgers equation in unbounded domains has been also considered in literature . in @xcite , it is analyzed the case of the half - line @xmath58 for the space variable @xmath7 , with constant initial and boundary data chosen so that speed of the shock generated at @xmath59 is stationary for the corresponding hyperbolic equation .
the presence of the viscosity generates a motion of the transition layer , which is precisely identified by means of the lambert s @xmath60 function .
later , the ( slow ) motion of a shock wave , with zero hyperbolic speed , for the burgers equation in the quarter plane has been considered in @xcite , where it is shown that the location of the wave front is of order @xmath61 ; the same result has been generalized in @xcite in the case of general fluxes ( for other contributions to the same problem , we refer also to @xcite ) .
the case of the whole real line has been examined in @xcite with emphasis on the generation of @xmath62wave like structures and their evolution towards nonlinear diffusion waves .
the analysis is based on the use of self - similar variables , suggested by the invariance of the burgers equation under the group of transformations @xmath63 ( for subsequent contributions in the same direction , see @xcite ) .
more recently , it has been shown in @xcite that the slow motion is determined by the presence of a one - dimensional center manifold of steady states for the equation in the self - similar variables ( corresponding to the diffusion waves ) and a relative family of one - dimensional global attractive invariant manifold . in a short - time scale , the solution approaches one of the attractive manifolds and remains close to it in a long - time scale .
at the present day , results relative to metastability in the case of systems appear to be rare .
slow dynamics analysis for systems of conservation laws have been considered in @xcite , basic model examples being the navier - stokes equations of compressible viscous heat conductive fluid and the keyfitz - kranzer system , arising in elasticity .
the approach is based on asymptotic expansions and consists in deriving appropriate limiting equations for the leading order terms , in the case of periodic data . in @xcite , the problem of proving convergence to a stationary solution for a system of conservation laws with viscosity
is addressed , with an approach based on a detailed analysis of the linearized operator at the steady state .
a recent contribution is the reference @xcite , where the authors consider the saint - venant equations for shallow water and , precisely , the phenomenon of formation of roll - waves .
the approach merges together analytical techniques and numerical results to present some intriguing properties relative to the dynamics of solitary wave pulses .
.15 cm summing up , apart for the formal expansions methods , the rigorous approaches used in the literature are largely based on typical scalar equations features .
the first of these properties is the direct link between the scalar burgers equation and the heat equation given by the hopf cole transformation : @xmath64 , and the consequent invariance of the burgers equation under the group of scaling transformations @xmath63 .
on the one hand , the presence of such a connection is an evident advantage , since it permits to determine optimal descriptions for the behavior under study ( see @xcite ) ; on the other hand , to use such exceptional property makes the approach very stiff and difficult to apply to more general cases .
a different `` scalar hallmark '' is the base of the approach considered in @xcite , where the authors make wide use of maximum principle and comparison arguments , taking benefit from the fact that the equation is second - order parabolic . in order to extend the results to more general settings and specifically for systems of pdes
, it is useful to determine strategies and techniques that are more flexible , paying , if necessary , the price of a less accurate description of the dynamics .
a contribution in this direction has been given in @xcite , where the location of the shock transition for a scalar conservation law in the quarter plane has been proved by means of weighted energy estimates , extending the result proved in @xcite , that used an explicit formula determined by means of the hopf
cole transformation for the green function of the linearization at the shock profile of the burgers equation . .15
cm the present article intends to contribute to the definition of a versatile language for metastability , suitable for general class of partial differential equations of evolutive type . with this direction in mind
, we follow an approach that it is strictly related with the _ projection method _ considered in @xcite and we go behind the philosophy tracked in the analysis of stability of viscous shock waves by k.zumbrun and co - authors ( see @xcite ) .
precisely , we separate three distinct phases : to choose a family of functions @xmath65 , considered as approximate solutions , and to measure how far they are from being exact solutions ; + to investigate spectral properties of the linearized operators at such states ; + to show that appropriate assumptions on the approximate solutions ( step * i * ) and on the spectrum of the linearized operators ( step * ii * ) imply the appearance of a metastable behavior . .15 cm with respect to the framework of shock waves stability analysis , there are two main differences .
first of all , we concentrate on the case of bounded domains and , therefore , the spectrum of the linearized operators is discrete . additionally , since the reference states @xmath66 are approximate solutions , the perturbations of such states satisfy at first order a _ non - homogeneous _ linear equation , with forcing term negligible as @xmath16 .
the defect of working in a neighborhood of a manifold that is not invariant has the counterpart of a wider flexibility in its construction that leads , in particular , to ( more or less ) explicit representations .
thus , it should be possible in principle to obtain numerical evidence of special spectral properties even in cases where analytical results appear to be not achievable .
the article is organized as follows . to start with , in section [ sect : abs ] , we consider a general framework containing scalar viscous conservation laws as a very specific case . given a family of approximate solutions @xmath67 ,
our approach consists in representing the solution to the initial - bondary value problem as the sum of an element @xmath68 moving along the family @xmath67 plus a perturbation term @xmath5 .
the equation for the unknown @xmath2 is chosen in such a way that the slower decaying terms in the perturbation @xmath5 are canceled out . in order to state a general result
, we consider an approximation of the complete nonlinear equations for the couple @xmath69 , obtained by disregarding quadratic terms in @xmath5 and keeping the nonlinear dependence on @xmath55 , in order to keep track of the nonlinear evolution along the manifold @xmath67 .
such reduced system for @xmath70 is called quasi - linearized system and it is the concern of theorem [ thm : metal ] , the main contribution of the paper . under appropriate assumptions on the manifold @xmath66 , the linearized operators at such states , and the coupling between the two objects , such result gives an explicit representation for the solution to the evolutive problem together with an estimate on the remainder , vanishing in the limit @xmath38 .
this gives a sound justification to the reduced equation for the unknown @xmath2 obtainable by neglecting also the linear term in @xmath5 .
dealing with the complete system for the couple @xmath70 brings into the analysis also the specific form of the quadratic terms . as a consequence , in case of parabolic systems of reaction - diffusion type , we expect that a results analogous to theorem [ thm : metal ] could be proved , under the assumption of an _ a priori _ @xmath71 bound on the solution .
differently , when a nonlinear first order space derivative term is present ( as is the case for viscous conservation laws ) , the quadratic term involve a dependence on the space derivative of the solution and a rigorous result needs an additional bound , which we are not presently able to achieve .
in section [ sect : appl ] we consider the application of the general framework to the case of viscous scalar conservation laws .
firstly , we present the dynamics of the hyperbolic equation obtained in the vanishing viscosity limit , proving a result on finite - time convergence to the one - parameter manifold of steady states ( theorem [ thm : stabhyp ] ) . then , we pass to consider the parabolic equation in under assumption and we build up a specific family @xmath67 by matching continuously stationary solutions at a given point @xmath55 . to apply the general result of section [ sect : abs ] , we need to measure how far are states @xmath66 from being stationary solutions , and this amounts in estimating the jump of the space derivative at the matching point .
such task is completed , showing that the residual has order @xmath72 , hence it is exponentially small in the limit @xmath16 . as a by - product , we deduce a formal equation for the motion of the shock layer , which generalizes the one known for the case of the burgers flux @xmath73 . in section
[ sect : spec ] , we analyze spectral properties of the diffusion - transport linear operator , arising from the linearization at the state @xmath74 .
we show that , under appropriate assumption on the limiting behavior of @xmath66 as @xmath16 , the spectrum can be decomposed into two parts : the first eigenvalue of order @xmath75 ; all of the remaining eigenvalues are less than @xmath76 ( where @xmath77 denotes a generic positive constant independent on @xmath15 ) .
additionally , precise asymptotics for the first eigenvalue are achieved by considering the linear operator with piecewise constant coefficient , obtained by taking the limit of functions @xmath74 as @xmath16 .
this analysis is needed to give evidence of the validity of the coupling assumption required in theorem [ thm : metal ] .
given @xmath9 , @xmath78 and @xmath79 , we consider the space @xmath80^n$ ] endowed with @xmath81 where @xmath82 denotes the usual scalar product in @xmath83 . given @xmath84 , we consider the evolutive cauchy problem for the unknown @xmath85 @xmath86,\qquad u\,\bigr|_{t=0}=u_0\ ] ] where @xmath87 denotes a nonlinear differential operator , complemented with appropriate boundary conditions .
we are interested in describing the dynamical behavior of @xmath88 , solution to , in the regime @xmath89 .
in particular , we have in mind the case of a singular dependence of @xmath87 with respect to @xmath15 , in the sense that the operator @xmath90 is of lower order with respect to @xmath87 .
the specific example , considered in detail in the subsequent sections , is the one - dimensional scalar viscous conservation laws with dirichlet boundary conditions ; at the same time , also the usual allen cahn parabolic equation fits into the framework . given a one - dimensional open interval @xmath91 ,
let @xmath92 be a one - parameter family in @xmath93 , whose elements can be considered as approximate stationary solutions to the problem in the sense that @xmath94 $ ] depends smoothly on @xmath15 and tends to @xmath95 as @xmath38 .
precisely , we assume that the term @xmath96 $ ] belongs to the dual space of the continuous functions space @xmath97 and there exists a family of smooth positive functions @xmath98 , uniformly convergent to zero as @xmath38 , such that , for any @xmath99 , there holds @xmath100\rangle| \leq \omega^\varepsilon(\xi)\,|\psi|_{{}_{\infty } } \qquad \forall\,\psi\in c(i).\ ] ] the family @xmath101 will be referred to as an approximate invariant manifold with respect to the flow determined by in @xmath93 .
generically , since an element @xmath102 is not a steady state for , the dynamics walk away from the manifold with a speed dictated by @xmath103 .
the dependence of @xmath103 on @xmath15 plays a relevant rle , since it drives the departure from the approximate invariant manifold .
next , we decompose the solution to the initial value problem as @xmath104 with @xmath105 and @xmath106^n$ ] to be determined . substituting , we obtain @xmath107 -\partial_{\xi}u^{\varepsilon}(\cdot;\xi)\,\frac{d\xi}{dt}+\mathcal{q}^\varepsilon[v,\xi]\ ] ] where @xmath108\,v\\ \mathcal{q}^\varepsilon[v,\xi]&:={\mathcal f}^\varepsilon[u^{\varepsilon}(\cdot;\xi)+v ] -{\mathcal f}^\varepsilon[u^{\varepsilon}(\cdot;\xi ) ] -d{\mathcal f}^\varepsilon[u^{\varepsilon}(\cdot;\xi)]\,v .
\end{aligned}\ ] ] next , we assume that the linear operator @xmath109 has a discrete spectrum composed by semi - simple eigenvalues @xmath110 with corresponding right eigenfunctions @xmath111 . denoting by @xmath112 the eigenfunctions of the adjoint operator @xmath113 and setting @xmath114 we can use the degree of freedom we still have in the choice of the couple @xmath70 in such a way that the component @xmath115 is identically zero , that is @xmath116 using equation , we infer @xmath117 - \partial_{\xi}u^{\varepsilon}(\cdot;\xi)\frac{d\xi}{dt}+{\mathcal q}^\varepsilon[v,\xi ] \rangle + \langle \partial_{\xi}\psi^\varepsilon_1(\xi,\cdot ) \frac{d\xi}{dt},v \rangle = 0\ ] ] since @xmath118 , we obtain a scalar differential equation for the variable @xmath55 , describing the reduced dynamics along the approximate manifold , that is @xmath119+\mathcal{q}^\varepsilon[v,\xi ] \rangle\ ] ] where @xmath120 together with the condition on the initial datum @xmath121 @xmath122 to rewrite equation in normal form in the regime of small @xmath5 , we assume @xmath123 for some @xmath21 independent on @xmath55 . such assumption gives a ( weak ) restriction on the choice of the members of the family @xmath1 asking for the manifold to be never transversal to the first eigenfunction of the corresponding linearized operator . from now on , we can renormalize the eigenfunction @xmath124 so that @xmath125 for any @xmath12 and for any @xmath99 . in the regime @xmath126 ,
we may expand @xmath127 as @xmath128 inserting in , we mat rewrite the nonlinear equation for @xmath55 as @xmath129,\ ] ] where @xmath130 \rangle}\\ \rho^\varepsilon[\xi , v]&:=\frac{1}{\alpha^\varepsilon(\xi , v ) } \bigl(\langle \psi^\varepsilon_1,\mathcal{q}^\varepsilon\rangle + \langle \partial_\xi\psi^\varepsilon_1 , v\rangle^2\bigr ) . \end{aligned}\ ] ] using , equation can be rephrased as @xmath131\ ] ] where @xmath132 -\partial_{\xi}u^{\varepsilon}(\cdot;\xi)\,\theta^\varepsilon(\xi),\\ { \mathcal m}^\varepsilon_\xi v&:=-\partial_{\xi}u^{\varepsilon}(\cdot;\xi ) \,\theta^\varepsilon(\xi)\,\langle\partial_{\xi } \psi^\varepsilon_1 , v \rangle,\\ \mathcal{r}^\varepsilon[v,\xi]&:=\mathcal{q}^\varepsilon[v,\xi ] -\partial_{\xi}u^{\varepsilon}(\cdot;\xi)\,\rho^\varepsilon[\xi , v].\end{aligned}\ ] ] let us stress that , by definition , there holds @xmath133 so that @xmath134 is the projection of @xmath94 $ ] onto the space orthogonal to @xmath135 .
summarizing , the couple @xmath70 solves the differential system - where the initial condition @xmath121 for @xmath55 is such that @xmath136 and the initial condition @xmath137 for @xmath5 is given by @xmath138 . neglecting the @xmath139 order terms
, we obtain the system @xmath140 with initial conditions @xmath141 from now on , we will refer to this system as the quasi - linearization of . our aim is to describe the behavior of the solution to in the regime of small @xmath15 .
shortly , the quasi - linearized system is determined by an appropriate combination of the term @xmath96 $ ] , measuring how far is the function @xmath66 from being a stationary solution , and the linear operator @xmath142 , controlling at first order how solutions to depart from @xmath66 when the latter is taken as initial datum . to state our first result , we need to precise the assumption on such terms . .15 cm * h1 . *
the family @xmath143 is such that @xmath96 $ ] belongs to the dual space of @xmath144 and there exists functions @xmath103 such that , denoting again with @xmath145 the duality relation , @xmath146\rangle| \leq \omega^\varepsilon(\xi)\,|\psi|_{{}_{\infty } } \qquad \forall\,\psi\in c(i).\ ] ] with @xmath103 converging to zero as @xmath38 , uniformly with respect to @xmath99 .
.15 cm * h2 . *
the eigenvalues @xmath147 of @xmath109 are semi - simple , @xmath148 is simple , real and negative , and @xmath149 for some constant @xmath150 independent on @xmath151 , @xmath12 and @xmath99 .
.15 cm * h3 . * the eigenfunctions @xmath152 and @xmath153 of @xmath154 and @xmath155 normalized so that @xmath156 where @xmath157 is the usual kronecker symbol , are such that @xmath158 for some constant @xmath77 independent on @xmath12 and @xmath99 . the last assumption we require relate the term @xmath159 to the first eigenvalue @xmath160 of the linearized operator @xmath161 at @xmath74 .
formally , if @xmath162 is an exact stationary solution , then @xmath163 = \mathcal{f}[u^\varepsilon(\cdot;\xi)]-\mathcal{f}[u^\varepsilon(\cdot;\bar\xi ) ] \approx\mathcal{l}_\xi^\varepsilon\partial_\xi u^\varepsilon(\cdot;\bar\xi)(\bar\xi-\xi).\ ] ] if @xmath164 is chosen to be approximately close to the first eigenfunction of @xmath161 , then @xmath165\rangle = \mathcal{f}[u^\varepsilon(\cdot;\xi)]-\mathcal{f}[u^\varepsilon(\cdot;\bar\xi ) ] \approx \lambda_1^\varepsilon(\xi)\langle \psi(\cdot),\partial_\xi u^\varepsilon(\cdot;\bar\xi)\rangle ( \bar\xi-\xi),\ ] ] so that , heuristically , there exists a constant @xmath150 such that @xmath166\rangle| \leq c|\lambda_1^\varepsilon(\xi)||\psi|_{{}_{\infty}}\ ] ] which gives the final form of our ultimate assumption .
[ thm : metal ] let hypotheses * h1 - 2 - 3 * be satisfied .
additionally , assume that @xmath167 for some constant @xmath150 independent on @xmath12 and @xmath99 .
then , denoted by @xmath168 the solution to the initial - value problem , for any @xmath15 sufficiently small , there exists a time @xmath169 such that for any @xmath170 the solution @xmath171 is given by @xmath172 where @xmath173 is defined by @xmath174 and the remainder @xmath175 satisfies the estimate @xmath176 for some constant @xmath150 independent on @xmath177 .
moreover , for initial data @xmath178 sufficiently small in @xmath179 , the final time @xmath169 can be chosen with order @xmath180 .
.15 cm the conclusion of the proof of theorem [ thm : metal ] is based on the following version of a standard nonlinear iteration argument .
.15 cm [ lem : nquadro ] let @xmath181 and @xmath182 be continuous functions for @xmath183 $ ] for some @xmath84 , such that @xmath184 let @xmath185 be a non - negative function satisfying the estimate @xmath186 for any @xmath187 .
if there holds @xmath188 } \int_0^t g^2(s)\,f(s)\,ds \,\cdot\ , \sup_{t\in[0,t ] } \frac{1}{g(t)}\,\int_0^t h(s , t)\,ds < \frac14\ ] ] for any @xmath183 $ ] , then @xmath189 } \int_0^\tau h(s,\tau)\,ds\ ] ] for any @xmath183 $ ] . the auxiliary function @xmath190 enjoyes the estimate @xmath191 where @xmath192 and @xmath193 .
the quantity @xmath194 } w(\tau).\ ] ] is such that for any @xmath183 $ ] there holds @xmath195 where @xmath196 } \int_0^t \alpha(s)\,ds,\qquad b = b(t):=\sup_{t\in[0,t ] } \int_0^t \beta(s , t)\,ds.\ ] ] since @xmath197 , if @xmath198 , then @xmath199 in term of @xmath200 , if holds , then @xmath201 } \frac{1}{g(\tau)}\int_0^\tau h(s,\tau)\,ds.\ ] ] the final estimate follows from the monotonicity of the function @xmath202 .
setting @xmath203 we obtain an infinite - dimensional differential system for the coefficients @xmath204 @xmath205 where , omitting the dependencies for shortness , @xmath206 and the coefficients @xmath207 , @xmath208 are given by @xmath209 convergence of the series is guaranteed by assumption .
differentiating the normalization condition on the eigenfunction , we infer @xmath210 thus , for the coefficients @xmath207 there hold @xmath211 so that , in particular , @xmath212 for any @xmath213 . thus , equation for @xmath214 becomes @xmath215 now let us set @xmath216 as a consequence of hypothesis * h2 .
* , there exists @xmath150 such that @xmath217 for any @xmath218 .
thus , the absolute value of @xmath219 , @xmath218 , can be estimated by @xmath220 from equalities and , choosing @xmath221 , there follow @xmath222 for @xmath218 .
such expressions suggest to introduce the function @xmath223 from the representation formulas for the coefficients @xmath224 , since @xmath225 for some constant @xmath150 depending on the @xmath226norm of @xmath227 , there holds @xmath228 since @xmath229 , we infer @xmath230 the assumption on the asymptotic behavior of the eigenvalues @xmath231 can now be used to bound the series .
indeed , there holds for some @xmath150 @xmath232 as a consequence , for unknown @xmath171 such that @xmath233 for some @xmath234 , we infer @xmath235 let us set @xmath236 } |w - z|_{{}_{l^2}}\,e_1(s,0)\ ] ] then , since @xmath237 , we infer @xmath238 since @xmath239 is negative . by assumption ,
@xmath240 for some @xmath150 , hence @xmath241 so that we obtain the inequality @xmath242 taking the supremum , we end up with the estimate @xmath243 hence , as soon as @xmath244 there holds @xmath245 that means , in term of the difference @xmath246 , @xmath247 condition gives a constraint on the final time @xmath169 .
since @xmath248 and @xmath249 , it is enough to require @xmath250 to assure condition is satisfied .
the latter constraint can be rewritten as @xmath251 so that we can choose @xmath169 of the form @xmath252 for @xmath178 sufficiently small . as a consequence of the estimate
, for @xmath253 for some @xmath234 , the function @xmath254 satisfies @xmath255 where the constant @xmath77 depends also on @xmath256 .
in particular , if @xmath15 and @xmath257 are small , the function @xmath258 has similar decay properties of the function @xmath259 , solution to the reduced cauchy problem @xmath260 this preludes to the following consequence of theorem [ thm : metal ] .
[ cor : metal2 ] let hypotheses * h1 - 2 - 3 * and be satisfied .
assume also @xmath261 then , for @xmath15 and @xmath257 sufficiently small , the estimate holds globally in time and the solution @xmath168 converges exponentially fast to @xmath262 as @xmath263 .
thanks to assumption * h1 * , for @xmath15 and @xmath257 sufficiently small , estimate holds .
hence , for any initial datum @xmath264 , the variable @xmath258 satisfies . and , as a consequence , it converges exponentially fast to @xmath265 as @xmath263 , i.e. there exists @xmath266 such that @xmath267 for any @xmath268 under consideration .
furthermore , from , we deduce @xmath269 setting @xmath270 , by the jensen s inequality , we infer the estimate @xmath271 let @xmath272 be such that @xmath273 ; then , if @xmath274 , there holds @xmath275 showing the exponential convergence to @xmath95 of the component @xmath171 .
let us also stress that in the regime @xmath276 , a linearization at the equilibrium solution @xmath277 would furnish a more detailed description of the dynamics , since the source term due to the approximation at an approximate steady state would not be present .
in fact , the description given by the quasi - linearization is meaningful in the regime far from equilibrium and its aim is to describe the slow motion around a manifold of approximate solutions .
next , our aim is to show how the general approach just presented applies to the case of scalar conservation laws with viscosity . specifically , given @xmath9 , we consider the nonlinear equation @xmath278 with initial and boundary conditions given by @xmath279 for some @xmath12 , @xmath280 .
we assume that the flux @xmath10 and the data @xmath23 satisfy the conditions @xmath281 the single value @xmath282 such that @xmath283 is denoted by @xmath284 . without loss of generality , we assume @xmath285 . to clarify the relevance of the requirements and to justify the subsequent choice for the manifold @xmath92 , we propose a digression on the dynamics determined by the problem - in the vanishing viscosity limit .
setting @xmath31 , equation reduces to the first - order equation of hyperbolic type @xmath286 to be considered together with .
the boundary conditions are understood in the sense of bardos leroux
ndlec @xcite , meaning that the trace of the solution at the boundary is requested to take values in appropriate sets . to be precise ,
let @xmath287 be such that @xmath288 and set @xmath289 then , skipping the details ( see @xcite ) , the conditions @xmath290 translate into @xmath291\cup\{u_-\},\qquad u(\ell-0,t)\in \{u_+\}\cup[\mathcal{r}u_+,+\infty)\ ] ] since @xmath37 , there holds @xmath292 , and the conditions can be rewritten as @xmath293\cup\{u_-\},\qquad u(\ell-0,t)\in \{u_+\}\cup[u_-,+\infty)\ ] ] from the boundary conditions , it follows that characteristic curves entering in the domain from the left side @xmath294 , respectively , from the right @xmath295 , possess speed @xmath296 , resp .
speed @xmath297 . for with conditions a _ finite - time stabilization phenomenon
_ holds , similar to the one showed for the first time in @xcite in the case of the cauchy problem .
[ thm : stabhyp ] let @xmath298 and @xmath10 be such that holds .
then , for any @xmath299 , the solution @xmath300 to the initial - boundary value problem is such that for some @xmath84 and @xmath301 $ ] , there holds @xmath302 for almost any @xmath7 in @xmath29 .
the proof of the statement relies on the _ theory of generalized characteristics _ , introduced in @xcite .
the convexity assumption on the flux function @xmath10 guarantees that for any point @xmath303 there exist a minimal , respectively maximal , backward characteristics , which are classical characteristic curves , hence straight lines with slope @xmath304 , resp .
@xmath305 . by means of such technique
it is possible to follow the evolution of the curves @xmath306 as an illustrative example , let us first consider the case of a non - increasing initial datum @xmath30 .
then , for any @xmath307 , @xmath308 is non - increasing .
if @xmath309 are classical characteristics , the difference between their speeds of propagation satisfies @xmath310\!]}=:-\phi(u ) , \end{aligned}\ ] ] for any @xmath311 . since @xmath312 is strictly positive ,
the two curves intersect at a time @xmath313 that is smaller than @xmath314 .
the complete rigorous proof of theorem [ thm : stabhyp ] requires more technicalities and it is reported here for completeness .
let @xmath315 be the solution to the initial - boundary value problem under consideration with initial datum @xmath30 . for later use , we set in particular , @xmath316 .
we are going to show that @xmath317 for some @xmath84 . .15
cm _ 1 .
there exists @xmath318 such that @xmath319 $ ] for any @xmath320 . _
.15 cm indeed , let @xmath321 be the solution to the riemann problem for with datum @xmath322 hence , the restriction of @xmath323 to @xmath324 is a super - solution to the initial boundary value problem under consideration and , by comparison principle for entropy solution , we infer @xmath325 . since @xmath326 for any @xmath327 , there holds @xmath328 a similar estimate from below can be obtained by considering as subsolution the restriction of @xmath329 to @xmath324 , where @xmath329 is the solution to with initial datum @xmath330 from now on , we assume that the solution @xmath300 takes values in the interval @xmath331 $ ]
. .15 cm _ 2 .
assume that @xmath332 for any @xmath268 ; then there exists @xmath333 such that @xmath334 and @xmath335 for any @xmath336 .
_ .15 cm if @xmath300 is continuous at @xmath337 for some @xmath338 , then @xmath339
. therefore , the maximal backward characteristic from @xmath337 is the straight line @xmath340 . for @xmath341 ,
such curve intersects the boundary @xmath294 at some @xmath342 . by continuity ,
all of the maximal backward characteristics from @xmath343 with @xmath344 and sufficiently close to @xmath345 intersect the boundary @xmath294 at some time @xmath346 smaller than @xmath347 and close to it .
because of the boundary conditions , this may happen if and only if @xmath348 .
hence , @xmath349 for @xmath350 for some @xmath12 , in contradiction with the definition of @xmath351 .
thus , continuity of @xmath300 at @xmath337 may happen only for @xmath352 .
a similar assertion holds for @xmath353 .
.15 cm _ 3 . there exist @xmath84 and @xmath301 $ ] such that @xmath354 for any @xmath355 .
_ .15 cm given @xmath356 , let @xmath357 be such that @xmath358 let @xmath359 be the maximal backward characteristic from @xmath360 , whose equation is @xmath361 . if @xmath359 hits the right boundary @xmath295 at some positive time , the solution @xmath300 coincides with @xmath362 . otherwise , there holds @xmath363 , which gives @xmath364 similarly , let @xmath365 be the maximal backward characteristic from @xmath366 , whose equation is @xmath367 .
if @xmath365 does not intersect the left boundary @xmath294 at some positive time , there holds @xmath368 .
hence , for any @xmath12 , we can choose @xmath369 sufficiently large so that @xmath370 and @xmath371 .
thus , we have @xmath372 which is uniformly negative for @xmath15 sufficiently small .
hence , the curves @xmath353 and @xmath351 intersect at some finite positive time @xmath84 . as soon as the viscosity term is switched on , i.e. for @xmath12 , the number of steady states for drastically reduces with respect to the corresponding hyperbolic case . indeed , stationary solution to the problem
are implicitly determined by the relation @xmath373 where @xmath374 is such that @xmath375 assumptions [ fluxhyp2 ] on the flux @xmath10 imply that @xmath376 is strictly decreasing and such that @xmath377 therefore , for any @xmath9 , there exists a unique steady state for . in the case of burgers equation , @xmath378 , the value @xmath25 coincides with @xmath379 and @xmath376 has the explicit form @xmath380 , so that the value @xmath347 determining the stationary solution is uniquely determined by the relation @xmath381 given @xmath382 , the steady state @xmath383 has the expression @xmath384 .
following the general approach introduced in the previous section , we build a one - parameter family of functions @xmath385 with @xmath99 converging to @xmath386 as @xmath38 .
in particular , the parameter set @xmath91 coincides with the interval @xmath29 @xmath302 there are many meaningful choices for @xmath66 ( see the traveling wave approach in @xcite ) ; here , we opt for matching at a given point @xmath387 the two stationary solutions of in @xmath388 and @xmath389 , denoted by @xmath390 and @xmath391 , satisfying the boundary conditions @xmath392 where @xmath284 is such that @xmath288 .
hence , we set @xmath393 given @xmath374 and @xmath311 , let us define @xmath394 similarly to the case of stationary states , the function @xmath376 is such that @xmath395 so that for any @xmath396 there are ( unique ) @xmath397 such that @xmath398 correspondingly , functions @xmath399 are implicitly given by @xmath400 by substitution , denoting by @xmath401 the dirac s delta distribution concentrated at @xmath402 , there holds in the sense of distributions @xmath403 = { [ \![}\partial_x u^{\varepsilon}{]\!]}_{{}_{x=\xi}}\delta_{{}_{x=\xi } } = \frac{1}{\varepsilon}\bigl(\kappa_-^\varepsilon(\xi)-\kappa_+^\varepsilon(\xi)\bigr)\delta_{{}_{x=\xi}}\ ] ] with @xmath404 implicitly defined by . as a consequence of the properties of function @xmath376 , the difference function @xmath405 is monotone decreasing and such that @xmath406 then , there exists unique @xmath407 such that @xmath408 and such a value is such that @xmath409 is the unique steady state of the problem
. from the bounds @xmath410,\\ f(u_\pm)-f'(u_-)(u_--u ) & \leq f(u)\leq \frac{f(u_\pm)}{u_--u_\ast}(u - u_\ast ) & \qquad & u\in[u_\ast , u_- ] , \end{aligned}\ ] ] we locate approximately the differences @xmath411 @xmath412 such bounds show that @xmath413 is exponentially small as @xmath16 , uniformly in any compact subset of @xmath56 ; therefore , for any @xmath414 , there exist @xmath415 , indipendent on @xmath15 , such that @xmath416\!]}_{{}_{x=\xi}}\bigr| \leq c_1\,e^{-c_2/\varepsilon } \qquad\qquad \forall\,\xi\in(-\ell+\delta,\ell-\delta).\ ] ] in particular , hypothesis * h1 * , stated in section 2 , is satisfied .
going further , retracing the definitions previously introduced and setting @xmath417 , we consider the operators @xmath418 where the adjoint operator @xmath113 is considered with dirichlet boundary conditions . for small @xmath15 and @xmath5 , the dynamics of the parameter @xmath55
is approximately given by @xmath419 \rangle}\\\ ] ] where @xmath124 is the first eigenfunction of the adjoint operator @xmath113 satisfying the normalization condition @xmath420 for @xmath89 , the eigenfunction @xmath421 is close to the eigenfunction of @xmath422 relative to the eigenvalue @xmath423 , with @xmath424 hence , we obtain the representation formula @xmath425 where @xmath426 for some @xmath427 .
in the limit @xmath38 , we obtain @xmath428 , provided @xmath55 is bounded away from the boundaries @xmath429 . with the approximation @xmath430
we infer @xmath431\!]}\,\chi_{{}_{(\xi,\xi+h)}}(x)\ ] ] so that we expect @xmath164 to converge to @xmath432\!]}\,\delta_{\xi}$ ] as @xmath38 in the sense of distributions .
hence , the normalization condition gives the choice @xmath433\!]}$ ] in .
therefore , we deduce an approximate expression for the function @xmath434 @xmath435\ ! ] } } \,\langle 1,{\mathcal f[u^{\varepsilon } ] \rangle } = \frac{1}{\varepsilon\,{[\![}u{]\ ! ] } } \left(\kappa_+^\varepsilon(\xi)-\kappa_-^\varepsilon(\xi)\right).\ ] ] estimate shows that the the function @xmath369 has order of magnitude @xmath436 . in the very special case @xmath437 , with @xmath438 and @xmath439 , the earlier estimates on @xmath440
are exact , so that @xmath441 in this case , the function @xmath434 is approximated by @xmath442 which gives @xmath443 in the regime @xmath16 .
.15 cm [ example : burgerseq ] for the burgers equation , @xmath378 , there holds @xmath444 given @xmath396 , the values @xmath440 can be approximated by @xmath445 determined by @xmath446 obtained by substituting the multiplicative term @xmath447 with @xmath448 . by computation
, we obtain the explicit expressions @xmath449 since , for @xmath450 , @xmath451 as @xmath16 , the function @xmath434 approaches @xmath452 which corresponds to the formula determined in @xcite .
our concern in the present section is to estabilish a precise description on the location of the eigenvalues of the linearized operator , in order to show that the general procedure developed in section [ sect : abs ] is indeed applicable in the case of scalar conservation laws with convex flux .
the problem of determining the limiting structure of the spectrum of the type of second order differential operators we deal with has been widely considered in the literature . among others ,
let us quote the approach , based on the use of prfer transform , used in @xcite , in the context of metastability analysis for the allen cahn equation . here
, we prefer to follow the strategy implemented in @xcite , for the linearization at the steady state of the burgers equation . in what follows ,
we show that the same kind of eigenvalues distribution holds in a much more general situation , the main ingredient being the resemblance of the coefficient @xmath453 to a step function @xmath454 , jumping from a positive to a negative value , as @xmath16 .
.15 cm fixed @xmath12 and linearizing the scalar conservation law at a given a reference profile @xmath455 , satisfying the boundary conditions @xmath456 , we end up with the differential linear diffusion - transport operator @xmath457 where @xmath458 .
the aim of this section is to describe the structure of the spectrum @xmath459 of the operator @xmath460 for @xmath15 sufficiently small . given the function @xmath453 , let us introduce the self - adjoint operator @xmath461 where @xmath462 a straightforward calculation shows that if @xmath300 is an eigenfunction of relative to the eigenvalue @xmath463 , then the function @xmath464 defined by @xmath465 ( with @xmath466 arbitrarily chosen ) is an eigenfunction of the operator @xmath467 relative to the eigenvalue @xmath468 . since @xmath469 is self - adjoint , we can state that the spectrum of the operator @xmath109 is composed by eigenvalues .
moreover , if @xmath300 is an eigenfunction of relative to the first eigenvalue @xmath470 , integrating in @xmath56 the relation @xmath471 , we deduce the identity @xmath472 assuming , without loss of generality , @xmath300 to be strictly positive in @xmath56 and normalized so that its integral in @xmath56 is equal to 1 , we get @xmath473 hence , for any choice of the function @xmath453 , there holds @xmath474 our next aim is to show that under appropriate assumption on the behavior of the family of functions @xmath453 as @xmath16 , it is possible to furnish a detailed representation of the eigenvalue distributions for small @xmath15 .
specifically , we are interested in coefficients @xmath453 behaving , in the limit @xmath16 as a step function of the form @xmath475 for some @xmath396 and @xmath476
. we will show that , under appropriate assumptions making precise in which sense @xmath453 `` resemble '' @xmath454 for @xmath15 small , the first eigenvalue @xmath470 turns to be `` very close '' to @xmath95 for @xmath15 small , and all of the others eigenavalues @xmath477 , with @xmath218 , are such that @xmath478 as @xmath16 .
we estimate the first eigenvalue @xmath479 of the operator @xmath469 by means of the inequality @xmath480 for smooth test function @xmath481 such that @xmath482 .
let us consider as test function @xmath483 , where @xmath484 a direct calculation shows that @xmath485 and , assuming the family @xmath486 to be uniformly bounded , we infer @xmath487 as soon as @xmath488 .
the opposite case being similar , let us assume @xmath489 . from the definition of @xmath490
, it follows @xmath491 therefore , we deduce @xmath492 where @xmath493 since @xmath453 converges to the step function @xmath454 as @xmath16 , it is natural to approximate the latter integral in term of the corresponding one for @xmath454 : @xmath494 since , for @xmath15 small , @xmath495}{a_-a_+}\,\varepsilon\,e^{a_-(\xi+\ell)/\varepsilon}. \end{aligned}\ ] ] the subsequent estimate holds @xmath496 whenever @xmath497 for some @xmath21 .
thus , we deduce for the first eigenvalue @xmath479 of the self - adjoint operator @xmath469 the estimate @xmath498 for some positive constant @xmath499 . as a consequence , since the spectrum @xmath500 coincides with @xmath501 , the next result holds . [ prop : firstevalue ] let @xmath453 be a family of functions satisfying the assumption : + there exists @xmath150 , indipendent on @xmath12 , such that @xmath502 if there exists @xmath396 , @xmath476 and @xmath150 for which @xmath503 , then there exist constants @xmath504 such that @xmath505 .
let us stress that the request @xmath476 is essential , even if hided in the proof .
if this is not the case , the term @xmath490 would not be small as @xmath16 and its @xmath179 norm would not be bounded by the @xmath179-norm of @xmath506 .
in fact , the statement in proposition [ prop : firstevalue ] may not hold when @xmath507 have the same sign , the easiest example being the case @xmath508 .
the next example gives an heuristic estimate for the first eigenvalue @xmath470 .
given @xmath509 and @xmath510 , let us set @xmath511 , @xmath512:=a_+-a_-$ ] and @xmath513 given @xmath514 , let us look for functions @xmath515 , such that @xmath516 in the sense of distributions . since @xmath517\,\delta_{0}$
] , this amounts in finding two functions @xmath14 such that @xmath518 and the following transmission conditions are satisfied @xmath519\,u_\pm(0)=0.\ ] ] the characteristic polinomial of @xmath520 is @xmath521 , with roots @xmath522 assume @xmath523 .
choosing @xmath23 in the form @xmath524 setting @xmath525 and @xmath526 , there holds @xmath527 therefore , the transmission conditions take the form of a linear system in @xmath528 @xmath529\right)\theta_+^+ -\left(2\varepsilon\,\mu_+^--[a]\right)\theta_+^-\bigr\}a_+\\ & \qquad + \bigl\{-\left(2\varepsilon\,\mu_-^++[a]\right)\theta_-^+ + \left(2\varepsilon\,\mu_-^-+[a]\right)\theta_-^-\bigr\}a_-=0 . \end{aligned}\right.\ ] ] after some manipulations , the determinant @xmath530 of system can be written as @xmath531-[\delta]\right)\theta_-^+\theta_+^++\left([a]+\{\delta\}\right)\theta_-^+\theta_+^- + \left([a]-\{\delta\}\right)\theta_-^-\theta_+^+-\left([a]+[\delta]\right)\theta_-^-\theta_+^-,\ ] ] where @xmath532:=\delta_+-\delta_-$ ] and @xmath533 . since @xmath534 ,
in the case @xmath476 there hold @xmath535\left(1-\frac{2\,\varepsilon\lambda}{a_+\,a_-}\right)+o(\varepsilon\lambda)\\ [ \delta]&=\sqrt{a_+^2 + 4\,\varepsilon\,\lambda}-\sqrt{a_-^2 + 4\,\varepsilon\,\lambda } = -\{a\}\left(1+\frac{2\,\varepsilon\lambda}{a_+ a_-}\right ) + o(\varepsilon\lambda ) \end{aligned}\ ] ] as @xmath536 , together with @xmath537 hence , for @xmath538 and @xmath536 , disregarding the exponentially small term @xmath539 keeping only the principal term in the expansions , we infer @xmath540\,\varepsilon\lambda}{a_+\,a_-}\,e^{(a_-\alpha - a_+\beta)/\varepsilon } + a_- e^{-a_+\beta/\varepsilon}.\ ] ] therefore , @xmath541 for @xmath542 in the regime @xmath543 small .
asymptotic representation permits to verify the relation between the first eigenvalue of the linearized operator and the term @xmath103 , controlling the size of @xmath544 $ ] ( see ) .
specifically , for the burgers equation , becomes @xmath545 the term @xmath544 $ ] given in for the burgers equation ( example [ example : burgerseq ] ) is such that @xmath546 therefore , the estimate @xmath547 holds and hypothesis is verified
. .15 cm for general scalar conservation it still possible to obtain an analogous bound .
indeed , for @xmath548 , @xmath549 and @xmath550 , expression becomes @xmath551 ( compare with lemma 3.2 in @xcite ) .
the bounded for @xmath103 can be obtained by proceeding as in section [ sect : abs ] , by means of a more detailed estimate on the functions @xmath552 starting from the inequalities @xmath553,\\ f(u)&\leq f(u_-)+f'(u_-)(u - u_-)+\frac{1}{2}c_0(u - u_-)^2 & \qquad & u\in[u_\ast , u_- ] , \end{aligned}\ ] ] a careful ( and tedious ) computation of the integrals in a the corresponding approximated form for the implicit relation , leads to the bound @xmath554 which , together with the asymptotic representation for @xmath470 , guarantees requirement in theorem [ thm : metal ] .
.15 cm controlling the location of the second ( and subsequent ) eigenvalue needs much more care and , also , a number of additional assumption on the limiting behavior of the function @xmath453 as @xmath16 .
precisely , we suppose @xmath555)$ ] satisfies the following hypotheses : + the function @xmath453 is twice differentiable at any @xmath556 and @xmath557 for any @xmath150 there exists @xmath558 such that , for any @xmath7 satisfying @xmath559 , there holds @xmath560 there exists the left / right first order derivatives of @xmath453 at @xmath55 and @xmath561 as a consequence , the function @xmath562 satisfies a number of corresponding properties , listed in the next statement .
[ lem : propb ] let the family @xmath453 be such that hypotheses a1 - 2 - 3 are satisfied , and let @xmath563 be such that @xmath564 then there exist @xmath565 such that , for @xmath566 , the functions @xmath562 , with @xmath486 defined in , enjoy the following properties : + the function @xmath562 is decreasing in @xmath388 and increasing in @xmath389 ; + there exist @xmath567 such that , for any @xmath7 with @xmath568 there holds @xmath569 ; + there exist the left / right limits of @xmath562 at @xmath55 and @xmath570 property b1 .
is an immediate consequence of assumption a1 , since @xmath571 from a2 , given @xmath150 , for @xmath572 , there holds @xmath573 from such inequality , by choosing @xmath150 sufficiently small , and combining with an analogous estimate on @xmath574 , property b2 . follows . for what concerns b3 ,
we observe that , since @xmath575 and @xmath576 , there holds @xmath577 thanks to a3 . for later reference
, we denote @xmath578 the zeros of @xmath562 , with @xmath579 .
since property b2 holds , we deduce that @xmath580 .
.15 cm assume the assumption of lemma [ lem : propb ] to hold , and let @xmath581 and @xmath582 be the second eigenvalue of the operators @xmath583 and @xmath584 , respectively , with corresponding eigenfunctions @xmath585 and @xmath586 . such eigenfunctions are linked together by the relation @xmath587 for some constants @xmath588 and @xmath589 . since @xmath581 is the second eigenvalue , the functions @xmath585 and @xmath586 possess a single root located at some point @xmath590 .
the sign properties of @xmath591 described in lemma [ lem : propb ] imply that @xmath592 .
then , @xmath585 and @xmath586 restricted to the intervals @xmath593 and @xmath594 are eigenfunctions relative to the first eigenvalue of the same operator considered in the corresponding intervals and with dirichlet boundary conditions . from now on ,
we drop , for shortness , the dependence on @xmath15 of @xmath595 , we assume , without loss of generality , @xmath596 and we restrict our attention to the interval @xmath597 . integrating on @xmath91 , we deduce @xmath598 having chosen @xmath599 positive in @xmath91 . assuming @xmath600 to be given as in with @xmath601 and @xmath602 , and normalized so that @xmath603 , from the latter inequality we infer the inequality @xmath604 where @xmath605 our next aim is to deduce an estimate from above on @xmath606 and an estimate from below for @xmath607 , in order to get a control on the size of the second eigenvalue @xmath608 . from the definition of @xmath606 , since @xmath596
, it follows @xmath609 whenever @xmath610 .
thus , estimate provisionally becomes @xmath611 for some positive constant @xmath77 , independent on @xmath15 .
let the value @xmath612 be such that @xmath613 , minimum with such property . from the assumption on the function @xmath614 , it follows @xmath615 .
then there exists @xmath616 such that @xmath617 since the function @xmath481 is concave in the interval @xmath618 , we deduce @xmath619 plugging into , we end up with @xmath620 , for some @xmath77 independent on @xmath15 . as a consequence , we can state a result relative to the second eigenvalue @xmath608 .
.15 cm [ prop : secondevalue ] let @xmath453 be a family of functions sastisfying a1 - 2 - 3 then there exists @xmath150 such that @xmath621 for any @xmath15 sufficiently small . collecting the results of propositions [
prop : firstevalue ] and [ prop : secondevalue ] give a complete description for the spectrum of operator @xmath622 for small @xmath15 , under assumptions a0 - 1 - 2 - 3 on the family @xmath453 .
.15 cm [ prop : spectral ] let @xmath453 be a family of functions satisfying the assumptions a0 - 1 - 2 - 3 for some @xmath396 , @xmath476 .
then there exist @xmath150 such that @xmath623 for any @xmath218 .
hypotheses a0 - 1 - 2 - 3 are satisfied in the case of a family of function @xmath453 that is a ( small ) perturbation of a function @xmath624 with the form @xmath625 for some decreasing smooth bounded functions @xmath528 , bounded together with their first and second order derivatives , and such that @xmath626 and @xmath627 .
the authors are thankful to the anonymous referees for their valuable comments and suggestions which helped them improving the content and the overall organization of the article .
de groen , p. p. n. ; karadzhov , g. e. ; _ slow travelling waves on a finite interval for burgers-type equations _ ,
advanced numerical methods for mathematical modelling .
j. comput .
appl . math . 132
( 2001 ) no .
1 , 155189 . sun , x. ; ward , m. j. ; russell , r. ; _ the slow dynamics of two - spike solutions for the gray - scott and gierer - meinhardt systems : competition and oscillatory instabilities _
, siam j. appl .
syst . 4 ( 2005 ) no .
4 , 904953 . ward , m.j . ; reyna , l.g . ; _ internal layers , small eigenvalues , and the sensitivity of metastable motion _ , perturbation methods in physical mathematics ( troy , ny , 1993 ) .
siam j. appl .
( 1995 ) no .
2 , 425445 . | the aim article is to contribute to the definition of a versatile language for metastability in the context of partial differential equations of evolutive type . a general framework suited for parabolic equations in one dimensional bounded domains
is proposed , based on choosing a family of approximate steady states @xmath0 and on the spectral properties of the linearized operators at such states .
the slow motion for solutions belonging to a cylindrical neighborhood of the family @xmath1 is analyzed by means of a system of an ode for the parameter @xmath2 , coupled with a pde describing the evolution of the perturbation @xmath3 .
we state and prove a general result concerning the reduced system for the couple @xmath4 , called quasi - linearized system , obtained by disregarding the nonlinear term in @xmath5 , and we show how such approach suits to the prototypical example of scalar viscous conservation laws with dirichlet boundary condition in a bounded one - dimensional interval with convex flux .
corrado mascia , marta strani 0.3 cm * key words * metastability ; slow motion ; spectral analysis ; viscous conservation laws .
0.3 cm * ams subject classification * 35b25 ( 35p15 , 35k20 ) |
this article is concerned with a programme that has as its goal the development of a theory of quantum space - time . in this programme ,
an outline of which will be given in more detail shortly , an important role is played by certain higher - dimensional analogues of spinors and twistors .
it will be useful to begin , therefore , by remarking that there are two distinct notions of how one extends the concept of spinor into higher dimensions .
this fundamental dichotomy arises in association with the fact that in four - dimensional space - time there is a local isomorphism between the lorentz group @xmath0 and the spin transformation group @xmath1 . in higher dimensions
, however , this relation breaks down and as a consequence we are left with two concepts of spinors
one for the group @xmath2 , and one for the group @xmath3 .
the spinors associated with @xmath2 , where we allow also for various possible signatures in the quadratic form defining these orthogonal or pseudo - orthogonal transformations when we specialise to the real subgroup @xmath4 with @xmath5 , are the so - called cartan spinors @xcite .
the study of cartan spinors has a long and interesting history , and there is a beautiful geometry associated with these spinors .
there are also various specific cases of great interest for example , the cartan spinors associated with the group @xmath6 are penrose s twistors ; and the cartan spinors associated with @xmath7 are intimately linked with the cayley numbers ( octonions ) and the exceptional lie groups .
there are also a number of interesting connections between cartan spinors and massless fields in higher dimensions @xcite .
the spinors associated with @xmath3 , which are usually now called ` hyperspinors ' , have the advantage of being more directly linked with quantum mechanics .
in fact , we shall show later that a naturally relativistic model for hyperspin arises when one considers ` multiplets ' of two - component spinors , i.e. expressions of the form @xmath8 and @xmath9 , where @xmath10 are standard spinor indices , and @xmath11 is an ` internal ' index . in the general case ( @xmath12 ) we then think of @xmath8 as an element of the tensor product space @xmath13 , where @xmath14 is the complex vector space of two - component spinors , and @xmath15 is an infinite - dimensional complex hilbert space .
there is also a link , arising through a further extension of this idea , between hyperspinor theory and the theory of multi - twistor ( hypertwistor ) systems .
indeed , we find that the theory of hyperspin constitutes a natural starting place for building up a theory of quantum geometry or , as we shall call it here , _
quantum space - time_. in summary , we shall be taking the left - hand path in the following diagram : ( 300,100)(0,0 ) ( 155,80)@xmath16 ( 125,8)@xmath3 ( 240,8)@xmath2 ( 105,-5)hyperspinors @xmath17 ( 225,-5)cartan spinors ( 196,72)(-1,-1)50 ( 204,72)(1,-1)50 the hyperspinor route has the virtue that the resulting space - time has a rich causal structure associated with it , and as a consequence is unusually well - positioned to form the geometrical basis of a physical theory .
to start , let us review briefly the role of two - component spinors in the description of four - dimensional minkowskian space - time geometry . in what follows we use bold upright roman letters to denote two - component spinor indices ,
and we adopt the standard conventions for the algebra of two - component spinors @xcite
. then we have the following correspondence between two - by - two hermitian matrices and space - time points , relative to some origin : @xmath18 more explicitly , in a standard basis this correspondence is given by @xmath19 we then have the fundamental relation @xmath20 from which it follows that two - component spinors are connected both with quantum mechanics and with the causal structure of space - time .
it is a peculiar aspect of relativistic physics that there is this link between ( a ) the spin degrees of freedom of spin one - half particles , and ( b ) the causal geometry of four - dimensional space - time .
let us pursue this idea now in a little more detail , and then extend it to higher dimensions .
for the interval @xmath21 between a pair of points @xmath22 and @xmath23 in minkowski space - time we write @xmath24 from which it follows that @xmath25 where @xmath26 is the antisymmetric spinor .
hence if we adopt the standard ` index clumping ' convention and write @xmath27 , @xmath28 , and so on , according to which a pair of spinor indices , one primed and the other unprimed , corresponds to a lower case space - time vector index , then we can write @xmath29 for the corresponding squared space - time interval , and thus we are able to identity @xmath30 as the metric of minkowski space .
there are essentially three different situations that can arise for the interval @xmath31 , each of which represents a certain level of degeneracy .
the first case is @xmath32 ; the second case is @xmath33 and @xmath34 ; and the third case is @xmath35 .
each of these cases gives rise to a canonical form for the interval @xmath36 , with various sub - cases , which can be summarised as follows : * @xmath32 : @xmath37 * @xmath38 : @xmath39 * @xmath40 : @xmath41 _ causal structure of four - dimensional space - time_. the canonical form of the spinor decomposition of the four - vector associated with a space - time interval determines its causal properties . ] here it is understood that in case ( iii ) the spinors @xmath42 and @xmath43 do not coincide in direction .
it is interesting to note that once the canonical form for @xmath21 is specified , then so is the causal relationship that it determines on space - time .
this correspondence is illustrated in figure [ fig:1 ] .
on the other hand , the specification of @xmath36 does not completely determine the spinors @xmath44 and @xmath43 .
in general there is some freedom , and this is expressed by a group of transformations .
in particular , if @xmath36 is null , then this freedom is the phase shift @xmath45 , and the relevant group is @xmath46 .
if @xmath36 is time - like , then the group is @xmath47 , and if @xmath36 is space - like , the group is @xmath48 .
the terminology ` hyperspinor ' is due to finkelstein @xcite . essentially the same concept ( although introduced for different purposes )
is also touched on in @xcite .
the idea of a hyperspinor is a simple one we replace the two - component spinors associated with four - dimensional space - time with @xmath49-component spinors .
thus we can regard hyperspin space as the vector space @xmath50 with some extra structure .
in particular , in addition to the original hyperspin space we have three other vector spaces the dual hyperspin space , the complex conjugate hyperspin space , and the dual complex conjugate hyperspin space . the theory of hyperspin has been pursued by a number of authors @xcite , and the material we describe here builds on various aspects of this work .
let us write @xmath51 and @xmath52 , respectively , for the complex @xmath49-dimensional vector spaces of unprimed and primed hyperspinors .
for hyperspinors we use italic indices to distinguish them from the boldface indices used exclusively for two - component spinors .
it is assumed that @xmath51 and @xmath52 are related by an anti - linear isomorphism under the operation of complex conjugation .
thus if @xmath53 , then under complex conjugation we have @xmath54 , where @xmath55 .
the dual spaces associated with @xmath51 and @xmath52 are denoted @xmath56 and @xmath57 , respectively . if @xmath53 and @xmath58 , then their inner product is denoted @xmath59 .
likewise if @xmath60 and @xmath61 then their inner product is @xmath62 .
we also introduce the totally antisymmetric hyperspinors of rank @xmath49 associated with the spaces @xmath51 , @xmath63 , @xmath52 , and @xmath57 .
these will be denoted @xmath64 , @xmath65 , @xmath66 , and @xmath67 , respectively .
the choice of these antisymmetric hyperspinors is unique up to an overall scale factor .
once a choice has been made for @xmath65 , then the other epsilon hyperspinors are determined by the relations @xmath68 where @xmath69 is the complex conjugate of @xmath65 .
now let @xmath70 denote the skew tensor product space @xmath71 . using an analogous notation
, we introduce the spaces @xmath72 , @xmath73 , and @xmath74 for each @xmath75 .
then once the epsilon hyperspinors have been fixed we have a collection of maps of the form @xmath76 as a consequence , a wide range of algebraic theorems can be formulated , which are useful in calculations .
for example , if @xmath77 then @xmath78 must be of the form @xmath79 for some @xmath80 .
now we introduce the complex matrix space @xmath81 .
an element @xmath82 is said to be _ real _ if it satisfies the ( weak ) hermitian property @xmath83 , where @xmath84 is the complex conjugate of @xmath85 .
we shall have more to say about weak versus strong hermiticity conditions in relation to the idea of symmetry breaking .
we denote the vector space of real elements of @xmath86 by @xmath87 .
the elements of @xmath88 constitute what we call the real quantum space - time @xmath89 of dimension @xmath90 .
we then regard @xmath91 as the complexification of @xmath89 .
many problems in @xmath89 are best first approached as problems in @xmath92 , and hence sometimes although we refer to @xmath89 our operations are actually carried out in @xmath92 .
let @xmath85 and @xmath93 be points in @xmath89 , and write @xmath94 for the corresponding separation vector , which is independent of the choice of origin . using the index - clumping convention we set @xmath95 , @xmath96 , @xmath97 , and for the separation of @xmath98 and @xmath99 in @xmath89 we write @xmath100 .
there is a natural causal structure induced on such intervals by the so - called ` chronometric tensor ' . making use of the index - clumping convention
, we define this fundamental tensor ( introduced by finkelstein @xcite ) by the following basic relation : @xmath101 the chronometric tensor , which is of rank @xmath49 , is totally symmetric and is nondegenerate in the sense that for any vector @xmath31 the condition @xmath102 implies @xmath103 .
we say that @xmath98 and @xmath99 in @xmath89 have a ` degenerate ' separation if the chronometric form @xmath104 vanishes for @xmath100 .
degenerate separation is equivalent to the vanishing of the determinant of the matrix @xmath105 , that is , @xmath106
if the hyperspin space has dimension @xmath107 , this reduces to the usual condition for @xmath98 and @xmath99 to be null - separated in minkowski space . for @xmath108 , however , the situation is more complicated since there are various degrees of degeneracy that can arise between two points of quantum space - time , of which ` nullness ' ( in the minkowskian sense ) is only the most extreme . as an example , consider @xmath109 .
in this case the quantum space - time has dimension nine , and the chronometric form is given by @xmath110 the different possibilities that can arise for the separation vector are as follows : * @xmath111 : @xmath112 * @xmath113 and @xmath114 : @xmath115 * @xmath116 and @xmath117 : @xmath118 * @xmath119 and @xmath120 : @xmath121 when the separation of two points of quantum space - time is degenerate , we define the ` degree ' of degeneracy by the rank of the matrix @xmath105 .
null separation is the case for which the degeneracy is of the first degree , i.e. where @xmath105 is of rank one , and thus satisfies a system of quadratic relations of the following form : @xmath122 or equivalently @xmath123 this implies that @xmath105 can be expressed in the ` null ' form @xmath124 for some @xmath125 . in the case of degeneracy of the second degree , @xmath105 is of rank two and satisfies a set of cubic relations given by @xmath126 or equivalently @xmath127 in this situation @xmath105 can be put into one of the following three canonical forms : * @xmath128 , * @xmath129 , * @xmath130 . in case ( a ) , @xmath98 lies to the future of @xmath99 , and @xmath31 can be thought of as a degenerate future - pointing time - like vector . in case
( b ) , @xmath31 can be thought of as a degenerate space - like separation . in case ( c )
, @xmath98 lies to the past of @xmath99 , and @xmath31 is a degenerate past - pointing time - like vector .
a similar analysis can be applied to degenerate separations of other ` intermediate ' degrees .
if the determinant of the @xmath49-by-@xmath49 weakly hermitian matrix @xmath105 is nonvanishing , and @xmath105 is thus of maximal rank , then the chronometric form @xmath131 is nonvanishing . in that case
the matrix @xmath105 can be represented in the following canonical form : @xmath132 with the presence of @xmath49 nonvanishing terms , where the @xmath49 hyperspinors @xmath133 are all linearly independent .
let us write @xmath134 for the numbers of plus and minus signs appearing in the canonical form for the matrix @xmath105 given above .
we call @xmath134 the ` signature ' of @xmath105 .
the hyperspinors @xmath135 are determined by the specification of @xmath105 only up to an overall unitary ( or pseudo - unitary ) transformation of the form @xmath136 where @xmath137 , and @xmath138 the signature @xmath134 is nevertheless an invariant of @xmath105 . in the cases for which the signature is @xmath139 or
@xmath140 we say that @xmath105 is future - pointing time - like or past - pointing time - like , respectively .
then recalling the definition ( [ eq:10 ] ) for the associated chronometric form , we define the ` proper time interval ' between the events @xmath98 and @xmath141 by the formula @xmath142 in the case @xmath107 we then recover the standard minkowskian proper - time interval between the given events .
a remarkable feature of the causal structure of quantum space - time is that the essential physical features of the causal structure of minkowski space are preserved . in particular
, the space of future - pointing time - like vectors forms a convex cone .
the same is true when we consider the structure of the associated momentum space , from which it follows that we can also give a good definition of what is meant by ` positive energy ' .
now suppose that @xmath143 defines a smooth curve @xmath144 in @xmath89 for @xmath145\subset{\mathbb r}$ ] .
then @xmath144 will be said to be time - like if the tangent vector along @xmath144 , @xmath146 is time - like and future - pointing . in that case
we define the proper time @xmath147 elapsed along @xmath144 by the integral @xmath148^{\frac{1}{r } } { { \rm d}}\lambda .
\label{eq:21}\end{aligned}\ ] ] in the case of a very small interval , we can also write this in the ` pseudo - finslerian ' form @xmath149 in the case @xmath107 this clearly reduces to the standard pseudo - riemannian expression for the line element . ( 300,100 ) ( 110,0)(150,80)(220,68 ) ( 220,68)(290,60)(300,80 ) ( 128,29 ) ( 255,66 ) ( 132,22)@xmath150 ( 245,52)@xmath151 ( 180,70)@xmath144 ( 118,2)@xmath152 now let us consider the condition @xmath144 must satisfy in order to be a geodesic in @xmath89 . since the geometry is not riemannian , the answer may not be entirely obvious . in the case of a time - like curve
, we can choose the proper time as the parameter along the curve , in which case the resulting affine parameterisation of the curve is determined up to transformations of the form @xmath153 where @xmath154 is a constant .
the equation of motion for the situation in which @xmath144 is a time - like geodesic is obtained by an application of the calculus of variations to formula ( [ eq:21 ] ) . as usual , we assume the variation vanishes at the endpoints .
writing @xmath155 a standard argument shows that @xmath152 describes a geodesic only if the velocity vector @xmath156 satisfies the euler - lagrange equation @xmath157 a calculation shows that this condition is given more explicitly by @xmath158 where @xmath159 if @xmath160 is chosen to be proper time , then @xmath161 and the geodesic equation takes to form @xmath162 where the dot denotes differentiation with respect to proper time .
in the case @xmath107 the geodesic equation ( [ eq:26 ] ) reduces to the familiar relation @xmath163 .
we shall now show that the geodesic equation also implies @xmath163 in the case of a general quantum space - time with @xmath108 .
it suffices to examine the case @xmath109 , which will indicate the relevant line of argument . for @xmath109
the geodesic equation takes the form @xmath164 which can be expressed in terms of hyperspinors in the form @xmath165 this relation can then be written @xmath166 because @xmath167 we know that @xmath168 has an inverse @xmath169 satisfying @xmath170 therefore , contracting ( [ eq:28 ] ) with @xmath171 we obtain @xmath172 this equation shows that if @xmath173 were not zero , then it would have to be proportional to @xmath174 .
however , if that were so , then @xmath175 would imply @xmath176 , contrary to the assumption that @xmath174 is time - like .
it follows that @xmath163 .
a similar argument shows that for all @xmath177 the geodesic equation ( [ eq:26 ] ) implies @xmath163 .
hence we have deduced the following result .
let @xmath99 and @xmath178 be quantum space - time points with the property that @xmath179 is time - like and future - pointing
. then the affinely parametrised geodesic @xmath144 connecting these points in @xmath89 is given by @xmath180^{1/r}}\ , s\end{aligned}\ ] ] for @xmath181 , where @xmath182
it is a straightforward exercise to verify that the chronometric form @xmath131 for the separation between two points is invariant when the points of @xmath89 are subjected to transformations of the following type : @xmath183 here @xmath184 represents an arbitrary translation in quantum space - time , @xmath185 is an element of @xmath186 , and @xmath187 is the complex conjugate of @xmath188 .
the relation of this group of transformations to the poincar group in the case @xmath107 should be apparent .
indeed , one of the attractive features of the extension of space - time geometry that we are putting forward here is that the hyper - poincar group allows such a description , which implies a wide range of intuitively plausible physical features .
more generally , we remark that the proper hyper - poincar group preserves the signature of any space - time interval , whether or not the interval is degenerate , and hence leaves the causal relations between events unchanged .
we refer to a transformation of the form @xmath189 as a ` hyper - lorentz transformation ' if @xmath190 for some element @xmath191 . in the case of minkowski space ( @xmath107 ) it is well known that a two - component spinor @xmath192 can be represented by a complex null bivector @xmath193 conversely , the bivector @xmath194 determines @xmath195 up to the transformation @xmath196 this geometrical ambiguity is often referred to as the fundamental ` two - valuedness ' of two - component spinors in relativity theory . in the case of a general quantum space - time of dimension @xmath90 , a hyperspinor @xmath197 can be represented by a complex null @xmath49-vector ( antisymmetric tensor of rank @xmath49 ) of the form @xmath198 the @xmath49-vector @xmath199 is ` null ' in the sense that @xmath200 then @xmath199 determines @xmath197 up to transformations of the form @xmath201 where @xmath202 .
hence we can say that hyperspinors have a fundamental ` @xmath49-valuedness ' ( cf .
holm @xcite ) .
the real dimension of the hyper - lorentz group is @xmath203 , and thus the real dimension of the hyper - poincar group is @xmath204 .
we observe that the dimension of the hyper - poincar group grows linearly with the dimension of the quantum space - time itself , which is given by @xmath90 .
this can be contrasted with the dimension of the group arising if we endow @xmath205 with a standard lorentzian metric with signature @xmath206 . in that case
the associated pseudo - orthogonal group has real dimension @xmath207 , which together with the translation group gives a total dimension of @xmath208 .
the parsimonious dimensionality of the hyper - poincar group arises from the fact that it preserves the system of causal relations holding between pairs of points in quantum space - time . in a flat four - dimensional space - time
the symmetries of the poincar group are associated with a ten - parameter family of killing vectors .
thus , for @xmath107 we have the minkowski metric ( [ eq:7 ] ) , and the poincar group is generated by the ten - parameter family of vector fields @xmath209 on @xmath210 satisfying @xmath211 where @xmath212 denotes the lie derivative with respect to @xmath209 . for any vector field @xmath209 and any symmetric tensor field @xmath213
we have @xmath214 if @xmath213 is the metric and @xmath215 denotes the associated covariant derivative satisfying @xmath216 we obtain the killing equation @xmath217 where @xmath218 .
the condition @xmath219 therefore implies that @xmath220 is a killing vector . for @xmath108
we have no riemannian metric , and the usual relations between symmetries and killing vectors are lost .
what survives , however , is of interest . specifically , to generate a symmetry of the quantum space - time the vector field @xmath209 has to satisfy @xmath221 where @xmath222 is the chronometric tensor . for a general vector field @xmath209 and a general symmetric tensor field @xmath222
we have @xmath223 in the case of the quantum space - time @xmath89 we let @xmath224 be the natural flat connection for which @xmath225 then to generate a symmetry of the chronometric structure of @xmath89 the vector field @xmath209 has to satisfy @xmath226 equation ( [ eq:4.5 ] ) can be written in an alternative form if we define a symmetric tensor @xmath227 of rank @xmath228 by setting @xmath229 then ( [ eq:4.5 ] ) says that @xmath227 satisfies the conditions for a symmetric killing tensor : @xmath230 thus we see that @xmath89 provides an example of a symmetry group generated by a family of killing tensors .
the symmetries of the quantum space - time are generated , more specifically , by a system of @xmath204 irreducible symmetric killing tensors of rank @xmath228 .
the significance of killing tensors is that they are associated with conserved quantities .
for other examples of killing tensors arising in a physical context , see , e.g. , @xcite . in the present
setting it follows that if the vector field @xmath231 satisfies the geodesic equation , which on a quantum space - time of dimension @xmath90 is given , as we have seen , by @xmath232 and if @xmath227 is the killing tensor of rank @xmath228 given by ( [ eq:4.6 ] ) , then we have the following conservation law : @xmath233 in other words , the quantity @xmath234 is a constant of the motion .
it follows from the material of the previous section that in higher - dimensional quantum space - times the main conservation laws and symmetry principles of relativistic physics remain intact .
in particular , the conservation of hyper - relativistic momentum and angular momentum for a system of interacting particles can be given a well - defined formulation , the basic principles of which are similar to those applicable in the minkowskian case .
for this purpose it will be useful to introduce the notion of an ` elementary system ' in hyper - relativistic mechanics .
such a system is defined by its hyper - relativistic momentum and angular momentum .
the hyper - relativistic momentum of an elementary system is given by a momentum covector @xmath235 .
the associated mass @xmath236 is given ( cf .
@xcite ) by the following natural expression : @xmath237 the hyper - relativistic angular momentum of an elementary system is given by a tensor @xmath238 of the form @xmath239 where the hyperspinor @xmath240 is required to be trace - free : @xmath241 . the angular momentum is defined with respect to a choice of origin in such a manner that under a change of origin defined by a shift vector @xmath242 we have @xmath243 in the case @xmath107 these formulae reduce to the usual expressions for relativistic momentum and angular momentum in a minkowskian setting .
the real covector @xmath244 is invariant under a change of origin , and carries the interpretation of the intrinsic spin of the elementary system .
the magnitude of the spin is then defined by @xmath245 in the case of a set of interacting hyper - relativistic systems we require that the total momentum and angular momentum should be conserved .
this then implies conservation of the total mass and spin . in short
, we see that the idea of ` relativistic mechanics ' carries through nicely to the case of a general quantum space - time . now what is the interpretation of these conservation laws ?
we shall show later , once we introduce the idea of symmetry breaking , that hypermomentum can be interpreted as the momentum operator for a relativistic quantum system .
conservation of hypermomentum then can be thought of as conservation of four - momentum in relativistic quantum mechanics in the heisenberg representation .
in four - dimensional space - time it is useful in many contexts to examine the geometry of the space of complex null vectors at a point in the space - time .
this has the effect of giving us a vivid picture of the local causal relationships in space - time , and by sticking with a complex picture we also retain the link with quantum mechanics .
thus we consider complex vectors @xmath178 satisfying the quadratic equation @xmath246 in spinor terms this relation implies that the corresponding complex matrix @xmath247 is of the special form @xmath248 the space of complex vectors at a point in minkowski space is @xmath249 .
the space of complex directions ( which results if we consider equivalence class of vectors modulo overall proportionality ) is the complex projective 3-space @xmath250 .
the null directions constitute a quadric @xmath251 in that space , which owing to the decomposition ( [ eq:7.1 ] ) has the structure of a doubly ruled surface @xmath252 we can identify the first set of lines ( the @xmath253-lines ) with the projective unprimed spinors , and the second set of lines ( the @xmath254-lines ) with the projective primed spinors .
the quadric @xmath251 is ruled in such a way that two lines of the same type do not intersect , whereas two lines of the opposite type intersect at a point in @xmath251 ( see figure [ fig:2 ] ) .
this point corresponds to the null direction they jointly determine .
_ complex null directions in a four - dimensional space - time_. the space of directions at a point in complex minkowski space has the geometry of a complex projective 3-space .
the complex null directions constitute a quadric which has the structure of a doubly ruled surface @xmath255 . ] in the case of a general quantum space - time of dimension @xmath90 , we consider the space of complex vectors at a point , and examine the corresponding space of directions , which has the structure of a complex projective space @xmath256 .
the space of degenerate complex directions , which is given by the vanishing of the chronometric form @xmath257 , is a hypersurface in @xmath256 , which we shall call @xmath258 .
the points of @xmath258 correspond to degenerate directions of degree @xmath228 .
the null directions in @xmath259 correspond to those directions for which the associated degenerate vectors are of minimal rank and hence of the form @xmath260 .
these constitute a subvariety @xmath261 defined by the mutual intersection of a system of quadrics , given by the equation @xmath262 in this situation we have @xmath263 and we can identify the two systems of @xmath264-planes by which @xmath265 is foliated , which we refer to as @xmath253-planes and @xmath254-planes , as the spaces of projective unprimed and primed hyperspinors , respectively . the various degenerate directions of intermediate degree correspond to points in @xmath258 lying on the linear spaces spanned by the joins of @xmath266 points in @xmath265 ( @xmath267 ) .
the degree of degeneracy is given by the integer @xmath266 .
_ the space of complex null directions at a point in a quantum space - time of dimension nine_. in this case the space of degenerate directions is given by a cubic surface @xmath268 in the complex projective space @xmath269 .
the complex null directions then constitute a four - dimensional doubly foliated subvariety @xmath270 in @xmath268 . ] in the case @xmath109 , illustrated in figure [ fig:3 ] , the space of complex directions at a point in @xmath271 is @xmath272 , and the degenerate directions constitute a cubic hypersurface @xmath273 .
the null directions lie in the doubly foliated surface @xmath274 in @xmath268 .
the points of @xmath268 all lie on the ` first join ' of @xmath275 with itself ; in other words , any point of @xmath276 lies on a line joining two points of @xmath275 .
thus we write @xmath277 .
the space @xmath278 then consists of degenerate directions that are strictly of the second degree .
note that any point of @xmath269 can be represented as the join of three points in @xmath275 , and hence @xmath279 .
in the case @xmath280 , illustrated in figure [ fig:4 ] , the space of complex directions at a point in @xmath281 is the complex projective space @xmath282 , and the degenerate directions constitute a quartic hypersurface @xmath283 .
the null directions ( degenerate direction of the first degree ) lie on the doubly foliated surface @xmath284 in @xmath285
. the degenerate directions of the second degree lie on the first join of @xmath286 with itself : @xmath287 the degenerate directions of the third degree lie in @xmath288 and constitute the general elements of @xmath285 . _
the space of complex null directions at a point in the quantum space - time _ @xmath289 .
the degenerate directions of degree three constitute a quartic hypersurface @xmath285 in @xmath282 .
the null directions belong to a six - dimensional doubly foliated subvariety @xmath290 .
degenerate directions of degree two constitute a submanifold @xmath291 consisting of points that lie on lines spanned by pairs of points in the subvariety @xmath286 . ]
the proceeding discussion of complex null directions can be seen as a ` warm - up ' exercise for the introduction of the concept of hypertwistors . for this purpose
we introduce a notation that closely parallels the standard notation for @xmath107 penrose twistors .
let us denote by @xmath292 the complex vector space of dimension @xmath293 given by the pair @xmath294 .
we write @xmath295 for a typical element of @xmath296 .
such elements will be referred to as ` hypertwistors ' ( also called ` generalised twistors ' @xcite ) .
let @xmath297 denote the space of dual hypertwistors .
a natural pseudo - hermitian structure can be introduced on the geometry of hypertwistors by means of the complex conjugation operation that maps @xmath298 to @xmath299 .
the corresponding pseudo - hermitian form is then given by @xmath300 it is straightforward to verify that the inner product @xmath301 is invariant under the action of the group @xmath302 . in the case
@xmath107 , the elements of @xmath303 are standard penrose twistors , and the relevant symmetry group is @xmath304 . _ the klein representation for a general quantum space - time_. the aggregate consisting of all @xmath264-planes in the complex projective space @xmath305 forms a manifold of dimension @xmath90 .
this space is the complexified , compactified version of the quantum space - time @xmath306 . ]
the space @xmath305 of projective hypertwistors is a good starting point for the analysis of the conformal geometry of complex quantum space - time , which can be regarded as the grassmannian variety of projective @xmath264-planes in @xmath307 , as illustrated in figure [ fig:5 ] .
more precisely , the aggregate of all projective @xmath264-planes in @xmath307 constitutes a compact manifold of dimension @xmath90 , which we identify as the complexified , compactified quantum space - time @xmath308 .
the ` finite ' points of @xmath308 correspond to those @xmath264-planes of @xmath305 that are determined by a linear relation of the form @xmath309 for some fixed @xmath85 .
thus for each ( complex ) point @xmath85 in @xmath92 we obtain , according to equation ( [ eq:54 ] ) , an @xmath264-plane in @xmath305 .
the aggregate of such @xmath264-planes constitute the points of @xmath92 .
the @xmath264-planes for which @xmath85 is hermitian then constitute the points of the real quantum space - time @xmath89 .
the conformal structure of a quantum space - time is implicit in the various possibilities arising for the intersections of @xmath264-planes in the projective hypertwistor space @xmath307 . in the case
@xmath107 ( penrose twistors ) the projective space @xmath310 contains a four - dimensional family of complex projective lines , the aggregation of which constitutes the associated space - time ( see figure [ fig:6 ] ) . in this case , a pair of space - time points @xmath311 and @xmath312 are null - separated in the minkowskian sense if and only if the corresponding lines in @xmath310 intersect .
the space of all complex projective lines in @xmath310 constitutes a four - dimensional quadric hypersurface @xmath313 in @xmath314 , which we identify as complexified , compactified minkowski space - time @xmath315 .
the quadric @xmath313 contains two systems of projective 2-planes , called @xmath253-planes and @xmath254-planes .
the @xmath253-planes are in one - to - one correspondence with the points of @xmath310 , whereas the @xmath254-planes are in one - to - one correspondence with the 2-planes of @xmath310 , as illustrated in figure [ fig:7 ] . the case @xmath109 has been studied closely by finkelstein @xcite and his collaborators . in this case , the nine - dimensional quantum space - time @xmath316 arises when we consider the grassmanian of complex projective 2-planes in complex projective 5-space . a pair of planes in @xmath310 in general will not intersect . if they do , however , then there are three possibilities : they can intersect in a point , a line , or a plane .
these situations correspond to the various levels of degeneracy obtainable for the separation vector .
in particular , if the two planes intersect in a line , then the separation vector is null ( see figure [ fig:8 ] ) .
_ twistor theory_. the lines of the complex projective space @xmath310 correspond to the points of the complex space - time @xmath317 .
two points in @xmath318 are null - separated if and only if the corresponding lines in @xmath310 intersect . ] a pair of @xmath264-planes in @xmath305 will not in general intersect .
this generic non - intersection property corresponds to the nonvanishing of the chronometric form for the separation vector of generic quantum space - time points .
an @xmath264-plane in @xmath305 is represented by a simple skew hypertwistor @xmath319 of rank @xmath49 . by a ` simple ' skew hypertwistor we mean one of the form @xmath320}\end{aligned}\ ] ] for some set @xmath321 of @xmath49 hypertwistors .
suppose that the simple skew hypertwistors @xmath322 and @xmath323 represent , respectively , the @xmath264-planes @xmath324 and @xmath325 in @xmath305 .
a necessary and sufficient condition for the vanishing of the chronometric form for the corresponding space - time points is @xmath326 where @xmath327 is the skew hypertwistor of rank @xmath293 .
we note that ( [ eq:5.4 ] ) is symmetric under the interchange of @xmath324 and @xmath325 if @xmath49 is even , and antisymmetric if @xmath49 is odd .
the vanishing of ( [ eq:5.4 ] ) is the condition that the planes @xmath324 and @xmath325 contain a point in common .
this means that the skew hypertwistors @xmath319 and @xmath328 contain at least one hypertwistor as a common factor .
thus , a necessary and sufficient condition for a pair of quantum space - time points to have a degenerate separation is that the corresponding @xmath264-planes in @xmath305 should intersect .
_ complexified , compactified minkowski space - time as the klein quadric in complex projective 5-space . _
the aggregate of complex projective lines in @xmath310 constitutes a nondegenerate quadric @xmath313 in @xmath314 .
the quadric contains two distinct systems of projective 2-planes , called @xmath253-planes and @xmath254-planes .
any two distinct planes of the same type in @xmath313 intersect at a point .
two planes of the opposite type in @xmath313 will in general not intersect , but if they do , they intersect in a line
. the points of @xmath310 correspond to @xmath253-planes in @xmath313 , and the 2-planes of @xmath310 correspond to @xmath254-planes in @xmath313 . ]
more generally , the degeneracy @xmath266 of the separation of a pair of quantum space - time events is given by @xmath329 , where @xmath236 is the dimensionality of the intersection of the corresponding @xmath264-planes in @xmath305 .
the possible degrees of degeneracy are @xmath330 .
if we interpret the case of no intersection as an intersection of dimension @xmath331 , then a nondegenerate separation can be interpreted as a ` degeneracy of degree @xmath49 ' .
thus separations of degree less than @xmath49 are degenerate , whereas a separation of degree @xmath49 is nondegenerate .
the degree of degeneracy is given by the rank of the separation matrix @xmath332 .
alternatively , given two skew hypertwistors @xmath333 and @xmath334 , each with @xmath49 indices , let us form the dual hypertwistor by @xmath335 then @xmath266 is the maximum number of index contractions we can make between @xmath322 and @xmath336 without obtaining the result zero .
if a single index contraction gives zero , this corresponds to the case where @xmath322 is proportional to @xmath334 .
thus @xmath337 ( separation of degree zero ) can be interpreted as the ` completely degenerate ' case where the two space - time points coincide .
_ hypertwistor theory for @xmath109_. the aggregate of complex projective @xmath338-planes in @xmath314 constitutes a nine - dimensional quantum space - time @xmath339 .
in general , two such planes will not intersect . if two planes intersect in a point , then the separation vector of the corresponding space - time points is degenerate .
if the intersection is in a line , then the corresponding points are null - separated . ]
now we present a few elementary applications of hypertwistor theory . in the case of a four - dimensional space - time , it is well known that twistors can be characterised as solutions of the differential equation @xmath340 the so - called ` twistor equation ' .
indeed , the solution of this equation is @xmath341 and the associated twistor is determined by the fixed spinor pair @xmath342 .
more generally , in the case of a quantum space - time of dimension @xmath90 we consider the equation @xmath343 it is not difficult to verify that in the case @xmath107 equation ( [ eq:2.1 ] ) reduces to the conventional twistor equation ( [ eq:73 ] ) .
the solutions of ( [ eq:2.1 ] ) can be expressed in terms of generalised twistors : @xmath344 where @xmath345 and @xmath346 are constant .
this result is established in @xcite .
we note that a necessary and sufficient condition for a hypertwistor field @xmath347 to satisfy the hypertwistor equation ( [ eq:2.1 ] ) is that it should satisfy @xmath348c'}\xi^c=0\end{aligned}\ ] ] for all values of @xmath349 . for @xmath107
the twistor equation ( [ eq:73 ] ) is a special case of a more general nonlinear partial differential equation known as the geodesic shear - free condition : @xmath350 when @xmath108 , however , the connection between the twistor equation and the geodesic shear - free condition is lost , and the appropriate generalisation of the latter is @xmath351b'}\xi_{[c}\right ) \xi_{d]}=0 , \label{eq:79}\end{aligned}\ ] ] which reduces to the geodesic shear - free condition ( [ eq:78 ] ) when @xmath107 .
solutions of the equation ( [ eq:79 ] ) can be generated through the consideration of the intersections of projective @xmath264-planes in @xmath305 with certain types of complex analytic varieties in @xmath305 .
now let us consider the dynamical equations for fields of totally null momentum on a general quantum space - time .
the relevant field equations take the form @xmath352b } \phi = 0 , \label{eq:1.43}\end{aligned}\ ] ] @xmath353\cdots c ' } = 0 , \label{eq:1.42}\end{aligned}\ ] ] and @xmath354\cdots c } = 0 .
\label{eq:1.425}\end{aligned}\ ] ] these equations generalise the zero rest - mass equations for each helicity @xmath355 ( where @xmath356 ) in four - dimensional space - time . the solutions can be described briefly as follows .
we write @xmath357 and let @xmath358 denote an analytic , homogeneous function of degree @xmath359 defined on some region of the hypertwistor space .
for @xmath360 we set @xmath361 where the differential form @xmath362 is given by @xmath363 note that @xmath362 is homogeneous of degree @xmath49 in @xmath346 , and thus that the quantity appearing in the integral sign in ( [ eq:3.2 ] ) is homogeneous of degree zero .
similarly , for helicity @xmath364 we define @xmath365 and for helicity @xmath366 we define @xmath367 here it is understood that first we differentiate @xmath368 with respect to @xmath345 , and we set @xmath369 before integrating .
a proposition established in @xcite shows that the contour integrals ( [ eq:3.2 ] ) , ( [ eq:3.25 ] ) , and ( [ eq:3.255 ] ) satisfy the totally null momentum conditions ( [ eq:1.43 ] ) , ( [ eq:1.42 ] ) , and ( [ eq:1.425 ] ) , respectively . as an illustration , let us first take the simplest case , given by equation ( [ eq:3.2 ] ) for @xmath109 . for our hypertwistor function we take @xmath370 where @xmath371 , @xmath372 , and @xmath373 denote fixed points in the dual hypertwistor space . for their hyperspinor decomposition
we write @xmath374 and set @xmath375 where @xmath376 , @xmath377 , and @xmath378 are solutions of the primed analogue @xmath379 of the hypertwistor equation ( [ eq:2.1 ] ) . inserting this expression into the contour integral formula
, we obtain the result @xmath380 where the contour is taken to be @xmath381 .
it is then a straightforward exercise to verify directly that ( [ eq:4.66 ] ) satisfies the scalar hyper - wave equation ( [ eq:1.43 ] ) .
this example generalises the so - called ` elementary states ' of standard penrose twistor theory @xcite .
one of the motivations for the idea of quantum space - time is that it provides a possible way forward towards the unification of cosmology and elementary particle physics .
clearly some such unification is required for a coherent discussion of the early stages of the universe but in what mathematical framework should this unification be pursued ?
an added impetus to these considerations comes from the dark matter / energy problem ( see , e.g. , @xcite and references cited therein ) , which has had the effect of encouraging physicists to rethink the foundations of cosmology .
if we are to consider new classes of models , then some criteria are required to limit the range of possibilities . in particular , it is reasonable to propose that : ( a ) the model should be geometrical in character ; and ( b ) the model should be rich enough to admit within its space the possibility of a geometrisation of quantum mechanics ( by ` geometrisation ' we mean an approach in the spirit , e.g. , of @xcite ) .
our intention now is to put forward a tentative approach to ` quantum cosmology ' in this spirit , based on the idea of quantum space - time . in doing so , we bear in mind that there are a number of distinct inter - related geometrical structures arising in cosmology , that need to be taken into consideration as we proceed .
these include : ( a ) reality structure ( as in the distinction between a real and a complex space - time ) ; ( b ) causality structure ( identification of the light cones ) ; ( c ) infinity structure ( is the universe open or closed ? ) ; ( d ) chronometric structure ( how are time , distance , and energy measured ? ) ; and ( e ) singularity structure ( how does one characterise the beginning and end ? ) .
the theory of hyperspin is well suited for addressing these issues in a mathematically compelling way ( see @xcite for earlier examples of models for hypercosmologies ) .
it is convenient to begin the present analysis with a discussion of the structure of infinity in the case of a general flat quantum space - time .
this will lead us to more general cosmological considerations . as indicated above , for any skew hypertwistor @xmath382 of rank @xmath49 in a quantum space - time of dimension @xmath90 we define its dual @xmath383 by the relation @xmath384 here @xmath385 is the skew hypertwistor of rank @xmath293 , which is unique up to scale . depending on whether @xmath49 is even or odd , we have the following interchange relations : @xmath386 thus if @xmath49 is even , then once the scale is fixed we obtain a symmetric inner product on the space of skew hypertwistors of rank @xmath49 , which we denote @xmath387 if @xmath49 is odd then the product ( [ eq:6.3 ] ) gives us a symplectic structure . under complex conjugation
the skew hypertwistor @xmath388 becomes @xmath389 .
if @xmath319 is simple , thus corresponding to an @xmath264-plane @xmath324 in @xmath305 , then we say that @xmath324 is real if @xmath390 is proportional to @xmath391 .
the real @xmath264-planes of @xmath305 correspond to the real points of quantum space - time .
the structure at infinity in the compactified quantum space - time @xmath392 can be described as follows . in the hypertwistor space
@xmath305 we choose a real @xmath264-plane @xmath393 represented by a simple skew hypertwistor @xmath394 .
the point @xmath393 in @xmath392 corresponding to the @xmath264-plane @xmath393 in @xmath305 will be called the ` point at infinity ' .
the locus in @xmath392 consisting of all points that have a degenerate separation from @xmath393 will be called ` infinity ' .
it should be evident that infinity has a rich structure , with various domains that can be classified according to the degree of degeneracy of their separation from the point @xmath393 .
the ` finite ' points of @xmath392 are those for which the separation from @xmath393 is nondegenerate , i.e. those points @xmath324 for which @xmath395 . in the case of two finite quantum space - time points the chronometric form @xmath131 is given as follows : @xmath396 equivalently we can write @xmath397 if @xmath324 and @xmath393 are not null - separated , then we can choose the scales of @xmath319 and @xmath394 such that @xmath398 , without loss of generality , and similarly for @xmath323 and @xmath394 .
we note that @xmath399 is independent of the scale of @xmath319 and @xmath323 .
on the other hand , @xmath399 does depend on the scale of @xmath400 and the scale of @xmath394 .
it has an epsilon ` weight ' of @xmath331 and an @xmath393 ` weight ' of @xmath401 ( cf .
if we form the ratio associated with four hypertwistors @xmath324 , @xmath325 , @xmath402 , and @xmath403 , given by @xmath404 where @xmath405 , @xmath406 , @xmath49 , and @xmath147 are the quantum space - time points corresponding to @xmath324 , @xmath325 , @xmath402 , and @xmath403 , respectively , then we obtain an expression that is absolute that is to say , a geometric invariant .
this is because @xmath399 has the ` dimensionality ' of time raised to the power @xmath49 ; whereas the ratio ( [ eq:6.5 ] ) arises as a comparison of two such time intervals , and thus is dimensionless .
the basic chronometric geometry , with infinity chosen as indicated above , admits no absolute or ` preferred ' unit of time : in this geometry only ratios of time intervals have an absolute meaning .
there is , on the other hand , no reason _ a priori _ why just such a structure should apply at infinity .
other choices are available for @xmath394 , and these have the effect of giving @xmath407 the structure of a cosmological model . in the case @xmath107 , for example , if @xmath408 is chosen to be real and non - simple , then the quadratic form @xmath409 which has an epsilon weight of one and an @xmath393-weight of two , has the dimensionality of inverse squared - time . hence in this case there _
is _ a preferred unit of time .
other time intervals can then be expressed in multiples of the preferred unit of time . _ time - like geodesics in minkowski space . _
each time - like geodesic in a flat four - dimensional space - time determines a quadric in projective twistor space .
a quadric in @xmath310 has two distinct systems of generators . in the case of a time - like geodesic
, the line @xmath393 belongs to one of the two systems .
the corresponding geodesic is a conic curve in the klein quadric @xmath410 , and has the property that it passes through the point @xmath393 .
the second system of generators then corresponds to a polar conic that lies on the null cone infinity and does not go through @xmath393 . ] to pursue this point further , we recall that @xmath411 has the structure of a quadric @xmath410 in @xmath314 . more specifically , for @xmath107 the space of skew rank two twistors is @xmath412 , which is projectively @xmath314 , and @xmath411 is the locus defined by the homogeneous quadratic equation @xmath413 infinity in @xmath411 can then be defined by the intersection of @xmath411 in @xmath314 with the projective 4-plane @xmath414 given by the equation @xmath415 if @xmath408 is simple , then @xmath414 is tangent to @xmath411 , and the intersection is a cone the null cone at infinity .
the geometry of this space plays a crucial role in determining the properties of time - like geodesics in minkowski space , as discussed in figure [ fig:9 ] . on the other hand ,
if @xmath408 is not simple , then the intersection @xmath416 is a 3-quadric . the resulting geometry ,
if @xmath408 is real , is that of de sitter space .
the metric on @xmath416 in this case is given by @xmath417 and the parameter @xmath160 defined by ( [ eq:105 ] ) has the interpretation of being the associated cosmological constant .
we note that @xmath418 is independent of the scale of @xmath419 , and has an @xmath393-weight of @xmath331 , as is appropriate for an element of time .
the de sitter group consists of transformations of @xmath314 that preserve both the quadric @xmath410 and the point @xmath393 . with the incorporation of some additional structure at infinity
, the entire class of robertson - walker cosmological models can be represented in a similar way @xcite .
the idea can be described briefly as follows .
we start with the complex projective space @xmath420 and in it the quadric @xmath410 defined by ( [ eq:99 ] ) .
the points of space - time are given by a reality structure by requiring that @xmath421 .
next we introduce a pencil of 4-planes in @xmath314 of the form @xmath422 where @xmath423 . for each such @xmath424 ,
the intersection of the 4-plane @xmath425 with the real quadric given by @xmath426 defines a certain subspace of the space - time . for certain choices of @xmath427
the resulting subspace can be interpreted ( in some cases with the deletion of certain elements ) as a constant - time space - like hypersurface @xmath428 in the space - time .
the corresponding cosmological model is then obtained by selecting a one - parameter ` chronological family ' of such surfaces , and choosing an appropriate conformal factor for the metric geometry . in more detail ,
the construction is as follows . for @xmath429 ( a closed universe )
we require the skew twistors @xmath430 and @xmath431 to be complex and to satisfy the following relations : @xmath432 we then let @xmath433 , @xmath434 for some @xmath435 $ ] .
the resulting one - parameter family of surfaces determined by @xmath436 has the property that each is a 3-sphere . for a @xmath437 cosmology
we require @xmath438 in this case the relevant family of 4-planes is given by ( [ eq:102 ] ) with @xmath439 , the overall scale of @xmath424 being unimportant .
for @xmath440 we set @xmath441 with @xmath134 as above in the @xmath437 case .
_ friedmann - robertson - walker cosmologies_. in the case @xmath429 depicted here the chronology of the frw cosmology is generated by a pencil of projective 4-planes hinged on an axis @xmath442 in @xmath443 .
the axis is chosen so that it does not impinge on any of the real points of the space - time . as a consequence the constant - time hypersurfaces in the cosmological model are compact . ] in each case we can consider the ` axis ' obtained by intersecting the elements of the given family of 4-planes . in the @xmath444 case , the axis itself does not intersect the associated real space - time , and as a consequence the resulting hypersurfaces of constant time are topologically 3-spheres ( see figure [ fig:10 ] ) . in the @xmath440 case
the axis ` touches ' the space - time at a point ( given by @xmath445 ) common to all of the intersection spaces .
if we remove this point ( or treat it as a point at infinity ) , then the resulting constant - time surfaces are each topologically @xmath446 . in the @xmath437 case ,
the common intersection region is a 2-sphere , and as a consequence an ` open ' cosmological model results in this case as well .
_ chronological foliation of an @xmath280 hypercosmology_. the compactified quantum space - time @xmath447 is a real submanifold of the complex projective space @xmath448 .
a pencil of @xmath449-hyperplanes hinged on a complex axis @xmath450 intersects the space - time in a set of hypersurfaces .
depending on the reality structure of the pencil , and its real subparameterisation , a variety of different possibilities can emerge for the global structure of the hypersurface family . ]
thus we see that the algebraic geometry of twistor theory gives us an essentially ` unified ' point of view over the various standard cosmological models this approach can be pursued at greater length , giving rise to a geometrical characterisation of the different types of situations that can occur , depending , in particular , on the global structure and topology of the space - time , on the equation of state of the fluid representation in the energy tensor , and on the type of cosmological constant ( if any ) in the model .
much more detail , along with specific examples for various choices of the equation of state , can be found in @xcite .
it is interesting to note that more or less the same state of affairs prevails in higher dimensions ( see figure [ fig:11 ] for an example of a sixteen - dimensional class of ` quantum cosmologies ' analogous to the @xmath429 friedmann - robertson - walker models ) . in other words ,
the choice of structure at infinity gives rise to various possible global structures for the quantum space - time , and in particular , to a chronometric form that is in general not flat , thus making @xmath392 a cosmological model . in the case of a standard four - dimensional cosmological model based on einstein s
theory , the existence of structure at ( or ` beyond ' ) infinity has a bearing on the geometry of space - time alone . in the case of a quantum cosmology ,
however , the structure at infinity also has implications for microscopic physics .
for instance , whereas in the four - dimensional de sitter cosmology the relevant structure at infinity contains the ` invariant ' information of one dimensional constant ( the cosmological constant ) , in the higher - dimensional situation there are in general a number of such constants that may emerge as geometrical invariants of the theory . thus within a single geometric framework one has the scope for introducing structure ( or what amounts to the same thing the breaking of symmetry ) both on a global or cosmological scale , as well as on the microscopic scales of distance , time , and energy associated with the phenomenology of elementary particles .
one might say that in these models the structure at infinity is playing the role of the higgs fields .
one can even envisage the possibility of explaining , in a purely geometrical language , the basis of the remarkable coincidences involving various fundamental constants of nature that have puzzled physicists for many decades .
as a prelude to our discussion of the idea of symmetry breaking in quantum space - time , we digress briefly to review the notions of weak and strong hermiticity .
this material is relevant to the origin of unitarity in quantum mechanics . intuitively speaking
, we observe that when the weak hermiticity condition is imposed on a hyperspinor @xmath85 representing a space - time event , then @xmath85 belongs to the real subspace @xmath87 . as such
, the hyper - relativistic symmetry of quantum space - time is not affected by the imposition of this condition .
if , however , we break the hyper - relativistic symmetry by selecting a preferred time - like direction , then we can speak of a stronger reality condition whereby an isomorphism is established between the primed and unprimed hyperspin spaces .
we begin with the weak hermitian property .
let @xmath51 denote , as before , an @xmath49-dimensional complex vector space .
we also introduce the associated spaces @xmath56 , @xmath451 , and @xmath57 . in general , there is no natural isomorphism between @xmath52 and @xmath56 .
therefore , there is no natural matrix multiplication law or trace operation defined for elements of @xmath452 .
nevertheless , certain matrix operations are well defined .
for example , the determinant of a generic element @xmath453 is given by @xmath454 the weak hermitian property is also well - defined .
in particular , if @xmath455 is the complex conjugate of @xmath453 , then we say that @xmath453 is weakly hermitian if @xmath456 .
as we have observed , for many applications , weak hermiticity suffices .
now we consider the strong hermitian property . in some situations there
may exist a natural map @xmath457 defined by the context of the particular problem .
such a map is called a hermitian correlation . in this case
, the complex conjugate of an element @xmath53 determines an element @xmath458 .
for any element @xmath459 we define the operations of determinant , matrix multiplication , and trace in the usual manner .
the determinant is @xmath460 and the hermitian conjugate of @xmath461 is @xmath462 .
the hermitian correlation is given by the choice of a preferred element @xmath463 .
then we write @xmath464 where @xmath465 is now called the complex conjugate of @xmath466 .
when there is a hermitian correlation @xmath467 , we call the condition @xmath468 the strong hermitian property . thus once we break the relativistic invariance by introducing a preferred element @xmath469 that determines a hermitian correlation , we may carry out specific calculations in that frame . to gain a better understanding of this ,
consider an event in complex minkowski space defined by its separation @xmath470 from the origin .
the complex conjugate of @xmath470 is @xmath471 , so @xmath470 is real if @xmath472 .
let @xmath473 be a fixed time - like vector satisfying @xmath474 .
with respect to this choice of @xmath473 , the trace of @xmath475 is defined by @xmath476 once @xmath473 is fixed , we may represent @xmath470 in terms of pauli matrices , according to which @xmath470 admits a matrix representation of the form ( [ eq:2 ] ) .
the time variable is then given by @xmath477 and the minkowski metric is given by the determinant ( [ eq:3 ] ) .
we can also define a commutator for a pair of space - time position vectors @xmath470 and @xmath478 by setting @xmath479 the geometrical meaning of ( [ eq:113 ] ) is that @xmath247 corresponds to the ordinary 3-space cross - product of the projection of the vectors @xmath470 and @xmath478 onto the space - like hyperplane orthogonal to @xmath473 that passes through the origin .
analogously , we can discuss the ` spectral ' properties of vectors with respect to a given choice of @xmath473 and a given choice of origin in space - time .
the two eigenvalues of @xmath470 are then given by @xmath480 .
thus two such vectors @xmath470 and @xmath478 are isospectral if and only if they lie on a common sphere about the origin lying in the given space - like hyperplane .
similar remarks apply to higher - dimensional quantum space - times .
now we proceed to introduce a natural mechanism for symmetry breaking that arises in the case of a standard ` flat ' quantum space - time endowed with the canonical structures associated with reality and infinity .
we shall make the point in particular that the breaking of symmetry in quantum space - time is intimately linked to the notion of quantum entanglement . according to this point of view the introduction of symmetry - breaking in the early stages of the universe
can be understood as a phase transition , or a sequence of phase transitions , the ultimate consequence of which is an approximate disentanglement of a four - dimensional ` classical ' space - time . in practical terms
the breaking of symmetry is represented in our framework by an ` index decomposition ' . in particular ,
if the dimension @xmath49 of the hyperspin space is not a prime number , then a natural method of breaking the symmetry arises by consideration of the decomposition of @xmath49 into factors .
the specific essential assumption that we shall make at this juncture will be that the dimension of the hyperspin space @xmath51 is _
even_. then we write @xmath481 , where @xmath482 , and set @xmath483 where @xmath484 is a standard two - component spinor index , and @xmath485 will be called an ` internal ' index @xmath486 .
thus we can write @xmath487 , where @xmath488 is a standard spin space of dimension two , and @xmath489 is a complex vector space of dimension @xmath490 .
the breaking of the symmetry then amounts to the fact that we can identify the hyperspin space with the tensor product of these two spaces .
we shall assume , moreover , that @xmath489 is endowed with a strong hermitian structure , i.e. we shall assume that there is a canonical anti - linear isomorphism between the complex conjugate of the internal space @xmath489 and the dual space @xmath491 .
if @xmath492 , then we write @xmath493 for the complex conjugate of @xmath494 , where @xmath495 .
we see therefore that @xmath496 is a complex hilbert space and indeed although here we consider for technical simplicity the case for which @xmath490 is finite , one should have in mind also the general infinite dimensional situation . for the other hyperspin spaces we write @xmath497 respectively .
these equivalences preserve the duality between @xmath51 and @xmath56 , and between @xmath451 and @xmath57 ; and at the same time are consistent with the complex conjugation relations between @xmath51 and @xmath52 , and between @xmath63 and @xmath57 . hence if @xmath498 then under complex conjugation we have @xmath499 , and if @xmath500 then @xmath501 .
in the case of a quantum space - time vector @xmath105 we have a corresponding structure induced by the identification @xmath502 when the quantum space - time vector is real , the weak hermitian structure on @xmath105 is manifested in the form of a standard weak hermitian structure on the two - component spinor index pair , together with a strong hermitian structure on the internal index pair .
in other words , the hermitian condition on the space - time vector @xmath105 is given by @xmath503 one consequence of these relations is that we can interpret each point in quantum space - time as being a space - time valued operator .
the ordinary classical space - time then ` sits ' inside the quantum space - time in a canonical manner
namely , as the locus of those points of quantum space - time that factorise into the product of a space - time point @xmath470 and the identity operator on the internal space : @xmath504 thus , in situations where special relativity is a satisfactory theory , we regard the relevant events as taking place on or in the immediate neighbourhood of this embedding of minkowski space in @xmath505 . this picture can be presented in more geometric terms as follows .
the hypertwistor space @xmath305 in the case @xmath481 admits a segr embedding of the form @xmath506 many such embeddings are possible , though they are all equivalent to one another under the action of the overall symmetry group @xmath507 . if the symmetry is broken and one such embedding is selected out , then following the conventions discussed earlier we can introduce homogeneous coordinates and write @xmath508 for the hypertwistor . here
the greek letter @xmath509 denotes an ordinary twistor index @xmath510 and @xmath485 denotes an internal index @xmath511 .
these two indices , when clumped together , constitutes a hypertwister index . the segr embedding consists of those points in @xmath512 for which we have a product decomposition of the associated hypertwistor given by @xmath513 the idea of symmetry breaking that we are putting forward here is related to the notion of disentanglement in standard quantum mechanics ( cf .
gibbons 1992 ; brody & hughston 2001 ) . that is to say , at the unified level
the degrees of freedom associated with space - time symmetry are quantum mechanically entangled with the internal degrees of freedom associated with microscopic physics .
the phenomena responsible for the breakdown of symmetry are thus analogous to the mechanisms of decoherence through which quantum entanglements are gradually diminished .
some readers may raise the objection that surely it is impossible to unify the unitary symmetries of elementary particle phenomenology with the symmetries of space - time ( cf .
, e.g. , @xcite )
. it should be noted , however , that our approach is not to attempt to embed a relativistic symmetry group in a higher - dimensional unitary group , but rather to embed the unitary group in a higher - dimensional relativistic symmetry group .
our methodology is consistent with the point of view put forward by penrose that for a coherent unification of general relativity and quantum mechanics , the rules of quantum theory must undergo ` profound modification ' @xcite .
the compactified complexified quantum space - time @xmath514 can be regarded as the aggregate of projective @xmath515-planes in @xmath512 .
now generically a @xmath516 in @xmath512 will not intersect the segr variety @xmath517 such a generic @xmath515-plane corresponds to a generic point in @xmath518 .
the @xmath515-planes that correspond to the points of compactified complexified minkowski space can be constructed as follows . for each line @xmath519 in @xmath520
we consider the subvariety @xmath521 where @xmath522 .
for any algebraic variety @xmath523 @xmath524 we define the _ span _ of @xmath525 to be the projective plane spanned by the points of @xmath525 .
we say a point @xmath526 in the ambient space @xmath527 lies in the span of the variety @xmath525 if and only if there exist @xmath236 points in @xmath525 for some @xmath528 with the property that @xmath526 lies in the @xmath529-plane spanned by those @xmath236 points .
the dimension @xmath154 of the span of @xmath525 satisfies @xmath530 ; however , the value of @xmath154 depends on the geometry of @xmath525 . the linear span of the points in @xmath531 , for any given @xmath519 , is a @xmath515-plane .
this is the @xmath532 in @xmath512 that represents the point in @xmath533 corresponding to the line @xmath519 in @xmath310 .
the aggregate of such special @xmath515-planes , defined by their intersection properties with the segr variety @xmath534 , constitutes a submanifold of @xmath514 , and this submanifold is compactified complexified minkowski space @xmath535 .
thus we see that once the symmetry of quantum space - time @xmath536 has been broken in the particular way we have discussed , then ordinary minkowski space can be identified as a submanifold .
let us now consider the implications of our symmetry breaking mechanism for fields defined on quantum space - time .
as an example , let @xmath537 be a scalar field on quantum space - time .
after we break the symmetry by writing @xmath538 , we consider a taylor expansion of the field around the embedded minkowski space - time .
specifically , for such an expansion we have @xmath539 where @xmath540 and @xmath541 therefore , the order zero term has the character of a classical field on minkowski space , and the first order term can be interpreted as a ` multiplet ' of fields , transforming according to the adjoint representation of the internal symmetry group @xmath542 .
_ symmetry breaking and matter formation in a hypertwistorial quantum cosmology_. in its earliest stages the universe is highly symmetrical .
eventually , symmetry is broken , and a four - dimensional ` classical ' cosmology freezes out , becoming largely disentangled from the reminder of the quantum space - time .
the formation of matter during the disentanglement process may not be confined to the four - dimensional subspace .
if such a scenario prevails , the bulk of this ` dark material ' is likely to remain outside the four - dimensional subspace , though nevertheless having an impact on its dynamics . ] in this connection we note that the symmetry breaking mechanism that we have proposed here has yet another representation namely , the expression of a hypertwistor as a multi - twistor system , i.e. as a multiplet of penrose twistors .
the physical and geometrical characteristics of such @xmath490-twistor systems have been analysed at great length by a number of authors ( see , e.g. , @xcite and references cited therein ) , and it is interesting therefore to see the direct link with hypertwistor theory and quantum space - time geometry .
it is fitting also to make a tribute here to the work of zoltan perjs , whose extensive contributions to relativity theory include , in particular , a number of important studies concerning the properties of @xmath490-twistor systems and their symmetries @xcite .
it is tempting to speculate that even in a more dynamic context some version of the symmetry breaking mechanism provided here will manifest itself . in this picture
we would envisage the earliest stages of the universe as being highly symmetrical , rather in the spirit of penrose s weyl curvature hypothesis @xcite , with no appreciable distinction between the conventional space - time degrees of freedom and the internal degrees of freedom associated with quantum theory .
nevertheless , the causal geometry of the universe remains well defined , and it is interesting to ask whether there might be some scenario within the rather rich causal structure of a quantum space - time that would allow us to account for the so - called ` horizon problem ' . in any event , once symmetry breaking takes place and this may happen in stages , corresponding to a successive factorisation of the underlying hypertwistor space then it makes sense to think of ordinary four - dimensional space - time as becoming more or less disentangled from the rest of the universe , and behaving in a way that is to some extent autonomous .
nonetheless , we might reasonably expect its global dynamics , on a cosmological scale , to be affected by the distribution of mass and energy elsewhere in the quantum space - time as well ( see , e.g. , figure [ fig:12 ] ) .
the embedding of minkowski space in the quantum space - time @xmath505 given by ( [ eq:11.4 ] ) implies a corresponding embedding of the poincar group in the hyper - poincar group .
this can be seen as follows .
the standard poincar group in @xmath543 consists of transformations of the form @xmath544 and the hyper - poincar transformations in @xmath505 are of the form @xmath545 with the identification @xmath546 , the general hyper - poincar transformation in the broken symmetry phase can be expressed in the form @xmath547 thus the embedding of the poincar group as a subgroup of the hyper - poincar group is given by @xmath548 bearing this in mind , we now construct a class of maps from the general even - dimensional quantum space - time @xmath505 to minkowski space @xmath543 .
it turns out @xcite that under rather general physical assumptions such maps are necessarily of the form @xmath549 where @xmath550 is a density matrix . as usual , by a density matrix we mean a positive semi - definite hermitian matrix with unit trace .
thus the maps arising here can be regarded as quantum expectations . in particular ,
let @xmath551 satisfy the following conditions : ( i ) @xmath552 is linear and maps the origin of @xmath505 to the origin of @xmath543 ; ( ii ) @xmath552 is poincar invariant ; and ( iii ) @xmath552 preserves causal relations
. then @xmath552 is given by a density matrix on the internal space .
[ theo:2 ] the general linear map from @xmath505 to @xmath553 preserving the origin is given by @xmath554 where @xmath555 is weakly hermitian .
now suppose that we subject @xmath505 to a poincar transformation of the form ( [ eq:12.2 ] ) , and require the corresponding transformation of @xmath556 should be of the form ( [ eq:12.1 ] ) .
if @xmath552 satisfies these conditions then we shall say that the map @xmath552 is poincar invariant .
clearly , poincar invariance holds if and only if @xmath557 for all @xmath558 , for all @xmath559 , and for all @xmath560 .
thus we have @xmath561 for all @xmath562 , and @xmath563 for all @xmath564 .
equation ( [ eq:12.8 ] ) implies that @xmath552 is of the form @xmath565 for some @xmath550 .
then ( [ eq:12.9 ] ) implies that @xmath552 must satisfy the trace condition @xmath566 .
finally we require that if @xmath567 and @xmath568 are quantum space - time points with the property that the interval @xmath569 is future - pointing then @xmath570 is also future pointing , where @xmath571 this is the requirement that @xmath552 should be a ` causal ' map .
this condition implies that @xmath552 must be positive semi - definite .
in particular , if @xmath572 is future - pointing then it must be of the form @xmath573 consider therefore the case for which @xmath574 is null .
then we require that the expression @xmath575 should be future - pointing ( or vanish ) for any choice of @xmath8 .
in particular , we require that the vector @xmath576 should be future - pointing if @xmath8 is of the form @xmath577 for any choice of @xmath192 and @xmath494 .
this means that the inequality @xmath578 holds for all @xmath494 , which shows that @xmath550 is positive semi - definite .
since we have shown that the trace of @xmath550 is unity , it follows that @xmath550 is a density matrix .
this result shows how the causal structure of quantum space - time is linked with the probabilistic structure of quantum mechanics .
the concept of a quantum state emerges when we ask for consistent ways of ` averaging ' over the geometry of quantum space - time in order to obtain a reduced description of physical phenomena in terms of the geometry of minkowski space .
we see that a probabilistic interpretation of the map from a general quantum space - time to minkowski space arises as a consequence of elementary causality requirements .
we can thus view the space - time events in @xmath505 as representing quantum observables , the expectations of which correspond to points of @xmath543 .
dcb gratefully acknowledges financial support from the royal society .
the work described here is based , in part , on ideas and suggestions arising in discussions with e. j. brody .
the authors are grateful to participants at the xixth max born symposium , institute of theoretical physics , wroclaw , poland , for helpful comments .
borowiec , a. 1993 _ g_-structure for hypermanifold , in _ spinors , twistors , clifford algebras and quantum deformations _
( sobtka castle , 1992 , z. oziewicz , b. jancewicz , & a. borowiec , eds . ) _ fund .
theories phys .
_ * 52 * , 75 - 79 , dordrecht : kluwer .
hughston , l.p .
1990 a remarkable connection between the wave equation and pure spinors in higher dimensions , in _ further advances in twistor theory , vol .
i : the penrose transform and its applications _ ( l. j. mason & l. p. hughston , eds . )
harlow : longman .
hurd , t. r. 1995 cosmological models in @xmath420 , in _ further advances in twistor theory , vol .
ii : integrable systems , conformal geometry and gravitation _ ( l. j. mason , l. p. hughston & p. z. kobak , eds . ) harlow : longman .
penrose , r. 1995 twistors for cosmological models , in _ further advances in twistor theory , vol .
ii : integrable systems , conformal geometry and gravitation _ ( l. j. mason , l. p. hughston & p. z. kobak , eds . ) harlow : longman .
pirani , f. a. e. 1965 introduction to gravitational radiation , in lectures on general relativity : 1964 brandeis summer institute in theoretical physics , vol . 1 ( s. deser & k. w. ford , eds . ) englewood cliffs , nj : prentice - hall . | the purpose of this paper is to present a model of a ` quantum space - time ' in which the global symmetries of space - time are unified in a coherent manner with the internal symmetries associated with the state space of quantum - mechanics .
if we take into account the fact that these distinct families of symmetries should in some sense merge and become essentially indistinguishable in the unified regime , our framework may provide an approximate description of or elementary model for the structure of the universe at early times .
the quantum elements employed in our characterisation of the geometry of space - time imply that the pseudo - riemannian structure commonly regarded as an essential feature in relativistic theories must be dispensed with .
nevertheless , the causal structure and the physical kinematics of quantum space - time are shown to persist in a manner that remains highly analogous to the corresponding features of the classical theory . in the case of the simplest conformally flat cosmological models arising in this framework
, the twistorial description of quantum space - time is shown to be effective in characterising the various physical and geometrical properties of the theory .
as an example , a sixteen - dimensional analogue of the friedmann - robertson - walker cosmologies is constructed , and its chronological development is analysed in some detail .
more generally , whenever the dimension of a quantum space - time is an even perfect square , there exists a canonical way of breaking the global quantum space - time symmetry so that a generic point of quantum space - time can be consistently interpreted as a quantum operator taking values in minkowski space . in this scenario ,
the breakdown of the fundamental symmetry of the theory is due to a loss of quantum entanglement between space - time and internal quantum degrees of freedom .
it is thus possible to show in a certain specific sense that the classical space - time description is an emergent feature arising as a consequence of a quantum averaging over the internal degrees of freedom .
the familiar probabilistic features of the quantum state , represented by properties of the density matrix , can then be seen as a by - product of the causal structure of quantum space - time .
address = blackett laboratory , imperial college , london sw7 2bz , uk address = department of mathematics , king s college london , the strand , london wc2r 2ls , uk |
magnetised plasma dynamics lies at the heart of many fascinating phenomena in astro , space and laboratory physics .
turbulence in the solar wind @xcite and in the interstellar medium @xcite , solar @xcite , stellar @xcite and accretion disk flares @xcite , substorms in the earth s magnetosphere @xcite , and turbulent transport and instabilities in magnetised fusion experiments @xcite , are just a few examples of remarkable physics problems whose solution is indeed determined by understanding the behaviour of plasmas in a magnetised environment . in many of these cases ,
( i ) the collision frequency is much lower than the typical frequencies of the physical phenomena of interest ( e.g. , turbulence , magnetic reconnection ) i.e. , the plasmas are weakly collisional ; and ( ii ) the size of the ion larmor orbit is several orders of magnitude smaller than the size of the system .
weak collisionality implies that on the timescales of interest the plasma can not be treated as a fluid , and instead a kinetic description that evolves the particles distribution functions is required .
this is rather unfortunate from the computational point of view , since fully kinetic models live on a six - dimensional phase - space ( each particle is characterised by its position and velocity vectors ) .
the strong magnetisation , however , implies that the plasma is highly anisotropic , with very different particle motions along and across the magnetic field direction .
this anisotropy can be explored analytically to yield reduced kinetic models , i.e. , asymptotic descriptions that reduce the phase - space to only 5d or even 4d .
this leads to tremendous computational savings and effectively renders possible calculations that would otherwise not be feasible on today s supercomputers .
gyrokinetics @xcite is a rigorous description of strongly magnetised , weakly - collisional plasmas .
the key idea behind the gyrokinetic formalism is that , because of the strong magnetic ( guide ) field , the particles larmor gyration frequency is much higher than the frequencies of dynamical interest , and can thus be averaged over .
this allows for the reduction of the dimensionality of the system , from 6d ( three position and three velocity coordinates ) to 5d ( three position coordinates , and velocities parallel and perpendicular to the magnetic field ) while retaining all the essential physical effects .
gyrokinetics was originally motivated by the attempt to model microinstabilities in magnetic fusion experiments ; in this respect it has been rather successful @xcite . as recognition of its usefulness
, the range of applications of gyrokinetics has broadened in recent years ; it is now routinely applied to the study of turbulence in magnetised astrophysical systems @xcite , and there have also been some studies pioneering its application to the problem of magnetic reconnection @xcite .
this reduction of the dimensionality of the system allowed by gyrokinetics is extremely advantageous from the numerical point of view .
nonetheless , intrinsically multiscale problems such as kinetic turbulence and reconnection remain formidable computational challenges .
further simplification where possible is therefore desirable .
one possible such simplification of gyrokinetics has recently been proposed by zocco and schekochihin @xcite : the kinetic reduced electron heating model ( krehm ) , a rigorous asymptotic limit of gyrokinetics valid for plasmas such that [ krehm_order ] _
e~m_e / m_i , where @xmath0 is the electron beta , @xmath1 are the background electron density and temperature , respectively , and @xmath2 is the background magnetic field .
under this assumption , ref .
@xcite shows that it is possible to reduce the plasma dynamics to a 4d phase - space position and velocity parallel to the magnetic field while retaining key physics such as phase - mixing and electron landau damping , ion finite larmor radius effects , electron inertia , electron collisions and ohmic resistivity .
this is a very significant simplification of the full kinetic description , which renders possible truly multiscale kinetic simulations . in particular , because no _ ad hoc _ fluid closure is employed , krehm can be used for detailed studies of energy conversion and dissipation in kinetic turbulence and reconnection , including electron heating via phase - mixing and landau damping . if taken literally , the ordering imposed by is somewhat restrictive , and obviously excludes many plasmas of interest .
examples of plasmas where it may hold are some regions of the solar corona @xcite , the large plasma device ( lapd ) experiment at ucla @xcite , and edge regions in some tokamaks @xcite does tend to be rather low in this region . ] .
however , one may legitimately expect that the plasma behaviour captured by krehm will qualitatively hold beyond its rigorous limits of applicability , as is so often the case with many other simplified plasma models ( mhd being a notorious example of a description known to work rather well far outside its strict limits of validity ) .
hints that this may indeed be the case are offered in , where a direct comparison of krehm with a ( non - reduced ) gyrokinetic model for the linear collisionless tearing mode problem yields very good agreement at values of @xmath3 significantly larger than @xmath4 .
this paper reports on the numerical methods and algorithms used in , the first numerical code to solve this particular set of equations .
an extensive series of tests and benchmarks is also presented .
considerable attention is devoted to orszag - tang - type decaying turbulence , both in the fluid and kinetic regimes .
the reader interested in the application of this code and physics model to the problem of magnetic reconnection is referred to @xcite , where the importance of electron heating via landau damping in reconnection is demonstrated for the first time .
a second set of equations solved by are the kinetic reduced mhd ( krmhd ) equations @xcite , which describe the evolution of compressible fluctuations ( density and parallel magnetic field ) in the regime @xmath5 ( @xmath6 being the wave number perpendicular to the guide - field of a typical perturbation , and @xmath7 the ion larmor radius . )
these equations are structurally identical to those of krehm , so their numerical implementation in is straightforward .
we also note that krehm reduces to the standard reduced - mhd ( rmhd ) set of equations @xcite in the appropriate limit ( i.e. , when the wave length of the fluctuations is much larger than all the kinetic scales ) .
thus , can also be used as a rmhd code ( in either 2d or 3d slab geometry ) .
finally , we remark that under an isothermal closure for the electrons , krehm reduces to the simple two - field gyrofluid model treated in refs .
@xcite ( which is a limit of the more complete models of snyder @xcite and of schep @xcite ) .
this paper is organized as follows .
presents the different sets of equations integrated by .
the kinetic equations are solved by means of a hermite expansion , which requires some form of closure ( or truncation ) .
this is discussed in , where an asymptotically exact nonlinear closure for the hermite - moment hierarchy is derived .
presents the energy evolution equation for the closed krehm model ; and the normalizations that we adopt are laid out in .
deals with the numerical discretization of the equations , including in a discussion of the implementation of a spectral - like scheme for the advection in the direction along the guide - field : a combination of an optimal third order total variation diminishing ( tvd ) runge kutta @xcite for the time derivative with a seventh - order upwind scheme for the fluxes @xcite . a series of linear and nonlinear benchmarks of the code are presented in , with emphasis on orszag - tang - type decaying turbulence test cases .
finally , the main points and results of this paper are summarised in .
also included for reference in appendix a is the recently proposed modification of the krehm model to allow for background electron temperature gradients @xcite .
solves two distinct sets of equations : ( i ) the kinetic reduced electron heating model ( krehm ) equations @xcite and ( ii ) the kinetic reduced mhd ( krmhd ) equations @xcite .
these models are briefly discussed below ; we refer the interested reader to the original references for a detailed derivation of the equations of each model . the kinetic reduced electron heating model ( krehm ) derived in ref .
@xcite is a rigorous asymptotic reduction of standard gyrokinetics @xcite applicable to plasmas that verify . in the slab geometry that we adopt , the background magnetic field ( the guide - field ) is assumed to be straight and uniform , @xmath8 .
the perturbed electron distribution function , to lowest order in @xmath9 , and in the gyrokinetic expansion , is defined as [ delta_f_e ] f_e= g_e + ( n_e / n_0e+2v_u_e / v_the^2)f_0e , where @xmath10 is the equilibrium maxwellian , @xmath11 is the electron thermal speed , @xmath12 is the velocity coordinate parallel to the guide - field direction , @xmath13 is the electron density perturbation ( the zeroth moment of @xmath14 ) , and [ uepar ] u_e =- j_/n_0e e=(e / c m_e)d_e^2_^2is the parallel electron flow ( the first moment of @xmath14 ) . in this expression , @xmath15 is the parallel current and @xmath16 is the parallel component of the vector potential ( note that , in this model , the parallel ion flow is zero to the order that is kept in the expansion ) ; and @xmath17 is the electron skin depth , where @xmath18 is the electron plasma frequency .
all moments of @xmath14 higher than @xmath13 and @xmath19 are contained in the `` reduced '' electron distribution function @xmath20 , e.g. , parallel temperature fluctuations are given by = d^3v g_e . for notational simplicity ,
let us introduce the following usual definitions : [ eq : ddt ] & = & + @xmath21 , + [ eq : bdotgrad ] & = & - @xmath22 , where @xmath23 is the electrostatic potential and @xmath24 $ ] denotes the poisson bracket .
the krehm equations are @xcite : @xmath25+\(1-\frac{2{v_{\parallel}}^2}{{v_{{\rm th}e}}^2}\)f_{0e}\hat{\bm b}\cdot\nabla\frac{e}{cm_e}d_e^2\lapperp\apar . \end{aligned}\ ] ] here , @xmath26 is the ohmic diffusivity and @xmath27 $ ] is the collision operator .
the perturbed electron density and the electrostatic potential are related via the gyrokinetic poisson law @xcite : [ eq : gk_poisson ] = @xmath28 , where @xmath29 and @xmath30 denotes the inverse fourier transform of @xmath31 , with @xmath32 the modified bessel function and @xmath33 ( @xmath34 is the ion larmor radius , with @xmath35 the ion thermal velocity and @xmath36 the ion gyrofrequency ) . is a kinetic equation for the reduced electron distribution function @xmath37 .
an important observation is that it does not contain an explicit dependence on @xmath38 .
if such a dependence is not introduced by the collision operator @xmath27 $ ] , then @xmath38 can be integrated out , and the reduced electron distribution function becomes effectively 4d only , @xmath39 .
the use of a hermite polynomial expansion of the distribution function is a well - known technique to simplify the numerical solution of kinetic equations such as ( [ eq : ge ] ) @xcite see @xcite in particular for an insightful discussion of this approach .
a very convenient aspect of the hermite formulation is that it enables a spectral representation of velocity space , a highly - desirable property when the available resolution is limited ( as is almost invariably the case ) .
it is worth pointing out , furthermore , that the advantages of the hermite representation transcend the numerical aspects , as it often enables one to make analytical headway in problems that are otherwise too complex : see , e.g. , refs .
@xcite .
perhaps for both these reasons , hermite formulations have gathered considerable interest recently ; we refer the reader to ref .
@xcite for a very comprehensive overview of recent and past work on the subject . the hermite expansion of @xmath20 is defined by [ hermite_expansion ] g_e(x , y , z , v_,t)=_m=2^h_m@xmath40 g_m ( x , y , z , t)f_0e(v _ ) , where @xmath41 denotes the hermite polynomial of order @xmath42 and @xmath43 is its coefficient .
note that @xmath44 because @xmath13 and @xmath19 have been explicitly separated in the decomposition of @xmath14 given in . introducing this expansion into , and choosing a modified lenard - bernstein collision operator @xcite , yields a set of coupled , fluid - like equations for the coefficients of the hermite polynomials : @xmath45 where @xmath46 is a kronecker delta and @xmath47 is the electron - ion collision frequency . in addition
, this choice for @xmath27 $ ] defines the resistive diffusivity : _
eid_e^2 . in the hermite formulation ,
@xmath42 is the velocity - space equivalent of @xmath48 in the usual fourier representation of position space .
thus , for example , the formation of fine scale structures in velocity space ( as arises from phase - mixing ) can be conveniently thought of as a transfer of energy to high @xmath42 s , much in the same way as the formation of fine scales in real space leads to energy being transferred to high wave numbers @xmath48 in the usual fourier representation . on the other hand
, the hermite representation introduces a closure problem , in that the equation for @xmath43 couples to the higher order moment @xmath49 .
we shall see in , however , that a rigorous , nonlinear closure can be obtained .
the well known reduced mhd ( rmhd ) equations @xcite can be obtained from by taking the collisional limit @xmath50 , @xmath51 , where @xmath52 and @xmath6 represent the typical frequencies and perpendicular wave numbers of the fluctuations , and @xmath53 is the ion sound larmor radius . in this limit , the isothermal approximation , @xmath54 , applies , and thus decouples from . for @xmath55 , becomes [ eq : rmhd - poisson ] = , where we have defined @xmath56 to make contact with the standard terminology .
further defining @xmath57 , we obtain : + @xmath58 & = & v_a+ @xmath59 , [ eq : vorticity ] + + @xmath60&= & + v_a , [ eq : epar ] where @xmath61 is the alfvn speed based on the guide - field , @xmath62 .
a different set of equations solved by is the kinetic reduced magnetohydrodynamics ( krmhd ) model , derived by expanding the gyrokinetic equation in terms of the small parameter @xmath63 @xcite in this sense , it is the long wavelength limit of gyrokinetics . in this limit , the alfvnic component of the turbulent fluctuations decouples from the compressive component .
the dynamics of the system are completely determined by the alfvnic fluctuations , which are governed by the reduced mhd .
the compressive fluctuations , on the other hand , evolve according to a kinetic equation : + v_g & = & d v_g , [ eq : slowmodekin ] where @xmath64 is related to the perturbed ion distribution function [ see equation ( 183 ) of schekochihin @xcite ] and @xmath65 is a one dimensional maxwellian . the parameter @xmath66 is a linear combination of the physical parameters ion - to - electron temperature ratio , plasma beta , and the ion charge [ see equation ( 182 ) of schekochihin @xcite ] . the structure of is mathematically similar to that of , the main difference being that this kinetic equation is decoupled from the alfvnic fluctuations , unlike its krehm counterpart .
similar to , one obtains the following set of equations by expanding in terms of hermite polynomials : @xmath67 , \quad m\ge2.\end{aligned}\ ] ] notice that , unlike , begin at @xmath68 .
additionally , since the term on the right hand side of is proportional to the first hermite polynomial , the parameter @xmath66 makes an appearance only in the equation for @xmath69 .
the hermite expansion transforms the original electron drift - kinetic equation , ( [ eq : ge ] ) , into an infinite , coupled set of fluid - like equations , ( [ eq : gms ] ) [ or , similarly for krhmd , into ] .
formally , the two representations are exactly equivalent , i.e. , no information is lost by introducing the hermite representation .
however , the numerical implementation of equations ( [ eq : gms ] ) obviously requires some form of truncation , i.e. , given a certain number of hermite moments , @xmath70 , it is necessary to specify some prescription for @xmath71 . in other words , as in the derivation of any fluid set of equations , the hermite expansion introduces a closure problem .
attempts to solve this problem have varied , from simply setting @xmath72 ( e.g. , @xcite ) , to polynomial closures in which @xmath71 is extrapolated from a number of previous moments @xcite . particularly noteworthy is the approach followed by hammett and co - workers @xcite where closures have been carefully designed to rigorously capture the linear landau damping rates ( as well as gyro - radius effects and dominant nonlinearities ) . in the system of equations under consideration here
, it turns out that an asymptotically exact closure can be obtained in the large @xmath70 limit .
let us consider that the collision frequency is small but finite .
then , there will be a range of @xmath42 s for which the collisional term is negligible one may think of this as the @xmath42 inertial range : energy is injected into low @xmath42 s via the coupling with ohm s law , and cascades ( phase - mixes ) to higher @xmath42 s .
however , as @xmath42 increases , a dissipation range is encountered , when the collisional term in [ or in ] is no longer subdominant with respect to the other terms . roughly speaking , in the dissipation range
, energy arrives at @xmath43 from @xmath73 and is mostly dissipated there ; only an exponentially smaller fraction is passed on to @xmath49 .
one thus expects that @xmath74 in the dissipation range , by definition .
the implication of this is that , for @xmath75 in the dissipation range , the dominant balance in the equation for @xmath71 must be [ eq : closure ] v_theg_m-_ei(m+1 ) g_m+1 . solving this equation for @xmath71 yields the sought closure @xcite .
the equation for @xmath76 therefore becomes : [ eq : last_g ] -_e @xmath77 + v_theg_m-1 = -_eim g_m , where @xmath78 is the parallel ( spitzer ) thermal diffusivity ( i.e. , the semi - collisional limit ) , then this equation needs to include the term proportional to the electron current [ the first term on the rhs of ] , becoming equation ( 99 ) of ref . @xcite . ] .
it is easy to see how the exact same reasoning leads to the equivalent closure for .
it can be useful to have an _ a priori _ estimate of the value of @xmath70 required to formally justify the asymptotic closure , for a given collision frequency .
one such _ linear _
estimate is provided in ref .
@xcite : if the hermite spectrum is in steady - state , then the collisional cutoff , @xmath79 , can be shown to occur at @xcite . if @xmath80 then the collisional cutoff is superseded .
this does not affect any of the considerations drawn here . ] : [ m_cutoff ] m_c=@xmath81 ^ 2/3 .
thus , we expect the hermite closure , , to be valid if @xmath82 .
the numerical implementation of introduces some difficulties and will be discussed in .
since our primary interest lies in weakly collisional plasmas , one finds that @xmath83 .
for example , a simple estimate using standard parameters for the solar corona suggests @xmath84 ; certain experiments on jet @xcite suggest @xmath85 in the edge region , considerably smaller than for the solar corona , but still quite large .
further noticing that such cases are invariably tied to a broad range of spatial scales , thereby also requiring high spatial resolutions , renders obvious the impracticability of such computations : not only must one solve a very large set of nonlinear , coupled pde s , as also the stiffness increases , due to the coefficients proportional to @xmath86 .
one possibility of avoiding this problem is to artificially enhance the value of the collision frequency .
note however that @xmath87 , i.e. , a relatively weak scaling , implying that cutting the number of necessary @xmath42 s down to computationally manageable sizes would require drastic increases in the collision frequency .
to make matters worse , the collision operator scales only linearly with @xmath42 , implying that in fact one needs to retain @xmath88 to adequately capture the dissipation range and validate the closure . one way to circumvent these difficulties
is to make use of a ` hyper - collision ' operator , i.e. , add a term of the form @xmath89 to the rhs of . here
, @xmath90 is the order of the hyper - diffusion operator ( a typical value would be @xmath91 ) and @xmath92 is a numerically - based coefficient defined such that energy arriving at @xmath75 can be dissipated in one timestep : [ hyper - coeff ] ~m^h_h g_m .
thus , in practice , one may simply set @xcite : _
h=1/(m^h ) .
it is worth remarking that if it is possible to choose a value of @xmath70 that is very deep into the dissipation range , then presumably the issue of which closure to implement becomes less sensitive , and it may be justified to simply set @xmath72 .
indeed , we have performed simulations with both closures and observed no differences ( not reported in this paper ) .
finally , we point out that in alternative to a hyper - collision operator one may use a spectral filter ( in @xmath42-space ) , such as the one of hou and li @xcite , as proposed by parker and dellar @xcite ( see also for a discussion of this filter in fourier space ) .
in the absence of collisions , conserve a quadratic invariant usually referred to as free energy @xcite .
this quantity can be defined as @xmath93 , where @xcite w_fluid=_k @xmath94(1-_0 ) + is the `` fluid '' ( electromagnetic ) part of the free energy , and h_e = d^3v is the electron free energy ( i.e. , the free energy associated with the reduced electron distribution function @xmath20 ) . upon introducing the hermite expansion of @xmath20 , , and allowing for finite collisions ( modelled by the lenard - bernstein collision operator )
, one finds that @xmath95 evolves according to the following equation @xcite : [ eq : energy ] w_fluid + _
m=2^g_m^2 = -n_0et_0e_ei_m=3^m g_m^2 -j_^2 .
the above equation is exact .
however , as discussed in , the numerical implementation of the hermite expansion requires that only a finite number of hermite polynomials are kept , and some form of closure to the expansion is required .
if we adopt the closure described by , and truncate the expansion at @xmath75 , adopts the truncated form : [ eq : trunc_energy ] & & w_fluid + _
m=2^mg_m^2 = + & & -n_0et_0e_ei_m=3^m m g_m^2 -n_0et_0e_e @xmath77 ^ 2 -j_^2 , + where the second term on the rhs is due to the specific closure that we have used ( and would vanish if , for example , we instead use the simpler closure @xmath72 . )
the same arguments that were invoked to motivate the hermite closure in apply here to justify the asymptotic equivalence of the full form of the energy balance , and its truncated version , that is , as long as @xmath70 is as large as required for @xmath76 to lie in the collisional ( i.e. , @xmath42- ) dissipation range , one expects the terms neglected in going from to to be exponentially small .
the corresponding equation for krmhd is equation ( 4.7 ) of ref .
the closure that we propose in can be implemented in this set of equations in a similar way , and it is straightforward to obtain the krmhd counterpart of .
the normalizations that we adopt for the krehm set of equations ( [ eq : mom],[eq : ohm],[eq : gk_poisson],[eq : gms ] ) are : * _ length scales : _ ( x , y)= ( x , y)/;z = z/ , where @xmath96 are , respectively , the perpendicular and parallel ( to the guide - field ) reference length - scales . *
_ times : _
t = t/ , where @xmath97 is the parallel alfvn time . * _ fields : _ ( n_e , g_m ) & = & _ i@xmath98 , + & = & , + & = & . under these normalizations , equations ( [ eq : mom ] ) , ( [ eq : ohm ] ) and ( [ eq : gms ] ) become : [ momentum ] & & = @xmath99 - , + [ ohm ] & & @xmath100 = + _ s^2@xmath101 + & & - + ^2@xmath102 , + [ g2_eq ] & & = \ { @xmath103- } + \{@xmath104- } , + [ gm_eq ] & & = \{@xmath105 - } + \{@xmath106 - } + & & - m_eig_m , m>2 . where now = + @xmath107 .
the normalized form of the quasi - neutrality is [ poisson ] n_e=@xmath108 .
it can immediately be seen that neglecting the @xmath109 reduces the above set of equations to the simpler two - field gyrofluid model treated in @xcite . for the krmhd set of the normalisation of space and time
are as above , upon which the only modification is the conversion of the prefactor @xmath110 into @xmath111 , where @xmath112 is the ion plasma beta .
the normalisation of the hermite moments @xmath43 is arbitrary since those equations are linear in @xmath43 .
the rhs of is conveniently separated into operators acting either in the direction perpendicular ( @xmath113 ) or parallel ( @xmath114 ) to the guide field .
this suggests that an efficient way of integrating those equations is to use operator splitting techniques such as to individually handle each class ( perpendicular or parallel ) of operators .
allows for both godunov @xcite or strang splitting @xcite .
although godunov splitting is formally only 1st - order accurate , direct comparisons of both splitting schemes performed by us ( not reported here ) yield undistinguishable results .
thus , by default , employs godunov splitting ( as it is computationally cheaper ) ; all results reported in are obtained with this option .
we now detail the algorithms employed for the perpendicular and parallel steps .
the numerical discretisation of is the straightforward generalisation of that derived in @xcite . for presentational simplicity ,
let us denote the nonlinear terms ( i.e. , the poisson brackets ) in by generalised operators , such that we have which is used in tearing mode simulations to prevent the resistive diffusion of the background ( reconnecting ) magnetic field . ] : & = & ( n_e , ) , + ( 1+^2d_e^2 ) & = & a(n_e , , g_2 ) - ^2(-a_,eq ) , + & = & _ 2(n_e,,g_2,g_3 ) , + & = & _ m(n_e,,g_m-1,g_m , g_m+1 ) -m_eig_m .
then , the integration scheme is as follows .
first we take a predictor step : n_e^n+1 , * & = & n_e^n + ( n_e^n,^n ) , + ^n+1 , * & = & e^-d_^n + @xmath115a_,eq + + & & (n_e^n,^n , g_2^n ) , + g_2^n+1 , * & = & g_2^n + _ 2(n_e^n,^n , g_2^n , g_3^n ) , + g_m^n+1 , * & = & e^-m_eig_m^n + + & & @xmath116_m(n_e^n,^n , g_m-1^n , g_m^n , g_m+1^n ) , where @xmath117 .
this is followed by the corrector step , which can be iterated @xmath118 times until the desired level of convergence is achieved : ^n+1,p+1 & = & e^-d_^n + @xmath115a_,eq + (n_e^n,^n , g_2^n ) + + & & (n_e^n+1,p,^n+1,p , g_2^n+1,p ) , + n_e^n+1,p+1 & = & n_e^n + @xmath119 + @xmath120 , + g_2^n+1,p+1 & = & g_2^n + @xmath121 + + & & _ 2(n_e^n+1,p+1,^n+1,p+1,g_2^n+1,p , g_3^n+1,p ) , + g_m^n+1,p+1 & = & e^-m_eig_m^n + e^-m_ei_m@xmath122 + + & & _ m@xmath123 . for presentational simplicity , we have not included here the hyper - diffusion and hyper - collisions operators , but it is trivial to do so : they are handled in the same way as the resistivity or the collisions are in the above equations . to deal with
the possibility of aliasing instability @xcite , offers two options .
one is the standard @xmath124 s rule @xcite , where the fourier transformed fields are multiplied by a step function @xmath125 defined by : [ two_thirds_rule ] ( k/)= 1 & , + 0 & , where @xmath126 for a grid with @xmath127 points .
the second option is the high - order fourier filter proposed by hou & li @xcite : [ hou_li_filter ] ( k/)=@xmath128 . compared to , the hou - li filter retains 12 - 15% more active fourier modes in each direction . for other advantages of this filter , and justification of its specific functional form ,
the reader is referred to ref .
tests reported in refs .
@xcite unanimously confirm the numerical superiority of the hou - li filter over the @xmath124 s rule dealiasing , as will our results presented in .
has inbuilt two distinct methods for the integration of the equations in the direction along the guide - field , @xmath114 : a maccormack scheme @xcite , and a combination of a third - order total variation diminishing ( tvd ) runge kutta method for the time derivative @xcite with a seventh - order upwind discretization for the fluxes @xcite ( tvdrk3uw7 for short ) .
the maccormack scheme is fairly standard ( see , e.g. , @xcite for textbook presentations ) and there is no need to detail it here .
the tvdrk3uw7 is not as conventional and is described below .
the @xmath114-advection step consists in solving the following set of equations : [ z_system ] = a , where = ( n_e , , g_2 , ... , g_m)^t is the solution vector and @xmath129 is tridiagonal matrix of size @xmath130 whose only non - zero entries are the coefficients of the @xmath114-derivatives , as follows : & & a_k , k+1 = \{^2 , ^2 , - , , - , } , + & & k=1 , , m , + & & a_k , k-1 = \{@xmath131 , ^2 , - , , - , } , + & & k=2 , , m. to be able to use upwind schemes , we need to write in characteristics form , i.e. , we need to diagonalize @xmath129 . to do so
, we introduce the matrix @xmath132 such that becomes p^-1 = p^-1 a p p^-1 .
we define @xmath133 and solve for @xmath132 requiring that p^-1ap = d , where @xmath134 is a diagonal matrix .
the equation for @xmath135 is now in characteristics form : [ z_charact ] = d , namely , if @xmath136 , @xmath137 is a right propagating wave field , and vice - versa . finally , since the entries of @xmath129 are independent of @xmath114 , so are the entries of @xmath134 .
can thus be written in flux - conservative form : [ z_fluxform ] = , where @xmath138 . as is well known from standard linear algebra , the diagonal entries of the matrix @xmath134 are the eigenvalues of @xmath129 , whereas @xmath132 is the matrix whose column vectors are the eigenvectors of @xmath129 . in , both eigenvalues and eigenvectors of @xmath129 are easily obtained with the linear algebra package lapack @xcite . as an example , let us consider the simplest possible case : the reduced - mhd limit .
matrix @xmath129 becomes : a= 0 & ^2 + & 0 .
it is a trivial exercise to obtain the matrices @xmath132 , @xmath139 and @xmath134 .
they are : p= -^2 & ^2 + 1 & 1 , p^-1= - & + & , d= -1 & 0 + 0 & 1 . in this case , the characteristic fields are = p^-1*u*=(- , + ) ^t . to relate this to a more familiar case , note that , using in the @xmath140 limit to express the electron density perturbation in terms of the electrostatic potential , @xmath141 , we immediately recognize the commonly used elsasser potentials : w^ = ( ) .
note that the entries of @xmath129 are constants , independent of either time or space .
thus , the matrices @xmath132 , @xmath139 and @xmath134 need only to be calculated once per run , with negligible impact on the overall code performance .
the derivative of the flux @xmath142 is computed using a seventh - order upwind scheme @xcite : @xmath143_i= , where , for the @xmath144th component of @xmath142 , we have [ flux1 ] f_i+1/2^j =- f_i+4^j+f_i+3^j - f_i+2^j + f_i+1^j + + f_i^j - f_i-1^j+f_i-2^j , d_j>0 , [ flux2 ] f_i+1/2^j =- f_i-3^j+f_i-2^j - f_i-1^j + f_i^j + + f_i+1^j - f_i+2^j+f_i+3^j , d_j<0 . for the time integration of we follow @xcite .
the time derivative is discretized using an optimal third - order total variation diminishing ( tvd ) runge kutta method @xcite : [ tvdrk ] & & * w*^(1 ) = * w*^(n ) + , + & & * w*^(2 ) = ^(n ) + ^(1 ) + , + & & * w*^(n+1 ) = ^(n ) + ^(2 ) + .
the final step is to compute @xmath145 .
compared to the maccormack method , the tvdrk3uw7 scheme just described has the disadvantage of being somewhat slower , as it requires three evaluations of the right hand side ( as opposed to only two for maccormack ) and there are more communications involved between different processors to compute the fluxes , .
this is partially offset by the fact that the tvdrk3uw7 scheme requires much fewer grid points per wavelength than the maccormack method for an adequate resolution , as will be exemplified in .
expanding the @xmath146 operator in the closure term in , we find that it becomes : [ eq : gm ] & & = -v_theg_m-1 + & & + _ e \{-@xmath147-@xmath148 + @xmath149 } + & & -_eim g_m .
as we have discussed in previous sections , the numerical algorithm employed in uses operator splitting methods to deal separately with the @xmath114-derivatives and with the poisson brackets ( i.e. , it splits the dynamics parallel and perpendicular to the magnetic guide - field ) .
this raises a difficulty when discretising the equation above , which contains mixed terms ( the second and third terms inside the curly brackets ) introduced by the closure , ; this is an especially subtle issue when the @xmath114-step scheme advects the equations in characteristics form , as is the case of the tvdrk3uw7 that we employ ( and would equally be the case for any other upwind scheme ) .
simple solutions to this problem require abandoning the operator splitting scheme and forsaking the use of the characteristics form for the @xmath114-derivative terms of the equations , both of which are not only highly convenient from the point of view of numerical accuracy and stability , but also physically motivated .
one possibility would be to treat this equation differently from all other equations solved by the code .
although this is certainly possible , at this stage we have chosen not to introduce this additional complexity .
as such , the actual form of implemented in is [ eq : gm_closed ] & = & -v_theg_m-1 + _ e \ { + @xmath149 } + & & -_eim g_m .
we emphasize that the dropping of the mixed terms is purely for algorithmic reasons . from the physical point of view
those terms are , _ a priori _ , as important as the closure terms which are kept ; their implementation is thus left to future work .
a serious drawback of this approach , for example , is that the semi - collisional limit of the krehm equations ( which results from setting @xmath150 , see section v.c of ref .
@xcite ) is , therefore , not correctly captured . on the other hand ,
note that : ( i ) for 2d problems , our implementation of the closure is exact ; ( ii ) for simple linear 3d problems [ where the background magnetic field is simply given by that guide - field ( which is the setup used to investigate alfvn wave propagation in ) , the numerical implementation of the closure is also exact ; ( iii ) in weakly collisional plasmas ( which are our main focus ) , provided that @xmath70 is sufficiently large to lie in the collisional dissipation range , one expects @xmath151 and thus the actual functional form of the closure may not be very important ; ( iv ) if we first apply the operator splitting scheme ( i.e. , the separation of the perpendicular and parallel operators ) and then impose our closure scheme on the parallel and perpendicular equations separately , we would obtain instead of .
finally , we remark that adopting as the evolution equation for @xmath76 changes the second term on the rhs of the energy balance equation , ( [ eq : trunc_energy ] ) , in the obvious way .
in this section , we report an extensive suite of linear and nonlinear benchmarks of . to illustrate the relative merits of the two numerical schemes for the @xmath114-advection available in , we carry out a simple test in the limit of isothermal electrons and cold ions . and reduce to [ iso_cold_ne ] & = & ^2 , + [ iso_cold_apar ] & = & .
the initial condition we adopt is : ( z , t=0 ) = .
are solved on a periodic box @xmath152 , with @xmath153 .
the grid step size is @xmath154 .
the time step is set by the cfl condition @xmath155 , where @xmath156 .
we chose @xmath157 , @xmath158 and @xmath159 .
there is no explicit dissipation in this test . a measure of how well resolved the wave front is is given by the parameter @xmath160 , where @xmath161 is the number of grid points per wavelength .
we test the behaviour of the maccormack and tvdrk3uw7 schemes for three representative values of @xmath162 ( note that the highest resolvable wave number corresponds to @xmath163 , i.e. , @xmath164 ) . for each of these cases ,
the equations are integrated for 10 transit times across the box , @xmath165 .
time traces of the energy conservation for both schemes are plotted in . as expected
, the tvdrk3uw7 scheme behaves remarkably better than maccormack .
notice , for example , that for the extreme case of @xmath166 , tvdrk3uw7 yields an amount of energy loss after 10 crossing times of @xmath167 , very similar to what is obtained with the maccormack scheme for the ten times better resolved case of @xmath168 .
-axis is the time normalized by the transit time across the simulation box , @xmath169 .
the @xmath170-axis is the variation in energy ( @xmath171 ) normalized by the initial energy , @xmath172 .
the parameter @xmath160 , where @xmath161 is the number of grid points per wavelength .
, width=340 ] ( top ) and @xmath173 ( bottom ) using the maccormack scheme ( left panels ) and the tvdrk3uw7 scheme ( right panels ) , for the case @xmath174 .
the maccormack scheme is seen to introduce strong gibbs oscillations , which are remarkably minimized by the tvdrk3uw7 scheme .
, width=453 ] besides much better energy conservation properties , we find the tvdrk3uw7 scheme to be very robust against spurious gibbs oscillations , even though it is not a shock - capturing scheme .
this is clearly visible in , where we plot the time history of the profiles of @xmath16 and @xmath173 obtained with both schemes for @xmath174 .
as can be seen , the tvdrk3uw7 scheme advects the initial condition with no visible deterioration , unlike the maccormack scheme .
the linearisation of in the collisionless limit yields the kinetic alfvn wave dispersion relation @xcite : @xmath175 \[1+\zeta z(\zeta)\]= \frac{1}{2}\kperp^2d_e^2,\ ] ] where @xmath176 , @xmath177 is the plasma dispersion function and @xmath178 . on the left plot of we show a comparison between the analytical values of the frequencies and damping rates , obtained by solving , and those computed by setting the number of hermite moments to @xmath179 and the number of grid points in the @xmath114-direction to @xmath180 .
very good agreement is observed over several orders of magnitude of the electron skin depth , @xmath181 ; the maximum relative error , obtained for the highest value of @xmath181 , is only a few percent .
the right plot shows the values of the frequency and damping rate for @xmath182 as a function of the number of hermite moments . for @xmath183
the damping rate converges to the analytical value ( @xmath184 ) , whereas for @xmath52 very little dependence on @xmath70 is observed .
, at fixed @xmath185 , @xmath186 , @xmath179 , as a function of the electron skin - depth @xmath17 .
lines are the exact solution of the analytical dispersion relation , , whereas data points are obtained from .
right : kaw frequency and damping rate obtained from for fixed @xmath182 , as a function of the total number of hermite moments kept , @xmath70 .
, title="fig:",scaledwidth=49.0% ] , at fixed @xmath185 , @xmath186 , @xmath179 , as a function of the electron skin - depth @xmath17 .
lines are the exact solution of the analytical dispersion relation , , whereas data points are obtained from .
right : kaw frequency and damping rate obtained from for fixed @xmath182 , as a function of the total number of hermite moments kept , @xmath70 .
, title="fig:",scaledwidth=49.0% ] the tearing mode @xcite is a fundamental plasma instability driven by a background current gradient .
tearing leads to the opening , growth and saturation of ( one or more ) magnetic island(s ) via the reconnection of a background magnetic field .
it is of intrinsic interest to magnetic confinement fusion devices , where it occurs either in standard or modified form ( i.e. , neoclassical tearing , microtearing ) .
it also represents the most basic paradigm for studies of magnetic reconnection . in this section
, we present the results of a linear benchmark of against the gyrokinetic code astrogk @xcite for the tearing mode problem .
we consider an in - plane magnetic equilibrium configuration given by @xmath187 , with @xmath188 , with @xmath189 the normalizing equilibrium scale length .
the simulations are performed in a doubly periodic box of dimensions @xmath190 , with @xmath191 and @xmath192 , such that @xmath193 yields the tearing instability parameter @xmath194 .
other parameters are @xmath195 , @xmath186 , @xmath196 .
all simulations keep @xmath197
. shows a plot of the linear growth rate of the tearing mode as a function of the lundquist number @xmath198 .
the @xmath199 case is obtained by setting @xmath200 , in which case the tearing mode is collisionless , i.e. , the frozen - flux condition is broken by electron inertia instead .
calculations with astrogk are done at three different values of @xmath3 and mass ratio : @xmath201 ( crosses , squares and circles , respectively ; this is the same data as plotted in fig . 2 of ref .
as seen , the agreement between the two codes improves for smaller @xmath3 , and is rather good for the smallest value of @xmath202 .
though it is expected that gyrokinetics will converge to krehm as @xmath3 is decreased , we note that , at least in this particular case , agreement is achieved for @xmath3 substantially larger than @xmath4 ( a factor of @xmath203 ) , suggesting that krehm may remain a reasonable approximation to the plasma dynamics outside its strict asymptotic limit of validity set by the requirement @xmath204 . and .
as expected , good agreement is obtained in the small @xmath3 limit.,width=302 ] a nonlinear benchmark is provided by the comparison of the tearing mode saturation amplitude with the prediction of mhd theory @xcite .
this was reported in ref .
@xcite , where it is shown that accurately reproduces the theoretical prediction in the parameter region where such prediction is valid [ i.e. , for @xmath205 and as long as islands are larger than the kinetic scales of the problem ( @xmath206 ) ] .
finally , see also figs . 1 and 3 of ref
. @xcite for more direct comparisons between and astrogk in the linear and nonlinear regime of a collisionless tearing mode simulation .
the orszag - tang ( ot ) vortex problem @xcite is a standard nonlinear test for fluid codes , and a basic paradigm in investigations of decaying mhd turbulence @xcite .
here we present results from a series of 2d and 3d runs , including a kinetic case . for easy reference ,
we summarise the main parameters of each simulation performed in .
+ .main parameters for decaying turbulence runs [ with the orszag - tang - type initial conditions of for the 2d runs , and of for the 3d runs ] . in all cases , @xmath207 and @xmath208 .
run e also includes @xmath209 hermite moments . [ cols="^,^,^,^,^,^",options="header " , ] + to avoid an overly symmetric initial configuration , we adopt the modification of ot initial conditions proposed in ref .
@xcite , namely : @xmath210 the runs are performed on a box of dimensions @xmath211 , at a resolution of @xmath212 collocation points . in the cases where no hyper - dissipation is used ( runs a and b ) , the resistivity is set to @xmath213 , and the magnetic prandtl number @xmath214 .
the kinetic scales @xmath215 are set to zero , so this is strictly a rmhd run .
magnetic ( @xmath216 ) and kinetic ( @xmath217 ) energy time traces for runs a and b are shown on the left - hand panel of .
we compare the results obtained using the hou - li high order fourier filter , , with those obtained with the standard @xmath124 s dealiasing rule of orszag @xcite , .
the agreement between the two sets of results is perfect , demonstrating that the hou - li filter does as good a job at conserving energy as the @xmath124 s rule .
the right - hand panel shows the time trace of the energy dissipation , normalized by the instantaneous total energy , i.e. , .
since no energy is being injected into the system , the rmhd equations should obey the conservation relation [ rmhd_inv ] = -d . in order to demonstrate the accuracy of the code
, we overplot a time trace of @xmath218 .
the very good agreement between the two curves is manifest ; in this particular run , is satisfied to better than @xmath219 . ) and kinetic ( @xmath220 ) energies , obtained from runs with different dealiasing methods : `` hou - li '' uses the high - order fourier filter of ref .
@xcite , given by ; `` @xmath124 dealia . '' uses the usual @xmath124 s rule of ref .
right panel : time trace of the energy dissipation rate ( for the hou - li run ) , normalized by the instantaneous total energy , @xmath221 .
overplotted is @xmath218 : code conserves energy to better than @xmath219 in this run .
the vertical dotted lines identify the times at which the contours of and spectra of are plotted . ]
contour plots of current and vorticity ( i.e. , @xmath222 ) at the times identified by the vertical lines in are plotted in ( top and bottom rows , respectively ) .
the formation of sharp current and vorticy sheets is observed , as expected . at @xmath223 one
can observe a plasmoid @xcite erupting from the current sheet on the lower right - hand corner of the plot , in what is perhaps the small - scale version of the observations reported in ref .
the role of the tearing instability of current sheets in 2d decaying turbulence has been previously discussed in refs .
. shows the total energy spectra obtained from the simulation with the hou - li filter ( run b ) , taken at the times identified by the vertical lines in .
there is no evidence of pile - up ( bottleneck ) at the small scales ( we note that the only dissipation terms present in this simulation are the standard laplacian resistivity and viscosity , i.e. , there is no hyper - dissipation ) . due to the relatively large values of the dissipation coefficients used in this simulation ,
the inertial range is very limited and it is not possible to clearly fit a unique power law ; for reference , @xmath224 is indicated in , following the iroshnikov - kraichnan prediction @xcite , and its numerical confirmation reported in refs .
@xcite ( although steeper power - laws @xmath225 have also been reported in the literature @xcite ) .
slope is shown for reference.,scaledwidth=70.0% ] a much longer and cleaner inertial range is obtained by replacing the standard ( laplacian ) dissipation terms with hyper - dissipation ( runs a1 and b1 ) . in that case , the spectra shown in are obtained ; the inertial range now shows an excellent agreement with the power - law slope of @xmath226 .
note also the extended inertial range obtained when the hou - li filter is used ( b1 ) instead of the standard @xmath124 s dealiasing . obtained with the hou - li filter ( blue , full line ) and with the standard @xmath124 s dealising rule ( red , dashed line ) .
the hou - li method results in an extended inertial range for the same number of collocation points , as expected .
neither spectra shows signs of energy pile - up at the small scales .
the power - law @xmath227 is indicated for reference.,scaledwidth=70.0% ] for the 3d simulations the initial conditions differ from the 2d case only in that they are modulated in the @xmath114-direction , as follows : @xmath228\sin(2\pi z / l_z),\\ \label{modot3d - psi } \psi(x , y ) = \[\cos(4\pi x / l_x+2.3 ) + \cos(2\pi y / l_y + 4.1)\]\cos(2\pi z / l_z).\end{aligned}\ ] ] we perform three different runs with these initial conditions ( runs c , d and e ) .
the first ( run c ) is just a straightforward extension to 3d of run b1 , except now with a resolution of @xmath229 . the second ( run d )
is designed to look at sub - ion - larmor radius turbulence ( i.e. , kinetic alfvn wave turbulence ) ; thus we set @xmath230 , where @xmath231 , and @xmath186 .
the resolution in this case is @xmath232 ( we use a smaller resolution here because the timestep , which is set by the cfl condition , is now also smaller , due to the dispersive nature of the kinetic alfvn waves ) . finally , run
e also includes the velocity - space dependence , represented with @xmath209 hermite moments ( meaning that it differs from run d in that the electrons are no longer isothermal , i.e. , @xmath233 ) .
a @xmath234 slope is shown for reference.,scaledwidth=70.0% ] the total energy spectrum obtained for run c is shown in .
the inertial range shows very good agreement with the goldreich - sridhar @xmath235 power law @xcite and again is clean of bottleneck effects . shows the magnetic , kinetic and electric energy spectra for run d , where we are now focussing on sub - ion larmor radius scales .
the slopes indicated refer to several power laws that have been widely discussed in the literature .
in particular , we see that the separation between electric and magnetic energy scalings , occurring at around @xmath236 , agrees quite well with the solar wind observations reported by bale _
et al . _
@xcite and with the gyrokinetic simulations of howes @xcite .
however , instead of the @xmath237 power law for the magnetic energy suggested in those works ( discussed in more detail in ref .
@xcite ) , we see that our data seems to more closely fit a @xmath238 scaling , which is a better fit to the @xmath239 slope often reported in observations ( e.g. , @xcite ) and in agreement with the recent work of boldyrev and perez @xcite on strong kinetic alfvnic turbulence . at @xmath240 .
the blue ( full ) line represents the perpendicular electric energy spectrum ; the red ( dashed ) line is the perpendicular magnetic field energy and the green ( dash - dot ) line is the kinetic energy .
the slopes @xmath241 , @xmath242 and @xmath243 are indicated for reference ( see text for discussion ) .
the vertical line indicates the ion larmor radius scale.,scaledwidth=70.0% ] hermite moments ) .
spectra at @xmath240 for ot - decaying kinetic turbulence .
lines represent the same quantities as in .
see text for a discussion of the power laws indicated.,scaledwidth=70.0% ] again shows energy spectra , this time for run e , which differs from run d in that it also includes hermite moments ( i.e. , it is a fully kinetic run , whereas d assumes isothermal electrons , @xmath244 ) .
comparing the magnetic spectra in the two cases ( i.e , runs d and e , both drawn at the same time ) , we see that its values increase at the larger ( spatial ) scales when adding the hermite moments , by about an order of magnitude , and run e s spectrum seems to be somewhat steeper than @xmath238 .
such differences may be due to landau damping , which is present in run e , but absent in run d. the hermite spectrum ( i.e. , the electron free energy spectrum , @xmath245 ) for run e is shown in , at different times .
a @xmath246 slope is indicated for reference ; this is the inertial - range slope predicted by zocco & schekochihin @xcite for the linear phase - mixing of kinetic alfvn waves .
since the number of hermite moments ( @xmath209 ) used is quite small we get an equivalently limited inertial range , and thus the agreement with the @xmath246 slope can only be regarded as indicative ; however , this tentative agreement lends credence to the idea that landau damping may be playing a significant role in this simulation
. a detailed analysis of kinetic turbulence in the krehm framework and , in particular , of the relative importance of the different energy dissipation mechanisms available , will be the subject of a future publication .
finally , for completeness we show in contour plots of the electron parallel velocity , @xmath247 , and of the density perturbations , @xmath173 , taken at the same time as the spectra of ( @xmath240 ) .
hermite moments ) .
electron free - energy spectra @xmath248 at different times .
an indicative power law of @xmath249 for the inertial range is also shown @xcite.,scaledwidth=70.0% ] ( left ) , and density perturbations @xmath173 ( right ) , at @xmath240.,title="fig:",scaledwidth=49.0% ] ( left ) , and density perturbations @xmath173 ( right ) , at @xmath240.,title="fig:",scaledwidth=49.0% ] we turn now to a benchmark of s implementation of the krmhd equations .
linearly , slow modes in krmhd are subject to collisionless damping via the barnes damping mechanism @xcite .
an initial perturbation damps at a rate that depends on the parameter @xmath66 . if slow mode fluctuations are constantly driven with an external force ( this is achieved by adding a forcing term to ) , then the system can be thought of as a plasma - kinetic langevin equation .
the mean - squared amplitude of the electrostatic potential for such a system reaches a steady - state saturation level , which can be derived analytically @xcite . , where @xmath250 .
the solid line is the analytical prediction @xcite , the red crosses are numerical results calculated using gandalf , and the green circles are calculated using .,width=340 ] in , we compare the steady - state saturation levels computed using with the analytical predictions , and the numerical results from another code gandalf ( a fully spectral gpu code that solves the krmhd equations ) .
slow mode fluctuations were driven using white noise forcing which injected energy into the system with unit power .
the spatial resolution was set to @xmath251 ; @xmath209 hermite moments of the distribution function were retained , @xmath252 .
the system was evolved until it reached a steady state .
the saturation level was then calculated by averaging over the steady state fluctuations for a few alfvn times .
it can be seen that the saturation amplitudes obtained using are in near perfect agreement with those calculated by gandalf , as well as with the analytical prediction .
has been used on a variety of computing clusters , with different architectures .
it is quite easy to install and run , having dependencies only on standard , widely - used libraries such as lapack @xcite and fftw @xcite .
its parallelization relies on standard mpi routines .
as described in detail in , the direction parallel to the field can be integrated by two different numerical methods , both of them fairly scalable , in terms of parallel performance .
in contrast , the direction perpendicular to @xmath253 uses standard pseudospectral techniques , which are plagued with well - known limits on scalability , due to the inherent non - locality of fourier transforms . for this reason , if one wishes to increase the number of processors for a given computation , it is more effective to do so by increasing the ratio between the number of processes for the parallel direction and the number of processes in the perpendicular direction .
the results of such a test , made on the helios machine ( an intel xeon e5 cluster ) , can be seen on , where the maccormack method was used in the parallel direction .
the initial conditions are the 3d orszag - tang vortex given by , with @xmath254 hermite moments .
we look at strong scaling , keeping the problem size fixed and varying the number of mpi processes , mainly in the parallel direction .
this produces a supralinear scaling , which breaks down after @xmath255 cores for the @xmath256 case and at @xmath257 cores for the @xmath258 one .
similar results have been obtained on other clusters , such as stampede ( a mixed intel xeon e5 and intel xeon phi coprocessor cluster ) , hopper ( a cray xe6 ) and edison ( a cray xc30 .
currently ongoing optimization work includes parallelizing the computation of the hermite moments via openmp .
cores for the @xmath256 case and at @xmath257 cores for the @xmath258 one .
the vertical axis gives the wall - clock time ( in seconds ) spent per timestep .
, width=302 ]
this paper describes , a novel code developed to investigate strongly magnetised , weakly - collisional , fluid - kinetic plasma dynamics in ( 2d or 3d ) slab geometry . solves two different sets of equations : the kinetic reduced electron heating model ( krehm ) of zocco & schekochihin @xcite ( which simplifies to conventional reduced - mhd @xcite in the appropriate limit ) and the kinetic reduced mhd ( krmhd ) equations of schekochihin @xcite .
the main numerical methods and the overall algorithm are described .
a noteworthy feature of is its spectral representation of velocity - space , achieved via a hermite expansion of the distribution function , as proposed in @xcite for krehm and in @xcite for the krmhd equations .
this representation has the attractive property of converting the kinetic equation for the distribution function into a coupled set of fluid - like equations for each hermite polynomial coefficient the advantage being that such equations are numerically more convenient to solve than the kinetic equation where they stem from . on the other hand ,
the hermite expansion introduces a closure problem ( in the sense that the equation for the hermite coefficient of order @xmath42 couples to that of order @xmath259 ) . to address this problem
, we present a nonlinear , asymptotically rigorous closure whose validity requires only that collisions are finite , but otherwise as small as required .
naturally , the smaller the collision frequency the higher the number of hermite moments that need to be kept to guarantee the accuracy of the closure .
realistic values of the collision frequency in the systems that are of primary interest to us ( e.g. , modern fusion devices , space and astrophysical environments ) lead to impractically large number of moments .
the adoption of a hyper - collision operator ( the direct translation into hermite space of the usual hyper - diffusion operators used in ( fourier ) @xmath48-space ) allows us to deal with this problem .
together with a pseudo - spectral representation of the plane perpendicular to the background magnetic field , and the option of a spectral - like algorithm for the dynamics along the field , the hermite representation of velocity space implies that is ideally suited to the investigation of magnetised kinetic plasma turbulence and magnetic reconnection , with the unique capability of allowing for the direct monitoring of energy flows in phase - space @xcite .
a series of linear and nonlinear numerical tests of is presented , with emphasis on orszag - tang - type decaying turbulence , both in the fluid and kinetic limits , where it is shown that recovers the theoretically expected power - law spectra . in this context ,
an interesting , novel result that warrants further investigation and will be discussed in a separate publication is the @xmath260 velocity - space ( hermite ) spectrum that is obtained in the 3d kinetic ( sub - ion larmor radius scales ) orszag - tang run presented in ( see ) .
this particular form of the hermite spectrum is indicative of linear phase mixing @xcite and suggests that this ( and ensuing landau damping ) may be a key energy transfer mechanism in kinetic decaying turbulence .
the authors are greatly indebted to alex schekochihin for many discussions and ideas that have been fundamental to this work .
nfl thanks paul dellar for pointing out the high - order fourier smoothing method of ref .
@xcite , ravi samtaney for discussions on high - order integration schemes for advection - type partial differential equations , and ryusuke numata for providing the data obtained with astrogk that appears in of this paper .
this work was partly supported by fundao para a cincia e tecnologia via grants uid / fis/50010/2013 , ptdc / fis/118187/2010 and if/00530/2013 , and by the leverhulme trust network for magnetised plasma turbulence .
simulations were carried out at hpc - ff ( juelich ) , helios ( iferc ) , edison and hopper ( nersc ) , kraken ( ncsa ) and stampede ( tacc ) .
a recent paper by zocco @xcite extends the krehm model to include a background electron temperature gradient .
this extension is also implemented in ; results of ongoing investigations exploring different instabilities introduced by these terms ( namely , the electron temperature gradient mode , and the microtearing instability ) will be reported elsewhere . for completeness
, we write below the krehm equations with this extension in normalised form ( see for the details of the normalisation adopted in ) .
they are : & & = @xmath99 - , + & & @xmath100 = + _ s^2@xmath101 - _ te + & & - + ^2@xmath102 + & & = \ { @xmath103- } + \{@xmath104- } + & & - _ te , + & & = \{@xmath105 - } + \{@xmath106 - } + & & - m_eig_m + _ m,3_te , m>2 , where @xmath261 , with @xmath262 the electron larmor radius and @xmath263 the electron temperature gradient scale length .
r. bruno , v. carbone , http://adsabs.harvard.edu/abs/2005lrsp....2....4b[the solar wind as a turbulence laboratory ] , liv .
solar phys . 2 ( 2005 ) 4 .
http://adsabs.harvard.edu/abs/2005lrsp....2....4b b. g. elmegreen , j. scalo , http://www.annualreviews.org/doi/abs/10.1146/annurev.astro.41.011802.094859[interstellar turbulence i : observations and processes ] , annu .
astrophys . 42 ( 1 ) ( 2004 )
211273 . http://dx.doi.org/10.1146/annurev.astro.41.011802.094859 [ ] .
d. a. uzdensky , http://link.springer.com/article/10.1023/b%3aastr.0000045064.93078.87[magnetic interaction between stars and accretion disks ] , astrophys .
space sci .
292 ( 1 - 4 ) ( 2004 ) 573585
. http://dx.doi.org/10.1023/b:astr.0000045064.93078.87 [ ] .
http://link.springer.com/article/10.1023/b%3aastr.0000045064.93078.87 e. a. frieman , l. chen , http://adsabs.harvard.edu/abs/1982phfl...25..502f[nonlinear gyrokinetic equations for low - frequency electromagnetic waves in general plasma equilibria ] , phys .
fluids 25 ( 1982 ) 502508 . http://dx.doi.org/10.1063/1.863762 [ ] . http://adsabs.harvard.edu/abs/1982phfl...25..502f g. g. howes , s. c. cowley , w. dorland , g. w. hammett , e. quataert , a. a. schekochihin , http://adsabs.harvard.edu/abs/2006apj...651..590h[astrophysical gyrokinetics : basic equations and linear theory ] , astrophys . j. 651 ( 2006 ) 590614 .
http://dx.doi.org/10.1086/506172 [ ] .
x. garbet , y. idomura , l. villard , t. watanabe , http://stacks.iop.org/0029-5515/50/i=4/a=043002[gyrokinetic simulations of turbulent transport ] , nucl .
fusion 50 ( 4 ) ( 2010 ) 043002 .
http://dx.doi.org/10.1088/0029-5515/50/4/043002 [ ] .
http://stacks.iop.org/0029-5515/50/i=4/a=043002 j. a. krommes , http://www.annualreviews.org/doi/abs/10.1146/annurev-fluid-120710-101223[the gyrokinetic description of microturbulence in magnetized plasmas ] , annu .
fluid mech .
44 ( 1 ) ( 2012 ) 175201 .
http://dx.doi.org/10.1146/annurev-fluid-120710-101223 [ ] .
i. g. abel , g. g. plunk , e. wang , m. barnes , s. c. cowley , w. dorland , a. a. schekochihin , http://stacks.iop.org/0034-4885/76/i=11/a=116201[multiscale gyrokinetics for rotating tokamak plasmas : fluctuations , transport and energy flows ] , rep .
76 ( 11 ) ( 2013 ) 116201 . http://dx.doi.org/10.1088/0034-4885/76/11/116201 [ ] . http://stacks.iop.org/0034-4885/76/i=11/a=116201 g. g. howes , w. dorland , s. c. cowley , g. w. hammett , e. quataert , a. a. schekochihin , t. tatsuno , http://link.aps.org/doi/10.1103/physrevlett.100.065004[kinetic simulations of magnetized turbulence in astrophysical plasmas ] , phys
100 ( 6 ) ( 2008 ) 065004 . http://dx.doi.org/10.1103/physrevlett.100.065004 [ ] .
a. a. schekochihin , s. c. cowley , w. dorland , g. w. hammett , g. g. howes , e. quataert , t. tatsuno , http://iopscience.iop.org/0067-0049/182/1/310[astrophysical gyrokinetics : kinetic and fluid turbulent cascades in magnetized weakly collisional plasmas ] , astrophys .
182 ( 1 ) ( 2009 ) 310 .
http://dx.doi.org/10.1088/0067-0049/182/1/310 [ ] .
g. g. howes , j. m. tenbarge , w. dorland , e. quataert , a. a. schekochihin , r. numata , t. tatsuno , http://link.aps.org/doi/10.1103/physrevlett.107.035004[gyrokinetic simulations of solar wind turbulence from ion to electron scales ] , phys .
107 ( 3 ) ( 2011 ) 035004 .
http://link.aps.org/doi/10.1103/physrevlett.107.035004 b. n. rogers , s. kobayashi , p. ricci , w. dorland , j. drake , t. tatsuno , http://pop.aip.org/resource/1/phpaen/v14/i9/p092110_s1?isauthorized=no[gyrokinetic simulations of collisionless magnetic reconnection ] , phys .
plasmas 14 ( 9 ) ( 2007 ) 092110 . http://dx.doi.org/10.1063/1.2774003 [ ] . http://pop.aip.org/resource/1/phpaen/v14/i9/p092110_s1?isauthorized=no m. j. pueschel , f. jenko , d. told , j. bchner , http://pop.aip.org/resource/1/phpaen/v18/i11/p112102_s1?isauthorized=no[gyrokinetic simulations of magnetic reconnection ] , phys . plasmas 18 ( 11 ) ( 2011 ) 112102 .
r. numata , w. dorland , g. g. howes , n. f. loureiro , b. n. rogers , t. tatsuno , http://pop.aip.org/resource/1/phpaen/v18/i11/p112106_s1?isauthorized=no[gyrokinetic simulations of the tearing instability ] , phys .
plasmas 18 ( 11 ) ( 2011 ) 112106
. http://dx.doi.org/10.1063/1.3659035 [ ] .
j. m. tenbarge , w. daughton , h. karimabadi , g. g. howes , w. dorland , collisionless reconnection in the large guide field regime : gyrokinetic versus particle - in - cell simulations , phys .
plasmas 21 ( 2 ) ( 2014 ) 020708 .
http://arxiv.org/abs/1312.5166 [ ] , http://dx.doi.org/10.1063/1.4867068 [ ] .
p. a. muoz , d. told , p. kilian , j. bchner , f. jenko , http://adsabs.harvard.edu/abs/2015arxiv150401351m[gyrokinetic and kinetic particle - in - cell simulations of guide - field reconnection .
part i : macroscopic effects of the electron flows ] , arxiv e - printshttp://arxiv.org / abs/1504.01351 [ ] .
http://adsabs.harvard.edu/abs/2015arxiv150401351m a. zocco , a. a. schekochihin , http://adsabs.harvard.edu/abs/2011phpl...18j2309z[reduced fluid - kinetic equations for low - frequency dynamics , magnetic reconnection , and electron heating in low - beta plasmas ] , phys .
plasmas 18 ( 2011 ) 2309 . http://dx.doi.org/10.1063/1.3628639 [ ] . http://adsabs.harvard.edu/abs/2011phpl...18j2309z m. j. aschwanden , a. i. poland , d. m. rabin , http://www.annualreviews.org/doi/abs/10.1146/annurev.astro.39.1.175[the new solar corona ] , annu .
39 ( 1 ) ( 2001 ) 175210 .
http://dx.doi.org/10.1146/annurev.astro.39.1.175 [ ] .
d. a. uzdensky , http://iopscience.iop.org/0004-637x/671/2/2139[the fast collisionless reconnection condition and the self - organization of solar coronal heating ] , astrophys .
j. 671 ( 2 ) ( 2007 ) 2139 .
http://dx.doi.org/10.1086/522915 [ ] .
w. gekelman , h. pfister , z. lucky , j. bamber , d. leneman , j. maggs , http://rsi.aip.org/resource/1/rsinak/v62/i12/p2875_s1?isauthorized=no[design , construction , and properties of the large plasma research device the lapd at ucla ] , rev .
62 ( 12 ) ( 1991 ) 28752883 . http://dx.doi.org/10.1063/1.1142175 [ ] .
http://rsi.aip.org/resource/1/rsinak/v62/i12/p2875_s1?isauthorized=no g. saibene , n. oyama , j. lnnroth , y. andrew , e. d. l. luna , c. giroud , g. t. a. huysmans , y. kamada , m. a. h. kempenaars , a. loarte , d. m. donald , m. m. f. nave , a. meiggs , v. parail , r. sartori , s. sharapov , j. stober , t. suzuki , m. takechi , k. toi , h. urano , http://iopscience.iop.org/0029-5515/47/8/031[the h mode pedestal , elms and tf ripple effects in jt-60u / jet dimensionless identity experiments ] , nucl .
fusion 47 ( 8) ( 2007 ) 969 .
http://dx.doi.org/10.1088/0029-5515/47/8/031 [ ] .
http://iopscience.iop.org/0029-5515/47/8/031 n. f. loureiro , a. a. schekochihin , a. zocco , http://link.aps.org/doi/10.1103/physrevlett.111.025002[fast collisionless reconnection and electron heating in strongly magnetized plasmas ] , phys .
( 2013 ) 025002 .
http://dx.doi.org/10.1103/physrevlett.111.025002 [ ] .
b. b. kadomtsev , o. p. pogutse , http://adsabs.harvard.edu/abs/1974jetp...38..283k[nonlinear helical perturbations of a plasma in the tokamak ] , sov . j. exp .
38 ( 1974 ) 283290 .
http://adsabs.harvard.edu/abs/1974jetp...38..283k h. r. strauss , http://pof.aip.org/resource/1/pfldas/v19/i1/p134_s1[nonlinear , threedimensional magnetohydrodynamics of noncircular tokamaks ] , phys .
fluids 19 ( 1 ) ( 1976 ) 134140 .
http://dx.doi.org/10.1063/1.861310 [ ] .
n. loureiro , g. hammett , http://www.sciencedirect.com/science/article/pii/s002199910800034x[an iterative semi - implicit scheme with robust damping ] , j. comp .
227 ( 9 ) ( 2008 ) 45184542 .
http://www.sciencedirect.com/science/article/pii/s002199910800034x p.
b. snyder , g. w. hammett , w. dorland , http://pop.aip.org/resource/1/phpaen/v4/i11/p3974_s1[landau fluid models of collisionless magnetohydrodynamics ] , phys .
plasmas 4 ( 11 ) ( 1997 ) 39743985 . http://dx.doi.org/10.1063/1.872517 [ ] . http://pop.aip.org/resource/1/phpaen/v4/i11/p3974_s1 t. j. schep , f. pegoraro , b. n. kuvshinov , http://pop.aip.org/resource/1/phpaen/v1/i9/p2843_s1?isauthorized=no[generalized twofluid theory of nonlinear magnetic structures ] , phys .
plasmas 1 ( 9 ) ( 1994 ) 28432852 .
http://dx.doi.org/10.1063/1.870523 [ ] .
s. pirozzoli , http://www.sciencedirect.com/science/article/pii/s002199910297021x[conservative hybrid compact - weno schemes for shock - turbulence interaction ] , j. comp .
178 ( 1 ) ( 2002 ) 81117 . http://dx.doi.org/10.1006/jcph.2002.7021 [ ] .
http://www.sciencedirect.com/science/article/pii/s002199910297021x a. zocco , n. f. loureiro , d. dickinson , r. numata , c. m. roach , http://stacks.iop.org/0741-3335/57/i=6/a=065008[kinetic microtearing modes and reconnecting modes in strongly magnetised slab plasmas ] , plasma phys .
fusion 57 ( 6 ) ( 2015 ) 065008 .
http://stacks.iop.org/0741-3335/57/i=6/a=065008 j. a. krommes , http://adsabs.harvard.edu/abs/2002phr...360....1k[fundamental statistical descriptions of plasma turbulence in magnetic fields ] , phys .
360 ( 2002 ) 1352 .
http://adsabs.harvard.edu/abs/2002phr...360....1k f. c. grant , m. r. feix , http://adsabs.harvard.edu/abs/1967phfl...10..696g[fourier-ermite solutions of the lasov equations in the linearized limit ] , phys . fluids 10 ( 1967 ) 696702 . http://dx.doi.org/10.1063/1.1762177 [ ] .
http://adsabs.harvard.edu/abs/1967phfl...10..696g f. c. grant , m. r. feix , http://adsabs.harvard.edu/abs/1967phfl...10.1356g[transition between landau and van kampen treatments of the vlasov equation ] , phys .
fluids 10 ( 1967 ) 13561357 . http://dx.doi.org/10.1063/1.1762288 [ ] .
g. joyce , g. knorr , h. k. meier , http://www.sciencedirect.com/science/article/pii/0021999171900349[numerical integration methods of the lasov equation ] , j. comp .
phys . 8 ( 1 ) ( 1971 ) 5363 .
g. knorr , m. shoucri , http://www.sciencedirect.com/science/article/pii/0021999174900011[plasma simulation as eigenvalue problem ] , j. comp .
14 ( 1 ) ( 1974 ) 17 .
http://www.sciencedirect.com/science/article/pii/0021999174900011 g. w. hammett , m. a. beer , w. dorland , s. c. cowley , s. a. smith , http://iopscience.iop.org/0741-3335/35/8/006[developments in the gyrofluid approach to tokamak turbulence simulations ] , plasma phys .
fusion 35 ( 8) ( 1993 ) 973 .
http://dx.doi.org/10.1088/0741-3335/35/8/006 [ ] .
s. a. smith , http://w3.pppl.gov/~hammett/sasmith/thesis.html[dissipative closures for statistical moments , fluid moments , and subgrid scales in plasma turbulence ] , ph.d . ,
princeton university ( 1997 ) .
h. sugama , t .- h .
watanabe , w. horton , collisionless kinetic - fluid closure and its application to the three - mode ion temperature gradient driven system , phys .
plasmas 8 ( 2001 ) 26172628 . http://dx.doi.org/10.1063/1.1367319 [ ] .
a. kanekar , a. a. schekochihin , w. dorland , n. f. loureiro , fluctuation - dissipation relations for a plasma - kinetic langevin equation , j. plasma phys .
81 ( 2015 ) 3004 . http://arxiv.org/abs/1403.6257 [ ] , http://dx.doi.org/10.1017/s0022377814000622 [ ] .
a. zocco , http://journals.cambridge.org/article_s0022377815000331[linear collisionless landau damping in hilbert space ] , journal of plasma physics firstview ( 2015 ) 17 .
http://dx.doi.org/10.1017/s0022377815000331 [ ] .
j. t. parker , p. j. dellar , fourier - hermite spectral representation for the vlasov - poisson system in the weakly collisional limit , j. plasma phys .
81 ( 2015 ) 3003 . http://arxiv.org/abs/1407.1932 [ ] , http://dx.doi.org/10.1017/s0022377814001287 [ ] .
l. gibelli , b. d. shizgal , http://dx.doi.org/10.1016/j.jcp.2006.06.017[spectral convergence of the hermite basis function solution of the vlasov equation : the free - streaming term ] , j. comput .
219 ( 2 ) ( 2006 ) 477488 .
http://dx.doi.org/10.1016/j.jcp.2006.06.017 [ ] .
http://dx.doi.org/10.1016/j.jcp.2006.06.017 g. w. hammett , f. w. perkins , http://link.aps.org/doi/10.1103/physrevlett.64.3019[fluid moment models for andau damping with application to the ion - temperature - gradient instability ] , phys .
( 1990 ) 30193022 .
http://link.aps.org/doi/10.1103/physrevlett.64.3019 p.
b. snyder , g. w. hammett , http://scitation.aip.org/content/aip/journal/pop/8/7/10.1063/1.1374238[a andau fluid model for electromagnetic plasma microturbulence ] , phys .
plasmas 8 ( 7 ) ( 2001 ) 31993216 .
http://dx.doi.org/10.1063/1.1374238 [ ] .
v. borue , s. a. orszag , http://link.aps.org/doi/10.1103/physreve.55.7005[spectra in helical three - dimensional homogeneous isotropic turbulence ] , phys .
e 55 ( 6 ) ( 1997 ) 70057009 .
http://dx.doi.org/10.1103/physreve.55.7005 [ ] .
t. y. hou , r. li , http://www.sciencedirect.com/science/article/pii/s0021999107001623[computing nearly singular solutions using pseudo - spectral methods ] , j. comp .
phys . 226 ( 1 ) ( 2007 ) 379397 . http://dx.doi.org/10.1016/j.jcp.2007.04.014 [ ] . http://www.sciencedirect.com/science/article/pii/s0021999107001623 s. godunov , http://mi.mathnet.ru/eng/msb4873[a difference method for numerical calculation of discontinuous solutions of the equations of hydrodynamics ] , mat .
( n. s. ) 47 ( 3 ) ( 1959 ) 271306 .
s. a. orszag , http://adsabs.harvard.edu/abs/1971jats...28.1074o[on the elimination of aliasing in finite - difference schemes by filtering high - wavenumber components . ] , j. atmosph .
28 ( 1971 ) 10741074 . http://dx.doi.org/10.1175/1520-0469(1971)028<1074:oteoai>2.0.co;2 [ ] . http://adsabs.harvard.edu/abs/1971jats...28.1074o t. grafke , h. homann , j. dreher , r. grauer , http://www.sciencedirect.com/science/article/pii/s0167278907004101[numerical simulations of possible finite time singularities in the incompressible euler equations : comparison of numerical methods ] , phys .
d : nonlin .
237 ( 1417 ) ( 2008 ) 19321936 . http://dx.doi.org/10.1016/j.physd.2007.11.006 [ ] .
t. y. hou , blow - up or no blow - up ?
a unified computational and analytic approach to 3d incompressible euler and navier-stokes equations , acta num .
18 ( 2009 ) 277346 . http://dx.doi.org/10.1017/s0962492906420018 [ ] .
j. dellar , http://www.sciencedirect.com/science/article/pii/s0021999112007012[lattice boltzmann magnetohydrodynamics with current - dependent resistivity ] , j. comp .
237 ( 2013 ) 115131 . http://dx.doi.org/10.1016/j.jcp.2012.11.021 [ ] . http://www.sciencedirect.com/science/article/pii/s0021999112007012 r. w. maccormack , http://www.worldscientific.com/doi/abs/10.1142/9789812810793_0002[the effect of viscosity in hypervelocity impact cratering ] , in : frontiers of computational fluid dynamics , world scientific , 2002 , ch . 2 ,
http://arxiv.org/abs/http://www.worldscientific.com/doi/pdf/10.1142/9789812810793_0002 [ ] , http://dx.doi.org/10.1142/9789812810793_0002 [ ] . http://www.worldscientific.com/doi/abs/10.1142/9789812810793_0002 e. anderson , z. bai , c. bischof , s. blackford , j. demmel , j. dongarra , j. du croz , a. greenbaum , s. hammarling , a. mckenney , d. sorensen , usersguide , 3rd edition , society for industrial and applied mathematics , philadelphia , pa , 1999 .
r. samtaney , http://iopscience.iop.org/1749-4699/5/1/014004[numerical aspects of drift kinetic turbulence : ill - posedness , regularization and a priori estimates of sub - grid - scale terms ] , comput . sci . disc . 5 ( 1 ) ( 2012 ) 014004 .
http://dx.doi.org/10.1088/1749-4699/5/1/014004 [ ] .
h. p. furth , j. killeen , m. n. rosenbluth , http://pof.aip.org/resource/1/pfldas/v6/i4/p459_s1?isauthorized=no[finite-resistivity instabilities of a sheet pinch ] , phys .
fluids 6 ( 4 ) ( 1963 ) 459484 .
http://dx.doi.org/10.1063/1.1706761 [ ] .
r. numata , g. g. howes , t. tatsuno , m. barnes , w. dorland , http://www.sciencedirect.com/science/article/pii/s0021999110005000[astrogk : astrophysical gyrokinetics code ] , j. comp .
229 ( 24 ) ( 2010 ) 93479372 .
http://dx.doi.org/10.1016/j.jcp.2010.09.006 [ ] .
f. militello , f. porcelli , http://pop.aip.org/resource/1/phpaen/v11/i5/pl13_s1[simple analysis of the nonlinear saturation of the tearing mode ] , phys .
plasmas 11 ( 5 ) ( 2004 ) l13l16 .
http://dx.doi.org/10.1063/1.1677089 [ ] .
d. escande , m. ottaviani , http://www.sciencedirect.com/science/article/pii/s0375960104001781[simple and rigorous solution for the nonlinear tearing mode ] , phys .
lett . a 323 ( 34 ) ( 2004 ) 278284 . http://dx.doi.org/10.1016/j.physleta.2004.02.010 [ ] .
http://www.sciencedirect.com/science/article/pii/s0375960104001781 n. f. loureiro , s. c. cowley , w. d. dorland , m. g. haines , a. a. schekochihin , http://link.aps.org/doi/10.1103/physrevlett.95.235003[x-point collapse and saturation in the nonlinear tearing mode reconnection ] , phys .
95 ( 23 ) ( 2005 ) 235003 . http://dx.doi.org/10.1103/physrevlett.95.235003 [ ] . http://link.aps.org/doi/10.1103/physrevlett.95.235003 r. numata , n. f. loureiro , ion and electron heating during magnetic reconnection in weakly collisional plasmas , j. plasma phys .
81 ( 2015 ) 3001 . http://arxiv.org/abs/1406.6456 [ ] , http://dx.doi.org/10.1017/s002237781400107x [ ] .
d. biskamp , h. welter , http://pop.aip.org/resource/1/pfbpei/v1/i10/p1964_s1[dynamics of decaying twodimensional magnetohydrodynamic turbulence ] , phys .
fluids b : plasma phys . 1 ( 10 ) ( 1989 ) 19641979 . http://dx.doi.org/10.1063/1.859060 [ ] . http://pop.aip.org/resource/1/pfbpei/v1/i10/p1964_s1 h. politano , a. pouquet , p. l. sulem , http://scitation.aip.org/content/aip/journal/pop/2/8/10.1063/1.871473[current and vorticity dynamics in threedimensional magnetohydrodynamic turbulence ] , phys . plasmas 2 ( 8) ( 1995 ) 29312939 .
http://dx.doi.org/10.1063/1.871473 [ ] .
d. biskamp , e. schwarz , http://pop.aip.org/resource/1/phpaen/v8/i7/p3282_s1?isauthorized=no[on two - dimensional magnetohydrodynamic turbulence ] , phys .
plasmas 8 ( 7 ) ( 2001 ) 32823292 .
n. f. loureiro , a. a. schekochihin , s. c. cowley , instability of current sheets and formation of plasmoid chains , phys .
plasmas 14 ( 10 ) ( 2007 ) 100703 .
http://arxiv.org/abs/astro-ph/0703631 [ ] , http://dx.doi.org/10.1063/1.2783986 [ ] .
n. f. loureiro , a. a. schekochihin , d. a. uzdensky , plasmoid and kelvin - helmholtz instabilities in sweet - parker current sheets , phys . rev .
e 87 ( 1 ) ( 2013 ) 013102 . http://arxiv.org/abs/1208.0966 [ ] , http://dx.doi.org/10.1103/physreve.87.013102 [ ] .
n. f. loureiro , d. a. uzdensky , a. a. schekochihin , s. c. cowley , t. a. yousef , turbulent magnetic reconnection in two dimensions , mon . not .
r. astron .
( 2009 ) l146l150 .
http://arxiv.org/abs/0904.0823 [ ] , http://dx.doi.org/10.1111/j.1745-3933.2009.00742.x [ ] .
r. kinney , j. c. mcwilliams , t. tajima , http://scitation.aip.org/content/aip/journal/pop/2/10/10.1063/1.871062[coherent structures and turbulent cascades in twodimensional incompressible magnetohydrodynamic turbulence ] , phys .
plasmas 2 ( 10 ) ( 1995 ) 36233639 . http://dx.doi.org/10.1063/1.871062 [ ] . http://scitation.aip.org/content/aip/journal/pop/2/10/10.1063/1.871062 s. bale , p. kellogg , f. mozer , t. horbury , h. reme , http://link.aps.org/doi/10.1103/physrevlett.94.215002[measurement of the electric fluctuation spectrum of magnetohydrodynamic turbulence ] , phys .
94 ( 2005 ) 215002 . http://dx.doi.org/10.1103/physrevlett.94.215002 [ ] . http://link.aps.org/doi/10.1103/physrevlett.94.215002 o. alexandrova , j. saur , c. lacombe , a. mangeney , j. mitchell , s. j. schwartz , p. robert , http://link.aps.org/doi/10.1103/physrevlett.103.165003[universality of solar - wind turbulent spectrum from mhd to electron scales ] , phys . rev . lett . 103
( 2009 ) 165003 .
http://dx.doi.org/10.1103/physrevlett.103.165003 [ ] .
s. boldyrev , j. c. perez , http://stacks.iop.org/2041-8205/758/i=2/a=l44[spectrum of kinetic - lfvn turbulence ] , astrophys . j. lett .
758 ( 2 ) ( 2012 ) l44 .
http://dx.doi.org/10.1088/2041-8205/758/2/l4 [ ] .
j. m. tenbarge , g. g. howes , w. dorland , g. w. hammett , http://www.sciencedirect.com/science/article/pii/s0010465513003664[an oscillating langevin antenna for driving plasma turbulence simulations ] , comp .
185 ( 2 ) ( 2014 ) 578589 .
http://dx.doi.org/10.1016/j.cpc.2013.10.022 [ ] . | we report on the algorithms and numerical methods used in , a novel fluid - kinetic code that solves two distinct sets of equations : ( i ) the kinetic reduced electron heating model ( krehm ) equations [ zocco & schekochihin , phys .
plasmas * 18 * , 102309 ( 2011 ) ] ( which reduce to the standard reduced - mhd equations in the appropriate limit ) and ( ii ) the kinetic reduced mhd ( krmhd ) equations [ schekochihin , astrophys . j. suppl . *
182*:310 ( 2009 ) ] .
two main applications of these equations are magnetised ( alfvnic ) plasma turbulence and magnetic reconnection
. uses operator splitting ( strang or godunov ) to separate the dynamics parallel and perpendicular to the ambient magnetic field ( assumed strong ) . along the magnetic field , allows for either a second - order accurate maccormack method or , for higher accuracy , a spectral - like scheme composed of the combination of a total variation diminishing ( tvd ) third order runge - kutta method for the time derivative with a 7th order upwind scheme for the fluxes .
perpendicular to the field is pseudo - spectral , and the time integration is performed by means of an iterative predictor - corrector scheme . in addition , a distinctive feature of is its spectral representation of the parallel velocity - space dependence , achieved by means of a hermite representation of the perturbed distribution function .
a series of linear and nonlinear benchmarks and tests are presented , including a detailed analysis of 2d and 3d orszag - tang - type decaying turbulence , both in fluid and kinetic regimes . |
filamentous networks @xcite are an important class of materials in which the interplay between stretching and bending energies and temperature lead to unique mechanical properties such as strong nonlinear response and strain stiffening @xcite , negative normal stress @xcite , non - affine response @xcite , crossover from non - affine to affine response @xcite , and power - law frequency dependence of the storage and loss moduli @xcite .
they are a part of such important components of living matter @xcite as the cytoskeleton , the intercellular matrix , and clotted blood and of industrial materials like paper @xcite . here
we introduce the diluted kagome lattice , shown in fig .
[ fig : diluted_kagome ] , in which elastic rods on nearest - neighbor ( @xmath12 ) bonds are removed with probability @xmath0 , as a model for filamentous networks in two - dimensions . lines of contiguous and colinear occupied bonds are identified as filaments , which are modeled , as in previous work @xcite as elastic rods with one - dimensional stretching modulus @xmath1 and bending modulus @xmath2 .
effective medium theories ( emts ) @xcite have proven to be a powerful tool for the study of random systems .
we use recently derived emts @xcite that treat bending as well as stretching forces in elastic networks to calculate the shear modulus @xmath5 as a function of @xmath1 , @xmath2 , and @xmath3 , and derive scaling forms for its behavior near @xmath4 and near the @xmath13 rigidity threshold at @xmath10 .
we also calculate @xmath5 using numerical simulations .
the simulations and emts are in qualitative agreement over the entire range of @xmath3 and in quantitative agreement near @xmath4 .
our results are also in general agreement with those for the mikado model @xcite , including , in particular , a crossover from bending - dominated nonaffine response at low density ( small @xmath3 ) to stretching dominated affine response at high density ( @xmath3 near @xmath14 )
. our model , however , highlights special properties of networks of straight filaments , including scaling collapse , described analytically by our emts , of the region near @xmath4 where filament length @xmath15 approaches infinity , and the underlying origins of affine response in this limit . and @xmath16 horizontal bonds and @xmath17 bonds along the axis at an angle @xmath18 to the @xmath19-axis .
the red sites on the lattice are crosslinks , and colinear sequences of bonds are filaments .
only two filaments , shown with thicker bonds , traverse the lattice from top to bottom and support external stresses .
one these makes an angle @xmath20 and one an angle @xmath21 with the @xmath19-axis , and the length of both is @xmath22 . in more general lattices ,
sample crossing filaments can orient at any angle @xmath23 to the @xmath19-axis , but the projection of their length onto the vertical axis is @xmath24 , and thus their length is @xmath25 . ] the building blocks of filamentous networks are typically semi - flexible polymers , characterized by a bending modulus @xmath26 , where @xmath27 is the persistence length @xmath28 the boltzmann constant , and @xmath29 the temperature . where filaments intersect , physical or chemical crosslinks bind them together but do not inhibit their relative rotation .
since only two filaments cross at a point , each crosslink is connected to at most four others .
the average distance @xmath30 between crosslinks is less that @xmath27 . a remarkably faithful description of experimentally measured linear @xcite and nonlinear @xcite elastic response is provided by a model in which elastic response is assumed to be affine and in which filamentous sections between crosslinks are treated as nonlinear central - force springs with a force - extension curve determined by the worm - like chain model @xcite augmented by an enthalpic stretching energy .
this simple model , however , misses some important properties of filamentous networks , including the existence of a rigidity percolation threshold @xcite , nonaffine response @xcite , and crossover from bending - dominated nonaffine response at small @xmath31 @xcite , or low frequency @xcite , to almost affine , stretching dominated response at large @xmath31 or high frequency .
the mikado model @xcite provides additional insights into the physics of filamentous networks . in this model ,
straight filaments of length @xmath15 are deposited with random position and orientation on a two - dimensional flat surface and crosslinked at their points of intersection .
filaments are treated as elastic rods with one - dimensional stretching and bending moduli , @xmath1 and @xmath2 , rather than as worm - like chains .
numerical simulations on this model reveal a rigidity percolation transition from a floppy network to one with non - vanishing shear and bulk moduli and with non - affine bending dominated elastic response . as @xmath31
is increased , response becomes more affine and stretching dominated , reaching almost perfectly affine response in the @xmath32 limit .
our emts and numerical simulations yield similar results for the diluted kagome model .
though the mikado and the diluted kagome lattice are quite similar with crosslinks of maximum coordination four and variable values of the ratio @xmath31 , they model slightly different parts of the phase space of possible two - dimensional filamentous networks . in particular , since the starting point of the diluted kagome lattice is the full lattice with all bonds occupied , it necessarily deals directly with the @xmath33 limit , which is not generally accessed in studies of the mikado model in which @xmath15 is restricted to be less than the sample width @xmath34 .
the mikado model is an off - lattice " model , whereas the kagome model is lattice based .
the latter property of the kagome model facilitates the application of lattice - based emts @xcite ( modified to include bending ) for all values of @xmath3 .
finally , in the mikado model , the distance between between crosslinks on a single filament follows a poisson distribution with no lower bound , whereas in the kagome lattice , this distance is an integral multiple of the lattice spacing with a lower bound of one lattice spacing .
this difference leads , as we shall see , to a different scaling of the shear modulus in the bending dominated regime near @xmath4 . as detailed in apps .
[ app : asym ] and [ sec : appc ] , our emts clearly show that there are three critical points ( or fixed points in the renormalization - group sense ) : the trivial empty lattice point at @xmath35 ( which we ignore ) , the rigidity - percolation point at @xmath10 , and the @xmath4 full lattice point .
it also provides analytic scaling relations for the the shear modulus @xmath5 as a function of @xmath1 , @xmath2 , and @xmath3 in the @xmath36 limit near both @xmath10 and @xmath4 .
of particular interest is the behavior near @xmath4 , where the ratio @xmath37 of the shear modulus to its @xmath4 value @xmath38 can be expressed as a function of the single scaling variable @xmath39 where @xmath40 is the bending length @xmath41 is the average filament length and @xmath42 is the average spacing between crosslinks along a filament . as @xmath43 , @xmath5 approaches @xmath38 , but as @xmath44 , there is crossover to a bending dominated , nonaffine regime in which @xmath45 this behavior , which appears in our simulations , has also been observed in simulations in diluted three - dimensional lattices of four - fold coordinated filaments @xcite . in sec .
[ sec : discussion ] , we speculate about the reasons for the same behavior appearing in both two- and three - dimensions .
@xcite developed an off - lattice emt for filamentous networks that predicts a bending - dominated non - affine regime with a different power of @xmath31 , @xmath46 , than the one we predict .
the origin of this different scaling is the absence of a lower cutoff in the poisson distribution of the distance between crosslinks in the mikado model compared to the fixed cutoff equal to the lattice spacing in the kagome model .
the analysis of refs .
@xcite yields the kagome lattice scaling law if the probability distribution for distances is replaced by one with a fixed lower cutoff . and the stretching dominated region in which @xmath47 .
this boundary between stretching- and bending - dominated response was calculated using a formula that interpolates between the analytic emt expressions for @xmath5 near @xmath10 and near @xmath4 . ]
our numerical simulations follow the emt prediction very closely with _ no _ adjustable parameters near @xmath4 if @xmath2 is not too large . for really small values of @xmath48 ,
finite - size effects become important , and simulations can be fit with a combination of the exactly calculable finite size results at @xmath49 and the emt scaling form valid at infinite @xmath34 . near @xmath10 , simulations are consistent with a scaling function of the form of eq .
( [ eq : scaling-2 ] ) but with @xmath50 approximately @xmath51 rather than @xmath14 and with different values of @xmath52 and @xmath53 . under periodic boundary conditions , the kagome lattice has exactly @xmath54 neighbors per site , where @xmath55 is the spatial dimension .
if sites interact only via central - force springs on bonds and not by bending forces along filaments , the lattice is _ isostatic _ in that it has exactly enough internal forces to eliminate zero modes according the simple maxwell count @xcite : @xmath56 , where @xmath57 is the number of zero modes and @xmath58 is the number of bonds . under
free rather than periodic boundary conditions , sites on the boundaries have fewer than four neighbors , and there is an overall bond deficiency of @xmath59 . as a result , there are of order @xmath60 zero floppy " internal modes of zero energy in the absence of bending forces . under periodic boundary conditions
, the simple maxwell count leads to @xmath61 , but the total number of zero modes ( as calculated by the dynamical matrix , whose normal modes are the infinitesimal " modes that result when the elastic energy is truncated to harmonic order ) , like the number under free boundary conditions , is of order @xmath60 rather than zero @xcite .
this discrepancy is a result of the existence of lattice configurations that can support stress in bonds while maintaining a zero force at each note @xcite .
associated with each such configuration @xcite , called a _ state of self stress _ " , is an additional infinitesimal zero mode so that the maxwell count becomes @xmath62 , where @xmath63 is the number of states of self stress .
thus , in the undiluted central - force kagome lattice under periodic boundary conditions , @xmath64 is simply the number of states of self - stress .
these states endow the undiluted central - force kagome lattice with affine response and non - zero bulk and shear moduli , proportional to the bond spring constant , and , thus , determine the approach to affine , stretching - dominated response as @xmath11 in the diluted lattice with nonzero @xmath2 .
in addition , the central - force zero modes associated with the states of self stress also control the form of the effective - medium equations and the scaling forms they predict .
the addition of bending forces lifts all but the two trivial zero modes of rigid translation to finite frequency .
this fact plays a central roll in our emt and is ultimately responsible for the form of the scaling function for @xmath5 near @xmath4 .
we are ultimately interested in three - dimensional filamentous networks , which are subisostatic with @xmath65 because their maximum coordination number , like the that of two - dimensional networks , is four .
if constituent filaments are straight , these networks have a state of self - stress for each filament in the undiluted limit , and their elastic moduli are non - zero , proportional to the bond - spring constant , and independent of @xmath2 @xcite . associated with each state of self - stress , there is an additional zero mode , but as in the kagome lattice , the addition of bending forces raises all but the three trivial zero modes of uniform translation to finite frequency and stabilizes the lattice .
we will argue in sec .
[ sec : discussion ] that these properties are the likely origin of the very similar behavior , seen in simulations , of the shear modulus in @xmath66- and @xmath67-dimensional networks with maximum coordination of four .
this paper consists of five sections and three appendices .
section [ sec : filnetworks ] compares and contrasts the mikado and kagome models and demonstrates how straight , sample - traversing filaments with the energy of an elastic beam produce affine elastic response .
section [ sec : filnetworks ] also defines the lattice model that we use .
section [ sed : emt - or ] describes emt procedures and presents their results .
section [ sec : numericalsimulations ] presents the results of our numerical simulations and compares them with those of the emts .
section [ sec : discussion ] reviews our results and speculates about application of our two - dimensional calculations to three - dimensional systems .
there are three appendices : appendix [ sec : app1 ] discusses general properties of the emt dynamical matrix .
appendices [ app : asym ] and [ sec : appc ] present details of the solutions to the emt equations near @xmath4 and @xmath10 for the bending emts described in refs .
@xcite and in refs .
@xcite , respectively .
networks composed of straight filaments with elastic - rod energies have special properties . in this section
, we will explore some general properties of these networks before setting up the energy for the kagome lattice .
the elastic energy of a one - dimensional elastic beam with stretching modulus @xmath1 and bending modulus @xmath2 is @xmath68 where @xmath69 is the local longitudinal displacement and @xmath70 the local angle of the tangent to the filament at the point @xmath71 .
the usual practice is to treat the stretching modulus as independent of @xmath2 even though at nonzero temperature the spring constant for the entire filament has an important entropic component that depends on @xmath2 .
the spring constant for a filament section of length @xmath72 is @xmath73 .
this leads to affine response in linear elasticity for any lattice in any dimension consisting of sample - traversing straight filaments with a sufficient number of orientations to ensure stability with respect to all strain @xcite . to see this , consider a crosslink ( node of the lattice ) at the origin on a filament parallel to the unit vector @xmath74 , and let it be connected to two other crosslinks on the same filament at respective positions @xmath75 and @xmath76 relative to the orign . under a uniform , affine deformation , the relative positions @xmath77 , @xmath78 transform to @xmath79 , where @xmath80 is the deformation tensor with components @xmath81 .
the forces that crosslinks @xmath14 and @xmath66 exert on the origin are then @xmath82 where @xmath83 and @xmath84 run over @xmath85 and the summation convention is understood .
the sum of these forces is zero because @xmath86 .
the same analysis applies to any site and filament in the lattice . under affine distortions
, filaments do not bend , so the energy of the affine distortion depends only on the central force and does not depend on @xmath2 . thus , under affine distortions of sample traversing filaments , the force on every intermediate crosslink is zero , and non - affine distortions are not introduced : the response is affine and independent @xmath2 .
when the lattice is diluted , not every crosslink is connected to two others , the above cancelation does not occur , and the result is nonaffine response with a bending component . with these observations , we can calculate the shear modulus of any network of straight filaments with stretching energy only , i.e. , @xmath6 . for simplicity ,
we restrict our attention to two - dimensional systems in a rectangular- or rhombus - shaped box with base - length @xmath87 and height @xmath24 as shown in fig .
[ fig : diluted_kagome ] , and we consider only shear deformations in which the only nonvanishing component of @xmath88 is @xmath89 . in this case , only those filaments that cross from the bottom to the top of the sample will contribute to the shear modulus .
the length of such a sample traversing filament oriented along a unit vector @xmath90 making an angle @xmath23 with the @xmath19-axis is @xmath91 . under a shear deformation @xmath88
, its length will change by @xmath92 to lowest order in @xmath88 . since the spring constant of a filament of length @xmath15 is @xmath93 ,
the elastic energy of the filament is @xmath94 ^ 2 } { \sin \phi}.\ ] ] and the total energy from all filaments is @xmath95 ^ 2/\sin\phi\rangle$ ] , where @xmath96 is the total number of sample traversing filaments and @xmath97 signifies an average over the orientation angles of the filaments .
the linearized energy density , @xmath98 , is @xmath99 where @xmath100 are the components of the elastic constant or elastic modulus tensor and @xmath101 are the components of the linearized strain tensor @xmath102 . in the isotropic limit
, the energy density reduces to @xmath103 where @xmath102 is the traceless part of @xmath104 , and @xmath105 and @xmath106 are the shear and bulk moduli , respectively .
these relations along with eq .
( [ eq : e - fil ] ) for the case in which the only nonvanishing component of @xmath107 is @xmath89 lead to @xmath108 for the shear modulus of a network of elastic rods with stretching modulus @xmath1 .
note that rods parallel to the @xmath19-axis ( @xmath109 ) do not contribute to @xmath5 . in the kagome lattice ,
bottom - to - top traversing filaments all have length @xmath110 and their angles with respect to the @xmath19-axis are restricted to @xmath111 for which @xmath112 .
there are @xmath113 sites along the @xmath19 axis from which a single filament aligned either along @xmath114 or @xmath115 can emerge , so @xmath96 is simply @xmath116 , where @xmath117 is the probability a filament emerging from a given site along the @xmath19-axis traverses the sample .
for simplicity , we take @xmath118 in which @xmath119 .
when @xmath4 , all bonds are occupied , and the shear modulus of the undiluted kagome lattice is @xmath120 .
when @xmath121 and @xmath122 , there is a nonzero stretching contribution to @xmath5 in finite samples : @xmath123 in the @xmath124 limit , the probability of sample traversing filaments vanishes for any @xmath121 , and @xmath125 is zero at @xmath122 for all @xmath121 .
thus , as @xmath126 , @xmath127 must become smaller than @xmath128 , and for sufficiently small @xmath2 and @xmath3 near one , @xmath129 , and , as we will show in sec .
[ sec : numericalsimulations ] , the simple interpolation formula @xmath130 provides and excellent description of the simulation data with the emt form for @xmath131 near @xmath4 where the finite size corrections are the most important .
the kagome lattice has three sites per unit cell , which we take to be the three sites , labeled @xmath14 , @xmath66 , and @xmath67 , on an elemental triangle in fig .
[ fig : simplekagome ] .
all bonds in the lattice are parallel to one of the vectors , @xmath132 , @xmath133 , and @xmath134 , specifying the direction of bonds in the unit elemental triangle .
we label each site by an index @xmath135 where @xmath136 labels the position of site @xmath14 in the unit cell at dimensionless ( i.e. , taking @xmath137 ) position @xmath138 , where the factor of two arises because the separation between unit cells is twice the bond length .
the index @xmath139 labels the site within the unit cell , and the position of site @xmath140 is @xmath141 , where @xmath142 , @xmath143 , and @xmath144 .
when the lattice is distorted , @xmath145 maps to a new position @xmath146 , where @xmath147 is the displacement vector of site @xmath140 .
the stretching energy of a bond @xmath148 connecting nearest - neighbor ( @xmath12 ) sites @xmath140 and @xmath149 is @xmath150 where @xmath151 and @xmath152 is the unit vector parallel to @xmath153 . bending energy
is determined by angles @xmath154 between @xmath12 contiguous bonds @xmath155 and @xmath156 parallel to one of the lattice directions @xmath157 . to harmonic order , @xmath158 , where @xmath159 is the vector perpendicular to @xmath160 , which specifies the directions of both bonds @xmath148 and @xmath161 , according the right hand rule about the outward normal to the plane .
the bending energy arising from a nonzero @xmath162 is @xmath163 , and the harmonic energy for our latticed is thus [ eq : hami ] @xmath164 where @xmath165 if the bond connecting sites @xmath140 and @xmath149 is occupied and @xmath166 otherwise and where the sum in @xmath167 is over all continuous triples of sites @xmath168 along each filament .
the combination @xmath169 is the stretching elastic constant , and the combination @xmath170 is the bending constant for contiguous pairs of bonds . , @xmath171 , and @xmath172 .
sites and @xmath12 bonds in our choice for the unit cell labeled @xmath14 , @xmath66 , and @xmath67 and @xmath173 in italics , respectively .
two of the six @xmath174 bonds in a unit cell are shown at the bottom of the figure .
the other four have a similar configurations aligned along @xmath171 and @xmath172 rather than @xmath175 . ]
effective medium theory @xcite is a well - established approximation for calculating properties such as electronic band structure and phonon or magnon dispersions of random media .
various formulations of emt exist , but all replace the random medium with a homogeneous one whose parameters ( such as @xmath12 hopping strength or spring constant ) are determined by a self - consistent equation . here we use the formulation , presented in greater detail in app .
[ sec : app1 ] , in which self - consistency equations for the effective - medium parameters are determined by requiring that the average multiple - scattering potential or @xmath29-matrix associated with a single random bond ( or group of bonds ) in the homogeneous effective medium vanishes .
the development of an emt for elastic networks with both stretching and bending forces presents challenges not encountered in networks with stretching forces only . in our system , stretching forces are associated with single @xmath12 bonds , but bending forces are associated with contiguous pairs of @xmath12 bonds that couple next - nearest - neighbor ( @xmath174 ) sites in what we call _ phantom _
@xmath174 bonds that only exist if both its constituent @xmath12 bonds are occupied .
thus , the replacement of a single @xmath12 bond , which we will refer to as the replacement bond , in an effective medium affects not only the stretching energy of that bond but also the bending of the two @xmath174 phantom " bonds containing that bond .
there are currently two versions @xcite of emt on lattices with bending energies , which we will refer to as emt i and emt ii , respectively , that deal with this problem in different ways . in both methods ,
the effective medium is characterized by homogeneous stretching and bending moduli @xmath176 and @xmath177 respectively . in emt
i @xcite , the replacement bond has a stretching modulus @xmath178 and a bending modulus @xmath179 that take on respective values @xmath1 and @xmath2 if the bond is occupied and @xmath180 if the bond is not occupied . thus , the probability distribution for @xmath178 and @xmath179 , @xmath181 exhibits strong correlation between the values of @xmath178 and @xmath179 . as discussed in detail in ref .
@xcite , the bending constant of the @xmath174 phantom bonds containing the replacement bond is calculated assuming it is composed of two elastic beams connected in series , one with the bending modulus @xmath177 of the effective medium and one with the bending modulus @xmath179 of the replacement bond .
this leads to a bending constant , @xmath182 , for both @xmath174 bond containing he replacement bond , that is a nonlinear function of @xmath179 and @xmath177 : @xmath183 the perturbation to the effective medium arising from the replacement bond then consists of a stretching energy on that bond with spring constant @xmath184 and bending constants on the two @xmath174 bonds of @xmath185 .
it turns out , as discussed more fully in refs .
( @xcite ) and in app .
[ app : asym ] , that the emt equations in method i do not close unless an additional term , with coupling constant @xmath186 , coupling the angles on neighboring @xmath174 bonds along a single filament is added to the effective medium energy .
this leads to an additional term in the perturbation arising from the replacement bond , with coupling constant @xmath187 , that couples the two angles on the two @xmath174 bonds containing the replacement bond .
thus this emt is characterized by three parameters @xmath188 , and @xmath186 . in emt
ii @xcite , the phantom @xmath174 bonds carrying the bending energy are elevated to the status of real bonds that exist whether or not the @xmath12 bonds of which they are composed are occupied .
this leads to the great simplification that stretching and bending are effectively decoupled , and there is no necessity of introducing @xmath186 .
there are separate probability distributions for the stretching modulus of the @xmath12 replacement bond and for the bending modulus of the @xmath174 replacement bond .
however , since the bending modulus of a @xmath174 bond is zero unless both @xmath12 bonds comprising it are occupied , the probability that the @xmath174 is present with a bending modulus @xmath2 is chosen to be @xmath189 , the probability that any two given @xmath12 bonds are occupied : @xmath190 at @xmath4 , so it is equivalent to @xmath191 ) obtained from our numerical simulations ( data points ) : emt i ( solid lines ) and emt ii ( dashed lines ) .
different colors mark different values of @xmath192 , and from top to bottom the corresponding values of @xmath192 are @xmath14 , @xmath193 , @xmath180 .
the dotted line indicates the finite size correction predicted by eq .
( [ eq : finite - size ] ) as @xmath3 increases toward @xmath14 the simulation curve for @xmath194 follows first the emt curves and then the finite - size @xmath6 curve beyond the point at which the latter curves meet in accord with eq .
( [ eq : finite - size ] ) . for the values of @xmath195 ,
the emt solution is larger that the @xmath6 finite - size result , and thus the simulation data follow the emt curves . ]
the predictions of these two methods are qualitatively similar : they both yield a rigidity percolation threshold , below which the lattice loses rigidity , and a smoothly varying shear modulus approaching the pure lattice value of @xmath38 as @xmath11 .
figure [ fig : compare ] plots values of @xmath196 calculated using both emt methods for different values of @xmath197 and @xmath3 along with the results of simulations .
the @xmath198 simulation curves follows the finite - size result of eq . .
the @xmath199 curve follows the emt curve , which corresponds to infinite @xmath34 , with increasing @xmath3 and then the @xmath6 finite size curve after the two cross in accord with eq . .
as is the case for the triangular lattice @xcite , emt ii predicts a value of @xmath200 that is close to that measured in simulations , whereas emt i predicts a considerably larger value .
in addition , for values of @xmath197 beyond @xmath201 , the emt i numerical solution ceases to exist for small @xmath3 , and as a result the @xmath202 curve is not included for emt i in fig .
[ fig : compare ] . for @xmath203 and @xmath195 , the two emt curves and the simulation curves are essentially indistinguishable , and near @xmath4 , they become analytically identical with the effective medium modulus @xmath176 satisfying a scaling form @xmath204 where the scaling variable is @xmath205 and the scaling function is @xmath206 with the constant @xmath207 ^ 2\simeq 53.9 $ ] .
this scaling function can be expanded in the following limits @xmath208 therefore , for the case of @xmath209 corresponding to @xmath210 near @xmath14 so that @xmath211 , the shear modulus is approximately @xmath212 indicating that the macroscopic elastic response of the network is dominated by the stretching stiffness of the filaments and reaches the affine pure lattice limit as @xmath213 , and we shall call this the stretching dominatedelastic regime . in the other limit @xmath214 corresponding to smaller @xmath2 or larger @xmath215
so that @xmath216 , we have the shear modulus @xmath217 indicating that the macroscopic elastic response of the network is dominated by the bending stiffness of the filaments , and we shall call this the bending dominatedelastic regime . the emt solution for @xmath3 near one is plotted in terms of the scaling variable @xmath218 together with the asymptotic scaling function @xmath219 in fig .
[ fig : scalingnearp=1 ] .
because the asymptotic solution assumes small @xmath220 , it requires very small @xmath197 to make the regime @xmath221 visible .
we conclude from this that the elasticity of the network is bending dominant as long as @xmath222 for different values of @xmath223 .
the data points for the shear modulus stem from our simulations for @xmath224 unit cells .
the continuous curves depict our numerical solutions for @xmath176 of the emt equations for the corresponding values of @xmath2 .
the dashed red curve shows the asymptotic scaling function predicted by the emt , see eq .
( [ eq : scalfunc ] ) .
as expected , the simulation data and the numerical solution of the emt equations deviate from this scaling function as @xmath0 becomes large.,width=317 ] as discussed in sec .
[ sec : intro ] , near the rigidity threshold , @xmath225 vanishes as @xmath226 , where @xmath227 and @xmath228 for emt i and @xmath229 for emt ii , times a scaling function of @xmath230 that is a constant at large @xmath231 and proportional to @xmath231 at small @xmath231 .
thus , at sufficiently small @xmath231 for all @xmath121 , response is bending dominated with @xmath232 ; at sufficiently large @xmath231 , response is stretching dominated with @xmath233 as shown in the phase diagram of fig .
[ fig : phasediagram ] .
in the numerical portion of our work , we study the elasticity of the filamentous kagome lattice by generating diluted lattice conformations on a computer and then calculating their mechanical response numerically .
practically this is done via deforming the network by imposing a certain @xmath88 and by then minimizing the elastic energy ( [ eq : hami ] ) over the non - affine displacements @xmath234 of the sites using a conjugate gradient algorithm . to explore the response to shear , e.g. , we set @xmath235 , where @xmath236 specifies the magnitude of the applied deformation .
we use the same small magnitude @xmath237 for all deformations . for a range of @xmath3-values ,
we generate up to @xmath238 random conformations , we calculate several measurable quantities for multiple @xmath2-values , and we average arithmetically over all conformations . the quantities that we calculate are the elastic moduli and the corresponding fractions of rigid conformations @xmath239 and non - affinity parameters @xmath240 .
@xmath239 is defined as the number of conformations with non - zero @xmath100 ( non - zero meaning larger than a small numerical threshold , here @xmath241 ) divided by @xmath242 .
the non - affinity parameters @xcite @xmath243 where @xmath96 is the total number of sites and @xmath244 is the equilibrium non - affine displacement of site @xmath140 in the presence of the @xmath88 that leads to @xmath100 , measure the deviation from a homogeneous strain field . to mitigate boundary effects , we apply periodic boundary conditions on all boundaries .
we simulate system sizes ranging from @xmath245 to @xmath224 unit cells which corresponds to @xmath246 and @xmath247 sites , respectively .
( normalized by the bulk modulus @xmath248 at @xmath4 ) from our simulations of @xmath249 unit cells . ]
figures [ fig : compare ] and [ fig : blogplot ] show log - linear plots of the shear modulus @xmath5 and the bulk modulus @xmath250 , respectively , as functions of the occupation probability @xmath3 for @xmath249 unit cells .
the emt predicts the bulk modulus to be proportional to @xmath176 , and hence , both @xmath1 and @xmath250 should follow the same curves when they are normalized by their respective affine values @xmath38 and @xmath248 at @xmath4 . within the numerical errors ,
this prediction is indeed borne out by figs .
[ fig : compare ] ( a ) and ( b ) . moreover , these figures are consistent with the emt prediction that the rigidity percolation threshold @xmath200 be the same for all @xmath7 . numerically , we find @xmath251 which is only slightly smaller than the emt ii prediction @xmath252 and about 15% smaller than the emt i prediction @xmath253 .
previous emt predictions for elastic networks have produced somewhat larger values for the rigidity threshold than found in corresponding numerical studies , see e.g. , refs .
@xcite , and this apparent trend is continued here . for @xmath122 ,
the emt predicts a first - order rigidity percolation transition at @xmath4 .
the curves for @xmath122 in figs .
[ fig : compare ] ( a ) and ( b ) rise from zero with finite slope below @xmath4 , a clear finite - size effect .
below , we will analyze this finite - size effect systematically , and we will find that our data is indeed consistent with a first - order transition at @xmath4 .
figure [ fig : gammacombined ] shows the non - affinity functions @xmath254 and @xmath255 extracted from the simulation runs producing the results for @xmath1 and @xmath250 discussed in the previous paragraph . for @xmath4 ,
the network is affine , and the non - affinity functions are zero . the curves for @xmath256 are expected to have their maxima at @xmath200 , and our data is consistent with that expectation . the data for @xmath257 the curves for @xmath5 and @xmath250 should show no indication of the existence of @xmath200 , and indeed they do not . and ( b ) @xmath255 for @xmath249 unit cells .
color - code is the same as in figs .
[ fig : compare ] and [ fig : blogplot].,width=317 ] next , we turn to the fractions of rigid conformations .
these quantities are convenient for the purposes of finite - size analysis , in particular for the expected first - order transition for @xmath49 at @xmath4 . in the following , we will with focus on @xmath258 .
figure [ fig : nmucombined](a ) displays @xmath259 for @xmath249 unit cells for @xmath49 and several values of @xmath260
. the curves for @xmath7 are consistent with the above statement that the rigidity percolation threshold for all non - vanishing values of @xmath2 is at @xmath261 .
unit cells and various values of @xmath2 [ with the same color - code as in figs .
[ fig : compare ] and [ fig : blogplot ] . ] , and ( b ) for @xmath122 and system sizes of @xmath262 unit cells with @xmath263 ( increasing from left to right ) .
the bold lines correspond to the theoretical result ( [ nmutheotetical ] ) . ]
figure [ fig : nmucombined](b ) shows @xmath259 for @xmath49 for a variety of system sizes of @xmath262 unit cells .
along with the data points , the figure follows predictions for @xmath264 that follow from the elementary combinatorics ( see sec .
[ sec : elasticfil ] ) . consider a system such as that shown in fig .
[ fig : diluted_kagome ] with @xmath265 bonds along its two sides . only those filaments , which consist of @xmath266 bonds , that extend from the top to the bottom of the sample will contribute to @xmath5 , and @xmath5 will be zero unless at least one of the filaments starting on bottom reaches the top , which occurs with probability @xmath267 this function becomes a step function at p=1 when @xmath268 .
thus , we expect the data for @xmath49 to be fit by the function @xmath269 figure [ fig : nmucombined ] ( b ) reveals that the data is fit by this prediction remarkably well even though there is not a single adjustable fit - parameter involved .
now , we are in position to discuss the infinite - size limit . for @xmath270 ,
@xmath271 approaches a unit - step function at @xmath4 .
this establishes that the rigidity percolation transition for @xmath49 is a first - order transition at @xmath4 where the shear modulus jumps discontinuously from zero to its affine value .
now , let us look at the elastic response near @xmath4 .
figure [ fig : scalingnearp=1 ] shows the shear modulus and @xmath176 as functions of the scaling variable @xmath218 defined in eq .
( [ eq : tau ] ) .
we find remarkably good agreement between the simulation data , the numerical solution of the emt equations as well as the emt prediction for the asymptotic scaling function @xmath272 including the predicted value for the constant @xmath273 .
note that this agreement is obtained without adjusting any fit - parameters .
figure [ fig : scalingnearpb ] , finally , displays the scaling behavior near the rigidity percolation threshold @xmath200 .
our numerical data collapses in full qualitative agreement with the emt scaling form ( [ eq : scaling-2 ] ) albeit with an exponent @xmath274 that is significantly smaller than the @xmath227 predicted by the emt .
the value @xmath275 , on the other hand , is in semi - quantitative agreement with the emt . .
the plot shows both our simulation results and our numerical solutions of the emt i equations relative to the asymptotic scaling form near @xmath200 predicted by the emt , cf .
( [ eq : scaling-2 ] ) .
the data points stem from our simulations for @xmath249 unit cells , and we use @xmath276 , @xmath277 , @xmath278 , and @xmath274 .
the lower branches correspond to values of @xmath3 below @xmath200 and are a typical finite - size effect .
the continuous curves represent the numerical solutions to the emt i equations , which reduce to eqs . to as @xmath279 tends to zero with @xmath280 , @xmath281 , and @xmath282.,width=317 ]
we have introduced the diluted kagome lattice in @xmath66-dimensions as a model for networks of semi - flexible polymers , whose maximum coordination number is four . we identify straight lines of contiguous occupied bonds as filaments that , following previous work @xcite , we endow with the energy of an elastic beam characterized by a stretching ( young s ) modulus @xmath1 and a bending modulus @xmath2 .
we contrast this model , which has filaments of arbitrary length , with the mikado lattice whose filaments are of finite ( and usually fixed length ) .
we show that when @xmath6 , this kagome model has affine response and non - vanishing elastic moduli so long as there are sample - traversing filaments , which is the case for samples with finite lengths and widths @xmath34 , even for bond - occupation probabilities @xmath3 less than one . as @xmath34 increases , the probability of sample - traversing filaments decreases , and in the @xmath124 limit , elastic moduli , which are nonzero for undiluted @xmath4 lattice , fall precipitously to zero for @xmath283 in a first - order transition .
the undiluted lattice has @xmath2-independent , and thus affine , elastic moduli .
the addition of bending forces restores rigidity for any @xmath2 for @xmath3 greater than @xmath200 , the rigidity percolation threshold , and elastic moduli approach the non - zero affine values of the @xmath4 lattice as @xmath11 .
we argue that this is the underlying cause for the affine limit found in the large @xmath284 limit ( @xmath3 near one ) in the mikado model .
we use two recently introduced lattice - based effective medium theories to calculate elastic moduli of the diluted kagome lattice as a function of @xmath1 , @xmath2 , and @xmath3 , we we calculate scaling forms for the shear modulus both near @xmath4 and @xmath10 .
both forms are a function of the unitless ratio @xmath230 , where @xmath285 is the lattice spacing , and yield a crossover for all @xmath286 between bending dominated response at small @xmath231 and stretching dominated response at large @xmath231 .
we supplement our emts with numerical simulations of the shear modulus and other functions , such as that measuring the degree of nonaffine response .
the results of these simulations agree qualitatively with emt predictions for all @xmath3 and quantitatively with them near @xmath4 .
our study of the kagome lattice provides insight into the behavior of @xmath67-dimensional filamentous lattices with maximum coordination number of four . with stretching forces only ,
these lattices are subisostatic with of order one zero mode per site , and one might , therefore , conjecture that their elastic moduli should vanish when @xmath6 .
however , simulations @xcite on two model lattices consisting of straight filaments yields curves of elastic moudli as a function of @xmath3 that look very similar to those of the kagome lattice presented here with an approach to affine @xmath2-independent response as @xmath11 .
the underlying cause of this result is that straight sample - traversing filaments in _ any _ dimension give rise to nonvanishing macroscopic elastic moduli .
indeed , exact calculations @xcite for one of the @xmath287-lattices with elastic - beam energies show that all elastic moduli are nonzero when @xmath6 and @xmath4 and that response is affine . near @xmath4 , simulations on both of the @xmath287 lattice show a crossover from a bending dominated regime with @xmath288 to a stretching dominated regime in which @xmath289 in agreement with the predictions of our kagome - based lattice emt .
in fact the results of one of the simulations @xcite track the scaling function of the our kagome emt .
we conjecture that the important common feature of the @xmath290 and @xmath287 lattices is that they both consist of sample - traversing straight filaments in the undiluted limit . away from @xmath4 , bending forces stabilize the lattices in a process that is not so sensitive to lattice dimension , even though the central - force @xmath287 lattice is subisostatic .
our emts provide some support for this conjecture .
the characteristic property of the kagome emt solutions , such as the form of the scaling function near @xmath4 , are a direct consequence of the existence of lines of zeros in the phonon spectrum at @xmath122 that get raised to nonzero frequency for @xmath256 as shown in fig .
[ fig : integral ] .
this feature causes the integral @xmath291 [ eqs .
( [ eq : h2pi2 ] ) to ( [ eq : h23final ] ) ] to diverge as @xmath292 as @xmath293 . in three dimensions ,
the lines of zero modes in the phonon spectrum become planes of zero - modes , and we believe that the @xmath287 version of @xmath291 will have a from analogous to eq .
( [ eq : h23final ] ) but with @xmath294 replaced by a similar function of @xmath295 and @xmath296 : @xmath297 when @xmath298 , the denominator vanishes for all @xmath295 and @xmath296 when @xmath299 , and the integral diverges .
nonzero @xmath300 yields a @xmath301 divergence just as in the @xmath290 kagome case .
unfortunately , the @xmath287 lattices are quite complicated , with a @xmath302-site unit cell in one case @xcite and complicated crossing configurations in the other @xcite , and we have not been able to set up a @xmath287 emt to verify this conjecture . much of the work on the elastic properties of filamentous networks ( with the notable exceptions of @xcite ) have focused on models , such as the mikado or kagome lattice presented her , consisting of straight filaments .
this work has also largely focused on what are really mechanical models with beam energies assigned to the filaments and the effects of thermal fluctuations ignored .
as our analysis shows , these are exceptional lattices because they can support stress with central forces only even when their coordination number @xmath303 is substantially less the maxwell stability limit of @xmath304 .
filaments in real biological gels are not straight , and their energies are not described by that of an elastic beam .
it is thus legitimate to ask how seriously one should take models such as those presented here as realistic descriptions of real bio - gels of semiflexible polymers . if constituent filaments are not straight , then in general , at least some of elastic moduli of even perfect lattices vanish in the absence of bending forces .
prime examples of this behavior are found in the two - dimensional honeycomb lattice and the three - dimensional diamond lattice whose bulk moduli are nonzero but whose shear moduli vanish in the @xmath305 limit , but it has also been observed in a more realistic model of a filamentous lattice @xcite . if @xmath2 is needed to stabilize the system , the elastic response to external stresses will involve bending and thus be nonaffine .
so , it would seem that the straight - filament models are not such good ones for real systems , though they do provide us with valuable insight into how network architecture influences elastic response .
one of the predictions of the straight - filament models is that response becomes more stretching dominated and more affine as @xmath2 increases .
thus , it is plausible that if @xmath2 is large enough , the elastic response of even those lattices whose shear moduli vanish if @xmath6 will exhibit stretching dominated , nearly affine response for sufficiently large @xmath2 .
we are grateful for helpful discussions with fred mackintosh and chase broedersz .
this work was supported in part by national science foundation under dmr-1104707 and under the materials research science and engineering center dmr11 - 20901 .
to implement the @xmath29-matrix version of emt , we deal with the dynamical matrix @xmath306 of the em , its associated green s funcition @xmath307 ( we consider only zero frequency ) , the perturbation @xmath308 associated with bond replacement , the dynamical matrix @xmath309 , and its associated green s function @xmath310 .
[ for more details of this formalism , see refs .
@xcite and @xcite ] because there are three sites per unit cell , all of these matrices are @xmath311 matrices , to be detailed further below , where @xmath96 is the number of unit cells in the lattice .
we will specify @xmath312 more completely below . with these definitions , @xmath313 where @xmath314 the emt self - consistency equation requires that the disorder average over the probability distribution of eq .
( [ eq : p - i ] ) for emt i or eq .
( [ eq : p - ii ] ) or emt ii of the perturbation vanishes @xmath315 all of the matrices discussed here can be expressed in a position or wavenumber representation .
the six independent displacements in a unit cell can be expressed as a six - dimensional vector @xmath316 for each @xmath317 or as its fourier transform @xmath318 .
the energy of the bond - replaced system is then @xmath319 where @xmath320 and @xmath321 are @xmath322 dimensional matrices for each pair @xmath323 and @xmath324 .
the effective medium is translationally invariant and @xmath325 the energy of the effective medium is constructed by occupying all bonds with identical beams with stretching and bending moduli @xmath176 and @xmath177 and adding ( for emt i ) an additional coupling of strength @xmath186 coupling angles in neighboring @xmath174 bonds along a filaments [ see ref .
@xcite for details ] .
the em energy is then @xmath326 where @xmath327 and @xmath328 are evaluated at @xmath329 for all bonds @xmath155 , where @xmath330 where it is understood that @xmath331 and @xmath332 are all contiguous sites along a single filament .
@xmath333 can be constructed from the fourier transforms of the stretching and bending energies on @xmath12 and phantom @xmath174 bonds .
for example , the stretching energy of bond @xmath334 in fig .
[ fig : simplekagome ] , connecting site @xmath335 at position @xmath336 with site @xmath337 at position @xmath338 such that @xmath339 is @xmath340 ^ 2 $ ] .
the fourier transform of @xmath341 is @xmath342 where @xmath343 is specified in detail in eq .
( [ eq : bs ] ) below . similar procedures apply to all stretching and bending bonds , and the em dynamical matrix can be expressed as @xmath344 where @xmath345 and @xmath346 where @xmath347 are the unit vectors perpendicular to the bonds , @xmath348 should be understood as @xmath349 , and the same for @xmath347 so all these vectors are @xmath350 dimensional .
the factor of @xmath66 is from the fact that the length of the bonds is @xmath351 .
the scattering @xmath352 can also be expressed in this form .
we assume that the changed bond is bond @xmath14 in the unit cell at @xmath353 as marked in fig .
[ fig : simplekagome ] .
then we have @xmath354
asymptotic solutions for the emt i self - consistency equation has been developed in ref .
@xcite in the limit of small @xmath355 . in this section
we review this asymptotic solution and apply it to the case of kagome lattice . for convenience
we define the notation @xmath356 and @xmath357 where the second line is derived from the definition of the dynamical matrix . the self - consistency equation can be solved by projection to the space @xmath358 , @xmath359 , @xmath360 that spans @xmath352 . in this basis
we can rewrite the emt matrix equation into three independent equations @xmath361 where @xmath362 and @xmath363 are the projections of the green s function defined as @xmath364 with the inner product defined as @xmath365\end{aligned}\ ] ] where the sum is over all @xmath96 vectors in the first brillouin zone of the kagome lattice , and the trace is understood to include the sum over these vectors ( along with the factor of @xmath366 in addition to the sum over the @xmath350 dimensional space of @xmath367 .
these equations are exactly equivalent to the matrix equation . for the special case of @xmath368 , equations ( [ eq : asymeqone ] , [ eq : asymeqtwo ] , [ eq : asymeqthree ] ) simplify and give @xmath369 as discussed in ref .
@xcite , it is straightforward from the definition @xmath370 and the fact that the central force undiluted kagome lattice is isostatic with @xmath304 to derive the relation @xmath371 the effective medium filament stretching stiffness is then @xmath372 for @xmath373 and @xmath374 for @xmath375
. therefore this @xmath368 emt solution indicates a first order rigidity transition at @xmath375 . in the following we solve these equations asymptotically at small @xmath376 near the two critical points @xmath375 and @xmath377 .
the emt solution for small @xmath378 can be calculated perturbatively from the @xmath368 solution .
near @xmath375 , as discussed in ref .
@xcite , we can make simplifications to the self - consistency equation ( [ eq : asymeqtwo],[eq : asymeqthree ] ) using the fact that @xmath379 so that @xmath380 ( which we shall verify later ) and therefore we only need to solve eq .
using perturbation .
in contrast to the perturbative calculation in the triangular lattice , the kagome lattice effective medium is isostatic as @xmath381 , and thus the phonon green s functions exhibit singularities .
these singularities correspond to the zero - frequency floppy modes of the dynamical matrix , and make diverging contributions to @xmath382 .
nevertheless , perturbation theory around the @xmath368 solution is still well defined .
we shall see below that the @xmath382 singularity is proportional to @xmath383 and thus all the terms in the self - consistency equations are non - singular .
the expansion of @xmath362 at small @xmath231 can be calculated using the following equality @xmath384 = 1 .\end{aligned}\ ] ] these relations follow because all of the @xmath350 @xmath12 and the @xmath350 @xmath174 bonds in a cell are , respectively , equivalent by symmetry allowing the trace over one set of bonds to be replaced by @xmath385 the trace over the sum of the bonds .
but the quantity in square brackets is just the inverse of @xmath386 , and the final result follows . employing the analysis of the phonon modes in ref .
@xcite , we find that @xmath359 has a nonzero projection onto the floppy mode branch , which we call @xmath387 , whereas @xmath388 does not .
the floppy modes branch has low frequencies that are proportional to @xmath292 along symmetry directions at @xmath389 in the first brillouin zone , as shown in fig .
[ fig : integral ] . in this calculation
we shall use the direction @xmath390 which correspond to @xmath391 as an example . near the @xmath391 line the floppy modes branch phonon green s function takes the form , @xmath392 where @xmath393 with @xmath394 and @xmath395 and thus @xmath294 takes value between @xmath66 and @xmath67 in the first brillouin zone .
this form of the green s function is shown in fig .
[ fig : integral ] .
it was derived in ref .
@xcite in which the weak additional interactions are nnn bonds rather than bending forces .
it is straightforward to calculate the singular part of @xmath396 along symmetry directions of the floppy modes , which involves the integral @xmath397 where @xmath398 is the area of the kagome lattice unit cell , and the condition @xmath399 confines the integral to be around the floppy mode directions @xmath390 and @xmath400 near which the expansion is valid .
the total contribution involves the singular part of all the isostatic directions @xmath401 all these terms exhibit similar behavior so we just discuss @xmath402 for convenience . in eq . , assuming @xmath403
, one can first integrate out @xmath404 using contour integrals , which gives @xmath405 indicating a leading order divergence @xmath383 of @xmath406 at small @xmath407 .
a calculation including all terms in eq .
yields @xmath408 where @xmath409 from this we get @xmath410 at small @xmath407 . plugging this back into eq .
, we arrive at the solution @xmath411 where @xmath412 and this solution is asymptotically accurate for @xmath379 and @xmath413 .
the rigidity threshold @xmath377 can be obtained by plugging @xmath372 into eqs .
( [ eq : asymeqone],[eq : asymeqtwo],[eq : asymeqthree ] ) , which leads to @xmath415 following the calculation in ref .
@xcite we arrive at the asymptotic solution for @xmath416 @xmath417 where @xmath418 with @xmath419
the self - consistency equation of the emt ii can be written as @xcite @xmath420 where the variables @xmath421 and @xmath422 correspond to the integrals we defined as @xmath423 as shown in ref .
@xcite .
close to @xmath4 , if we take @xmath425 , then @xmath426 and we have @xmath427 plugging this back into eq .
we get @xmath428 which is a quadratic equation in @xmath429 , and the solution to this leads to the same asymptotic solution as in eqs .
( [ eq : mumasym],[eq : muscale ] ) .
next we discuss asymptotic behaviors near the rigidity threshold @xmath430 ( the subscript @xmath66 is used to distinguish it from the threshold @xmath377 in emt i ) .
the value of @xmath430 in emt ii can be readily obtained from @xmath431 along with the condition that @xmath432 , as discussed in the app .
d of ref .
we then have @xmath433 the emt ii self - consistency equations [ eq : dasemt ] can then be expanded at small @xmath434 and this leads to @xmath435 with @xmath436 this is in the same scaling form as in emt i [ eqs . and ] , apart from different constant factors
( , ) link:\doibase 10.1103/physrevlett.95.178102 [ * * , ( ) ] link:\doibase 10.1021/jp807749f [ * * , ( ) ] link:\doibase 10.1038/nmat1810 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.105501 [ * * , ( ) ] link:\doibase 10.1103/physreve.76.031906 [ * * , ( ) ] link:\doibase 10.1103/physreve.78.051801 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.91.108102 [ * * , ( ) ] link:\doibase 10.1103/physreve.68.061907 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.93.188102 [ * * , ( ) ] , ed .
( , ) link:\doibase 10.1103/physrevlett.91.108103 [ * * , ( ) ] link:\doibase 10.1103/physrev.178.1136 [ * * , ( ) ] link:\doibase 10.1103/physrevb.1.3250 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.46.465 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.038101 [ * * , ( ) ] link:\doibase 10.1038/nphys2127 [ * * , ( ) ] link:\doibase e35939 10.1371/journal.pone.0035939 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.75.4425 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.52.216 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.75.4051 [ * * , ( ) ] link:\doibase 10.1103/physreve.82.061902 [ * * , ( ) ] link:\doibase 078102 10.1103/physrevlett.108.078102 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.103.205503 [ * * , ( ) ] \doibase doi : 10.1016/0020 - 7683(78)90052 - 5 [ * * , ( ) ] link:\doibase 10.1209/0295 - 5075/87/24001 [ * * , ( ) ] link:\doibase 10.1103/physreve.72.066619 [ * * , ( ) ] link:\doibase 088301 10.1103/physrevlett.106.088301 [ * * , ( ) ] link:\doibase 10.1103/physreve.83.011111 [ * * , ( ) ] | the diluted kagome lattice , in which bonds are randomly removed with probability @xmath0 , consists of straight lines that intersect at points with a maximum coordination number of four .
if lines are treated as semi - flexible polymers and crossing points are treated as crosslinks , this lattice provides a simple model for two - dimensional filamentous networks .
lattice - based effective medium theories and numerical simulations for filaments modeled as elastic rods , with stretching modulus @xmath1 and bending modulus @xmath2 , are used to study the elasticity of this lattice as functions of @xmath3 and @xmath2 . at @xmath4 , elastic response
is purely affine , and the macroscopic elastic modulus @xmath5 is independent of @xmath2 . when @xmath6 , the lattice undergoes a first - order rigidity percolation transition at @xmath4 . when @xmath7 , @xmath5 decreases continuously as @xmath3 decreases below one , reaching zero at a continuous rigidity percolation transition at @xmath8 that is the same for all non - zero values of @xmath2 .
the effective medium theories predict scaling forms for @xmath5 , which exhibit crossover from bending dominated response at small @xmath9 to stretching - dominated response at large @xmath9 near both @xmath4 and @xmath10 , that match simulations with no adjustable parameters near @xmath4 .
the affine response as @xmath11 is identified with the approach to a state with sample - crossing straight filaments treated as elastic rods . |
studies of some epidemics , for example , the spread of the black death in europe from 13471350 @xcite , the past influenza pandemics rvachev95 , spread of fox rabies in europe , or spread of rabies among raccoons in eastern united states and canada @xcite , indicate that host and infective interactions and spatial distributions of their populations should play an important role in the dynamics of spread of many infectious diseases .
until recently most mathematical models of spread of epidemics have described interactions of large number of individuals in aggregate form and often these models have neglected aspects of spatial distribution of populations , importance of which have been addressed in @xcite . adopting methodologies like cellular automata , coupled map lattices , lattice gas cellular automata or agent based simulations , new classes of models
have been proposed and studied @xcite , to incorporate with various levels of abstraction and details : direct interactions among individuals ; spatial distribution of population types ( i.e. , infective , susceptible , removed ) ; individuals movement ; effects of social networks on spread of epidemics .
the goal of our work is to study the effects of population interactions and mixing on the spatio - temporal dynamics of spread of epidemics of sir ( susceptible - infected - removed ) type in a realistic population distribution . for the purpose of our study we developed a fully discrete individually - based simulation model that incorporates the random nature of disease transmission .
the key feature of this model is the fact that for each individual the set of all individuals with whom he / she interacts may change with time .
this results in time varying small world network structure . as a case study we considerder census data obtained from statistic canada @xcite for southern and central ontario .
the data set specifies population of small areas composed of one or more neighbouring street blocks , called dissemination areas . using these data , we study the effects of two types of interactions among individuals on the spread of epidemics .
the first type of interaction is the one among individuals located only in adjacent dissemination areas .
the second type of interaction is the one among individuals who in addition to being in contact with members of their own and adjacent dissemination areas may also be in contact with individuals located in remote , non adjacent , dissemination areas .
this last case can be seen as a case of `` short - cuts '' among multiple far away dissemination areas .
we investigate spatial correlations in our model and how they can be destroyed by the `` short - cuts '' in population contacts .
additionally , we derive a mean field description of our individually based simulation model and compare the results of the two models .
the presented work is continuation and expansion of our work in paper15,paper16,paper24 and contributes to better understanding of spread of epidemics of sir type , including influenza .
in order to study how population interactions ( `` population mixing '' ) affects the spread of epidemics we construct an individually based model in which each individual is represented by a particle , as in our earlier work paper15,paper16 .
models of this type take various forms , ranging from stochastic interacting particle systems @xcite to models based on cellular automata or coupled map lattices schon93,bc93,duryea99,benyo2000 . in our model
, we consider a set of @xmath0 individuals , labelled with consecutive integers @xmath1 .
this set of labels is denoted by @xmath2 .
we assume that each individual , at any given time can be in one of three , mutually exclusive , distinct states : susceptible ( s ) , infected ( i ) or removed ( r ) .
an individual can change state only in two ways : a susceptible individual who comes in direct contact with an infected individual can become infected with probability @xmath3 ; an infected individual can become removed with probability @xmath4 .
the precise description of the model is as follows .
the state of the @xmath5-th individual at the time step @xmath6 is described by a boolean vector variable @xmath7 , where @xmath8 if the @xmath5-th individual is in the state @xmath9 , where @xmath10 , and @xmath11 otherwise .
thus , @xmath12we assume that @xmath13 and @xmath14 , that is , the time is discrete .
hence , the only allowed values of the vector @xmath15 are @xmath16 no other values of @xmath15 are possible in sir epidemic model . in sir model
an individual can only be at one state at any give time and transitions occur only from susceptible to infected and from infected to removed .
a removed individual does not become susceptible or infected again in sir model .
therefore , sir model is suitable for studying spread of influenza in the same season because the same type of influenza virus can infect an individual only once and once the individual is recovered from the flu it becomes immune to this type of virus .
we further assume that at time step @xmath6 the @xmath5-th individual can interact with individuals from a subset of @xmath2 , to be denoted by @xmath17 .
using this notation , after one time iteration the @xmath18 becomes @xmath19where @xmath20 and @xmath21 is a sequence of iid boolean random variables such that @xmath22 , @xmath23 , and @xmath24 and @xmath25 is a sequence@xmath26of iid boolean variables such that @xmath27 , @xmath28 .
we assume that the sequences @xmath29 and @xmath30 of random variables are independent of each other and of the random variables @xmath31 observe that , if the @xmath5-th individual interacts with an infected @xmath32-th individual at time step @xmath6 and @xmath33 then the infection is transmitted from the @xmath32-th individual to the @xmath5-th individual at this time step .
thus , if some product @xmath34 takes the value 1 , then @xmath35 , meaning that the @xmath5-th individual has changed its state from susceptible to infected .
the key feature of this model is the set @xmath17 , representing all individuals with whom the @xmath5-th individual may have interacted at time step @xmath6 . in a large human population , it is almost impossible to know @xmath17 for each individual , so we make some simplifying assumptions . first of all , it is clear that the spatial distribution of individuals must be reflected in the structure of @xmath17 .
we have decided to use realistic population distribution for southern and central ontario using census data obtained from statistic canada @xcite .
the selected region is mostly surrounded by waters of great lakes , forming natural boundary conditions
. the data set specifies population of so called dissemination areas , that is , small areas composed of one or more neighbouring street blocks
. we had access to longitude and latitude data with accuracy of roughly @xmath36 , hence some dissemination areas in densely populated regions have the same geographical coordinates .
we combined these dissemination areas into larger units , to be called modified dissemination areas(mda ) .
we now define the set @xmath17 using the concept of mdas .
this set is characterized by two positive integers @xmath37 and @xmath38 .
let us label all mdas in the region we are considering by integers @xmath39 , where in our case @xmath40 . for the @xmath5-th
individual belonging to the @xmath41-th mda , the set @xmath17 consists of all individuals belonging to the @xmath41-th mda plus all individuals belonging to the @xmath37 mdas nearest to the @xmath41-th mda and the @xmath38 mdas randomly selected among all remaining mdas .
while the close neighbours , that is , the @xmath37 nearest mdas , will not change with time , the far neighbours , that is , the @xmath38 randomly selected mdas , will be randomly reselected at each time step .
the model described in the previous section involves strong spatial coupling between individuals .
before we describe consequences of this fact , we first construct a set of equations which approximate dynamics of the model under the assumption of perfect mixing , in other words , neglecting the spatial coupling . the state of the system described by eq .
( [ sirdyn1][sirdyn3 ] ) at time step @xmath6 is determined by the states of all individuals and is described by the boolean random field @xmath42 . under the assumptions of our model ,
the boolean field @xmath43 is a markov stochastic process . by taking the expectation @xmath44 of this markov stochastic process
when the initial configuration is @xmath45 we get the probabilities of the @xmath5-th individual being susceptible , or infected , or removed at time @xmath6 , that is , @xmath46 $ ] for @xmath47 .
since the sequences of random variables @xmath29 and @xmath30 are independent of each other and of the sequences of the random variables @xmath48 , assuming additionally independence of random variables @xmath48 , the expected value of a product of these variables is equal to the product of expected values . under these mean field assumptions , taking expected values of both sides of equations ( [ sirdyn1][sirdyn3 ] ) we obtain @xmath49since mean field approximations neglect spatial correlations , we further assume that @xmath50 is independent of @xmath5 , that is @xmath51 . even though sets @xmath17 have different number of elements for different @xmath5 and @xmath6 , for the purpose of this approximate derivation we assume that they all have the same number of elements @xmath52 , where @xmath53 is the average mda population .
all these assumptions lead to @xmath54the third equation in the above set is obviously redundant , since @xmath55 . similarly to the classical kermack - mckendrick model , mean field equations ( [ mf1])-([mf3 ] )
exhibit a threshold phenomenon .
depending on the choice of parameters , we can have @xmath56 for all @xmath6 , meaning that the infection is not growing and eventually it will die out because in our model no new individuals are being born or arrive from outside the area under consideration during the time of the epidemic .
alternatively , we can have @xmath57 for some @xmath6 , meaning that the epidemic is spreading .
the intermediate scenario of constant @xmath58 will occur when @xmath59 , that is , when @xmath60assuming that initially the entire population consists only of susceptible and infective individuals , that is , there are no individuals in the removed group at @xmath61we have @xmath62 .
furthermore , if @xmath52 is large , we can assume @xmath63 .
solving eq .
( [ trecond ] ) for @xmath4 under these assumptions we obtain @xmath64thus , assuming the mean field approximation the epidemic can occur only if @xmath65 .
the mean - field equations derived in the previous section depend only on the sum of @xmath37 and @xmath38 .
this means , for example , that the model with @xmath66 , @xmath67 and the model with @xmath68 , @xmath69 will have the same mean field equations .
however , the actual dynamics in these two cases are very different , see figure [ fig : frontlow ] and figure fig : frontdest . depending on the relative size of @xmath38 and @xmath37 ,
the epidemic may propagate or die out , as the following analysis shows . in order to make the subsequent analysis more convenient
, we introduce parameter @xmath70 , defined as @xmath71let @xmath72 be the expected value of the total number of individuals belonging to class @xmath73 , that is , @xmath74we say that an epidemic occurs if there exists @xmath75 such that @xmath76 . for fixed @xmath3 , @xmath38 and @xmath37 ,
there exists a threshold value of @xmath4 to be denoted by @xmath77 , such that for each @xmath78 an epidemic occurs , and for @xmath79 it does not occur .
obviously @xmath77 depends on @xmath3 , and this is illustrated in figure [ phasetran1 ] , which shows graphs of @xmath77 as a function of @xmath3 for several different values of @xmath70 , where @xmath80 .
the graphs were obtained numerically by direct computer simulations of the model .
the condition @xmath80 means that the size of the neighbourhood is kept constant , but the proportion of far neighbours(represented by @xmath70 ) varies . , and @xmath81 .
the first line from the bottom represents mean field approximation . ]
figure [ phasetran1 ] also shows the mean - field line given by eq .
( separatmf ) .
we observe that the parameter @xmath70 controls dynamics of the epidemic process in a significant way , shifting the critical line up or down .
when @xmath82 , that is , when there are no interactions with far neighbours , the epidemic process has a strictly local nature , and we can observe well defined epidemic fronts propagating in space , regardless at which mda the epidemic starts at @xmath83 .
this is illustrated in figure [ fig : frontlow ] , where the epidemic starts at a single centrally located mda with low population density ( figure fig : frontlowa ) and on figure [ fig : fronthigh ] , where the epidemic starts in a mda with high population density ( figure [ fig : fronthigh]a ) .
the simulations were done for the same parameters in both cases except for the different locations of the onsets of epidemics .
the figures display mdas that are represented by pixels colored according to the density of individuals of a given type .
the red component of the color represents density of infected individuals , green density of susceptibles , and blue density of removed individuals . by density we mean the number of individuals of a given type divided by the size of the population of the mda . , @xmath84 , @xmath85 , with ( a ) @xmath83 , ( b ) @xmath86 , ( c ) @xmath87 and ( d ) @xmath88 .
the initial outbreak is located in an area with low population density .
modified dissemination areas are represented by pixels colored according to density of individuals of a given type , such that the red component represents density of infected , green density of susceptibles , and blue density of removed individuals . ]
the epidemic waves propagating outwards can be clearly seen on figure fig : frontlow and figure [ fig : fronthigh ] , in the successive snapshots ( b ) , ( c ) and ( d ) . the fronts are mostly red .
this means that the bulk of infected individuals is located at the fronts .
after these individuals gradually recover the centers become blue .
let us now consider slightly modified parameters , taking @xmath89 .
this means that we now replace one close mda by one farmda .
this does not seem to be a significant change , yet the effect of this change is truly noticeable . , @xmath84 , @xmath85 , with ( a ) @xmath83 , ( b ) @xmath90 , ( c ) @xmath91 and ( d ) @xmath87 .
colour coding is the same as in the previous figure . ]
as we can see in figure [ fig : frontdest ] , the epidemic propagates much faster , and there are no visible fronts .
the disease quickly spreads over the entire region and large metropolitan areas become red in a short time , as shown in figure [ fig : frontdest](b ) .
this suggests that infected individuals are more likely to be found in densely populated regions , and their distribution is dictated by the population distribution unlike in figure [ fig : frontlow ] or figure [ fig : fronthigh ] , where infected individuals are to be found mainly at the propagating front .
in order to quantify the observations of the previous section , we use a spatial correlation function for densities of infected individuals defined as @xmath92where @xmath93 is the distance between @xmath5-th and @xmath32-th individual , and @xmath94 represents averaging over all pairs @xmath5 , @xmath32 satisfying condition @xmath95 . in the following considerations we take @xmath96 .
the distance between two individuals is defined as the distance between mdas to which they belong .
consider now a specific example of the epidemic process described by eq .
( [ sirdyn1]-[sirdyn3 ] ) , where @xmath97 , @xmath98 , and @xmath99 . for this choice of parameters epidemics always occur as long as @xmath100 .
figure [ cor3d ] shows graphs of the correlation functions @xmath101 at the peak of each epidemic , so that @xmath102 is the time step at which the number of infected individuals achieves its maximum value . for different values of @xmath103 , where @xmath97 , @xmath98 , and @xmath104 .
] an interesting phenomenon can be observed in the figure under consideration : while the increase of the proportion of far neighbours does destroy spatial correlations , one needs very high proportion of farneighbours to make the correlation curve completely flat . in bonabeau1998
it is reported that for influenza epidemics @xmath105 .
if we fit @xmath106 curve to the correlation data shown in figure [ cor3d ] , we obtain values of the exponent @xmath107 as shown in figure [ alpha ] . in order to obtain @xmath107 of comparably small magnitude as reported in @xcite , one would have to take @xmath70 equal to at least @xmath108 , meaning that vast majority of neighbours would have to be far neighbours . for different values of the parameter @xmath103 .
the exponent has been obtained by fitting @xmath109 to simulation data . ] in reality , this would require that the vast majority of all individuals one interacted with were not his / her neighbours , coworkers , etc . , but individuals from randomly selected and possibly remote geographical regions .
this is clearly at odds with our intuition regarding social interactions , especially outside large metropolitan areas .
this prompted us to investigate further and to find out what is responsible for this effect .
upon closer examination of spatial patterns generated in simulations of our individually - based model , we reach the conclusion that the inhomogeneity of population sizes in neighbourhoods @xmath17 makes spatial correlations so persistent .
since different mdas have different population sizes , we expect that some individuals will have larger neighbourhood populations than others , and as a result they will be more likely to get infected , even if the proportion of infected individuals is the same in all mdas .
this will build up clusters of infected individuals around populous mdas . to test if this is indeed the factor responsible for strong spatial correlations in our model , we replaced all mda population sizes with a constant population size @xmath53 , that is , the average mda population size .
as expected , graphs of the correlation functions obtained in this case were are all essentially flat , with the exponent @xmath107 close to zero even in the case of @xmath69 , when we obtained @xmath110 .
we conclude that spatial correlations are difficult to destroy if neighbourhood sizes are inhomogeneous
. very significant amount of long - range interactions ( i.e. , very strong mixings ) is required to obtain flat correlations curves . however , for homogeneous neighbourhood sizes , even relatively small long - range interaction immediately forces the process into the perfect - mixing regime , resulting in the lack of spatial correlations .
the model that we developed can be applied to other realistic population distributions and geographical regions . | _ abstract : - _ we study dynamics of spread of epidemics of sir type in a realistic spatially - explicit geographical region , southern and central ontario , using census data obtained from statistics canada , and examine the role of population mixing in epidemic processes .
our model incorporates the random nature of disease transmission , the discreteness and heterogeneity of distribution of host population.we find that introduction of a long - range interaction destroys spatial correlations very easily if neighbourhood sizes are homogeneous . for inhomogeneous neighbourhoods , very strong long - range coupling is required to achieve a similar effect .
our work applies to the spread of influenza during a single season . |
the recent interest in nuclei beyond the proton drip line , prompted the study of proton emission phenomenon from the ground state of such nuclei @xcite .
like in the case of @xmath7 decay , this process consists in the tunneling of the proton across a potential barrier .
however for this kind of radioactivity it is vital to add the centrifugal potential to the coulomb barrier since the majority of emitted protons are likely to decay from states with @xmath8 . in last years a series of theoretical investigations , based on the quasi - classical approximation
have been carried out in this field @xcite .
the proton half - lives of observed heavy proton emitters were calculated and compared with the experimental ones and in some cases a good agreement was found . in the gamow approach
the decay is treated as a stationary process , the penetrability being given by the ratio of probabilities of finding the quantum particle on each side of the potential barrier . in this image
the dynamical aspects are neglected .
however quantities like the tunneling time are important in decay processes , in fission or in fusion reactions . in order to include the time evolution of a wave packet propagating in a classically forbidden region one need to solve the time dependent schr " odinger equation ( tdse ) .
there have been some attempts to incorporate the time - dependency in the study of the proton decay @xcite by expanding the wave function into a term describing the bound - state of the prepared nucleus @xmath9 and a set of exit channel wave functions @xmath10 orthogonal to it which decomposed into the intrinsic states of the daughter nucleus @xmath11 and of the proton @xmath12 . however , imposing that the decay width is smaller than the resonance ( quasi - stationary ) energy ( @xmath13 ) the treatment reduce to the stationary schr " odinger equation for @xmath14 with a complex energy @xmath15 .
afterwards the usual computational procedure is carried on @xcite . in this paper ,
based on a previous application of tdse to the study of @xmath7 decay @xcite , we address the question of proton - decay from orbitals with @xmath8 for the spherical nucleus @xmath0pb .
this process is different from the ground - state proton emission beyond the proton drip line .
however the theoretical techniques that we employ in this paper can be applied in both cases without major changes . in our time - dependent approach the description of the decaying system is fully contained in the state vector .
the time evolution of this vector enables us to determine the decay probability at any moment @xmath16 .
we are interested to study the influence of the potential on the tunneling quantities and to determine to what extent the errors of the wkb method depends on the energy of the quasi - stationary state and on the angular momentum .
the interaction between the proton and the daughter nucleus @xmath17tl is described by an average woods - saxon ( ws ) field which accounts for the nuclear potential @xmath18 a coulomb potential , which is approximated by the interaction between the point proton and the uniformly charged spherical core of charge @xmath19
@xmath20 , r\le r_{0}^c \\ \nonumber & = & \frac{(z-1)e^2}{r } , \hskip 4.65 cm r\ge r_{0}^c \end{aligned}\ ] ] and the centrifugal barrier @xmath21 the ws interaction is defined by the nuclear radius @xmath22 , with @xmath23= 1.25 fm , the diffuseness @xmath24=0.7 fm , and the depth of the central potential @xmath25= 58 mev @xcite .
note that in other papers , the depth of the potential is not fixed , its value being adjusted to reproduce the experimental energy of the quasi - bound state @xcite . the coulomb radius is given by @xmath26 .
the initial wave function of the proton was chosen to correspond to a quasi - stationary state of the potential @xmath27 , with positive energy @xmath28 .
the decay width of such a metastable state can be calculated using an analytic method developed by gurwitz et al .
@xcite . in this approach
it is assumed that the proton occupies a bound eigenstate in the potential well @xmath29 which represents a slight modification of @xmath30 .
therefore we take @xmath31 where , @xmath32 is an eigenstate of energy @xmath33 corresponding to the hamiltonian @xmath34 where @xmath35 is the reduced mass of the daughter - proton system .
the modification @xmath36 reads @xmath37 where @xmath38 gives the slope of the potential barrier beyond the point @xmath39 at which @xmath30 attain its maximum . in what follows we shall consider only the wave functions @xmath40 with the highest eigenvalue @xmath33 bellow the barrier @xmath41 .
in table i we list the heights of the barriers @xmath42 , their locations @xmath43 and the selected eigenvalues @xmath44 for a given value of the angular momentum @xmath1 .
the next step consists in the resolution of the time dependent schrdinger equation : @xmath45 where @xmath46 a numerical procedure based on the iterated leap - frog method , provides the solution of this equation @xcite .
the equidistant spatial grid used in this method is large enough such that no interference of the outgoing and ingoing wave functions will take place during the time intervals employed in the calculations .
once we get the wave - function which describes the time evolution of the proton packet through the potential barrier we are able to compute relevant quantities of the decay process . the tunneling probability can be expressed as the probability of finding the proton beyond a certain point @xmath47 on the border which separates the zone inside the barrier from the external one @xmath48 the decay rate is calculated according to the relation @xmath49 it is also interesting to calculate the average value of the proton wave packet position operator , @xmath50 , inside the nucleus @xmath51
in figure 1 a , b we present the time evolution of the proton wave function @xmath52 for two angular momenta @xmath53 at four different moments .
we see the tendency of the wave function to decrease its amplitude in the interior of the barrier when time goes on .
eventually we observe that the fraction of the wave function which passed across the barrier behaves like a spreaded wave packet ( at least on the spatial interval that we considered , i.e. up to 60 fm ) .
this tendency is more pronounced for the wave function with @xmath1=5 , because it faces a thiner barrier ( see table 2 ) .
it is worthwhile to mention that although the wave function amplitude decreases constantly in time its shape does not change to much inside the barrier . in figures 2 a
, b the time - dependent decay rate @xmath54 is ploted for angular momentum @xmath1 = 0 , 2 , 5 and 8 .
as has been pointed in an earlier work @xcite the decay rate undergoes two regimes . in the first one , @xmath54 oscillates but increases on the average . the fact that at small times the decay rate is not constant as characteristic for exponential decay , but varies with time is typical for a quantum mechanical description @xcite .
this fact contrasts to the usual classical image which portrays the radioactive system as an ensamble of nuclei decaying independently one of each other with a probability which does not depend on time . in the second regime
@xmath54 performs small fluctuations around a constant value , that we call asymptotic value @xmath55 .
the decay rates presented in figure 2 a correspond to @xmath56 fm , i.e. the inferior limit of integration is choosed to lay between the two turning points as can be observed from table 2 . if @xmath57 is increased , the irregularities occuring in the first regime are smoothed out .
this fact is pictured in figure 2 b , where @xmath47 is choosed to be 25 fm . in figure 3
we represented the behaviour of the wave packet s average position inside a potential region defined by @xmath58 fm .
as in the case of the decay rate we deal with two regimes . whereas in the first regime @xmath50 increases up to a certain limit , in the second one it performs small - amplitude fluctuations around this limiting value as we expected since , according to a previous remark , the wave function does not change to much its shape .
the period of oscillations in this second regime is denoted by @xmath3 and its value can be deduced simply by measuring the distance between two maxima of @xmath50 ( right column in fig.3 ) .
as has been noted in previous papers @xcite , the quasi - stationary state tends to penetrate the potential barrier by performing these small - amplitude oscillations instead of simply crossing the barrier from one side to the other .
for this reason it makes sense to associate the frequency of collisions in the formula of gamow with @xmath59 instead of @xmath60 , which is defined by the quasi - classical expression ( see eq.(13 ) bellow ) . in the wkb approximation
the decay rate @xmath61 is constant , its value resulting from the product of the barrier penetrability @xmath62 and the collision frequency @xmath63 @xmath64 where @xmath63 is given by the inverse of the classical period of motion @xmath65 the wave number @xmath66 reading @xmath67^{1/2}\ ] ] and the penetrability @xmath68 the stationary states computed in the modified potential , become quasi - stationary when we turn on the real potential ( without the modification @xmath69 ) and their energy is no longer well defined .
therefore , in all the above formulas the decay energy @xmath70 was taken to be the energy of the quasi - stationary state @xmath71 , computed as the average energy of the time dependent scr " odinger equation @xmath72 in what concerns the decay rate , it can be inferred from table 2 that the decay rates calculated in the time dependent approach @xmath54 are in relatively good agreement with the wkb values @xmath73 especially when the quasi - stationary energy decreases with respect to the barrier height .
except the last case , with the highest angular momentum , in all other cases the wkb results overestimates the time - dependent values .
the comparison between the times @xmath3 and @xmath74 ( see table 2 ) shows that they have very close values .
however they are refering to different types of motion , the first describing the small oscillations of the proton wave function between the walls of the barrier during tunneling , the second , the classical movement inside the potential well . the transient time is defined as the time interval between the moment when @xmath54 starts to increase up to the moment when it reaches the limiting value @xmath55 .
it depends on the energy of the quasi - stationary state .
its value can be deduced by inspecting figures 2 a and b. the tunneling time is related to the shift in time of the transition point between the two above mentioned regimes , for the decay rate .
therefore it can be associated to the time necessary for the wave function to cross the barrier .
computing the decay rate for two different values of @xmath47 , i.e. the barrier s turning points , and measuring the time delay between the two maxima of the two curves one gets the tunneling time ( see fig.4 a and b ) .
notice that for @xmath75 the tunneling time decreases with increasing @xmath55 .
however this does not happens for the state with higher angular momentum @xmath76 , where , although the decay rate is smaller , the tunneling seems to take place faster .
in fact the tunneling time is correlated with the imaginary time , which is nothing else than the time necessary for the proton packet to cross the inverse potential , i.e. @xmath77
motivated by recent theoretical and experimental investigations on proton radioactivity phenomenon , we studied the time dependent characteristics of the proton tunneling from excited states in the spherical nucleus @xmath0pb .
although this process is different from the ground - state proton emission beyond the proton drip line , the approach employed in this paper could be easily extended to the above mentioned case of proton radioactivity . since
other theoretical approaches are based on the semiclassical approximation we were interested to compare our exact results with the wkb ones .
we found that the discrepancy between the two methods decreases when the difference between the top of the barrier and the energy of the quasi - stationary state increases .
our study does not concern a certain angular momentum state which could be measured in the decay reaction @xmath0pb@xmath78@xmath79tl @xmath80 . rather for a fixed set of woods - saxon parameters we investigated the dependence of the proton tunneling on time choosing one quasi - stationary state for every angular momentum .
it seems that the accuracy of the wkb approximation increases with @xmath1 . for a comparison with the experiment one should fit some of the ws parameters , e.g. the potential depth , in such a way to reproduce the observed energy .
however the present comparison between the wkb and the time - dependent approaches gives an idea of the error involved in the stationary approach and provides a good starting point for future investigations of proton decay using tdse which could eventually answer to some questions related to this phenomenon .
99 1.5truecm hofmann s 1989 _ particle emission from nuclei _ , edited by m.ivacu and d.n.poenaru ( crc , boca raton , fl ) .
kadmenskii s g and khlebostroev v g 1974 _ sov.j.nucl.phys .
_ @xmath81 505 .
+ bugrov v p and kadmenskii s g 1989 _ sov.j.nucl.phys .
_ @xmath82 967 .
buck b , merchant a c and perez s m 1992 _ phys.rev .
c _ @xmath83 1688 .
aberg s , semmens p b and nazarewicz w 1997 _ phys.rev .
c _ @xmath84 1762 .
feix w f and e r hilf 1983 _ phys.lett . _
@xmath85 14 .
de vries r m 1976 _ comput.phys.commun .
_ @xmath86 249 .
serot o and carjan n 1994 _ zeit.phys.a _ * 349 * 353 .
serot o , carjan n and strottman h 1994 _ nucl.phys .
_ @xmath87 562 .
rost e 1968 _ phys.lett.b _ * 26 * 184 .
gurwitz s a and kalbermann g 1987 _ phys.rev.lett . _
@xmath88 262 .
serot o 1992 _ dissertation thesis , cenbg:9216 _ ( universit de bordeaux i ) .
grotz k and klapdor h v 1984 _ phys.rev .
c _ @xmath89 2098 . .
the values of the barrier heights @xmath90 , their locations @xmath39 , the selected eigenvalues @xmath91 and the wave functions number of nodes for different angular momenta @xmath1 . [ cols="^,^,^,^,^,^,^",options="header " , ] @xmath92 time evolution of the squared wave function @xmath93 for angular momentum ( a ) @xmath94 and ( b ) @xmath95 at moments @xmath96 0 , 1.5@xmath9710@xmath98 , 3@xmath9710@xmath98 and 4.5@xmath9710@xmath98 seconds . in the second case
the barrier through which the proton undergoes tunneling is thiner and therefore the probability to find it outside the barrier at a later time is larger .
the wave function amplitude decreases in time and later on , the part of the wave function which already tunneled manifests itself as a well - spreaded wave packet .
1.0truecm @xmath99 the time dependent decay rates @xmath54 for the four selected quasi - stationary states of angular momentum @xmath100 . in the eq.(10 ) we choosed ( a ) @xmath56 fm and ( b ) @xmath101 fm . in all four cases we observe that after a certain time the decay rate will fluctuate around an asymptotic value . in the second case
these fluctuations are smoother .
1.0truecm @xmath102 the average value of the wave packet position operator @xmath50 for three quasi - stationary states ( @xmath103 ) .
on the left side the asymptotic behaviour of @xmath50 is observed .
whereas on the right side we foccused on the small amplitude oscillations of the wave packet on its way to tunneling
. 1.0truecm @xmath104 the decay rates for ( a ) @xmath105 and ( b ) @xmath106 when @xmath57 is choosed to be the internal turning point ( full line ) and the external turning point ( dotted line ) .
the difference between the maxima of the two curves gives the tunneling time @xmath5 . | based on a single particle model which describes the time evolution of the wave function during tunneling across a one dimensional potential barrier we study the proton decay of @xmath0pb from excited states with non - vanishing angular momentum @xmath1 .
several quantities of interest in this process like the decay rate @xmath2 , the period of oscillation @xmath3 , the transient time @xmath4 , the tunneling time @xmath5 and the average value of the proton packet position @xmath6 are computed and compared with the wkb results . |
inspection of the galaxy distribution in redshift surveys ( e.g. in the sloan digital sky survey , york et al .
2000 ) or of the distribution of dark matter particles in numerical simulations ( e.g. millennium simulations , springel et al .
2005 ) reveals a striking ` cosmic web ' ( bond et al .
1996 ) , composed of clusters , filaments , sheets and voids as primary building blocks . in the matter or galaxy
distributions , one can clearly see large volumes of ` empty ' regions ( voids ) which are surrounded by thin denser sheetlike structures ( sheets ) .
at even higher density contrast a network of filaments dominates the matter or galaxy distributions .
finally , the massive clumps at the intersections of filaments typically correspond to rich galaxy groups and clusters .
the human eye is readily capable of identifying these morphological features of the cosmic large - scale structure , but for quantitative analysis objective techniques for structure analysis are needed .
the most commonly employed statistical measure for the distribution of matter and galaxies are low - order clustering statistics ( e.g. , the two - point correlation function and its fourier transform , the power spectrum ; peebles 1980 ) .
apart from these standard statistical tools , one may also use other statistics , for example based on higher - order correlation functions , or measures for the abundance of halos as a function of mass or of voids of different sizes , to characterize cosmic structure .
since halos and voids are approximately spherically symmetric structures and relatively easy to model , several very successful methods to extract these structures from simulations and observations have been developed . in n - body simulations ,
halos are most commonly found as groups of particles by the friends - of - friends ( fof ) algorithm with a linking length set equal to some fraction ( @xmath1 ) of the mean particle separation ( e.g. davis et al .
halos detected in this way show a mean enclosed density of about @xmath2 times the average mass density of the universe .
observationally , galaxy groups can also be identified using the fof method , but with two linking lengths in order to take redshift distortions into account ( e.g. , eke et al .
one may also use models for the density profile and velocity dispersion of dark matter halos to help extracting galaxy groups ( e.g. , yang et al . 2005 ) .
there are numerous successful void finders , many of them have been compared in detail in a recent study by colberg et al .
( 2009 ; and references therein ) .
compared to clusters and voids , filaments and sheets are more complicated geometric structures .
both of them are genuine 3-dimensional objects associated with a distinct orientation in space .
their density contrast is often quite close to the mean cosmic density , making their identification difficult and sometimes ambiguous .
we also note that from a dynamical point of view there exists a sequence of transformations that can cast one of these structures into another , which blurs any clear distinction between these different morphological structures .
matter tends to flow out of the voids , accretes onto the sheets , collapses to the filaments , and finally accumulates onto the large clumps at the intersection of the filaments .
these processes are expected to impact the properties of the halos and the galaxies formed therein , leading to correlation between halo properties and large - scale environment . in a recent study , using high - resolution n - body simulations , gao et al .
( 2005 ) found that the halo clustering strength not only depends on the masses of the halos but also on their formation times ( see also sheth & tormen 2004 ; wang et al .
2007 ; dalal et al . 2008 ; hahn et al . 2009 ) .
in addition to the age , other halo properties such as concentration and spin have also been found to correlate with the local environment ( avila - reese et al .
2005 ; wechsler et al . 2005 ; bett et al . 2007 ; gao & white 2007 ;
macci et al .
2007 ) . based on semi - analytical galaxy evolution models , croton et al .
( 2007 ) showed that besides the properties of the host halo the large scale environment has to be taken into account to fully characterize the galaxies at halo centers ( see also reed et al .
observations indicate a similar dependence of galaxy properties on environment . using a large galaxy redshift catalog , yang et al .
( 2006 ) showed that groups ( of similar masses ) with red central galaxies are more strongly clustered than those with blue central galaxies ( see also wang et al .
2008 ) . using the same data set , wang et al .
( 2009b ) detected numerous red dwarf galaxies which were isolated , i.e. which were not belonging to any neighboring larger association or group of galaxies .
the origin of those red isolated dwarfs still remains unknown ( see also cooper et al . 2007 ) . to shed some more light on the impact of large - scale structure on halo / galaxy properties we here adopt the following strategy .
we first identify and classify the large - scale environment , then we determine whether a given halo resides in a filament or sheet , and finally we try to find correlations between the halo spin and shape orientations and their large - scale environment . up to the present day , a number of different approaches have been suggested to find filaments ( and/or sheets ) in simulations as well as in observations . among these methods
two different classes of techniques can be distinguished : the first uses discrete distributions of objects ( e.g. , galaxies ) and the second is based on continuous density fields . in the following , we briefly summarize the basic ideas of these two approaches ( see also aragn - calvo et al . 2007a ; and references therein ) . a discrete point set allows the use of a minimum spanning tree technique to link the points ( barrow et al .
1985 ; colberg 2007 ) .
however , cosmic structures identified by this method often show web - like features and it is difficult to define sizes and directions of the extracted filaments . finding filaments joining neighboring clusters has also been carried out .
pimbblet ( 2005 ) searched the 2df galaxy redshift survey catalog for filamentary structures using the orientations of inter - cluster galaxies . a related approach based on the inter - cluster dark matter distribution derived from n - body simulations
is discussed in colberg et al .
more mathematically , stoica et al .
( 2005 ; 2007 ) used a so called _ candy model _ to locate the filaments in galaxy surveys .
all these approaches have the advantage that they directly deal with the original point sources and do not require the creation of continuous density fields .
however , in general these techniques incorporate a large number of free parameters and the specific assignment of these parameters is , to a certain degree , arbitrary .
the second type of techniques usually require the creation of a continuous density field .
these approaches include the full characterization of the topology of the matter distribution in terms of minkowski functionals ( mecke et al .
1994 ; schmalzing et al .
1999 ) , the genus of the density field ( gott et al . 1986 ; hoyle & vogeley 2002 ) and the shape - finder statistics in three ( sahni , sathyaprakash & shandarin 1998 ) and two dimensions ( pandey & bharadwaj 2008 ) . a more rigorous classification of filaments and other structural elements is obtained with the skeleton analysis of density fields by novikov et al .
( 2006 ) and sousbie et al .
( 2008 ) , where the maxima and saddle points in the density field are specified using the hessian matrix ( i.e. the second partial derivative ) of the density field .
recently , hahn et al . ( 2007a ; b ) have quantified the cosmic web using the hessian matrix of the potential field where according to the number of positive eigenvalues a region was classified as belonging either to a cluster , filament , sheet or void .
the only free parameter in this analysis is the smoothing length of the density field . in a similar spirit , aragn - calvo et al .
( 2007a ; b ) computed the hessian matrix based on the density field constructed with a delaunay triangulation field estimator ( schaap & van de weygaert 2000 ) .
the methods in this second category are mathematically more rigorous than the techniques based on discrete point sets . only a few ( or even no ) free parameters are needed , and all halos ( galaxies ) can be classified since essentially every volume element and hence the objects within it can be classified .
however , a disadvantage lies in the necessity for constructing a density ( potential ) field , and in the difficulties in assigning points , halos , or galaxies , to the appropriate volume element . in this study , we compare two types of filament finding algorithms , one using the overall matter distribution and one just the halos obtained from a high - resolution cosmological n - body simulation .
method slowromancap1@ employs the hessian matrix of the mass density field , where halos are differentiated into four types using a combination of criteria from hahn et al .
( 2007a ) and aragn - calvo et al .
( 2007a ) .
method slowromancap2@ uses the halo distribution directly .
our development of this approach has been inspired by the candy model approach suggested by stoica et al .
( 2005 ) and the connecting cluster technique described in colberg et al .
based on either method we extract filaments and sheets from our high resolution n - body simulation .
we then correlate the characteristics of the cosmic structures , like sizes and orientations , with the spins and orientations of dark matter halos .
this paper is organized as follows . in section [ sec_data ] ,
we give a brief description of our n - body simulation and the halo catalog we used .
we present a detailed description of the filament - finding methods based on either the dark matter density field or the distribution of dark matter halos , in section [ sec_method ] . in section [ sec_results ] , we analyze various alignment signal measurements for halos with respect to the filaments ( and sheets ) .
finally , we summarize our results in section [ sec_summ ] .
in this study , we make use of a n - body simulation carried out at the shanghai supercomputer center using the massively parallel gadget2 code ( springel et al . 2001 ; 2005 ) .
the simulation evolved @xmath3 dark matter particles in a periodic box of @xmath4 on a side from redshift @xmath5 to the present epoch .
the particle mass and softening length are @xmath6 and @xmath7 , respectively .
the cosmological parameters used in the simulation are @xmath8 , @xmath9 , @xmath10 , @xmath11 , @xmath12 , and @xmath13 . in the upper - left panel of fig .
[ fig : simulation ] , we show the distribution of dark matter particles in a slice of thickness @xmath14 at redshift @xmath15 . for clarity ,
only 0.15% of the dark matter particles are plotted .
this representation clearly shows the well - known features of the non - linear cosmic density field , in particular the prominent filamentary structures that coined the term ` cosmic web ' are nicely seen .
dark matter haloes were identified from the simulation at redshift @xmath15 using the standard friends - of - friends ( fof ) algorithm ( davis et al .
1985 ) with a linking length of 0.2 times the mean interparticle separation .
haloes obtained with this linking length have a mean over - density of @xmath16 ( porciani , dekel & hoffman 2002 ) . as an illustration ,
we show in the upper - right , lower - left and lower - right panels of fig .
[ fig : simulation ] the halo distributions for a sequence of decreasing lower mass limits : @xmath17 , @xmath18 and @xmath19 , as indicated in the panels .
the sizes of the dots are scaled to be proportional to the virial radius of the dark matter halos .
these plots indicate that the distribution of halos with masses @xmath20 can only resolve the high density regions at the nodes of the cosmic web , while halos with masses @xmath21 trace the filamentary structures quite well . including smaller halos down to @xmath19 can reveal subtle features even in void - like regions . as discussed in yang et al .
( 2009 ; fig .
1 ) , the sdss observations can completely resolve the dark matter halos with a mass limit of @xmath20 at redshift @xmath22 , and of @xmath23 at redshift @xmath24 .
thus the halo - based probe for filamentary structures investigated in this paper can be reliably applied to the sdss observations at low redshifts .
the main purpose of this paper is to probe the orientations and spins of dark matter halos with respect to the filaments and sheets within the cosmic web .
however , a reliable measurement of these halo properties , in particular of the spin , requires high mass resolution .
therefore we only retain halos with at least @xmath25 particles for further analysis , resulting in a catalog with @xmath26 halos .
the angular momentum @xmath27 of a fof halo containing @xmath28 particles is @xmath29 where @xmath30 is the particle s mass , @xmath31 is the position vector of the @xmath32-th particle relative to the center of mass , and @xmath33 is its velocity relative to the bulk velocity of the halo . in order to determine the shape of a fof halo , we use the moment of inertia tensor ) represents the second moment of the mass tensor .
the correct definition of the moment - of - inertia tensor deviates from this expression ( see e.g. , eq . 1 in hahn et al .
however , in the context of cosmological alignment studies it has become a convention to call the second moment of mass tensor inertia tensor . ]
@xmath34 with @xmath35 where @xmath36 denotes the component @xmath37 of the position vector of particle @xmath32 with respect to the center of mass .
the axis ratios @xmath38 , @xmath39 and @xmath40 @xmath41 are proportional to the square roots of the corresponding eigenvalues , @xmath42 , @xmath43 and @xmath44 , and the orientation of the halo is determined by the corresponding eigenvectors .
in this section , we investigate two distinct methods to find the filaments in the simulation : one using the mass density field and the other using the distribution of dark matter halos .
the first method we examine , hereafter referred to as method slowromancap1@ , is based on the local hessian matrix @xmath45 of the smoothed mass density field , defined as @xmath46 where @xmath47 is the smoothed density field . @xmath37 and @xmath48 denote the hessian matrix indices with values of 1 , 2 or 3 .
thus , at the location @xmath49 of each halo we can quantify the local `` shape '' of the density field by calculating the eigenvalues of the hessian matrix ( e.g. , aragn - calvo et al .
the number of positive eigenvalues of @xmath50 can be used to classify the possible environments in which a halo can reside into four regions , according to : * @xmath51 : a region with no positive eigenvalue ; * @xmath52 : a region with one positive and two negative eigenvalues ; * @xmath53 : a region with two positive and one negative eigenvalues ; * @xmath54 : a region with three positive eigenvalues . the direction of a filament can be identified with the eigenvector corresponding to the single positive eigenvalue of the hessian matrix in a filament region .
we note that this method has only one free parameter , namely the smoothing scale @xmath55 of the gaussian filter employed in constructing the density field .
different from the multi - scale morphology filter ( mmf ) method discussed in aragn - calvo et al .
( 2007a ) , we follow hahn et al .
( 2007a ) and adopt a fixed smoothing length of @xmath56 .
hahn et al . ( 2007a ) pointed out that the relative volume fractions of the four categories depend on the choice of @xmath55 .
for @xmath56 ( corresponding to halo masses of @xmath57 ) they find a good agreement between their orbit - stability criterion and a visual classification of the large - scale structure .
, which is derived from the matter density distribution via poisson s equation , @xmath58 where the poisson s equation is solved using a fast fourier transform on @xmath3 grid cells . as a consequence of that , their eigenvalues have opposite signs compared to those based on the corresponding density field that we use here . ] in order to compute the hessian matrix @xmath50 , we first construct a continuous density field from the discrete distribution of particles in the n - body simulation , using the cloud - in - cell ( cic ) technique with a @xmath3 grid .
then the smoothed density field @xmath59 is produced by smoothing the cic generated density field @xmath60 with a spherically symmetric gaussian filter @xmath61 , @xmath62 where @xmath49 corresponds to the location of a given halo and the gaussian filter with smoothing scale , @xmath55 , is given by @xmath63 from equations ( [ hessianmatrix ] ) , ( [ smoothdensity ] ) and ( [ gassianfilter ] ) , we find @xmath64 \nonumber\\ & & \rho_{\rm cic}(\boldsymbol{y } ) g_{r_s}(\boldsymbol{y},\boldsymbol{x}),\end{aligned}\ ] ] where @xmath65 is the kronecker delta ( aragn - calvo et al .
finally , the eigenspace structure of the symmetric hessian matrix can be computed at the center of mass of each halo . according to the number of positive eigenvalues at the locations of the dark matter halos we classify them into four categories , as outlined above .
the numbers of halos in cluster , sheet , filament and void regions are , respectively , @xmath66 , @xmath67 , @xmath68 and @xmath69 , corresponding to @xmath70 , @xmath71 , @xmath72 and @xmath73 per cent of the total number of halos . in the left panel of fig .
[ fig : fila1 ] , we show the distribution of halos in a slice of thickness @xmath14 , indicating the classification according to their environments with different colors . in addition , the directions of the filaments are marked with arrows for those halos associated with them .
the right panel of fig .
[ fig : fila1 ] shows the environmental classification of @xmath74 grid point put down regularly in the mid - plane of the slice . according to visual inspection , the general appearance of the filamentary structures identified with this method is remarkably good .
as shown in the previous section , filamentary structures can be well identified using the hessian matrix of the density field . however , this approach relies on an accurate knowledge of the matter density field , which is in many cases highly non - trivial to obtain .
in particular , for observational data one may wish to use the distribution of galaxies , groups or halos directly to identify the filaments of the cosmic web .
for this purpose , we now present a halo - based method for filament finding which is based on a slightly modified version of the ` candy model ' proposed by stoica et al .
the candy model reconstructs filaments by connecting individual segments that are found in a basic point distribution ( galaxies , halos , etc . ) . in this study , we only aim to compare halo shape orientations with the orientations of the segments they are residing in , thus we do not discuss in detail the problem of composing individual segments into long connected filaments .
hereafter , the segment extraction method derived from the candy model will be referred to as method slowromancap2@. in brief , there are two fundamental differences between method slowromancap1@ and slowromancap2@. one is the use of a biased tracer of the density field in method slowromancap2@. the other is that halos in method slowromancap2@ are automatically grouped into given filaments .
that is , we know what halo belongs to what filament . the requirements which a group of points has to fulfill to be considered a candidate segment have to be adjusted to the problem in question .
here we focus on dark matter halos as building blocks for the filamentary structure of the cosmic web .
therefore , a candidate segment is assumed to be a cylinder , with a length in the range of @xmath75 $ ] and a radius in the range @xmath76 $ ] .
the mean mass density within the segment should be at least @xmath77 times that of the average mass density of all halos , @xmath78 , where @xmath79 is the total mass of the halos with more than @xmath25 particles and @xmath80 is the volume of the simulation .
finally , a segment should have at least @xmath81 member halos .
we set these free parameters as follows : @xmath82 note that the values chosen for these parameters are somewhat arbitrary , but the results are robust to substantial changes in these parameters .
in particular , we tested a number of different reasonable sets of parameter values , all of them led to general agreement between the results . specifically , we found that changing any of the above listed parameters ( except for @xmath77 ) by 50% will only result in a less than 5% change in the number of halos identified as belonging to filaments .
if @xmath77 is varied to @xmath83 or @xmath84 , about 59% or 31% of the halos , respectively , are classified as belonging to filaments .
even for these two cases , the final results do not change qualitatively , however . in the following
we describe the successive steps of the segment extraction method .
* step 1 . * the dark matter halos are ranked according to their masses .
* step 2 . *
starting from the most massive halo to ever smaller ones , we search around each halo @xmath32 ( node ) all other halos @xmath85 with distance in the range [ @xmath86 .
these halo pairs form candidates for our filament segments .
* step 3 . *
we calculate for each candidate segment the filament strength ( average mass density ) , @xmath87 , which is defined as @xmath88 here , @xmath89 is the length of the candidate segment ( distance between the two halos @xmath32 and @xmath85 ) , @xmath28 is the total number of halos within the cylinder around this segment with radius @xmath55 , and @xmath90 is the mass of halo @xmath91 .
the candidate segment radius , in the range @xmath92 $ ] , is determined such that @xmath93 reaches its maximum .
* step 4 . * for all the candidate segments , we rank their strengths @xmath94 .
the one with the largest strength , with at least @xmath81 member halos and @xmath95 , is defined as the first candidate segment associated with node halo @xmath32 .
* we rank all the member halos within that segment according to their masses , except the node halo @xmath32 .
the most massive one @xmath96 with at least @xmath97 halos residing between halos @xmath32 and @xmath96 is defined as the second , terminal node of the segment .
we remove those member halos that are between the two node halos ( belonging to segment @xmath98 ) from the halo list .
thus member halos can only belong to one segment , while node halos can belong to more than one segment .
the direction of the segment ( filament ) is defined to be the connecting line between the two node halos
. * step 6 .
* once the segment @xmath99 is determined , we return to step 1 and search for other segments associated with node halo @xmath32 until no further segments can be found for this node halo . as an illustration
, we show in fig . [
fig : test ] the halo distribution in a @xmath100 cube centered on the most massive halo @xmath101 , together with the node and member halos in segment @xmath99 which are marked with colored dots .
the first four segments associated with node halo @xmath32 ( 6 in total ) are illustrated using colored dots in the four panels of fig .
[ fig : test ] . * step 7 .
* we turn to the next node halo to search for its associated segments .
we iteratively search the halo catalog until no further segments can be found .
applying the segment extraction method to our halo catalog we find that 45% of all halos are classified as members of segments , and 2% as node halos .
these two populations are substantially smaller than those obtained with method slowromancap1@ , where we found 63% and 19% to be classified as filament and cluster halos , respectively .
note , however , in a recent paper forero - romero et al .
( 2008 ) introduced an additional threshold @xmath102 for the classification of environment instead of just using the number of positive ( negative ) eigenvalues . increasing this barrier from @xmath103 to some finite value
greatly reduces the volume ( mass ) occupied by the identified filaments , and in particular , reduces the number of ambiguous detections of feeble or spurious filaments .
the discrepancy may hence just be a result of the higher detection threshold of method slowromancap2@ compared with method slowromancap1@. as a more quantitative comparison , in fig .
[ mass_function ] , we show the mass functions of filament halos derived with method slowromancap1@ ( red dashed ) and method slowromancap2@ ( green dotted ) .
the mass functions obtained from the two methods are very similar , especially at the high - mass end . while at the low mass end , because of the higher detection threshold of method slowromancap2@ , less halos are specified as belonging to the filaments . for reference
, we also show the total mass function from our simulation ( blue dash - dotted ) , which is in good argeement with the analytic model prediction by sheth et al .
( 2001 ; smt ) ( black solid ) . the right panel of fig .
[ fig : fila1 ] shows the distribution of halos within the filaments obtained with method slowromancap2@ in a slice of thickness @xmath14 .
compared to the filaments detected by the hessian matrix approach in the left panel , only strong filaments are extracted by this method , while less prominent structures are not identified . yet , the advantage of this method is that it does not require knowledge of the density field .
this is a significant advantage especially in the context of observations , where density reconstruction is often difficult .
to quantify the impact of filaments and sheets onto halos , we investigate two types of alignment signals : one based on a halo s spin and the other on its orientation .
these two vectors correlate with the orientation of the filament or the normal vector of the sheet in which the halo resides .
finally , we also compare the results of the two filament - finding methods considered here . for simplicity
, we will often use the terms ` shape ' and ` spin ' , as well as ` filament ' and ` sheet ' , as an indication of direction in the obvious sense . in order to quantify the spin and shape orientations of halos relative to filaments , we compute the probability distribution function @xmath104 ( alignment signal ) , where @xmath105 is the angle between the orientations of halo and filament or sheet .
@xmath106 here , @xmath107 ) is restricted to the range @xmath108 $ ] , and @xmath109 implies that the halo orientations are parallel to the filaments while @xmath110 indicates perpendicular orientations . for the null hypothesis of random orientations of halos relative to the environment one
expects @xmath111 . to assess the poisson sampling errors in our alignment signals , we generate 500 random samples in which the orientations of filaments are kept fixed , but the halo spin or major axes orientations are randomized . for each of these samples
we compute the @xmath112 , which we use to compute the significance of any detected alignment signal . lines . ] in addition , we also calculate the mean cosine @xmath113 of the alignment angle . in the absence of any alignment , @xmath114 . the significance of any alignment can be assessed in terms of @xmath113 and @xmath115 , which is the standard deviation of @xmath116 for the 500 random samples .
we first examine possible alignment signals between the spin of halos and the filaments they reside in .
method slowromancap1@ , which is based on the hessian matrix , allows to distinguish cluster , sheet , filament and void halos . for filament halos ,
the eigenvector of the positive eigenvalue of the hessian matrix at the location of the halo indicates the direction of the filament .
method slowromancap2@ determines the direction of the filament for each member halo in a segment as the connecting line between the two terminal node halos of that segment .
[ spin_filament ] shows the probability distribution @xmath104 of the cosine of the angle between the halo spins and the filaments .
the left and right panels display the results for methods slowromancap1@ and slowromancap2@ , respectively .
we find that the halo spins tend to lie parallel to the filaments .
the results are robust against the filament detection method .
it has been argued that the angular momentum of halos originates from the tidal field exerted by the surrounding dark matter distribution ( peebles 1969 ; doroshkevich 1970 ; white 1984 ) .
the spin directions are expected to be preferentially aligned with the planes of the sheets and the directions of the filaments ( e.g. , lee 2004 ) , although hydrodynamical simulations have also suggested that the spin axis may aline with the intermediate axis at turnaround ( navarro et al .
2004 ) . in order to characterize the strength of the alignment between angular momenta and filaments we calculate their average cosine , @xmath117 .
these average values together with their errors are displayed in the panels of fig .
[ spin_filament ] .
although the alignment signals for the two methods are slightly different , the overall strength of the alignment detected with methods slowromancap1@ and slowromancap2@ agrees well .
another interesting question is whether the alignment signal and strength depend on the mass of the filament halos .
[ spin_massbin ] presents the probability distribution @xmath104 for filament halos in different mass ranges .
results are shown for methods slowromancap1@ and slowromancap2@ in the upper and lower panels , respectively .
the overall alignment signals obtained from the two methods are very similar .
there is a weak mass dependence , in the sense that the alignment is somewhat weaker for massive halos . according to the values of @xmath113 , filament halos with masses @xmath118 are consistent with being randomly oriented at a 2-@xmath119 confidence level . in a recent study , hahn et al .
( 2007b ) reported that the spin of halos with mass greater than the characteristic halo mass tends to lie perpendicular to the host filaments .
this trend has been confirmed by aragn - calvo et al .
( 2007b ) . in a study comparing simulations and observations paz et al .
( 2008 ) found an indication of this behaviour based on sdss data .
the characteristic mass for gravitational collapse at redshift @xmath15 is @xmath120 , calculated for the cosmological parameters used in our simulation . due to the small box size of our simulation , we find less than 100 filament halos with mass @xmath121 , independent of the filament finding method applied .
most of these massive halos are classified as clusters or node halos . owing to the sparse number of filament halos , the statistics is too poor to obtain a robust measurement of the alignment signal , especially with respect to a possible transition from alignment to anti - alignment
. however , for the well constrained halo mass ranges @xmath122 , our alignment signals and strengths are in very nice agreement with those found by hahn et al .
( 2007b ) , and slightly larger than those predicted by aragn - calvo et al .
( 2007b ) . in method slowromancap2@ ,
the filaments are defined via segments extracted from the distribution of dark matter halos .
associated with each segment , there are two node halos , one of which is the most massive one among all the associated halos .
thus we can probe the alignment signals separately for halos in filaments with different most massive node halos . in fig .
[ spin_masstop ] , we show the results for four mass bins .
there is a hint for a very weak positive mass dependence on the most massive halo in the segment , but this is statistically not significant .
in addition to the mass dependence itself , we can further investigate the alignment signals at different separations to the most massive halo .
the solid line in fig .
[ spin_distance ] displays the alignment signals , the mean cosine of the angle between the spin and filament , for halos at different distances .
we find that the strength of the alignment is slightly suppressed if the halos are either very close to or far away from the most massive node halos .
to better understand this , we also measure the alignment signals for the mean cosine of the angle between spin and direction of the most massive node halo .
the results are shown in fig .
[ spin_distance ] as the dotted line . at large separation , since the direction of the most massive node halo and the filament direction is almost parallel , the two kinds of alignment signals are very similar . at very small separation , however , the spin and direction of the most massive halo show opposite alignment signal compare to that between the spin and filament .
this feature clearly indicates a _ transition _ of the 2-d collapse phase of the filaments to the 3-d collapse phase of the cluster / node halos at small separation .
next , we probe another important structural parameter , the orientation of the halo shape with respect to the direction of the filament .
similar to the last section , we measure the alignment signals between the shapes and filaments for filament halos . in fig .
[ major_filament ] , we show the probability distribution of the cosine of the angle between the halo major axis and the direction of the filament .
the left and right panels show the results for methods slowromancap1@ and slowromancap2@ , respectively .
we find significant alignment signals with both methods .
in fact , the shapes of dark matter halos tend to be parallel to the filaments .
again , we use the average value of @xmath113 to quantify the strength of the alignment signal . from method slowromancap1@ , we obtain an average cosine of @xmath123 , whereas method slowromancap2@ results in a slightly smaller value of @xmath124 .
similar alignment trends are reported in other recent studies ( e.g. , altay et al .
2006 ; aragn - calvo et al .
2007b ; hahn et al .
2007b ) . in analogy to our investigation of spin - filament alignment
, we now examine the dependence of the alignment strength on mass and separation .
[ major_massbin ] shows the results for the filament halos in three mass bins .
the upper and lower panels display results derived with methods slowromancap1@ and slowromancap2@ , respectively .
the strength of the alignment grows significantly with halo mass .
interestingly , the observed mass dependence shows an opposite trend compared to the spin - filament alignment .
these trends agree well with results obtained by hahn et al .
( 2007b ) and aragn - calvo et al .
( 2007b ) . in fig .
[ major_masstop ] , we show the alignment signals for halos in segments with most massive halos in four mass bins ( note that this can only be done for method slowromancap2@ ) .
an obvious mass dependence of the alignment signals is visible .
halos in segments with more massive node halos exhibit stronger alignment signals .
the solid line in fig .
[ major_distance ] displays the alignment signals between shape and filament for halos at different separations to the most massive node halo .
one can see a pronounced distance dependence .
halos at smaller separations to the most massive node halos , except at the smallest distance bin , tend to have stronger alignment strength .
similar to the spin of the halos , we also measure the alignment signals between shape and direction of the most massive node halo .
the results are shown in fig .
[ major_distance ] as the dotted line . obviously in this measure the distance dependence is much enhanced and _
monotonic_. again , this feature indicates the _ transition _ of the 2-d collapse phase of the filaments to the 3-d collapse phase of the cluster / node halos at small separation .
this distance dependence , if not restricted to the filament members , is in general agreement with the alignment signals measured by faltenbacher et al .
( 2008 ) for central and satellite halos as a function of radius and pereira et al .
( 2008 ) for substructures with their host halos .
method slowromancap1@ differentiates between four cosmic environments : clusters , filaments , sheets and voids .
filaments are distinguished by the condition that their hessian matrix has only one single positive eigenvalue , and the corresponding eigenvector determines a unique direction .
sheets on the other hand are defined by having only one single negative eigenvalue .
the associated eigenvector also determines a unique direction , which can be identified with the normal to the sheet .
the alignment signal for the angle between the halo spin and the normal of the sheet is shown in the left panel of fig .
[ sheet ] .
we obtain an anti - alignment signal , which means that there is a trend for sheet halos to have their angular momentum vector parallel to the plane of the sheet .
the alignment strength , quantified by the average of the cosine is @xmath125 .
this alignment strength is in very good agreement with that obtained by aragn - calvo et al .
( 2007b ) .
the right panel of fig .
[ sheet ] shows the probability distribution of the cosine of the angle between the halo major axis vector and the vector perpendicular to the sheet .
the major axes of halos in sheets are strongly aligned with the sheet planes .
the average of the cosine is @xmath126 .
a similar tendency has also been found in the shells of voids , as reported by brunino et al .
( 2007 ) .
having specified the spin and shape ( major axis ) alignments with respect to the large - scale environments ( filaments and sheets ) , we now proceed to examine the spin - shape alignment within the halos themselves . fig .
[ spin_major ] shows the alignment signals between the spin - major axes ( left panel ) , spin - middle axes ( middle panel ) and spin - minor axes ( right panel ) , respectively .
measurements of these signals in different environments are displayed with different line styles : results are given for clusters ( orange dotted ) , filaments ( red dash - dotted ) , sheets ( blue dashed ) , voids ( green dash - dot - dot ) and for all halos ( black solid ) .
in each case , the average of the cosine of the alignment angle is also indicated in each panel .
we find no significant environmental dependence of the alignment signal between the spin and shape of halos .
the halo spin vector appears to be preferentially perpendicular to the halo major axis , and has a strong tendency to be parallel to the halo minor axis .
this behavior is in good agreements with previous findings obtained , e.g. , by faltenbacher et al .
( 2002 ) , bailin & steinmetz ( 2005 ) ; allgood et al .
( 2006 ) and bett et al .
we have also separately investigated the alignment signal for halos with mass @xmath127 and @xmath20 , and basically found no dependence on mass besides a marginal enhancement of the alignment signal for the @xmath17 halos . according to our analysis above , both the spin and the major axes of halos
are preferentially aligned with the directions of the filaments or the planes of the sheets . on the other hand , within halos
the spin axes are strongly aligned with the _ minor axes _ of the halos . at first glance , these two results seem contradictory . in case of a perfect alignment between halo spin and minor axes , an alignment between halo orientations and filaments would cause an anti - alignment between spin axes and filaments . to have a better understanding of these two sets of ` contradictory ' results , we perform the following additional test .
we first rank all the ( member ) halos within the filaments according to the cosine of their angles between the major axes and the filament directions , @xmath128 , here we use @xmath129 to represent the angle between the major axis and the filament directions . ] .
these halos are then split into four bins with equal numbers according to the values of @xmath128 .
we then measure the cosines of the angles between the spin axis and the filament directions for the halos in each @xmath128 bin .
the results are shown in fig .
[ spin - shape ] , with upper panels for method slowromancap1@ and lower panels for method slowromancap2@. each panel in fig .
[ spin - shape ] from left to right corresponds to results for 1/4 of the halos in the filaments within a different @xmath128 range , as indicated .
the most right hand panels of this figure clearly demonstrate that a _ strong _ alignment between the major axes and the directions of the filaments produces an anti - alignment between the spin axes and the filaments .
this result is in agreement with simple geometric considerations .
note however , theoretically a _
strong _ anti - alignment between the major axes and the directions of the filaments , as addressed in the left hand panels of fig .
[ spin - shape ] , _ does not _ guarantee an alignment between the spin axes and the filaments .
in the case of @xmath130 and perfect alignment between minor axis and spin , the angle between the spin ( minor ) axis and the filament direction can assume any value within @xmath131 .
the very strong alignment signals apparent in the most left hand panels must be induced purely by the influence of the filaments .
combining the signals from the four panels , one would obtain the alignment signals shown in fig .
[ spin_filament ] . on the other hand , note the range of the @xmath128 in each panel , the average @xmath128 for all halos in the four panels is larger than @xmath132 , again indicating an alignment between the major axis and filament directions ( as shown in fig .
[ major_filament ] ) .
thus we conclude that the two sets of alignments are indeed not contradictory , and that the large - scale environment , i.e. the filaments and the sheets , can impact both the orientation and the spin of halos while allowing still for an internal correlation of these quantities within halos . throughout the paper
we find very good agreement between various alignment signals for methods slowromancap1@ and slowromancap2@ , however it is not necessory that the same halos in the two methods have the same filament directions . in order to make a more profound understanding of the agreement results of the two methods ,
we compute the probability distribution @xmath133 of the angle between the directions of the filaments from method slowromancap1@ and the filament vectors from method slowromancap2@. as shown in fig .
[ fila1-fila2 ] , the filament directions for the same halos in the two methods agree very well and more than 60% halos have @xmath134 . in addition
, we did not find any significant halo mass dependence for these filament angles .
as we have also checked the local density and distance to the most massive node halo of each common halo , we do not find any significant dependence on this angle @xmath105 .
using the dark matter and halo distributions from a high resolution n - body simulation , we have identified the filamentary structures in the cosmic web using two different filament - finding algorithms .
the first method is based on the hessian matrix of the density field ( method slowromancap1@ ) , where the halos are classified into four categories according to the signs of the three eigenvalues of the hessian matrix : clusters , filaments , sheets and voids .
the advantage of this method is that the large - scale environments of the halos can be characterized unambiguously , and that there is only one free parameter in the analysis .
however , this method needs detailed information about the local density field , which is difficult but not impossible to obtain based on observations .
indeed , some recent studies are devoted to density field reconstructions based on galaxy and halo distributions ( e.g. erdogdu et al .
2006 from 2mass redshift survey ; and wang et al .
2009a from halo distribution in simulations ) .
on the other hand , method slowromancap2@ directly uses the distribution of halos .
thus it eliminates the need for a reconstruction of the local mass density field , at the price of a larger number of tunable parameters .
filamentary structures are here traced by connecting single segments which are identified according to their relative mass over - density . in this study
we were not interested in explicitly reconstructing long coherent filaments , rather we only explore the orientations of the segments relative to the halos orientations associated with them .
the main shortcomings of method slowromancap2@ are the relatively large number of parameters and a lack of sensitivity for the detection of less prominent filaments . nevertheless , method slowromancap2@ is easy to implement on top of observational data , including also galaxy group catalogs such as that of yang et al .
this makes it a highly useful approach in practice , provided its results are consistent with those obtained with method slowromancap1@. this is indeed the case , as we have demonstrated in this study .
moreover in method slowromancap2@ , halos are automatically grouped into filaments , and we know what halo belongs to what filament .
based on the classification of the large - scale environment around halos that we obtained , we examined the spin - filament and the shape - filament alignments , as well as the corresponding quantities for halos and sheets .
overall , we find a tendency for halo spins and shapes to be oriented parallel to filaments and perpendicular to the normal vector of sheet like structures .
this trend becomes obvious and is clearly significant when compared to results from isotropic ( randomized ) orientations .
both the spin and shape alignment strengths are stronger for halos close to more massive node ( cluster ) halos , and at smaller separations to the node halos . yet , the spin and shape alignment strengths show an opposite mass dependence : smaller halos display stronger ( weaker ) spin ( shape ) alignments with the filaments . for the halos in different environments , we have also investigated their intrinsic spin - shape alignment .
we found that the spin axes are preferentially aligned with the minor axes of halos , and are perpendicular to the major axes , independent of the halo environments . at first glance , this result seems to be at odds with our finding that _ both _ the spin and major axes tend to align with the filament directions . however , there is not really a contradiction here , because the alignments _ are not perfect _ in both cases .
in fact , the distributions of the alignment angles are sufficiently broad that both of these seemingly contradictory alignment signals can coexist .
our results are in good agreement with recent n - body studies where different filament finding methods have been employed .
hahn et al . ( 2007b ) applied a hessian matrix approach to the gravitational potential field ( instead of the density field ) , and also found an opposite mass dependence for the alignment strengths of spin- and shape - filament alignments .
moreover , they found that the spin axes of halos with mass @xmath121 are preferentially perpendicular to the filament directions , which we can not confirm in this study because of the limited volume of our simulation .
a similar behavior has been detected by aragn - calvo et al .
( 2007b ) using a multi - scale morphology filter method for the classification of the large - scale environments . quite interestingly
, we find in method slowromancap2@ according to the spin and shape alignment signals of the halos in the filaments , there is a transit influence scale of large scale environments : from the 2-d collapse phase of the filament to the 3-d collapse phase of the cluster / node halo at small separation to the most massive node halo .
the general trends found from our n - body simulation reveal a substantial interplay between the large - scale environments and the internal properties of the dark matter halos
. it would be interesting to see whether similar trends can be observed in real galaxy samples ( e.g. , lee & pen 2001 ; lee & erdogdu 2007 ) .
the two methods outlined in this study should be straightforwardly applicable to observational data sets ( e.g. sdss ) .
this should provide for interesting tests of galaxy formation models , and of cold dark matter cosmologies .
we are grateful to the anonymous referee for useful and insightful comments that greatly helped to improve the presentation of this paper .
the simulation was done at shanghai supercomputer center .
this work was supported by 973 program ( nos .
2007cb815401 , 2007cb815402 ) , 863 program ( 2006aa01a125 ) , the cas knowledge innovation program ( grant no .
kjcx2-yw - t05 ) and grants from nsfc ( nos .
10533030 , 10633049 , 10821302 , 10873027 , 10925314 ) . | clusters , filaments , sheets and voids are the building blocks of the cosmic web .
forming dark matter halos respond to these different large - scale environments , and this in turn affects the properties of galaxies hosted by the halos .
it is therefore important to understand the systematic correlations of halo properties with the morphology of the cosmic web , as this informs both about galaxy formation physics and possible systematics of weak lensing studies . in this study
, we present and compare two distinct algorithms for finding cosmic filaments and sheets , a task which is far less well established than the identification of dark matter halos or voids .
one method is based on the smoothed dark matter density field , the other uses the halo distributions directly .
we apply both techniques to one high resolution n - body simulation and reconstruct the filamentary / sheet like network of the dark matter density field .
we focus on investigating the properties of the dark matter halos inside these structures , in particular on the directions of their spins and the orientation of their shapes with respect to the directions of the filaments and sheets .
we find that both the spin and the major axes of filament - halos with masses @xmath0 are preferentially aligned with the direction of the filaments .
the spins and major axes of halos in sheets tend to lie parallel to the sheets .
there is an opposite mass dependence of the alignment strengths for the spin ( negative ) and major ( positive ) axes , i.e. with increasing halo mass the major axis tends to be more strongly aligned with the direction of the filament whereas the alignment between halo spin and filament becomes weaker with increasing halo mass .
the alignment strengths as a function of distance to the most massive node halo indicate that there is a transit large scale environment impact : from the 2-d collapse phase of the filament to the 3-d collapse phase of the cluster / node halo at small separation .
overall , the two algorithms for filament / sheet identification investigated here agree well with each other .
the method based on halos alone can be easily adapted for use with observational data sets . |
-5 mm dans ce texte , @xmath0 dsigne toujours une varit close et oriente .
les champs dhyperplans tangents quon considre sur @xmath0 sont coorients , donc aussi orients puisque @xmath0 lest .
un tel champ @xmath1 est le noyau dune forme @xmath2 , appele _ quation _ de @xmath1 , unique multiplication prs par une fonction positive .
on dit que @xmath1 est une _ structure de contact _
si @xmath3 induit sur @xmath1 en tout point une forme symplectique directe , _
i.e. _ si @xmath0 est de dimension impaire @xmath4 et si @xmath5 est en tout point un lment de volume direct pour lorientation de @xmath0 .
dautre part , un _
livre ouvert _
dans @xmath0 est un couple @xmath6 form des objets suivants : * une sous - varit close @xmath7 de codimension deux fibr normal trivial ; * une fibration @xmath8 qui , dans un voisinage @xmath9 de @xmath10 , concide avec la coordonne angulaire normale .
on peut aussi voir les livres ouverts autrement .
soit @xmath11 un diffomorphisme dune varit compacte gal lidentit prs du bord @xmath12 .
sa suspension , savoir la varit compacte @xmath13 \bigr ) \big/ \mathord\sim , \quad \text{o\`u \ $ ( p,1 ) \sim ( \phi(p),0)$,}\ ] ] est borde par @xmath14 car @xmath15 et la varit close @xmath16 possde un livre ouvert vident .
en outre , tout livre ouvert @xmath6 dans @xmath0 identifie @xmath0 @xmath17 , o @xmath18 est une fibre de @xmath19 ( un peu rtrcie ) et @xmath20 lapplication de premier retour sur @xmath18 dun flot transversal aux fibres de @xmath19 et constitu , prs de @xmath21 , de rotations autour de @xmath21 .
le diffomorphisme @xmath20 , dfini seulement conjugaison et isotopie prs , est la _ monodromie _ de @xmath6 .
toute la discussion venir tourne autour de la dfinition suivante : [ d : porteur ] _ une structure de contact @xmath1
sur @xmath0 est dite _ porte _ par un livre ouvert @xmath6 si elle admet une quation @xmath2 ayant les proprits suivantes : _
* @xmath2 induit sur @xmath21 une forme de contact ; * @xmath3 induit sur chaque fibre @xmath18 de @xmath19 une forme symplectique ; * lorientation de @xmath21 dfinie par la forme de contact @xmath2 concide avec son orientation comme bord de la varit symplectique @xmath22 .
une telle forme @xmath2 sera dite _ adapte _ @xmath6 .
[ x : porteurs ] soit @xmath23 une fonction holomorphe ayant lorigine un point critique isol et soit @xmath24 lhypersurface ( singulire ) @xmath25 .
il existe une boule ferme lisse @xmath26 autour de lorigine dans @xmath27 et un feuilletage de @xmath28 par des sphres strictement pseudoconvexes @xmath29 , o @xmath300,1]$ ] et @xmath31 , tels que , pour @xmath32 assez petit , les proprits suivantes soient satisfaites : * la sphre @xmath29 est transversale @xmath24 , de sorte que @xmath33 est une sous - varit close de @xmath29 de codimension deux et fibr normal trivial ; * lapplication @xmath34 est une fibration qui fait de @xmath6 un livre ouvert ; * le livre ouvert @xmath6 porte la structure de contact sur @xmath29 dfinie par le champ des tangentes complexes . autrement dit , chaque livre ouvert donn dans la sphre par le thorme de fibration de j. milnor porte , isotopie prs , la structure de contact standard .
-5 mm en dimension trois , divers travaux ont depuis longtemps fait apparatre des connivences entre les structures de contact et les livres ouverts sans toutefois tablir aucun lien formel
. dans @xcite , w. thurston et h. winkelnkemper construisent des formes de contact sur toute varit close @xmath0 partir dun livre ouvert dans @xmath0 .
avec les termes de la dfinition [ d : porteur ] , ils dmontrent en fait que tout livre ouvert dans @xmath0 porte une structure de contact .
dans @xcite dautre part , pour transformer en thorme de gomtrie de contact son rsultat sur les tresses fermes , d. bennequin met en vidence la proprit suivante : toute courbe transversale la structure de contact standard @xmath35 dans @xmath36 structure dquation @xmath37 est isotope , parmi les courbes transversales , une tresse ferme cest - - dire une courbe transversale au livre ouvert form par laxe des @xmath38 et la coordonne angulaire @xmath19 . or cette proprit vient de ce que ce livre ouvert porte @xmath35 .
enfin , dans @xcite , i. torisu a clairement dgag les relations entre les livres ouverts et les configurations de thorie de morse considres dans @xcite pour tudier les structures de contact _ convexes _ au sens de @xcite . la premire observation qui montre ltroitesse des liens imposs par la dfinition [ d : porteur ] et dcoule de la stabilit des structures de contact est la suivante : [ p : isotopie ] sur une varit close de dimension trois , toutes les structures de contact portes par un mme livre ouvert sont isotopes .
quant la question de savoir quelles structures de contact possdent un livre ouvert porteur , la rponse est simple : [ t : existence3 ] sur une varit close de dimension trois , toute structure de contact est porte par un livre ouvert .
cependant , comme lillustre lexemple des fibrations de milnor , le livre ouvert qui porte une structure de contact donne est loin dtre unique mme isotopie prs .
pour apprhender ce phnomne , quelques dfinitions sont utiles .
soit @xmath39 une surface compacte bord et @xmath40 un arc simple et propre . on dit quune surface compacte @xmath41 sobtient partir de @xmath18 par le _ plombage positif _
_ ngatif _ ) dun anneau
le long de @xmath42 si @xmath43 o @xmath44 est un anneau ayant les proprits suivantes : * @xmath45 est un voisinage rgulier de @xmath42 dans @xmath18 ; * @xmath46 est inclus dans une boule ferme @xmath26 do nt lintersection avec @xmath18 est rduite @xmath45 et lenlacement des deux composantes de @xmath47 dans @xmath26 vaut @xmath48 ( resp .
@xmath49 ) .
un rsultat de j. stallings affirme que , si @xmath6 est un livre ouvert dans @xmath0 et si @xmath18 est ladhrence dune fibre de @xmath19 , alors , pour toute surface @xmath50 obtenue partir de @xmath18 par le plombage dun anneau , il existe un livre ouvert @xmath51 tel que @xmath52 soit le bord de @xmath50 et que @xmath50 soit ladhrence dune fibre de @xmath53 .
dans la suite , on dira que le livre @xmath51 et lentrelacs @xmath52 sont eux - mmes obtenus par plombage partir respectivement de @xmath6 et de @xmath21 .
en outre , on dira quun livre ouvert @xmath51 est une _ stabilisation _ dun autre @xmath6 sil sobtient partir de @xmath6 par une suite finie de plombages positifs .
[ t : unicite3 ] dans une varit close de dimension trois , deux livres ouverts quelconques qui portent une mme structure de contact ont des stabilisations isotopes . les thormes [ t : existence3 ] et [ t : unicite3 ] permettent de traduire nombre de questions sur les structures de contact en questions sur les livres ouverts , autrement dit sur les diffomorphismes des surfaces compactes bord . en ce sens , ce sont les analogues des thormes de s. donaldson @xcite sur les pinceaux de
lefschetz dans les varits symplectiques de dimension quatre .
ils admettent cependant , la diffrence de ceux - ci , des dmonstrations purement topologiques do nt on dcrit brivement les ides ci - dessous , aprs avoir introduit loutil essentiel . on supposera
le lecteur familier avec certaines notions de gomtrie de contact en dimension trois ( structures de contact vrilles / tendues , invariant de thurston - bennequin des courbes legendriennes , surfaces @xmath1-convexes ) .
on appelle _ cellule polydrale _
dans @xmath0 limage dun polydre convexe compact euclidien par un plongement topologique .
une telle cellule possde une structure affine induite par son paramtrage et son _ intrieur _ est , par dfinition , limage de lintrieur du polydre , cest - - dire de son intrieur topologique dans son enveloppe affine .
une _ cellulation polydrale _ de @xmath0 dsigne ici un recouvrement fini de @xmath0 par des cellules polydrales ayant les proprits suivantes : * les intrieurs des cellules forment une partition de @xmath0 ; * le bord de chaque cellule @xmath54 est une union de cellules @xmath55 et les inclusions @xmath56 sont affines ; * les cellules de dimension deux ( et moins ) sont lisses , _
i.e. _ sont les images de plongements lisses . les cellulations polydrales ont cet avantage sur les triangulations dtre trs faciles
subdiviser : toute subdivision dun sous - complexe se prolonge trivialement .
en outre , elles jouent un rle cl dans la dmonstration du thorme de reidemeister - singer donne dans @xcite , dmonstration qui sert de guide pour tablir le thorme [ t : unicite3 ] .
soit @xmath1 une structure de contact sur @xmath0 .
on construit dabord dans @xmath57 une _ cellulation de contact _ , cest - - dire une cellulation polydrale @xmath58 ayant les proprits suivantes : * chaque cellule de dimension @xmath48 est un arc legendrien * chaque cellule de dimension @xmath59 est @xmath1-convexe et linvariant de thurston - bennequin de son bord vaut @xmath49 * chaque cellule de dimension @xmath60 est contenue dans le domaine dune carte de darboux .
on paissit ensuite le @xmath48-squelette @xmath61 de @xmath58 en une surface compacte @xmath62 ( presque ) tangente @xmath1 le long de @xmath61 et on choisit un voisinage rgulier @xmath63 de @xmath61 assez petit pour que @xmath64 soit une surface proprement plonge dans @xmath63 .
quitte prendre @xmath63 plus petit , @xmath1 admet une quation @xmath2 vrifiant les conditions suivantes : * @xmath3 induit sur @xmath18 une forme daire ; * @xmath2 est non singulire sur @xmath12 et oriente @xmath21 comme le bord de @xmath22 .
dautre part , pour toute cellule @xmath54 de dimension @xmath59 , la proprit 2 ) dit que le bord de @xmath65 intersecte @xmath21 en deux points ( isotopie prs ) .
il en rsulte quil existe une fibration
@xmath66 ayant @xmath67 pour fibre .
quitte rogner @xmath63 , on peut supposer que @xmath63 est une union de fibres de @xmath19 sur lesquelles @xmath3 induit une forme daire .
il reste voir que @xmath1 est isotope , relativement @xmath63 , une structure de contact porte par @xmath6 .
le point cl est que @xmath1 est tendue sur @xmath68 et que @xmath69 est une surface @xmath1-convexe do nt le dcoupage est fourni par @xmath21 .
soit @xmath58 une cellulation de contact de @xmath57 .
on dira ici quun livre ouvert porteur @xmath6 est _ associ _ @xmath58 si , comme dans la dmonstration du thorme [ t : existence3 ] , lune des fibres de @xmath19 contient le @xmath48-squelette de @xmath58 et se rtracte dessus par une isotopie de contact . en imitant @xcite , on montre dabord que
tout livre ouvert porteur admet une stabilisation associe une cellulation de contact .
on se ramne ainsi considrer le cas de deux livres ouverts porteurs associs des cellulations de contact @xmath70 et @xmath71 en position gnrale .
daprs @xcite , @xmath70 et @xmath71 possdent une subdivision commune @xmath72 qui sobtient , partir de @xmath70 comme de @xmath71 , par des bissections . on dforme
alors @xmath72 , relativement lunion des @xmath48-squelettes de @xmath70 et @xmath71 , en une cellulation vrifiant les proprits 1 ) et 3 ) des cellulations de contact et ayant des @xmath59-cellules @xmath1-convexes .
il suffit ensuite de subdiviser le @xmath59-squelette de @xmath72 pour obtenir une cellulation de contact @xmath58 et on montre pour finir que le livre ouvert associ @xmath58 est une stabilisation de ceux associs @xmath70 et @xmath71 .
on discute maintenant quelques corollaires des thormes [ t : existence3 ] et [ t : unicite3 ] .
on rappelle dabord quun thorme de m. hilden et j. montesinos affirme que toute varit close @xmath0 de dimension trois est un revtement trois feuillets de la sphre @xmath73 simplement ramifi au - dessus dun entrelacs ( simplement signifie que le degr local aux points de ramification dans @xmath0 vaut deux ) .
on obtient le mme rsultat pour les varits de contact closes : [ c : revetements ] toute varit
de contact close de dimension trois est un revtement trois feuillets de la sphre de contact standard @xmath74 simplement ramifi au - dessus dun entrelacs transversal @xmath35 .
un autre corollaire concerne la dynamique des flots de reeb .
un _ flot de reeb _ sur une varit de contact est un flot qui prserve la structure de contact tout en lui tant transversal et en pointant du ct positif .
un exemple typique est le flot godsique sur le fibr cotangent unitaire dune varit riemannienne .
les flots de reeb dune structure de contact donne @xmath1 sont en bijection avec les quations de @xmath1 : toute forme @xmath2 correspond lunique champ de vecteurs @xmath75 qui engendre le noyau de @xmath3 et sur lequel @xmath2 vaut @xmath48 . en prenant une quation de @xmath1 adapte un livre ouvert porteur , on obtient : sur toute varit de contact close de dimension trois ,
il existe un flot de reeb qui admet une section de poincar - birkhoff , cest - - dire une surface compacte qui rencontre toutes les orbites , do nt lintrieur est transversal au flot et do nt chaque composante du bord est une orbite priodique . en fait , il nest
pas exclu que tout flot de reeb admette une telle section @xcite ( ceci impliquerait la conjecture de weinstein selon laquelle tout flot de reeb a une orbite priodique ) mais cest l un problme de nature diffrente , certainement inaccessible par des mthodes topologiques .
une question naturelle au vu des thormes [ t : existence3 ] et [ t : unicite3 ] est de savoir comment lire sur la monodromie de ses livres ouverts porteurs si une structure de contact est tendue , ou remplissable en un quelconque sens .
la seule rponse connue concerne les structures de contact _ holomorphiquement remplissables _ , cest - - dire ralisables comme champs des tangentes complexes au bord de varits de stein compactes .
le corollaire suivant prcise un rsultat de a. loi et r. piergallini : [ c : remplissage ] une structure de contact sur une varit close de dimension trois est holomorphiquement remplissable si et seulement si elle est porte par un livre ouvert do nt la monodromie est un produit de twists de dehn droite .
pour finir , on donne un corollaire de pure thorie des nuds .
on appelle ici _ entrelacs fibr _
dans @xmath0 tout entrelacs orient @xmath21 pour lequel il existe une fibration @xmath8 qui fait de @xmath76 un livre ouvert et induit sur @xmath21 lorientation prescrite .
lorsque @xmath0 est une sphre dhomologie , un thorme de f. waldhausen assure que cette fibration , si elle existe , est unique isotopie prs .
le rsultat suivant rpond une question pose par j. harer dans @xcite : [ c : entrelacs ] deux entrelacs fibrs quelconques dans une sphre dhomologie entire @xmath0 sobtiennent lun partir de lautre par une suite de plombages et de ( oprations inverses ) .
une trivialisation de @xmath0 tant choisie , les classes dhomotopie de champs de plans tangents @xmath0 sont repres par leur _ invariant de hopf _ , savoir lenlacement des fibres des applications @xmath77 correspondantes .
on considre alors un livre ouvert quelconque @xmath6 dans @xmath0 et on note @xmath51 un livre ouvert obtenu partir de @xmath6 par un plombage ngatif .
les trois observations suivantes dmontrent le corollaire : * toute structure de contact @xmath78 porte par @xmath51 est vrille
car lme de lanneau plomb est isotope une courbe legendrienne non noue dans @xmath79 do nt linvariant de thurston - bennequin vaut @xmath80 ; * linvariant de hopf de @xmath78 est suprieur dune unit celui des structures de contact portes par @xmath6 ( voir @xcite ) ; * si deux structures de contact vrilles ont le mme invariant de hopf , elles sont isotopes daprs @xcite et deux livres ouverts quelconques qui les portent ont donc des stabilisations isotopes . cet argument borne en outre par @xmath81 le nombre des ( d)plombages ngatifs ncessaires pour passer dun livre ouvert un autre , o @xmath82 dsigne la diffrence entre les invariants de hopf correspondants .
-5 mm en dimension suprieure trois , les livres ouverts porteurs de structures de contact ne sont pas quelconques : leurs fibres ont une structure symplectique invariante par la monodromie .
pour prciser ce point , quelques dfinitions sont utiles .
soit @xmath18 une varit compacte , bord @xmath12 .
une forme symplectique exacte @xmath83 sur @xmath67 est _ convexe linfini _ sil existe sur @xmath67 un champ de liouville ( champ de vecteurs @xmath83-dual dune primitive de @xmath83 ) qui est transversal toutes les hypersurfaces @xmath84 , @xmath850,1]$ ] , o @xmath86 $ ] est un voisinage collier de @xmath87 .
on dit en outre que @xmath88 est une _ varit de weinstein _ @xcite sil existe un tel champ de liouville qui , de plus , est le ( pseudo)gradient dune fonction de morse @xmath89 constante et sans points critiques sur @xmath21 .
lexemple typique de varit de weinstein est lintrieur dune _ varit de stein compacte_. on nomme ainsi toute varit complexe compacte @xmath18 qui admet une fonction strictement pluri - sous - harmonique @xmath90 constante et sans points critiques sur le bord .
la @xmath59-forme @xmath91 dfinit alors une structure symplectique .
il ressort en fait du travail de y. eliashberg @xcite que toute varit de weinstein est symplectiquement diffomorphe lintrieur dune telle varit de stein compacte .
si maintenant @xmath2 est une forme de contact adapte un livre ouvert @xmath6 , sa diffrentielle @xmath3 induit sur chaque fibre de @xmath19 une structure symplectique exacte convexe linfini .
celle - ci dpend du choix de @xmath2 mais sa compltion @xcite est bien dfinie isotopie prs .
le thorme de w. thurston et h. winkelnkemper et la proposition [ p : isotopie ] stendent alors ainsi en grande dimension : [ p : construction ] soit @xmath18 une varit compacte avec , sur @xmath67 , une forme symplectique exacte convexe linfini et soit @xmath11 un diffomorphisme symplectique gal lidentit prs de @xmath12 .
il existe alors sur @xmath17 une structure de contact porte par le livre ouvert vident . de plus , deux structures de contact portes par un mme livre ouvert et qui induisent sur ses pages des structures symplectiques ayant des compltions isotopes sont isotopes .
soit @xmath1 une structure de contact , @xmath2 une quation de @xmath1 et @xmath92 une structure presque complexe sur @xmath1 calibre par @xmath93 .
on note @xmath75 le champ de reeb associ @xmath2 et @xmath94 la mtrique riemannienne sur @xmath0 qui vaut @xmath95 sur @xmath1 et rend @xmath75 unitaire et orthogonal @xmath1 . en
termes lmentaires , le thorme principal de @xcite montre quil existe des constantes @xmath96 et des fonctions @xmath97 , @xmath98 , vrifiant les conditions suivantes : ( ici , @xmath103 et @xmath104 sont les parties respectivement @xmath92-linaire et @xmath92-antilinaire de @xmath105 . ) en termes plus parlants , les fonctions @xmath106 sont des sections approximativement holomorphes et quitransversales du fibr @xmath107 , o @xmath61 est le fibr hermitien trivial @xmath108 muni de la connexion unitaire dfinie par la forme @xmath109 .
les estimations ci - dessus entranent dabord que , pour @xmath110 , lensemble @xmath111 est une sous - varit et que la forme @xmath112 induite par @xmath2 sur @xmath113 est une forme de contact ( voir @xcite ) .
en effet , @xmath112 est non singulire pour @xmath114 assez grand puisque son noyau est gal au noyau de @xmath105 et que @xmath115 tandis que @xmath116 .
mieux , ces ingalits montrent que , pour @xmath114 grand , le noyau de @xmath112 est proche dun sous - espace @xmath92-complexe de @xmath1 si bien que @xmath117 y est non dgnre .
lobservation suivante est que lapplication @xmath118 est une fibration do nt les fibres sont transversales au champ de reeb @xmath75 en tout point o @xmath119 . pour le voir , on note que lestimation sur @xmath120 implique que @xmath121 ainsi , en un point @xmath100 o @xmath122 et pour @xmath114 assez grand , @xmath123 est proche de @xmath124 , _ i.e. _ est non nul et presque orthogonal @xmath125
. par suite , les sous - varits @xmath126 sont transversales au champ de reeb @xmath75 .
ces arguments montrent que le livre ouvert @xmath127 , pour @xmath114 assez grand , porte la structure de contact @xmath128 .
il reste vrifier que les fibres de @xmath19 sont des varits de weinstein .
pour simplifier , on prouve ci - dessous lassertion analogue en gomtrie symplectique .
[ p : weinstein ] soit @xmath63 une varit close , @xmath83 une forme symplectique entire sur @xmath63 et @xmath129 une sous - varit symplectique de @xmath63 en dualit de poincar avec @xmath130 et obtenue par la construction de donaldson , partir dun fibr hermitien en droites @xmath61 muni dune connexion unitaire de courbure @xmath131 .
pour @xmath114 assez grand , @xmath132 est une varit de weinstein . en reprenant les arguments de @xcite , on peut supposer que @xmath129 est le lieu dannulation dune section @xmath133 qui vrifie , en tout point de @xmath63 , @xmath134 dans la trivialisation de @xmath135 donne au - dessus de @xmath136 par la section unitaire @xmath137 , la connexion est dfinie par une @xmath48-forme @xmath138 o @xmath139 .
si on pose @xmath140 , lingalit ci - dessus donne @xmath141 ce qui montre que @xmath142 est plus loin de @xmath143 que de @xmath144 .
le champ de liouville dual de @xmath145 est alors un pseudogradient de @xmath146 .
comme en dimension trois , le livre ouvert porteur dune structure de contact donne nest pas unique .
on dcrit dans @xcite une opration de _ plombage le long dun disque lagrangien _ dans laquelle les twists de dehn - seidel viennent remplacer les twists de dehn qui permet dtablir des analogues du thorme [ t : unicite3 ] et du corollaire [ c : remplissage ] .
ces rsultats ramnent ltude des structures de contact celles des diffomorphismes symplectiques des varits de stein compactes qui sont lidentit prs du bord .
ils permettent peut - tre ainsi de rapprocher les travaux de y. eliashberg , h. hofer et a. givental sur la thorie symplectique des champs de ceux de , par exemple , de p. seidel sur lhomologie de floer et les groupes de diffomorphismes symplectiques .
on peut aussi se demander si le thorme [ t : existence ] cache des obstructions lexistence dune structure de contact sur les varits closes .
daprs @xcite , toute varit close @xmath0 de dimension @xmath4 possde un livre ouvert do nt chaque fibre a le type dhomotopie dun complexe cellulaire de dimension @xmath147 . il est probable que , si @xmath0 admet un champ dhyperplans tangents muni dune structure presque complexe , il existe un tel livre ouvert pour lequel chaque fibre est une varit presque complexe et est donc , daprs @xcite , lintrieur dune varit de stein compacte .
toute la difficult serait donc vraiment de raliser la monodromie par un diffomorphisme symplectique ... dans cet ordre dide , voici un corollaire concret du thorme [ t : existence ] obtenu par f. bourgeois et qui montre , en rponse une vieille question , que tout tore de dimension impaire possde une structure de contact : soit @xmath1 une structure de contact sur @xmath0 , soit @xmath2 une quation de @xmath1 adapte un livre ouvert porteur @xmath6 et soit @xmath149 un voisinage de @xmath87 dans lequel @xmath19 est la coordonne angulaire normale . on note @xmath32 la
coordonne radiale normale dans @xmath150 et on pose @xmath151 o la fonction@xmath152 vaut @xmath32 pour @xmath153 , @xmath48 pour @xmath154 et vrifie @xmath155 .
un calcul montre que , si on choisit @xmath156 assez petit , @xmath157 est une forme de contact sur @xmath148 . | on dcrit ici des relations entre la gomtrie globale des varits de contact closes et celle de certaines varits symplectiques , savoir les varits de stein compactes . lorigine de ces relations est lexistence de livres ouverts adapts aux structures de contact .
4.5 mm * 2000 mathematics subject classification : * 57m50 , 53d35 .
* mots cls : * structures de contact , livres ouverts .
la gomtrie de contact en dimension trois a connu un essor important durant la dernire dcennie grce au dveloppement de mthodes topologiques adquates .
dans le prolongement des travaux de d. bennequin @xcite et de y. eliashberg @xcite , la thorie des @xcite et ltude des rocades ( bypasses ) @xcite ont men une classification complte des structures de contact sur quelques varits simples et , plus rcemment , une classification grossire sur toutes les varits closes @xcite .
en fait , comme on essaiera de le montrer plus loin , les structures de contact en dimension trois sont des objets purement topologiques , un peu comme les structures symplectiques en dimension deux . en termes prcis , sur toute varit close @xmath0 de dimension trois , les classes disotopie des structures de contact se trouvent en correspondance bijective avec les classes disotopie et de stabilisation des livres ouverts dans @xmath0 , lopration lmentaire de stabilisation tant un plombage positif @xcite . en dimension suprieure , des mthodes radicalement diffrentes permettent de mettre en vidence une correspondance similaire @xcite et , au - del , de faire apparatre des liens troits entre la gomtrie globale des varits de contact closes et celle de certaines varits symplectiques compactes .
les livres ouverts quon associe une structure de contact sont en effet particuliers : leurs pages sont des varits de stein compactes , leur monodromie est un diffomorphisme symplectique support dans
lintrieur et lopration lmentaire de stabilisation qui les unifie est un plombage lagrangien positif .
en outre , loutil essentiel pour les construire est la thorie des fibrs positifs que s. donaldson a introduite et dveloppe en gomtrie symplectique dans
@xcite et qui a t adapte en gomtrie de contact dans @xcite . |
the quantum kicked rotor @xcite consists of a massive quantum particle moving freely on a circumference and periodically kicked by a cosine pulse that modifies its angular momentum .
this kind of dynamics is generated by the following time - dependent singular hamiltonian @xmath0 where @xmath1 is a coupling constant with the dimension of an action , @xmath2 is the momentum of inertia and @xmath3 the kicking period .
the classical and quantum kicked rotors have been deeply investigated in connection with classical chaos and its signature in quantum systems @xcite and have also been physically implemented @xcite .
further interest in this model came from the possibility of being used to investigate the anderson localization @xcite and transition from the metallic to the insulating regime @xcite .
the quantum kicked rotor localizes the system in angular momentum space when the period @xmath4 , with irrational @xmath5 , while in the anderson model an electron is localized on a @xmath6-dimensional lattice in presence of spatial noise .
there exists a procedure to map the kicked rotor into an anderson localization model and vice versa @xcite .
there exists another model , known as quantum linear kicked rotor @xcite , also known as _ maryland model _
, where the kinetic term in the hamiltonian is taken to be linear in angular momentum , @xmath7 this model has the advantage of being analytically solvable and can also be mapped into a model exhibiting anderson localization when @xmath8 , with irrational @xmath5 @xcite . in the following
, we shall consider the scenario in which the quantum linear kicked rotor is in contact with an external environment of classical type whose presence is effectively taken into account by treating the kick strengths as stochastic variables .
the ensuing decoherence weakens the destructive quantum interferences responsible for the angular momentum localization leading to a diffusive behaviour exhibited by the momentum variance increasing with the square root of time @xcite .
we shall focus upon the kick strengths forming a discrete - time markov process with varying degree of memory between the realization of the stochastic kick strength at a certain time and the successive one .
the resulting dynamics is described by a discrete family of completely positive maps that compose as a semigroup in absence of memory and are otherwise connected by intertwining maps that are not completely positive as witnessed by the non - monotonic behaviour of the hilbert - schmidt norm .
this discrete - time dissipative quantum process appears to be non - markovian according to the approach to non - markovianity through the absence of divisibility @xcite @xcite ( for another approach to non - markoviaity see @xcite and @xcite ) .
furthermore , in the present model it can be proved that memory effects make the intertwining maps not even positive .
as outlined in the introduction , we shall focus upon the following hamiltonian @xmath9 where the @xmath10 are stochastic variables taking values in @xmath11 that perturb the otherwise constant kick strength at discrete times @xmath12 ; we shall take them to form a one - step markov process @xmath13 that we will specify later on . in the following
, we shall frequently use angle representation ( for sake of simplicity @xmath14 ) , where @xmath15 and the operator @xmath16 has eigenvectors @xmath17 because of the dirac delta , the time - evolutor from @xmath18 and @xmath19 is given by the following unitary operator on the hilbert space @xmath20,{\rm d}\theta)$ ] of the system : @xmath21 let us consider an initial state , namely a density matrix @xmath22 acting on @xmath23 ; the state @xmath24 at discrete time @xmath25 is thus obtained as @xmath26 using that @xmath27 one rewrites @xmath28 the variance of the angular momentum with respect to a time - evolving state initially localized in momentum space is a measure of how much the latter spreads . in the heisenberg picture
@xmath29 setting @xmath30 and choosing as initial state the zero - angular momentum eigenstate @xmath31 of @xmath16 , one gets @xmath32\ . \label{sm27}\end{aligned}\ ] ] as this result depends on the explicit realization of the stochastic process up to time @xmath25 , in order to get a physically testable quantity one has to average over all possible realizations . before going to that , let us consider the case when the kick strengths are not stochastically perturbed , namely @xmath33 for all @xmath34 ; then , the variance is a bounded function of @xmath35 when @xmath8 with @xmath5 irrational , with corresponding localization , that is there is no spreading in angular momentum as illustrated in fig .
[ im0 ] . on the other hand , fig .
[ im0a ] shows the increasing of the angular momentum variance for @xmath36 .
as function of @xmath37 for @xmath38 and @xmath39 without stochastic kicks.,scaledwidth=70.0% ] as function of @xmath37 for @xmath38 and @xmath40 without stochastic kicks.,scaledwidth=70.0% ] from now on , the attention will be focused on the case @xmath8 with @xmath5 irrational .
as already stated , we take as @xmath13 a discrete @xmath6-step markov process with each @xmath10 taking the values @xmath41 with probabilities @xmath42 .
the stochastic strengths are chosen in such a way that , at each tick of time of period @xmath3 , the rotor is kicked or not with equal probability .
furthermore , the probability for the realization of the stochastic process @xmath43 will be denoted by @xmath44 of the form @xmath45 with transition coefficients @xmath46 for the value @xmath47 be followed by @xmath48 , given by @xmath49 so that the @xmath50 transition matrix @xmath51 $ ] reads @xmath52 the first contribution to the transition matrix represents a markov process with full memory , in the sense that only the first kick strength is stochastic and if it takes the value @xmath1 @xmath53 , it will keep it @xmath1 @xmath53 at all other times . on the other hand ,
the second contribution to the transition matrix gives rise to a bernoulli process with no memory , that is successive kick strengths are independent of one another .
therefore , by increasing @xmath54 from @xmath55 to @xmath6 one interpolates between a markov process with no memory and one with full memory .
the form of the transition matrix is particularly simple and yields @xmath56 furthermore , the probability vector @xmath57 is left invariant by the transition matrix @xmath58 so that the markov process is stationary : @xmath59 the characteristic function of such a process reads @xmath60 we shall denote by @xmath61 the averages with respect to the probability distribution ; then , the first moments @xmath62 while the second moments @xmath63 can be computed as follows : suppose @xmath64 , then @xmath65 from the stationarity of the process ( [ stationary ] ) and ( [ tmn ] ) one gets @xmath66 then , the momentum variance with respect to the angular momentum eigenstate @xmath67 reads @xmath68\ .
\label{sm31markov}\end{aligned}\ ] ] figure [ imm1 ] shows a diffusive behaviour which tends to disappear as @xmath54 approaches @xmath6 . as function of @xmath37 for the markov process with @xmath38 and @xmath39.,scaledwidth=70.0% ] in the case of @xmath69 , with probability @xmath70 there are no kicks and with the same probability there are always kicks of same strength @xmath1 ; therefore , the full memory case corresponds to the deterministic maryland model ( divided by two as this is the probability for a process with kicks ) as showed in the following figure for irrational @xmath71 .
as function of @xmath37 in the deterministic maryland model and in the stochastic markov model with with full memory : @xmath38 , @xmath39 and @xmath69.,scaledwidth=70.0% ]
like for the physical variance , a physical time - evolution for the density matrix of the linear kicked rotor with stochastic kicks is obtained from ( [ sm5 ] ) by averaging with respect to the realizations @xmath72 of the stochastic process up to time @xmath25 .
the averaging process yields the following maps on the density matrices of the quantum linear kicked rotor : @xmath73=\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,\prod_{j=1}^n\hat{u}_{j\tau}(x_j)\,\hat{\rho}\ , \prod_{j=1}^n\hat{u}^\dag_{j\tau}(x_j)\ , \label{sm32}\ ] ] where the unitary operators @xmath74 are as in ( [ sm4 ] ) and depend on the realization of the stochastic process at @xmath75 , @xmath76 is the probability distribution ( [ markov ] ) .
the maps @xmath77 are unital and trace - preserving : @xmath78&=&\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,\prod_{j=1}^n\hat{u}_{j\tau}(x_j)\,1\ , \prod_{j=1}^n\hat{u}^\dag_{j\tau}(x_j)=\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,=\ , 1\\ \nonumber { { { \rm tr}}}(\phi_n[\hat{\rho}])&=&\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,{{{\rm tr}}}\left[\prod_{j=1}^n\hat{u}_{j\tau}(x_j)\right]\,\hat{\rho}\ ,
\left[\prod_{j=1}^n\hat{u}^\dag_{j\tau}(x_j)\right]=\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,{{{\rm tr}}}(\hat{\rho})\\ \label{trace0 } & = & \,{{{\rm tr}}}(\hat{\rho})\ .\end{aligned}\ ] ] [ rem1 ] the dynamical maps @xmath77 evolve any given initial density matrix into a density matrix at time @xmath25 ; indeed , because of the positivity of the probability measure , they are a convex combination of unitary actions and thus a ( generalized ) kraus - stinespring form of completely positive maps on the space of states @xcite . therefore , they preserve the positivity not only of @xmath22 , but the composite maps @xmath79 also preserve the positivity of generic density matrices acting on the hilbert spaces @xmath80 , for all @xmath81 . this latter fact is essential for the physical consistency of the maps as dynamical maps ; indeed , one could always couple the quantum linear kicked rotor with an @xmath82-level system in such a way that the states of the coupled system evolve according to @xmath79 : if @xmath77 were not completely positive , one could always find an @xmath82 and an entangled initial state @xmath83 of the compound system such that @xmath84 $ ] develops for some @xmath35 negative eigenvalues then losing positivity and thus its physical interpretation as a state . if the stochastic process @xmath85 consisted of independent identically distributed stochastic variables , then @xmath86 and @xmath87 , @xmath35 times , where @xmath88=\sum_{x=0,1}\,e^{-ik(1-x)\cos\hat{\theta}}\,e^{-i\hat{p}\tau}\,\hat{\rho}\ , e^{i\hat{p}\tau}\,e^{ik(1-x)\cos\hat{\theta}}\ . \label{sm46}\ ] ] independence of @xmath89 means the absence of memory among realizations of the noise at different times which is reflected by the fact that the family of dynamical maps is a discrete semigroup : @xmath90 .
the presence of memory between kicks at different times spoils the semigroup property , but one can still wonder whether there exist intertwining maps @xmath91 on the space of states such that @xmath92 so that the issue at stake is to invert maps as in ( [ sm32 ] ) .
let us introduce the real @xmath35-dimensional vectors @xmath93 , @xmath94 and denote by @xmath95 the scalar products of such vectors ; then , using ( [ sm10 ] ) , one rewrites @xmath96&=&e^{-i\hat{p}\,n\,\tau}\,\psi_n[\hat{\rho}]\,e^{i\hat{p}\,n\,\tau}\\ \psi_n[\hat{\rho}]&=&\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,e^{-i\langle\bar{j}_n(\bar{x}_{n})\vert\bar{\nu}_{n}(\hat{\theta})\rangle}\,\hat{\rho}\ , e^{i\langle\bar{j}_n(\bar{x}_{n})\vert\bar{\nu}_{n}(\hat{\theta})\rangle}\ . \label{sm33a}\end{aligned}\ ] ] in order to find the inverse of the map @xmath97 , one can proceed as follows : from @xmath98\vert\theta_{2}\rangle\,=\,\langle\theta_{1}\vert\hat{\rho}\vert\theta_{2}\rangle\,\sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\ , e^{i\langle\bar{j}_n(\bar{x}_{n})\vert\bar{\nu}_{n}(\theta_{2})-\bar{\nu}_{n}(\theta_{1})\rangle}\ , \label{sm67}\ ] ] one can formally invert the expression , yielding @xmath99\vert\theta_{2}\rangle } { \sum_{\bar{x}_n\in\{0,1\}^n } p(\bar{x}_n)\,e^{-i\langle\bar{x}_{n}|\bar{\nu}_{n}(\theta_{2})-\bar{\nu}_{n}(\theta_{1})\rangle}}\ . \label{sm68}\ ] ] by means of a @xmath35-dimensional dirac delta @xmath100 used to represent the fraction , one can write the representation independent operator expression @xmath101\ , e^{-i\langle \bar{i}_n+\bar{v}_{n}\vert\bar{\nu}_{n}(\hat{\theta})\rangle}\ , \label{sm71}\ ] ] where @xmath102 is the characteristic function of the probability distribution over the noise - realizations @xmath103 . therefore , the formal inverse of the map @xmath77 in ( [ sm32 ] ) is the following linear map @xmath104 acting on the operators @xmath105 on @xmath106 , @xmath107=\int \frac{{\rm d}\bar{u}_{n}{\rm d}\bar{v}_{n}}{(2\pi)^{n}}\,\frac{e^{i\langle\bar{v}_{n}\vert\bar{u}_{n}\rangle}}{\lambda_n(\bar{u}_n)}\ , e^{i\langle \bar{i}_{n}+\bar{v}_{n}\vert\bar{\nu}_{n}(\hat{\theta})\rangle}\,e^{i\hat{p}\,n\,\tau}\,\hat{x}\,e^{-i\hat{p}\,n\,\tau}\ , e^{-i\langle \bar{i}_{n}+\bar{v}_{n}\vert\bar{\nu}_{n}(\hat{\theta})\rangle}\ .\ ] ] let @xmath108 be the fourier transform of the inverse of the characteristic function ; one rewrites @xmath109=\int{\rm d}\bar{v}_n\ , \tilde{\lambda}_n(\bar{v}_n)\,e^{i\langle \bar{i}_{n}+\bar{v}_{n}\vert\bar{\nu}_{n}(\hat{\theta})\rangle } \,e^{i\hat{p}\,n\,\tau}\,\hat{x}\,e^{-i\hat{p}\,n\,\tau}\,e^{-i\langle \bar{i}_{n}+\bar{v}_{n}\vert\bar{\nu}_{n}(\hat{\theta})\rangle}\ .\ ] ] like ( [ sm32 ] ) , the expression above resembles a continuous kraus - stinespring representation of a completely positive map ; however , unlike the fourier transform of the characteristic function @xmath102 which yields the probability distribution and is thus positive , @xmath110 need not be positive , whence @xmath104 and thus @xmath111 need not be completely positive .
[ ncprem ] the fourier transform @xmath112 need not exist as a function .
however , since it appears within an integration with respect to @xmath113 it can always be interpreted as a distribution over a suitable class of test functions .
these integration functions immediately appear when computing quantities as @xmath114\vert\psi\rangle=\int{\rm d}\bar{v}_n\ , \tilde{\lambda}_n(\bar{v}_n)\ , \langle\varphi\vert\hat{v}_n(\bar{v}_n ) \,e^{i\hat{p}\,n\,\tau}\,\hat{x}\,e^{-i\hat{p}\,n\,\tau}\ , \hat{v}^\dag_n(\bar{v}_n)\vert\psi\rangle\ , \ ] ] where the operators are as in ( [ inverse1 ] ) , with suitable choices of vectors @xmath115 .
notice that @xmath104 is unital and trace preserving , namely @xmath116&=&\int{\rm d}\bar{v}_n\ , \tilde{\lambda}_n(\bar{v}_n)\ , 1\ , = \
, 1\\ \nonumber { { { \rm tr}}}(\phi^{-1}_n[\hat{\rho}])&=&\int{\rm d}\bar{v}_n\ , \tilde{\lambda}_n(\bar{v}_n)\ , { { { \rm tr}}}\big(\hat{v}_n(\bar{v}_n ) \,e^{i\hat{p}\,n\,\tau}\,\hat{\rho}\,e^{-i\hat{p}\,n\,\tau}\ , \hat{v}^\dag_n(\bar{v}_n)\big)\\ \label{trace } & = & \int{\rm d}\bar{v}_n\ , \tilde{\lambda}_n(\bar{v}_n)\ , { { { \rm tr}}}(\hat{\rho})={{{\rm tr}}}(\hat{\rho})\ .\end{aligned}\ ] ] in the case of a continuous - time family of completely positive maps @xmath117 , while markovianity is identified by the composition law @xmath118 a criterion of non - markovianity is taken to be the indivisibility of the one - parameter family of maps @xmath119 ( [ sm32 ] ) , namely the absence of completely positive intertwining maps @xmath120 such that @xmath121 the absence of completely positive intertwining maps can be witnessed by the non - monotonic behavior of quantities that , under completely positive maps , would either be always non - decreasing or non - increasing . consider a function @xmath122 on the space of states which is monotonically non - increasing under completely positive maps @xmath123 , that is @xmath124 .
since the maps @xmath119 are completely positive , it follows that @xmath125 for all @xmath126 ; however , if @xmath120 is not completely positive , then @xmath127 need not be smaller than @xmath128 and @xmath129 may then increase in the time - interval @xmath130 $ ] ( see @xcite for an updated review ) . in the following we study the discrete - time occurrence of such a non - markovianity criterion .
in the continuous case , several witnesses of non - markovianiy have been proposed @xcite-@xcite based on the absence of monotonic behaviour . in the following , we will consider the hilbert - schmidt norm on the space of bounded operators on @xmath106 , @xmath131 all unital completely positive maps @xmath123 , that is such that @xmath132=1 $ ] , satisfy the so - called schwartz - positivity , namely @xmath133\geq ( \psi[x])^\dag\psi[x]\ ] ] for all operators @xmath105 on @xmath106 . indeed , @xmath123 completely positive means that @xmath134 is positive ; then , setting @xmath135 , on the positive operator on @xmath136 @xmath137 one gets @xmath138 & -(\psi[\hat{x}])^{\dagger}\cr -\psi[\hat{x } ] & 1 \end{pmatrix}\,\geq\,0\ .\ ] ] it then follows that , for all @xmath139 , @xmath140)^\dag\right ) \begin{pmatrix } \psi[\hat{x}^\dagger\hat{x } ] & -\psi[\hat{x}^\dag]\\ -\psi[\hat{x } ] & 1 \end{pmatrix } \begin{pmatrix } \psi[x]|\varphi\rangle \end{pmatrix } = \langle\varphi\vert\psi[\hat{x}^\dag\hat{x}]-(\psi[\hat{x}])^\dag\psi[\hat{x}]\vert\varphi\rangle\ge0\ .\ ] ] from schwartz - positivity it follows that the hilbert - schmidt norm must decrease under completely positive unital and trace preserving maps ; indeed , @xmath141\|^2_{hs}={{{\rm
tr}}}\big((\psi[\hat{x}])^\dag\psi[\hat{x}]\big)\leq{{{\rm tr}}}\big(\psi[\hat{x}^\dag\hat{x}]\big)={{{\rm tr}}}\big(\hat{x}^\dag\hat{x}\big ) = \|\hat{x}\|^2_{hs}\ .\ ] ] let us now consider the hilbert - schmidt norm of a density matrix @xmath142 $ ] that evolves in time according to the discrete family of dynamical maps @xmath77 in ( [ sm32 ] ) that compose as in ( [ intertwining ] ) . if @xmath91 were completely positive , then , because of the previous argument @xmath143 .
indeed , @xmath144 is unital and trace preserving for such are @xmath77 and @xmath145 as follows from ( [ unital0 ] ) , ( [ trace0 ] ) and ( [ unital ] ) , ( [ trace ] ) . a sufficiently explicit expression for the hilbert - schmidt norm @xmath146 that is amenable to numerical investigations is readily achieved . using ( [ sm33 ] ) , one explicitly finds @xmath147\|^2_{hs}={{{\rm tr}}}\big(\psi^2_n[\hat{\rho}]\big)\\ \label{app1 } & & = \sum_{\bar{x}_n,\bar{y}_n\in\{0,1\}^n}\,p(\bar{x}_n)\,p(\bar{y}_n)\ , { { { \rm tr}}}\left(\hat{\rho}\,e^{i\langle\bar{x}_n-\bar{y}_n\vert\bar{\nu}_n(\hat{\theta})\rangle}\,\hat{\rho}\ , e^{-i\langle\bar{x}_n-\bar{y}_n\vert\bar{\nu}_n(\hat{\theta})\rangle}\right)\ .\end{aligned}\ ] ] by going to the angle representation , whereby @xmath148 , one expresses the hilbert - schmidt norm in terms of the probability distribution characteristic function : @xmath149\|^2_{hs}= \int_0^{2\pi}{\rm d}\theta_1\,\int_0^{2\pi}{\rm d}\theta_2\,|\langle\theta_1\vert\hat{\rho}\vert\theta_2\rangle|^2\ , \big|\lambda\big(\bar{\nu}_n(\theta_1)-\bar{\nu}_n(\theta_2)\big)\big|^2\ .\ ] ] in the case of the markov process one uses ( [ charfunctmarkov ] ) with @xmath150 figure @xmath151 shows various time behaviours of the hilbert - schmidt norm for an initial momentum eigen - projector @xmath152 : in absence of memory ( @xmath153 ) , the hilbert - schmidt distance behaves monotonically , whereas the higher is @xmath54 , the higher is the memory in the process , the more it oscillates .
\|^2_{hs}$ ] for markovian kicks : @xmath38 and @xmath154.,scaledwidth=70.0% ] with respect to definition of non - markovianity corresponding to the absence of intertwining completely positive maps , the above result is an evidence of the fact that this choice makes sense also when there are no generators for the dynamics .
moreover , in the present case the non - markovianity witness is the hilbert - schmidt norm while in @xcite is the fidelity and in @xcite the relative entropy of entanglement .
this is due to the occurring maps being not only trace - preserving , but also unital . from the previous numerical evidence , some of the maps @xmath91 can not be @xmath155-positive , for instance @xmath156 can not be positive , and thus @xmath157 can not be completely positive . in this section ,
we address the properties of @xmath158 in more detail : namely , we also consider the problem of its positivity as a linear map , that is whether @xmath159\geq 0 $ ] on all projectors @xmath160 on @xmath106 .
we set @xmath161 . by means of ( [ sm33 ] ) with @xmath162 , of ( [ inverse1 ] ) and of the fact that @xmath163 with @xmath164 one finds @xmath165= \sum_{x_{1,2}\in\{0,1\}}p(x_1,x_2)\ , \int{\rm d}u{\rm d}v\,\frac{e^{iuv}}{2\pi\lambda_1(u)}\ , e(x_1,x_2,v,\hat{\theta})\,e^{-i\hat{p}\tau}\,\rho\ , e^{i\hat{p}\tau}\,e^\dag(x_1,x_2,v,\hat{\theta})\\ \label{gamma2 } & & \hskip -.5 cm e(x_1,x_2,v,\hat{\theta})=e^{ik(x_{1}+v)\cos(\hat{\theta}-\tau)}e^{-ik(1-x_{2})\cos\hat\theta}.\end{aligned}\ ] ] in the angle representation where @xmath166 , @xmath167= \int_0^{2\pi}{\rm d}\theta_1\int_0^{2\pi}{\rm d}\theta_2\,e^{ik(\cos\theta_2-\cos\theta_1)}\ , g(\theta_1,\theta_2)\ , \vert\theta_1\rangle\langle\theta_1\vert\,e^{-i\hat{p}\tau}\,\rho\ , e^{i\hat{p}\tau}\,\vert\theta_2\rangle\langle\theta_2\vert\\ & & \hskip -.5 cm g(\theta_1,\theta_2)=\frac{\lambda_2\big(k[\cos(\theta_2-\tau)-\cos(\theta_1-\tau)],k[\cos\theta_2-\cos\theta_1]\big ) } { \lambda_1\big(k[\cos(\theta_2-\tau)-\cos(\theta_1-\tau)]\big)}\ . \label{gamma4}\end{aligned}\ ] ] let @xmath22 be the following pure state projection @xmath168&=&\frac{1}{2\pi } \int_0^{2\pi}{\rm d}\theta_1\int_0^{2\pi}{\rm d}\theta_2\ , g(\theta_1,\theta_2)\ , \vert\theta_1\rangle\langle\theta_2\vert .\end{aligned}\ ] ] let then @xmath169 be a not normalized vector , with @xmath170}(\theta)\ , \quad j=1,2\ ] ] the characteristic functions of intervals around @xmath171 , @xmath172 , of size @xmath173 . then one considers the mean value @xmath174\vert\varphi\rangle = i_1(\varepsilon)+i_2(\varepsilon)-i_{12}(\varepsilon ) -i_{21}(\varepsilon)\quad\hbox{where}\\ \nonumber\\ \label{positivity2a } i_1(\varepsilon)&=&\frac{1}{\varepsilon^2 } \int_{\bar{\theta}_1-\varepsilon/2}^{\bar{\theta}_1+\varepsilon/2}{\rm d}\theta_1\int_{\bar{\theta}_1-\varepsilon/2}^{\bar{\theta}_1+\varepsilon/2}{\rm d}\theta_2\ , g(\theta_1,\theta_2)\\ \nonumber\\ \label{positivity2b } i_2(\varepsilon)&=&\frac{1}{\varepsilon^2 } \int_{\bar{\theta}_2-\varepsilon/2}^{\bar{\theta}_2+\varepsilon/2}{\rm d}\theta_1 \int_{\bar{\theta}_2-\varepsilon/2}^{\bar{\theta}_2+\varepsilon/2}{\rm d}\theta_2\ , g(\theta_1,\theta_2)\\ \nonumber\\ \label{positivity2c } i_{12}(\varepsilon)&= & \frac{1}{\varepsilon^2 } \int_{\bar{\theta}_1-\varepsilon/2}^{\bar{\theta}_1+\varepsilon/2}{\rm d}\theta_1\int_{\bar{\theta}_2-\varepsilon/2}^{\bar{\theta}_2+\varepsilon/2}{\rm d}\theta_2\ ,
g(\theta_1,\theta_2)\\ \nonumber\\ \label{positivity2d } i_{21}(\varepsilon)&=&\frac{1}{\varepsilon^2}\int_{\bar{\theta}_2-\varepsilon/2}^{\bar{\theta}_2+\varepsilon/2}{\rm d}\theta_1 \int_{\bar{\theta}_1-\varepsilon/2}^{\bar{\theta}_1+\varepsilon/2}{\rm d}\theta_2\ , g(\theta_1,\theta_2)\ .\end{aligned}\ ] ] by letting @xmath175 one gets @xmath176 using ( [ charfunctmarkov ] ) , the quantity @xmath177 between parenthesis reads @xmath178 for a suitable choice of angles @xmath179 , the quantity @xmath177 can become strictly larger than @xmath155 as showed in fig .
@xmath180 , the positive and negative peaks correspond to zeroes of the denominator in the above expressions .
therefore , the quantity @xmath181 in ( [ nonppositivity ] ) becomes negative showing that the intertwining map , besides not being completely positive , is not even a positive map . in the absence of memory , that is when @xmath153 , the function @xmath181 is instead positive , as it must be .
[ remfinal ] the functions in ( [ positivity1 ] ) tend to dirac deltas when @xmath175 and thus exit from the hilbert space @xmath106 ; however , because of the continuity of the functions entering the integral defining @xmath182 in ( [ positivity2 ] ) , the previous limit ensures that one can always find a suitably small , but finite @xmath183 such that @xmath184 .
: @xmath185\times[0;2\pi]\to\mathbb{r}$ ] for markovian kicks : @xmath186 , @xmath38 and @xmath154.,scaledwidth=70.0% ]
we studied a stochastic version of the quantum linear kicked rotor , obtained by letting the kick strengths become a discrete markov process .
the dynamics becomes a dissipative quantum process consisting of a discrete family of completely positive maps . in case of absence of memory ,
when the markov process becomes of bernoulli type , it has been shown that these maps compose as a discrete semigroup .
on the other hand , when memory effects are present , non - markovianity is witnessed by the non monotonic behaviour of the hilbert - schmidt norm .
the origin of this behaviour is to be found in the fact that the intertwining maps connecting completely positive maps at different ticks of time are trace - preserving but not completely positive , linking non - markovianity to non - divisibility ( in a discrete sense in the present case ) .
a more detailed study of the intertwining maps shows that they are not even positive . | the so - called maryland model is a linear version of the quantum kicked rotor ; it exhibits anderson localization in momentum space . by turning the kicks into a markovian stochastic process ,
the dynamics becomes a dissipative quantum process described by a discrete family of completely positive maps that allows to explicitly study the relation between divisibility of the maps and the degree of memory of the process . |
recent observations of high redshift quasars from the sloan digital sky survey ( sdss ) suggest that the cosmological reionization was completed at @xmath2 ( e.g. , fan et al .
2001 ; becker et al .
2001 , barkana 2001 ; djorgovski , castro , & stern 2001 ; cen & mcdonald 2002 ; litz et al .
2002 ) , implying a largely opaque universe beyond @xmath2 . in the context of the standard cold dark matter model ( bahcall et al .
1999 ) there is by now a clear consensus that galaxies of mass @xmath8 are largely responsible for the final cosmological reionization ( e.g. , barkana & loeb 2001 ; madau 2002 ; wyithe & loeb 2003a , b ; mackey , bromm , & hernquist 2003 ; cen 2003a , b ; venkatesan , tumlinson , & shull 2003 ; ciardi , ferrara , & white 2003 ; somerville & livio 2003 ; fukugita & kawasaki 2003 ; sokasian et al .
2003 ) . the recent wilkinson microwave anisotropy probe ( wmap ) observations ( kogut et al .
2003 ) appear to confirm a much more extended reionization process , with the first reionization occurring at a redshift possibly as high as @xmath9 ( cen 2003a ) . as a result
, our observational frontier is quickly being pushed towards high redshifts not thought possible just a few years ago .
it is of great interest to directly detect the sources of reionization and the ionization state of the intergalactic medium from @xmath10 to @xmath11 . to date the observed lyman alpha emitters ( lae ) at @xmath12 appear to be quite luminous with star formation rate ( sfr ) probably in excess of @xmath13yr@xmath14 ( hu et al .
2002 ; kodaira et al .
2003 ) , which , if assuming a starburst of duration of @xmath15yrs and star formation efficiency of @xmath16 , would give @xmath17 of stars .
the masses of these galaxies most likely exceed @xmath18 .
these observed laes are thus much more massive than the typical galaxies thought to produce most of the ionizing photons at @xmath3 current observations are just scraping the tip of an iceberg . because of the damping wing of the gunn - peterson trough cast by the significantly neutral intergalactic medium ( igm ) before reionization ( miralda - escud 1998 ) , the intrinsic luminosities ( after correcting for possible gravitational lensing effects ) for these observed laes may be still larger .
consistent with this , typical laes observed at somewhat lower redshift appear to be less luminous by a factor of ten or more ( yan , windhorst , & cohen 2003 ; rhoads et al .
2003 ; taniguchi et al .
2003 ) . for smaller galaxies of mass @xmath8
typically with an intrinsic ly@xmath0 line width narrower than @xmath19km / s , the absorption by the damping wing of intervening neutral igm is dramatically more severe , probably rendering them undetectable ( haiman 2002 ) . in this _
we suggest that targeted , very narrowband observations of ly@xmath0 emission at strmgren spheres of known luminous quasars at @xmath3 may provide a unique window to directly detect the bulk of the sources for cosmological reionization , because a large fraction of galaxies in such regions can transmit their ly@xmath0 emission lines largely unattenuated .
in the absence of hydrogen recombination with a static density field , the proper radius @xmath20 of the strmgren sphere of a quasar of age @xmath21 is such that the number of hydrogen atoms in the sphere equals the total number of ionizing photons produced by the source ( cen & haiman 2000 ) : @xmath22^{1/3}=4.3x_{\rm igm}^{-1/3}\left(\frac{\dot n_{\rm ph}}{1.3\times 10^{57}{\rm s^{-1}}}\right)^{1/3}\left(\frac{t_q}{2\times10 ^ 7{\rm yr}}\right)^{1/3}\left(\frac{1+z_q}{7.28}\right)^{-1}~{{\ , { \rm mpc}}},\end{aligned}\ ] ] where @xmath23 is the mean hydrogen density within @xmath20 , @xmath24 is the neutral hydrogen fraction of the igm , and @xmath25 is emission rate of ionizing photons from the quasar . for @xmath23 ,
we have adopted the mean igm density with @xmath26 , and @xmath27 is the observed flux of sdss 1030 + 0524 ( haiman & cen 2002 ) . in the realistic case with an evolving cosmological density field and a finite hydrogen recombination time ,
@xmath20 can be computed in detail ( cen & haiman 2000 ) . but @xmath20 estimated in equation ( 1 ) gives a reasonable approximation to the truth ( haiman & cen 2002 ) at redshift @xmath28 . at sufficiently large separations from the ly@xmath0 line center ( @xmath29 )
the optical depth of the damping wing is ( miralda - escud 1998 ) @xmath30 where @xmath31 is baryon density parameter , @xmath32 is the helium mass fraction of the primordial gas , @xmath33 and @xmath34 are the hubble constant at redshift zero and @xmath35 , respectively , @xmath36 and @xmath37 are the observed wavelength separation from the line center and the observed wavelength of the ly@xmath0 line , respectively . replacing @xmath38 by @xmath39 , where @xmath40 is the distance between the surface of the strmgren sphere and the point @xmath41 inside the strmgren sphere in question along the line of sight and @xmath42 is the speed of light , and approximating @xmath43 , we obtain @xmath44 where @xmath45 is the proper quasar - centric distance of the point in question and @xmath46 is the angle between the line segment from the quasar to the observer and the line segment from the quasar to @xmath41
. the gunn - peterson optical depth due to the residual neutral hydrogen inside the strmgren sphere , in the absence of other ionizing sources , can be expressed as ( cen & haiman 2000 ) @xmath47 , due to a very small residual neutral fraction @xmath48 . as a result of small @xmath49
, the local mini - strmgren spheres produced by ly@xmath0 galaxies would render the absorption due to all residual neutral hydrogen inside the strmgren sphere negligible .
the total ly@xmath0 optical depth at @xmath41 is hence just @xmath50 . from equation ( 3 )
we see that , for a luminous quasar like sdss 1030 + 0524 , ly@xmath0 lines from galaxies located within the central region region of @xmath51mpc , would suffer a total optical depth of @xmath52 corresponding @xmath46 from @xmath53 to @xmath54 , i.e. , a flux reduction of @xmath55 , assuming @xmath56 .
it is evident that there is a significant region within the strmgren sphere where most of the ly@xmath0 emission galaxies are able to transmit the ly@xmath0 emission lines largely unattenuated .
there are several potentially very interesting applications from direct observations of regions around luminous quasars prior to the completion of cosmological reionization . before turning to the possible applications it is useful to estimate the abundance of possibly detectable galaxies at @xmath3 . assuming that the fraction of all mass having formed into halos with masses @xmath57 is @xmath58 and that starbursts each with a duration of @xmath59 has occurred once in each halo in the last half hubble time
, then we find that the comoving number density of such galaxies is @xmath60mpc@xmath61 . assuming a star formation efficiency @xmath62 we obtain the star formation rate in each galaxy @xmath63/yr .
this simple estimate , however , involves two currently unknown parameters , @xmath62 and @xmath59 . perhaps a better way to estimate
the abundance and luminosity of lae is to utilize both hierarchical cdm model and lae survey observations at highest accessible redshift , currently at @xmath64 .
figure 1 shows the predicted cumulative number density of galaxies at @xmath65 ( solid curve ) and @xmath66 ( dashed curve ) for a cosmological constant dominated cold dark matter model with @xmath67 , @xmath68 , @xmath69km / s / mpc , @xmath70 and @xmath71 . the observational data point for lae shown as the solid dot in figure 1
is taken from rhoads et al .
( 2003 ) , who give @xmath72mpc@xmath61 at @xmath64 .
we have normalized " the observational data point to the theoretical prediction at @xmath65 by requiring the same abundance , which then gives an implied mass threshold of the observed lae by rhoads et al .
clearly , even the lae observed at @xmath64 by rhoads et al .
are quite massive with masses about @xmath73 , if the adopted cold dark matter model is correct .
the computed lower bound of star formation rate for the observed lae at @xmath64 by rhoads et al . of @xmath74 corresponds to the flux sensitivity of @xmath75erg/@xmath76/sec of their lala survey .
if we assume that the ly@xmath0 luminosity of an lae is simply proportional to its mass , it would require a flux sensitivity of @xmath77)erg/@xmath76/sec to detect galaxies of mass of @xmath78 with corresponding abundance of @xmath79mpc@xmath61 at @xmath66 ( open circles in figure 1 ) , respectively . between night sky lines
there are a few good observational windows at @xmath80 ( corresponding to @xmath81 for the ly@xmath0 line ) , most notably the one center at @xmath64 and the other at @xmath82 .
however , even within those good " windows the required integration time to detect fainter lae using keck telescopes is quite long , @xmath83/sec)@xmath84hours ( gjorgovski 2003 , private communications ; fan 2003 , private communications ) , implying that approximately a hundred hours is needed on keck to detect galaxies of mass @xmath85 , _ even in the absence of significantly attenuation by neutral igm_. the current observational limit for ly@xmath0 flux roughly corresponds to @xmath86 a few @xmath87yrs at @xmath2 ( hu et al .
2002 ; yan , windhorst , & cohen 2003 ; rhoads et al . 2003
; taniguchi et al .
detecting even lower mass lae may be quite infeasible using keck telescopes .
furthermore , for ground - based observations where the primary background noise in the interested spectral regime comes from sky background , it is a very tedious task to search for usable windows between numerous strong sky lines to find high redshift galaxies and quasars .
hence , it is likely necessary to resort to james webb space telescope ( jwst ) for detecting the bulk of galaxies ( of mass @xmath88 ) at @xmath7 . as pointed out by oh ( 1999 )
, jwst should be able to detect the h@xmath89 line ( and possibly h@xmath90 and other balmer lines ) of galaxies of mass @xmath88 at @xmath7 .
a joint detection in ly@xmath0 and h@xmath89 would not only firmly determine the redshift of each galaxy but also shed very useful light on important physical processes of each galaxy including dust content .
in addition , the ratio of ly@xmath0 to h@xmath89 lines would provide a cross - check of the damping wing modulated spatial distribution of lae or a tool to break possible degeneracies in the spatial distribution of lae from ly@xmath0 line alone .
we note that three factors could boost the effectiveness of star formation at high redshift in terms of ly@xmath0 emission .
first , the gas at high redshift may be much metal poor ; for example , taking the same salpeter imf would give about a factor of @xmath91 more ionizing photons hence ly@xmath0 photons for stars with @xmath92 compared to stars with solar metallicity .
second , if the metal poor gas at high redshift forms relatively more massive stars , the efficiency of producing ionizing photons would be higher .
third , the dust content in the high - z metal - poor galaxies should be lower , possibly allowing for relatively more ly@xmath0 photons to escape .
these factors may render high redshift lae more luminous that their low redshift counterparts at a star formation rate .
figure ( 2 ) shows the ratio of the number of galaxies at a distance @xmath93 ( with width @xmath94 ) from the quasar on the blue side over the number of galaxies at the same distance from the quasar on the red side within a quasar - centric transverse radius of @xmath95mpc proper , as a function of @xmath93 .
three panels correspond to three flux sensitivities of @xmath96erg/@xmath97/sec ( panel a ) , @xmath98erg/@xmath97/sec ( panel b ) and @xmath99erg/@xmath97/sec ( panel c ) , which are then translated into the threshold masses of detectable galaxies using figure 1 to infer the corresponding abundances . in each panel
two sets of cases are shown .
case ( 1 ) : @xmath100mpc proper with @xmath101 , respectively , shown as thick solid , thick dotted and thick dashed curves . case ( 2 ) : @xmath102mpc proper with @xmath101 , respectively , shown as thin solid , thin dotted and thin dashed curves .
we choose to plot the ratio instead of the direct number density to avoid ambiguity due to a possibly non - uniform , quasi spherical distribution of galaxies around quasars .
galaxies will still likely be clumpy and clustered on small scales but a large volume considered would smooth out most of the fluctuations .
let us now estimate the accuracy of the data points .
the comoving volume enclosed by a cylindrical region within a transverse radius @xmath103 and a line - of - sight distance of @xmath104 ( from @xmath105 to @xmath106 ) is @xmath107 comoving at @xmath108 , which is then multiplied by the number density of galaxies using the respective @xmath109 from figure 1 to yield the corresponding total number of galaxies above the respective mass threshold .
five data points are used for the indicated @xmath93 range .
the errorbars are shown in each panel for the case with @xmath102mpc proper . from panel
( a ) of figure 2 it is clear that @xmath110 and @xmath111 are distinguishable at @xmath112 level even for galaxies with mass @xmath113 ( panel a ) , detectable by keck telescope with an integration time of order of @xmath19 hours for each galaxy ( assuming one spectrum per observation ) with total @xmath114 galaxies .
this might be a costly project but the potential payoff is attractive .
cases with @xmath111 and @xmath115 can only be marginally differentiated at this sensitivity level . at the same time , it is evident that the models with different @xmath116 can be more easily distinguished , allowing for a precise measurement of @xmath116 .
in addition , a precise measurement of @xmath116 , in conjunction with the spatial information ( see 3.2 ) , may allow a determination of the age of the quasar .
with jwst , if one can reach flux levels @xmath117erg/@xmath76/sec , very precise measurements of @xmath24 can be made , as can be seen from panels ( b , c ) in figure 2 .
we note that @xmath106 corresponds to a line - of - sight comoving distance from the quasar of @xmath118mpc at @xmath108 .
the gravitational influence of the quasar host galaxy at that distance is likely to be negligible , alleviating some possible complications including infall .
in addition , the typical peculiar velocities of galaxies at high redshift is likely be @xmath119km / s , which would correspond to a possible displacement of @xmath120 , too small to significantly matter .
finally , we note that , in the case of a lone lae ( not sitting inside a large strmgren sphere of a luminous quasar ) , an lae estimated to have an sfr about an order of magnitude higher than the solid dot in figure 1 ( hu et al .
2002 ) can only transmit a small fraction of ly@xmath0 flux due to attenuation by the neutral igm ( haiman 2002 ) .
lower mass lae are further devastated by their lower intrinsic line widths in addition to smaller strmgren spheres .
as haiman ( 2002 ) pointed out , no ly@xmath0 line of an lae narrower than @xmath19km / s can be observed .
therefore , it is estimated that no lae galaxies less massive than @xmath121 will be detectable in a random field outside a large strmgren sphere of a luminous source . fitting the data to obtain @xmath116 and @xmath24 ( 3.1 ) would yield precise information on @xmath50 , which in turn would allow a determination of a region on the red side of the quasar where attenuation due to intervening neutral igm is negligible
consequently , one can construct an unbiased luminosity function given the likely availability of up to thousands of galaxies in that region , which would further allow for corrections to be made for missing " galaxies in other regions where @xmath50 is appreciable .
moreover , the spatial distribution of galaxies in this region and others in other regions would allow an accurate determination of the clustering properties of galaxies up to a scale of @xmath122mpc comoving .
given the determined luminosity function and @xmath24 , along with follow - up spectroscopic observations of the galaxies detected which would produce the line widths for these galaxies , we could self - consistently determine a fraction of these galaxies that would be directly observed even sitting in an average region and their possible observable features . taking the farthest red region as an unbiased region
, this would then allow a determination of luminosity - dependent biased distributions of galaxies , down to the lowest luminosity of galaxies in the farthest red region .
this may shed useful light on the formation of quasars and their host galaxies . whether or not the farthest red region is an unbiased
, average " region can be checked using the same data , where one may expect to see a diminished radial gradient .
the three - dimensional information of the distribution of galaxies around a quasar makes it possible to probe the angular distribution of emission from the quasar .
it should be relatively straightforward to build self - consistent models to account for possible outcomes from observations .
for example , if galaxy abundance drops off rapidly at a transverse distance significantly smaller than the inferred @xmath116 , it would indicate that the quasar emission is bi - polar , whose angles can be inferred . if , in addition , there is a strong suppression of galaxies on the red side of the quasar even at small @xmath123 , it would imply that the quasar emission is largely beamed towards us .
we will not attempt to make more detailed models here .
if the standard unified model for agn / quasars ( e.g. , antonucci 1993 ) is correct , the ensuing consequence on the distribution of laes around quasars will clearly be detectable .
this method may provide the first and accurate way to test this picture . in summary
, we show that , with a tunable , very narrowband filter of @xmath124 or a narrowband filter of @xmath5 to survey a luminous quasar strmgren sphere with follow - up spectroscopic identifications of the ly@xmath0 emission lines of individual galaxies inside the quasar strmgren sphere , we could obtain a wealth of information for the high redshift galaxies that are responsible for reionizing the universe at @xmath2 . this same technique may be applicable to quasars at any high redshift . at higher redshift ,
galaxies are somewhat less massive , but they may have a more top - heavy imf , possibly a higher star formation efficiency , a shorter starburst duration and much lower dust content .
these positive , compensatory effects could possibly render high redshift galaxies with mass @xmath125 intrinsically as luminous as galaxies of mass @xmath8 at @xmath10 .
we expect that jwst should be able to detect these galaxies at @xmath126 ( eisenstein 2003 ) , provided that a tunable , narrowband filter at appropriate wavelengths is available .
we show that targeted observations at the strmgren spheres of known luminous quasars with tunable very narrowband filters of width @xmath127 or a narrowband filter of @xmath5 to survey a luminous quasar strmgren sphere with follow - up spectroscopic identifications of the ly@xmath0 emission lines of individual galaxies would provide a direct way to observe the ly@xmath0 emission lines of the bulk of galaxies at @xmath7 in the mass range @xmath8 that are believed to be responsible for the second cosmological reionization . at present
it appears that ground - based observations , such as by keck , can only detect a relatively small fraction , high end tail of galaxies with mass @xmath128 .
but even such observations may already be able to differentiate between a fully neutral and a 10% neutral intergalactic medium at @xmath7 .
jwst should provide the unique opportunity to detect the bulk of galaxies at high redshift .
the likely gain in sensitivity with a tunable , very narrowband filter may allow jwst to detect the pop iii galaxies of mass @xmath125 with a top - heavy imf at @xmath126 . in this case , the key is to search for luminous enough quasars at these very high redshifts , perhaps requiring an infrared survey comparable to sdss .
this is presently technologically challenging but the potential scientific payoff is immense . we show that such observations will allow for accurate determinations of the ionization state of the igm , the sizes of the strmgren spheres , the ages of the quasars , the luminosity function of high redshift galaxies and its evolution , the spatial distribution of galaxies and its evolution , the biased distribution of galaxies around quasars and the anisotropy of quasar emission .
if the universe was indeed reionized twice ( cen 2003a ) , then there will be a window of opportunity somewhere at @xmath129 , where galaxies can transmit ly@xmath0 lines ( as well as optical - uv continuum and other emission lines ) unabsorbed ; they should be directly observable by jwst even in the absence of luminous quasars .
a signature continuum cutoff of the spectra of these galaxies shortward of @xmath130 would pin down the redshift of the second cosmological recombination epoch @xmath131 and then a rapid thinning out of the number density of ly@xmath0 galaxies at some redshift @xmath132 would be indicative of the first reionization epoch .
this research is supported in part by nsf grant ast-0206299 .
i thank george djorgovski , xiaohui fan and rogier windhorst for useful comments and discussions , michael strauss for pointers to useful astronomy websites , and the referee for useful suggestions which significantly strengthen the paper .
antonucci , r.r.j .
1993 , araa , 31 , 473 bahcall , n.a . , ostriker , j.p . , perlmutter , s. , & steinhardt , p.j .
1999 , science , 284 , 1481 barkana , r. , 2002 , newa , 7 , 85 barkana , r. , & loeb , a. 2001 , phys . rep .
, 349 , 125 becker , r.h .
, et al .
2001 , aj , 122 , 2850 cen , r. 2003a , apj , 591 , 12 cen , r. 2003b , apj , 591 , l5 cen , r. , & haiman , z. 2000 , apj , 542 , l75 cen , r. , & mcdonald , p. 2002
, apj , 570 , 457 ciardi , b. , ferrara , a. , & white , s.d.m .
2003 , astro - ph/0302451 djorgovski , s.g . , castro , s. , stern , d. , & mahabal , a.a .
2001 , apj , 560 , l5 eisenstein , d. 2003 , tucson reionization workshop fan , x. , et al .
2002 , aj , 123 , 1247 fukugita , m. , & kawasaki , m. 2003 , mnras , 343 , l25 haiman , z. 2002 , apj , 567 , l1 haiman , z. , & cen , r. 2002 , apj , 578 , 702 hu , e.m . , et al .
2002 , apj , 568 , l75 kodaira , k. , et al .
2003 , astro - ph/0301096 litz , a. , hui , l. , zaldarriaga , m. , & scoccimarro , r. 2002 , astro - ph/0111346 , apj , in press mackey , j. , bromm , v. , & hernquist , l. 2003 , apj , 586 , 1 madau , p. 2002 , astro - ph/0210268 madau , p. , & rees , m.j . 2000 , apjl , 542 , l69 miralda - escud , j. 1998 , apj , 501 , 15 oh , s.p . 1999 ,
apj , 527 , 16 rhoads , j.e .
, et al .
2003 , aj , 125 , 1006 sokasian , a. , abel , t. , hernquist , l. , & springel , v. 2003 , astro - ph/0303098 somerville , r.s .
, & livio , m. 2003 , astro - ph/0303017 taniguchi , y. , et al . 2003 , apj , 585 , l97 wyithe , s. , & loeb , a. 2003a , apj , 586 , 693 wyithe , s. , & loeb , a. 2003b , apj , 588 , l69 yan , h. , windhorst , r.a . , & cohen , s.h . 2003 , apj , 585 , l93 | the intrinsic ly@xmath0 emission lines of normal galaxies before reionization are much absorbed by the damping wing of the gunn - peterson trough , rendering their direct detection nearly impossible , if their intrinsic line widths are less than @xmath1km / s .
high redshift luminous quasars prior to the completion of cosmological reionization at @xmath2 , on the other hand , are capable of producing large hii regions around them ( strmgren spheres ) to allow their intrinsic ly@xmath0 emission lines to be transmitted without overwhelming absorption ( cen & haiman 2000 ) .
we suggest that targeted observations at the strmgren spheres of known luminous quasars at @xmath3 would be able to detect ly@xmath0 emission lines of galaxies inside the strmgren spheres largely unattenuated .
a tunable , very narrowband filter of @xmath4 or a narrowband filter of @xmath5 with follow - up spectroscopic identifications will be required .
such observations could directly observe the sources of cosmological reionization including possibly the pop iii galaxies at @xmath6 by jwst .
possible applications include determinations of the ionization state of the intergalactic medium , the sizes of the strmgren spheres , the ages of the quasars , the luminosity function of high redshift galaxies and its evolution , the spatial distribution of galaxies and its evolution , the biased distribution of galaxies around quasars and the anisotropy of quasar emission .
observations using keck - class telescopes may already be made to enable a differentiation between a fully neutral and a 10% neutral intergalactic medium at @xmath7 . |
the existence of local intersite pairs in cuprate superconductors is inferred from the very short coherence lengths in these materials , which is in turn determined from measurements of @xmath10 . given that the pair dimensions @xmath11 can not exceed the coherence length , i.e. @xmath12 , we may infer that any possible lattice distortions associated with pairing have a finite range , defined by the extent of the pair @xmath13 .
however , the particular form of the electron - phonon interaction leading to these distortions and particular type of phonon mode responsible for pairing should be determined from experiments . according to inelastic neutron scattering data
a large ( 30% ) anomaly in @xmath7 phonon mode corresponding to the ( @xmath14 ) direction has been observed in a number of doped cuprates ( but not in the parent compound)@xcite .
the observed anomaly takes place at an incomensurate wavevector and is suggested to be associated with pairing and/or stripe formation .
the electron phonon interaction can then be written in the form : @xmath15 where @xmath16 is a constant describing the strength of coupling , @xmath17defines the wavevector asociated with the interaction and its range in @xmath0-space , while @xmath18 defines its width . in this paper
we analyse the @xmath19 coupling of phonon modes with @xmath1using group theory applicable to la@xmath2sr@xmath3cuo@xmath4 and discuss the relevant phonon modes by examining the available experimental data from neutron diffraction , esr and exafs .
we find that the experimentally observed features can be described by the coupling of a @xmath7 - symmetry in - plane o - mode along the @xmath20 direction to holes occupying @xmath6 or @xmath5 - symmetry planar o states .
the brillouin zone ( bz ) corresponding to the tetragonal point group @xmath21 applicable for la@xmath2sr@xmath3cuo@xmath4 is shown in figure 1 . to consider local pairs and/or stripes forming along the cu - o bond direction or along 45@xmath22 to it , we need to consider the @xmath20 and the @xmath23 points , corresponding to the ( @xmath24 and ( @xmath25 directions respectively .
the special symmetry points ( @xmath26 and @xmath27 etc . ) give rise to commensurate stripes@xcite . before proceeding with analysis of the e - p coupling for the case of general @xmath28
, it should be pointed out since the symmetrised cross product of the representations at the @xmath29 point , @xmath30=[e_{g}\times e_{g}]=a_{1g}+b_{1g}+b_{2g}\ ] ] the electrons can couple only with @xmath31 modes ( there are no @xmath32 and @xmath33 at the @xmath29 point of la@xmath2sr@xmath3cuo@xmath4 ) .
( there are two such modes in @xmath21 and are associated either with apex oxygen o(2 ) ions or la ions . )
these do not lead to any symmetry breaking .
it follows that stripe formation or further symmetry breaking can not be associated with the @xmath29 point ( @xmath34 , but can only be _ associated with finite _ @xmath35 . for the case of tetragonal symmetry
, this corresponds to the @xmath36 and the @xmath23 points in the bz .
since the @xmath20 and @xmath23 points have a four pronged star , we propose the following form of coupling to electrons in single non - degenerate electronic states : @xmath37 where @xmath38 is the site label and @xmath39 where @xmath40 are the 4 wavevectors corresponding to the prongs of the star associated with stripe ( pair ) formation
. we should stress that in this case the nondegenerate electronic states are associated with @xmath41-orbitals of planar oxygens which transform as @xmath42 or @xmath43 .
the hamiltonian above on its own does not lead to symmetry breaking , but phase separation can in principle still take place as a result of competition of short - range attraction and long - range coulomb repulsion forces . however , the symmetry of both phases would be the same .
a more interesting case arises when two - fold degenerate levels ( for example the two @xmath5 states corresponding to the planar o @xmath44 and @xmath45 orbitals or the @xmath5 and @xmath6 states of the apical o ) interacts with @xmath46 phonons .
again , we are interested in the phonons which lead to symmetry breaking and allow the formation of stripes .
the invariant form of hamiltonian has a simple form for phonons which transforms as the @xmath7 representations of the group of wave - vector @xmath47 .
taking into account that @xmath6 and @xmath5 representations are real and pauli matrices @xmath48 corresponding to the doublet of @xmath6 or @xmath5 transform as @xmath49 ( @xmath50 ) for @xmath51 , @xmath32 ( @xmath52 ) for @xmath53 , @xmath33 ( @xmath54 ) for @xmath55 , and @xmath56 ( @xmath57 ) for @xmath58representations respectively , we can construct an invariant form of the hamiltonian : @xmath59 where @xmath60 the first term in ( 5 ) describes the symmetric coupling and is identical to the non - degenerate case .
the second and third terms describe the e - p interaction corresponding to the @xmath20 and @xmath23 directions respectively , while the last term describes the _ coupling to spins_. the proposed interaction ( 5 ) on its own results in splitting of the degenerate states ( a jahn - teller - like effect ) , breaking tetragonal symmetry and resulting in a local orthorhombic distortion .
it can therefore lead to the formation of pairs and stripes with no further interaction .
of course the stability and size of such a distortion will be determined by the balance of short - range attraction and long - range coulomb repulsion@xcite .
to determine the relevant wavevector @xmath40 applicable in ( 5 ) and ( 6 ) we need to look at the experimental data .
it is well known that for @xmath61 there are no significant phonon anomalies ( in raman or infrared spectroscopies for example ) which can be associated either with @xmath62 or the pseudogap @xmath63
. however , local probe experiments like neutron scattering@xcite , exafs@xcite and esr@xcite give a different picture , and point to the existence of sizeable lattice anomalies for @xmath64 .
the lattice distortions associated with these experimental observations are shown in figure 2 .
particularly the inelastic neutron data ( fig .
2a ) should be singled out here because of the unambiguity of the raw data . in these experiments
a large renormalisation of phonon frequency is ubiquitously observed along the @xmath20 direction in many cuprates , including 214 and 123 structure superconductors , but _ not _ in their parent insulators .
moreover , its temperature dependence clearly suggests a connection with pair formation@xcite .
the phonon @xmath7 modes corresponding to the @xmath20 direction which can couple to the electrons are shown in figure 3 .
one of these , o(1)1 can be clearly identified as the mode observed in neutron scattering experiments @xcite and may thus be indentified as the mode responsible for pairing or stripe formation .
we also note that from the experiments , the width in @xmath0-space of the anomaly is of the same order as the magnitude@xmath65 the interaction given in eqs.(5 ) and ( 6 ) leads directly to phase segregation of distorted and undistorted domains , where the size of domains is determined by the interplay of short - range attraction and screened coulomb repulsion as already mentioned@xcite .
the incommensurability of the interaction leads to modulation of charge density within the stripe .
importantly since the number of charged carriers in the stripe is less than the number of sites occupied by stripe , we may expect to observe itinerant behaviour in such stripes , i.e. the charge carriers are not localised within the stripe itself . at low doping when screening is reduced ,
the formation of the stripes of the length 2 ( pairs ) is most probable , while at lower temperatures and higher doping clearly longer stripes are more probable .
these pairs , contrary to the case of the holstein interaction is more extended in space and its size is governed by @xmath40 .
the central idea that the _ pair _ is the ground state entity which defines the extent of the e - p interaction leads to an interaction hamiltonian of degenerate modes with @xmath64 phonons which is suggested to describe the pairing and stripe formation in the cuprates .
the splitting of the degeneracy of planar o orbitals gives rise to a deformation either along the bond axes or along the diagonals , depending on which mode is chosen , but inelastic neutron scattering experiments suggest that the important interaction is along the cu - o bonds , giving rise to an @xmath66 - like form of the interaction as well as coupling to spins via an @xmath56-symmetry term .
remarkably , the interaction which splits the degeneracy of the o states via a jt - like effect is in spirit , if not in detail similar as was originally suggested by bednorz and muller in their original paper on la@xmath2ba@xmath3cuo@xmath4@xcite .
figure 1 . the brillouin zone of the tetragonal structure corresponding to la@xmath67sr@xmath3cuo@xmath68 according to experiment , the @xmath20 directions is of relevance for pairing and stripe formation .
a ) the wavevectors corresponding to the @xmath69representation at the @xmath70point of the point group @xmath21 . note that the relative magnitude of the displacements shown depends on the value of @xmath71 .
b ) the @xmath72 mode corresponding to the distortion suggested in ref . 4 ( fig . 2
b ) can also be viewed as a linear combination of two @xmath73 modes at the @xmath70point with different @xmath0 .
r.j.mcqueeney et al phys.rev.lett.82 628 ( 1999 ) and references therein .
the anomaly was originally reported by pinschovius et al .
, see for example physica , 162c-164c ( 1989 ) or ibid , 185c-189c 156 ( 1991 ) . | we consider finite-@xmath0 electron - phonon coupling appropriate for a superconductor with a very short coherence length and local intersite pairs . from group
theoretical analyis we find that @xmath1phonons can couple to degenerate electronic states in la@xmath2sr@xmath3cuo@xmath4 and a jahn - teller - like distortion is found to be operative for @xmath5 and @xmath6-symmetry o states . experimentally observed anomalies in inelastic neutron scattering , esr and exafs are found to be consistent with such an interaction of the @xmath7 symmetry @xmath8point mode with the planar o p@xmath3 and p@xmath9 states .
the proposed interaction is found to naturally facilitate symmetry breaking associated with pairing and/or stripe formation .
[ theorem]acknowledgement [ theorem]algorithm [ theorem]axiom [ theorem]claim [ theorem]conclusion [ theorem]condition [ theorem]conjecture [ theorem]corollary [ theorem]criterion [ theorem]definition [ theorem]example [ theorem]exercise [ theorem]lemma [ theorem]notation [ theorem]problem [ theorem]proposition [ theorem]remark [ theorem]solution [ theorem]summary |
the observed neutrino properties are in agreement with the minimal extension of the standard model ( sm ) resulting from the sole addition of neutrino masses @xcite .
neutrinos are massless within the sm because there are not neutrino singlet counterparts , higgs fields transform as an electroweak doublet , and the theory is renormalizable in the old sense . however , neutrino masses are generated relaxing any of these three conditions .
the addition of three right - handed ( rh ) neutrinos @xmath0 allows for arbitrary dirac neutrino masses after electroweak symmetry breaking , @xmath1 , @xmath2 similarly as for up quarks but with yukawa couplings @xmath3 much smaller .
this lepton number conserving ( lnc ) term and the corresponding neutrino mass matrix then provide the observed neutrino masses and charged current mixing upon diagonalisation , @xmath4 @xmath5 neutrino oscillations , which is the only manifestation of neutrino masses and mixing up to now , are well described by two mass splittings @xmath6 and the pmns mixing matrix @xcite , which equals @xmath7 when @xmath8 are written in the charged lepton mass eigenstate basis .
a global fit to available data gives @xcite @xmath9 the variation ranges correspond to 3@xmath10 errors , and the negative and positive @xmath11 figures stand for inverted and normal hierarchy , respectively .
there are excellent reviews on neutrino oscillation experiments @xcite and global fits @xcite , which are in good agreement within the available precision .
( we ignore lsnd data @xcite . )
note that these experiments do not distinguish between dirac and majorana masses , because in both cases the neutrino charged gauge interactions are given by eq .
( [ eq : chargedcurrent ] ) , and neutral gauge interactions also only involve left - handed ( lh ) neutrinos and are universal at lowest order @xcite ( see also @xcite ) . in this scenario and in the absence of other light fields further new physics can be parameterized by the corresponding effective lagrangian .
current limits on non - standard operators @xcite are presented in next section . with the minimal sm fermion content neutrinos
can have majorana masses if we add higgs triplets or we allow for higher order operators .
this second case in particular includes the first one when the higgs triplet is integrated out . at any rate
there is lepton number violation ( lnv ) , and neutrino masses result from the famous dimension five weinberg operator @xcite @xmath12 after electroweak symmetry breaking @xmath13 in this case the tiny neutrino masses are due to the very small operator coefficients @xmath14 multiplying @xmath15 , which are so minuscule because @xmath16 are extremely small ( @xmath17 for @xmath18 ) or @xmath19 is very large ( @xmath20 for @xmath21 ) . in the latter case , as @xmath19 is the effective scale in the lagrangian expansion , all higher order effects from higher dimensional operators are negligible .
hence , the phenomenologically relevant situation at lhc is the first one , with new physics near the tev scale and the dimensionless operator coefficients quite small on other grounds . in this set - up
there are two possibilities , too : that the see - saw messengers generating @xmath12 are near the electroweak scale , @xmath22 , or that the new physics near this scale does not mediate the see - saw mechanism , @xmath23 but is related to it within a given model .
we will discuss both cases in section [ majorananeutrinos ] . as already emphasized ,
majorana masses give the same neutrino oscillation predictions as dirac ones in this minimal sm extension with only @xmath12 .
but in this case the neutrino mass matrix is symmetric and @xmath24 in eq .
( [ eq : diracdiagonalization ] ) , which plays no rle in neutrino oscillations , is equal to @xmath25 ( see eq . ( [ op5 ] ) ) . independently of the neutrino mass character
any realistic model must reproduce the observed spectrum .
although many models can accommodate the values in eq .
( [ eq : limitsonmassesandmixing ] ) , there is no compelling , simple and predictive theory of lepton flavor .
but , as we emphasize in section [ models ] , in contrast with the quark sector lepton mixing is rather close to tri - bi - maximal mixing @xcite , what seems to indicate that a flavor symmetry slightly broken is at work .
neutrino masses are bounded to be less than 0.1 ev @xcite , thus they are very small compared to other mass parameters in the theory , and in particular to the electroweak scale @xmath26 gev .
this makes them unobservable in laboratory experiments where the relevant energies are much larger , and generically in experiments sensitive to electroweak interactions ranging from muon decay to particle collisions at lhc .
thus , the question is if light neutrinos have further observable interactions beyond their masses .
this can be answered considering the most general effective lagrangian up to the relevant dimension to be fixed by the available experimental accuracy , and fitting it to present data . in general
it is enough to go to the next order beyond the sm , typically up to dimension six .
however , the analysis depends , as do the extra operators , on the fields assumed to be light and the symmetries preserved .
hence , it does depend on the dirac or majorana neutrino mass character , or more precisely on whether the effective lagrangian involves or not rh neutrinos and on whether lepton number ( ln ) is conserved or not .
as emphasized above , neutrino oscillations and then light neutrino masses and mixing are compatible with any neutrino mass character , dirac or majorana .
hence , we will discuss them in turn .
the list of dimension six operators preserving the sm gauge symmetry and ln can be found in @xcite for operators not involving rh neutrino singlets . as a matter of fact
, the @xmath27 gauge symmetry alone implies that all dimension six operators involving only sm fields are lnc if they also conserve baryon number @xcite .
this list can be extended to include @xmath0 @xcite .
neutrino masses are the only vestige of lnv within the sm if they originate from the weinberg operator in eq .
( [ op5 ] ) .
as the neutrino mass scale is so small , it is appropriate to assume that new physics parameterized by dimension six operators involving only sm fields or light rh neutrinos is lnc .
limits on those operators for lh neutrinos can be found in @xcite , being typically at the per cent level in definite models and near the expected sensitivity in neutrino oscillation experiments .
model independent bounds can be one order of magnitude larger .
they are in general derived assuming only one new operator beyond the sm at a time . on the other hand , although requiring a precise cancellation ( thus at least two new dimension six operators , besides the extension of the operator set to include light rh neutrinos ) , there is still a ( small ) window for new interactions with observable effects in a near detector at a neutrino factory @xcite .
neutrino oscillations can not decide on the neutrino mass character without further interactions .
however , if the neutrino mass generation mechanism is mediated by new particles near the electroweak scale , these and the associated mechanism could be established at the lhc . this has been often reviewed for see - saw neutrino masses @xcite .
there are three different tree level particle exchanges generating the see - saw operator in eq .
( [ op5 ] ) .
the lepton and higgs doublets can couple to heavy fermions transforming as singlets , @xmath28 , or triplets , @xmath29 , under @xmath30 .
these are known as see - saw of type i @xcite and of type iii @xcite , respectively . on the other hand
, the two lepton doublets can couple to a scalar transforming as their symmetric product .
this means an @xmath30 scalar triplet , @xmath31 , what we refer to as see - saw of type ii @xcite .
if there is no further new physics at the tev than the see - saw mediators , the scalar and fermion triplets will be pair produced with electroweak strength at lhc for they transform non - trivially under @xmath32 . as a consequence ,
their discovery limits for an integrated luminosity of 30 fb@xmath33 at 14 tev are above half a tev ( see table [ tab : discoverylimits ] ) .
.lhc reach for see - saw mediators with an integrated luminosity of 30 fb@xmath33 at 14 tev . for a comparison with other heavy lepton sm additions giving multi - lepton signals
see @xcite . [ cols="<,<,<,<,<,<,<,<",options="header " , ] this is quite different from the heavy neutrino singlet case because @xmath28 can only decay through its mixing with the sm leptons @xmath34 , suppressing the electroweak cross - section for single production by the corresponding quadratic factor @xmath35 .
current limits on this mixing , @xmath36 @xcite , make them difficult to observe .
the lhc reach and the main signals for the three types of see - saw mechanisms are gathered in table [ tab : discoverylimits ] .
fermion singlets and triplets can be dirac ( d ) or majorana ( m ) , in which case events can be lnv as the samples to look at .
on the other hand , the lhc potential for scalar triplets depends on the neutrino mass hierarchy , normal ( nh ) or inverted ( ih ) , because this determines their coupling to @xmath37 leptons , which do not allow for an efficient scalar mass reconstruction @xcite .
several comments are in order .
the possibility of observing lnv events at large hadron colliders due to the production of heavy majorana neutrinos was emphasized long ago @xcite , but in the decay of extra charged gauge bosons in left - right models @xcite .
this is still a viable possibility and heavy neutrinos are observable at lhc , for the reference luminosity and energy above , up to @xmath28 masses @xmath38 tev for @xmath39 masses up to @xmath40 tev @xcite .
analogously , lhc can probe new neutral gauge boson masses up to 2.5 tev and @xmath28 masses up to 800 gev for a leptophobic @xmath41 @xcite .
hence , heavy neutrinos transforming trivially under the sm can be observed at lhc but as products of new interactions .
otherwise , the large backgrounds @xcite for lnv and lnc signals and the small mixings with sm leptons make their significance too low for discovery .
present limits on these mixings follow by comparison with electroweak precision data and rare flavor changing processes @xcite , being more stringent in particular for muons due to the better determination of the ckm mixing matrix , in very good agreement with the sm prediction @xcite . at any rate
, there is some debate about the naturalness of so large mixings for a heavy majorana neutrino because their size should be given by the see - saw value @xmath42 .
although large mixings are parametrically possible @xcite , they still require some ( accidental ) symmetry @xcite . finally ,
although lnv signals , as same sign di - lepton events , have much smaller backgrounds than the corresponding lnc ones , in this case opposite sign di - lepton events , this does not mean that lnc signals are in general less significant when samples with larger lepton multiplicities are taken into account @xcite .
alternatively , even if the interactions directly involved in the neutrino mass generation are too suppressed to manifest at large colliders , observables with a different physical origin can be related to neutrino masses in specific models .
in such a case lhc could also give new insights on neutrino masses .
for example , the flavor dependence of the neutralino decay rates to charged leptons can provide a determination of the corresponding neutrino mixing in specific supersymmetric models @xcite , or the observation of new vector - like lepton doublets decaying only to @xmath37 leptons can signal to a strongly coupled electroweak symmetry breaking sector , as recently shown in the context of a holographic composite higgs model @xcite .
we can learn on neutrino physics from lepton flavor changing processes @xcite , even if we do not observe new resonances at lhc .
these transitions require new lepton interactions not banished to very high energies to be sizeable .
this is the case , for instance , of sm extensions tackling the hierarchy problem , like supersymmetric , little higgs , or extra dimensional models @xcite . in general
they predict lepton flavor violating transitions at an observable level in forthcoming experiments @xcite .
let us comment on models of neutrino masses to conclude . in general they do accommodate their tiny scale compared with all other masses in the theory , their hierarchical splitting , and their mixing matrix quite close to tri - bi - maximal mixing @xcite , @xmath43 but there is no simple predictive model of such a pattern , on the other hand rather different from that for quarks , considered in general compelling .
lacking a theory of flavor , the challenge is often to make this pattern of neutrino masses compatible with a given class of models , for instance , constructed to solve the hierarchy problem or to unify the gauge interactions @xcite .
a popular solution is to realize the discrete symmetry @xmath44 on the leptonic sector @xcite , for it allows to enforce tri - bi - maximal mixing automatically , reducing the challenge to prove that the extended scalar sector breaks @xmath44 along the correct direction and that higher order corrections give small deviations from @xmath45 @xcite . in summary , there is still a lot of experimental and theoretical work ahead to have a satisfactory understanding of lepton masses and interactions . in particular , we expect to know if @xmath46 is different from 0 and cp violation in the leptonic sector is sizeable @xcite .
we are grateful to the corfu institute 2009 organizers for their habitual kind hospitality , and to j.a .
aguilar - saavedra , m. prez - victoria , r. pittau , j. wudka , and m. zraek for earlier collaboration .
this work has been partially supported by micinn project fpa2006 - 05294 and junta de andaluca projects fqm 101 , fqm 03048 . the work of j.s .
( a.c . ) has been partially supported by a micinn ( me ) ramn y cajal contract ( fpu fellowship ) . c. amsler _ et al . _
[ particle data group ] , phys .
b * 667 * , 1 ( 2008 ) and 2009 partial update for the 2010 edition . for monographs see : r. n. mohapatra and p. b. pal , massive neutrinos in physics and astrophysics . second edition , world sci .
notes phys .
* 60 * , 1 ( 1998 ) [ world sci .
notes phys .
* 72 * , 1 ( 2004 ) ] ; m. fukugita and t. yanagida , physics of neutrinos and applications to astrophysics ( springer - verlag , heidelberg , 2003 ) , 593 pp ; g. c. branco , l. lavoura and j. p. silva , cp violation , int .
phys . * 103 * , 1 ( 1999 ) .
b. pontecorvo , sov .
jetp * 6 * , 429 ( 1957 ) [ zh . eksp .
fiz . * 33 * , 549 ( 1957 ) 549 ] ; z. maki , m. nakagawa and s. sakata , prog .
* 28 * , 870 ( 1962 ) .
m. c. gonzalez - garcia and m. maltoni , phys .
rept . * 460 * , 1 ( 2008 ) [ arxiv:0704.1800 [ hep - ph ] ] .
b. kayser in @xcite ; v. barger , d. marfatia and k. whisnant , int .
j. mod .
e * 12 * , 569 ( 2003 ) [ arxiv : hep - ph/0308123 ] ; r. n. mohapatra and a. y. smirnov , ann .
nucl . part .
sci . * 56 * 569 ( 2006 ) [ arxiv : hep - ph/0603118 ] ; a. strumia and f. vissani , arxiv : hep - ph/0606054 . m. maltoni , t. schwetz , m. a. tortola and j.w.f .
valle , new j. phys .
* 6 * , 122 ( 2004 ) [ arxiv : hep - ph/0405172 ] ; g. l. fogli , e. lisi , a. marrone and a. palazzo , prog .
phys . * 57 * , 742 ( 2006 ) [ arxiv : hep - ph/0506083 ] ; m. c. gonzalez - garcia , m. maltoni and j. salvado , arxiv:1001.4524 [ hep - ph ] . for
a useful tool see p. huber , j. kopp , m. lindner and w. winter , pos * nufact08 * , 145 ( 2008 ) .
a. aguilar _ et al . _
[ lsnd collaboration ] , phys .
d * 64 * , 112007 ( 2001 ) [ arxiv : hep - ex/0104049 ] .
f. del aguila , j. syska and m. zralek , phys .
d * 76 * , 013007 ( 2007 ) [ arxiv : hep - ph/0702182 ] .
j. w. f. valle , phys .
b * 199 * , 432 ( 1987 ) ; m. c. gonzalez - garcia _ et al .
_ , phys .
* 82 * , 3202 ( 1999 ) [ arxiv : hep - ph/9809531 ] ; s. bergmann , y. grossman and e. nardi , phys .
d * 60 * , 093008 ( 1999 ) [ arxiv : hep - ph/9903517 ] ; n. fornengo , m. maltoni , r. t. bayo and j.w.f .
valle , phys .
d * 65 * , 013010 ( 2002 ) [ arxiv : hep - ph/0108043 ] ; m. blennow , t. ohlsson and w. winter , eur .
j. c * 49 * , 1023 ( 2007 ) [ arxiv : hep - ph/0508175 ] ; j. kopp , m. lindner , t. ota and j. sato , phys .
d * 77 * , 013007 ( 2008 ) [ arxiv:0708.0152 [ hep - ph ] ] .
s. antusch , c. biggio , e. fernandez - martinez , m. b. gavela and j. lopez - pavon , jhep * 0610 * , 084 ( 2006 ) [ arxiv : hep - ph/0607020 ] ; a. abada , c. biggio , f. bonnet , m. b. gavela and t. hambye , jhep * 0712 * , 061 ( 2007 ) 061 [ arxiv:0707.4058 [ hep - ph ] ] ; m. b. gavela , d. hernandez , t. ota and w. winter , phys .
d * 79 * , 013007 ( 2009 ) [ arxiv:0809.3451 [ hep - ph ] ] ; c. biggio , m. blennow and e. fernandez - martinez , jhep * 0908 * , 090 ( 2009 ) [ arxiv:0907.0097 [ hep - ph ] ]
. f. del aguila , j. de blas , r. szafron , j. wudka and m. zralek , phys .
b * 683 * , 282 ( 2010 ) [ arxiv:0911.3158 [ hep - ph ] ] ; p. langacker and d. london , phys .
d * 39 * , 266 ( 1989 ) . for
a neutrino factory study see a. bandyopadhyay _ et al .
_ [ iss physics working group ] , rept . prog .
phys . * 72 * , 106201 ( 2009 ) [ arxiv:0710.4947 [ hep - ph ] ] .
s. weinberg , phys .
lett . * 43 * , 1566 ( 1979 ) .
p. f. harrison , d. h. perkins and w. g. scott , phys .
b * 530 * , 167 ( 2002 ) [ arxiv : hep - ph/0202074 ] .
u. seljak , a. slosar and p. mcdonald , jcap * 0610 * , 014 ( 2006 ) [ arxiv : astro - ph/0604335 ] ; see also a. osipowicz _ et al .
_ [ katrin collaboration ] , arxiv : hep - ex/0109033 .
w. buchmuller and d. wyler , nucl .
b * 268 * , 621 ( 1986 ) .
a. de gouvea and j. jenkins , phys .
d * 77 * , 013008 ( 2008 ) [ arxiv:0708.1344 [ hep - ph ] ] .
f. del aguila , s. bar - shalom , a. soni and j. wudka , phys .
b * 670 * , 399 ( 2009 ) [ arxiv:0806.0876 [ hep - ph ] ] .
f. del aguila , j.a .
aguilar - saavedra and r. pittau , j. phys .
ser . * 53 * , 506 ( 2006 ) [ hep - ph/0606198 ] .
f. del aguila , j.a .
aguilar - saavedra and j. de blas , acta phys .
2901 ( 2009 ) [ arxiv:0910.2720[hep - ph ] ] . f. del aguila _
_ , eur .
j. c * 57 * , 183 ( 2008 ) [ arxiv:0801.1800 [ hep - ph ] ] .
a. atre , t. han , s. pascoli and b. zhang , jhep * 0905 * , 030 ( 2009 ) [ arxiv:0901.3589 [ hep - ph ] ] .
p. minkowski , phys .
b * 67 * , 421 ( 1977 ) .
m. gell - mann , p. ramond and r. slansky , in : supergravity , edited by p. van nieuwenhuizen and d.z .
freedman ( north holland , amsterdam , 1979 ) , p.315 ; s.l .
glashow , in : quarks and leptons , cargese , edited by m. levy et al .
( plenum , new - york , 1980 ) , p. 707 ; t. yanagida , workshop on the unified theory and baryon number in the universe , edited by o. sawada and a. sugamoto ( kek report 79 - 18 , tsukuba , japan , 1979 ) , p.95 ; r. n. mohapatra and g. senjanovic , phys . rev .
lett . * 44 * 912 ( 1980 ) .
r. foot , h. lew , x. g. he and g. c. joshi , z. phys .
c * 44 * , 441 ( 1989 ) ; e. ma , phys .
81 * , 1171 ( 1998 ) [ hep - ph/9805219 ] .
m. magg and c. wetterich , phys .
b * 94 * , 61 ( 1980 ) ; t. p. cheng and l. f. li , phys .
d * 22 * , 2860 ( 1980 ) ; g. b. gelmini and m. roncadelli , phys .
b * 99 * , 411 ( 1981 ) ; g. lazarides , q. shafi and c. wetterich , nucl .
b * 181 * , 287 ( 1981 ) ; r. n. mohapatra and g. senjanovic , phys .
d * 23 * , 165 ( 1981 ) .
aguilar - saavedra , nucl .
b * 828 * , ( 2010 ) 289 [ arxiv:0905.2221 [ hep - ph ] ] .
f. del aguila and j.a .
aguilar - saavedra , phys .
b * 672 * , 158 ( 2009 ) [ arxiv:0809.2096 [ hep - ph ] ] .
f. del aguila , j.a .
aguilar - saavedra and r. pittau , jhep * 0710 * , 047 ( 2007 ) [ hep - ph/0703261 ] . f. del aguila and j.a .
aguilar - saavedra , nucl .
b * 813 * , 22 ( 2009 ) [ arxiv:0808.2468 [ hep - ph ] ] .
a. datta , m. guchait and a. pilaftsis , phys .
d * 50 * , 3195 ( 1994 ) [ hep - ph/9311257 ] ; f. m. l. almeida , y. a. coutinho , j.a .
martins simoes and m. a. b. do vale , phys .
d * 62 * , 075004 ( 2000 ) [ hep - ph/0002024 ] ; o. panella , m. cannoni , c. carimalo and y. n. srivastava , phys .
d * 65 * , 035005 ( 2002 ) [ hep - ph/0107308 ] ; t. han and b. zhang , phys .
lett . * 97 * , 171804 ( 2006 ) [ hep - ph/0604064 ] ; s. bray , j. s. lee and a. pilaftsis , nucl .
b * 786 * , 95 ( 2007 ) [ hep - ph/0702294 ] .
k. huitu , j. maalampi , a. pietila and m. raidal , nucl .
b * 487 * , 27 ( 1997 ) [ hep - ph/9606311 ] ; j. f. gunion , c. loomis and k. t. pitts , [ hep - ph/9610237 ] ; a. g. akeroyd and m. aoki , phys .
d * 72 * , 035011 ( 2005 ) [ hep - ph/0506176 ] ; a. hektor , m. kadastik , m. muntel , m. raidal and l. rebane , nucl .
b * 787 * , 198 ( 2007 ) [ arxiv:0705.1495 [ hep - ph ] ] ; p. fileviez perez , t. han , g. y. huang , t. li and k. wang , phys .
d * 78 * , 071301 ( 2008 ) [ arxiv:0803.3450 [ hep - ph ] ] ; phys .
d * 78 * 015018 ( 2008 ) [ arxiv:0805.3536 [ hep - ph ] ] .
see also e. j. chun , k. y. lee and s. c. park , phys .
b * 566 * , 142 ( 2003 ) [ hep - ph/0304069 ] ; j. garayoa and t. schwetz , jhep * 0803 * , 009 ( 2008 ) [ arxiv:0712.1453 [ hep - ph ] ] ; m. kadastik , m. raidal and l. rebane , phys . rev .
d * 77 * , 115023 ( 2008 ) [ arxiv:0712.3912 [ hep - ph ] ] ; a. g. akeroyd , m. aoki and h. sugiyama , phys .
d * 77 * , 075010 ( 2008 ) [ arxiv:0712.4019 [ hep - ph ] ] ; a. g. akeroyd and c. w. chiang , phys .
d * 80 * , 113010 ( 2009 ) [ arxiv:0909.4419 [ hep - ph ] ] .
r. franceschini , t. hambye and a. strumia , phys .
d * 78 * , 033002 ( 2008 ) [ arxiv:0805.1613 [ hep - ph ] ] ; a. arhrib , b. bajc , d. k. ghosh , t. han , g. y. huang , i. puljak and g. senjanovic , arxiv:0904.2390 [ hep - ph ] ; t. li and x. g. he , phys .
d * 80 * , 093003 ( 2009 ) [ arxiv:0907.4193 [ hep - ph ] ] .
f. del aguila , j. de blas and m. perez - victoria , phys .
d * 78 * , 013010 ( 2008 ) [ arxiv:0803.4008 [ hep - ph ] ] . see for previous limits p. langacker and d. london , phys .
d * 38 * , 907 ( 1988 ) ; e. nardi , e. roulet and d. tommasini , phys .
b * 327 * , 319 ( 1994 ) [ hep - ph/9402224 ] ; s. bergmann and a. kagan , nucl .
b * 538 * , 368 ( 1999 ) [ hep - ph/9803305 ] ; b. bekman , j. gluza , j. holeczek , j. syska and m. zralek , phys .
d * 66 * , 093004 ( 2002 ) [ hep - ph/0207015 ] .
d. tommasini , g. barenboim , j. bernabeu and c. jarlskog , nucl .
b * 444 * , 451 ( 1995 ) [ hep - ph/9503228 ] ; j. g. korner , a. pilaftsis and k. schilcher , phys .
b * 300 * , 381 ( 1993 ) [ hep - ph/9301290 ] ; a. ilakovac and a. pilaftsis , nucl .
b * 437 * , 491 ( 1995 ) [ hep - ph/9403398 ] ; j. i. illana and t. riemann , phys .
d * 63 * , 053004 ( 2001 ) [ hep - ph/0010193 ] .
w. y. keung and g. senjanovic , phys .
lett . * 50 * , 1427 ( 1983 ) .
j. c. pati and a. salam , phys .
d * 10 * , 275 ( 1974 ) [ erratum - ibid . * 11 * , 703 ( 1975 ) ] ; r. n. mohapatra and j. c. pati , phys . rev .
d * 11 * , 2558 ( 1975 ) ; g. senjanovic and r. n. mohapatra , phys .
d * 12 * , 1502 ( 1975 ) .
a. ferrari et al . , phys .
d * 62 * , 013001 ( 2000 ) ; s. n. gninenko , m. m. kirsanov , n. v. krasnikov and v. a. matveev , phys .
nucl . * 70 * , 441 ( 2007 ) .
f. del aguila and j.a .
aguilar - saavedra , jhep * 0711 * , 072 ( 2007 ) [ arxiv:0705.4117 [ hep - ph ] ] .
k. huitu , s. khalil , h. okada and s. k. rai , phys .
* 101 * , 181802 ( 2008 ) [ arxiv:0803.2799 [ hep - ph ] ] ; p.
fileviez perez , t. han and t. li , phys .
d * 80 * , 073015 ( 2009 ) [ arxiv:0907.4186 [ hep - ph ] ] .
z. sullivan and e. l. berger , phys .
d * 78 * , 034030 ( 2008 ) [ arxiv:0805.3720 [ hep - ph ] ] .
f. del aguila , j.a .
aguilar - saavedra , j. de blas and m. zralek , acta phys .
b * 38 * , 3339 ( 2007 ) [ arxiv:0710.2923 [ hep - ph ] ] ; x. g.
he , s. oh , j. tandean and c. c. wen , phys .
d * 80 * , 073012 ( 2009 ) [ arxiv:0907.1607 [ hep - ph ] ] .
w. buchmuller and c. greub , nucl .
b * 363 * , 345 ( 1991 ) ; g. ingelman and j. rathsman , z. phys .
c * 60 * , 243 ( 1993 ) ; f. del aguila , j.a .
aguilar - saavedra , a. martinez de la ossa and d. meloni , phys .
b * 613 * , 170 ( 2005 ) [ arxiv : hep - ph/0502189 ] ; a. pilaftsis and t. e. j. underwood , phys .
d * 72 * , 113001 ( 2005 ) [ hep - ph/0506107 ] ; j. kersten and a. y. smirnov , phys .
d * 76 * , 073005 ( 2007 ) [ arxiv:0705.3221 [ hep - ph ] ] . see for quasi - dirac neutrinos d. wyler and l. wolfenstein , nucl .
b * 218 * , 205 ( 1983 ) ; r. n. mohapatra and j.w.f .
valle , phys .
d * 34 * , 1642 ( 1986 ) ; for a review and further references e. ma , arxiv:0908.1770 [ hep - ph ] .
w. porod , m. hirsch , j. romao and j.w.f .
valle , phys .
d * 63 * , 115004 ( 2001 ) [ hep - ph/0011248 ] .
f. del aguila , a. carmona and j. santiago , arxiv:1001.5151 [ hep - ph ] ; see also c. csaki , c. delaunay , c. grojean and y. grossman , jhep * 0810 * , 055 ( 2008 ) [ arxiv:0806.0356 [ hep - ph ] ] . m. raidal
_ et al . _ ,
eur . phys .
j. c * 57 * , 13 ( 2008 ) [ arxiv:0801.1826 [ hep - ph ] ] .
see for a review a. j. buras , arxiv:0910.1032 [ hep - ph ] .
a. maki , aip conf .
* 981 * , 363 ( 2008 ) .
i. de medeiros varzielas , s. f. king and g. g. ross , phys .
b * 648 * , 201 ( 2007 ) [ arxiv : hep - ph/0607045 ] .
e. ma and g. rajasekaran , phys .
d * 64 * , 113012 ( 2001 ) [ arxiv : hep - ph/0106291 ] . see for a review g. altarelli and f. feruglio , arxiv:1002.0211 [ hep - ph ] . | we briefly summarise the current status of neutrino masses and mixing , paying special attention to the prospects for observing new leptonic interactions . |
low energy supersymmetry ( susy)@xcite has been playing a central role in the quest for physics beyond the standard model ( sm ) . since phenomenological consistency requires susy to be broken , and broken softly ( so as not to reintroduce any quadratic divergence ) , the lagrangian of the minimal supersymmetric standard model ( mssm)@xcite includes soft and gauge invariant susy breaking terms . while the generic mssm lagrangian may contain many arbitrary soft terms , specific models for susy breaking
have been proposed that provide relationships between the mssm parameters . incorporating well - motivated new interactions and particles at high mass scales
, such scenarios drastically reduce the large number of unknown parameters in the mssm to only a few , thereby making the model more predictive .
we will focus here only on supergravity ( sugra)@xcite type of models where susy is considered as a local symmetry .
these models incorporate a hidden sector wherein susy is broken , and a visible sector where the mssm fields reside and to which the breaking is communicated by gravitational interactions . in @xmath4 sugra , which incorporates grand unification , one has a choice of three functions in building a model@xcite , namely the gauge kinetic energy function @xmath5 , the khler potential @xmath6 , and the superpotential @xmath7 , where @xmath8 refer to matter fields . in msugra , the minimal version of the model , one has a flat khler potential and a flat gauge kinetic energy function .
the corresponding soft susy breaking sector is characterized by only a few parameters , normally specified at the scale of the grand unified theory ( gut ) viz .
@xmath9 gev@xcite .
these are the universal gaugino mass @xmath10 , the universal scalar mass @xmath11 , the universal trilinear coupling @xmath1 and the universal bilinear coupling @xmath12 .
in addition to these , there is a superpotential parameter , namely the higgs mixing term @xmath13 . unlike in the sm , where the breaking of the electroweak symmetry necessitates the explicit introduction of a negative valued scalar mass - squared , in a generic sugra model , the said breaking can be realized even for a positive mass - squared term in the bare lagrangian , thanks to radiative corrections@xcite . in other words ,
the renormalization of the soft susy breaking terms as one moves from the unification scale down to the electroweak scale automatically engenders a negative mass - squared thereby breaking the symmetry@xcite . in a similar vein
, the low energy parameters of the mssm ( which are quite large in number ) are obtained from only a few unification scale parameters via the renormalization group equations ( rge)@xcite integrated from @xmath14 to the electroweak scale ( @xmath15 ) .
the two minimization conditions for the higgs potential then eliminate @xmath13 ( except for its sign ) on the one hand , and , on the other , relate @xmath12 to @xmath16 ( @xmath17 ) , the ratio of higgs vacuum expectation values .
thus msugra may be characterized by @xmath16 , @xmath18 , @xmath11 , @xmath1 and sign(@xmath19 ) and @xmath1 follows the standard convention of ref@xcite . ] . with all the low energy parameters of the mssm
being generated in terms of these few parameters , one has a considerable amount of predictivity for the mssm spectrum .
a different problem remains though , namely that of the susy cp violating phases .
many phases of sugra models can be rotated away . in an universal scenario like msugra
, the gaugino masses can be considered real with the result that only two combinations of phases ( beyond the cabibbo - kobayashi - maskawa quark mixing ( ckm ) phase already present in the sm ) are physical . a convenient choice for the two is given by @xmath20 for @xmath1 ( at @xmath14 ) and @xmath21 for the @xmath0-parameter at the electroweak scale .
it should be noted though that many analyses prefer to work with @xmath22 , the phase of @xmath19 , instead of @xmath21 .
an advantage of this latter choice is that @xmath23 since @xmath22 does not run up to the one - loop level .
these different descriptions can be understood in terms of @xmath24 and @xmath25 ( peccei - quinn ) symmetries and the choice of reparametrization invariant combinations of phases , a discussion of which may be found in refs.@xcite .
a selection of past analyses using @xmath21 as an input parameter may be seen in refs.@xcite . here
we note that a choice of @xmath21 instead of @xmath22 as a phase parameter makes the entire set of input parameters to be of soft - breaking origin . a few important points need to be noted in the context of the susy cp problem .
the latter arises from the fact that the phases are highly constrained by the experimental limits on the electric dipole moments ( edm ) of the electron and the neutron@xcite .
consequently , we are forced to admit one of the three eventualities : 1 .
the phase @xmath21 is very small@xmath26 or @xmath27if the superpartners are not considered to be very heavy may reach up to @xmath28 in the focus point zone@xcite . ] .
in addition , the phases of the @xmath29-parameters at the electroweak scale are also constrained . in msugra with phases ,
the requirement of having a very small @xmath21 typically translates into a relatively large but _ highly fine - tuned _ value for @xmath30 ( i.e. , @xmath0 at @xmath14 ) .
this , in turn , constrains the phase @xmath20 of @xmath1 , although to a somewhat lesser degree . the fact that the issue of fine - tuning in phases at the gut scale arises out of the combined requirement of satisfying the edm constraints and the radiative electroweak symmetry breaking was discussed in great detail in refs.@xcite as well as in refs.@xcite . in this paper
we try to focus our attention on this problem by looking at suitable models beyond msugra that can have unique features in the evolution of @xmath0 .
2 . the phases are large and less fine - tuned but the sparticles are massive .
of course , fully ameliorating the susy cp problem in this fashion requires that the sfermions be super - massive , thereby aggravating the problem of the little mass hierarchy in the higgs sector .
we will investigate whether the amount of fine - tuning can be reduced even while one considers a lighter sparticle spectra .
3 . finally there is the possibility that the susy breaking parameters may have special pockets where there can be a large amount of internal cancellations between the diagrams contributing to the electric dipole moments of electron and neutron@xcite .
this means that phases could be large while sparticle masses are significantly light .
this scenario is highly parameter dependent and clearly depends on very delicate cancellations .
hence we will not include this in our work while trying to focus on generic behaviors . as mentioned above
, we would like to address the first and the second issues in this analysis .
we are particularly interested in exploring the possible role of non - universal gaugino masses ( nugm ) in reducing the fine - tuning in the phase @xmath31 .
to quantify the latter , we consider a _ naturalness like _ measure of the form = [ 0/_b]__b 0 .
[ sensitivityphi ] a large value for @xmath32 would mean a lesser degree of fine - tuning of @xmath31 with respect to a variation in @xmath21 satisfying the edm constraints .
the phase - derivative is evaluated at @xmath33 with the choice being dictated by the fact that the edm constraints force @xmath34 to be close to zero .
thus , this is a restrictive definition compared to the type of fine - tuning defined in ref.@xcite .
we will see that the issue of such fine - tuning of phase can be addressed by focusing on scenarios where there is a large evolution of the bi - linear higgs coupling parameter @xmath0 between the electroweak scale and the gut scale .
the evolution of @xmath0 depends on the @xmath35 and the @xmath36 gaugino masses , the trilinear couplings and @xmath16 . within msugra , in addition to the evolution of @xmath37 being typically small
, the phase @xmath31 also turns out to be quite fine - tuned ( i.e. @xmath32 tends to be small ) .
in other words , for a given @xmath31 satisfying the edm constraints , the variation @xmath38 that still is consistent with the constraints is generally much smaller than the variation @xmath39 allowed at the electroweak scale@xcite .
as we will see , the evolution in @xmath37 may be enhanced by appropriate mass relationships between the gauginos that are away from universality at @xmath14 . at the same time
, these would help in reducing the above - mentioned fine - tuning so that @xmath32 can be significantly increased in specific nugm scenarios .
we , however , desist from choosing an arbitrary non - universal gaugino mass scenario since that will introduce new phases@xcite . as we will see in sec.[nugmdesc ] , non - universalities in gaugino masses may originate from a non - trivial gauge kinetic energy function .
the latter is a function of chiral superfields and transforms as a symmetric product of the adjoint representations of the underlying gauge group .
this leaves @xmath40 with the possibility of being in one or more of several representations , one of which is the singlet .
while the choice of the singlet corresponds to msugra , the non - singlet representations give rise to non - universalities in the gaugino masses .
it is possible to identify a suitable non - singlet representation in isolation ( _ i.e. _ , we will not combine a non - singlet representation with the singlet or other non - singlet representations ) whose gaugino mass pattern is effective in generating a large evolution in @xmath0 . at the same time
, there will be no additional phases to worry about since the overall phase of the gaugino masses can be rotated away in a fashion similar to that in msugra . in this paper
, we will analyze the consequences of a large evolution of the @xmath0-parameter ( mostly in the presence of such non - universalities ) on the cp violating phases . here , the basic input parameters are @xmath16 , @xmath11 , @xmath10 ( providing with definite nugm patterns ) , @xmath41 along with its phase @xmath20 and the phase @xmath21 of @xmath0 given at the electroweak scale ( @xmath15 ) .
note that @xmath37 at the electroweak scale is obtained via radiative electroweak symmetry breaking ( rewsb ) condition .
subsequently , @xmath42 , the gut scale magnitude of the @xmath0-parameter along with its phase @xmath31 is obtained via rges .
we will identify broad but correlated regions of parameter space where there can be a significant degree of reduction of the phase sensitivity while going from msugra to a type of nugm models .
the paper is organized as follows . in sec.[nugmdesc ] , we discuss the non - universal gaugino mass models .
the study of the relevant contributions from different sectors in the associated rges of @xmath0 and @xmath29 parameters allows us to identify the non - singlet representations which provide with a large evolution in @xmath0 .
we will probe the parameter space that is suitable for reducing the amount of fine - tuning in the cp violating phases . in sec.[results
] , we present the numerical results for the evolution of @xmath0 .
an analysis in the absence of phases points us to the favored regions of parameter spaces . on inclusion of phases , this facilitates the identification of the regions with significantly reduced level of fine - tuning , finally , we conclude in sec.[secconclusion ] .
non - universality in gaugino masses may originate from a non - trivial gauge kinetic energy function @xmath43 which , in turn , is a function of the chiral superfields in the theory . the indices @xmath44 run over the generators of the gauge group ( for example , @xmath45 for su(5 ) ) .
the gaugino mass matrix is given by @xmath46 where @xmath47- \kappa^2 k$ ] .
here , @xmath48 is the superpotential , @xmath49 is the khler potential , @xmath50 are the complex scalar fields , and @xmath51 gev@xmath52 with @xmath53 being newton s constant
. the functions @xmath43 may have non - trivial field contents , or in other words , may contain combinations of field transforming as either singlet or non - singlet irreducible representations@xcite . with the gauginos being majorana particles , @xmath43 , of necessity , must be contained in the symmetric product of the adjoint representations of the gauge group .
for example , in the case of su(5 ) , @xmath54 for the singlet case , one has @xmath55 which indeed leads to universality of gaugino masses .
similarly , the non - singlet representations will give rise to non - universal gaugino masses .
in general @xmath56 , where @xmath57 s give the relative weights of each contributing representation and @xmath58 , for the subgroup @xmath59 , are essentially the clebsch - gordan coefficients corresponding to the breaking by the adjoint higgs field@xcite . for the case of @xmath60 , the coefficients @xmath58
are displayed in table [ tabrelativeweights ] . clearly , the non - singlet representations have characteristic mass relationships for the gaugino masses at the gut scale .
past analyses exploring various phenomenological implications of such non - universality may be found in refs.@xcite .
as we shall argue later , the adjoint representation @xmath61 for @xmath40 ( nugm:24 in the notation of table [ tabrelativeweights ] ) is the most interesting one in the context of the present investigation .
consequently , we will analyze this case in isolation , or , in other words , assume that the sole contribution to @xmath43 is from a 24-plet structure .
apart from reducing the number of free parameters , this has the additional advantage that no new phase degree of freedom for the gaugino masses is introduced . with the gaugino mass ratios at the gut scale now being given by @xmath62 , for a positive gluino mass , the other two gaugino mass parameters are negative , a signature different from msugra .
this indeed would turn out to be useful in our quest . as mentioned earlier ,
we only consider either @xmath63 ( msugra ) or @xmath64 ( nugm:24 ) with all other @xmath57 s assumed to be zero .
.__the coefficients @xmath58 as pertaining to the @xmath65 , @xmath36 and @xmath35 gaugino masses at the gut scale for different representations of @xmath60 . _
_ [ cols="^,^,^,^,^",options="header " , ] an analogous analysis with so(10 ) as the underlying gauge group is also possible@xcite , though we will not investigate it in this paper .
similar to eq.[su5breakup ] here , one has @xmath66 .
if the symmetry breaking pattern is @xmath67 , one finds from the @xmath68-plet that @xmath69 .
this pattern is quite similar to nugm:24 as can be ascertained from table [ tabrelativeweights ] .
we would like to comment at this point that , in general , such non - universal gaugino mass scenarios change the gauge coupling unification conditions@xcite .
however , it is still possible to find specific conditions@xcite under which the usual gauge coupling unification condition remains unaltered and we consider this in our work .
note though that our results are quite robust and have very little dependence on the exact details of the spectrum .
we now identify the differences between msugra and nugm:24 in regard to the evolution of the @xmath0-parameter _ in the absence of cp violating susy phases_. this , in turn , will help us in understanding the evolution of @xmath21 upon the inclusion of the phases ( see refs.@xcite for past analyses discussing phase evolutions ) .
note that @xmath70 and @xmath0 are determined via the rewsb condition , viz .
\sin(2\beta ) & = & 2|b\mu|/(m_{h_d}^2+m_{h_u}^2 + 2\mu^2+\sigma_1+\sigma_2 ) \ , \label{eqnbewsb } \end{array}\ ] ] where @xmath72 represent the one - loop corrections @xcite .
the higgs scalar mass parameters @xmath73 and @xmath74 , and thereby @xmath70 and @xmath0 depend quite strongly on @xmath11 as well as on @xmath10 .
to one - loop order , the running of the @xmath0 parameter has two additive components , the first proportional to the gaugino masses and the second depending on a combination of the trilinear couplings and the yukawa couplings @xcite , namely , @xmath75 where @xmath76 with @xmath77 being the renormalization scale .
@xmath78 are the scaled gauge coupling constants ( with @xmath79 ) and @xmath80 for @xmath81 are the running gaugino masses .
furthermore , @xmath82 represent the squared yukawa couplings , e.g , @xmath83 where @xmath84 is the top yukawa coupling . in a similar vein ,
the evolution of the trilinear terms is given by @xmath85 \displaystyle { \frac{da_b}{{dt } } } & = & \displaystyle -(\frac { 16}{3}\tilde{\alpha}_3{\tilde{m}_3 } + 3\tilde{\alpha}_2{\tilde{m}_2 } + \frac { 7}{15}\tilde { \alpha}_1 { \tilde{m}_1 } ) -y_ta_t -6y_b a_b - y_\tau a_\tau \\[1.5ex ] \displaystyle { \frac{da_\tau}{{dt } } } & = & \displaystyle -(3\tilde{\alpha}_2{\tilde{m}_2 } + \frac { 9}{5}\tilde { \alpha}_1{\tilde{m}_1 } ) -3y_b a_b-4y_\tau a_\tau \ .
\label{eqnatrge } \end{array}\ ] ] for small @xmath16 , the contributions from the bottom quark and tau yukawa couplings @xmath86 and @xmath87 may be neglected , and the rges approximately integrated to obtain @xcite @xmath88 where @xmath89 with @xmath90 corresponding to the electroweak scale .
the functions @xmath91 and @xmath92 encapsulate the running of the gauge coupling constants , viz , @xmath93 f(t ) & \equiv & \displaystyle \int_{0}^{t } e(t ' ) dt ' \end{array}\ ] ] where @xmath94 and @xmath95 are the coefficients in the respective one - loop beta - functions . of course
, unification imposes the boundary condition that @xmath96 . at the top mass scale ( @xmath97 )
, @xmath98 is indeed a very good approximation .
the function @xmath99 , in eq.[eqnbzvsb0 ] , on the other hand , is given by @xmath100 where @xmath101 h_3(t ) & \equiv & \displaystyle \int_{0}^{t } { e(t ' ) h_2(t')}dt ' \\[2ex ] h_2(t ) & \equiv & \displaystyle { \tilde\alpha}(0 ) \left({\frac{16}{3 } } h_3 + 3h_2 + { \frac{13 } { 15 } } h_1 \right ) \ .
\end{array}\ ] ] for the generic ( nugm ) case , the above results remain the same except that @xcite h_i(t ) _
i(t ) h_i(t ) .
note that , in @xmath102 , the gaugino contribution is positive for msugra , but negative for nugm:24 .
thus , it is useful to understand the nature of evolution of trilinear couplings in either scenario so as to evaluate their role in the evolution of @xmath0 . for the msugra case ,
the gaugino contributions to @xmath103 are always negative ( vide eq.[eqnatrge ] ) .
hence , it is obvious that if @xmath1 not be too large , then @xmath104 would typically turn negative by the electroweak scale . in fact , the large gluino contributions render both @xmath105 and @xmath106 negative well above the electroweak scale .
this implies , that in this case ( msugra ) , the two pieces in @xmath102 would tend to cancel each other , an effect also manifested by the smallness of @xmath107 in eq.[eqnbzvsb0 ] . in turn , this leads to a small value for @xmath108 in msugra . comparing the evolution of the trilinear terms in nugm:24 with that in msugra ,
it turns out that a qualitative difference arises only in the case of @xmath109 , while for @xmath105 and @xmath106 the difference between the scenarios is only a quantitative one .
this is easy to understand given the overwhelming dominance , in the last two cases , of the gluino contribution over those from the electroweak gauginos .
specifically , for @xmath110 , @xmath109 at the weak scale comes to be negative for msugra while it is positive ( with usually a larger magnitude ) for nugm:24 .
given the relative weights of the @xmath104 terms in eq.[eqnbrge ] , it is thus quite apparent that the total contribution from the trilinear couplings to the evolution of @xmath0 is quite similar in the two models . on the other hand , since the signs of @xmath111 are reversed in nugm:24 , the aforementioned cancellations in @xmath102 would no longer be operative ; rather , the different contributions would enhance each other leading to a large @xmath112 .
this is the very reason why we choose to concentrate on models like nugm:24 .
we note in passing that although the rge for @xmath0 does not explicitly include the su(3 ) gaugino mass , it implicitly depends on the latter via the contributions from trilinear couplings .
we now discuss the dependence of @xmath0 and @xmath12 on @xmath11 and the other parameters .
being obtained from the rewsb condition of eq.[eqnbewsb ] , @xmath0 ( and hence @xmath12 ) evidently depends on @xmath11 quite strongly .
the structure of eq.[eqnbrge ] suggests that , to one - loop order , @xmath112 should not depend on @xmath11 .
however , a subsidiary dependence arises through the determination of the scale at which the minimizations of higgs potential ( i.e. rewsb ) is to be performed .
canonically , this scale is determined by demanding that the contribution , to @xmath70 , of the 1-loop correction terms of the effective potential be small . in our analysis
this scale is approximately halfway between the lowest and highest mass of the spectra and , generally , is not very far from the average stop mass scale @xmath113 ( see ref.@xcite ) .
since this scale does depend on @xmath11 , it leads to a small dependence in @xmath112 as well by virtue of being a limit of integration for the rges . even on inclusion of phases for the @xmath29 and @xmath0 parameters , the rges
formally remain the same as in eqs.([eqnbrge]&[eqnatrge ] ) . the evolution of the phases
can then be extracted by comparing the real and imaginary parts of the said equations . clearly , unlike in the case of the real parts , the imaginary parts of the beta functions for @xmath29 s and @xmath0
do not depend on the gaugino masses and hence there is no cancellation between the different contributions .
furthermore , even a vanishing @xmath31 can lead to a non - zero @xmath21 provided @xmath1 has a non - trivial phase .
for example , in the small @xmath16 limit , the explicit analytical solution gives @xmath115 we examine now the interdependence between the phases , their evolution ( also see ref.@xcite ) and the phase sensitivity @xmath32 for different values of @xmath16 and other parameters both within msugra as well as nugm:24 .
as we have already mentioned , the edm constraints limit @xmath21 to be tiny ( @xmath116 , and typically much smaller ) .
now , if either of @xmath41 or @xmath20 is small ( actually , if @xmath117 ) , then @xmath31 would be determined essentially by @xmath37 , @xmath42 and @xmath21 . in this case
, @xmath20 would be quite unconstrained .
the dependence on @xmath16 is crucial and is best understood by considering the two opposite limits , namely small and large values : * for a small @xmath16 ( @xmath118 or so ) , @xmath119 is large , and therefore @xmath37 is appreciably large ( see eq.[eqnbewsb ] ) .
within msugra , for not too large a value of @xmath41 , the gut scale value @xmath42 is then quite comparable to @xmath37 .
this can be understood by recognizing the cancellations between the various terms in eq.[eqnforc ] that keeps @xmath107 small and thereby keep @xmath120 relatively small ( courtesy eq.[eqnbzvsb0 ] ) .
consequently , in such a scenario , @xmath31 is not too different from @xmath21 .
this remains true even for @xmath121 which maximizes the edm values@xcite .
+ on the contrary , the situation in nugm:24 is quite different . here
, a larger difference between @xmath37 and @xmath42 is generated by the enhancement in @xmath107 .
consequently , @xmath31 becomes appreciably different from ( and numerically larger than ) @xmath21 .
* for a large value of @xmath16 , on the other hand , @xmath122 is quite small .
thus , unless @xmath123 is extremely tiny ( as happens , for example , in hyperbolic branch / focus point@xcite scenarios ) , @xmath37 is constrained to be small and has only sub - dominant influence on the evolution of @xmath21 .
this , in turn , implies that the value of @xmath31 becomes strongly correlated with that of @xmath20 .
in other words , a high degree of fine - tuning in one will necessitate a similar degree of fine - tuning in the other .
we now focus on the issue of phase sensitivity . as eq.[eqnbimag ] suggests , the range allowed to @xmath21 ( i.e. @xmath39 ) imposes rather strong limits in the @xmath31@xmath20 plane . adopting the measure of phase naturalness @xmath32 ( as espoused in eq.[sensitivityphi ] ) , one may estimate , from eq.[eqnbimag ] , the amount of fine - tuning associated with the phase @xmath31 .
now , as the rges suggest , the implicit dependence of @xmath32 on @xmath1 occurs primarily through the dependence of @xmath12 itself on @xmath1 .
thus , to the leading order , one has an approximate relation of the form@xcite @xmath124 we would like to point out that although the above simplification ( as also those of neglecting @xmath86 and @xmath87 ) is quite illustrative , we do not take recourse to it .
rather we solve the complete set of rges numerically and also compute @xmath32 numerically directly from its definition ( eq.[sensitivityphi ] ) .
note that , as obtained from eq.[sensitivityphi ] and the first of eqs.[eqnbimag ] , the measure @xmath32 actually involves a factor of @xmath125 in the denominator .
this causes @xmath32 to be very large when @xmath126 is close to @xmath127 , as also a change of sign for @xmath32 when @xmath126 crosses @xmath127 .
we will see that this is indeed the case for nugm:24 where @xmath126 can easily cross @xmath127 owing to a large degree of phase evolution . in the msugra scenario , on the other hand ,
such a feature rarely appears .
as we have already discussed , msugra is associated with a relatively small degree of evolution in @xmath0 , and hence @xmath128 .
this leads to a low value of @xmath129 or , equivalently , to a high degree of fine - tuning in @xmath126 .
on the other hand , a non - universal gaugino mass scenario like nugm:24 can provide us with a large evolution of @xmath37 .
this , of course , can generate either @xmath130 or @xmath131 .
the parameter space corresponding to the latter case ( which is typically satisfied better for smaller @xmath16 zones ) reduces fine - tuning in @xmath31 .
we will see that the said reduction can be as large as a factor of 10 to 20 compared to msugra . and
finally , the very same large evolution of @xmath37 also implies that @xmath132 could be a possibility within such scenarios . in nugm:24 where the evolution of @xmath0 is large
, the above reduction of @xmath42 toward zero is possible when @xmath37 is large i.e. when @xmath16 is small . in msugra
too this is possible , but only to a limited degree , as the aforesaid evolution is smaller in extent .
so @xmath37 needs to be closer to zero in order to have a tiny @xmath42 . in this sense , a requirement of a smaller @xmath37 would then favor large values of @xmath16 for msugra .
this we explore numerically in the next section .
we show our numerical results in two stages . to begin with
, we examine the difference between the evolution of @xmath0 in msugra and the nugm:24 scenarios in the absence of any phases . building on the lessons drawn from this exercise , we investigate
next the core issue at hand , namely the behavior of the phase naturalness measure @xmath32 in each of the scenarios and the differences therein .
focusing first on msugra , we begin with the value of @xmath0 as determined , by the rewsb conditions , in terms of the other parameters of the model , viz , @xmath11 , @xmath10 , @xmath1 and @xmath133 .
this study , coupled with that for the derived value at the gut scale , @xmath12 , would serve to indicate the regions of the parameter space for which the phase sensitivity can be significantly reduced .
fig.[sugbgraphsa ] shows the variation of @xmath0 and correspondingly @xmath12 with respect to @xmath10 . with an illustrative choice of parameters , viz .
gev , @xmath110 and @xmath135 , we exhibit our results for @xmath136 and @xmath137 .
one finds that @xmath0 , determined through the rewsb condition , is almost linear with @xmath10 .
the dependence on @xmath133 , on the other hand , is quite nonlinear ; but as already touched upon in the previous section , the rewsb condition implies that , for a given @xmath10 , @xmath0 decreases with increase in @xmath138 . as for the evolution of @xmath0 , we find that @xmath139 unless @xmath10 is quite large .
this is reflective of the aforementioned cancellations between the gaugino and trilinear terms of eq.[eqnbrge ] in msugra . for our choice of @xmath110 ,
this is same as the cancellations between the terms of @xmath107 of eq.[eqnforc ] .
once @xmath10 becomes large , the contributions from the gaugino part of eq.[eqnbrge ] dominates and the cancellations are no longer as effective .
this causes @xmath12 to supersede @xmath0 as is shown in fig.[sugbgraphsa ] .
the information regarding the evolution of @xmath0 can also be parametrized in terms of the ratio @xmath140 and this is displayed in fig.[sugbgraphsb ] as a function of @xmath10 .
this ratio is of particular interest on account of its relatively straightforward relation with the phase naturalness measure @xmath32 ( note that @xmath141 ) . as could have been guessed from fig.[sugbgraphsa ] itself , the variation with @xmath10 is nearly monotonic .
the shallow dip at small @xmath10 values is a consequence of the variation in the degree of cancellation between contributions to @xmath102 and is difficult to see analytically from the leading terms alone . for large @xmath10 , the ratio @xmath140 is seen to increase with @xmath138 , while for small @xmath10 the behavior is opposite .
this , within msugra , indicates that a small value of @xmath10 , coupled with a large @xmath138 seems to be best suited for achieving a low degree of fine - tuning in the phases . in fig.[sugbgraphsc ]
, we display the dependence of the same ratio on @xmath11 .
while the behavior may seem intriguing at first , note that @xmath0 depends on @xmath11 only via the requirement of rewsb .
as fig.[sugbgraphsa ] has already shown us , for the reference value of @xmath142 , @xmath12 is typically somewhat smaller than @xmath0 .
now , @xmath0 grows smaller as @xmath11 decreases .
thus , for small @xmath11 and large @xmath16 , @xmath0 can be very small and the aforesaid evolution implies that @xmath12 would have been negative . on the other hand , for large @xmath11 values , @xmath0 is large and thus the relatively small evolution leaves the ratio @xmath140 very close to unity .
the dependence of @xmath140 on the trilinear coupling parameter @xmath1 is quite linear ( fig.[sugbgraphsd ] ) .
this , again , can be deduced from eq.[eqnbzvsb0 ] where fixing @xmath16 , @xmath10 and @xmath11 will give rise to a linear relation between @xmath140 and @xmath1 .
note that progressively larger values for @xmath138 increases the importance of the trilinear term contributions to @xmath143 , thereby increasing the slope of the curve .
we now repeat the analysis for the case of nugm:24 choosing @xmath110 as before . however , since the sign of the electroweak gaugino mass parameters are now reversed , the gaugino contribution to eq.[eqnbrge ] would now enhance the trilinear contribution instead of cancelling it . and
since the sign inversion affects only the sub - dominant contributions to the evolution of @xmath144 , the latter remain close to their msugra values with the result that the total trilinear contribution to @xmath102 suffers only a small relative change .
the result is then a monotonic decrease of @xmath12 with an increase in @xmath10 , and hence , in an appreciably large amount of evolution ( fig.[nonbgraphsa ] ) .
a further consequence is that the ratio @xmath145 too is monotonic in @xmath10 ( fig.[nonbgraphsb ] ) .
the slope though decreases with @xmath10 , leading to a flat behavior for moderately large @xmath10 values .
this can be understood by realizing that , apart from @xmath0 being approximately linear in @xmath10 @xmath112 too is approximately linear especially for large @xmath10 .
while the steep slope for small @xmath10 might seem intriguing given the almost linear behavior of both @xmath0 and @xmath12 in fig.[nonbgraphsa ] , it should be noted that @xmath0 is very small for such @xmath10 and consequently any departure from linearity would be magnified in the ratio . that the slopes at small @xmath10 values grow with @xmath138 is understandable too , as for larger @xmath138
, the trilinear term contributions to @xmath112 assume greater significance .
the abrupt ending of the curves , especially for larger @xmath138 values might seem curious .
however , note that @xmath109 is appreciably larger in nugm:24 than in msugra ( see sec.[analysissubsec1 ] ) .
this leads to a rapid suppression of @xmath146 , the mass of the lighter stau .
while the latter also sees an enhancement on account of the @xmath36 gaugino mass being significantly larger in nugm:24 in comparison that within msugra for an identical value of @xmath10 , this effect is sub - dominant .
consequently , for such parameter values , the lighter stau would have a mass smaller than the lightest of the neutralinos thereby becoming the lightest supersymmetric particle .
since this is phenomenologically unacceptable , such regions of the parameter space have to be discarded .
note though that the extent of the allowed parameter range in the @xmath10@xmath138 plane does depend on the value of @xmath11 .
fig.[nonbgraphsc ] displays @xmath145 for different values of @xmath138 as @xmath11 is varied .
as discussed before , @xmath0 increases with increase of @xmath11 and diminishes with increasing @xmath16 . for most of the region ( except when @xmath11 is large and @xmath16 quite small )
the ratio can be large and negative because of a large degree of evolution of @xmath0 in nugm:24 . for larger
@xmath138 , @xmath0 itself is much smaller .
hence a large evolution results into a large negative @xmath12 . on the other hand , a larger value for @xmath11 pushes @xmath0 higher and @xmath12 would then be dragged down to a value near zero . additionally , we like to clarify that the larger @xmath16 curves really end near 2 tev or so in fig.[nonbgraphsc ] because of the rewsb requirement .
this is unlike the smaller @xmath16 contours that span the entire @xmath11 range displayed . as for the dependence on @xmath1 ( see fig.[nonbgraphsd ] ) , the relationship is once again linear , as predicted by eq.[eqnbzvsb0 ] , for either of the two models under discussion . having analyzed the simple case of @xmath147 , we may now consider the effect of phases . to start with , we continue to maintain @xmath148 , but now consider @xmath149 , or , in other words , a maximal phase in the trilinear coupling .
this choice maximizes the edm values@xcite . to study the generic features and compare with the results of sec.[resultssubsec1 ] , we first choose a relatively small value of @xmath41 ( @xmath150 gev ) .
thus , @xmath151 and @xmath152 ) would not be very different from the analysis of sec.[resultssubsec1 ] because of the absence of any phase in the gaugino parts of eqns.([eqnbrge ] & [ eqnatrge ] ) and the smallness of @xmath41 . with this choice of inputs ,
the only contributions to @xmath153 or @xmath154 arise from @xmath155 themselves , and hence there is no occasion for cancellations / enhancements unlike in the case for the real parts .
in addition , the effect of @xmath20 on @xmath42 would be limited even for maximal @xmath20 unless @xmath41 is quite large .
this is reflected by figs.[compbgraph ] , wherein we display the variation of both @xmath37 and @xmath42 with @xmath10 for either model .
the results are seen to be consistent with the no - phase cases of fig.[sugbgraphsa ] and fig.[nonbgraphsa ] .
we now invoke a non - zero @xmath21 and analyze the resulting evolution of the same from the electroweak scale to the gut - scale . in figs.[phaseall ] , we display this for both msugra and nugm:24 , and in each case for two values of @xmath138 , namely 3 and 10 .
again , for illustrative purposes , we choose , for the other relevant parameters , @xmath156 gev , @xmath157 gev and @xmath158 gev with @xmath121 .
although the constraints from the edm measurements restrict @xmath34 to very small values ( @xmath159 ) , we display the functional dependence for a wider range of @xmath21 .
the apparent discontinuities for the nugm:24 curves are not physical and have only been occasioned by the choice for the domain of @xmath31 , namely @xmath160 $ ] .
clearly , the amount of phase evolution in nugm:24 is seen to be higher than that in msugra .
having established that the degree of fine tuning could , in principle , be smaller in the nugm:24 case , we now perform a scan of the parameter space for both msugra and nugm:24 so as to quantify the extent of this reduction . in each case
, we consider two different values of @xmath138 ( @xmath161 ) while maintaining @xmath162 so as to maximize the edm values .
allowing @xmath11 , @xmath10 and @xmath41 to vary up to 1 tev ( with the lower end set in accordance with the current limits on super - particle masses ) , we show , in figs.[phigraphall ] , the scatter plots in the @xmath32@xmath10 plane . it is interesting to note that , for low to moderate values of @xmath16 , the measure @xmath32 rarely becomes negative in the msugra case , whereas in the non - universal scenario it is more evenly distributed . while @xmath163 does tend to concentrate around zero ( fig.[phigraphc ] ) , note that , for small @xmath138 , the nugm:24 case does have a significantly dense distribution up to @xmath164 and values as large as @xmath165 are also obtained , albeit with a reduced frequency . in contrast , the msugra case barely registers a presence even for @xmath166 ( fig.[phigrapha ] ) .
thus , in going from msugra to nugm:24 , the fine - tuning can be reduced by a factor as large as @xmath167 . for the @xmath168 case
though , the improvement is much more moderate . as fig.[phigraphb ] shows , the msugra scatter reaches up to @xmath169 , whereas the non - universal scenario admits @xmath170 ( fig.[phigraphd ] ) , or , in other words , a reduction of the maximal fine tuning by a factor of @xmath171 .
more important , though , is that the density of points at higher @xmath32 is much larger in the nugm:24 case than for msugra . in other words , it is far more likely to have a less fine - tuned point in the parameter space for nugm:24 .
concentrating on nugm:24 , we present , in fig.[slopecontours ] , contour plots for @xmath32 in the @xmath172 plane for two different values of @xmath138 .
note that the limits on @xmath11 and @xmath10 are @xmath173 tev , higher than what was chosen for fig.[phigraphall ] .
once again , @xmath41 is fixed at @xmath174 with @xmath121 .
a comparison of the two plots clearly reinforces our earlier result that the fine - tuning is less severe for low @xmath138 .
furthermore , the values of @xmath11 and @xmath10 leading to a particular @xmath32 are highly correlated .
note that both signs for @xmath32 are possible .
the region where @xmath32 changes sign is associated with a parameter point where @xmath31 is @xmath175 . to summarize , the results displayed in fig.[phigraphall ] and fig.[slopecontours ] show that it is indeed possible to obtain a surprisingly large amount of reduction of phase sensitivity even for relatively small sparticle masses .
we now explore , in detail , the range of @xmath16 that is associated with very low level of phase sensitivity or , in other words , a very large @xmath163 .
as has been argued earlier , @xmath37 itself strongly depends on @xmath16 .
moreover , @xmath112 , and thereby @xmath12 too , has a nontrivial dependence on @xmath16 .
thus it is understandable that a very large @xmath163 would indeed prominently highlight such a dependence . rather than attempting a full , but very computing - intensive ,
scan over the entire parameter space , we choose to restrict ourselves to the subset of the parameter space that would naturally produce very large values for @xmath163 , namely the region with small @xmath42 and small @xmath41 .
hence we adopt a framework with given values for @xmath42 instead of @xmath16 .
the requirement of rewsb determines @xmath138 once @xmath12 , @xmath11 , @xmath10 and @xmath1 are fixed .
note however , that the point @xmath176 would imply the absence of any susy cp phase _
at all scales_. thus , it is not surprising to obtain very large values of @xmath163 in this scenario .
however , in this part of our work the focus is simply to study , the effect of @xmath16 on @xmath32 in detail , more importantly for large @xmath163 values . to quantify our study of this issue ,
we choose small representative values viz . @xmath177 and @xmath178 , along with @xmath121 so as to maximize the edm contributions as before . in figs.[slopesmallb0 ] , we present various scatter plots for @xmath32 as @xmath11 and @xmath10 are varied over a wide range ( 0 to 2 tev ) .
note that the results of this analysis have a significant dependence on @xmath41 .
for example , increasing @xmath41 to @xmath179 gev may reduce @xmath32 by a factor of 10 to 20 . as fig.[sugslopesmallb0 ] shows , within msugra , @xmath163 could be as large as @xmath179 while most of the points lie between @xmath137 to @xmath180 .
the situation is qualitatively different in nugm:24 ( fig.[nonslopesmallb0 ] ) where @xmath163 may go up to @xmath181 while typically ranging between @xmath182 to @xmath183 .
thus , nugm:24 is much better able to accommodate low phase - sensitivity solutions than do the universal gaugino mass scenarios .
it is curious to note that , unlike what fig.[phigraphall ] suggested , @xmath32 could assume negative values within msugra ( see fig.[sugslopesmallb0 ] ) .
this prompts us present a scatter plot of @xmath32 against the derived quantity @xmath138 . as fig.[sugtanslopesmallb0 ] shows , msugra admits negative @xmath32 only for large @xmath138 .
in fact , even for the positive branch , large values of @xmath163 are typically concentrated in the large @xmath16 ( 20 to 45 ) region .
in contrast , for nugm:24 , @xmath32 assumes larger values typically for low @xmath16 values ( 2 to 5 ) .
it should be remembered in this context that , within nugm:24 , the large @xmath138 domain is significantly restricted from considerations of the lsp ( see sec.[resultssubsec1 ] ) . that the favored range for @xmath16 is different in the two scenarios is attributable to the interplay between the cancellations / enhancements in the rge evolution of @xmath0 on the one hand and the requirement of rewsb on the other . finally , we comment on the case of @xmath184 .
it turns out that for this branch of @xmath19 and @xmath121 , one has @xmath185 for almost all the parameter space of nugm:24 . as a result
one finds no advantage toward reducing the phase sensitivity .
as is well known , the experimental upper bounds on the electric dipole moments of the neutron and the electron impose strong constraints on any source of @xmath186 violation in supersymmetric models , in particular on the weak scale phase parameters .
for example , in the minimal supergravity model , @xmath21 , the phase of the bilinear higgs coupling parameter is constrained to be typically smaller than @xmath187 , with only some very limited regions ( such as the focus point scenario ) in the parameter space admitting slightly larger ( @xmath116 ) values .
this , however , implies a severe fine - tuning condition for @xmath31 , the value of the same phase parameter at the unification scale . in turn ,
@xmath20 , the phase of the trilinear coupling parameter is also severely fine - tuned .
this has been a longstanding problem with msugra - like scenarios . to quantify this problem
, we define a _ phase naturalness _
measure @xmath32 as the ratio of the spread of the phase @xmath31 at the unification scale that is consonant with the spread @xmath21 allowed , at the electroweak scale , by the electric dipole moment constraints a larger @xmath32 would imply a lower degree of phase sensitivity .
one finds that , unless @xmath16 is very large , @xmath32 may be approximated to @xmath188 for much of the parameter space . in this analysis
, we have demonstrated that models admitting a large rg evolution of the bilinear higgs coupling could be interesting in the context of a reduction in the fine - tuning of phases . in particular , we choose a supergravity - inspired scenario wherein non - universal gaugino masses arise from a gauge kinetic energy function @xmath43 transforming as a particular non - singlet representation of @xmath60 ( nugm:24 of table [ tabrelativeweights ] ) . as in the msugra ( singlet @xmath40 )
case , this representation , considered in isolation , introduces no additional phase for the gaugino masses . studying the nature of the evolution of @xmath0 to understand the correspondence with phase - sensitivity
, we identify the large cancellations in the rge for @xmath0 as being primarily responsible for the high degree of fine - tuning within msugra . in the nugm:24 ,
on the other hand , the said cancellations are replaced by enhancements ( on account of the reversal in the sign of the gaugino mass terms ) and this translates into a reduction of the above - mentioned fine - tuning .
in fact , @xmath32 can be significantly increased in nugm:24 ( by a factor of 10 to 20 ) with respect to comparable msugra type of models .
the said improvement is typically more pronounced for small @xmath16 values .
a particularly interesting result is the identification of extended regions in the nugm:24 parameter space which admit a low degree of phase - sensitivity even for relatively small super - particle masses .
this feature is absent in msugra as well as in most other models with high scale inputs for susy breaking .
we further explored the dependence of our results , on @xmath16 , by specifically concentrating on the parameter space corresponding to very large @xmath32 ( or very small phase sensitivity ) so as to compare the two models .
naturally , this occurs close to vanishing @xmath1 and @xmath12 values .
we adopt a scheme where @xmath12 itself is given as an input parameter instead of @xmath16 , given the more direct relationship of @xmath12 with @xmath32 .
our analysis shows that , even here , the values of @xmath32 in nugm:24 are typically larger by a factor of 10 to 20 in comparison to those in msugra . and
whereas msugra generically requires large @xmath16 ( 20 to 40 ) for @xmath163 to be large , the nugm:24 scenario prefers a smaller @xmath16 ( 2 to 5 ) instead . finally ,
while our analysis has focussed on @xmath60 as the gut gauge group , similar considerations hold for @xmath189 as well .
a suitable non - singlet representation resulting in a similar gaugino - mass pattern as in nugm:24 would also produce such a reduction of phase sensitivity .
* acknowledgments * + dc acknowledges financial assistance from the department of science and technology , india under the swarnajayanti fellowship grant .
dd would like to thank the council of scientific and industrial research , govt
. of india for the support received as a junior research fellow . 1
see for example textbooks like : `` supersymmetry and supergravity '' by j. wess and j. bagger , princeton univ . press , 1992 ; `` introduction to supersymmetry and supergravity '' by p. c. west , world scientific , singapore , 1990 ; j. d. lykken , arxiv : hep - th/9612114 ; `` theory and phenomenology of sparticles : an account of four - dimensional n=1 supersymmetry in high energy physics '' , m. drees , r. godbole and p. roy , world scientific , singapore , 2004 . for reviews see p. nath , r. arnowitt and a.h .
chamseddine , _ applied n = 1 supergravity _
( world scientific , singapore , 1984 ) ; h.p .
nilles , phys . rep .
* 110 * , 1 ( 1984 ) ; r. arnowitt and p. nath , proc . of vii j.a .
swieca summer school ed .
e. eboli ( world scientific , singapore , 1994 ) . l.e .
ibaez , phys . lett . *
b118 * , 73 ( 1982 ) ; p. nath , r. arnowitt and a. h. chamseddine , phys .
b * 121 * , 33 ( 1983 ) ; r. arnowitt , a. h. chamseddine and p. nath , phys .
b * 120 * , 145 ( 1983 ) ; j. ellis , d. v. nanopoulos and k. tamvakis , phys .
b121 * , 123 ( 1983 ) .
k. inoue et al . , prog .
theor . phys . *
68 * , 927 ( 1982 ) ; l. ibaez and g. g. ross , phys .
b110 * , 215 ( 1982 ) ; l. alvarez - gaum , j. polchinski and m. b. wise , nucl .
* b221 * , 495 ( 1983 ) ; j. ellis , j. hagelin , d. v. nanopoulos and k. tamvakis , phys . lett . *
b125 * , 275 ( 1983 ) ; l. e. ibaez and c. lopez , phys . lett . *
b126 * , 54 ( 1983 ) ; nucl . phys .
* b233 * , 511 ( 1984 ) ; l. e. ibaez , c. lopez and c. muos , nucl . phys . * b256 * , 218 ( 1985 ) ; j. ellis and f. zwirner , nucl .
* b338 * , 317 ( 1990 ) .
s. p. martin and m. t. vaughn , phys .
b * 318 * , 331 ( 1993 ) [ arxiv : hep - ph/9308222 ] ; phys .
d * 50 * , 2282 ( 1994 ) [ arxiv : hep - ph/9311340 ] ; i. jack , d. r. jones , s. p. martin , m. t. vaughn and y. yamada , phys .
d * 50 * , r5481 ( 1994 ) [ arxiv : hep - ph/9407291 ] .
t. ibrahim and p. nath , phys .
d * 58 * , 111301 ( 1998 ) , [ erratum - ibid .
d * 60 * , 099902 ( 1999 ) ] [ arxiv : hep - ph/9807501 ] ; m. brhlik , g. j. good and g. l. kane , phys .
d * 59 * , 115004 ( 1999 ) [ arxiv : hep - ph/9810457 ] .
t. ibrahim and p. nath , phys .
d * 61 * , 093004 ( 2000 ) [ arxiv : hep - ph/9910553 ] ; m. brhlik , g. j. good and g. l. kane , phys .
d * 59 * , 115004 ( 1999 ) [ arxiv : hep - ph/9810457 ] ; t. ibrahim and p. nath , phys .
d * 57 * , 478 ( 1998 ) [ erratum - ibid .
d * 58 * , 019901 ( 1998 errat , d60,079903.1999 errat , d60,119901.1999 ) ] [ arxiv : hep - ph/9708456 ] .
a. bartl , w. majerotto , w. porod and d. wyler , phys .
d * 68 * , 053005 ( 2003 ) [ arxiv : hep - ph/0306050 ] ; s. abel , s. khalil and o. lebedev , nucl .
b * 606 * , 151 ( 2001 ) [ arxiv : hep - ph/0103320 ] ; a. bartl , t. gajdosik , w. porod , p. stockinger and h. stremnitzer , phys .
d * 60 * , 073003 ( 1999 ) [ arxiv : hep - ph/9903402 ] ; d. chang , w. y. keung and a. pilaftsis , phys .
lett . * 82 * , 900 ( 1999 ) [ erratum - ibid . *
83 * , 3972 ( 1999 ) ] [ arxiv : hep - ph/9811202 ] j. ellis , k .
enqvist , d .
v. nanopoulos , and k. tamvakis , phys . lett . *
155b * , 381(1985 ) ; m. drees , phys .
lett . * b158*,409(1985 ) ; k. huitu , y. kawamura , t. kobayashi and k. puolamaki , phys .
d * 61 * , 035001 ( 2000 ) .
k. huitu , j. laamanen , p. n. pandita and s. roy , arxiv : hep - ph/0502100 ; u. chattopadhyay and d. p. roy , phys .
d * 68 * , 033010 ( 2003 ) [ arxiv : hep - ph/0304108 ] ; u. chattopadhyay and p. nath , phys .
d * 65 * , 075009 ( 2002 ) [ arxiv : hep - ph/0110341 ] .
g. gamberini , g. ridolfi and f. zwirner , nucl .
b * 331 * , 331 ( 1990 ) ; v. d. barger , m. s. berger and p. ohmann , phys .
d * 49 * , 4908 ( 1994 ) ; for two - loop effective potential see : s. p. martin , phys .
d * 66 * , 096001 ( 2002 ) . | the phases in a generic low - energy supersymmetric model are severely constrained by the experimental upper bounds on the electric dipole moments of the electron and the neutron .
coupled with the requirement of radiative electroweak symmetry breaking , this results in a large degree of fine tuning of the phase parameters at the unification scale . in supergravity type models , this corresponds to very highly tuned values for the phases of the bilinear higgs coupling parameter @xmath0 and the universal trilinear coupling @xmath1 .
we identify a cancellation / enhancement mechanism associated with the renormalization group evolution of @xmath0 , which , in turn , reduces such fine - tuning quite appreciably without taking recourse to very large masses for the supersymmetric partners .
we find a significant amount of reduction of this fine - tuning in non - universal gaugino mass models that do not introduce any new phases .
+ pacs numbers:13.40.em,04.65.+e,12.60jv,14.20.dh -1 cm 16.5 cm 21.5 cm -0.5 cm # 1 # 1 # 1 # 1 # 1#2#3#4#1 * # 2 * , # 3 ( # 4 ) 0_b_0 22 2 ^ 2_w 2m_z^2 2b2 v#1v_#1 # 1 ^ 0_#1 # 1m_#1 # 1m_#1 # 1m_^0_#1 # 1m_^+_#1 2wsec^2_w 2(g-2 ) _ = 11 # 1#2 # 1#1| # 1| # 1 # 1#1 # 1#1 versim#1#2 #
1#2 * large evolution of the bilinear higgs coupling parameter in susy models and reduction of phase sensitivity + * 0.5 cm utpal chattopadhyay@xmath2 , debajyoti choudhury@xmath3 and debottam das@xmath2 0.2 cm _
@xmath2department of theoretical physics , indian association for the cultivation of science , raja s.c .
mullick road , kolkata 700 032 ,
india + _ _ @xmath3department of physics and astrophysics , university of delhi , delhi 110 007 , india + _ |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.