article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the rapid proton capture ( rp process ) is a sequence of rapid proton capture reactions and @xmath2 decays passing through proton rich nuclei .
it is a dominant nucleosynthesis process in explosive hydrogen burning conditions @xcite .
reaction rates govern energy release and final isotopic abundances . rp
process rates are crucial for x ray burst models @xcite . some of the capture rates along rp
process are not determined experimentally .
but many rely on theoretical input .
previous predictions of reaction rates have generally assumed that all interacting nuclei are in the ground state @xcite .
schatz et al .
@xcite showed that the capture on the low - lying first - excited state of @xmath0cl was crucial for the @xmath0cl(p,@xmath3)@xmath4ar reaction rate .
the energies and spin in @xmath4ar were provided by experiment , but input for the proton and gamma decay widths come from theory .
the impact of the excited states in the target nuclei that are thermally populated in astrophysical plasma is expressed in terms of a `` stellar enhancement factor '' ( sef ) .
the sef for @xmath0cl was found to be around five for temperatures in the range 0.1 to 0.8 gk . in this paper
we investigate other sd - shell nuclei where low - lying excited states may lead to large sef .
in case of high level densities and high @xmath5 values ( q@xmath65 mev ) , reaction rates can be calculated by statistical models @xcite . for nuclei close to proton drip line @xmath5 values of ( p,@xmath3 )
reactions drop strongly , the compound nuclei are produced at low densities @xcite .
proton capture reaction rates for low level densities in the compound nucleus are determined by contributions of direct capture and single resonances correlated with unbound states in the compound nucleus .
nonresonant cross section is expressed in terms of the astrophysical @xmath7factor , which contains the nuclear component of the transition probability to the ground states and low excited states of final nuclei .
@xmath7factors are calculated using radcap code @xcite , which uses a woods saxon nuclear potential ( central + spin orbit ) and a coulomb potential of a uniform charge distribution .
the nuclear potential parameters are chosen to reproduce the bound state energies
. direct proton capture rate on the target nucleus in state @xmath8 is given by @xmath9de } , \end{aligned}\ ] ] where b , which arises from the barrier penetrability , is given by @xmath10 here @xmath11 is reduced mass in entrance channel in amu , @xmath12 is the boltzmann constant , @xmath13 is a charge number of target nucleus , and @xmath14 is proton energy .
@xmath15 is the sum of the individual @xmath7factors of the transitions from the initial state @xmath8 into all bound states in the final nucleus . for a given stellar temperature @xmath16 ,
nuclear reactions take place in relatively narrow energy window around the effective burning energy @xmath17 .
the @xmath7factor is nearly a constant over gamov window @xmath18 , where @xmath19 and @xmath20 hence , the capture rate @xmath21 is usually approximated by a constant @xmath7factor .
resonance decay widths depend on the coulomb barrier , the orbital momentum barrier and partial widths of the compound state .
the resonant reaction rate for capture on a nucleus in an initial state @xmath8 , @xmath22 for isolated narrow resonances is calculated as a sum over all relevant compound nucleus states @xmath23 above the proton threshold @xcite : @xmath24 here @xmath25 is the temperature in @xmath26 , @xmath27 is the resonance energy in the center of mass system , the resonance strengths in @xmath28 for proton capture are @xmath29 @xmath30 is a total width of the resonance level and @xmath31 and @xmath32 are target and projectile spins .
proton decay width depends exponentially on the resonance energy and can be calculated from proton spectroscopic factor @xmath33 and the single particle proton width @xmath34 as @xmath35 .
the single particle proton widths were calculated from @xmath36 @xcite with @xmath37 and where the channel radius @xmath38 was chosen to match the width obtained from an exact evaluation of the proton scattering cross section from a woods - saxon potential well for @xmath39 , @xmath40 mev .
this simple model matches exact calculations in the sd - shell t within about 10% , and has advantage that it is fast and can be easily extrapolated to energies below 0.1 mev where the scattering calculation becomes computationally difficult .
we use a coulomb penetration code from barker @xcite .
the proton spectroscopic factors and the gamma widths are obtained with the usd interaction @xcite in the full sd - shell model space .
the gamma decay widths were obtained with b(m1 ) values based on free - nucleon g - factors and b(e2 ) values with harmonic oscillator radial wave functions and effective charges of @xmath41 and @xmath42 @xcite .
the uncertainty of the spectroscopic factors and gamma decay widths is on the order of 20@xmath43 . as we will emphasize
the largest uncertainty for the present work is the precise location of the resonances near the proton decay thresholds .
the total reaction rate is the sum of the capture rate on all thermally excited states in the target nucleus weighted with their individual population factors : @xmath44 here @xmath45 and @xmath46 are the resonance energies .
dccdcdc e_x ( mev ) & @xmath47 & @xmath48 & c^2s & @xmath49 & c^2s^ * & @xmath50 + 0 & @xmath51 & 2s@xmath52 & 0.080 & 6.58e-3 7.00e-3 & & + & & 1d@xmath53 & 0.672 & 3.78e-3 4.43e-3 & 1.127 & 4.15e-3 5.32e-3 + & & 1d@xmath54 & & & 0.001 & 4.09e-6 6.38e-6 + 1.359 & @xmath55 & 2s@xmath52 & 0.001 & 8.65e-5 9.63e-5 & 0.006 & 5.76e-4
6.19e-4 + & & 1d@xmath53 & 0.185 & 1.54e-3 1.82e-3 & 0.119 & 6.35e-4
8.17e-4 + & & 1d@xmath54 & 0.004 & 5.74e-5 7.76e-5 & 0.009 & 7.73e-5 1.20e-4 + 1.798 & @xmath56 & 2s@xmath52 & & & 0.002 & 2.42e-4 2.69e-4 + & & 1d@xmath53 & 0.145 & 1.60e-3 1.91e-3 & 0.620 & 4.33e-3 5.60e-3
+ & & 1d@xmath54 & 0.006 & 1.07e-4 1.46e-4 & 0.021 & 2.25e-4 3.49e-4 + 2.439 & @xmath55 & 2s@xmath52 & 0.031 & 3.68e-3 5.20e-3 & 0.024 & 1.83e-3 2.36e-3
+ & & 1d@xmath53 & 0.167 & 9.65e-4 1.18e-3 & 0.128 & 4.68e-4
6.18e-4 + & & 1d@xmath54 & 0.014 & 1.19e-4 1.66e-4 & 0.016 & 8.99e-5
1.42e-4 + 3.154 & @xmath55 & 2s@xmath52 & 0.068 & 6.92e-3 1.53e-2 & 0.001 & 4.01e-5 7.36e-5 + & & 1d@xmath53 & 0.516 & 1.92e-3 2.67e-3 & 0.169 & 4.11e-4
5.91e-4 + & & 1d@xmath54 & 0.003 & 1.55e-5 2.40e-5 & & + & & & & 2.95e-2 3.74e-2 & & 1.37e-2 1.63e-2 + [ cols="^,^,^,^,^,^,^,^,^,^ " , ] @xmath57p is also an odd - odd nucleus , @xmath58 .
the first excited state of @xmath57p , spin @xmath59 , is at 105.64 kev .
spin of the ground state is @xmath60 .
the second excited state is at 877 kev and spin @xmath61 is assigned from usd shell model prediction .
usd shell model gives energies of 142 kev and 840 kev .
proton separation energy of @xmath62s is 3.290 mev . as seen from fig .
[ 34res ] and fig .
[ fig3 ] , @xmath57p(p,@xmath3)@xmath62s reaction rate is dominated by fourth @xmath56 state resonance at 3.817 mev ( usd shell model energy prediction ) .
proton capture on a ground state and first excited state of @xmath57p is equally important .
[ fig3 ] shows the sef of about 20 around 0.8 gk due to first excited state .
this comes from the final state ( 11,5/2@xmath63,4 ) at 3.817 mev in @xmath62s .
it is connected to the @xmath64 excited state with a relatively strong @xmath65 spectroscopic factor .
there are no known excited states in @xmath62s , but in the mirror nucleus @xmath62al , the fourth 5/2@xmath63 state at 3.641 mev is in reasonable agreement with the theoretical energy .
the analogue energy of this state in @xmath62s should be within 100 kev of 3.64 mev , but this is enough to give an order of magnitude change in the rp - process reaction rate around 0.8 gk .
reaction rate at relevant temperature range @xmath66 gk agrees well with statistical model @xcite , however for higher stellar temperatures up to 10 gk direct capture starts to dominate reaction rate and total reaction rate exceeds statistical model prediction .
results for @xmath57p@xmath67s .
the middle panel shows the rp - reaction rate for the @xmath57p @xmath68 ground state ( dashed line ) and the sum of the contributions from the ground and the @xmath64 excited state at 0.106 mev ( full line ) .
see caption to fig .
[ fig1 ] for other details.,width=302 ]
although very few @xmath69 shell nuclei have low lying excited states , contributing significantly to total rates , proton capture on a first excited state can not be neglected .
in addition to the previously studied case of @xmath0cl@xmath1ar , we find at least two other cases @xmath57p@xmath70s and @xmath71cl@xmath72ar for which the low - lying excited state in initial nucleus dominates the cross section .
unfortunately , the excited states are not known for these cases .
initial states above about 300 kev are not important unless there is a particularly strong spectroscopic factor that connects to low - lying final states in the daughter .
direct capture rates depend on astrophysical s factor , which changes slightly with temperature at temperatures relevant to rp
process for x ray bursts , @xmath73 gk , and may have resonances for higher temperatures .
therefore knowing at what energies resonances exist is important for quality of such reaction rate predictions and so for isotopic abundances and energy release .
an important task for the near future is to quantify the errors associated with these and other cases where the energies and spectroscopic factors are not known experimentally and must be taken from theory .
recent efforts have been me to establish improved hamiltonians for the sd - shell .
calculation of the corss section with several realistic hamiltonians will help to establish theoretical error bands for important cross sections . | we investigate the importance of proton capture on low - lying excited states that are thermally excited in hot stellar enviroments .
in addition to the previously known case of @xmath0cl@xmath1ar we find several other sd - shell nuclei with large stellar enhancement factors . we discuss the uncertainty this introduces into rp - process nucleosynthesis . |
elliptical galaxies define a fundamental plane ( fp ) on the 3-parameter space of the half - light radius @xmath10 , the mean surface brightness within that radius @xmath11 , and the central velocity dispersion @xmath12 ( djorgovski & davis 1987 ; dressler et al .
the scatter around the fp is only @xmath13 dex in log@xmath10 for cluster ellipticals .
the existence of this scaling relation provides one of the most important constraints to model the formation and evolution of elliptical galaxies ( bender et al 1993 ; guzmn et al .
1993 ; renzini & ciotti 1993 ; zepf & silk 1996 ) .
also , it can be used as a distance indicator to trace the non - hubble motions in the local universe ( dressler et al 1987b ; lucey & carter 1988 ; lynden - bell et al .
1988 ) , and potentially as a cosmological evolutionary probe when studying the fp for cluster ellipticals at higher redshifts ( franx et al . 1996
; van dokkum & franx 1996 ; barger et al .
1996 ) .
the biparametric nature of elliptical galaxies suggests that the virial theorem is the main constraint on their structure ( faber et al .
1987 ; djorgovski 1987 ; but see guzmn et al .
the tilt of the fp relative to the virial equation implies a systematic variation either in the imf , the shape of the light profile , the age and metallicity or the dark and luminous matter distribution with galaxy mass ( guzmn et al .
1993 ; ciotti et al .
our understanding of the physical mechanisms responsible for such variation is still at a very early stage .
moreover , substantial fine - tuning is required to produce the tilt while preserving the small scatter of the galaxy distribution along the fp ( ciotti et al .
the tightness of the fp for cluster ellipticals is clear evidence for a very standardized and synchronized production of ellipticals , with the vast majority of their stellar populations being formed at redshifts @xmath14 , at least for those ellipticals in clusters ( renzini 1995 ; ellis et al .
. however , there is growing evidence that ellipticals outside the core of rich clusters tend to have a younger stellar component ( rose 1985 ; pickles 1985 ; bower et al .
1990 ) . as a result ,
non - cluster ellipticals at a given @xmath10 and/or @xmath12 have , on average , bluer colours , higher effective surface brightnesses , and lower @xmath15 line strength indices as compared to their cluster counterparts ( larson et al .
1980 ; guzmn et al .
1992 ; de carvalho & djorgovski 1992 ) .
this systematic variation of stellar population properties due to the environment may in turn translate into a zero - point offset of the fp - based distance indicators ( such as @xmath16 ) and thus lead to spurious peculiar velocities ( kaiser 1988 ; silk 1989 ; de carvalho & djorgovski 1992 ; but see burstein et al .
1990 ) . indeed , using stellar population evolutionary models , guzmn & lucey ( 1993 )
have combined the fp with the @xmath17 relation to create a new distance indicator that is independent of age / environmental differences . when the new distance indicator is applied to the ellipticals with a younger stellar component ,
the large positive peculiar motions previously reported are greatly removed ( guzmn & lucey 1993 ) .
most studies on the fp have been done at optical wavelengths .
observations in the infrared may shed light on the origin of the tilt and scatter of the fp by probing a wavelength region with a different sensitivity to stellar population effects .
in particular , the effect of age and/or metallicity variations in ellipticals is expected to be greatly reduced in the k - band ( i.e. , 2.2@xmath0 ) since the galaxy light at these wavelengths is mainly sensitive to the old stellar component and is less affected by line - blanketing . moreover , the k - band is significantly less sensitive to both the presence of dust in ellipticals and galactic extinction .
the new k - band fp can also be used as an independent distance indicator to check for stellar - population effects on the peculiar velocity measurements derived using the fp in the optical . early work on the infrared fp relied on single element photometry and hence , lacked detailed surface photometry ( gregg 1995a ; recillas - cruz et al .
1990 , 1991 ) .
recently , pahre et al .
( 1996 ) have shown that the coefficients and scatter of the fp in the k - band for a sample of ellipticals in five nearby clusters are similar to those derived in the optical .
they conclude that the observed small differences are consistent with the reduction of metallicity effects in the near - infared bandpass . in this paper
, we present the first results of a large - scale k - band survey of elliptical galaxies located in different environments ( i.e. , clusters , field and great attractor region ) . here
we study the coefficients and scatter of the k - band fp , using accurate surface photometry for a homogeneous sample of 48 galaxies in the coma cluster . by considering a single cluster ,
we avoid introducing artificial scatter due to distance errors or non - universality of the fp .
the aim of this study is three fold : 1 ) to shed light on the origin of the fp by using a wavelength region with a different sensitivity to the stellar population effects than the optical ; 2 ) to establish the intrinsic scatter on the k - band fp ; and 3 ) to set the zero - point of the new infrared distance scale using the k - band fp for coma cluster ellipticals . in the next section ,
the observations and data reduction are discussed .
this is followed by the study of the near - infrared scaling laws in section 3 .
the implications for the origin of the fp and the new distance scale are discussed in section 4 .
finally , our conclusions are summarized in section 5 .
the observations used in this study were made at the united kingdom infrared telescope ( ukirt ) in the period 28 - 30th april 1994 .
near - infrared k - band ( 2.2 @xmath0 ) images of 48 elliptical galaxies in the coma cluster were obtained , using the @xmath18 infrared array detector ( ircam3 ) with a pixel size of @xmath19 .
the galaxies are selected to be ellipticals ( i.e. no lenticulars ) from the sample in lucey et al ( 1991 , table 6 ) with reliable velocity dispersion and optical photometry .
they are located at different distances from the center of the coma cluster ( @xmath20 12 57 19 @xmath21 28 15 51 ( 1950 ) ) . also , they have optical half - light diameters smaller than @xmath22 , fitting the field of view of the ircam3 .
an exposure time of 10 sec . was used with 6 co - adds . for each object ,
nine exposures were taken at different positions , separated by @xmath23 , giving a total integration time of 9 mins . per object .
this technique was employed to avoid the effect of bad pixels and cosmic rays , to construct reliable flatfields for each frame and to measure the sky background .
surface photometry of smaller galaxies can be affected by the seeing condition which was estimated to be around @xmath24 .
the sensitivity of the results to this effect will be explored in the following analysis .
observations of faint ukirt standards were also carried out during the night and were used to monitor the accuracy of the photometry .
observations were dark subtracted using the dark frames closest ( in time ) to each object . for each galaxy , the nine mosaic frames were then median filtered and the result normalised to its median pixel value to construct the flatfield .
the dark subtracted , flatfielded frames were then registered and mosaiced to increase the s / n .
the data reduction was carried out , using the ccdpack software in the starlink environment .
the median sky value was then estimated for each frame individually and used to sky - subtract that frame .
the mosaic frames , providing a larger sky coverage than single frames , were used for this purpose and the sky values were measured as far away from the center of galaxy as possible .
the variations in sky estimates at different positions on the frame were monitored , giving a measure of the uncertainties in these values .
this did not exceed 0.02 mag .
, even for the bigger galaxies for which the uncertainties are larger . therefore , this is taken as our estimated error due to sky subtraction .
the standard stars were reduced in the same way and used to establish the extinction relation for each night .
changes in the zero - points , estimated from the standards throughout the run ( internal photometric error ) , is 0.02 mag .
aperture photometry was performed on the final images and used to construct the near - ir growth curves for individual galaxies .
the magnitudes were corrected for atmospheric extinction , galactic absorption ( taking @xmath25 ( pahre et al .
1996 ) and the median @xmath26 for the coma galaxies from faber et al ( 1989 ) ) and redshift effect ( assuming @xmath27 for ellipticals at the k - band ( persson et al ( 1979 ) ) .
the photometric accuracy is checked by comparing magnitudes of the objects in our sample with the same galaxies from other independent observations .
the k - band magnitudes for 18 galaxies in the present sample , in common with bower et al ( 1992 ) , are measured over an aperture of @xmath28 diameter , used by these authors .
the two magnitudes are compared in figure 1a and show a mean offset of @xmath29 with an rms scatter of 0.03 mag . assuming equal errors in both the data sets , this corresponds to an internal error of @xmath30 mag . per galaxy ,
consistent with that claimed by bower et al ( 1992 ) .
the cause of the zero - point offset between the two data sets is not clear but is likely to be due to different detectors used .
however , since the data here are 2-d and bower et al s is based on single element photometry , we expect to have a more accurate control over the sky subtraction in our data .
neverthless , the offset is relatively small .
the growth curves , corrected for galactic absorption and redshift effects , were used to construct near - infrared surface brightness profiles for each galaxy , corrected for the @xmath31 surface brightness dimming .
near - infrared isophotal diameters , corresponding to a mean surface brightness of 16.5 mag / arcsec@xmath32 are then estimated .
this isophote was adopted to be directly comparable with the mean optical surface brightness of 19.8 mag / arcsec.@xmath32 ( assuming @xmath33 mag . for ellipticals ) used in the optical @xmath34 relation . to examine the accuracy of our surface photometry , we have also obtained data for a sub - set of 12 ellipticals in our sample , using the iris at the anglo - australian telescope ( march 1995 ) .
these diameters , measured at a mean k - band surface brightness of 16.5 mag./arcsec@xmath32 , are compared with the present measurements in figure 1b and show a mean offset of @xmath35 with an rms scatter of 0.019 dex .
assuming that the uncertainty in these measurements is shared equally between both the data sets , we estimate an observational error of 0.01 dex in log(@xmath36 ) values .
the diameters of smaller galaxies in our sample might be affected by the seeing condition . to estimate this effect
, we used the models in lucey et al ( 1991 ) and estimated the effect of 1 arcsec seeing on the aperture photometry ( the observations here were done in 0.8 - 1.2 arcsec seeing conditions ) .
the correction was then applied to one of our smaller galaxies and its photometric parameters were recalculated .
we find that at 1 arcsec seeing conditions , an increase in log(@xmath36 ) of only 0.01 dex was needed due to seeing .
the diameters of larger galaxies were almost insensitive to this effect .
the range in surface brightness values between 16 and 17 mag / arcsec@xmath32 was explored in 0.1 intervals .
the results in the following sections were found to be insensitive to the actual choice of the isophote .
the near - infrared effective diameter and surface brightness ( i.e. the diameter containing half the total luminosity of a galaxy and the mean surface brightness within that diameter ) were calculated for galaxies in the present sample by fitting the observed k - band curve of growths to the de vaucouleurs @xmath37 law .
the accuracy of this fitting procedure was estimated by generating simulated profiles following the @xmath37 law with known effective parameters and artificially added noise , analogous to the photometric errors .
an @xmath37 fit to these profiles recovers both the input effective parameters , giving errors of @xmath38 and 0.03 mag / arcsec@xmath32 in the effective diameter and surface brightness values respectively .
the total k - band magnitudes are then estimated by extrapolating the @xmath37law fit for individual galaxies to large radii , using their corresponding values of the effective diameter and surface brightness .
these magnitudes are corrected for galactic absorption and redshift , using the relations discussed above .
also , the effective surface brightness estimates are corrected for the @xmath31 dimming effect . finally , the k - band magnitudes , measured over an aperture of 20 arcsec .
diameter are estimated and combined with their corresponding v - band data from literature , measured over the same aperture ( lucey et al 1991 ) .
the @xmath39 colours were then estimated and corrected for galactic extinction and redshift , using the above prescription for the near - ir and the relations from faber et al ( 1989 ) for the optical data . the sample of 48 elliptical galaxies in the coma cluster , observed in this study , is presented in table 1 .
column 2 gives the distance ( in degrees ) of galaxies from the center of the cluster . in columns 3 and 4 , the near - infrared ( k - band )
isophotal diameters in arcsec .
( corresponding to a mean surface brightness of @xmath40 mag / arcsec@xmath32 ) and total magnitudes of ellipticals are presented .
these are corrected for redshift effect , surface brightness dimming and galactic absorption .
column 5 gives the optical v - band diameters ( in arcsec . ) , corresponding to a mean surface brightness of @xmath41 mag / arcsec@xmath32 , taken from lucey et al .
the velocity dispersion of ellipticals ( log(@xmath12)- in km / sec ) , listed in column 6 , are taken from lucey et al .
( 1997 ) . for galaxies with no available velocity dispersion measurements from this source , the values in lucey et al .
( 1991 ) were converted to the above system using the common galaxies between the two samples .
the velocity dispersions have an associated error of @xmath42 .
columns 7 and 8 present the near - infrared effective diameters ( in arcsec . ) and effective mean surface brightnesses .
this is followed in columns 9 and 10 by the k - band aperture magnitudes and v@xmath9k colours respectively , measured over an aperture of 20 arcsec .
diameter and corrected for galactic extinction and redshift .
the mg@xmath8 line strengths are listed in column 11 .
these are also taken from lucey et al .
( 1997 ) . for galaxies with no such measurements from this reference , the mg@xmath8 indices from lucey et al .
( 1991 ) are used after conversion to this system .
[ cols="<,^,^,^,^,^,^,^,^,^,^ " , ]
the near - infrared fundamental plane of ellipticals in the coma cluster is established by fitting a plane to the distribution of galaxies on the effective diameter ( a@xmath43 ) , effective mean surface brightness ( @xmath44 ) and velocity dispersion ( @xmath12 ) space , using the data from table 1 .
a simultaneous 3-parameter fit , using 48 galaxies , gives @xmath45 with an rms scatter of 0.072 dex in @xmath46 . an edge - on view of the near - ir fp is shown in figure 2 .
the shape of the near - infrared fp and its scatter is in good agreement with that found by pahre et al ( 1996 ) , using 12 ellipticals in the coma cluster .
the presence of curvature on the infrared fp is explored by investigating the dependence of the scatter in figure 2 on the fp parameters .
no relation has been found between the residuals around the fp ( @xmath47 ( fp ) @xmath48 ) and the effective diameter or surface brightness .
the observed rms scatter in the near - infrared fp corresponds to an uncertainty of @xmath1 in distances to individual galaxies from this relation . to investigate differences between the infrared and optical fps
, we have also constructed the optical ( v - band ) fp , using the same sample of 48 ellipticals listed in table 1 .
the coefficients of the optical fp , estimated by performing a 3-parameter fit to these data , correspond to 1.44@xmath49 0.04 and 0.32@xmath49 0.01 for log(@xmath12 ) and @xmath50 respectively .
the rms scatter in the optical fp here is 0.074 dex , which is similar to the rms scatter of 0.08 dex ( djorgovski and davies 1987 ; jorgensen et al 1996 ) and 0.07 dex ( de carvalho and djorgovski 1992 ) for the b - band and 0.075 dex ( lucey et al 1991 ) for the v - band fp , using other independent samples .
differences between the fp coefficients from different studies , found in literature , are mainly due to the adopted fitting mathods ( i.e. taking @xmath51 or @xmath12 as the independent variables in the least - squares solutions ) or sample selection , with consideration to minimise the bias in the fit .
for example , it has been estimated that a least - squares fit in @xmath52 , produces a bias of the order of 5% for the @xmath53 coefficient , caused by observational errors ( jorgenson et al 1996 ) . in this study , it is attempted to reduce such biases by performing a 3 parameter fit to the fp . however , adopting different fitting methods , the fp coefficients here lie in the range 1.321.50 ( @xmath53 ) and 0.300.32 ( @xmath54 ) at the near - infrared and 1.231.44 ( @xmath53 ) and 0.300.35 ( @xmath54 ) at the optical wavelengths .
this shows the sensitivity of the fp coefficients to the adopted fitting technique , with the values in the above range , being consistent with the present study .
the fp parameters found here , also agree closely with other independent estimates of the near - infrared ( pahre et al 1996 ) and optical ( faber et al 1987 ; guzman et al 1993 ) fps .
the coefficients of the near - ir and v - band fps are similar despite different sensitivities to the stellar population and metallicity at these wavelengths . however , a smaller rms .
scatter might have been expected in the near - ir fp since at this wavelength , the light samples a more uniform stellar population and is less affected by differences in line blanketing among the ellipticals .
these will be addressed later in this section .
studies of the optical fp have revealed the presence of a young stellar population in ellipticals to be partly responsible for the observed scatter in this relation ( gregg 1995b ) .
moreover , at a given @xmath12 , coma ellipticals with lower mg@xmath8 values have , on average , slightly larger @xmath55 diameters , with most of them located outside the core of the cluster ( i.e. r@xmath561 deg . ) , indicating contributions from an intermediate age stellar population , preferentially in ellipticals in low density environments ( guzmn et al 1992 ) .
this confirms that mg@xmath8 features are sensitive to both the ` young ' stellar population and local environment of ellipticals with the galaxies having larger contributions from the ` young ' stellar populations satisfying @xmath57 ( guzmn and lucey 1993 ) .
there are only five galaxies in our near - infrared sample which satisfy this relation ( n4876 , n4807,n4789 , 159 - 83 and n4673 ) with four of them located at the outskirts of the cluster ( r@xmath58 deg . ) .
these galaxies are indistinguishable from the rest of the ellipticals on the near - ir fp in the coma cluster ( fig 2 ) .
the sensitivity of the near - ir fp to changes in the stellar population and metallicity among the ellipticals are further investigated by comparing the @xmath47 ( fp ) values with the mg@xmath8 index and v - k colour residuals ( at a given @xmath12 ) from the mg@xmath59 and ( v - k)@xmath60 relations respectively ( figure 3 ) .
no relation has been found .
a similar study of the residuals diagram , using the optical fp consisting of both field and cluster ellipticals , shows a significant trend which is mainly interpreted as a stellar population effect ( prugniel and simien 1996 ) .
however , using the sample of cluster ellipticals in this study , we find no relation between the residuals from the v - band fp and @xmath47 ( mg@xmath8 ) or @xmath47 ( v@xmath9k ) estimates . this result has been confirmed independently by jorgensen et al .
( 1996 ) , using a different sample of elliptical galaxies in clusters .
therefore , it is not clear if the lack of correlation on the residuals diagrams in figures 3a and 3b , using the near - infrared fp in the coma cluster , is due to the uniformity of our cluster galaxies or , represents a genuine effect , implying that the observed scatter on this relation is not affected by metallicity or contributions from the young stellar population .
these scenarios are currently explored by extending the study of the near - infrared fp to the field ellipticals .
however , it is likely that changes in contributions from the intermediate age asymptotic giant branch ( agb ) stars to the near - infrared light of ellipticals is partly responsible for the observed scatter in this relation .
spectroscopic observations of near - ir co ( 2.3 @xmath0 ) absorption features of ellipticals in clusters have revealed these objects to have a smaller population of agb stars than those in general field ( mobasher and james 1996 ) .
therefore , it may be possible to introduce near - ir co features as an extra parameter to reduce the scatter on the infrared fp in the same way the mg@xmath8 indices were introduced to the optical fp to correct the relation for contributions from the young population ( guzmn and lucey 1993 ) . considering the measurement errors of 0.02 dex in log(a@xmath43 ) , 0.03 dex in log(@xmath12 ) , 0.03 mag / arcsec@xmath32 . in the @xmath44 and 0.02 mag . due to sky subtraction , and considering the fact that the errors in log(a@xmath43 ) and @xmath44 are not independent ( kormendy 1977 ) , we estimate an observational error of 0.040 dex for the near - ir fp . compared with a total rms error of 0.072 dex
, this gives an intrinsic scatter of 0.060 dex in the near - ir fp .
the @xmath61 relation for our sample of 48 ellipticals in the coma cluster is shown in figure 4 .
taking @xmath12 as the independent variable ( lynden - bell et al .
1988 ) , a least squares fit to the @xmath61 relation , using all the 48 galaxies , gives @xmath62 with an rms scatter of 0.073 dex . to explore the wavelength dependence of this relation , the optical @xmath34 relation , using the same sample of 48 ellipticals from table 1 ,
is constructed .
this is consistent with a slope of @xmath63 and rms scatter of 0.077 dex , in agreement with studies based on other independent samples ( jorgensen et al 1996 ; lucey et al 1991 ; djorgovski and davies 1987 ) .
the similarity in the rms scatter between the two relations implies that dynamical parameters or the presence of an intermediate - age stellar population contributing to the near - ir light of ellipticals are responsible for the observed scatter in the @xmath61 relation .
the slight wavelength dependence of the slope of the isophotal diameter - velocity dispersion relation ( @xmath64 effect ) leads to a correlation between @xmath65 and @xmath53 ( figure 5 ) as @xmath66 this implies that over the range of velocity dispersions covered in our sample ( @xmath67 ) , the diameters change by @xmath68 ( ie .
more massive galaxies have correspondingly larger near - infrared diameters ) . using the stellar synthesis models of worthey ( 1994 ) ,
the slope of the log@xmath69log@xmath70 relation , due to changes in metallicity and age , is predicted and compared with the observed relation in fig 5 .
the models correspond to a range in age from 5 to 17 gyr at a constant metallicity of [ fe / h]=0.25 and variations in metallicity from @xmath71 to 0.5 at a fixed age of 12 gyr .
the two models are normalised to have log@xmath72 at 12 gyr and [ fe / h]=0.25 .
these normalisations are chosen to be close to the values expected for local ellipticals .
the predicted relations are consistent with the observed slope , implying that changes in age ( ie .
stellar population ) , metallicity or a combination of them among the ellipticals is responsible for the wavelength dependence of the slope of their diameter - velocity dispersion relation .
however , it is not possible to disentangle the effects of age and metallicity using these models .
the total near - infrared luminosities of elliptical galaxies in the coma cluster are used from table 1 to establish the near - ir @xmath73 relation .
the total magnitudes are employed here in order to have this relation on the same magnitude scale as the previous studies of its optical counterpart .
the @xmath73 relation is presented in figure 6 .
a least squares fit , taking @xmath12 as the independent variable and using 48 ellipticals in the coma , gives @xmath74 with an rms scatter of 0.50 mag . using the aperture magnitudes ( measured over an aperture of 20 arcsec .
diameter- column 4 of table 1 ) gives a shallower slope of @xmath75 and a smaller rms scatter of 0.32 mag .
the optical relation , using the total v magnitudes for the same sample of ellipticals has a slope and rms scatter of @xmath76 and 0.49 mag . respectively .
the similarity in the slope and rms scatter between the near - infrared and optical @xmath3 relations implies that dynamical and structural characteristics in ellipticals are probably as important as the photometric parameters ( i.e. stellar population , metallicity and age ) in defining the mass luminosity relation in these systems .
this will be explored in the following discussion .
taking the observational errors in k ( 0.03 mag . ) , @xmath53-(0.03 dex ) and sky background estimate ( 0.02 mag . ) , gives an observational error of 0.19 mag . in the @xmath73 relation . compared with the observed scatter of 0.50 mag .
, an intrinsic scatter of 0.46 mag . in the @xmath73 relation is estimated .
it is known that the colours of elliptical galaxies are correlated with their absolute magnitudes and indirectly , with their mass , in both optical ( faber 1973 ; visvanathan 1983 ) and infrared ( persson et al 1979 ; mobasher et al 1986 ) wavelengths .
the existence of such a relation has important implications for studies of evolution of these galaxies .
this colour - absolute magnitude ( mass ) correlation can therefore be translated into a more direct relation between the colours and velocity dispersions in ellipticals .
using the v@xmath9k colours measured over an aperture of 20 arcsec .
diameter for the sample of 47 galaxies with available such data from table 1 , the @xmath77 relation is presented in figure 7 .
a least squares fit gives @xmath78 with an rms scatter of 0.083 mag . in @xmath39 colours .
using the stellar synthesis models of worthey ( 1994 ) , the @xmath79 relations due to changes in age and metallicity are predicted and compared with the observed relation in figure 7 .
the models are calculated for a given metallicity , [ fe / h]=0.25 , and a change in age from 5 to 17 gyrs , and a range in metallicity,@xmath80 [ fe / h ] @xmath81 , at a constant age of 12 gyrs .
the models are normalised to have [ fe / h]=0.25 and an age of 12 gyrs at @xmath82 mag .
, the values expected for local ellipticals .
they are consistent with the observed @xmath79 relation , again indicating that changes in either age or metallicity ( or a combination of them ) among the ellipticals are , at least , partly responsible for this relation .
however , these models can not disentangle the effects of age and metallicity on the @xmath79 plane ( fig .
recent studies also propose the presence of a mass - metallicity relation among the cluster ellipticals ( kodama & arimoto 1996 ) with the metallicity being the more fundamental parameter . to explore
if the above results are model dependent , we investigate the behaviour of the observable parameters in ellipticals to changes in age and metallicity , using independent stellar synthesis models ( bruzual and charlot 1996 ) . at a fixed age of 12 gyrs and over the metallicity range @xmath83 [ fe / h ] @xmath84
, we predict a slope of @xmath85 on the ( v - k)mg@xmath8 plane which is close to @xmath86 found for a fixed metallicity of [ fe / h]=0.093 and a range in age from 5 to 12 gyrs .
these results are used to predict the @xmath87 slope ( ie . the relation in fig .
7 ) , using the empirical mg@xmath8-@xmath12 relation derived from the present sample .
we find that the ( v@xmath9k)-@xmath12 relations , predicted by bruzual and charlot ( 1996 ) model , have similar slopes due to changes in metallicity ( at a given age ) or age ( at a given metallicity ) .
this confirms that the results in figures 5 and 7 are independent of the models used ( ie .
the metallicity and age trends are degenerate ) .
the @xmath79 relation is independent of distance and hence , can be used to study evolution of ellipticals at different environments ( i.e. in field and clusters ) and redshifts .
this relation can also be used to estimate velocity dispersion of ellipticals at high redshifts ( for which a spectroscopic measurement of @xmath12 is difficult ) , using their v@xmath9k colours and allowing for evolution with redshift . the accuracy in log(@xmath12 )
values , estimated from the ( v@xmath9k)-@xmath12 relation , is 0.11 dex .
the near - infrared fp found here differs , at a high confidence level , from that expected from a pure virial theorem if light directly traces the mass and if ellipticals are dynamically and structurally homologous ( @xmath88 ; where @xmath89 is the effective surface brightness in units of luminosity per area and @xmath90 is the effective diameter ) .
this leads to a relation between the near - ir mass - to - luminosity ratios @xmath91 and mass @xmath92 of the ellipticals ( fig .
8)- ( faber et al 1987 ; djorgovski 1987 ) .
the @xmath7 and mass estimates in figure 8 are in solar units , assuming @xmath93 for the coma and @xmath94 km / sec / mpc .
this relation is consistent with @xmath95 where @xmath96 , in excellent agreement with @xmath97 estimated from a k - band study of 59 ellipticals in 5 nearby clusters ( pahre et al 1996 ; see also recillas - cruz et al 1990 , 1991 ) .
to investigate the origin of this relation and explore if it is caused by differences in the dark - to - luminous matter distribution among the ellipticals or is due to changes in metallicity , age or stellar population , we extend this study to optical wavelengths , using the same sample of galaxies ( fig .
the optical relation has a steeper slope of @xmath98 , indicating that over the same range in mass , the optical and infrared m / l ratios change by factors of 3 and 2.3 respectively .
the slight wavelength dependence of the slope of the m / l @xmath99 m relation implies that changes in the luminous - to - dark matter distribution and/or luminosity profile among the ellipticals is not the only parameter responsible for the observed relations .
other observable parameters ( see below ) , are also likely to be as important . in order to have an entirely independent measurement of the slope of the m / l @xmath99 m relations in the two passbands , we estimate both the m / l ratios and masses of individual galaxies , using their respective effective diameters in the near - ir and optical wavelengths .
however , this introduces the unphysical effect of the dependence of the dynamical mass on the wavelength in figure 8 .
therefore , the shift in the masses of the same object between the optical and near - infrared ( m / l ) @xmath99 m relations in figure 8 is due to differences in the shape of the surface brightness profiles in the two bands , and to observational uncertainties in the values of the effective diameters .
nevertheless , using the effective diameters measured only in one passband to estimate the mass of galaxies , the slope of the m / l @xmath99 m relations remain the same as above , indicating that the conclusions in this section do not depend on different effective diameters for individual galaxies , used to estimate the mass .
the effects of stellar population , metallicity and age on the near - infrared mass - to - luminosity ratios of the ellipticals are investigated in fig 9 , using v - k colours and mg@xmath8 line indices . employing the stellar synthesis models of worthey ( 1994 ) , changes in the ( m / l)@xmath100 with age and metallicity
are calculated for 5 gyrs @xmath101 17 gyrs at constant metallicities of [ fe / h]=-0.25 , 0 and 0.25 and for a range in metallicity corresponding to @xmath102<0.5 $ ] at a constant age of 12 gyrs .
the models , normalised to have @xmath103 at 12 gyrs and [ fe / h]=0 , are compared with the data in figs 9a and 9b .
there is only a weak dependence of the near - ir mass - to - luminosity ratio on metallicity .
furthermore , the effects of age and metallicity in ellipticals , as shown in fig . 9 , decouple when using their ( m / l)@xmath100 estimates ( ie .
the age and metallicity sequences are orthogonal ) .
assuming the same ` collapse factor ' ( the ratio of the radius covering half the total mass to the effective radius ) for all the elliptical galaxies in our sample .
figures 9a and 9b suggest that the age is responsible for the observed trends on both the ( m / l)@xmath104mg@xmath8 and ( m / l)@xmath104(v@xmath9k ) diagrams while metallicity mainly contributes to the scatter in these relations .
the near - infrared fp is less affected by differences in line blanketing , metallicity and stellar population among the ellipticals , compared to its optical counterpart .
this implies a fp relation with a potentially reduced scatter at the near - infrared wavelength .
therefore , the similarity of the rms scatter between the near - ir ( 0.072 dex ) and optical ( 0.074 dex ) fps found here , was unexpected and could have important implications for constraining models of formation of ellipticals .
understanding the origin of this scatter provides a challenge in studying the evolution of elliptical galaxies .
study of the scatter around the optical fp with ellipticity ( ie .
shape ) has found no relation between the two quantities ( jorgensen et al 1996 ; jorgensen et al 1993 ) .
as discussed in section 3.4 , it is likely that changes in the @xmath7 among ellipticals is responsible for at least part of the scatter in their near - infrared fp .
this may be caused by differences in age and/or metallicity ( figure 9 ) , contributions from the asymptotic giant branch ( agb ) population to their near - infrared light or differences in matter distribution and the internal dynamics ( ie .
orbital anisotropy or rotation ; bender et al 1993 ) among the ellipticals .
however , similarity of the scatter between the optical and near - infrared fp relations indicates that structural and dynamical differences rather than changes in age / metallicity or the agb contributions among the ellipticals are probably responsible for the observed scatter in their near - infrared fp . the tightness of the fp has been exploited in estimating relative distances between elliptical galaxies in the field or clusters .
however , the optical fp is affected by contributions from the young population or residual star formation .
indeed , it is likely that the relatively blue ellipticals , found in the great attractor region , are responsible for the proposed streaming motion ( guzman and lucey 1993 ) .
the near - infrared fp may be less sensitive to this effect ( see section 3.1 ) and hence , the fp at this wavelength is expected to have a more stable slope and zero - point .
therefore , this relation can be used to test the reality of the streaming motion and its dependence on the non - uniformity of stellar populations among the ellipticals . for this purpose
, we have performed near - ir ( k - band ) imaging observations of a sub - sample of ellipticals in the great attractor region . using the present sample of ellipticals in the coma cluster , we establish the zero - point of the field near - ir fp in order to use this to explore the sensitivity of the streaming motion to changes in stellar populations among the ellipticals .
these results will be presented in a forthcoming paper .
recent studies have indicated that elliptical galaxies are not the homogeneous population once assumed and are likely to have a complex evolutionary history involving interaction , starburst and infall of material ( white and frenk 1991 ; cole et al .
this leads to changes in the luminosity , mass and morphology of these galaxies with time , implying that their progenitors may be distinctly different from the galaxies we see today .
a study of the fp of ellipticals at high redshift is needed to address these scenarios .
recently , using the high resolution imaging capability of the hubble space telescope and spectroscopic facilities on the largest ground - based telescopes ( mmt and keck ) , it has become possible to explore the evolution of the optical fp and m / l of ellipticals with redshift ( franx et al 1996 ; van dokkum & franx 1996 ; barger et al 1996 ) .
this reveals a fp similar to those at low redshift with a comparable scatter .
however , the observed change in the optical m / l is found to be smaller than that expected from population synthesis models based on a given metallicity , imf and formation redshift ( van dokkum & franx 1996 ) . with the installation of the near - infrared ( nicmos ) detectors on the hst , it is now possible to extend this study to longer wavelengths .
the near - infrared light mainly measures the integrated star formation and hence , is more closely related to the mass function of the ellipticals .
therefore , study of the evolution of near - infrared fp and m / l of ellipticals with redshift constrains the infall and merging in these galaxies and is needed for a careful interpretation of the evolution of the luminosity function of galaxies .
the results of this study can be summarised as follows : 1 .
the near - infrared ( 2.2 @xmath0 ) fundamental plane of elliptical galaxies in the coma cluster is studied , using a sample of 48 galaxies .
this shows an rms scatter of 0.072 dex , similar to its optical counterpart , corresponding to an accuracy of @xmath1 in distances derived from this relation . +
no relation is found between the scatter around the near - infrared fp and the v@xmath9k colours or mg@xmath8 line strengths in ellipticals , probably implying negligible contribution to this relation due to changes in metallicity or the young population among the coma cluster ellipticals .
the agb stars are proposed as the likely source responsible for the observed scatter in the near - ir fp .
allowing for observational uncertainties , we estimate an intrinsic scatter of 0.060 dex in this relation .
2 . the @xmath61 and @xmath73 relations are studied .
these have an rms scatter of 0.073 dex and 0.50 mag . respectively , compared to scatters of 0.077 dex and 0.49 mag . for their optical counterparts .
the wavelength dependence of the slopes of these relations lead to correlations between @xmath105 and ( v - k ) with the mass ( @xmath12 ) of ellipticals . using stellar synthesis models
, we interprete these as due to an age or metallicity ( or a combination of these two ) effect . however , it is not possible to disentangle the effect of age and metallicity , using these parameters .
the relation between the m / l ratio and the mass of ellipticals ( @xmath106 ) in the coma cluster is studied .
the slight difference in the slope of this relation between the infrared ( @xmath5 ) and optical ( @xmath6 ) wavelengths is caused by a mass - metallicity relation and not due to changes in the luminous - to - dark matter distribution among the ellipticals .
4 . using the stellar synthesis models
, we disentangle the effects of age and metallicity on the ( m / l)@xmath104mg@xmath8 and ( m / l)@xmath104(v@xmath9k ) diagrams . the trend in these relations
is interpreted as an age sequence while metallicity mainly contributs to its scatter .
we propose the use of ( m / l)@xmath100 to separate the effects of age and metallicity in elliptical galaxies .
we propose to use the near - infrared fp of ellipticals in the coma cluster to fix the zero - point of its corresponding field sample in the great attractor region .
this is required in order to explore the effect of stellar population and metallicity on the estimates of the streaming motion of galaxies .
* acknowledgements : * we acknowledge support from the pparc ( bm ) , the nsf through grant ast 95 - 29098 ( rg ) , the royal society ( aas ) and the space telescope science institute through a hubble fellowship grant hf-1055.01 - 93a ( sz ) .
we are grateful to an anonymous referee for careful reading of the manuscript .
barger , a.j .
, aragn - salamanca , a. , butcher , h.r . ,
couch , w.j . ,
dressler , a. , ellis , r.s . ,
oemler , a. , sharples , r.m . ,
smail , i. 1996 , in preparation bender , r. , bursterin , d , faber , s. m. 1993 ap.j,411 , 153 bower , r.g . ,
lucey , j.r . , ellis , r.s .
1992 , mnras , 254 , 589 bower , r.g . ,
ellis , r.s . , rose , j.a . ,
sharples , r.m .
1990 , aj , 99 , 530 bruzual , g. , charlot , s. 1997 in preparation burstein , d. , faber , s.m . & dressler , a. 1990 , ap.j .
354 , 18 .
capelato , h. v. , de carvalho , r. r. , carlberg , r. g. 1995 , apj .
451 , 525 ciotti , l. , lanzoni , b. , renzini , a. 1996 , mnras , 282 , 1 cole , s. , aragn - salamanca , a. , frenk , c.s . ,
navarro , j.f .
, zepf , s. 1994 , mnras , 271 , 781 de carvalho , r.r . , djorgovski , s. 1992 , apjl , 387 , l49 djorgovski , s. , davis , m. 1987 apj , 313 , 59 djorgovski , s. 1987 , in starbursts and galaxy evolution , proceedings of the moriond astrophysics workshop , ed . t. x. thuan ( gif sur yvette : editions frontieres ) , 549 dressler , a. 1987 , ap.j . 317 , 1 .
dressler , a .
, lynden - bell , d . ,
burstein , d . ,
davies , r.l .
, faber , s.m . , terlevich , r.j . ,
wegner , g. 1987a apj , 313,42 dressler , a .
, faber , s.m . , burstein , d . ,
davies , r.l . ,
lynden - bell , d .
terlevich , r.j . ,
wegner , g. 1987b apj , 313 , l37 ellis , r.s .
1996 , apj , in press , astro - ph/9607154 faber , s.m .
1973 ap.j .
179 , 731 faber , s.m . ,
dressler , a . ,
davies , r.l . , burstein , d . ,
lynden - bell , d . ,
terlevich , r.j . ,
wegner , g. 1987 , in nearly normal galaxies , ed .
faber , ( new york : springer - verlag ) , 175 faber , s.m . , wegner , g. , burstein , d .
, davies , r.l . ,
dressler , a .
, lynden - bell , d . ,
terlevich , r.j .
1989 , apjs , 69 , 763 faber , s.m . , burstein , d .
1988 , in large - scale motions in the universe . a vatican study week , eds .
v.c . rubin & g.v.coyne ( princeton : princeton university press ) , 115 franx et al .
1996 mnras . in press
gregg , m. 1995a aj .
110 , 1052 .
gregg , m. 1995b apj .
443 , 527 .
guzmn 1994 ph.d . thesis .
university of durham guzmn , r. , lucey , j.r . ,
carter , d. , & terlevich , r. j. 1992 mnras , 1992 , 257 , 187 .
guzmn , r. , lucey , j.r . ,
bower , r.g .
1993 mnras , 265 , 731 guzmn , r. , lucey , j.r .
1993 mnras , 263 , l47 jorgensen , i. , franx , m. , kjaergaard , p. 1993 ap.j .
411 , 34 jorgensen , i. , franx , m. , kjaergaard , p. 1996
mnras , 280 , 167 .
kaiser , n. 1988 in : large scale structure and motions in the universe , ictp .
kodoma & arimoto 1997 , a & a , in press kormendy j. 1977 ap.j .
218 , 333 larson , r.b . , tinsley , b.m . ,
caldwell , c.n .
1980 ap.j .
237 , 692 lucey , j.r . , guzman , r. , carter , d. & terlevich , r.j .
1991 mnras . 253 , 584 lucey , j.r . ,
guzman , r. , steel , i. , & carter , d. 1997 mnras .
287 , 899 lynden - bell , d . ,
faber , s.m . ,
burstein , d . ,
davies , r.l . ,
dressler , a .
, terlevich , r.j . ,
wegner , g. 1988 apj , 326 , 19 mobasher , b. , ellis , r.s .
& sharples , r.m . 1986 mnras 223 , 11 mobasher , b. & james , p.a .
1996 mnras 280 , 895 pahre , m.a . , djorgovski , s.g .
, de carvalho , r.r .
1996 apjl .
, 453 , l17 .
persson , s.e .
, frogel , j.a . , aaronson , m. 1979 , apjs , 39 , 61 pickles , a.j .
1985 , ap.j .
296 , 340 prugniel , p. & simien , f. 1996 , a & a in press recillas - cruz , e. , carrasco , l. , serrano , p.g . , cruz - gonzalez , i. 1990 , a & a , 229 , 64 recillas - cruz , e. , carrasco , l. , serrano , p.g .
, cruz - gonzalez , i. 1991 , a & a , 249 , 312 renzini , a. , 1995 , in gilmore g. , van der kruit , p. , eds , stellar populations .
kluwer , dordrecht , p. 325
renzini , a. , ciotti , l. 1993 , apj 416 , l49 rose , j. a. 1985 a.j .
90 , 1927 silk , j. 1989 , ap.j . 345 ,
l1 van dokkum , p. , franx , m. 1996 mnras , 281 , 985 .
visvanathan , n. 1983 ap.j .
275 , 430 white , s.d.m . &
frenk , c.s .
1991 ap.j .
379 , 52 worthey , g. 1994 apjs , 95 , 107 zepf , s. , silk , j. 1996 apj 466 , 114
* figure 1 . ( a ) * comparison between the k - band magnitudes of the coma ellipticals in common with galaxies observed in bower et al ( 1992 ) .
the magnitudes are measured over an aperture of @xmath28 diameter , used in bower et al .
* figure 1 .
( b ) * comparison between the k - band isophotal diameters ( at 16.5 mag / arcsec.@xmath32 ) of a sub - set of 12 ellipticals in the coma cluster observed both at the ukirt ( ircam3 ) and the aat ( iris ) . *
figure 3 .
* deviations from the near - infrared fundamental plane ( @xmath107 ) are plotted against @xmath47 ( v@xmath9k ) colour residuals from the ( v@xmath9k)-log(@xmath12 ) relation .
the lack of correlation implies that metallicity or age do not significantly contribute to the obseved scatter on the near - infrared fp of ellipticals in the coma cluster .
* figure 3 .
( b ) . * the same as fig .
3a for the @xmath47 ( mg@xmath8 ) residuals from the mg@xmath8log(@xmath12 ) relation .
again , lack of correlation indicates that changes in the young population or possible environmental dependence among the ellipticals in the coma cluster do not contribute to the scatter on their near - infrared fp
. * figure 5 ) . * the log(d@xmath108/d@xmath100)-log(@xmath12 ) relation for ellipticals in the coma cluster .
the lines are calculated from the stellar synthesis models of worthey ( 1994 ) .
they correspond to changes in age ( from 5 to 12 gyrs ) at a constant metallicity of [ fe / h]=0.25 ( dotted line ) and changes in metallicity ( @xmath109 [ fe / h]@xmath81 ) at a constant age of 12 gyrs ( solid line ) .
the models are transformed to the observed parameters , using the empirical log(@xmath12)-mg@xmath8 and log(d@xmath108/d@xmath100)- ( v - k ) relations ( from table 1 ) , with the mg@xmath8 and v - k values for a given age and metallicity taken from worthey ( 1994 ) models .
the models are normalised to have log(d@xmath108/d@xmath100)=@xmath110 at [ fe / h]=0.25 and 12 gyr age .
* figure 7 ) . *
the ( v@xmath9k)@xmath9log(@xmath12 ) relation for the coma ellipticals .
colours are measured over an aperture of @xmath23 diameter .
the lines are calculated from the stellar synthesis models of worthey ( 1994 ) .
they correspond to changes in age from 5 to 17 gyrs at a constant metallicity of [ fe / h]=0.25 ( dotted line ) and changes in metallicity from [ fe / h]=-0.25 to [ fe / h]=0.5 at a constant age of 12 gyr ( solid line ) .
the model estimates for @xmath12 are calculated using the empirical relation as in fig .
it is not possible to disentangle the effects of age and metallicity from this relation .
* figure 8) .
* correlation between the near - infrared ( @xmath112 ) and optical ( @xmath113 ) mass - to - luminosity ratio ; @xmath114 and mass ; log ( m ) = @xmath115 for the coma ellipticals .
the same sample of galaxies are used in both cases . the zero - points correspond to @xmath116 and @xmath117 for the near - infrared and optical wavelengths respectively and @xmath118 .
both the mass - to - luminosity ratio and mass are in solar units .
for each wavelength , the mass corresponding to the effective diameter for that wavelength is plotted , leading to a slight difference in the mass estimated for the same objects in the near - infrared and optical relations .
* figure 9 ( a ) . * changes in the near - infrared mass - to - luminosity ratios in solar units ( calculated as in fig . 8) with the v@xmath9k colours .
the lines are predictions from the stellar synthesis model of worthey ( 1994 ) and correspond to a change in age from 5 to 17 gyrs at a constant metallicity of [ fe / h]=-0.25 ( dashed line ) and [ fe / h]=0 ( dotted line ) and a change in metallicity from [ fe / h]=@xmath119 to [ fe / h]=0.25 at a constant age of 12 gyrs ( solid line ) .
the models are normalised to have log(m / l)@xmath100=@xmath120 at [ fe / h]=0 and an age of 12 gyrs . *
figure 9 ( b ) .
* the same as fig 9a for the mg@xmath8 features .
the lines are predictions from the stellar synthesis model of worthey ( 1994 ) and correspond to a change in age from 5 to 17 gyrs at a constant metallicity of [ fe / h]=0.25 ( dashed line ) and [ fe / h]=0 ( dotted line ) and a change in metallicity from [ fe / h]=@xmath119 to [ fe / h]= 0.5 at a constant age of 12 gyrs ( solid line ) .
the models are normalised to have log(m / l)@xmath100=@xmath120 at [ fe / h]=0 and an age of 12 gyrs . | near - infrared ( 2.2 @xmath0 ) observations of a sample of 48 elliptical galaxies in the coma cluster have been carried out and used to study the near - ir fundamental plane ( fp ) of ellipticals in this cluster .
an rms scatter of 0.072 dex was found for this relation , similar to that of its optical counterpart , using the same sample of galaxies .
this corresponds to an uncertainty of @xmath1 in distances to individual galaxies derived from this relation .
the sensitivity of the near - infrared fp to the star formation or changes in metallicity and stellar population among the ellipticals was explored and found to be small .
although , a likely source of scatter in this relation is contributions from the asymptotic giant branch ( agb ) population to their near - infrared light . allowing for observational uncertainties , we find an intrinsic scatter of 0.060 dex in the near - ir fp .
the cluster galaxies presented here , provide the zero - point for the peculiar velocity studies , using the near - infrared fp .
changes in the slopes of the @xmath2 and @xmath3 relations of ellipticals between the optical and near - infrared wavelengths were investigated and found to be due to variations in metallicity or age ( or a combination of them ) . however , it was not possible to disentangle the effects of age and metallicity in these relations .
we find @xmath4 with @xmath5 at the near - ir and @xmath6 at the optical wavelengths , using the same sample of galaxies .
this relation is interpreted as due to a mass - metallicity effect or changes in age or the imf slope with mass . using evolutionary population synthesis models , we find that the effects of age and metallicity decouple on the @xmath7 vs. mg@xmath8 and @xmath7 vs. ( v@xmath9k ) diagrams .
the models suggest that the observed trends on these relations may be due to an age sequence while metallicity mainly contributing to the scatter .
galaxies : elliptical galaxies : photometry galaxies : fundamental parameters galaxies : stellar content galaxies : structure galaxies : evolution galaxies : clusters : individual : coma infrared : galaxies |
since its first detections in the early 2000 s ( @xcite ) , weak gravitational lensing , the coherent distortion of distant galaxies by intervening dark matter , has become a premier tool to constrain the cosmological model ( for reviews , see e.g. @xcite ) and has been shown to be the most promising probe of dark energy ( @xcite ) .
surveys optimization and systematics minimizations in both software and hardware have been investigated ( @xcite ) , favoring wide surveys , with well - controlled , stable point spread function , with comprehensive photometric redshift follow - up .
those characteristics are shared by ambitious upcoming large area surveys , such as the large synoptic survey telescope ( lsst ) , the panoramic survey telescope & rapid response system ( pan - starrs ) , euclid and the joint dark energy mission ( jdem ) .
they will provide us with a large amount of high quality imaging , well fitted to measure dark energy with weak lensing . anticipating a significant improvement in removing systematics , we are left with the question of how to best
extract the cosmological information out of the data .
for instance , one must decide how to optimally capture non - gaussianities and break the degeneracies between cosmological parameters as constrained by the extensively studied and measured weak lensing power spectrum . in this paper , we ignore primordial non - gaussianities , and consider non - gaussianities due to the growth of structures only .
the power spectrum , the fourier transform of the 2-point correlation function of the shear field , has been the most used measurement so far , both from ground ( e.g. @xcite ) and from space ( e.g. @xcite ) . the introduction of tomography , the three - dimensional information of the shear field , that captures structure evolution , has tightened constraints on cosmological parameters , in particular the matter density @xmath0 and the amplitude of density fluctuations @xmath1 ( @xcite ) . despite its success ,
the power spectrum leaves us with well known degeneracies between parameters , that we must break by combining it with other measurements and/or probes . the power spectrum capturing only the gaussian features of the field , it is natural to introduce measures of non - gaussianities , and combine them with the power spectrum , to break degeneracies and tighten constraints .
higher - order correlations measure the matter density field s non - gaussian features .
the lowest one , the weak lensing 3-point correlation function ( 3pcf ) , and its fourier transform , the bispectrum , have been given a lot of attention in the past few years ( @xcite ) .
however , a clean measurement of the 3pcf is difficult , thus few papers have reported observational measurements so far . @xcite and @xcite have measured the skewness , but a full measurement of the shear 3pcfs , or of the convergence bispectrum , is still to be done .
several efforts are underway , including an algorithm to measure the convergence bispectrum directly in fourier space ( @xcite ) . large - scale structures such as clusters of galaxies , the result of the non - linear evolution of density fluctuations , are the non - gaussian features that we want to take into account .
assessing their abundance as a function of various parameters , such as redshift , has been known as a powerful probe and used as this for several years ( e.g. @xcite ) .
in particular , tight constraints can be obtained on @xmath0 and @xmath1 . because the weak lensing clusters selection function is rapidly evolving with redshift , counting weak - lensing - selected clusters is less sensitive than using catalogs of x - ray , optical , or sunyaev - zeldovich ( sz ) clusters . despite this fact ,
constraining cosmology is possible with weak - lensing clusters only ( @xcite ) , as shown on real data by @xcite .
although such counts give weaker constraints than the weak lensing power spectrum , combining them with the power spectrum efficiently breaks degeneracies ( @xcite ) . in this paper ,
we show how combining the weak lensing power spectrum with the weak lensing bispectrum or counts of weak - lensing - selected clusters of galaxies will allow one to tighten constraints on cosmological parameters with upcoming weak lensing surveys . in particular , we show how we will be able to measure the equation of state @xmath2 , where @xmath3 is the pressure , and @xmath4 the density of the dark energy , down to the percent level . while other authors studied the combination of the power spectrum with cluster counts ( e.g. @xcite ) and with the bispectrum ( e.g. @xcite ) , we present , for the first time , a detailed and consistent comparison of both combinations at once . constraints on cosmological parameters have been studied by several authors ( e.g. @xcite ) .
most constraints predictions have so far been made using a fitting function for the non - linear power spectrum , such as @xcite or @xcite .
however , fitting functions tuned for a @xmath5cdm universe must be used with caution when one is varying @xmath6 ( @xcite ) .
we use the halo model to compute the weak lensing statistics . despite its simplicity ,
it is well suited to cosmological parameters prospects , since it allows us to vary the dark energy without extrapolating fitting functions out of @xmath7 .
appendix [ sect_hm ] summarizes the halo model description that we use in this paper .
the halo model allows us to define the power spectrum as the separated contributions of the resolved and unresolved parts of the weak lensing field .
such a separation , since it gets rid of highly non - gaussian features , can help improve constraints on cosmological parameters .
we consider four different types of cluster counts , with more or less intrinsic cosmological information : counts as a function of shear signal - to - noise ratio ( s / n ) only , counts as a function of mass , counts as a function of redshift , and counts as a function of mass and redshit .
we show that when combined with the power spectrum , counting clusters just as a function of their s / n gives constraints similar to those obtained when combining the power spectrum with counts of clusters taking the full mass and redshift information .
we then show how the weak lensing bispectrum captures non - gaussian features and breaks the power spectrum degeneracies as well as clusters counts .
we assume a euclid - like survey , 20,000 deg@xmath8 wide , with @xmath9 with median redshift @xmath10 , without any external ( e.g. from cmb ) priors , so as to show what can be done with a weak lensing survey alone .
section [ sect_method ] summarizes the methods we employ to constrain cosmological parameters with different weak lensing statistics .
section [ sect_results ] presents our results .
we conclude in section [ sect_ccl ] .
our halo model code will be made part of the public icosmo package ( @xcite ) .
the weak lensing selection function for clusters of galaxies has already been extensively studied , by using different kinds of smoothing functions or matched - filters ( e.g. @xcite ) .
we derive a simplified , ideal selection function , based on a matched - filter approach for the signal - to - noise ratio created by a halo .
we neglect projection effects and intrinsic alignments , which have been investigated e.g. by @xcite , @xcite and @xcite .
that selection function has been introduced , without the mathematical justification that follows , and tested on real data in @xcite .
we consider a spherically symmetric weak lens , decoupled from its surrounding and alone along the line of sight , characterized by its convergence @xmath11 , where @xmath12 is the distance from its center ( the convergence can be replaced by the shear without any loss of generality ) .
the number density of background sources is @xmath13 .
we assume that the noise @xmath14 , originating from poisson and intrinsic shape noise only , averages to 0 in circular shells and that its variance in the circular shell @xmath15 is given by @xmath16 , where @xmath17 is the r.m.s error on the shape measurement and @xmath18 is the number of lensed galaxies in shell @xmath15 .
we fit a theoretical model @xmath19 to the observable @xmath20 .
the signal - to - noise ratio created by the lens is defined as the ratio of an estimator of @xmath19 to its associated error , @xmath21 .
the estimator is defined by @xmath22 , where @xmath23 is a weight function to be adjusted so as to optimize the s / n . this is the case when @xmath24 , giving the optimized s / n as galaxies per arcmin@xmath8 , distributed along equation [ nz ] .
contours denote the s / n of galaxy clusters.,width=302 ] @xmath25 which is consistent with previously published expressions .
we will consider nfw halos ( @xcite ) only , the s / n of which is expressed as ( appendix [ app_nfw ] ) @xmath26 where @xmath27 is a function of the halo s concentration @xmath28 only , well fitted by @xmath29 for @xmath30 , and the function @xmath31 is defined by eq .
( [ eq_g ] ) . the quantity @xmath32 in eq .
( [ eq_snnfw ] ) describes the effect of the distribution of galaxy sources . in the above equations , @xmath33 is the angular - diameter distance to the lens , @xmath34 and @xmath35 parametrize the nfw halo ( see appendix [ sect_hm ] ) , and @xmath36 is defined in appendix [ app_nfw ] .
we parametrize the redshift distribution of source galaxies by ( @xcite ) @xmath37\ ] ] with @xmath38 , @xmath39 , and @xmath40 , where @xmath41 is the median redshift of the survey .
figure [ fig_mlim ] shows our fiducial survey s selection function , in the mass - redshift plane .
contours originate from the s / n of halos in that plane .
the halo model allows us to decompose the full three - dimensional matter power spectrum along @xmath42 where @xmath43 is computed on the _ unresolved _ part of the halos distribution , defined as the ensemble of halos with s / n smaller than a certain threshold @xmath44 ; @xmath45 is computed on the _ resolved _ part of the halos distribution , the ensemble of halos with @xmath46 .
those names come naturally from the separation made on a mass map between peaks ( defined as resolved structures with @xmath47 ) and the rest , unresolved part of the map .
the resolved part can be seen as the most non - gaussian features of the field .
( dashed ) , @xmath48 ( dash - dot ) and @xmath49 ( dash - dot - dot - dot).,width=302 ] the full weak lensing power spectrum @xmath50 follows the same decomposition .
figure [ fig_ps ] shows the full power spectrum together with power spectra computed on unresolved parts defined by the thresholds @xmath51 , @xmath48 and @xmath49 .
the effect of removing significant clusters is mostly apparent on intermediate scales .
it is less pronounced than when cutting the power spectrum with respect to halo masses , as shown by @xcite , because of the weak lensing selection function , that mixes masses when integrating the matter power spectrum on redshift .
the power spectrum is a measure of the gaussian field only , and one must rely on direct measures of non - gaussianity to complement it , like the bispectrum or cluster counts .
three point statistics are the lowest statistics to capture non - gaussian features in a statistical field .
the dark matter and the weak lensing bispectra have been the object of numerous studies ( e.g. @xcite ) , most commonly based on the perturbation theory on large scales and on the hyper - extended perturbation theory ( hept ) on smaller scales ( e.g. @xcite for a review ) .
here , we compare and combine the constraints from the weak lensing power spectrum and bispectrum evaluated in the same halo model .
this fills a gap in the literature , where such constraints come from the hept approximation only .
the halo model allows us to separate the contribution from the unresolved and the resolved parts of the field when computing the bispectrum , like we did for the power spectrum .
however , discarding the unresolved part from the bispectrum removes too much useful signal , and thus suppresses much power in the bispectrum .
therefore , we will always evaluate the bispectrum on the entire ( resolved and unresolved ) data field .
clusters of galaxies , tracing dark matter halos , have emerged from the primordial gaussian field through non - linear gravitational clustering , and are those non - gaussianities that one can detect with higher order statistics . or one can simply estimate their abundance to take non - gaussianities into account . in this paper , we concentrate on how to break the degeneracies from the power spectrum by combining it with non - gaussian features . as a result
, we will not focus on the constraints that halo counts alone can bring , but on how they improve the constraints from the power spectrum .
weak - lensing - selected clusters are detected on a mass map derived from the data used to measure the power spectrum and the bispectrum .
hence , they come at no extra cost , contrary to x - ray or optically - selected clusters .
defined as peaks with s /
n greater than a certain threshold @xmath44 , they are the resolved part introduced above .
once halos are selected , one can add information to their single s / n distribution , e.g. by introducing their redshift and mass , and thus measure their abundance as a function of those quantities .
hereafter , we investigate the constraints on cosmological parameters provided by four kinds of halo counts . with increasing information , we look at ( 1 ) counts as a function of halo s s / n only ; ( 2 ) counts as a function of halo s mass ; ( 3 ) counts as a function of halo s redshift ; and ( 4 ) counts as a function of halo s redshift and mass . in the remainder of this paper , we assume that the weak lensing selection function is perfectly known .
we neglect the nuisance parameters commonly used to account for the imperfect knowledge of the relation between an observable and a cluster s mass ( e.g. @xcite ) .
fisher matrices allow one to characterize how a set of observables @xmath52 is able to constrain a set of parameters @xmath53 around a fiducial model .
the associated fisher matrix is defined by ( e.g. @xcite ) @xmath54 where @xmath55 is the associated likelihood .
given a fiducial model , the inverse of the fisher matrix @xmath56 estimated in its neighborhood provides a lower limit to the parameters covariance matrix .
it thus quantifies the best statistical error that can be reached on the parameters , @xmath57 , where @xmath58 is the 1@xmath59 error on parameter @xmath60 marginalized over other parameters @xmath61 . the fisher matrices for the weak lensing power spectrum , bispectrum and cluster counts can be found e.g. in @xcite and @xcite .
we aim to determine how to best capture non - gaussianities in order to break the degeneracies from the power spectrum alone .
we thus need to combine the power spectrum with the bispectrum or the halo counts . to combine two observables @xmath62 and @xmath63
, we create the data vector @xmath64 .
the associated fisher matrix is @xmath65^{-1 } { \rm d}_{,\beta}$ ] . the rigorous estimation of the combined observables covariance matrix
, @xmath66 requires the knowledge of the cross - covariance @xmath67 .
the weak lensing power spectrum is dominated by the most massive halos contribution , those which are the most likely to be detected and taken into account in cluster counts .
hence , the power spectrum and halo counts are not independent observables : their cross - covariance is non - zero , and must be accounted for .
nonetheless , @xcite showed that neglecting it changes the errors by only a few percent . here
, we take a different approach , by separating the contribution of different halos to the power spectrum ( eq .
[ eq_gng ] ) .
we combine the counts of clusters with s / n greater than the threshold @xmath44 ( the resolved part ) with the power spectrum created by all halos with s / n smaller than the same threshold @xmath44 ( the unresolved part ) . if we neglect the clustering between halos , the power spectrum in the unresolved part and the resolved clusters are uncorrelated : the covariance between cluster counts and the power spectrum vanishes , and we can simply add the fisher matrices , @xmath68 where the superscript @xmath69 stands for the power spectrum evaluated on the unresolved part .
as shown by @xcite , the cross - covariance between the power spectrum and the bispectrum has no gaussian feature , but arises from the five - point correlation function of the shear field . in such a case ,
it is safe to approximate the fisher matrix of the combination between the power spectrum and the bispectrum by the sum of the individual fisher matrices , @xmath70
in this section , we compute and compare the errors on cosmological parameters , that can be reached by measuring the power spectrum alone , and when combining it with either counts or the bispectrum .
we assume a fiducial cosmology described by the parameters @xmath71 .
the universe s curvature is a free parameter , set by the combination of @xmath0 and @xmath72 .
the evolution of the dark energy equation of state is parametrized by @xmath73 , @xmath74 being the scale factor .
we assume a euclid - like survey , 20,000 deg@xmath8 wide , with @xmath9 with median redshift @xmath10 .
we assume that the intrinsic shape r.m.s is @xmath75 . to count halos , we use 20 linear redshift bins , spanning the entire interval accessible to our fiducial survey ( @xmath76 ) - the highest bins being empty - and 15 logarithmic mass bins ( @xmath77 - where the lower bound is small enough so that all clusters detectable by their weak lensing signal are accounted for ) to compute halo counts as a function of redshift and/or mass , and 15 linear s / n bins for s / n peak counts ( @xmath78 ) .
we estimate the power spectrum in 10 redshift bins ( @xmath76 ) . for numerical reasons
, we can not use more than 3 redshift bins to compute the tomography bispectrum .
this is not problematic , since @xcite showed that the tomography bispectrum s / n quickly converges to its maximum value , and almost reaches it for three redshift bins . before reporting our results
, we would like to refer to the work of @xcite in which he shows the limitations of the fisher matrix formalism . in particular , one must pay particular attention to ill - conditioned fisher matrices , the inversion of which is likely to be wrong . here , we check the condition number of all of our fisher matrices , and consider as good only those with a small enough condition number so that their inversion can be trusted .
we find that the fisher matrices for counts as a function of s / n , of mass and of redshift , are all ill - conditionned .
therefore , such halo counts can not give reliable fisher constraints by themselves .
this is true for our particular set of parameters though , and could not be true on other parameter spaces . looking for such spaces
is beyond the scope of this paper , and we will give constraints from halo counts only when combined with the power spectrum , for which the condition number is low enough to be safe .
ccccccccccc & @xmath79 & @xmath80 & @xmath81 & @xmath82 & @xmath83 & @xmath84 & @xmath85 & @xmath86 & @xmath87 & @xmath88 + @xmath0 ( 0.3 ) & 0.0049 & 0.0027 & 0.0047 & 0.070 & 0.0024 & 0.0024 & 0.0026 & 0.0026 & 0.0021 & 0.0023 + @xmath72 ( 0.7 ) & 0.024 & 0.012 & 0.019 & 0.254 & 0.0099 & 0.010 & 0.011 & 0.011 & 0.0087 & 0.0094 + @xmath89 ( 0.04 ) & 0.015 & 0.015 & 0.016 & 0.656 & 0.013 & 0.014 & 0.014 & 0.014 & 0.008 & 0.008 + @xmath1 ( 0.9 ) & 0.0086 & 0.0039 & 0.0065 & 0.112 & 0.0032 & 0.0034 & 0.0036 & 0.0036 & 0.0029 & 0.0033 + @xmath90 ( 0.7 ) & 0.091 & 0.086 & 0.122 & 3.59 & 0.077 & 0.079 & 0.079 & 0.085 & 0.048 & 0.051 + @xmath91 ( -1 ) & 0.040 & 0.041 & 0.031 & 0.283 & 0.023 & 0.024 & 0.025 & 0.025 & 0.016 & 0.016 + @xmath92 ( 0 ) & 0.147 & 0.133 & 0.156 & 2.179 & 0.062 & 0.063 & 0.064 & 0.064 & 0.055 & 0.063 + @xmath93 & 0.022 & 0.015 & 0.013 & 0.252 & 0.014 & 0.014 & 0.015 & 0.015 & 0.0049 & 0.0045 + @xmath94 ( 1 ) & 0.020 & 0.021 & 0.038 & 1.00 & 0.018 & 0.018 & 0.019 & 0.020 & 0.011 & 0.012 + + + the goal of this paper is to compare how cluster counts and the bispectrum capture non - gaussianities and break degeneracies left over by the power spectrum .
consequently , we ignore combining cluster counts with the bispectrum , as well as combining the three observables ( power spectrum , bispectrum and cluster counts ) at a time .
we evaluate fisher matrices for power spectra estimated both on all the data and on the unresolved parts defined by different @xmath44 , for the corresponding halo counts ( halos with @xmath46 ) , and for the bispectrum , and the combinations introduced in section [ sect_fisher ] .
figure [ fig_errors ] shows the expected errors that they provide , marginalized over all eight parameters , as a function of the s / n threshold @xmath44 . the power spectrum @xmath95 is estimated on the unresolved part ( dashed red ) , then combined with clusters counts of clusters ( made as a function of mass and redshift - solid green- , redshift only -solid blue- , mass only -solid cyan- and s / n only -solid purple ) .
the dotted black flat line shows the errors brought by the measurement of the power spectrum on the entire data set . discarding the contribution of the most massive halos ( i.e. the most non - gaussian features ) from the power spectrum improves the errors on @xmath0 and @xmath1 , the errors being smallest for an optimal threshold @xmath96 .
although the counter - intuitive increase in the performance that we observe when discarding the resolved part of the data is not so surprising , since the impact of non - gaussianities on the power spectrum is lowered ( @xcite observe a similar trend for sunyaev - zeldovich clusters ) , we should note here that it could be exacerbated if our analysis is close to the fisher matrices formalism s limits .
moreover , there may be an optimal weighting scheme that would both allow for the signal in the resolved part and minimize its non - gaussian errors .
smaller errors on the cosmological parameters would then be expected , since more information ( with minimal noise ) would be taken into account than that we use here with the unresolved part .
investigating these issues is beyond the scope of this paper , and we defer it to a later study .
the dark energy equation of state parameters @xmath91 and @xmath92 are less affected by the separation of the power spectrum into the resolved and the unresolved parts . although cluster counts do not provide strong constraints by themselves ( see below , table [ table_constraints ] ) , combining them with the power spectrum improves the errors it provides , in particular when low s / n clusters are taken into account .
when the clusters considered are significant enough , all four kinds of counts perform as well at capturing non - gaussianities : the power spectrum provides most of the fisher information , and clusters help by breaking parameters degeneracies , independently of the way they are considered .
figure [ fig_errors ] shows that this is true for @xmath97 , a safe threshold to discard false detections ( @xcite ) .
therefore , that makes s / n peaks counts a direct and efficient probe to combine with the power spectrum , as shown e.g. by @xcite .
the dash - dotted yellow line on fig . [ fig_errors ] shows the errors given by the bispectrum , and the long - dashed orange line shows those given when combining it with the power spectrum . at low s
/ n thresholds , when combined with the power spectrum , the bispectrum and clusters counts give similar constraints on the four parameters @xmath0 , @xmath1 , @xmath91 and @xmath92 . at higher thresholds
the bispectrum gives better constraints than clusters counts on the dark energy parameters , yet errors remain comparable for @xmath0 and @xmath1 .
table [ table_constraints ] lists the marginalized errors on all of our eight parameters , as well as those of the dark energy equation of state @xmath93 at the pivot point ( section [ sect_de ] ) , for the measurements and combinations considered , with @xmath98 .
halo counts ( excepted counts as a function of redshift and mass ) are plagued by ill - conditioned fisher matrices , and we do not consider them alone . nevertheless , since counts as a function of s / n , of mass alone and redshift alone contain less information than counts as a function of mass and redshift , their expected constraints are weaker than those listed in table [ table_constraints ] for counts as a function of mass and redshift .
we also list in the table the errors provided when combining the bispectrum with the power spectrum estimated on all the data .
although this combination is not explicitly shown on figure [ fig_errors ] , it can be drawn from the high s / n threshold - limit of the combination of the bispectrum with the power spectrum estimated on the unresolved part ( long - dashed orange line ) .
the table emphasizes the conclusions given by figure [ fig_errors ] .
the four combinations of clusters counts with the power spectrum give similar constraints , that are comparable to those provided by the combination of the bispectrum and the power spectrum .
furthermore , although the bispectrum and the power spectrum give comparable errors , combining them breaks degeneracies and lowers errors by a factor @xmath99 .
@xcite looked at how several statistics ( clusters counts , bispectrum , skewness , kurtosis ) break the @xmath0-@xmath1 degeneracy , and reported the high efficiency of cluster counts .
they found that the skewness could also discriminate against models , but that the bispectrum , when using only equilateral triangles , gave poor results .
we expand their work by taking into account all triangle configuration when computing the bispectrum , and add it to clusters counts as an efficient way to break degeneracies ( see section [ sect_csand ] ) . in particular , fig .
[ fig_errors ] shows that this conclusion is true not only for the @xmath0-@xmath1 degeneracy , but also for the dark energy equation of state . ) , as a function of the threshold between the resolved and the unresolved parts , for the same measurements as in fig .
[ fig_errors ] , labeled with the same colors and line styles .
thick lines denote measurements made with the power spectrum estimated on the unresolved part , and thin lines take the entire information into account in the power spectrum estimation.,width=302 ] we now specialize the discussion to the dark energy figure of merit , as defined e.g. by @xcite , @xmath100 where @xmath93 is the dark energy equation of state at the pivot redshift , where the dark energy is best constrained .
the errors on @xmath93 are listed in table [ table_constraints ] .
figure [ fig_fom_detf ] shows the dependence of this figure of merit as a function of the s / n threshold @xmath44 between the resolved and the unresolved parts of the data .
the color code and line styles are the same as those used for the previous figure .
thick lines show the foms when the power spectrum is estimated on the unresolved part , and thin lines show the foms when it is estimated on all the data ( we neglect the covariance between the power spectrum and cluster counts ) .
considering the power spectrum on the unresolved part only ( dashed red ) slightly improves the fom .
this holds true when combining clusters with @xmath101 .
the bispectrum gives a higher figure of merit than cluster counts , and depends very weakly on @xmath44 . only at low s
/ n do the figures of merit compare .
clusters counts , when the selection function and the contamination are controlled well enough to allow one to consider a very low s / n threshold , and the bispectrum are comparable at extracting non - gaussian aspects in a weak lensing survey and at constraining dark energy . on the other hand ,
if we use a safe threshold for clusters detection ( @xmath102 ) , the bispectrum becomes better than clusters counts to break degeneracies and to measure dark energy .
-@xmath1 plane , when all other parameters are kept constant , for a 20,000 deg@xmath8 survey , and no tomography .
filled ellipses show constraints from the power spectrum ( cyan ) , s / n peak counts ( grey ) , cluster counts as a function of mass and redshift ( green ) and bispectrum when all triangle configuration are taken into account ( yellow ) ; the constraints obtained from the bispectrum when taking into account equilateral triangles only get out of the frame .
open ellipses are constraints from combining the power spectrum with the bispectrum ( all triangle configurations - blue ) , with peak counts ( black ) and cluster counts ( red).,width=302 ] in two recent papers , @xcite and @xcite show with numerical simulations how powerful weak lensing peaks counting is at constraining @xmath0 and @xmath1 , and @xmath6 , respectively .
these appear to contradict our results , in which we saw that if indeed , counting peaks efficiently captures non - gaussianity and breaks power spectrum degeneracies , it does it only when combined with the power spectrum , but gives poor constraints when used alone ( table [ table_constraints ] ) .
a concern expressed by the aforementioned authors is that they varied only their parameters of interest ( @xmath0 and @xmath1 or @xmath6 ) , and defer marginalizing on other parameters for later works . in our analysis
, we indeed marginalized on eight parameters . to show that this considerably lowers the constraints
, we make the same kind of analysis , by varying @xmath0 and @xmath1 only , as done by @xcite . for simplicity , we do not separate the power spectrum into resolved and unresolved parts , but consider the total power spectrum only , without tomography .
since our analytical model can not fully reproduce their taking into account cosmologically significant false detections ( peaks that do not correspond to a single , localized , virialized cluster , but that are instead due to mass alignments along the line of sight ) , we use counts as a function of s / n as a proxy of their measure .
we use a detection threshold @xmath51 .
figure [ fig_dp ] shows how counting peaks ( grey filled ellipse ) compares to the power spectrum ( cyan filled ellipse ) in that case .
the black open ellipse shows the combination of the two . as shown by @xcite , peak counts indeed is an efficient cosmological tool when varying @xmath0 and @xmath1 only .
marginalizing on other parameters hampers its ability to perform , as we showed above , unless combined with the power spectrum
. however , since our model does not catch filaments and significant false detections , as @xcite and @xcite do , our conclusion about cluster counts alone may be over - pessimistic .
therefore , it will be valuable to adapt their simulations to higher dimensional parameter estimations . looking for the best statistics to discriminate models , @xcite conclude that peak counting is the best way of breaking the @xmath103 degeneracy , with other third - order statistics ( such the skewness ) also being able to break degeneracies . in that work
the bispectrum in particular was found to be weak at breaking degeneracies .
this apparent discrepancy with our findings here , where we find that combining the bispectrum with the power spectrum is comparable to combining peak counting with the power spectrum , comes from the fact that the bispectrum presented here uses all triangle configurations , while the bispectrum in @xcite used only equilateral triangles . to emphasize the importance of considering all triangle configurations when computing the bispectrum , we compute the constraints expected on @xmath0 and @xmath1 , while keeping all other parameters fixed , as in @xcite , from the power spectrum , the bispectrum with all triangles information or only equilateral triangles , and cluster counts .
we use one redshift bin .
our analytical model does not allow us to compute @xcite s wavelet peak counts ( counts of clusters in different wavelet scales ) , so we use the counts as a function of mass and redshift as a proxy .
figure [ fig_dp ] shows how constraints from these measurements compare . the cyan filled ellipse shows constraints from the power spectrum , the green one , constraints from cluster counts , and the red open ellipse their combination .
as shown by @xcite , cluster counts are very efficient at grabbing non - gaussianities and breaking the power spectrum degeneracies in that particular case .
the yellow filled ellipse shows constraints from the bispectrum , when the information from all triangle configurations is taken into account , and the blue open ellipse shows constraints from combining this bispectrum with the power spectrum . as mentioned above , the bispectrum ( with information from all triangle configurations ) provides tight constraints , and is as efficient as cluster counts at capturing non - gaussian features .
however , we find that when taking into account only equilateral triangles when computing constraints from the bispectrum , the errors on the parameters reach 100% , making such a measurement extremely low efficiency . the frame of figure [ fig_dp ] is much too small to show the associated ellipse .
consequently , we concur with @xcite when claiming that the bispectrum measured with equilateral triangles only does not allow one to break the power spectrum degeneracy between @xmath0 and @xmath1 . coming back to our 8-parameters analysis
, we notice the same behavior as that mentioned in the above paragraph .
to capture non - gaussian features and efficiently break degeneracies left over by the power spectrum , one has to take into account all triangle configurations when measuring the weak lensing bispectrum .
equilateral triangles only do not provide any cosmological constraints .
we investigated how to best capture non - gaussianities in weak lensing surveys so as to break degeneracies left over by the power spectrum in weak lensing analysis , by means of combining it with galaxy clusters counts and with the weak lensing bispectrum . to this end , we computed , in the halo model framework , the power spectrum , the bispectrum and four kinds of halo counts ( as a function of s / n , mass , redshift , and mass and redshift ) , in a euclid - like survey .
we have computed the errors on cosmological parameters brought by the combination of the power spectrum with the bispectrum , and of the power spectrum with clusters counts . we have emphasized our discussion on the varying dark energy equation of state , the mass density parameter @xmath0 and the amplitude of density fluctuations @xmath1 , marginalized on a 8-parameter cosmological model .
we did not take any prior from outside weak lensing ( e.g. cmb priors ) , so as to show how a weak lensing survey alone can allow us to capture non - gaussianities , and what constraints we can expect from it on cosmological parameters . when computing the power spectrum , we have separated the contribution of the resolved and the unresolved parts of the data , delimited by a s / n threshold @xmath44 .
doing so , we noted an improvement in errors on cosmological parameters .
although peak counts alone are powerful at constraining @xmath0 and @xmath1 when other parameters are kept constant ( as shown by @xcite ) , we have seen that marginalizing on a 8-parameter space make them provide only weak constraints , unless combined with the power spectrum .
we have indeed shown that when combined with the power spectrum , cluster counts provide percent level marginalized errors on cosmological parameters .
furthermore , provided that the threshold is high enough ( @xmath102 ) , all four kinds of cluster counts we consider give similar constraints .
hence , combining the power spectrum with s / n peak counts gives constraints nearly as tight as combining the power spectrum with clusters counted as a function of redshift and mass .
galaxy clusters provide the same non - gaussian information , independently of the way they are taken into account , as long as they are combined with the power spectrum : different cluster counts break degeneracies in a similar way .
this will allow us to put stringent constraints on cosmological parameters from weak lensing alone , by combining s / n peaks with the power spectrum , without any extra cost and requirement of follow - up for clusters redshift and mass measurement .
we have shown that the weak lensing bispectrum efficiently captures non - gaussianities , and is marginally more efficient than cluster counts at breaking degeneracies left over by the power spectrum .
furthermore , the bispectrum alone provides constraints comparable to those given by the power spectrum .
we have expanded @xcite s conclusions by adding the bispectrum to clusters counts as efficient tools to capture non - gaussianities .
we also noted that equilateral triangles alone do not bring significant information . to provide percent level constraints ,
all triangle configurations must be accounted for when measuring the weak lensing bispectrum .
all our results are based on a euclid - like survey , and may not be directly applicable to current surveys such as the canada - france - hawaii telescope legacy survey ( cfhtls ) or the cosmic evolution survey ( cosmos ) . in particular , on current surveys , the cosmic variance could play an important role in lowering the ability of the bispectrum to be more efficient than s / n peak counts . how those two measures compare to capture non - gaussian features
could then depend on the survey s characteristics .
this question is beyond the scope of this paper , and will be addressed in a subsequent work .
we have assumed that all systematics were well accounted for so as not to bias our results .
although that can seem over - optimistic , the amount of efforts currently underway to correct for various systematics gives us good reasons to think that our assumption is likely to be met when a euclid - like survey is undertaken .
moreover , allowing for systematics requires a more sophisticated analysis than the fisher matrix ( @xcite ) .
we defer it to a future paper , looking at how systematics enter in the correlations between different statistics . in light of our results
, it appears that weak lensing surveys alone will be able to reach the percent accuracy on the dark energy equation of state parameters .
moreover , it provides us with ways to cross - check the parameters measurements . for instance , after measuring the power spectrum , counting s / n peaks provides an easy and fast way to optimally capture non - gaussianities and tighten constraints .
then , the bispectrum and its combination with the power spectrum should give consistent parameters estimation and similar constraints .
however , measuring the bispectrum on half the sky will constitute a real challenge .
along with already existing robust and fast algorithms ( @xcite ) , new ones are being developed ( e.g. berg et al in prep and semboloni et al in prep ) to measure either the 3pcf in real space or the bispectrum in fourier space ( as proposed by @xcite ) . in the meantime ,
more and more precise numerical simulations will be a needed tool to test and tune those algorithms , and current surveys such as the cfhtls and cosmos are ideal benchmark to test codes in the real world .
presented as an alternative to fitting functions based on numerical simulations , the halo model for large - scale structures describes structures as spherical haloes - see @xcite for a review . in this framework ,
the large - scale structures statistics can be described by correlations within a same halo ( the profile of which affects them ) and between different halos ( the clustering of halos is thus naturally taken into account ) .
although being simplistic , the model agrees fairly well with common numerical simulations . here , we summarize its key ingredients : the mass function , the profile of halos and the halo bias .
we also give the expressions for the 3d dark matter power spectrum and bispectrum .
the comoving number of virialized halos , whose mass ranges between @xmath104 and @xmath105 and redshift ranges between @xmath106 and @xmath107 , is given by ( @xcite ) : @xmath108 where @xmath109 is the universe s current background density and @xmath110 .
@xmath111 is the non - linear overdensity of a halo collapsing at redshift @xmath106 . in a dark energy universe
, @xmath111 depends weakly on cosmology , especially on the dark energy s equation of state @xmath6 .
we use @xcite s fitting function to compute @xmath112 .
this fitting function is valid for constant @xmath6 only , therefore we compute it using an effective equation of state @xmath113 .
the r.m.s of the mass fluctuations in a sphere containing a mass @xmath104 at redshift @xmath106 is defined by @xmath114 where @xmath115 is the three - dimension linear matter power spectrum , @xmath116 is the radius of the considered sphere , and @xmath117 is the mean background density .
the window function @xmath118 $ ] .
a couple of mass functions have been described in the literature ( @xcite ) .
we use that of @xcite , @xmath119 , where @xmath120 , @xmath121 and @xmath122 .
the top - hat collapse model for structure formation describes the matter infall on the gravitational wells until virialization . at this stage ,
non - linear physics takes place , that further modifies the virialized halo profile , and that must be assessed through numerical simulation .
@xcite showed that the mass profile of such virialized halos can be described by a
universal " profile , @xmath123 where @xmath35 and @xmath34 correspond to a characteristic radius and density , respectively , and ( @xmath124)=(1,2 ) for the usual nfw profile , that we use in this paper .
the mass of a nfw halo is then given by @xmath125 , where @xmath126 is the halo s virial radius . in theory ,
a nfw is infinite , but we assume here that it is truncated at @xmath126 .
the virial and characteristic radius are linked by @xmath127 , where @xmath28 is the halo s concentration , that can be parametrized , following @xcite and @xcite , by @xmath128 in this equation , @xmath129 is the non - linear mass scale , defined as @xmath130 .
in contrast to earlier works ( @xcite ) ,
we add an extra redshift - dependence in the parameter @xmath131 .
based on @xcite and @xcite , it allows us to avoid the catastrophic drop in concentration observed on simulations for high redshift , massive halos , as parametrized by @xcite with a constant @xmath131 .
we find that @xmath132 and @xmath133 , besides agreeing with @xcite , and with @xcite at low redshift , gives us the best agreement between the halo model and the @xcite s power spectrum and bispectrum .
one should note that the concentration has a significant impact on the 1-halo term of the halo model power spectrum and bispectrum , as shown by @xcite , and thus a thorough description is needed .
the bias describes how halos cluster with respect to the matter distribution .
it is well described at first order by fitting functions of the form ( @xcite ) , @xmath134 with @xmath135 and @xmath136 .
we neglect the higher order terms of the bias .
the halo model describes the correlation functions of the density field , in real - space , as the sum of the correlation between points belonging to a same halo and of the correlation between points in different halos .
the same contributions appear in fourier space , when defining the power spectrum and bispectrum , which are thus the sum of a 1-halo term and a 2-halo term ( and 3-halo term for the bispectrum ) ( e.g. @xcite ) , @xmath137 the 1-halo component is given by @xmath138 and the 2-halo term is given by @xmath139 the function @xmath140 is the fourier transform of the dark matter density profile .
the term @xmath141 represents the power spectrum of halos of mass @xmath142 and @xmath143 and is approximated by @xmath144 , where @xmath145 is halo @xmath15 s bias and @xmath146 is the matter linear power spectrum .
we use the @xcite s transfert function to evaluate @xmath146 .
the 3d matter bispectrum can be decomposed into terms coming from one , two and three halos : @xmath147 with @xmath148 @xmath149 and @xmath150 \\
\times i_1 ^ 1(k_1)i_1 ^ 1(k_2)p_\delta(k_1)p_\delta(k_2 ) + c.p.,\ ] ] where @xmath151 denotes circular permutations .
the @xmath152 function is given by the second - order perturbation theory ( @xcite ) : @xmath153.\ ] ] the function @xmath154 is defined through the fourier transform of the halo profile @xmath155 ( eq . [ eq_u ] ) : @xmath156 with @xmath157 .
here , we specialize the discussion of section [ sect_selfct ] to an nfw halo . to evaluate equation ( [ eq_nu1 ] ) in this case , we first need to derive the convergence of such a halo .
the mass density , projected along the line - of - sight , of a nfw halo at redshift @xmath106 , with concentration @xmath28 , is given by : @xmath158 where @xmath159 .
the function @xmath160 depends only on the distance to the halo s center , and is given by ( as long as @xmath161 ) @xcite in equation ( [ eq_sigma_crit ] ) , @xmath28 is the speed of light , @xmath166 the gravitation constant , and @xmath167 , @xmath33 and @xmath168 are the angular - diameter distances of the source , of the lens , and between the lens and the source , respectively . to take the redshift distribution @xmath169 of source galaxies , we follow @xcite , @xcite and @xcite and introduce the function @xmath170 where @xmath171 is the lens redshift , and @xmath172 is the critical density for a source at infinite redshift
this function allows us to project sources with a certain redshift distribution on a single redshift @xmath163 verifying @xmath173 , with @xmath174 .
one can then link the halo s convergence with a source at @xmath163 to that with infinite redshift source , @xmath175 ( @xcite ) .
the convergence given by a source population distributed in redshift along @xmath169 is then given by @xmath176 .
the signal - to - noise ratio ( eq . [ eq_nu1 ] ) , for a nfw halo is then @xmath177 where @xmath27 is a function of the halo s concentration only , and is well fitted , for @xmath30 ( the maximum deviation being of order 0.5% ) , by @xmath178 99 albrecht , a. , et al .
, 2006 , arxiv : astro - ph/0609591 albrecht , a. , et al . , 2009 , arxiv:0901.0721 amara , a. , & rfrgier , a. 2007 , , 381 , 1018 amara , a. , & rfrgier , a. 2008 , , 391 , 228 amara , a. , refregier , a. , & paulin - henriksson , s. 2009 , arxiv:0905.3176 bacon d. , refregier a. , ellis r. , 2000 , mnras , 318 , 625 bartelmann , m. , schneider , p. , 2001
, 340 , 291 battye , r. a. , & weller , j. 2003 , , 68 , 083506 benabed , k. , & scoccimarro , r. 2006 , , 456 , 421 benjamin , j. , et al . , 2007 , , 381 , 702 berg , j. , et al .
2008 , , 385 , 695 bernardeau , f. , colombi , s. , gaztaaga , e. , & scoccimarro , r. 2002 , , 367 , 1 bernardeau , f. , van waerbeke , l. , & mellier , y. 2003 , , 397 , 405 bullock , j.s . , et al . , 2001 , , 321 ,
559 bridle , s. , et al .
, 2009 , annals of applied statistics , vol .
3 , no . 1 , 6 - 37 , arxiv:0802.1214 cooray , a. , hu , w. , 2001a , , 548 , 7 cooray , a. , hu , w. , 2001b , , 554 , 56 cooray , a. , sheth , r. , 2002 , phys . rep . , 372 , 1 dietrich , j. p. , & hartlap , j. 2009 , arxiv:0906.3512 dolag , k. , et al . , 2004 , a&a , 416 , 853 duffy , a. r. , schaye , j. , kay , s. t. , & dalla vecchia , c. 2008 , , 390 , l64 eke , v. r. , cole , s. , frenk , c. s. , & patrick henry , j. 1998 , , 298 , 1145 eisenstein , d , hu , w. , 1998 , , 498 , 497 fan , z .- h . , 2007 , , 669 , 10 fu , l. , et al . 2008 , , 479 , 9 gao , l. , navarro , j. f. , cole , s. , frenk , c. s. , white , s. d. m. , springel , v. , jenkins , a. , & neto , a. f. 2008 , , 387 , 536 gladders , m. d. , yee , h. k. c. , majumdar , s. , barrientos , l. f. , hoekstra , h. , hall , p. b. , & infante , l. 2007 , , 655 , 128 hamana , t. , takada , m. , yoshida , n. , 2004 , 350 , 893 hennawi , j.f . , spergel , d.n .
, 2005 , , 624 , 59 heymans , c. , et al . , 2005 , , 361 , 160 heymans , c. , et al . , 2006 , , 368 , 1323 hoekstra , h. , et al . , 2006 , , 647 , 116 hoekstra , h. , & jain , b. 2008 , annual review of nuclear and particle science , 58 , 99 holder , g. , haiman , z. , & mohr , j. j. 2001 , , 560 , l111 horellou , c. , berge , j. , 2005 , , 360 , 1393 hu , w. 2003 , , 67 , 081304 hu , w. , jain , b. , 2004 , phreviewd , 70 , 043009 hu , w. , tegmark , m. , 1999 , , 514 , l65 huffenberger , k. m. , & seljak , u. 2003 , , 340 , 1199 jarvis , m. , bernstein , g. , & jain , b. 2004 , , 352 , 338 jarvis , m. , jain , b. , bernstein , g. , & dolney , d. 2006 , , 644 , 71 jenkins , a. , et al .
2001 , , 321 , 372 joachimi , b. , shi , x. , & schneider , p. 2009
, , 508 , 1193 joudaki , s. , cooray , a. , & holz , d. e. 2009 , , 80 , 023003 kratochvil , j. m. , haiman , z. , & may , m. 2009 , arxiv:0907.0486 lima , m. , hu , w. , 2004 , phreviewd , 70 , 043504 ma , c .- p . ,
fry , j.n . , 2000 , , 543 , 503 majumdar , s. , & mohr , j. j. 2004 , , 613 , 41 mantz , a. , allen , s. w. , ebeling , h. , & rapetti , d. 2008 , , 387 , 1179 marian , l. & bernstein , g .
, 2006 , phreviewd , 73 , 123525 marian , l. , smith , r. e. , & bernstein , g. m. 2009 , , 698 , l33 massey , r. , refregier , a. , bacon , d. j. , ellis , r. , & brown , m. l. , 2005 , , 359 , 1277 massey , r. , et al . , 2007a , , 376 , 13 massey , r. , et al . , 2007b , , 172 , 239 maturi , m. , angrick , c. , bartelmann , m. , & pace , f. 2009 , arxiv:0907.1849 maturi , m. , meneghetti , m. , bartelmann , m. , dolag , k. , moscardini , l. , 2005 , a&a , 442 , 851 mellier , y. , 1999 , araa , 37 , 127 mo , h. j. , & white , s. d. m. 1996 , , 282 , 347 munshi , d. , valageas , p. , van waerbeke , l. , heavens , a. , 2008 , phys . rep . , 462 , 67 navarro , j.f . ,
frenk , c.s . ,
white , s.d.m .
, 1996 , , 462 , 563 oukbir , j. , & blanchard , a. 1992 , , 262 , l21 pace , f. , maturi , m. , meneghetti , m. , bartelmann , m. , moscardini , l. , & dolag , k. 2007 , , 471 , 731 paulin - henriksson , s. , amara , a. , voigt , l. , refregier , a. , & bridle , s. l. 2008 , , 484 , 67 peacock , j.a . , dodds , s.j . , 1996 , , 280 , l19 peacock , j. a. , schneider , p. , efstathiou , g. , ellis , j. r. , leibundgut , b. , lilly , s. j. , & mellier , y. 2006 , esa - eso working group on fundamental cosmology , edited by j.a .
peacock et al .
esa , 2006 .
, pierpaoli , e. , borgani , s. , scott , d. , & white , m. 2003 , , 342 , 163 pires , s. , starck , j .- l . , amara , a. , rfrgier , a. , & teyssier , r. 2009b , , 505 , 969 pires , s. , starck , j .- l . ,
amara , a. , teyssier , r. , rfrgier , a. , & fadili , j. 2009a , , 395 , 1265 press , w.h . ,
schechter , p. , 1974
, , 187 , 425 refregier , a. , valtchanov , i. , & pierre , m. 2002 , , 390 , 1 refregier a. , 2003 , araa , 41 , 645 refregier , a. , amara , a. , kitching , t. , & rassat , a. 2008 , arxiv:0810.1285 rhodes , j. , refregier , a. , collins , n. r. , gardner , j. p. , groth , e. j. , & hill , r. s. 2004 , , 605 , 29 sahln , m. , et al .
2009 , , 397 , 577 schneider , p. , kilbinger , m. , & lombardi , m. 2005 , , 431 , 9 schneider , p. , & lombardi , m. 2003 , , 397 , 809 schrabback , t. , et al .
2007 , , 468 , 823 seitz , c. , schneider , p. , 1997 ,
a&a , 318 , 687 semboloni , e. , et al .
2006 , , 452 , 51 semboloni , e. , heymans , c. , van waerbeke , l. , & schneider , p. 2008
, , 388 , 991 shaw , l. d. , zahn , o. , holder , g. p. , & dor , o. 2009 , , 702 , 368 sheth , r. k. , mo , h. j. , & tormen , g. 2001 , , 323 , 1 sheth , r. k. , & tormen , g. 1999 , , 308 , 119 smail , i. , ellis , r. s. , & fitchett , m. j. 1994 , , 270 , 245 smith , r.e . , et al . , 2003 , , 341 , 1311 takada , m. , bridle , s. , 2007 , njph , 9 , 446 takada , m. , & jain , b. 2003a , , 583 , l49 takada , m. , jain , b. , 2003b , , 340 , 580 takada , m. , & jain , b. , 2003c , , 344 , 857 takada , m. , jain , b. , 2004 , , 348 , 897 tinker , j. , kravtsov , a. v. , klypin , a. , abazajian , k. , warren , m. , yepes , g. , gottlber , s. , & holz , d. e. 2008 , , 688 , 709 vafaei , s. , lu , t. , van waerbeke , l. , semboloni , e. , heymans , c. , & pen , u. l. 2010 , astroparticle physics , 32 , 340 vallisneri , m. , 2008 , phreviewd , 77 , 2001 van waerbeke , l. et al . , 2000 , a&a , 358 , 30 van waerbeke , l. , mellier , y. , & hoekstra , h. 2005 , , 429 , 75 wang , l. , & steinhardt , p. j. 1998 , , 508 , 483 wang , s. , khoury , j. , haiman , z. , & may , m. 2004 , , 70 , 123008 wang , s. , haiman , z. , & may , m. 2009 , , 691 , 547 wittman , d.m . , tyson , j.a . , kirkman , d. , dellantonio , i. , bernstein , g. , 2000 , nature , 405 , 143 weinberg , n.n . , kamionkowski , m. , 2002 , , 337 , 1269 weinberg , n.n . ,
kamionkowski , m. , 2003 , , 341 , 251 wright , c.o . , brainerd , t.g .
, 2000 , , 534 , 34 zaldarriaga , m. , & scoccimarro , r. 2003 , , 584 , 559 zhang , l. l. , & pen , u .- l .
2005 , new astronomy , 10 , 569 zhao , d. h. , jing , y. p. , mo , h. j. , & boerner , g. 2003 , , 597 , l9 zhao , d. h. , jing , y. p. , mo , h. j. , & boerner , g. 2009 , , 707 , 354 | we compare the efficiency of weak lensing - selected galaxy clusters counts and of the weak lensing bispectrum at capturing non - gaussian features in the dark matter distribution .
we use the halo model to compute the weak lensing power spectrum , the bispectrum and the expected number of detected clusters , and derive constraints on cosmological parameters for a large , low systematic weak lensing survey , by focusing on the @xmath0-@xmath1 plane and on the dark energy equation of state .
we separate the power spectrum into the resolved and the unresolved parts of the data , the resolved part being defined as detected clusters , and the unresolved part as the rest of the field .
we consider four kinds of clusters counts , taking into account different amount of information : signal - to - noise ratio peak counts ; counts as a function of clusters mass ; counts as a function of clusters redshift ; and counts as a function of clusters mass and redshift .
we show that when combined with the power spectrum , those four kinds of counts provide similar constraints , thus allowing one to perform the most direct counts , signal - to - noise peaks counts , and get percent level constraints on cosmological parameters . we show that the weak lensing bispectrum gives
constraints comparable to those given by the power spectrum and captures non - gaussian features as well as clusters counts , its combination with the power spectrum giving errors on cosmological parameters that are similar to , if not marginally smaller than , those obtained when combining the power spectrum with cluster counts .
we finally note that in order to reach its potential , the weak lensing bispectrum must be computed using all triangle configurations , as equilateral triangles alone do not provide useful information .
the appendices summarize the halo model , and the way the power spectrum and bispectrum are computed in this framework . |
we report here on a study of new physics ( np ) predictions for @xmath4 .
the standard model ( sm ) prediction for @xmath5 is currently smaller than the experimental branching fraction limit @xcite of @xmath6 by about a factor of @xmath7 .
this presents a window of opportunity for observing new physics ( np ) effects in this mode .
this topic is particularly timely in view of experimental indications of np effects in both the exclusive decay @xmath8 @xcite ( for recent cdf results , also see ref .
@xcite ) as well as the inclusive like - sign dimuon asymmetry observed in @xmath9 @xcite .
moreover , future work at lhc - b , @xmath10 super b - factories and ongoing cdf @xmath11 d0 measurements at fermilab ( see the discussion following eq .
( [ bmm ] ) ) is expected to markedly improve the current branching fraction bound .
our strategy in this paper is somewhat reminiscent of our recent study @xcite noting that in some np models the @xmath12 mixing and @xmath13 decay amplitudes have a common dependence on the np parameters .
if so , one can predict the @xmath13 branching fraction in terms of the observed @xmath14 provided that much or all of the mixing is attributed to np .
this is a viable possibility for @xmath12 mixing because the standard model ( sm ) signal has large theoretical uncertainties and because many np models can produce the observed mixing @xcite .
for @xmath0 the situation is very different . here , the sm prediction is in accord with the observed value ( _ e.g. _ see refs . and papers cited therein ) .
in fact , the analysis described below ( _ cf . _
see eqs .
( [ av - np]),([av - onesig ] ) ) gives @xmath15 , which demonstrates just how well the sm prediction agrees with the experimental value of @xmath0 . in view of this
, our sm expression for @xmath0 will be given at nlo whereas lo results will suffice for np models . as regards the corresponding width difference @xmath16 , the experimental and theoretical uncertainties are still rather significant ( _ viz _ sect .
ii - c ) . in those np models where mixing and @xmath5 arise from a common set of parameters , the severe constraint on any np signal to @xmath2 mixing places strong bounds on its contribution to @xmath17 .
in fact , we shall find the constraint can be so strong that for some np models the predicted @xmath5 branching fraction lies well below the sm prediction .
the first step in our study ( _ cf _ section
ii ) will be to revisit the sm predictions for mixing in the @xmath18-quark system by using up - to - date inputs .
we carry this out for the two mixing quantities @xmath0 and @xmath1 .
the former in turn yields phenomenological bounds on np mixing contributions which in certain models can be used to bound the magnitude of the @xmath4 decay mode .
we also update the sm branching fraction for @xmath4 by using the observed @xmath2 mixing as input .
then , in section iii we discuss general properties of np models with tree - level amplitudes . in section iv , we explore various np models such as extra @xmath19 bosons , family symmetry , r - parity violating supersymmetry , flavor - changing higgs models , and models with the fourth sequential generation .
our concluding remarks appear in section v , and some technical details are relegated to the appendix .
l|l @xmath20 mev @xcite & @xmath21 s @xcite + @xmath22 & @xmath23 @xcite + @xmath24 @xcite & @xmath25 @xcite + @xmath26 @xcite & @xmath27 @xcite + @xmath28 gev @xcite & @xmath29 mev @xcite + @xmath30 @xcite & @xmath31 @xcite + .05 in
we begin by considering the sm predictions for @xmath2 mixing .
this step is crucial to obtaining bounds on np contributions .
we also use the @xmath2 mixing signal as input to a determination of the branching fraction for @xmath5 .
the work in this section takes advantage of recent progress made in determining several quantities used in the analysis .
we summarize our numerical inputs in table i , along with corresponding references . included in table
i is an updated determination of the top quark pole mass @xcite @xmath32 which in turn is used to determine the corresponding running mass @xmath33 @xcite along with several decay constants and b - factors as evaluated in lattice qcd . for definiteness , we have used values appearing in ref . @xcite .
this area is , however , constantly evolving and one anticipates further developments in the near future @xcite .
our values for the cabibbo - kobayashi - maskawa ( ckm ) matrix elements @xmath34 and @xmath35 are taken from ref .
similar values occur for the global fits cited elsewhere ( _ e.g. _ refs . ) .
the pdg value for @xmath0 , m_b_s^(expt ) = ( 117.0 0.8 ) 10 ^ -13 gev , [ mexpt ] is a very accurate one the uncertainty amounts to about @xmath36% . the nlo sm formula , m_b_s^(sm ) = 2 g_f^2 m_w^2 m_b_s f^2_b_sb_b_s 12 ^2 |v_ts^ * v_tb|^2 _ b_s
s_0(|x_t ) , [ delmsm ] is arrived at from an operator product expansion of the mixing hamiltonian .
the short - distance dependence in the wilson coefficient appears in the scale - insensitive combination @xmath37 , where the factor @xmath38 is an inami - lin function @xcite ( with @xmath39 ) and @xmath40 is the running top - quark mass parameter in @xmath41 renormalization . in particular , we have @xmath42 which leads to @xmath43 . using the same matching scale , we obtain @xmath44 for the nlo qcd factor .
our evaluation for @xmath45 then gives & & m_b_s^(sm ) = ( 125.2^+13.8_-12.7 ) 10 ^ -13 gev , [ me ] which is in accord with the experimental value of eq .
( [ mexpt ] ) . the theoretical uncertainty in the sm prediction of eq .
( [ me ] ) is roughly a factor of sixteen larger than the experimental uncertainty of eq .
( [ delmsm ] ) .
the largest source of error occurs in the nonperturbative factor @xmath46 , followed by that in the ckm matrix element @xmath47 .
the asymmetry in the upper and lower uncertainties in @xmath45 arises from the corresponding asymmetry in the value of @xmath47 cited in ref .
@xcite . finally , we note in passing that for the ratio @xmath48 the experimental value is @xmath49 whereas the sm determination gives @xmath50 .
this good agreement is not surprising since the ratio @xmath48 contains less theoretical uncertainty than @xmath51 or @xmath0 separately .
the above discussion of @xmath45 sets the stage for analyzing np contributions to @xmath5 .
there is , in principle , a second approach which instead utilizes @xmath53 .
the pdg value for the @xmath2 width difference is @xmath54 . together with eq .
( [ mexpt ] ) , this gives implies @xmath55 , consistent with the value in eq .
( [ bsratio ] ) . ]
r^(expt ) = 0.062^+0.034_-0.037 10 ^ 12 s^-1 ( 17.77 0.12 ) 10 ^ 12 s^-1 = ( 34.9 20.0 ) 10 ^ -4 .
[ bsratio ] whereas the corresponding sm prediction from ref .
@xcite is @xmath56 .
in contrast to the mass splitting @xmath0 , the theoretical uncertainty in the ratio @xmath1 is much smaller than in the current experimental determination .
nonetheless , this situation is expected to change once lhcb gathers sufficient data . as such , we would expect a highly accurate value of @xmath57 to eventually become available .
we propose that it could be applied to the kind of analysis used in this paper as follows .
we define a kind of mass difference @xmath58 as m_b_s _ b_s^(expt ) .
[ ratio ] the point is that if np contributions are neglected in @xmath59 transitions , then @xmath60 is purely a sm effect .
in addition , the ratio @xmath61 will be less dependent on hadronic parameters than either factor separately .
this quantity is also important in the scenarios where np contributes a significant cp - violating phase to @xmath0 . in this situation ,
@xmath62 will be reduced compared to its sm value @xmath63 by a factor of @xmath64 , where @xmath65 is related to the relative phase between the sm and np contributions to @xmath0 @xcite . at the very least , the relation in eq .
( [ ratio ] ) would be of interest to analyze the np issue using both quantities @xmath0 and the above @xmath66 . pdg entries for @xmath67 are _
b_s ^+^-^(expt ) < 4.7 10 ^ -8 _ b_s e^+e^-^(expt ) < 2.8 10 ^ -7 , [ bmm ] with no experimental limit currently for the @xmath68 transition .
data collected by the d0 and cdf collaborations will improve the above brancing fraction limit .
for example , the d0 collaboration reports @xmath69 , with an anticipated limit of eleven times the sm prediction and similarly for the cdf collaboration @xcite .
since the long distance ( ld ) estimate for the branching fraction of @xmath5 in the sm gives @xmath70 @xcite , we consider only the short distance ( sd ) component in the following .
using eq .
( [ delmsm ] ) as input to the sd - dominated @xmath5 transition ( see also ref .
@xcite ) we arrive at _ b_s
^+^-^(sm ) = m_b_s _ b_s 3 g_f^2 m_w^2 m_^2 4 _
b_s b_b_s ^3 ^1/2 _
y^2 y^2(|x_t )
s_0(|x_t ) , where @xmath71 is another inami - lin function @xcite . expressing @xmath72 in this manner serves to remove some of the inherent model dependence .
numerical evaluation gives _
b_s ^+^-^(sm ) = ( 3.33 0.21 ) 10 ^ -9 . [ bmmsm ] the major sources of uncertainty , ordered by magnitude , arise from the factors @xmath73 , then the lifetime @xmath74 and finally the top - quark mass value .
in this section , we first obtain a numerical ( 1@xmath75 ) bound on any possible new physics contribution to @xmath0 .
we then use this to constrain couplings in a variety of np models and thereby learn something about the @xmath5 transition . as shown in ref .
@xcite , new physics in @xmath59 interactions can _ in principle _ markedly affect @xmath76 .
the logic is similar to that used in ref .
@xcite regarding the possible impact of np on @xmath77 .
since , however , in @xmath2 mixing such models are not easy to come up with , one can simply assume that @xmath59 processes are dominated by the sm interactions .
thus we can write [ deltabnp0 ] & & m_b_s = m_b_s^(sm ) + m_b_s^(np ) , if the @xmath59 sector were to contain significant np contributions , then the above relation would no longer be valid due to interference between the sm and np components . as can be seen from eq .
( [ deltabnp0 ] ) , interference between the sm and np components may also occur in the presence of a cp - violating phase @xmath78 in the np part of the mixing amplitude @xcite .
this large np phase could markedly affect @xmath57 even in the absence of a np contribution to the on - shell @xmath59 transitions ( recall that @xmath57 depends explicitly on the cosine of the cp - violating phase @xmath65 @xcite ; the explicit relation between @xmath78 and @xmath65 can be found in @xcite ) .
it is therefore more reasonable to use @xmath57 in studying those scenarios with a large np phase .
the appropriate strategy here would be to use @xmath57 and @xmath63 to extract the phase @xmath65 , eliminate @xmath79 from eq .
( [ deltabnp0 ] ) , and then extract @xmath80 in order to relate it to the rare leptonic decay rate . to do so , however , will require a significant reduction in the experimental uncertainty of @xmath57 .
alternatively , cp - violating phases could be extracted at lhcb from the studies of @xmath81 transition @xcite .
we shall defer those studies to a future publication @xcite .
in this paper we shall assume that the phase in the np component of @xmath0 is sufficiently small ( although not necessarily negligible ) , [ deltabnp1 ] & & m_b_s = m_b_s^(sm ) + m_b_s^(np ) .
accounting for np as an additive contribution , m_b_s^(expt ) = m_b_s^(sm ) + m_b_s^(np ) , we have from eqs .
( [ mexpt]),([me ] ) , & & m_b_s^(np ) = ( - 8.2_-12.7^+13.8 ) 10 ^ -13 gev .
[ av - np ] the error in @xmath82 has been included , but it is so small compared to the theoretical error in @xmath83 as to be negligible .
the @xmath84 range for the np contribution is thus & & m_b_s^(np ) = ( - 20.9 + 5.6)10 ^ -13 gev .
[ av - onesig ] to proceed further without ambiguity , we would need to know the relative phase between the sm and np components . lacking this
, we employ the absolute value of the largest possible number , & & |m_b_s^(np)| 20.9 10 ^ -13 gev , [ bnd ] to constrain the np parameters . new physics can affect both @xmath2 mixing and rare decays like @xmath5 by engaging in these two transitions at tree level . in this section we will , for generality , consider a generic spin-1 boson v or a spin-0 boson s with flavor - changing and flavor - conserving neutral current interactions that couple both to quarks and leptons .
the bosons v and s can be of either parity .
this situation is frequently realized , as in the interactions of a heavy @xmath85 boson or in multi - higgs doublet models without natural flavor conservation . _
spin-1 boson v _ : assuming that the spin-1 particle @xmath86 has flavor - changing couplings , the most general lagrangian can be written as will be @xmath87 .
] [ vecham ] h_v = g_v1^_l^__l v^+ g_v2^_r^__r v^ + g_v1 b_l _ s_l v^+ g_v2 b_r _
s_r v^+ . here
@xmath88 is the vector field and the flavor of the lepton @xmath89 might or might not coincide with @xmath90 .
it is not important whether the field @xmath88 corresponds to an abelian or non - abelian gauge symmetry group . using methods similar to those in ref .
@xcite , we obtain [ dmv2 ] m_b_s^(v ) & = & f_b_s^2 m_b_s 3 m_v^2 re , where the superscript on @xmath91 denotes propagation of a vector boson in the tree amplitude .
the wilson coefficients evaluated at a scale @xmath92 are related to the couplings @xmath93 and @xmath94 as l c_1 ( ) = r ( , m_v ) g_v1 ^ 2 , + c_2()= 2 r(,m_v)^1/2 g_v1 g_v2 , l c_3()= g_v1 g_v2 , + c_6()= r ( , m_v ) g_v2 ^ 2 , [ zwilsons ] where ( presuming that @xmath95 and @xmath96 ) , r(,m)&= & ( ) ^2/7 ( ) ^6/23 .
[ wilson ] similar calculations can be performed for the @xmath97 decay .
the effective hamiltonian in this case is _
bq^+^-^(v ) = , where the operators @xmath98 can be read off from those in ref .
@xcite with the label changes @xmath99 and @xmath100 .
this leads to the branching fraction , [ gammav ] b_b_s^0 ^+^-^(v ) = |g_v1 - g_v2|^2 |g_v1 - g_v2|^2 .
clearly , eqs .
( [ dmv2]),([gammav ] ) can be related to each other only for a specific set of np models . _
spin-0 boson s _ : analogous procedures can be followed if now the fcnc is generated by quarks interacting with spin-0 particles .
again , the most general hamiltonian can be written as [ sfcnc ] h_s = g_s 1^_l _ r s + g_s 2^_r _
l s + g_s 1 b_l s_r s + g_s 2 b_r s_l s+ .
evaluation of @xmath101 at scale @xmath102 gives m_b_s^(s ) = re + [ dms2 ] with the wilson coefficients defined as [ scalarcoeff ] & & c_3 ( ) = - 2 r(,m_s ) ^-4 g_s 1 g_s 2 c_3 ( ) g_s 1 g_s 2 + & & c_4 ( ) = - g_s 2 ^ 2 c_4 ( ) g_s 2 ^ 2 + & & c_5 ( ) = g_s 2 ^ 2 c_5 ( ) g_s 2 ^ 2 [ cs1 ] + & & c_7()= - g_s 1 ^ 2 c_7 ( ) g_s 1 ^ 2 + & & c_8 ( ) = g_s 1 ^ 2 c_8 ( ) g_s 1 ^ 2 , where for notational simplicity we have defined @xmath103 .
note that eq .
( [ dms2 ] ) is true only for the real spin-0 field @xmath104 .
if @xmath104 is a complex field , then only operator @xmath105 will contribute to eq .
( [ dms2 ] ) . the effective hamiltonian for the @xmath106 decay via a heavy scalar s with fcnc interactions is then _
bs^+^-^(s ) = - , and from this , it follows that the branching fraction is [ gammas ] b_b_s^0 ^+^-^(s ) & = & |g_s 1 - g_s 2|^2 + & & . note that if the spin-0 particle @xmath104 only has _
fcnc couplings , _
i.e. _ @xmath107 , no contribution to @xmath97 branching ratio is generated at tree level ; the non - zero contribution to rare decays is instead produced at one - loop level .
this follows from the _ pseudoscalar _ nature of the @xmath2-meson .
let us now consider specific models where the correlations between the @xmath108 mixing rates and ( in particular ) the @xmath5 rare decay can be found .
_ @xmath2 mixing : _
the @xmath2 mixing arising from the @xmath19 pole diagram has the same form as in @xmath12 mixing @xcite , m_b_s^(z ) = m_b_s f_b_s^2 b_b_s r_1(m_b , m_z ) 3 , where @xmath109 is a qcd factor which we take to be r_1(m_b , m_z ) 0.79 .
this is a compromise between @xmath110 and @xmath111 . solving for the @xmath19 parameters
, we have = 3 |m_b_s^(np)| m_b_s f_b_s^2 b_b_s r_1(m_b , m_z ) 3.02 10 ^ -11 gev^-2 upon using the constraint from @xmath2 mixing . _
@xmath5 decay : _ this has already been calculated for @xmath112 decay in ref .
@xcite . inserting obvious modifications for @xmath113
, we have from the branching fraction relation eq . ( 39 ) of ref . @xcite , [ gammazpr ] b_b_s ^+^-^(z ) = g_zs|b^2 m_z^2 . upon inserting numbers ,
we obtain _
b_s ^+^-^(z ) 0.30 10 ^ -9 ( 1 tev m_z)^2 .
this value is already below the corresponding sm prediction ( @xmath114 ) even if we take a @xmath19 mass as light as @xmath115 tev .
one of the models of new physics that has a rich flavor phenomenology is r - parity violating ( rpv ) susy .
the crucial difference between studies of rpv susy contributions to phenomenology of the up - quark ( see @xcite ) and down - type quark sectors is the possibility of tree - level diagrams contributing to @xmath2-mixing and @xmath116 decays @xcite .
if one allows for r - parity violation , the following terms should be added to the superpotential , [ rpvsup ] w _ r= _ ijk l_i l_j e^c_k + _ ijk^l_i q_j d^c_k + _ ijk^ u_i^c d^c_j d^c_k . here
@xmath117 and @xmath118 denote @xmath119 doublet quark and lepton superfields , and @xmath120 , @xmath121 and @xmath122 stand for the @xmath119 singlet up - quark , down - quark and charged lepton superfields .
also , @xmath123 are generation indices .
we shall require baryon number symmetry by setting @xmath124 to zero .
also , we will assume cp - conservation , so all couplings @xmath125 and @xmath126 are treated as real . _ @xmath127 mixing _ : neglecting the baryon - number violating contribution , the lagrangian describing rpv susy contribution to @xmath127 mixing can be written as @xmath128 where @xmath129 is a generational index for the sneutrino . matching to eq .
( [ sfcnc ] ) implies that the only non - zero contribution comes from the operator @xmath105 .
taking into account renormalization group running , we obtain for @xmath130 from the r - parity violating terms , & & m_b_s^(r ) = f_b_s^2 m_b_s f(c_3,b_3 ) _ i , where @xmath131 denotes the mass of the sneutrino of @xmath132th generation and the function f(c_3,b_3 ) = c_3 ( , m__i ) b_3 , is defined in terms of reduced wilson coefficient of eq .
( [ scalarcoeff ] ) and the b - factor is defined in table [ tab : bagconstants ] of the appendix . _
@xmath5 decay _ : in rpv - susy , the underlying transition for @xmath5 is @xmath133 via tree - level @xmath134-squark or sneutrino exchange . in order to relate the rare decay to the mass difference contribution from rpv susy @xmath135 , we need to assume that the up - squark contribution is negligible .
this can be achieved in models where sneutrinos are much lighter than the up - type squarks , which are phenomenologically viable . employing this assumption leads to the predicted branching fraction [ y2 ] b^(r)_b_s^+^- & = & ( ) ^2 ( 1- ) + & & ( |_i|^2 + |_i|^2 ) . in order to relate @xmath5 to @xmath130 in the framework of rpv susy
, we need to make additional assumptions .
in particular , we shall assume that the sum is dominated by a single sneutrino state , which we shall denote by @xmath136 .
in addition , we will assume that @xmath137 , which will reduce the number of unknown parameters .
this assumption is not needed , however , if one wishes to set a bound on a combination of coupling constants directly from the experimental bound on @xmath138 .
then , neglecting cp - violation , [ y2a ] b^(r)_b_s^+^- = k ( ) ^2 ( ) ^2 ( 1- ) , where @xmath139 if an assumption that @xmath137 is made , and @xmath140 otherwise .
since no @xmath5 signal has yet been seen , we can use the experimental bound to obtain an updated constraint on the rpv couplings , [ y4 ] _
k22 ^_k32 5.5 10 ^ -6 ( ) ^2 . now , assuming @xmath137 , one can relate the branching ratio @xmath138 to @xmath141 , [ relate3 ] b^(r)_b_s ^+^- = ( ) ^2 ( 1- ) x_b_s^(r ) .
it is possible to plot the dependence of @xmath138 on @xmath142 for different values of @xmath131 , which we present in fig .
[ rpvgraph ] .
[ tb ] as a function of rpv leptonic coupling @xmath142 and sneutrino mass @xmath143 gev , @xmath144 gev , and @xmath145 gev ( solid , dashed , and dash - dotted lines ) .
the yellow shaded area represents excluded parameter space.,width=340 ] the gauge sector in the standard model has a large global symmetry which is broken by the higgs interaction @xcite . by enlarging the higgs sector ,
some subgroup of this symmetry can be imposed on the full sm lagrangian and the symmetry can be broken spontaneously .
this family symmetry can be global @xcite as well as gauged @xcite .
if the new gauge couplings are very weak or the gauge boson masses are large , the difference between a gauged or global symmetry is rather difficult to distinguish in practice @xcite . in general
there would be fcnc effects from both the gauge and scalar sectors .
here we study the gauge contribution .
consider the family gauge symmetry group @xmath146 acting on the three left - handed families .
spontaneous symmetry breaking renders all the gauge bosons massive .
if the su(3 ) is broken first to su(2 ) before being completely broken , we may have an effective ` low ' energy symmetry @xmath147 .
this means that the gauge bosons @xmath148 are much lighter than the @xmath149 . for simplicity
we assume that after symmetry breaking the gauge boson mass matrix is diagonal to a good approximation .
if so , the light gauge bosons @xmath150 are mass eigenstates with negligible mixing .
the lh doublets @xmath151 transform as @xmath152 under @xmath147 , as do the lepton doublets @xmath153 and the right - handed fermions are singlets under @xmath147 . in the above
, the superscript @xmath154 refers to the fact that these are weak eigenstates and not mass eigenstates .
the couplings of fermions to the light family gauge bosons @xmath150 is given by @xmath155 \
\ , \ ] ] where @xmath156 denotes the coupling strength and @xmath157 are the generators of @xmath147 the fermion mass eigenstates are given by , first for quarks , @xmath158 and then for leptons , @xmath159 the four matrices @xmath160 and @xmath161 are unknown , except for @xmath162 where @xmath163 is the maki - nakagawa - sakata - pontcorvo lepton mixing matrix .
the couplings of the gauge bosons relevant for the @xmath2 system in the mass basis are : @xmath164\end{aligned}\ ] ] the contribution to @xmath165 mixing is given by @xmath166 \label{famsymm}\ ] ] where @xmath167 \\ b = -{\rm re}\left[\left ( u_{b1 } \ u^*_{s2 } - \ u_{b2 } \ u^*_{s1 } \right ) ^2\right ] \label{abc } \\ c = { \rm re}\left[\left ( u_{b1 } u^*_{s1 } - u_{b2}u^*_{s2}\right)^2\right ] \end{array}\end{aligned}\ ] ] in a simple scheme of symmetry breaking @xcite , one obtains @xmath168 and the square bracket in eq .
( [ famsymm ] ) becomes @xmath169 \ \ .\ ] ] although the matrices @xmath170 in principle are unknown , it has been argued that a reasonable ansatz @xcite , which is incorporated in many models is @xmath171 . in this case
one can simplify @xmath172 and @xmath173 further : @xmath174 thus the @xmath2 mixing becomes m_b_s^(fs ) f^2 m_1 ^ 2 1.610 ^ -3 , [ fammix ] so that , substituting experimental bound @xmath175 , .
[ fm1 ] the same above ansatz also implies that @xmath176 and @xmath177 .
then the coupling of the gauge bosons to muon pairs is given by @xmath178 \bar{\mu}_l \gamma_\lambda \mu_l \ \ . \label{gmm}\end{aligned}\ ] ] the branching ratio for @xmath179 is given by @xmath180 next we employ the approximation ( well - supported empirically ) that @xmath181 , where @xmath182 is the tri - bi - maximal matrix @xcite .
then eq .
( [ gmm ] ) becomes @xmath183 \bar{\mu}_l \gamma_\mu \mu_l \
\ .\end{aligned}\ ] ] with this , the contribution to the branching ratio for @xmath184 becomes @xmath185 ^ 2 { 1 \over m_1 ^ 4 } \nonumber \\ & \simeq & \frac{m_{b_s } f^2_{b_s } m_\mu^2 f^4}{32 \pi \gamma_{b_{s}}}~ { 1.4 \times 10^{-4 } \over m_1 ^ 4 } \ \ .
\label{brmumu}\end{aligned}\ ] ] the dependence on unknown factors in eq .
( [ brmumu ] ) ( _ i.e. _ @xmath186 ) can be entirely removed by using the bound in eq .
( [ fm1 ] ) to yield & & b_b_s ^+ ^-^(fs ) |m_b_s^(np)|^2 .
[ fsbr ] from the bounds of eqs .
( [ av - np]),([av - onesig ] ) , we obtain & & b_b_s ^+ ^-^(fs ) 0.92 10 ^ -12 .
many extensions of the standard model contain multiple scalar doublets , which increases the possibility of fcnc mediated by flavor non - diagonal interactions of neutral components . while many ideas exist on how to suppress those interactions ( see , _
@xcite ) , the ultimate test of those ideas would involve direct observation of scalar - mediated fcnc . consider a generic yukawa interaction consisting of a set of @xmath187 higgs
doublets @xmath188 @xmath189 with sm fermions , _ y = _
ijn^u q_li u_rj h_n + _
ijn^d q_li d_rj h_n + _
ijn^e l_li e_rj h_n + , [ int1 ] where @xmath190 and @xmath191 ( @xmath192 ) are respectively the left - handed weak doublets of an @xmath132th - generation of quarks ( leptons ) . restricting the discussion to @xmath193 mixing and @xmath194 decay , we find that eq .
( [ int1 ] ) reduces to [ fcnchiggs ] h^h_y = _
23n^d s_l b_r ^0_n + _ 32n^d b_l s_r ^0_n + _ 22n^e _ l _ r ^0_n + , where @xmath195 .
bringing this to the form of eq .
( [ sfcnc ] ) and confining the discussion only to the contribution of the lightest @xmath196 and @xmath197 states , we obtain [ fcnchiggs2 ] h^h_y & = & b_r s_l ^0 + b_l s_r ^0 + _ l _ r ^0 + & - & i b_r s_l a^0 + i b_l
s_r a^0 + i _ l _ r a^0 + ... + , where ellipses stand for the terms containing heavier @xmath196 and @xmath197 states whose contributions to @xmath0 and @xmath17 will be suppressed .
if the matrix of coupling constants in eq .
( [ fcnchiggs2 ] ) is hermitian , _ e.g. _
@xmath198 , then we can identify the couplings of eq .
( [ sfcnc ] ) as [ scouplings ] & & g_s_1 = g_s_2 = , g_s_1^= g_s_2^= for _ scalar _ interactions and [ pcouplings ] & & g_s_1 = - g_s_2 = , g_s_1^= - g_s_2^= for _ pseudoscalar _ interactions . to proceed
, we need to separate two cases : ( i ) the lightest fcnc higgs particle is a scalar , and ( ii ) the lightest fcnc higgs particle is pseudoscalar . the case of relatively light scalar higgs state is quite common , arising most often in type - iii two - higgs doublet models ( models without natural flavor conservation ) @xcite . _
@xmath199-@xmath200 mixing _ : given the general formulas of eq .
( [ dms2 ] ) , it is easy to compute the contribution to @xmath201 of an intermediate scalar ( @xmath78 ) with fcnc couplings , m_b_s^ ( ) & = & ( _ 32^d m _ ) ^2 , [ dmh2 ] + f_(c_i , m_b ) & & c_3(m_b ) b_3 - ( c_4(m_b ) b_4 + c_7(m_b ) b_7 ) + ( c_5(m_b ) b_5 + c_8(m_b ) b_8 ) , with reduced wilson coefficients @xmath202 given in eq .
( [ cs1 ] ) .
_ @xmath203 decay _ : comparing eq .
( [ scouplings ] ) to eq .
( [ gammas ] ) , we can easily see that the branching fraction for the rare decay @xmath204 is zero for the intermediate scalar higgs , _
b_s^0^+^-^ ( ) = 0 .
this is consistent with what was already discussed in sec .
[ sec : treelevel ] and implies that the fcnc higgs model does not produce a contribution to @xmath205 at tree level . the non - zero contribution to @xmath205 decay
is produced at one - loop level @xcite .
the case of a lightest _ pseudoscalar _ higgs state can occur in the non - minimal supersymmetric standard model ( nmssm ) @xcite or related models @xcite . in nmssm
, a complex singlet higgs is introduced to dynamically solve the @xmath92 problem .
the resulting pseudoscalar can be as light as tens of gev .
this does not mean , however , that it necessarily gives the dominant contribution to both @xmath206 mixing and the @xmath203 decay rate since there can be loop contributions from other higgs states . in the following , we shall work in the region of the parameter space where it does . _ @xmath199-@xmath200 mixing _ : the contribution to @xmath207 due to intermediate pseudoscalar with flavor - changing couplings can be computed using the general formula in eq .
( [ dms2 ] ) along with the identification given in eq .
( [ pcouplings ] ) , [ dma ] m_b_s^(a ) & = & ( _ 32^d m_a ) ^2 , + f_a(c_i , m_b ) & = & with ` reduced ' wilson coefficients @xmath208 again being defined in eq .
( [ cs1 ] ) . _
@xmath203 decay _ : the branching ratio for rare decay can be computed with the help of the general formula of eq .
( [ gammas ] ) , [ gammaa ] b_b_s^0 ^+^-^(a ) & = & ( 1-)^1/2 ( _ 32^d _ 22^e m_a^2 ) ^2 .
we can now eliminate one of the three unknown parameters ( @xmath209 , @xmath210 , and @xmath211 ) which appear in eqs.([dma ] ) and ( [ gammaa ] ) .
we choose to eliminate @xmath209 , so [ arel ] b_b_s^0 ^+^-^(a ) & = & ( 1-)^1/2 ( ) ^2 , where @xmath212 .
as one can see , the unknown factors enter eq .
( [ arel ] ) in the combination @xmath213 .
it is , however , more convenient to plot the dependence on @xmath211 for different values of @xmath210 , which we present in fig .
[ agraph ] .
[ tb ] as a function of pseudoscalar higgs mass @xmath211 . left : @xmath214 ( solid , dashed , dash - dotted lines ) .
right : @xmath215 ( solid , dashed , dash - dotted lines ) . in each figure , the yellow shaded area represents excluded parameter space.,title="fig:",width=340 ] as a function of pseudoscalar higgs mass @xmath211 .
left : @xmath214 ( solid , dashed , dash - dotted lines ) .
right : @xmath215 ( solid , dashed , dash - dotted lines ) . in each figure , the yellow shaded area represents excluded parameter space.,title="fig:",width=340 ] it must be emphasized that the discussion above assumed the absence of large destructive interference of the np and sm contributions to @xmath127 mixing .
concrete models where such interference is present ( and thus the new physics contribution is larger than the sm one ) can be constructed @xcite .
in such models possible contribution to @xmath5 could be large .
one of the simplest extensions of the standard model involves addition of the sequential fourth generation of chiral quarks @xcite , denoted for the lack of the better names by @xmath216 and @xmath217 .
the addition of the sequential fourth generation of quarks leads to a 4@xmath2184 ckm quark mixing matrix @xcite .
this implies that the parameterization of this matrix requires six real parameters and three phases .
besides providing new sources of cp - violation , the two additional phases can affect the branching ratios considered in this paper due to interference effects @xcite .
there are many existing constraints on the parameters related to the fourth generation of quarks .
in particular , a fit of precision electroweak data ( s and t parameters ) @xcite implies that the masses of the new quarks are strongly constrained to be @xcite m_t-m_b ( 1 + ) 50 , with @xmath219 gev . here
@xmath220 is the sm higgs mass , which we take for simplicity to be 120 gev .
we also used updated constraints on ckm matrix elements @xcite .
the relationship between @xmath0 and @xmath17 in the model with four generations of quarks has been previously studied in detail in @xcite . here
we update their result .
the branching ratio of @xmath5 can be related to the experimentally - measured from table [ tab : corr ] , as the separation of np and sm contributions used in the rest of this paper , @xmath221 , is not possible due to loops with both @xmath216 and @xmath222 , @xmath223 , or @xmath134 quarks . ] @xmath224 as @xcite _
b_s ^+^- = , where the parameter @xmath225 is a @xmath2-mixing loop parameter @xcite , ^= _ t s_0(x_t ) + _
t r_tt^2 s_0(x_t^ ) + 2 _ t^ r_tt s_0(x_t , x_t^ ) , and @xmath226 .
@xmath227 can be obtained from table [ tab : corr ] .
the definition of the function @xmath228 can be found in ref .
the wilson coefficient @xmath229 is defined as c_10^tot ( ) = c_10 ( ) + r_tt c_10^t ( ) with @xmath230 obtained by substituting @xmath231 into the sm expression for @xmath232 @xcite .
the results can be found in fig .
[ fourgengraph ] .
as one can see , the resulting branching ratios are still lower than the current experimental bound of eq .
( [ bmm ] ) , but for the values of the four - generation ckm matrix @xmath233 of about @xmath234 , disfavored by @xcite , but still favored by @xcite , can be quite close to it .
[ tb ] as a function of the top - prime mass @xmath231 for different values of the phase @xmath235 ( solid , dashed , dash - dotted lines ) and @xmath236 @xcite ( see also @xcite ) .
right : branching ratio of @xmath237 as a function of the ckm parameter combination @xmath238 with @xmath239 and different values of @xmath240 gev ( solid ) , @xmath241 gev ( dashed ) , and @xmath242 gev ( dash - dotted).,title="fig:",width=340 ] as a function of the top - prime mass @xmath231 for different values of the phase @xmath235 ( solid , dashed , dash - dotted lines ) and @xmath236 @xcite ( see also @xcite ) .
right : branching ratio of @xmath237 as a function of the ckm parameter combination @xmath238 with @xmath239 and different values of @xmath240 gev ( solid ) , @xmath241 gev ( dashed ) , and @xmath242 gev ( dash - dotted).,title="fig:",width=340 ]
experiment has determined @xmath0 exceedingly well .
the standard model determination provides a consistent value , although with a markedly greater uncertainty ( due mainly to the dependence on the nonperturbative quantity @xmath243 and to a lesser extent on the ckm mixing element @xmath47 ) .
we have argued that this fact can be used to constrain np predictions for other processes , such as the @xmath3 transition considered here .
we expect this kind of correlation to be a rather general feature of new physics models , provided there is an overlap between the np parameters which describe @xmath0 and ( for our purposes here ) @xmath3 .
however , given the abundance of new physics scenarios , each with its particular structure , it is not reasonable to expect any _ universal _
correlation between @xmath2-mixing and @xmath5 . instead ,
what we have done in this paper is to analyze several np models in detail . in each case , we have first determined the set of unknown np parameters and then , using dynamical assumptions , have been able to reduce ( or entirely eliminate ) the arbitrariness .
analyzing specific np models this way has two purposes : to serve as an instructive example for further study and to see what kinds of numerical predictions these particular models yield .
not surprisingly , the simplest model ( with a single @xmath19 boson ) provides a strong correlation between @xmath0 and @xmath3 in which the latter is determined in terms of @xmath244 .
an even stronger prediction occurs in the particular version of the family symmetry model discussed earlier , where a clean determination of @xmath3 is obtained . in this instance
, a set of reasonable assumptions allows for the initial presence of unknown parameters to be totally overcome . a similar , but not quite as fortunate , situation occurs for r - parity violating supersymmetry , wherein a reasonable assumption partially reduces the np parameter set . in this case , @xmath3 can be expressed in terms of a ratio of a coupling constant and sneutrino mass @xmath245 .
the flavor - changing higgs model turns out to be less accommodating in that no set of assumptions known to us can reduce the original set of three unknown parameters .
thus , the constraint from @xmath2 mixing still leaves one with two unknowns ( see fig .
[ agraph ] ) .
we also updated constraints on the models with fourth sequential generation of quarks .
of course , additional np models are available for study , _ e.g. _ r - parity conserving supersymmetry @xcite , and work proceeds on these . finally , as discussed in sect .
iii , it would be of interest to address the impact of np cp - violating contributions to @xmath2 mixing . indeed , we plan do so in a future project , but first await more accurate data on @xmath76 or studies of @xmath81 transition at lhcb .
we give our warm thanks to aida el - khadra for useful remarks on the present status of various qcd - lattice predictions .
the work of e.g. was supported in part by the u.s .
national science foundation under grant phy0555304 , j.h . was supported by the u.s .
department of energy under contract de - ac02 - 76sf00515 , s.p . was supported by the u.s .
department of energy under contract de - fg02 - 04er41291 and a.a.p . and g.k.y .
were supported in part by the u.s .
national science foundation under career award phy0547794 , and by the u.s .
department of energy under contract de - fg02 - 96er41005 .
there are eight @xmath246 effective operators that can contribute to @xmath2-mixing .
the operator basis we shall employ is [ setofoperators ] where quantities enclosed in parentheses are color singlets , _
e.g. _ @xmath247 .
these operators are generated at a scale @xmath248 where the np is integrated out .
a non - trivial operator mixing then occurs via renormalization group running of these operators between the heavy scale @xmath248 and the light scale @xmath92 at which hadronic matrix elements are computed .
we need to evaluate the @xmath199-to-@xmath249 matrix elements of these eight dimension - six basis operators .
this introduces eight _ non - perturbative _ b - parameters @xmath250 that require evaluation by means of qcd sum rules or qcd - lattice simulation .
we express these in the form @xmath251 where @xmath252 is the @xmath2 meson decay constant and @xmath253 . ref .
@xcite has performed a qcd - lattice determination ( quenched approximation ) of the b - parameters in an operator basis @xmath254 which is distinct from the @xmath255 of eq .
( [ setofoperators ] ) , @xmath256 three more operators @xmath257 can be obtained by substituting right - handed chiral projection operators with the left - handed ones @xmath258 in eq .
( [ setofoperators1 ] ) .
the @xmath199-to-@xmath249 matrix elements of these operators have been parameterized in ref . @xcite as & & several of the quantities introduced above are scale dependent , _
i.e. _ @xmath262 , @xmath263 and @xmath264 . throughout this paper
, we shall understand all these quantities to be renormalized at a common scale @xmath102 and to simplify notation , we shall denote them simply as @xmath265 , @xmath266 and @xmath267 .
in particular , our evaluation at scale @xmath102 of the quantity @xmath268 yields r_s^2 = m_b_s^2/(|m_b ( |m_b ) + |m_s ( |m_b))^2 = 1.57_-0.10^+0.04 , [ rs ] where we have used the input values @xmath269 gev @xcite and @xmath270 gev @xcite .
l||c|c list of @xmath275 & @xmath275 from lattice qcd & @xmath276 in mvs + ( in @xmath277 basis ) & ( from ref .
@xcite ) & ( from eq . ( [ me_mvs ] ) ) + @xmath278 & @xmath279 & @xmath279 + @xmath280 & @xmath281 & @xmath282 $ ] + @xmath283 & @xmath284 & @xmath285 $ ] + @xmath286 & @xmath287 & @xmath288 + @xmath289 & @xmath290 & @xmath288 + alternatively , the b - parameters can be estimated using the ` modified vacuum saturation ' ( mvs ) approach , wherein all matrix elements in eq .
( [ me ] ) are written in terms of ( known ) matrix elements of @xmath291 and @xmath292 matrix elements @xmath293 and @xmath294 , @xmath295 \ , \\
\langle q_3 \rangle = f_{{\rm b}_s}^2 m_{{\rm b}_s}^2 b_{b_s } \left [ \displaystyle{1 \over 4 n_c } + \displaystyle{\eta \over 2 } \right ] \ , \\
\langle q_4 \rangle = - \displaystyle{2 n_c - 1 \over 4 n_c } f_{{\rm b}_s}^2 m_{{\rm b}_s}^2 b_{b_s}~ \eta \
, \end{array } \qquad \qquad \begin{array}{l } \langle q_5 \rangle = \displaystyle{3 \over n_c } f_{{\rm b}_s}^2 m_{{\rm b}_s}^2 b_{b_s } ~ \eta \ , \\
\langle q_6 \rangle = \langle q_1 \rangle \ , \\ \langle q_7 \rangle = \langle q_4 \rangle \ , \\
\langle q_8 \rangle = \langle q_5 \rangle \ \ , \end{array } \ ] ] where we take @xmath296 as the number of colors and define [ bbar ] & & r_s^2 b_b_s^(s ) = b_b_s .
it is instructive to compare how well the mvs approximation estimates the recent lattice results .
we provide such a comparison in table [ tab : bagconstants ] .
m. bona _ et al . _
[ utfit collaboration ] , pmc phys . a * 3 * , 6 ( 2009 ) [ arxiv:0803.0659 [ hep - ph ] ] . `
measurement of @xmath297 at cdf ' , talk by louise oakes at flavor physics and cp violation conference 2010 , turin , italy ( may 25 - 29 , 2010 ) ; ` new measurement of @xmath2 mixing phase at cdf ' , talk by gavril giurgiu talk delivered at ichep 2010 ( paris , france ) .
v. m. abazov _ et al . _
[ d0 collaboration ] , phys .
d * 82 * , 032001 ( 2010 ) [ arxiv:1005.2757 [ hep - ex ] ] .
e. golowich , j. hewett , s. pakvasa and a. a. petrov , phys .
d * 79 * , 114030 ( 2009 ) [ arxiv:0903.2830 [ hep - ph ] ] .
e. golowich , j. hewett , s. pakvasa and a. a. petrov , phys . rev .
d * 76 * , 095009 ( 2007 ) [ arxiv:0705.3650 [ hep - ph ] ] .
a. j. buras , phys
. lett .
b * 566 * , 115 ( 2003 ) [ arxiv : hep - ph/0303060 ] .
a. lenz and u. nierste , jhep * 0706 * , 072 ( 2007 ) [ arxiv : hep - ph/0612167 ] .
a. j. buras , m. jamin and p. h. weisz , nucl .
b * 347 * , 491 ( 1990 ) .
j. urban , f. krauss , u. jentschura and g. soff , nucl .
b * 523 * , 40 ( 1998 ) [ arxiv : hep - ph/9710245 ] .
`` top quark physics '' , plenary talk by mousumi datta at the 2009 meeting of the division of particles and fields of the american physical society , wayne state university detroit , mi ( 7/26/09- 7/31/09 ) .
s. bethke , eur .
j. c * 64 * , 689 ( 2009 ) [ arxiv:0908.1135 [ hep - ph ] ] .
j. laiho , e. lunghi and r. s. van de water , phys .
d * 81 * , 034503 ( 2010 ) [ arxiv:0910.2928 [ hep - ph ] ] . k. melnikov and t. v. ritbergen , phys .
b * 482 * , 99 ( 2000 ) [ arxiv : hep - ph/9912391 ] .
aida el - khadra , private communication .
t. inami and c.s .
lim , prog .
* 65 * , 297 ( 1981 ) [ erratum - ibid .
* 65 * , 1772 ( 1981 ) ] .
y. grossman , phys .
b * 380 * , 99 ( 1996 ) [ arxiv : hep - ph/9603244 ] .
casey , _ new upper limit on the decay @xmath5 from d0 _ , talk delivered at ichep 2010 ( paris , france ) ; see also v. m. abazov _ et al .
_ [ d0 collaboration ] , phys .
b * 693 * , 539 ( 2010 ) [ arxiv:1006.3469 [ hep - ex ] ] .
g. burdman , e. golowich , j. hewett and s. pakvasa , phys .
d * 66 * , 014009 ( 2002 ) [ arxiv : hep - ph/0112235 ] .
a. badin , f. gabbiani and a. a. petrov , phys .
b * 653 * , 230 ( 2007 ) [ arxiv:0707.0294 [ hep - ph ] ] ; see also a. badin , f. gabbiani and a. a. petrov , arxiv:0909.4897 [ hep - ph ] .
e. golowich , s. pakvasa and a.a .
petrov , phys .
* 98 * , 181808 - 1 ( 2007 ) [ arxiv : hep - ph/0610039 ] .
p. ball and r. fleischer , eur .
j. c * 48 * , 413 ( 2006 ) [ arxiv : hep - ph/0604249 ] .
e. golowich , j. hewett , s. pakvasa , a. a. petrov , and g. k. yeghiyan , to be published .
a. kundu and j. p. saha , phys .
d * 70 * , 096002 ( 2004 ) [ arxiv : hep - ph/0403154 ] .
y. kao and t. takeuchi , arxiv:0910.4980 [ hep - ph ] . h. k. dreiner , m. kramer and b. oleary , phys .
d * 75 * , 114016 ( 2007 ) [ arxiv : hep - ph/0612278 ] . j. p. saha and a. kundu , phys .
d * 66 * , 054021 ( 2002 ) [ arxiv : hep - ph/0205046 ] ; we concur with a conclusion of ref .
@xcite regarding the missing factor of 4 in this paper .
m. sher , phys .
rept . * 179 * , 273 ( 1989 ) .
t. maehara and t. yanagida , lett .
nuovo cim .
* 19 * , 424 ( 1977 ) ; m. a. b. beg and a. sirlin , phys .
lett . * 38 * , 1113 ( 1977 ) ; c. l. ong , phys .
d * 19 * , 2738 ( 1979 ) ; f. wilczek and a. zee , phys . rev .
lett . * 42 * , 421 ( 1979 ) ; a. davidson , m. koca and k. c. wali , phys .
d * 20 * , 1195 ( 1979 ) , phys .
* 43 * , 92 ( 1979 ) .
v. a. monich , b. v. struminsky and g. g. volkov , phys .
b * 104 * , 382 ( 1981 ) [ jetp lett . *
34 * , 213 ( 1981 ) ] [ zh .
34 * , 222 ( 1981 ) ] .
bjorken , s. pakvasa and s.f .
tuan , phys .
d * 66 * , 053008 ( 2002 ) [ arxiv : hep - ph/0206116 ] .
p. f. harrison , d. h. perkins and w. g. scott , phys .
b * 530 * , 167 ( 2002 ) [ arxiv : hep - ph/0202074 ] .
d. becirevic , v. gimenez , g. martinelli , m. papinutto and j. reyes , jhep * 0204 * , 025 ( 2002 ) [ arxiv : hep - lat/0110091 ]
. l. j. hall and s. weinberg , phys .
d * 48 * , 979 ( 1993 ) [ arxiv : hep - ph/9303241 ] .
t. p. cheng and m. sher , phys .
d * 35 * , 3484 ( 1987 ) .
a. pich and p. tuzon , phys .
d * 80 * , 091702 ( 2009 ) [ arxiv:0908.1554 [ hep - ph ] ] . v. d. barger , j. l. hewett and r. j. n. phillips , phys .
d * 41 * , 3421 ( 1990 ) .
d. atwood , l. reina and a. soni , phys .
d * 55 * , 3156 ( 1997 ) [ arxiv : hep - ph/9609279 ] .
r. a. diaz , r. martinez and c. e. sandoval , eur .
j. c * 41 * , 305 ( 2005 ) [ arxiv : hep - ph/0406265 ] ; m. j. savage , phys .
b * 266 * , 135 ( 1991 ) ; m. e. luke and m. j. savage , phys .
b * 307 * , 387 ( 1993 ) [ arxiv : hep - ph/9303249 ] .
a. e. blechman , a. a. petrov and g. yeghiyan , jhep * 1011 * , 075 ( 2010 ) [ arxiv:1009.1612 [ hep - ph ] ] .
h. p. nilles , m. srednicki and d. wyler , phys .
b * 120 * , 346 ( 1983 ) ; j. m. frere , d. r. t. jones and s. raby , nucl .
b * 222 * , 11 ( 1983 ) ; j. p. derendinger and c. a. savoy , nucl . phys .
b * 237 * , 307 ( 1984 ) .
j. r. ellis , j. f. gunion , h. e. haber , l. roszkowski and f. zwirner , phys .
d * 39 * , 844 ( 1989 ) .
u. ellwanger , m. rausch de traubenberg and c. a. savoy , nucl .
b * 492 * , 21 ( 1997 ) [ arxiv : hep - ph/9611251 ] .
g. hiller , phys .
d * 70 * , 034018 ( 2004 ) [ arxiv : hep - ph/0404220 ] .
b. a. dobrescu , phys .
d * 63 * , 015004 ( 2001 ) [ arxiv : hep - ph/9908391 ] .
m. endo and n. yokozaki , arxiv:1012.5501 [ hep - ph ] .
m. bobrowski , a. lenz , j. riedl and j. rohrwild , phys .
d * 79 * , 113006 ( 2009 ) [ arxiv:0902.4883 [ hep - ph ] ] .
v. a. novikov , l. b. okun , a. n. rozanov _ et al .
_ , mod .
lett . *
a10 * , 1915 - 1922 ( 1995 ) .
j. erler and p. langacker , phys .
lett . * 105 * , 031801 ( 2010 ) [ arxiv:1003.3211 [ hep - ph ] ] .
g. d. kribs , t. plehn , m. spannowsky and t. m. p. tait , phys .
d * 76 * , 075016 ( 2007 ) [ arxiv:0706.3718 [ hep - ph ] ] .
a. k. alok , a. dighe and d. london , arxiv:1011.2634 [ hep - ph ] .
a. soni , a. k. alok , a. giri , r. mohanta and s. nandi , phys .
d * 82 * , 033009 ( 2010 ) [ arxiv:1002.0595 [ hep - ph ] ] .
a. j. buras , m. munz , phys .
rev . * d52 * , 186 - 195 ( 1995 ) .
[ hep - ph/9501281 ] . | we perform a study of the standard model ( sm ) fit to the mixing quantities @xmath0 , and @xmath1 in order to bound contributions of new physics to @xmath2 mixing .
we then use this to explore the branching fraction of @xmath3 in certain models of new physics ( np ) . in most cases
, this constrains np amplitudes for @xmath3 to lie below the sm component . |
suppose @xmath5 is an injective homomorphism from a group @xmath1 to itself .
the associated hnn extension @xmath6 is said to be _ ascending _ ; this extension is called _ strictly ascending _ if @xmath7 is not onto . in @xcite ,
drutu and sapir give examples of residually finite @xmath8-relator groups which are not linear ; that is , they do not embed in @xmath9 for any field @xmath10 and dimension @xmath11 .
all of their examples are ascending hnn extensions of free groups , and indeed this is how they know the groups are residually finite ( by @xcite ) . motivated by looking at the @xmath0 representations of their examples , they asked whether @xmath0 contains a strictly ascending hnn extension of a non - abelian free group ( * ? ? ?
* problem 8) .
in particular , they asked if there is a non - abelian free group @xmath1 in @xmath0 which is conjugate to a proper subgroup of itself . in this note , we answer their question in the affirmative . [
main - thm ] the group @xmath0 contains a strictly ascending hnn extension of a free group of rank @xmath2 .
we first outline a general method of finding such groups , and then describe a specific example in more detail .
a large part of our example comes from the fundamental group of a hyperbolic -manifold which fibers over the circle .
let @xmath13 be an open surface of finite type , that is , a closed surface with a non - empty finite set removed .
assume @xmath13 has negative euler characteristic . given a homeomorphism @xmath14 we can form the mapping torus @xmath15 $ } \big/ \raisebox{-4pt}{$(s,1 ) \sim ( \phi(s),0)$}\ ] ] thurston showed that provided @xmath7 is pseudo - anosov ( i.e. @xmath7 is not homotopic to a finite order or reducible homeomorphism )
, @xmath16 admits a _
complete _ hyperbolic structure @xcite . by mostow rigidity , this hyperbolic structure is in fact unique .
if @xmath16 is orientable , then after choosing an orientation , the hyperbolic structure determines a discrete , faithful representation @xmath17 it turns out that @xmath18 always lifts to a representation into @xmath0 ( see e.g. @xcite ) , and henceforth we will regard @xmath0 as the target of @xmath18 . now @xmath19 is an hnn extension @xmath20 since @xmath13 is an open surface of finite type , its fundamental group @xmath21 is a finitely generated free group .
take @xmath22 which is a free subgroup of @xmath0 .
note that for any @xmath23 , the element @xmath24 conjugates @xmath1 to itself . we now restrict attention to @xmath16 with a very special property ;
we will give an example of such an @xmath16 in section [ sec : example ] .
[ requirement ] there is a @xmath25 such that @xmath24 has trace @xmath26 for some integer @xmath27 .
equivalently , we want @xmath24 to have eigenvalues @xmath28 and @xmath29 .
the image of @xmath24 in @xmath0 is a hyperbolic isometry of @xmath30 , and has has two fixed points in @xmath31 .
after conjugating @xmath18 , we can assume these fixed points are @xmath32 and @xmath33 , and @xmath24 acts on @xmath34 by @xmath35 .
for @xmath36 , let @xmath37 denote the parabolic element of @xmath0 which acts on @xmath34 by @xmath38 .
note that @xmath24 conjugates @xmath37 to @xmath39 for each @xmath40 , define @xmath41
. then @xmath42 is finitely generated , and @xmath24 conjugates it to a proper subgroup of itself .
the following lemma , proved in the next section , shows that @xmath42 is free for generic values of @xmath40 .
[ lemma ] let @xmath1 be a finitely generated subgroup of @xmath0 .
consider the group @xmath43 for some @xmath36 .
suppose that no non - trivial element of @xmath1 fixes @xmath44 .
then for generic @xmath40 , the group @xmath42 is isomorphic to the free product @xmath45 . for the action of the fundamental group of a finite volume hyperbolic manifold on @xmath34 , the stabilizer of any point is either empty , an infinite cyclic group consisting of hyperbolic elements , or @xmath46 consisting of parabolic elements .
now @xmath24 is not parabolic as its trace is not @xmath47 .
thus the stabilizer of @xmath33 in @xmath19 is infinite cyclic .
so if some @xmath48 also fixed @xmath33 , then @xmath49 for suitable non - trivial powers @xmath11 and @xmath50 .
since no power of @xmath51 lies in @xmath21 , the element @xmath24 has no fixed point in common with any element of @xmath1 .
thus the lemma applies in our context and @xmath42 is free for most choices of @xmath40 .
thus , given a hyperbolic 3-manifold satisfying the requirement , we get a free group @xmath42 which is conjugate to a proper subgroup of itself by @xmath24
. then the subgroup @xmath52 of @xmath0 is isomorphic to the ascending hnn extension of @xmath42 by the map induced via conjugation by @xmath24 .
this completes our construction and proves theorem [ main - thm ] modulo two things : proving the lemma , and exhibiting an @xmath16 satisfying requirement [ requirement ] where @xmath21 has rank 5 .
these remaining tasks are carried out in the next two sections .
the reader may wonder how plausible it is to expect a manifold satisfying requirement [ requirement ] .
for any finite volume hyperbolic -manifold , the geometric representation @xmath53 can be conjugated so that its image lies in @xmath54 for some finite extension @xmath55 of @xmath56 .
this is because if one looks at representations @xmath57 where the cusp group acts by parabolics , then the representation @xmath18 coming from the hyperbolic structure is an isolated point if we mod out by conjugation @xcite .
the set of conjugacy classes of such representations is an algebraic variety defined over @xmath56 , and hence isolated points have coordinates in a number field .
thus it should not be surprising that the trace of @xmath51 has a particular value in a quadratic field . for our examples ,
the monomorphism of @xmath42 is necessarily reducible , i.e. it preserves a free product decomposition of @xmath42 into proper factors .
mark sapir asked whether there are also examples where this monomorphism is irreducible . for our examples
@xmath42 is necessarily indiscrete .
in fact this must always be the case .
potyagailo and ohshika proved that any topologically tame non - elementary 3-dimensional kleinian group can not be conjugate in @xmath0 to a proper subgroup @xcite .
since all finitely generated kleinian groups are topologically tame @xcite , this implies that there are no examples where the free group base of the hnn extension is discrete .
both authors were partially supported by the u.s . national science foundation ( grant # dms-0405491 ) and the sloan foundation .
the basic idea here is to use the bass - serre tree to show that the `` universal '' group @xmath58 , where @xmath40 is viewed as parameter , is isomorphic to the free product @xmath45 .
it will then easily follow that @xmath42 is also isomorphic to @xmath45 for generic choices of @xmath40 . for notational convenience , we change our parameterization of the parabolic @xmath37 from @xmath59 to @xmath60 . in matrix form
@xmath61 now let @xmath10 be the field of rational functions over @xmath62 .
consider the homomorphism @xmath63 which is just the inclusion of @xmath0 into @xmath64 on the first factor , and sends the generator of @xmath65 to the above parabolic matrix .
we claim that @xmath66 is injective .
consider the discrete valuation on @xmath10 whose value on a rational function @xmath67 is the order of vanishing of @xmath67 at @xmath68 .
let @xmath69 be the valuation ring of @xmath10 ; i.e. the set of rational functions which do not have a pole at @xmath32 .
note that @xmath69 is a local ring , and @xmath70 is its maximal ideal .
then the action of @xmath64 on the bass - serre tree associated to this valuation gives us a splitting of @xmath64 as a free product with amalgamation as follows ( see @xcite or @xcite for details ) .
let @xmath71 and @xmath72 be the conjugate of @xmath73 by @xmath74 if we set @xmath75 then @xmath76 .
( see the top of page 81 of @xcite for this precise splitting in the analogous case of @xmath77}$ ] . )
note that the map @xmath63 takes @xmath1 into the first factor of the amalgam and takes @xmath65 into the other .
the condition that no element of @xmath1 fix @xmath33 exactly corresponds to the condition that the image @xmath78 in @xmath64 does not intersect the edge group @xmath79 joining the two factors of the amalgam . as @xmath80 is also disjoint from @xmath79 , the subgroup of @xmath64 generated by @xmath78 and @xmath80 is isomorphic to @xmath81 . thus @xmath66 is an isomorphism . to complete the proof of the lemma , we will show that for a generic choice of @xmath40 , the composite of @xmath66 with the induced map @xmath82 is also injective .
let @xmath83 and consider the set of @xmath40 for which @xmath84 maps to the identity under the map @xmath85 .
such @xmath40 are the solutions to some polynomial equations in @xmath40 ; as @xmath66 is injective , these equations are nontrivial and have a finite solution set
. as @xmath1 is finitely generated , there are only countably many such @xmath83 and hence only countably many `` bad '' values for @xmath40 .
thus if we select @xmath36 from outside this countable set we have that @xmath42 is the free product @xmath45 , as desired .
if , as is the case in our application , the entries of the elements of @xmath1 live in some number field @xmath55 , then one can simply take @xmath40 to be any transcendental number , e.g. @xmath86 or @xmath87 .
our example of a fibered @xmath88-manifold satisfying requirement [ requirement ] is the orientable double cover of the hyperbolic manifold @xmath89 from the callahan - hildebrand - weeks census @xcite that comes with snappea @xcite .
the manifold @xmath3 is nonorientable and has one torus cusp .
it has an ideal triangulation with @xmath90 tetrahedra .
the shapes of these tetrahedra all lie in the field @xmath91 .
the fundamental group of @xmath3 has a presentation @xmath92 where @xmath93 and @xmath94 .
the _ faithful _
action of @xmath4 on @xmath34 is given by : @xmath95 where @xmath96 , a primitive @xmath97 root of unity .
notice here that @xmath98 is orientation reversing , and that expressed as a matrix @xmath99 which has trace @xmath100 , as desired . in a minute
, we will pass to the orientation cover @xmath101 of @xmath3 to produce the needed example , but first let s observe that @xmath3 fibers over the circle .
first notice that the defining relation for @xmath4 is in the commutator subgroup of the free group on @xmath98 and @xmath102 , and hence @xmath103 .
consider the homomorphism @xmath104 given by @xmath105 and @xmath106 .
if we set @xmath107 and @xmath108 we see that the defining relation of @xmath4 becomes @xmath109 since the highest and lowest subscripts appear only once each , we see by magnus rewriting @xcite that the kernel of @xmath7 is freely generated by @xmath110 . thus by a theorem of stallings @xcite , the map @xmath111 induced by @xmath7
can be homotoped to a fibration where the fundamental group of the fiber is the kernel of @xmath7 .
( actually , for theorem [ main - thm ] , we do nt really need stallings theorem since we have explicitly exhibited the desired free normal subgroup . however , ultimately the justification that this subgroup of matrices is free comes from identifying the full group of matrices with @xmath4 ; that is , the representation given above is faithful .
this is seen by deriving the representation from a hyperbolic structure on @xmath3 , described by a geometric solution to the gluing equations for some ideal triangulation . )
now pass to the orientable double cover @xmath101 of @xmath3 , whose fundamental group is @xmath114 .
( for a dehn surgery description of @xmath101 , see figure [ fig ] . )
note that the kernel of @xmath7 restricted to @xmath115 hasnt changed and is still the free group @xmath1 of rank 5 .
moreover , @xmath116 is not in @xmath1 as @xmath117 and , as we mentioned , the trace of @xmath116 is @xmath100 .
thus @xmath101 satisfies requirement [ requirement ] , and this completes the proof of theorem [ main - thm ] .
it is easy to use brown s elegant algorithm @xcite for computing the bieri - neumann - strebel invariant of a 1-relator group to see that any @xmath118 except @xmath119 and @xmath120 corresponds to a fibration .
( here @xmath121 is the dual basis of @xmath122 with respect to the basis @xmath123 of @xmath124 . )
this gives infinitely many other examples , where the rank of the free group in question grows more slowly than in examples obtained by taking subgroups of finite index in the kernel of @xmath125 . of the @xmath126
cusped hyperbolic 3-manifolds in the callahan - hildebrand - weeks census , @xmath127 seems to be the only one that satisfies requirement [ requirement ] .
there are a handful of other examples ( mostly also non - orientable ) which have an element with the right trace , but these are not fibered .
however , it is plausible that they have finite covers which are fibered .
john stallings , _ on fibering certain @xmath88-manifolds _ , topology of 3-manifolds and related topics ( proc . the univ . of georgia institute , 1961 ) , prentice - hall , englewood cliffs , n.j .
, 1962 , pp .
mr 28 # 1600 | we give an example of a subgroup of @xmath0 which is a strictly ascending hnn extension of a non - abelian finitely generated free group @xmath1 .
in particular , we exhibit a free group @xmath1 in @xmath0 of rank @xmath2 which is conjugate to a proper subgroup of itself .
this answers positively a question of drutu and sapir @xcite .
the main ingredient in our construction is a specific finite volume ( noncompact ) hyperbolic 3-manifold @xmath3 which is a surface bundle over the circle . in particular , most of @xmath1 comes from the fundamental group of a surface fiber .
a key feature of @xmath3 is that there is an element of @xmath4 in @xmath0 with an eigenvalue which is the square root of a rational integer .
we also use the bass - serre tree of a field with a discrete valuation to show that the group @xmath1 we construct is actually free . |
we would like to acknowledge useful discussions with t. devereaux and a.f .
ph and djs would like to thank the stanford institute for materials and energy sciences for their hospitality . tam and djs would like to acknowledge support from the center for nanophase materials sciences , which is sponsored at oak ridge national laboratory by the office of basic energy sciences , u.s .
department of energy .
this work was supported by doe de - fg02 - 05er46236 ( pjh ) , the dfg through trr80 and the free state of bavaria through the bacatec program ( sg ) . | neutron spin - flip scattering observations of a resonance in the superconducting state is often taken as evidence of an unconventional superconducting state in which the gap changes sign @xmath0 for momentum transfers @xmath1 which play an important role in the pairing .
recently questions regarding this identification for the fe - pnictide superconductors have been raised and it has been suggested that @xmath2 .
here we propose that inelastic neutron or x - ray scattering measurements of the spectral weight of a phonon of momentum @xmath1 can distinguish between these two pairing scenarios .
determining the superconducting gap structure in the fe - based superconductors has proven difficult due to the complicated multisheeted , multiorbital nature of the fermi surface , as well as problems with fabricating clean films and interfaces . at present the most popular candidate for pairing in most of these materials is the @xmath3 state which changes sign between hole and electron pockets and arises naturally from a spin - fluctuation interaction@xcite .
however , there are few phase - sensitive tests of the relative sign change of @xmath4 between the different fermi surfaces .
josephson tunneling experiments @xcite , fourier transform stm in a magnetic field @xcite , and the observation of a resonance mode in inelastic neutron scattering experiments@xcite have been offered as limited evidence for such a sign change .
recently , an electron - phonon interaction enhanced by orbital fluctuations has been proposed as the mechanism which is responsible for pairing in the fe - based superconductors @xcite .
this interaction leads to a so - called @xmath5 gap which is positive on both the hole and electron fermi surfaces . while in both cases
, the gap can be anisotropic and accidental nodes can even occur , the relative sign of the gap on regions of the hole and electron fermi surfaces which have the same dominant orbital weight will be negative if the pairing is driven by spin - fluctuations and positive if driven by phonon - orbital - fluctuations .
thus it is important to determine the relative sign between @xmath6 and @xmath7 on the hole and electron fermi surfaces , respectively , for momentum transfers @xmath1 which connect such regions .
one test of this is the occurrence of a spin resonance peak in the neutron spin - flip scattering in the superconducting state @xcite .
the bcs coherence factor which enters the spin - flip scattering is @xmath8 here @xmath9 is the quasi - particle energy and near threshold , where @xmath10 goes to zero , this coherence factor goes to 1 if @xmath7 and @xmath6 have opposite signs and zero if they have the same sign .
a resonance peak is expected in the former case @xcite .
however , arguments @xcite have been given that for the case of bafe@xmath11co@xmath12as@xmath13 the experimentally observed peak @xcite may be consistent with an @xmath5 gap , if the scattering rate collapses sufficiently rapidly in the superconducting state .
thus one would like to find additional experimental probes which provide a way of determining the relative sign of the gap between the hole and electron pockets .
as is known , there exist anomalies in the phonon spectral weight of conventional superconductors .
various authors @xcite have shown that if the phonon frequency @xmath14 is greater than twice the gap @xmath15 , but not too far above , significant spectral weight is transferred below @xmath16 into a resonance .
in addition , the phonon peak shifts to higher frequencies and broadens .
these effects have been convincingly observed in neutron scattering experiments on the borocarbide superconductor yni@xmath13b@xmath13c @xcite .
this behavior reflects the fact that in the superconducting state , the coherence factor that enters the coupling of a phonon to electron polarization fluctuations is similar to eq .
( [ eq:1 ] ) except that it has a plus sign . in this case
the coherence factor will go to 1 at threshold if @xmath7 and @xmath6 have the same sign .
this will lead to an anomalous resonance and hardening of the phonon peak for an @xmath5 gap . here
we point out that for an @xmath3 gap with @xmath17 , the anomalous resonance is absent and the phonon peak shifts down ( softens ) slightly .
thus neutron and x - ray scattering @xcite measurements of the phonon spectral weight at specific @xmath1 values can provide information on the relative sign of the gap @xcite . to explore the use of this effect to determine the relative gap signs in the fe - based materials
, we will use the same 5-orbital tight binding fit of the lda bandstructure that we previously used to discuss the possible occurrence of a neutron scattering spin resonance @xcite . in this case , with an electron doping @xmath18 one finds the fermi surfaces and orbital occupations illustrated in fig .
[ fig:1 ] .
here there are two hole fermi surfaces @xmath19 and @xmath20 around the @xmath21 point and two electron fermi surfaces @xmath22 and @xmath23 around the @xmath24 and @xmath25 points of the unfolded ( 1fe / cell ) brillouin zone .
we will use an orbital notation @xmath26 to denote the @xmath27 , @xmath28 , @xmath29 , @xmath30 and @xmath31 orbits , respectively , where the axes are along the fe - fe bonds .
the dominant orbital weight of the bloch states on the fermi surfaces is indicated in fig .
[ fig:1 ] .
[ htbp ] , green=@xmath28 , blue=@xmath29 ) . here
@xmath32 connects @xmath27 parts of the hole @xmath20 and @xmath28 parts of the electron @xmath22 fermi surfaces and @xmath33 connects @xmath27 parts of @xmath19 to the @xmath29 parts of @xmath22 .
[ fig:1],title="fig:",width=264 ] as kontani and onari @xcite have noted , local fe - ion displacements couple to orbital fluctuations . the polarization susceptibility associated with these orbital fluctuations enters the phonon self - energy and is reflected in the phonon spectral weight .
here we examine the change in the phonon spectral weight @xmath34 associated with the @xmath35 and @xmath36 momenta shown in fig .
[ fig:1 ] when the system goes from a normal state to a superconducting state with a nominal @xmath5 gap or an @xmath3 gap .
we begin with @xmath32 and will assume that there is an electron - phonon coupling @xmath37 for scattering an electron between the dominantly @xmath27-like states on the @xmath20 hole fermi surface to the @xmath28-like states on the @xmath22 electron fermi surfaces . in the following , the real part of the normal state phonon self - energy will be absorbed to give the dressed normal state phonon frequency @xmath38 .
then at low energies , the imaginary part of the phonon self - energy for momentum @xmath32 dominantly comes from scattering between the nearly nested @xmath27-like states on @xmath20 and the @xmath28-like states on @xmath22 .
similarly , the change between the superconducting and normal state phonon self - energy is also determined by the @xmath27@xmath28 orbital fluctuations . in this case
the important orbital contribution to the phonon self - energy is @xmath39 with @xmath40^{-1}\right\}_{1212 } \label{eq:3}\ ] ] and similarly for @xmath41 . here
the coulomb interaction matrix is @xmath42 and @xmath43 the notation follows ref .
@xcite with @xmath44 and @xmath45 the intra- and inter - orbital coulomb interactions , @xmath46 the exchange coupling and @xmath47 the pair hopping interactions .
we will consider a typically rotationally invariant set of interaction parameters which are similar to those previously used to study the occurrence of a spin resonance in the @xmath3 superconducting state @xcite .
these are such that @xmath48 in units of the hopping @xmath49 , @xmath50 , @xmath51 , and @xmath52 @xcite . in the superconducting state the normal and anomalous gorkov green s functions are given as @xmath53 with @xmath54 .
for the normal state calculations , the gap is set to zero in eq .
( [ eq:5 ] ) . here , because we are calculating the orbital - fluctuations , the sign between the normal and anomalous green s function products in eq .
( [ eq:5 ] ) is negative
. it would be positive if we were studying the spin susceptibility and in addition the orbital interaction matrix @xmath55 would be changed to @xmath56 . in eq .
( [ eq:5 ] ) we use a notation in which @xmath57 and @xmath58 . the calculations were carried out for a low temperature @xmath59 in units of @xmath60 .
[ htbp ] versus @xmath61 for the normal state ( red , solid ) , the @xmath5 superconducting state with @xmath62 ( blue , dashed ) and the @xmath3 state with @xmath63 ( green , dot - dashed ) .
[ fig:2],title="fig:",width=302 ] figure [ fig:2 ] shows the imaginary part of @xmath64 versus @xmath61 with @xmath65 for the normal state ( solid ) , the superconducting states with an @xmath66 gap ( dashed ) and an @xmath67 gap ( dash - dot ) @xcite .
the imaginary part of the orbital response is similar to what was found for the spin susceptibility with the @xmath3 and @xmath5 gaps interchanged .
thus , if the phonon - orbital - fluctuations provide the dominant pairing interaction in the fe - based superconductors then the gap will have @xmath5 character and one will find an anomalous resonance in the orbital susceptibility in the superconducting state rather than in the spin susceptibility . although it would be interesting to have a direct scattering probe of the orbital susceptibility spectral weight shown in fig .
[ fig:2 ] , one can still access its structure by measuring the phonon spectral weight . using the orbital susceptibility , we can evaluate the phonon self - energies that enter the normal and superconducting phonon progagators @xcite , @xmath68 and @xmath69 here @xmath70 and @xmath71 and @xmath72 are the superconducting and normal phonon self - energies obtained from eq .
( [ eq:2 ] ) , with the gap set to zero in eq .
( [ eq:5 ] ) for the normal case and set to @xmath73 and @xmath74 for the @xmath5 and @xmath3 superconducting states , respectively .
[ htbp ] versus @xmath61 for the normal state ( red , solid ) , the @xmath5 ( blue , dashed ) and @xmath3 ( green , dot - dashed ) superconducting gaps .
( b ) the real part @xmath75 ( see text ) for the @xmath5 ( red , solid ) and @xmath3 ( blue , dashed ) superconducting gaps .
[ fig:3],title="fig:",width=302 ] the phonon damping @xmath76 versus @xmath77 is plotted in fig .
[ fig:3]a .
here we have taken @xmath78 and set @xmath79 . as @xmath38 approaches @xmath80 ,
the effects we are discussing are enhanced and diminish when @xmath81 becomes large compared with @xmath80 . as seen in fig .
[ fig:3]a , for an @xmath5 gap there is an enhanced phonon damping above @xmath80 .
this arises from the increase in the density of states in this region due to the opening of the superconducting gap .
this enhancement also depends upon the coherence factor and as seen for the @xmath3 case , if the coherence factor vanishes , the onset of the damping rises slowly as @xmath61 exceeds @xmath80 and is smaller than in the normal state . in fig .
[ fig:3]b , we have plotted the difference between the real part of the phonon self - energy in the superconducting and the normal states asociated with the @xmath82@xmath83 scattering processes .
the @xmath5 case shows the familiar hardening of the phonon that occurs when @xmath38 exceeds @xmath80 .
that is , in the @xmath5 superconducting state , @xmath84 is positive for @xmath85 .
as we will see when we look at the phonon spectral weight , the negative dip for @xmath86 leads to what allen et al.@xcite have called a vibrational / superelectronic " resonance .
the structure of @xmath84 is quite difference in the @xmath3 state . here for @xmath85 , @xmath87 is negative and this leads to a softening of the @xmath38 phonon .
in addition , there will not be a resonance feature below @xmath80 in the phonon spectral weight . using the results for the self - energy shown in fig .
[ fig:3 ] , we have plotted the phonon spectral weight @xmath88 versus @xmath77 in fig . [ fig:4]a .
the solid line shows the result for the normal state while the long dashed and the dash - dot curves are for the superconducting @xmath5 and @xmath3 states , respectively .
here one sees that for an @xmath5 gap there is the expected resonance just below @xmath80 and the phonon peak has shifted up in frequency and broadened . for the @xmath3 gap ,
the anomalous resonance below @xmath80 is absent and the phonon peak has slightly softened as well as narrowed .
similar results for the @xmath33 phonon spectral weight are shown in fig .
[ fig:4]b .
[ htbp ] versus @xmath77 for wavevectors ( a ) @xmath32 and ( b ) @xmath33 for the normal state ( red , solid ) , the @xmath5 ( blue , dashed ) and @xmath3 ( green , dot - dashed ) superconducting gaps .
note that we have added a finite damping in the calculation of the orbital susceptibility @xmath89 .
[ fig:4],title="fig:",width=321 ] in rpa calculations @xcite there were parameter ranges in which the @xmath3-like gap was found to have nodes on the @xmath90 fermi surfaces . in this case , the nominally @xmath3 gap will have the same sign on the parts of the fermi surfaces connected by @xmath33 .
however , they would differ for @xmath32 .
thus , it will be important to measure the phonon spectral weight for various near nesting @xmath1 values .
however , one will need phonon modes that couple to the relevant orbital scattering and in addition have frequencies @xmath81 of order the sum of the magnitudes of the superconducting gaps @xmath91 associated with the nearby nested regions . in principle , with sufficient @xmath92 and @xmath61 resolution and sensitivity , neutron and x - ray scattering offer the possibility of mapping out the relative structure of the gap between various parts of the fermi surfaces in fe - based as well as other unconventional superconductors . |
high - energy muon beams have been proposed as uniquely powerful and incisive sources for neutrino scattering and oscillation studies .
they may also enable energy - frontier lepton - antilepton colliders and may have unique advantages for studying the physics of electroweak symmetry breaking . the production of high - energy muon beams at the intensities needed for these applications will require muon - beam cooling @xcite .
to accelerate a secondary or tertiary beam it is desirable first to decrease its size so that a reasonable fraction of the produced particles will fit inside the apertures of the beamline .
it is well known that a focusing element ( _ e.g. _ a pair of quadrupole magnets with opposed field gradients ) can decrease the area of a charged - particle beam while increasing its spread in transverse momentum and , consequently , its divergence .
this relationship is an example of liouville s theorem : conservative forces can not increase or decrease the volume occupied by a beam in six - dimensional phase space @xcite .
focusing alone does not suffice for efficient acceleration of a secondary or tertiary beam , since the resulting increase in divergence means the beam will exceed some other aperture further downstream .
what is needed instead is a process by which _ both _ the beam size and divergence can be reduced . by analogy with refrigeration , which decreases the random relative motions of the molecules of a gas ,
this is known as beam _
cooling_. it is convenient to represent the volume of phase space occupied by a beam by the beam s _ emittance_.
the emittance in a given coordinate can be expressed as @xmath0 where @xmath1 designates root - mean - square , @xmath2 , and the factor @xmath3 is introduced so as to express emittance in units of length ( @xmath4 is the mass of the beam particle and @xmath5 the speed of light ) . neglecting possible correlations among the coordinates and momenta @xcite
, we then have @xmath6 for the six - dimensional emittance .
the subscript @xmath7 distinguishes these _ normalized _ emittances from the frequently used _
unnormalized _ emittance @xmath8 where @xmath9 and @xmath10 are the usual relativistic factors . in terms of ( unnormalized ) emittance , the transverse beam sizes are given by @xmath11 where @xmath12 are the transverse amplitude functions of the focusing lattice in the @xmath13 and @xmath14 directions , which characterize the focusing strength along the lattice ( low @xmath15 corresponds to strong focusing in the @xmath16th direction ) . since liouville s theorem tells us that normalized emittance is a constant of the motion , beam cooling requires a violation " of liouville s theorem .
this is possible by means of dissipative forces such as ionization energy loss @xcite , as described in more detail below .
cooling of the transverse phase - space coordinates of a muon beam can be accomplished by passing the beam through energy - absorbing material and accelerating structures , both embedded within a focusing magnetic lattice ; this is known as ionization cooling @xcite .
other cooling techniques ( electron , stochastic , and laser cooling ) are far too slow to yield a significant degree of phase - space compression within the muon lifetime @xcite .
ionization of the absorbing material by the muons decreases the muon momentum while ( to first order ) not affecting the beam size ; by eq . [ eq : eps - n ] , this constitutes cooling . at the same time
, multiple coulomb scattering of the muons in the absorber increases the beam divergence , heating the beam . differentiating eq .
[ eq : eps - n ] with respect to path length , we find that the rate of change of normalized transverse emittance within the absorber is given approximately by @xcite @xmath17 where angle brackets denote mean value , muon energy @xmath18 is in gev , @xmath19 is evaluated at the location of the absorber , and @xmath20 is the radiation length of the absorber medium @xcite .
( this is the expression appropriate to the cylindrically - symmetric case of solenoidal focusing , where @xmath21 . ) the first term in eq .
[ eq : cool ] is the cooling term and the second is the heating term . to minimize the heating term , which is proportional to the @xmath10 function and inversely proportional to radiation length , it has been proposed @xcite to use hydrogen as the energy - absorbing medium , giving @xmath22mev / m and @xmath23 m @xcite , with superconducting - solenoid focusing to give small @xmath24 cm @xcite .
key issues in absorber r&d include coping with the large heat deposition by the intense ( @xmath25s ) @xcite muon beam and minimizing scattering in absorber - vessel windows @xcite .
an additional technical requirement is high - gradient reacceleration of the muons between absorbers to replace the lost energy , so that the ionization - cooling process can be repeated many times . even though it is the absorbers that actually cool the beam , for typical radio - frequency ( rf ) accelerating - cavity gradients ( @xmath26mev / m ) , it is the rf cavities that dominate the length of the cooling channel ( see _ e.g. _ fig .
[ fig : sfofo ] ) , and the achievable rf gradient determines how much cooling is practical before an appreciable fraction of the muons have decayed .
we see from eq .
[ eq : cool ] that the percentage decrease in normalized emittance is proportional to the percentage energy loss , thus cooling in one transverse dimension by a factor 1/@xmath27 requires @xmath28100% energy loss and replacement .
the expense of rf power favors low beam energy for cost - effective reacceleration , as do also the increase of @xmath29 and the decreased width of the @xmath30 distribution at low momentum @xcite , and most muon - cooling simulations to date have used @xmath31mev/@xmath5 .
we are pursuing r&d on high - gradient normal - conducting rf cavities suitable for insertion into a solenoidal - focusing lattice @xcite .
transverse ionization cooling causes the longitudinal emittance @xmath32 to grow .
several effects contribute to this growth : fluctuations in energy loss in the absorbers ( energy - loss straggling , or the landau tail " ) cause growth in the energy spread of the beam , as does the negative slope of the @xmath29 momentum dependence in the beam - momentum regime ( below the ionization minimum ) that we are considering @xcite .
moreover , these low - momentum , large - divergence beams have a considerable spread in propagation velocity through the cooling lattice , causing bunch lengthening .
these effects result in gradual loss of particles out of the rf bucket .
they could be alleviated by longitudinal cooling .
longitudinal ionization cooling is possible in principle , but it appears to be impractical @xcite .
its realization would call for operation _ above _ the ionization minimum , where the @xmath29 slope with momentum is positive @xcite , but that slope is small and the resulting weak cooling effect is overcome by energy - loss straggling .
instead what is envisioned is emittance _ exchange _ between the longitudinal and transverse degrees of freedom , decreasing the longitudinal emittance while at the same time increasing the transverse .
conceptually , such emittance exchange can be accomplished by placing a suitably shaped absorber in a lattice location where there is dispersion , _
i.e. _ , using a bending magnetic field to spread the muons out in space according to their momenta , and shaping the absorber so as to absorb more energy from the higher - momentum muons and less from the lower - momentum ones .
( one can see that this is emittance _ exchange _ rather than longitudinal cooling _
per se _ , since to the extent that the muon momentum spread has been reduced by the shaped absorber , the beam can no longer be reconverged to a small spot by a subsequent bend . )
this is followed by transverse ionization cooling , the combined process being effectively equivalent to longitudinal cooling @xcite .
a variety of focusing - lattice designs for transverse muon cooling have been studied , most using solenoids as focusing elements . especially for the large ( @xmath33 m ) aperture required at the beginning of a muon cooling channel , stronger focusing gradients are possible using solenoids than using quadrupoles , and unlike quadrupoles , solenoids have the virtue of focusing both transverse dimensions simultaneously , giving a more compact lattice . while a high - field solenoid can produce a small ( and constant ) @xmath19
, it is straightforward to see that a single such solenoid is not sufficient for muon cooling @xcite . a charged particle entering a solenoid off - axis receives a transverse magnetic kick from the fringe field , such that the particle s straight - line motion in the field - free region becomes helical motion within the solenoid .
the exit fringe field must thus impart an equal and opposite kick so that the particle resumes its straight - line motion in the subsequent field - free region .
if within the solenoid the particle loses energy in an absorbing medium , the angular momentum of its helical motion must decrease , resulting in an imbalance between the entrance and exit kicks .
the particle then exits the magnet with a net angular momentum , implying that a parallel beam entering an absorber - filled solenoid will diverge upon exiting . to cancel this net angular momentum ,
the field direction must alternate periodically .
the simplest case conceptually is focusing by a constant solenoidal field , but with one field flip " halfway along the cooling channel @xcite .
the length of a uniform section can be of order @xmath34 m .
better performance can be achieved by adding a second field flip @xcite . at an opposite extreme
, the solenoidal - field direction can be flipped every meter or so , leading to a variety of solenoidal - focusing lattices dubbed alternating solenoid @xcite , fofo @xcite , dfofo @xcite , sfofo ( see fig . [
fig : sfofo ] ) , etc .
detailed six - dimensional simulations show that enough transverse cooling can be achieved to build a high - performance neutrino factory @xcite , using either a double - flip or sfofo cooling lattice .
for example , in palmer s recent sfofo design an initial transverse normalized emittance of @xmath35mm@xmath36rad is cooled in a 400-m - long cooling system to a final emittance of @xmath37mm@xmath36rad with @xmath2875% muon loss @xcite .
such a facility would produce @xmath38 neutrinos per year aimed at a far detector that could be thousands of km from the source , giving oscillation sensitivity at least two orders of magnitude beyond that of long - baseline experiments now under construction @xcite . without longitudinal - transverse emittance exchange , transverse cooling reaches a point of diminishing returns as emittance growth in the longitudinal phase plane causes muons to be lost from the rf bucket . while emittance exchange would be helpful but not essential for a neutrino factory , to achieve the considerably smaller emittances required in a muon collider , it is mandatory .
r&d on emittance exchange is ongoing , and several promising ideas are being actively explored @xcite .
however , _ nonlinear _ conservative forces can cause phase - space filamentation , " the practical effect of which is essentially growth of the beam s occupied phase - space volume , and stochastic cooling can be thought of as the opposite effect , in which particles are moved in phase space to make occupied regions of phase space more contiguous and move empty regions of phase space outside of the beam envelope .
g. k. oneill , phys .
rev . * 102 * , 1418 ( 1956 ) ; a. a. kolomensky , sov . atomic energy * 19 * , 1511 ( 1965 ) ; g. i. budker , a. n. skrinsky , sov .
* 21 * , 277 ( 1978 ) ; d. neuffer , fnal report fn-319 ( 1979 ) ; a. n. skrinsky , v. v. parkhomchuk , sov .
j. nucl .
* 12 * , 223 ( 1981 ) ; d. neuffer , part .
* 14 * , 75 ( 1983 ) ; e. a. perevedentsev , a. n. skrinsky , in _ proc .
12th int .
conf . on high energy accelerators _ ,
f. t. cole , r. donaldson , eds .
( fermilab , 1984 ) , p. 485 .
however , at sufficiently high energy , a proposed technique known as optical stochastic cooling " may become practical ; see a. zholents , m. zolotorev , w. wan , phys .
st accel .
beams * 4 * , 031001 ( 2001 ) .
another possibility is liquid - lithium lenses , which could serve as both the focusing element and the absorbing material .
prototype studies in novosibirsk have so far been unsuccessful .
a third possibility is high - pressure gaseous absorbers @xcite .
a. moretti _ et al .
_ , proc .
20th international linac conference ( linac 2000 ) , monterey , california , 21 - 25 aug 2000 , p. 896
( econf c000821:thc18 , 2000 ) ; j. corlett _
et al . _ , to appear in proc .
pac2001 ( 2001 ) .
the numbers quoted above include the minicooling " section , located upstream of the buncher , " which precedes the sfofo channel . in the @xmath28110-m - long sfofo channel itself the emittance is cooled from @xmath2813@xmath42 to 2.8@xmath42mm@xmath36rad with @xmath2850% muon loss . | starting from elementary concepts , muon - beam cooling is defined , and the techniques by which it can be accomplished introduced and briefly discussed . |
since the late 1990s , an accelerated expansion of the late universe has gradually been accepted as a new paradigm @xcite . in order to explain this accelerated expansion , various cosmological models
have been suggested @xcite .
( see , e.g. , refs .
@xcite and the references therein . ) as one such model , easson , frampton , and smoot @xcite recently proposed an entropic - force scenario called ` entropic cosmology ' . in entropic cosmology , an extra driving term ,
i.e. , an entropic - force term , should be added to the friedmann
lematre acceleration equation , without introducing new fields @xcite .
the entropic - force term is derived from the usually neglected surface terms on the horizon of the universe , assuming that the horizon has an entropy and a temperature due to the information holographically stored there @xcite .
entropic cosmology has been extensively examined from various viewpoints @xcite . as an entropy on the horizon , i.e.
, the bekenstein entropy @xcite is always used , substituting the horizon of the universe for the event horizon of a black hole .
in fact , the bekenstein entropy is proportional to area ( or horizon ) and is additive .
however , self - gravitating systems exhibit peculiar features @xcite , such as nonequilibrium thermodynamics and nonextensive statistical mechanics @xcite .
accordingly , for example , the tsallis entropy @xcite and the renyi entropy @xcite have been proposed for nonadditive ( nonextensive ) generalized entropies and have been investigated from astrophysical viewpoints @xcite . in particular , tsallis and
cirto recently suggested a generalized black - hole entropy proportional to its volume based on appropriate nonadditive generalizations @xcite .
using the generalized entropy instead of the bekenstein entropy will provide new insight into entropic cosmology .
therefore , it is important to derive entropic - force models from the two entropies and examine the properties of the two models in order to understand entropic cosmology more deeply .
note that power - law and logarithmic entropic - corrections have been discussed @xcite .
in addition , in entropic cosmology , the entropy on the horizon of the universe can increase during the evolution of the universe @xcite , even if we consider a homogeneous and isotropic universe .
however , a bulk viscosity of cosmological fluids @xcite is usually the only thing that can generate an entropy in the homogeneous and isotropic universe @xcite .
( such a cosmological model is referred to as bulk viscous cosmology . ) through the study of entropic cosmology , we may be able to discuss the classical entropy generated by bulk viscous stresses . in this context
, we examine entropic cosmology using a generalized black - hole entropy ( proportional to its volume ) proposed by tsallis and cirto @xcite . in the present study
, we derive entropic - force terms not only from the bekenstein entropy but also from the generalized entropy .
moreover , using an effective description similar to bulk viscous cosmology , we formulate the modified friedmann , acceleration , and continuity ( conservation ) equations for entropic cosmology .
note that the entropic - force considered here is different from the idea that gravity itself is an entropic force @xcite .
since we focus on background evolutions of the late universe , we do not discuss the inflation of the early universe .
the remainder of the present paper is organized as follows . in sec .
[ cosmological models ] , we present a brief review of three cosmological models , i.e. , @xmath0cdm ( lambda cold dark matter ) cosmology , bulk viscous cosmology , and entropic cosmology . in sec . [ derivation of entropic - force ] , we derive entropic - force terms from both the bekenstein entropy and a generalized black - hole entropy . we also formulate the modified friedmann , acceleration , and continuity equations for two entropic - force models based on the obtained entropic - force terms . in sec .
[ combined model ] , we examine a model that combines the two entropic - force models . in sec . [ comparison ] , we discuss the evolution of the universe in the two entropic - force models using solutions of the combined model . finally , in sec .
[ conclusions ] , we present a discussion and our conclusions .
in the present paper , we consider a homogeneous , isotropic , and spatially flat universe and examine the scale factor @xmath1 at time @xmath2 in the friedmann lematre robertson
walker metric . in this section
, we present a brief review of three cosmological models , i.e. , @xmath0cdm , bulk viscous , and entropic cosmologies , focusing on the friedmann , acceleration , and continuity equations .
( a spatially non - flat universe is discussed in the last paragraph of in sec .
[ combined model ] . )
we first introduce the well - known @xmath0cdm models @xcite . in the standard @xmath0cdm model ,
the friedmann equation is given as @xmath3 and the acceleration equation is @xmath4 in addition , the continuity equation is given by @xmath5 where the hubble parameter @xmath6 is defined by @xmath7 and @xmath8 , @xmath0 , @xmath9 , @xmath10 , and @xmath11 are the gravitational constant , a cosmological constant , the speed of light , the mass density of cosmological fluids , and the pressure of cosmological fluids , respectively . equations ( [ eq : frw1_lcdm ] ) and ( [ eq : frw2_lcdm ] ) include the extra driving terms , i.e. , @xmath12 , which can explain the accelerated expansion of the late universe .
the continuity equation is consistent with the friedmann and acceleration equations , because two of the three equations are independent @xcite .
in other words , if the friedmann and acceleration equations are used , eq . ( [ eq : fluid_lcdm ] )
is derived from these two equations .
note that eq .
( [ eq : fluid_lcdm ] ) can be derived from the first law of thermodynamics as well , assuming adiabatic ( isentropic ) processes , without using the friedmann and acceleration equations @xcite . in bulk viscous cosmology , a bulk viscosity @xmath13 of cosmological fluids
is assumed @xcite . usually , the bulk viscosity is the only thing that can generate an entropy in the homogeneous and isotropic universe @xcite .
such a cosmological model is referred to as bulk viscous cosmology .
for example , in the 1980s , barrow @xcite , davies @xcite , and lima _ et al_. @xcite investigated bulk viscous cosmology in order to discuss the inflation of the early universe .
recently , a number of studies have examined not only the inflation but also the accelerated expansion of the late universe , based on bulk viscous cosmology @xcite .
generally , an effective pressure @xmath14 for bulk viscous cosmology is given by @xmath15 where , for simplicity , @xmath13 is a constant .
( it is possible to assume a variable bulk - viscosity .
for instance , see refs .
@xcite and the references therein . ) in bulk viscous cosmology , the friedmann equation is given by @xmath16 unlike the @xmath0cdm model [ eq .
( [ eq : frw1_lcdm ] ) ] , eq .
( [ eq : bulk_frw1 ] ) does not include an extra term such as a cosmological constant . using the effective pressure @xmath14 , the acceleration equation for bulk viscous cosmology
is given by @xmath17 substituting eq .
( [ eq : bulk_p ] ) into eq . ( [ eq : bulk_frw2pprime ] ) and rearranging , we obtain the acceleration equation as @xmath18 the last term , @xmath19 , corresponds to the extra driving term @xcite .
accordingly , instead of the cosmological constant , the extra term due to the bulk viscosity can explain the accelerated expansion of the universe .
note that the extra term considered here is proportional to the hubble parameter @xmath6 .
the continuity equation for bulk viscous cosmology is given by @xmath20 substituting eq .
( [ eq : bulk_p ] ) into eq .
( [ eq : bulk_fluid0 ] ) and rearranging , we have @xmath21 equation ( [ eq : bulk_fluid1 ] ) has a non - zero right - hand side related to a classical entropy generated by bulk viscous stresses @xcite .
we now derive the continuity equation using a different approach . to this end
, we use the generalized continuity equation obtained from the general friedmann and acceleration equations @xcite , because two of the three equations are independent . the general friedmann and acceleration equations
are given , respectively , by @xmath22 and @xmath23 where @xmath24 and @xmath25 are general functions .
using eqs .
( [ eq : general_frw01_f ] ) and ( [ eq : general_frw02_g ] ) , we obtain the generalized continuity equation @xcite given by @xmath26 comparing [ eqs .
( [ eq : general_frw01_f ] ) and ( [ eq : general_frw02_g ] ) ] with [ eqs .
( [ eq : bulk_frw1 ] ) and ( [ eq : bulk_frw2 ] ) ] , we can set the general functions as @xmath27 and @xmath28 . substituting these functions into eq .
( [ eq : drho_general0 ] ) , we obtain @xmath29 equation ( [ eq : bulk_fluid_g1 ] ) is the same as eq .
( [ eq : bulk_fluid1 ] ) .
this indicates that eq .
( [ eq : bulk_fluid1 ] ) obtained from the effective pressure is consistent with eq .
( [ eq : bulk_fluid_g1 ] ) obtained from the friedmann and acceleration equations .
the consistency plays an important role in formulating the entropic - force models discussed in sec .
[ derivation of entropic - force ] . in the entropic cosmology suggested by easson _
@xcite , the horizon of the universe is assumed to have an associated entropy and an approximate temperature .
the entropy considered here is the bekenstein entropy . in the present study
, we refer to this cosmology as the standard entropic cosmology .
( the bekenstein entropy is discussed in detail in sec .
[ entropic - force from the bekenstein entropy ] . ) in a study by koivisto _
@xcite , the modified friedmann and acceleration equations are summarized as @xmath30 and @xmath31 the four coefficients @xmath32 , @xmath33 , @xmath34 , and @xmath35 are dimensionless constants .
the @xmath36 and @xmath37 terms with the dimensionless constants correspond to the extra driving terms , i.e. , entropic - force terms .
the entropic - force terms are derived , taking into account the entropy and temperature on the horizon of the universe due to the information holographically stored there @xcite .
we can solve the two equations assuming the single - fluid - dominated universe @xcite .
note that we neglect high - order terms for quantum corrections , because we do not discuss the inflation of the early universe in the present paper .
( the dimensionless constants were expected to be bounded by @xmath38 and @xmath39 , and typical values for a better fitting were @xmath40 and @xmath41 @xcite .
it was argued that the extrinsic curvature at the surface was likely to result in something like @xmath42 and @xmath43 @xcite . ) as examined in ref .
@xcite , we can simplify the two modified friedmann equations , assuming a non - adiabatic - like expansion of the universe .
the simple modified friedmann and acceleration equations are summarized as @xmath44 @xmath45 equations ( [ eq : smfrw01s ] ) and ( [ eq : smfrw02s ] ) do not include @xmath37 terms .
in fact , easson _ et al . _
first proposed that the entropic - force terms of the modified acceleration equation are @xmath36 or @xmath46 ; i.e. , @xmath37 terms are not included in the entropic - force terms @xcite . in other words , eq .
( [ eq : smfrw02s ] ) is consistent with their original acceleration equation , as discussed in sec .
[ entropic - force from the bekenstein entropy ] . in the present paper , we select eqs .
( [ eq : smfrw01s ] ) and ( [ eq : smfrw02s ] ) as standard entropic - force models .
( note that it may be possible to neglect the entropic - force terms of the modified friedmann equation , i.e. , @xmath47 . in the standard entropic cosmology @xcite ,
the entropic - force terms are added not only with the acceleration equation but also with the friedmann equation , as in the case of the @xmath0cdm models . ) in ref .
@xcite , easson _ et al .
_ considered that the continuity equation was given by @xmath48 = 0 $ ] , assuming an adiabatic ( isentropic ) expansion . however , in entropic cosmology , since the entropy on the horizon is assumed , the entropy can increase during the evolution of the universe .
therefore , in a previous study , we derived the continuity equation from the first law of thermodynamics , taking into account a non - adiabatic - like process caused by the entropy and the temperature on the horizon @xcite . consequently , the modified continuity equation was written as @xmath49 because of the non - zero right - hand side , eq .
( [ eq : fluid_thermo(add ) ] ) was likely consistent with the continuity equation obtained from the modified friedmann and acceleration equations @xcite .
in the present study , we do not use eq .
( [ eq : fluid_thermo(add ) ] ) as the modified continuity equation .
instead , in the next section , we derive the modified continuity equation from an effective pressure for entropic cosmology .
we point out that a similar non - zero right - hand side of the continuity equation appears not only in bulk viscous cosmology but also in ` energy exchange cosmology ' , in which the transfer of energy between two fluids is assumed @xcite , e.g. , the interaction between matter and radiation @xcite , matter creation @xcite , interacting quintessence @xcite , the interaction between dark energy and dark matter @xcite , and dynamical vacuum energy @xcite . in particular , cosmological equations for dynamical vacuum energy models are very similar to those for entropic - force models . therefore , in the following paragraph , we discuss a similarity between the two models . instead of a cosmological constant
, a variable cosmological term @xmath50 is assumed in dynamical vacuum energy models , i.e. , @xmath50cdm models @xcite .
various @xmath50 terms ( e.g. , @xmath36 , @xmath51 , and constant terms ) have been examined in those works .
for example , see refs .
recently , @xmath36 and @xmath37 terms have been investigated in refs .
[ the @xmath36 and @xmath37 terms are the same as the entropic - force terms shown in eqs .
( [ eq : smfrw01 ] ) and ( [ eq : smfrw02 ] ) . ]
consequently , the influence of @xmath37 terms was found to be similar to that of @xmath36 terms . in other words , essentially new properties of the late universe
were not obtained from @xmath37 terms .
this implies that the @xmath37 terms can be neglected as if eqs .
( [ eq : smfrw01s ] ) and ( [ eq : smfrw02s ] ) were the approximated friedmann and acceleration equations , respectively .
in addition , in @xmath50cdm models , the continuity equation for matter ` @xmath52 ' is arranged as @xmath53 @xcite .
the continuity equation is equivalent to eq .
( [ eq : fluid_thermo(add ) ] ) , when @xmath50 is proportional to @xmath36 . in this way ,
cosmological equations for @xmath50cdm models are similar to those for entropic - force models .
accordingly , the cosmological equations and their solutions discussed in the present paper are similar to those for @xmath50cdm models , especially in the work of basilakos , plionis , and sol @xcite .
however , a theoretical background of the @xmath50cdm model is different from that of the entropic - force model .
finally , we discuss a fundamental problem of the standard entropic cosmology which includes @xmath36 and @xmath37 terms [ eqs .
( [ eq : smfrw01 ] ) and ( [ eq : smfrw02 ] ) ] . in fact ,
@xmath36 and @xmath37 terms of entropic - force terms can not describe a decelerating and accelerating universe predicted by the standard @xmath0cdm model .
this problem has been discussed in refs . @xcite and our previous paper @xcite . in particular , basilakos , polarski , and sol
have shown that not the @xmath36 and @xmath37 terms but extra constant terms are important for describing the decelerating and accelerating universe @xcite .
we emphasize that the standard entropic cosmology has such a fundamental problem , since entropic - force terms are usually considered to be @xmath36 and @xmath37 terms .
however , a generalized black - hole entropy @xcite is expected to solve the problem , as discussed later .
( it should be noted that the above problem does not occur in @xmath50cdm models , as examined in refs .
this is because the extra constant term is naturally obtained from an integral constant of the renormalization group equation for the vacuum energy density . for details ,
see a summarized review @xcite and a recent thorough review @xcite . )
in this section , in order to discuss entropic cosmology , we derive entropic - force terms from the bekenstein entropy and a generalized black - hole entropy proposed by tsallis and cirto @xcite . in sec .
[ entropic - force from the bekenstein entropy ] , we derive the standard entropic - force term from the bekenstein entropy , which is proportional to the surface area of a sphere with the hubble radius . moreover , using an effective description for pressure , we reformulate the entropic - force model . in sec .
[ entropic - force from a generalized entropy ] , we assume the generalized entropy proportional to the volume .
we derive an entropic - force term from the generalized entropy and propose a new entropic - force model . in the present paper ,
we use the hubble radius as the preferred screen , because the apparent horizon coincides with the hubble radius in the spatially flat universe @xcite .
in the standard entropic cosmology , the modified friedmann and acceleration equations include @xmath36 terms [ eqs .
( [ eq : smfrw01])([eq : smfrw02s ] ) ] , as entropic - force terms .
we derive the entropic - force term from the bekenstein entropy , according to the work of easson _ et al .
we also reformulate the entropic - force model , using an effective pressure for entropic cosmology .
for entropic cosmology , we assume that the hubble horizon has an approximate temperature @xmath54 and an associated entropy @xmath55 , where the hubble horizon ( radius ) @xmath56 is given by @xmath57 the temperature @xmath54 on the hubble horizon is given by @xmath58 where @xmath59 and @xmath60 are the boltzmann constant and the reduced planck constant , respectively .
the reduced planck constant is defined by @xmath61 , where @xmath62 is the planck constant .
the temperature considered here is obtained by multiplying the horizon temperature , @xmath63 , by @xmath64 @xcite . in the present study , @xmath64 is a non - negative free parameter on the order of @xmath65 , typically @xmath66 or @xmath67 .
a similar parameter for the screen temperature has been discussed in refs .
( note that the temperature on the horizon can be evaluated as @xmath68 , which is slightly lower than the temperature of our cosmic microwave background radiation , @xmath69 .
we use the temperature on the horizon , assuming thermal equilibrium states based on a single holographic screen @xcite . ) as an associated entropy on the hubble horizon , the bekenstein entropy @xmath55 is given as @xmath70 where @xmath71 is the surface area of the sphere with the hubble radius @xmath56 . as shown in eq .
( [ eq : sh(add ) ] ) , the bekenstein entropy is proportional to @xmath71 . substituting @xmath72 into eq .
( [ eq : sh(add ) ] ) and using @xmath73 as given in eq .
( [ eq : rh ] ) , we obtain @xmath74 where @xmath75 is a positive constant @xcite given by @xmath76 the entropic - force @xmath77 can be given by @xmath78 where the minus sign indicates the direction of increasing entropy or the screen corresponding to the horizon @xcite .
substituting eqs .
( [ eq : t0 ] ) and ( [ eq : sh(add ) ] ) into eq .
( [ eq : fr ] ) and using @xmath79 , we have the entropic - force as @xmath80 \notag \\ & = - \gamma \frac{c^{4}}{g } .
\label{eq : f(add)}\end{aligned}\ ] ] therefore , the pressure @xmath81 derived from the bekenstein entropy is given by @xmath82 the above derivation is based on the original idea of easson _
et al_. @xcite .
note that eq .
( [ eq : p(add ) ] ) includes @xmath64 used in this study , where @xmath64 is a free parameter for the temperature . in the following , in order to reformulate entropic cosmology , we consider an effective description similar to bulk viscous cosmology .
in other words , we assume an effective pressure for entropic cosmology
. we will discuss the reason for this in sec .
[ entropic - force from a generalized entropy ] .
we assume that the effective pressure @xmath14 based on the bekenstein entropy is given by @xmath83 using @xmath14 , the acceleration equation can be written as @xmath84 accordingly , substituting eq .
( [ eq : p_eff ] ) into eq .
( [ eq : frw2p_eff ] ) , we have @xmath85 the @xmath86 term is the entropic - force term , which can explain the accelerated expansion of the universe . when @xmath87 , eq .
( [ eq : frw2_p(add ) ] ) corresponds to the modified acceleration equation derived by easson _
_ @xcite .
we now examine the continuity equation for entropic cosmology . using the effective pressure @xmath14
, the continuity equation is expected to be given by @xmath88 substituting eq .
( [ eq : p_eff ] ) into eq .
( [ eq : fluid_eff ] ) and rearranging , we have @xmath89 this is the modified continuity equation obtained from the effective pressure for entropic cosmology . in the present paper ,
we consider eq .
( [ eq : fluid_(add ) ] ) as the modified continuity equation for entropic cosmology derived from the bekenstein entropy .
[ note that eqs .
( [ eq : fluid_(add ) ] ) and ( [ eq : fluid_thermo(add ) ] ) are consistent with each other when @xmath90 .
the universe for @xmath90 corresponds to the empty universe , as discussed in sec .
[ combined model ] . ]
we can determine two dimensionless constants @xmath91 and @xmath92 included in the modified friedmann and acceleration equations [ eqs .
( [ eq : smfrw01s ] ) and ( [ eq : smfrw02s ] ) ] using two continuity equations .
the first continuity equation is eq .
( [ eq : fluid_(add ) ] ) , whereas the second continuity equation can be derived from the modified friedmann and acceleration equations .
in order to obtain the second continuity equation , we use the generalized continuity equation , i.e. , eq .
( [ eq : drho_general0 ] ) . comparing [ eqs .
( [ eq : general_frw01_f ] ) and ( [ eq : general_frw02_g ] ) ] with [ eqs .
( [ eq : smfrw01s ] ) and ( [ eq : smfrw02s ] ) ] , we can set general functions as @xmath93 and @xmath94 .
substituting these functions into eq .
( [ eq : drho_general0 ] ) and rearranging , we have the second continuity equation , which is given by @xmath95 we expect that the two modified continuity equations , i.e. , eqs .
( [ eq : fluid_(add ) ] ) and ( [ eq : drho_2nd_(add ) ] ) , are consistent with each other .
consequently , we obtain @xmath91 and @xmath92 as @xmath96 therefore , we can neglect the entropic - force term @xmath97 of the modified friedmann equation shown in eq .
( [ eq : smfrw01s ] ) .
this is because , in the present paper , we assume the effective pressure for entropic cosmology .
as mentioned previously , we consider @xmath64 as a free parameter for the temperature .
moreover , a free parameter for the entropy may be required for calculating @xmath98 .
however , we do not use the free parameter for the entropy , because we assume that the bekenstein entropy is given by @xmath99 [ eq . [ eq : sh(add ) ] ] . in sec .
[ derivation of entropic - force from the bekenstein entropy ] , we derived an entropic - force term from the bekenstein entropy and obtained an entropic - force model .
consequently , the modified friedmann , acceleration , and continuity equations are summarized as @xmath100 @xmath101 and @xmath102 where @xmath91 and @xmath92 are given by @xmath103 note that we leave @xmath104 in eq .
( [ eq : smfrw01s_summ ] ) in order to discuss a combined model later .
thus far , we have considered the bekenstein entropy to discuss the standard entropic cosmology . instead of the bekenstein entropy ,
we now assume a generalized black - hole entropy proposed by tsallis and cirto @xcite .
recently , tsallis and cirto @xcite examined a black - hole entropy using appropriate nonadditive generalizations for @xmath105-dimensional systems and suggested a generalized black - hole entropy . in the following ,
we introduce the generalized black - hole entropy , according to the work of tsallis and cirto @xcite . in their study ,
a nonadditive entropy ( for a set of @xmath106 discrete states ) is defined by @xmath107 where @xmath108 is a probability distribution @xcite .
when @xmath109 , @xmath110 recovers the boltzmann
gibbs entropy given by @xmath111 if we compose two probabilistically independent subsystems @xmath112 and @xmath113 , the boltzmann
gibbs entropy @xmath114 is additive : @xmath115 however , when @xmath116 , @xmath110 is nonadditive : @xmath117 this is because , for @xmath118 , @xmath119 is given by @xmath120 ^{1/\delta } + \left [ \frac { s_{\delta}(b ) } { k_{b } } \right ] ^{1/\delta } \right ) ^{\delta } .\ ] ] in ref .
@xcite , tsallis and cirto demonstrated that a generalized black - hole entropy can be written as @xmath121 where the bekenstein entropy @xmath122 is given by @xmath123 and @xmath124 is the event horizon area of a black hole .
for details , see ref .
@xcite .
we now apply the generalized black - hole entropy to an entropy for entropic cosmology . to this end , substituting eq .
( [ eq : sbh ] ) into eq .
( [ eq : s - sbh ] ) , replacing @xmath124 with @xmath71 , and rearranging , the entropy @xmath55 on the hubble horizon can be evaluated as @xmath125 where @xmath71 is the surface area of the sphere with the hubble radius @xmath56 , and , therefore , @xmath79 .
because of nonadditive generalizations , the generalized entropy is proportional to @xmath126 , i.e. , volume .
accordingly , from eqs .
( [ eq : sbh ] ) and ( [ eq : sprop ] ) , we assume a generalized entropy @xmath127 given by @xmath128 where @xmath129 is a non - negative free - parameter .
note that @xmath129 is a dimensional constant .
we hereinafter refer to @xmath127 as the generalized entropy on the hubble horizon . substituting @xmath73 [ eq . ( [ eq : rh ] ) ] into eq .
( [ eq : s(nonadd ) ] ) , we obtain @xmath130 where @xmath75 is @xmath131 [ eq .
( [ eq : k - def ] ) ] . in order to calculate an entropic - force
, we use the generalized entropy [ eq .
( [ eq : s(nonadd ) ] ) ] and the temperature @xmath54 on the horizon [ eq . ( [ eq : t0 ] ) ] . substituting eqs .
( [ eq : t0 ] ) and ( [ eq : s(nonadd ) ] ) into eq .
( [ eq : fr ] ) and using @xmath73 , we obtain the entropic - force as @xmath132 \notag \\ % & = - \gamma \frac{c^{4}}{g } \left ( \frac{3}{2 } \zeta r_{h } \right ) = - \gamma \frac{c^{4}}{g } \left ( \frac{3 c \zeta}{2 } \frac{1}{h } \right ) .
\label{eq : f(nonadd)}\end{aligned}\ ] ] therefore , the pressure @xmath133 derived from the generalized entropy is given as @xmath134 in this equation , @xmath135 is shown separately in order to clarify the difference between eqs .
( [ eq : p(add ) ] ) and ( [ eq : p(nonadd ) ] ) .
similarly , @xmath135 is shown separately , in the following . in the present study , we assume an effective pressure @xmath14 for entropic cosmology . using eq .
( [ eq : p(nonadd ) ] ) , the effective pressure @xmath14 based on the generalized entropy is given by @xmath136 substituting eq .
( [ eq : p_eff(nonadd ) ] ) into eq . ( [ eq : frw2p_eff ] ) , we have @xmath137 this equation is the modified acceleration equation derived from the generalized entropy .
the @xmath138 term on the right - hand side corresponds to an extra driving term to explain the accelerated expansion of the universe .
the extra driving term is found not to be the @xmath139 term but rather to be the @xmath51 term , unlike in the case of the standard entropic cosmology .
the continuity equation is expected to be given by @xmath140 substituting eq .
( [ eq : p_eff(nonadd ) ] ) into eq .
( [ eq : fluid_eff2 ] ) and rearranging , we have @xmath141 in this paper , we consider eq .
( [ eq : fluid_(nonadd ) ] ) as the modified continuity equation based on the generalized entropy @xcite .
interestingly , the effective pressure , eq .
( [ eq : p_eff(nonadd ) ] ) , is similar to eq .
( [ eq : bulk_p ] ) , i.e. , @xmath142 , for bulk viscous cosmology . therefore , eqs .
( [ eq : frw2_p(nonadd ) ] ) and ( [ eq : fluid_(nonadd ) ] ) are also similar to eqs .
( [ eq : bulk_frw2 ] ) and ( [ eq : bulk_fluid1 ] ) for bulk viscous cosmology .
this is probably because the generalized entropy assumed here is proportional not to the surface area but to the volume .
however , the similarity may be interpreted as a sign that the generalized entropy behaves as if it were a classical entropy generated by bulk viscous stresses . in other words ,
the generalized entropy may be related to a bulk viscosity of cosmological fluids through a holographic screen .
in fact , the bulk viscosity is usually the only thing that can generate an entropy in the homogeneous and isotropic universe .
accordingly , this interpretation may help to explain the origin of the bulk viscosity of cosmological fluids . as an alternative interpretation
, the bulk viscosity may be derived from an extra entropy that is proportional to the volume of the universe . in sec .
[ derivation of entropic - force from a generalized entropy ] , @xmath51 terms are derived from a generalized entropy on the horizon .
accordingly , we expect that the modified friedmann and acceleration equations are summarized , respectively , as @xmath143 and @xmath144 where @xmath145 and @xmath146 are defined by @xmath147 the two coefficients @xmath148 and @xmath149 are dimensionless constants , and @xmath150 is the present value of the hubble parameter .
as shown in eq .
( [ eq : fluid_(nonadd ) ] ) , the modified continuity equation is given by @xmath151 we can determine @xmath145 and @xmath146 , which are included in eqs .
( [ eq : mfrw01(nonadd ) ] ) and ( [ eq : mfrw02(nonadd ) ] ) , from two continuity equations . in this case
, the first continuity equation is eq .
( [ eq : fluid_(nonadd)2 ] ) .
the second continuity equation ( derived from the modified friedmann and acceleration equations ) is calculated from the generalized continuity equation , i.e. , eq .
( [ eq : drho_general0 ] ) . comparing [ eqs .
( [ eq : general_frw01_f ] ) and ( [ eq : general_frw02_g ] ) ] with [ eqs .
( [ eq : mfrw01(nonadd ) ] ) and ( [ eq : mfrw02(nonadd ) ] ) ] , we can set general functions as @xmath152 and @xmath153 .
therefore , substituting these functions into eq .
( [ eq : drho_general0 ] ) and rearranging , we have the second continuity equation given by @xmath154 since eqs .
( [ eq : fluid_(nonadd)2 ] ) and ( [ eq : drho_2nd ] ) are expected to be consistent with each other , we obtain @xmath145 and @xmath146 as @xmath155 accordingly , the dimensionless constants @xmath156 and @xmath157 are given as @xmath158 as expected , we can neglect the entropic - force term @xmath159 of the modified friedmann equation , eq .
( [ eq : mfrw01(nonadd ) ] ) .
this is because , in the above discussion and in the present study , we assume an effective pressure for entropic cosmology . note that we leave @xmath159 in eq .
( [ eq : mfrw01(nonadd ) ] ) in order to discuss a combined model in the next section .
.entropic - force terms @xmath160 for the modified acceleration equation .
the bekenstein and generalized columns indicate the information for the entropic - force models derived from the bekenstein and generalized entropies , respectively . [
cols="<,<,<,<,<",options="header " , ] + for the two entropic - force models , we consider the matter - dominated universe given by @xmath161 in sec .
[ combined model ] , we obtained solutions for the combined model using four dimensionless constants , i.e. , @xmath32 , @xmath148 , @xmath34 , and @xmath149 . in other words , by setting the four constants , we can obtain the solution of each model .
the four constants for each model are listed in table [ tab - parameter ] . for the bekenstein entropic - force model , @xmath32 , @xmath148 , and @xmath149
are @xmath162 , as described in sec .
[ entropic - force model for the bekenstein entropy ] ( and shown in table [ tab - parameter ] ) .
therefore , substituting these values and @xmath163 into eq .
( [ eq : c3 ] ) , we have @xmath164 . on the other hand , @xmath92 is set to be @xmath165 , in order to obtain @xmath166 .
this is because , as reported in ref .
@xcite , when we consider the equation @xmath167 , the solution with @xmath166 was consistent with the observed supernova data . here
, @xmath166 was calculated from the hawking temperature description @xcite , without using a fitting with the supernova data @xcite . of course , we accept that @xmath92 should be a free parameter .
for example , @xmath168 is approximated as @xmath169 , if a fitting method discussed later herein is used to determine @xmath168 .
the value of @xmath169 is close to @xmath165 , or , more specifically , to @xmath170 . in ref .
@xcite , @xmath171 was calculated from the anticipated surface term order @xcite .
we have confirmed that the properties of the universe for @xmath171 or @xmath172 are similar to the properties for @xmath173 .
accordingly , in the present paper , we select @xmath165 as @xmath168 for the bekenstein entropic - force model .
in contrast , for the generalized entropic - force model , @xmath32 , @xmath148 , and @xmath34 are @xmath162 , as examined in sec .
[ entropic - force model for a generalized entropy ] ( and shown in table [ tab - parameter ] ) . substituting these values and @xmath163 into eq .
( [ eq : c1 ] ) , we obtain @xmath174 .
( in the standard cosmology , the universe for @xmath174 corresponds to the matter - dominated universe @xcite . ) on the other hand , @xmath157 is a free parameter .
therefore , @xmath157 is determined through fitting with a fine - tuned standard @xmath0cdm model based on the planck 2013 results @xcite . to this end
, we use the luminosity distance .
for the two entropic - force models , we calculate the luminosity distance @xmath175 from eqs .
( [ eqa : dlc1c3 ] ) and ( [ eq : dlc1(c3=0 ) ] ) and table [ tab - parameter ] . for the standard @xmath0cdm model ,
the luminosity distance of the spatially flat universe is given as @xmath176^{-1/2 } , \label{eq : dl(cdm)}\end{aligned}\ ] ] where @xmath177 and @xmath178 @xcite . @xmath179 and @xmath180 represent the density parameters for matter and @xmath0 , respectively .
moreover , @xmath181 represents the critical density , and @xmath182 is the density of matter , which includes baryon and dark matter .
the universe in which @xmath183 is a fine - tuned standard @xmath0cdm model , which takes into account the recent planck 2013 best fit values @xcite .
note that we assume @xmath184 and neglect the density parameter @xmath185 for the radiation @xcite .
in the present study , @xmath157 for the generalized entropic - force model is determined through fitting with the fine - tuned standard @xmath0cdm model .
in fact , substituting @xmath186 into eq .
( [ eq : c3 ] ) , @xmath187 is equal to @xmath157 .
accordingly , instead of @xmath157 , we determine @xmath187 through fitting with the fine - tuned standard @xmath0cdm model , minimizing the following function @xcite given by @xmath188^{2 } } , \label{eq : min}\ ] ] where @xmath189 and @xmath190 are the luminosity distances for the fine - tuned standard @xmath0cdm model and the combined ( entropic - force ) model , respectively . for the generalized entropic - force model , @xmath168 is set to be @xmath191 . in order to determine @xmath187 ,
the redshift is varied from @xmath192 to @xmath193 in increments of @xmath194 .
in other words , @xmath195 is given by @xmath196 , where @xmath197 is varied from @xmath162 to @xmath198 . through fitting , @xmath187 is approximately determined to be @xmath199 , as shown in table [ tab - parameter ] .
since @xmath200 , @xmath157 is set to be @xmath199 .
( it may be possible to consider @xmath168 and @xmath187 as free - parameters simultaneously . in this case ,
@xmath168 and @xmath187 are approximately @xmath201 and @xmath202 , respectively .
we do not discuss such conditions in the present paper . )
[ t ] we now observe the luminosity distance @xmath175 . as shown in fig .
[ fig - dl - z ] , the two entropic - force models agree well with supernova data points and the fine - tuned standard @xmath0cdm model . for the generalized entropic - force model , @xmath203 is determined through fitting with the fine - tuned standard @xmath0cdm model , as mentioned previously .
therefore , it is not so surprising that the generalized entropic - force model agrees well with the standard @xmath0cdm model . for the bekenstein entropic - force model , @xmath204 is selected , where the value of @xmath205 was calculated based on the description of the hawking temperature @xcite .
however , it is demonstrated that the two entropic - force models can describe the present accelerating universe without adding the cosmological constant or dark energy . as shown in fig .
[ fig - dl - z ] , the difference between the two entropic - force models is not clear when we observe the luminosity distance at late times .
therefore , we examine the time evolutions of the normalized scale factor @xmath206 for longer time ranges . for the two entropic - force models , we calculate @xmath206 from eqs .
( [ eq : a - t ] ) and ( [ eq : a - t(c3=0 ) ] ) and table [ tab - parameter ] . for the standard @xmath0cdm model
, we use the following equation @xcite : @xmath207 we integrate eq .
( [ eq : a_cdm ] ) numerically in order to obtain the time evolution of the normalized scale factor @xmath208 . in the present study , for a spatially flat universe ,
we assume @xmath209 and neglect the influence of radiation , i.e. , @xmath210 @xcite .
[ t ] figure [ fig - a - t ] shows the time evolutions of the normalized scale factor @xmath208 . for the bekenstein entropic - force model ,
@xmath208 increases uniformly , because @xmath208 is given by @xmath211 [ eq . ( [ eq : a - t(c3=0 ) ] ) ] when @xmath212 and @xmath213 .
in fact , a temporal deceleration parameter @xmath214 is given by @xmath215 , as shown in eq .
( [ eq : ddota_a_c1c3h ] ) .
accordingly , @xmath214 is constant , when @xmath168 is constant and @xmath216 .
in contrast , for the generalized entropic - force model , @xmath214 can vary when @xmath217 , even if @xmath168 is constant .
therefore , we can expect that @xmath208 for the generalized model does not increase uniformly , unlike that for the bekenstein model . in order to observe this phenomenon ,
we focus on the generalized entropic - force model . as shown in fig .
[ fig - a - t ] , the increase in @xmath208 tends to become gradually slower for @xmath218 .
however , for @xmath219 , the increase in @xmath208 tends to become gradually faster . in other words ,
the generalized entropic - force model predicts a decelerated and accelerated expansion of the universe , which is similar to @xmath208 for the fine - tuned standard @xmath0cdm model .
interestingly , the generalized entropic - force model can describe the decelerated and accelerated expansion , without adding a constant term such as a cosmological constant . as discussed in sec .
[ standard entropic - force models ] , the standard entropic cosmology can not describe a decelerating and accelerating universe @xcite .
however , the standard entropic cosmology can describe the decelerating and accelerating universe if the modified friedmann and acceleration equations include extra constant terms , as shown in ref .
@xcite . in the present study
, we do not discuss such entropic - force models that include constant terms , because the origin of the constant terms is not clear .
( in @xmath50cdm models , the constant terms are naturally obtained from an integral constant of the renormalization group equation for the vacuum energy density @xcite . ) of course , bulk viscous cosmology can describe the decelerated and accelerated expansion @xcite because the modified acceleration equation includes the @xmath51 term as an extra driving term .
however , in the generalized entropic - force model , the @xmath51 term is derived from the generalized entropy , without using a bulk viscosity of cosmological fluids .
we now examine a transition time @xmath220 @xcite , which represents the time of transition from the decelerated expansion epoch to the accelerated expansion epoch . calculating the second derivative of eq .
( [ eq : a - t ] ) , equating to @xmath162 , and solving the result , we obtain the transition time for the combined model as @xmath221 substituting @xmath222 and @xmath223 into eq .
( [ eq : trans - time ] ) , we have @xmath224 , for the generalized entropic - force model .
note that @xmath220 does not exist for the bekenstein entropic - force model , because @xmath206 increases uniformly when @xmath225 . as shown in fig .
[ fig - a - t ] , the two entropic - force models are consistent with the fine - tuned standard @xmath0cdm model at the present time . in order to discuss this more closely ,
we examine the deceleration parameter @xmath226 .
for example , we can estimate @xmath226 for the fine - tuned standard @xmath0cdm model as @xmath227 , from @xmath228 @xcite , where @xmath229 is set to be @xmath230 .
for the two entropic - force models , @xmath226 is calculated from @xmath231 , as shown in eq .
( [ eq : q0c1c3 ] ) .
consequently , for the bekenstein and generalized entropic - force models , we have @xmath232 and @xmath233 , respectively .
therefore , at the present time , the acceleration for the two entropic - force models is slightly slower than that for the fine - tuned standard @xmath0cdm model .
[ t ] finally , we observe the evolutions of the bekenstein and generalized entropies . in order to calculate the entropy , @xmath234 [ eq . ( [ eq : sh(add)2 ] ) ]
is used for the bekenstein entropic - force model , whereas @xmath235 [ eq .
( [ eq : s(nonadd)2 ] ) ] is used for the generalized entropic - force model . in the following ,
the two entropies are normalized as @xmath236 and @xmath237 , respectively .
we numerically calculate @xmath238 and @xmath239 of the fine - tuned standard @xmath0cdm model , in order to compare the two entropic - force models with the @xmath0cdm model .
this is because @xmath238 and @xmath239 correspond to @xmath240 and @xmath241 , respectively .
figure [ fig - s - a ] shows the evolutions of the bekenstein and generalized entropies .
we first focus on the entropies for the fine - tuned standard @xmath0cdm model . for @xmath242 ,
the entropies for the standard @xmath0cdm model increase rapidly , whereas , for @xmath243 , the increase in the entropies tends to become gradually slower .
since the standard @xmath0cdm model has been widely examined , similar results have been reported @xcite .
we now examine the two entropic - force models . for @xmath244 ,
the entropy for the bekenstein entropic - force model is consistent with @xmath240 for the standard @xmath0cdm model .
similarly , for @xmath244 , the entropy for the generalized entropic - force model agrees well with @xmath241 for the @xmath0cdm model .
this indicates that the evolutions of the late universe for the two entropic - force models are consistent with the evolution for the @xmath0cdm model .
we have confirmed this from figs .
[ fig - dl - z ] and [ fig - a - t ] as well .
however , for @xmath243 , the entropy for the bekenstein entropic - force model increases uniformly , whereas the increase in the entropy for the generalized entropic - force model tends to become gradually slower [ fig .
[ fig - s - a ] ] .
therefore , for @xmath245 , the entropy for the bekenstein entropic - force model is clearly different from @xmath240 for the standard @xmath0cdm model . on the other hand , for @xmath245 ,
the entropy for the generalized entropic - force model increases more rapidly than @xmath241 for the @xmath0cdm model .
however , the increase in the entropy for the generalized entropic - force model tends to become gradually slower .
accordingly , the evolution of the entropy for the generalized entropic - force model is similar to the evolution for the @xmath0cdm model .
this is because the generalized entropic - force model can describe a decelerating and accelerating universe .
in order to examine entropic cosmology , we have derived extra entropic - force terms not only from the bekenstein entropy but also from a generalized entropy , assuming that the horizon of the universe has an entropy and a temperature .
because of nonadditive generalizations , the generalized entropy is proportional to volume , unlike for the bekenstein entropy .
consequently , as entropic - force terms , @xmath51 terms are derived from the generalized entropy , whereas @xmath139 terms are derived from the bekenstein entropy .
interestingly , the @xmath51 term is similar to an extra driving term for bulk viscous cosmology .
this similarity may be interpreted as a sign that the generalized entropy behaves as if it were a classical entropy generated by bulk viscous stresses .
this is probably because the generalized entropy considered here is proportional not to area but rather to volume .
however , as one interpretation , the generalized entropy may be related to a bulk viscosity of cosmological fluids , through a holographic screen .
this interpretation does not contradict the following description ; i.e. , the bulk viscosity is considered to be the only thing that can generate an entropy in the homogeneous and isotropic universe . in the present study
, we assume an entropy on the horizon of the universe , using the holographic principle .
however , if we assume another entropy that is proportional to the volume of the universe , its entropic - force can likely explain the accelerated expansion of the universe . in other words , the concept of the entropic - force may be applied to the bulk of the universe , without using the holographic principle . in the present paper , based on effective pressure ,
we have formulated the modified friedmann , acceleration , and continuity equations for two entropic - force models derived from the bekenstein and generalized entropies .
the friedmann equation is shown not to include the entropic - force term , whereas the continuity equation has a non - zero right - hand side . in order to examine the properties of the universe in the two entropic - force models ,
we have considered a combined model , in which the friedmann and acceleration equations include both @xmath139 and @xmath51 terms simultaneously as the entropic - force terms .
we have obtained solutions of the combined model , assuming a homogeneous , isotropic , spatially flat universe , focusing on the single - fluid - dominated universe .
we have confirmed that the two entropic - force models can describe the present accelerating universe , without adding the cosmological constant or dark energy .
the bekenstein entropic - force model is found to predict a uniformly accelerating universe , whereas the generalized entropic - force model predicts a decelerating and accelerating universe , as in the case for a fine - tuned standard @xmath0cdm model .
similar properties of the universe with the extra driving terms , i.e. , @xmath139 and @xmath51 terms , have been examined and discussed in not only bulk viscous cosmology but also @xmath50cdm models with a variable cosmological term @xmath50 . however , in the present study , the extra driving terms are derived from the bekenstein and generalized entropies , without using a bulk viscosity of cosmological fluids and dark energy .
the present study has revealed fundamental properties of the expanding universe in entropic cosmologies based on the bekenstein and generalized entropies . in particular , the entropic - force derived from the generalized entropy will help in discussing the accelerating universe in other cosmological models .
for example , when an entropy is assumed to be proportional to @xmath246 , the obtained entropic - force term is a constant , as if it were a cosmological constant .
although such an entropy has probably not yet been suggested , we can discuss cosmological models from various entropic - force viewpoints . in the present study , in order to examine entropic cosmology , we assume an entropy on the horizon of the universe . at least in principle , it is possible to assume an entropy that is proportional to the volume of the universe in order to discuss entropic - force models . in the present study , we have not examined cosmological fluctuations , since we have focused on background evolutions of the late universe .
however , through linear density perturbations , it has been discussed that bulk viscous models are difficult to reconcile with astronomical observations of structure formations @xcite . in addition , as shown in ref .
@xcite , @xmath50cdm models similar to the combined model are not consistent with the structure formation data .
( the structure formation in @xmath50cdm models has been examined in detail @xcite . )
the previous works imply that the present entropic - force models are difficult to reconcile with the structure formation data .
we expect that another type of entropy , e.g. , @xmath247 , helps to overcome the difficulty .
this task is left for future research .
s. perlmutter _
et al . _ , nature ( london ) * 391 * , 51 ( 1998 ) ; astrophys .
j. * 517 * , 565 ( 1999 ) .
a. g. riess _
et al . _ ,
j. * 116 * , 1009 ( 1998 ) ; astrophys .
j. * 607 * , 665 ( 2004 ) .
a. g. riess _
et al . _ ,
j. * 659 * , 98 ( 2007 ) ; http://braeburn.pha.jhu.edu/~ariess/r06/sn_sample .
m. tegmark _
_ , astrophys .
j. * 606 * , 702 ( 2004 ) ; d. n. spergel _
et al . _ ,
* 170 * , 377 ( 2007 ) ; w. m. wood - vasey _ et al .
_ , astrophys .
j. * 666 * , 694 ( 2007 ) ; m. kowalski _
et al . _ ,
j. * 686 * , 749 ( 2008 ) ; m. hicken , w. m. wood - vasey , s. blondin , p. challis , s. jha , p. l. kelly , a. rest , and r. p. kirshner , astrophys . j. * 700 * , 1097 ( 2009 ) ; e. komatsu _ et al .
_ , astrophys
. j. suppl .
ser . * 180 * , 330 ( 2009 ) .
n. jarosik _
et al . _ ,
. ser . * 192 * , 14 ( 2011 ) ; e. komatsu _ et al .
_ , astrophys .
* 192 * , 18 ( 2011 ) .
p. a. r. ade _ et al .
_ , arxiv:1303.5076v1 [ astro-ph.co ] .
b. ryden , _ introduction to cosmology _
( addison - wesley , reading , ma , 2002 ) . s. weinberg , _ cosmology _ ( oxford university press , new york , 2008 ) .
g. f. r. ellis , r. maartens , and m. a. h. maccallum , _
relativistic cosmology _ ( cambridge university press , cambridge , england , 2012 ) .
l. miao , l. xiao - dong , w. shuang , and w. yi , commun .
. phys . * 56 * , 525 ( 2011 ) .
k. bamba , s. capozziello , s. nojiri , and s. d. odintsov , astrophys .
space sci . * 342 * , 155 ( 2012 ) .
d. a. easson , p. h. frampton , and g. f. smoot , phys .
b * 696 * , 273 ( 2011 ) .
d. a. easson , p. h. frampton , and g. f. smoot , int .
a * 27 * , 1250066 ( 2012 ) .
t. s. koivisto , d. f. mota , and m. zumalacrregui , j. cosmol .
astropart . phys .
02 ( 2011 ) 027 .
n. komatsu , and s. kimura , phys .
d * 87 * , 043531 ( 2013 ) . n. komatsu , in proceedings of the 12th asia pacific physics conference , appc12 , chiba , japan , 2013 ( _ to be published _ ) . y. f. cai , j. liu , and h. li , phys .
b * 690 * , 213 ( 2010 ) .
y. f. cai and e. n. saridakis , phys .
b * 697 * , 280 ( 2011 ) .
t. qiu and e. n. saridakis , phys .
d * 85 * , 043504 ( 2012 ) .
r. casadio and a. gruppuso , phys .
d * 84 * , 023503 ( 2011 ) ; u. h. danielsson , arxiv:1003.0668v1 [ hep - th ] ; y. s. myung , astrophys .
space sci . *
335 * , 553 ( 2011 ) ; f. e. m. costa , j. a. s. lima , and f. a. oliveira , arxiv:1204.1864v1 [ astro-ph.co ] .
s. basilakos , d. polarski , and j. sol , phys .
d * 86 * , 043010 ( 2012 ) .
j. binney and s. tremaine , _ galactic dynamics _
( princeton university press , princeton , nj , 1987 ) .
c. tsallis , _ introduction to nonextensive statistical mechanics : approaching a complex world _ ( springer , new york , 2009 ) . j. l. lebowitz and e. h. lieb , phys .
* 22 * , 631 ( 1969 ) ; w. thirring , z. phys .
* 235 * , 339 ( 1970 ) ; m. k. h. kiessling , j. stat . phys .
* 55 * , 203 ( 1989 ) ; t. padmanabhan , phys . rep . * 188 * , 285 ( 1990 ) ; h. a. posch and w. thirring , phys .
lett . * 95 * , 251101 ( 2005 ) . c. tsallis , j. stat
. phys . * 52 * , 479 ( 1988 ) .
a. renyi , _ probability theory _
( north - holland , amsterdam , 1970 ) .
a. plastino and a. r. plastino , phys .
a * 174 * , 384 ( 1993 ) .
s. abe , phys .
a * 263 * , 424 ( 1999 ) .
d. f. torres , h. vucetich , and a. plastino , phys .
* 79 * , 1588 ( 1997 ) .
r. s. mendes and c. tsallis , phys .
a * 285 * , 273 ( 2001 ) .
v. latora , a. rapisarda , and c. tsallis , phys .
e * 64 * , 056134 ( 2001 ) .
p. h. chavanis , astron . astrophys . * 386 * , 732 ( 2002 ) .
a. taruya and m. sakagami , phys .
lett . * 90 * , 181101 ( 2003 ) .
a. nakamichi and m. morikawa , physica a * 341 * , 215 ( 2004 ) .
a. taruya and m. sakagami , mon . not .
. soc . * 364 * , 990 ( 2005 ) .
b. liu and j. goree , phys .
lett . * 100 * , 055003 ( 2008 ) .
n. komatsu , s. kimura , and t. kiwata , phys .
e * 80 * , 041107 ( 2009 ) ; j. phys .
: conf . ser . *
201 * , 012009 ( 2010 ) .
n. komatsu , t. kiwata , and s. kimura , phys .
e * 82 * , 021118 ( 2010 ) ; phys .
e * 85 * , 021132 ( 2012 ) .
k. karami , a. sheykhi , n. sahraei , and s. ghaffari , europhys .
lett . * 93 * , 29002 ( 2011 ) ; a. sheykhi and s. h. hendi , phys .
d * 84 * , 044023 ( 2011 ) ; a. sheykhi , phys . rev .
d * 81 * , 104011 ( 2010 ) . h. m. sadjadi and m. jamil , europhys .
* 92 * , 69001 ( 2010 ) .
s. weinberg , _ gravitation and cosmology _
( john wiley & sons , new york , 1972 ) .
g. l. murphy , phys .
d * 8 * , 4231 ( 1973 ) . j. d. barrow , phys .
b * 180 * , 335 ( 1986 ) .
j. d. barrow , nucl .
b310 * , 743 ( 1988 ) .
p. c. w. davies , class .
quantum grav . * 4 * , l225 ( 1987 ) ; ann .
henri poincar * 49 * , 297 ( 1988 ) .
j. a. s. lima , r. portugal , and i. waga , phys .
d * 37 * , 2755 ( 1988 ) .
i. brevik , phys .
d * 65 * , 127302 ( 2002 ) .
i. brevik and o. gorbunova , gen .
. gravit . * 37 * , 2039 ( 2005 ) .
s. nojiri and s. d. odintsov , phys .
d * 72 * , 023003 ( 2005 ) .
j. ren and x .- h .
meng , phys .
b * 633*,1 ( 2006 ) . j. c. fabris , s. v. b. goncalves , and r. de s ribeiro , gen .
. gravit . * 38 * , 495 ( 2006 ) .
r. colistete , jr . ,
j. c. fabris , j. tossa , and w. zimdahl , phys .
d * 76 * , 103516 ( 2007 ) .
b. li and j. d. barrow , phys . rev .
d * 79 * , 103521 ( 2009 ) .
meng and x. dou , commun .
. phys . * 52 * , 377 , ( 2009 ) ; arxiv:0812.4904v1 [ astro - ph ] .
a. avelino and u. nucamendi , j. cosmol .
phys . 04 ( 2009 ) 006
. w. s. hipolito - ricaldi , h. e. s. velten , and w. zimdahl , j. cosmol .
astropart . phys .
06 ( 2009 ) 016 .
a. avelino and u. nucamendi , j. cosmol .
phys . 08 (
2010 ) 009 .
o. f. piattella , j. c. fabris , and w. zimdahl , j. cosmol .
phys . 05 ( 2011 ) 029 .
x. dou and x .- h .
meng , advances in astronomy * 2011 * , 829340 ( 2011 ) .
o. pujols , i. sawicki , and a. vikman , j. high energy phys .
11 ( 2011 ) 156 .
j. d. barrow and t. clifton , phys .
d * 73 * , 103520 ( 2006 ) .
w. davidson , mon . not .
r. astron .
soc . * 124 * , 79 ( 1962 ) ; m. szydlowski , phys . lett .
b * 632 * , 1 ( 2006 ) .
j. a. s. lima , a. s. m. germano , and l. r. w. abramo , phys .
d * 53 * , 4287 ( 1996 ) ; j. a. s. lima , s. basilakos , and f. e. m. costa , phys .
d * 86 * , 103534 ( 2012 ) .
l. amendola , phys .
d * 62 * , 043511 ( 2000 ) ; w. zimdahl , d. pavn , and l. p. chimento , phys .
b * 521 * , 133 ( 2001 ) .
b. wang , y. gong , and e. abdalla , phys .
b * 624 * , 141 ( 2005 ) ; b. wang , c .- y .
lin , d. pavn , and e. abdalla , phys .
b * 662 * , 1 ( 2008 ) .
k. freese , f. c. adams , j. a. frieman , and e. mottola , nucl .
b287 * , 797 ( 1987 ) ; h. a. borges and s. carneiro , gen .
. gravit . * 37 * , 1385 ( 2005 ) ; s. carneiro , c. pigozzo , h. a. borges , and j. s. alcaniz , phys .
d * 74 * , 023532 ( 2006 ) ; c. pigozzo , m. a. dantas , s. carneiro , and j. s. alcaniz , j. cosmol .
phys . 08 (
2011 ) 022 ; j. s. alcaniz , h. a. borges , s. carneiro , j. c. fabris , c. pigozzo , and w. zimdahl , phys . lett .
b * 716 * , 165 ( 2012 ) ; h. fritzsch , and sol , class .
quantum grav .
* 29 * 215002 , ( 2012 ) ; j. p. mimoso and d. pavn , phys .
d * 87 * , 047302 ( 2013 ) ; j. a. s. lima , s. basilakos , and j. sol , mon . not .
. soc . * 431 * , 923 ( 2013 ) .
i. l. shapiro and j. sol , j. high energy phys .
02 ( 2002 ) 006
. i. l. shapiro , j. sol , c. espaa - bonet , and p. ruiz - lapuente , phys .
b * 574 * , 149 ( 2003 ) .
c. espaa - bonet , p. ruiz - lapuente , i. l. shapiro , and j. sol , j. cosmol .
. phys . * 02 * , 006 ( 2004 ) .
s. basilakos , m. plionis , and j. sol , phys .
d * 80 * , 083511 ( 2009 ) .
j. grande , j. sol , s. basilakos , and m. plionis , j. cosmol .
. phys . * 08 * , 007 ( 2011 ) .
j. sol , j. phys .
: conf . ser . *
283 * , 012033 ( 2011 ) .
e. l. d. perico , j. a. s. lima , s. basilakos , and j. sol , phys .
d * 88 * , 063531 ( 2013 ) .
j. sol , arxiv:1306.1527 [ gr - qc ] .
s. basilakos and j. sol , arxiv:1307.4748 [ astro-ph.co ] .
we can obtain an alternative continuity equation from the first law of thermodynamics . according to ref .
@xcite , using eq .
( [ eq : s(nonadd ) ] ) , the modified continuity equation is calculated as @xmath248 = - \gamma \frac{3}{4 \pi g } \frac{3 c \zeta}{2 } \dot{h } $ ] .
this equation agrees with eq .
( [ eq : fluid_(nonadd ) ] ) , when @xmath90 . in the present paper
, we do not discuss the continuity equation obtained from the first law of thermodynamics .
_ , _ cosmology i _ , modern astronomy series vol . 2 , edited by k. sato and t. futamase ( nippon hyoronsha co. , tokyo , 2008 ) , in japanese . s. m. carroll and w. h. press , annu .
. astrophys . * 30 * , 499 ( 1992 ) .
p. c. w. davies and t. m. davis , found .
phys . * 32 * , 1877 ( 2003 ) ; t. m. davis , p. c. w. davies , and c. h. lineweaver , class . quantum grav . * 20 * , 2753 ( 2003 ) .
b. wang , y. gong , and e. abdalla , phys .
d * 74 * , 083520 ( 2006 ) ; y. gong , b. wang , and a. wang , j. cosmol .
astropart . phys .
01 ( 2007 ) 024
. m. r. setare , phys .
b * 641 * , 130 ( 2006 ) .
j. m. cline , a. r. frey , and g. holder , phys .
d * 77 * , 063520 ( 2008 ) . | an entropic - force scenario , i.e. , entropic cosmology , assumes that the horizon of the universe has an entropy and a temperature . in the present study , in order to examine entropic cosmology , we derive entropic - force terms not only from the bekenstein entropy but also from a generalized black - hole entropy proposed by c. tsallis and l.j.l .
cirto [ eur .
phys .
j. c * 73 * , 2487 ( 2013 ) ] .
unlike the bekenstein entropy , which is proportional to area , the generalized entropy is proportional to volume because of appropriate nonadditive generalizations .
the entropic - force term derived from the generalized entropy is found to behave as if it were an extra driving term for bulk viscous cosmology , in which a bulk viscosity of cosmological fluids is assumed . using an effective description similar to bulk viscous cosmology , we formulate the modified friedmann , acceleration , and continuity equations for entropic cosmology . based on this formulation , we propose two entropic - force models derived from the bekenstein and generalized entropies . in order to examine the properties of the two models
, we consider a homogeneous , isotropic , and spatially flat universe , focusing on a single - fluid - dominated universe .
the two entropic - force models agree well with the observed supernova data .
interestingly , the entropic - force model derived from the generalized entropy predicts a decelerating and accelerating universe , as for a fine - tuned standard @xmath0cdm ( lambda cold dark matter ) model , whereas the entropic - force model derived from the bekenstein entropy predicts a uniformly accelerating universe . |
to shape the beam so that it matches the characteristics of the specific treatment in radiation therapy ( thus achieving the delivery of the prescribed dose to the target ( tumour ) and maximal protection of the surrounding healthy tissue and vital organs ) , beam - limiting and beam - shaping devices ( bl / bsds ) are routinely used .
generally speaking , the beam is first restricted ( in size ) by the primary collimator , a beam - limiting device giving it a rectangular shape .
the beam may subsequently encounter the multi - leaf collimator ( mlc ) , which may be static or dynamic ( i.e. , undergoing software - controlled motion during the treatment session ) .
more frequently than not , the desirable beam shaping is achieved by inserting a metallic piece ( with the appropriate aperture and thickness ) into the beamline , directly in front of the patient ; this last beam - shaping device is called a patient collimator or simply a block .
being positioned close to the patient , the block achieves efficient fall - off of the dose ( sharp penumbra ) outside the target area .
( the simultaneous use of mlc and block is not common . ) to provide efficient attenuation of the beam outside the irradiated volume , the bl / bsds are made of high-@xmath1 materials .
the primary collimator and the mlc are fixed parts of the treatment machine , whereas the block not only depends on the particular patient , but also on the direction from which the target is irradiated .
therefore , there may be several blocks in one treatment plan , not necessarily corresponding to the same thickness or ( perhaps ) material .
the presence of bl / bsds in treatment plans induces three types of physical effects : * \a ) confinement of the beam to the area corresponding to full transmission ( i.e. , to the aperture of the device ) . *
\b ) effects associated with the nonzero thickness of the device ( geometrical effects ) . *
\c ) effects relating to the scattering of the beam off the material of the device .
type-(a ) effects ( direct blocking of the beam ) are dominant and have always been taken into account .
the standard way to do this is by reducing the bl / bsd into a two - dimensional ( 2d ) object ( i.e. , by disregarding its thickness ) and assuming no transmission of the beam outside its aperture .
type-(b ) and type-(c ) effects induce corrections which , albeit at a few - percent level of the prescribed dose , may represent a sizable fraction of the _ local _ dose ; due to their complexity and to time restrictions during the planning phase , these corrections have ( so far ) been omitted in clinical applications .
the subject of the slit scattering in beam collimation was first addressed by courant ( 1951 ) .
courant extracted analytical solutions for the effective increase in the slit width ( attributable to scattering ) by solving the diffusion equation inside the collimator . to fulfill the boundary conditions , he introduced the negative - image technique , which was later criticised ( e.g. , by burge and smith ( 1962 ) ) . despite the debatable usefulness of the practical use of courant s work , that article set forth definitions which were employed in future research ; for example , by categorising the scattered particles as : * those impinging upon the upstream face of the collimator and emerging from its inner surface ( bore ) , * those entering the bore and scattering out of it , and * those entering the bore and leaving the downstream face of the collimator .
in the present paper , courant s type-1 particles will correspond to our ` outer tracks ' ( ots ) , type-2 particles to our ` bore - scattered inner tracks ' ( bsits ) , and type-3 particles to our ` going - through inner tracks ' ( gtits ) , see fig . [
fig : model1 ] . the tracks which do not hit the block will comprise the ` pristine ' beam .
aiming at the determination of the optimal material for proton - beam collimation and intending to provide data for experiments at the linear accelerator of the national institute for research in nuclear science ( nirns ) at harwell , burge and smith ( 1962 ) re - addressed the slit - scattering problem and obtained a solution via the numerical integration of the diffusion equation .
burge and smith reported considerable differences to courant s results . in the last section of their article
, the authors discussed alternative approaches to the slit - scattering problem .
a monte - carlo ( mc ) method , as a means to study the collimator - scattering effects , was introduced by paganetti ( 1998 ) . using the geant code to simulate a proton beamline at the hahn - meitner institut ( hmi ) in berlin , paganetti introduced simple parameterisations to account for the changes in the energy and angular divergence of the beam as it traverses the various beamline elements . judging from fig .
6 of that paper , one expects the scattering effects to be around the @xmath2 level .
in the last section of his article , paganetti correctly predicted that ` monte - carlo methods will become important for providing proton phase - space distributions for input to treatment - planning routines , though the calculation of the target dose will still be done analytically . '
our strategy is similar to that of paganetti : the dose , delivered by the pristine beam , will be corrected for scattering effects on the basis of results obtained via mc runs prior to planning ( actually , when the particular proton - treatment machine is configured ) .
block scattering was also investigated , along the general lines of paganetti s paper , in the work of van luijk ( 2001 ) .
the ( same version of the ) geant code was used , to simulate a proton beamline at kernfysisch versneller instituut ( kvi ) in groningen , and the characteristics of the scattered protons were studied . to validate their approach ,
the authors obtained dose measurements for several field sizes and at several distances from the block . unlike other works
, the effect of scattering in air was also included in that study , and turned out to be more significant than previously thought ( its contribution to the angular divergence of the beam exceeded @xmath3 mrad per @xmath3 m of air ) .
one of the interesting conclusions of that paper was that the penumbra of the dose distribution is mostly accounted for by the lateral spread of the undisturbed beam ; that conclusion somewhat allayed former fears that the extraction of the effective - source size ( @xmath4 ) and of the effective source - axis distance ( sad ) , both obtained from measurements conducted in the presence of a block , might be seriously affected by block - scattering contributions .
finally , the paper addressed the asymmetries induced by a misalignment of the block , concluding that they might be sizable . in their recent work , kimstrand ( 2008 ) put the emphasis on the scattering off the multi - leaf collimator ( mlc ) . in their approach , they also obtain the values of the parameters ( involved in their corrections ) by using a mc geant - based method . in their section ` discussion and conclusions ' , the authors , setting forth the future perspectives , mention that ` a challenge is to implement collimator scatter for a pencil beam kernel dose calculation engine . '
we agree on this being ` a challenge ' .
the authors then advance to pre - empt that ` the methods presented can straightforwardly be applied to arbitrary shaped collimators of different materials , such as moulded patient - specific collimators used in passively scattered proton beams . '
one should at least remark that their paper does not contain adequate information supporting the thesis that the proposed approach is of practical use in a domain where the execution time is a important factor .
in the present paper , the standard icru ( 1987 ) coordinate system will be used ; in beam s eye view , the @xmath5 axis points to the right , @xmath6 upward , and @xmath7 toward the source . the origin of the coordinate system is the isocentre . assuming vanishing transmission outside the aperture of the bl / bsd , all relevant contributions to the fluence involve 2d integrals over its area .
the evaluation of such integrals is time - consuming ; at a time when serious efforts are made to reduce the overall time allotted to each patient , the generous allocation of resources in the evaluation and application of _ corrections _ to the primary dose is unacceptable . to expedite the extraction of these corrections
, one must find a fast way to decompose the effects of a 2d object ( e.g. , of the aperture of the bl / bsd ) into one - dimensional , easily - calculable , easily - applicable contributions .
an example of one such 2d object is shown in fig .
[ fig : miniblock1 ] ; the area b represents the aperture ( corresponding to full transmission of the beam ) and is separated from the area a , representing the ( high-@xmath1 ) material , via the contour c. full attenuation of the beam is assumed in the area a. the contour c and the ` outer ' contour of the area a ( which is not shown ) may have any arbitrary shape ( e.g. , rectangular , circular , etc . ) ; the only requirement is that the contour c be contained within the contour of the area a. let us assume that the aim is to derive the influence of the bl / bsd at one point q ( fig . [
fig : miniblock2 ] ) ; the point q is projected to the point p , on the bl / bsd plane .
the point p may be contained in the interior or the exterior of the area b. to enable the evaluation of the effects of the bl / bsd at the point q , a set of @xmath8 directions on the bl / bsd plane , intersecting at the point p , are chosen ; an example with @xmath9 is shown in fig .
[ fig : miniblock3 ] .
the effects of the bl / bsd at the point q will be evaluated from the elementary contributions of the line segments corresponding to the intersection of @xmath8 straight lines with the contour c. these line segments , which are contained within the area b and are bound by the contour c , will be called _ miniblocks_. the effect of the bl / bsd at the point q will be evaluated by averaging the elementary contributions of the miniblocks created around the point p. in fig .
[ fig : miniblock3 ] , one of these miniblocks is represented by the line segment @xmath10 .
obviously , the number of miniblocks in each direction depends on the number of intersections of the straight line ( drawn through the point p , parallel to the chosen direction ) and the contour c. as in the previous section , a set of @xmath8 directions , intersecting at the point p , are chosen ; an example with @xmath9 is shown in fig .
[ fig : miniblock4 ] .
again , the effects of the bl / bsd at the point q will be evaluated on the basis of the elementary contributions in these @xmath8 directions . in fig .
[ fig : miniblock4 ] , one of the miniblocks is denoted by @xmath11 .
evidently , the number of directions @xmath8 is arbitrary . for fixed @xmath8 ,
the accuracy of the evaluation depends on the details of the contour c and on the proximity of the point p to it .
the reliability of the estimation is expected to increase with @xmath8 .
( the evaluation is exact for @xmath12 . )
there is one difference between the cases described in sections [ sec : within ] and [ sec : without ] . in case
that the point p lies within the area b , there will always be at least one miniblock per direction ; if the point p lies outside the area b , there might be no intersections with the contour c in some directions ( in which case , the corresponding elementary contributions vanish ) .
let us assume that the physical characteristics of the beam ( lateral spread , angular divergence , energy , etc . ) and the entire geometry ( also involving the bl / bsd ) are fixed .
apart from the original point q , the elementary contributions will involve ( the coordinates of ) three additional points : the two end points of the particular miniblock and the point p. due to the fact that the @xmath8 directions are created around the point p , the two end points of every miniblock and the point p will always lie on one straight line . if the point p lies within the area b , it may lie within or outside a miniblock . on the other hand ,
if the point p lies outside the area b , it will always lie outside any corresponding miniblock . in any case
, the elementary contribution of a miniblock will generally be a function of the distance of its two end points to the point p. additionally , the elementary contribution of a miniblock to the point q will also involve the distance of the point q to the bl / bsd plane , denoted as @xmath7 in fig . [
fig : miniblock2 ] .
a few remarks are worth making .
* the angle between consecutive directions may be constant or variable .
if the angle is not constant , weights ( to the elementary contributions ) have to be applied . * the values of @xmath8 in sections [ sec : within ] and [ sec : without ] do not have to be the same .
furthermore , different values of @xmath8 might be used for the points which are projected onto the interior ( or exterior ) of the area b , e.g. , depending on the proximity of the point p to the contour c. at present , the method of this paper applies only to bl / bsds with a constant aperture profile throughout their thickness ; to derive the corrections in case of other shapes ( e.g. , for bl / bsds with a tapered aperture profile ) , substantial modifications are required .
the approach is applicable to any type of bl / bsd , be it the primary collimator , the mlc , or the block ; to derive the corrections , the only input pieces are the physical characteristics of the material ( of which the bl / bsd is made ) and , naturally , geometrical details . from now on ,
however , we will restrict ourselves to the effects induced by the block , which ( given its proximity to the patient ) are expected to be of greater interest and importance in clinical applications .
this choice , meant to emphasise importance , should not be seen as a restriction of the method . a method for the evaluation of the thickness corrections
was recently proposed by slopsema and kooy ( 2006 ) ; we will follow their terminology . thus , ` downstream ( upstream ) projection ' will indicate the projection of the downstream ( upstream ) face of the block on the ( @xmath5,@xmath6 ) plane at the specified @xmath7 position ( depth ) . similarly , the ` extension ' of the block will correspond to the physical block translated in space ( from its actual position to the specified depth , parallel to the central beam axis ) . in medical applications , the formal beam - shaper object ( e.g. , the so - called dicom block ) , which is created at planning time to represent the physical block , comprises the projection of the _ downstream face _ ( of the block ) to a plane ( perpendicular to the beam ) at isocentre depth . the physical block ( extension )
is therefore obtained via a simple scaling involving the source - axis and block - isocentre ( ibd ) distances .
having retrieved the extension of the block , it is straightforward to obtain the downstream projection at any @xmath7 ; in order to obtain the upstream projections , one has to use , in addition to the aforementioned quantities , the block thickness ( @xmath13 ) .
the relation between the dicom block , the extension , and the downstream projection at a specified @xmath7 position is shown in fig .
[ fig : geometry1 ] .
the upstream projection is obtained by projecting the extension from depth @xmath14 .
the thin - block approximation , which is currently used in clinical applications when evaluating the dose , corresponds to @xmath15 mm ( the downstream and upstream projections coincide at all @xmath7 values ) . assuming that the coordinates of the projected ( downstream- or upstream - face , as the case might be ) ends of a miniblock are denoted as @xmath16 and @xmath17 , the contribution of the miniblock to the fluence at a point @xmath18 ( lying on the straight line defined by @xmath16 and @xmath17 ) on the calculation plane is given by the formula @xmath19 where @xmath20 denotes the error function and @xmath21 stand for the rms ( lateral ) spreads of the beam at the specified depth .
the quantities @xmath21 are obtained via source mirroring according to the method of slopsema and kooy ( 2006 ) ; as points on the downstream or upstream faces of the block are used , the resulting @xmath22 values in eq .
( [ eq:1dcontributions ] ) are equal only when one block face is involved in the mirroring process . one simple example of the projections on the calculation plane is shown in fig . [ fig : geometry2 ] ; the block extension is contained within the downstream projection , which ( in turn ) is contained within the upstream projection .
in reality , depending on the complexity of the shape of the block aperture and on the relative position of the central beam axis , these three contours might intersect one another .
the central beam axis intersects the calculation plane at the origin of the ( @xmath5,@xmath6 ) coordinate system .
one miniblock is shown ( the line segment contained within the block extension ) , along with two points , one within the miniblock ( p ) , another outside the miniblock ( p@xmath23 ) .
the fluence contributions to both points may be evaluated by using eq .
( [ eq:1dcontributions ] ) with the appropriate @xmath24 and @xmath21 values .
the essentials for the evaluation of the contribution of a miniblock to the fluence at a specified point on the calculation plane are to be found in fig .
[ fig : geometry3 ] .
although , in the general case , the miniblock does not contain the intersection of the @xmath5 and @xmath6 axes of fig .
[ fig : geometry2 ] , all lengths ( which are important to our purpose ) scale by the same factor , thus enabling the simplified picture of fig .
[ fig : geometry3 ] . according to slopsema and kooy ( 2006 ) ,
the source is mirrored onto the calculation plane by using points either on the downstream or on the upstream face of the block .
the values of the mirrored - source size , corresponding to these two options , are given by @xmath25 and @xmath26 the coordinates of the projected points @xmath27 , @xmath28 , @xmath29 , and @xmath30 are obtained on the basis of fig . [ fig : geometry3 ] via simple operations .
the last item needed to derive the contributions to the fluence may be found on page 5444 of the paper of slopsema and kooy ( 2006 ) : ` protons whose tracks project inside the aperture extension onto the plane of interest see the upstream face as the limiting boundary . only for protons
whose tracks end up outside the aperture on the plane of interest is the downstream face the limiting aperture boundary . ' in fact , this statement applies to the case of a half - block ( one - sided block ) .
the modification , however , in case of a miniblock is straightforward .
* points within the extension of the aperture see the upstream face of the miniblock as the limiting boundary .
* points outside the extension of the miniblock see one upstream _ and one downstream _ edge as limiting boundaries . obviously , in order to evaluate the fluence at the point p ( fig .
[ fig : geometry3 ] ) , one has to use the coordinates of the projected points @xmath27 and @xmath30 , along with @xmath31 in eq .
( [ eq:1dcontributions ] ) . on the other hand , to evaluate the fluence at p@xmath23 , one has to use @xmath27 along with @xmath31 ( contribution of the ` left ' part of the miniblock ) and @xmath29 along with @xmath32 ( contribution of the ` right ' part of the miniblock ) .
this simplified picture , featuring what a point ` perceives ' as limiting boundaries , suffices in obtaining the appropriate fluence contributions . finally , the number of contributions which a point will receive depends on the geometry and on the shape of the block .
one example of a point receiving two contributions in a given direction is shown in fig .
[ fig : geometry4 ] ; the point p lies within the extension of the miniblock on the left and outside the extension of the miniblock on the right .
additionally , one might have to deal with a block with more than one apertures ( hence contours ) . to determine in the present work the appropriate thickness corrections for arbitrary block - aperture shape and beamline geometry , dedicated software was developed . before entering the details of the derivation of the scattering corrections
, it is worth providing a concise outline of our approach .
first of all , our aim is to obtain a _ fast and reliable _ solution to the scattering problem ; exact analytical solutions are welcome as long as they fulfill this requirement .
second , the solution has to be general enough for direct application to all proton - treatment machines . to expedite the derivation and application of the corrections at planning time
, we will introduce a two - step approach .
* all the parameters which are independent of the specific details of plans will be evaluated during the beam - configuration phase ; given a fixed ( hardware ) setup , every treatment machine is configured only once . * the corrections for each particular plan
will be derived ( at planning time ) from the existing results ( i.e. , those obtained at beam - configuration phase ) via simple interpolations .
given that the physics of multiple scattering is known , it is possible to obtain the exact solution for the relevant beam properties ( lateral spread , angular divergence ) from the beamline characteristics of each treatment machine .
however , it is unrealistic to introduce a dedicated process for each supported machine , especially in order to derive corrections to the delivered dose .
additionally , if a dedicated ( per - case ) approach is implemented in a software product which is intended to support a variety of machine manufacturers , one has to be prepared to allot the necessary resources whenever a new product appears . to avoid these problems and to retain the generality of the approach , one has no other choice but to introduce a simple , adjustable model to account for the beam optics .
the model of the present paper has only one parameter , which will be fixed from half - block fluence measurements .
the parameters which achieve the description of the various distributions of the scattered protons are determined via mc runs .
these runs take account of the variability in the block material , block thickness , incident energy ( energy at nozzle entrance ) , and nozzle - equivalent thickness ( net ) for all the options ( combinations of the hardware components of the beamline , leading to ranges of available energies and of nets , as well as imposing restrictions on the field size ) for which a machine is configured . to enable the easy use of the results ,
the output is put in the form of expansion parameters in two geometrical quantities which are involved in the description of the scattering effects .
the scattering corrections for all the blocks in a plan are determined ( at planning time ) from the aforementioned results via simple interpolations .
the application of the corrections involves the concept of miniblocks , as they have been introduced in section [ sec : miniblocks ] of this paper .
one model which is frequently used in beam optics features the bivariate gaussian distribution in the lateral direction @xmath6 ( distance to the central beam axis ) and the ( small ) angle @xmath33 ( with respect to that axis ) .
( rotational symmetry is assumed here . )
@xmath34 with @xmath35 the parameters @xmath36 and @xmath37 represent twice the variance in @xmath6 and @xmath33 , respectively .
the ( @xmath6,@xmath33 ) correlation is defined as @xmath38 the quantity @xmath39 ( which is bound between @xmath40 and @xmath3 ) is a measure of the focusing in the beam .
positive @xmath39 values indicate a defocusing system , negative a focusing one ; this becomes obvious after one puts eq .
( [ eq : bivariate ] ) in the form @xmath41 we now touch on the variation of @xmath36 , @xmath42 , and @xmath37 along the beam - propagation direction . assuming that the quantities @xmath43 , @xmath44 , and @xmath45 denote the corresponding values at isocentre depth and that the beam propagates in air ( without scattering ) , @xmath36 , @xmath42 , and @xmath37 at distance @xmath7 from the isocentre ( see fig . [ fig : geometry1 ] ) are given by the expressions @xmath46 with these transformations , the joint probability distribution of eq .
( [ eq : bivariate ] ) is invariant under translations in @xmath7 . in case that the beam traverses some material , eqs .
( [ eq : abcz ] ) have to be modified accordingly , to take account of the beam broadening due to multiple scattering .
the accurate modelling of the beam may be obtained on the basis of formulae such as those given in the previous section . in section [ sec : scattering ] , however , we reasoned that a simplified parameterisation of the beamline is desirable ; one additional argument may be put forth . currently , as far as proton therapy is concerned , four dose - delivery techniques are in use : single - scattering , double - scattering , uniform - scanning ( formerly known as wobbling ) , and modulated - scanning ( formerly simply known as scanning ) . in the modulated - scanning technique ,
magnets deflect a narrow beam onto a sequence of pre - established points ( spots ) on the patient ( for pre - determined optimal times ) , thus ` scanning ' the ( cross section of the ) region of interest .
uniform scanning involves the spread - out of the beam using fast magnetic switching .
the broadening of the beam in the single - scattering technique is achieved by one scatterer , made of a high-@xmath1 material and placed close to the entrance of the nozzle .
currently , the most ` popular ' technique involves a double - scattering system . in a double - scattering system ,
a second scatterer is placed downstream of the first scatterer in order to achieve efficient broadening of the beam ; studies of the effects of the second scatterer may be found in the literature , e.g. , see takada ( 2002 ) and more recent reports by the same author and gottschalk .
the second scatterer is usually made of two materials : a high-@xmath1 ( such as lead ) material at the centre ( i.e. , close to the central beam axis ) , surrounded by a low-@xmath1 ( such as aluminium , lexan , etc . )
material ( which is frequently , but not necessarily , shaped as a concentric ring ) .
the arrangement produces more scattering at the centre than the periphery , leading ( after sophisticated fine - tuning ) to the creation of a broad flat field at isocentre . to simulate the effect of the second scatterer in the present work , ( @xmath6,@xmath33 )
events are generated at @xmath47 as follows .
the variable @xmath6 is sampled from the gaussian distribution with mean @xmath48 and variance @xmath49 ( the source - size calibration must precede this step ) ; @xmath4 depends on the incident energy and net . to account for the lateral limits of the beam ,
@xmath6 is restricted within the interval @xmath50 $ ] , where the characteristic length @xmath51 is taken herein to be the radius of the second scatterer .
the variable @xmath33 is first sampled from the gaussian distribution with mean @xmath48 and a @xmath6-dependent variance @xmath52 according to the formula @xmath53 where @xmath54 denotes the lateral position as a fraction of @xmath51 ; @xmath55 . to account for the @xmath6-dependent bias in @xmath33
, we then use the transformation @xmath56 , where @xmath57 stands for the distance between the first scatterer and the source .
obviously , @xmath58 denotes the ratio of two @xmath59 values , i.e. , the value at the centre of the second scatterer over the one at the rim ; @xmath58 is the only free parameter of the model . the value of @xmath60 is obtained from the incident energy and net ; the angular divergence of the beam at nozzle entrance is ( currently ) assumed to be @xmath48 .
it is not difficult to prove that , given the aforementioned rules for generating the ( @xmath6,@xmath33 ) events , the open - field fluence at a lateral position @xmath6 at depth @xmath61 is given by @xmath62 where @xmath63 .
it has to be emphasised that this definition of the fluence does not involve an overall @xmath64-factor ( @xmath65 being the distance between the calculation plane and the source ) ; therefore , this formula is compatible with the format in which the lateral fluence measurements , used during the beam - configuration phase , appear . it can be shown that the only modification in case that a half - block is inserted into the beamline ( e.g. , as shown in fig .
[ fig : model1 ] ) involves the upper limit @xmath66 of integration in eq .
( [ eq : fluenceopen ] ) ; instead of @xmath63 , one must now use @xmath67 if @xmath68 , or otherwise @xmath69 .
although the method of the present paper was originally developed for the double - scattering technique , it is also applicable in single scattering and uniform scanning ; in both cases , one simply has to fix the parameter @xmath58 to @xmath3 . as the method is applicable only in case of broad fields , it has no bearing on modulated scanning .
the elements needed for the description of the passage of particles through matter may be found ( in a concise form ) in yao ( 2006 ) , starting on page 258 ; most of the deflection of a charged particle traversing a medium is due to coulomb scattering off nuclei . despite its incompleteness ( e.g. ,
see the discussion in the geant4 physics - reference manual , section on ` multiple scattering ' , starting on page 71 ) , the multiple - scattering model of molire is used here .
the large - angle scattering is not taken into account ; the angular distribution of the traversing beam is assumed gaussian .
highland s logarithmic term , appearing in the expression of @xmath70 , will be approximated by a constant factor involving the block thickness ; a similar strategy was followed in gottschalk ( 1993 ) .
the lynch - and - dahl ( 1991 ) values will be used in the expression for @xmath70 : @xmath71 where @xmath72 denotes the depth along the original trajectory , @xmath73 the velocity and @xmath74 the momentum of the proton , and @xmath75 the radiation length in the material of the block ( for a convenient parameterisation of @xmath75 , see yao ( 2006 ) , page 263 ) .
equation ( [ eq : theta0 ] ) applies to ` thin ' targets . for ` thick ' targets
, the dependence of the proton momentum on the depth @xmath72 must be taken into account . omitting the logarithmic term on the right - hand side
, one may put eq .
( [ eq : theta0 ] ) in the form @xmath76 where @xmath77 on the practical side , eq .
( [ eq : theta0simple ] ) with a constant ( or at least ` not too complicated ' ) @xmath78 factor would be attractive as one would then be able to obtain fast analytical solutions for the propagation of a simulated track inside the material of the block .
therefore , it is worth determining the range of thickness values within which the constancy of the @xmath78 factor remains a reasonable assumption .
the direct comparison with the data of gottschalk ( 1993 ) revealed that the ` thick - target ' corrections are unfortunately indispensable at depths exceeding about @xmath79 of the range of @xmath80-mev protons , incident on a variety of materials .
to abide by the original goal of obtaining a fast solution , we had to follow an alternative approach ( to that of using eq .
( [ eq : fq ] ) ) , by parameterising the @xmath72-dependence of the @xmath78 factor in a simple manner ; at present , the best choice seems to be to make use of the empirical formula @xmath81 where @xmath82 denotes the ( energy - dependent ) range of the incident proton in the material of the block and @xmath83 @xmath84 is now a constant , depending only on the initial value of @xmath85 .
the validation of eqs .
( [ eq : theta0new ] ) and ( [ eq : f ] ) was made on the basis of a comparison with the experimental data of gottschalk ( 1993 ) , namely the measured @xmath86 values of that paper .
good agreement with the data was obtained for four materials which are of interest in the context of the present study ( carbon , lexan , aluminum , and lead ) . only at one entry ( one of the largest depth values in lead , i.e. , the measurement at @xmath87 ) , was a significant difference ( of slightly less than @xmath88 ) found ;
the origin of that difference was not sought .
figure [ fig : model2 ] shows an example of a trajectory of one proton inside a block .
the incident proton , an ot in this figure , hits the upstream face of the block at angle @xmath33 with respect to the beam axis .
two new _ in - plane _ variables are introduced to describe the kinematics at depth @xmath72 ( along the direction of the original trajectory ) : the deflection @xmath65 ( off the original - trajectory course ) and the angle @xmath89 ( with respect to the direction of the original trajectory ) .
although the proton moves in an irregular path inside the block , the ` history ' of the actual motion will be replaced by a smooth movement leading to the same value of @xmath65 at @xmath72 .
this ` smooth - deflection ' approximation will enable the association of the energy loss in the material of the block with the doublet of ( @xmath72,@xmath65 ) values .
since the path length , which is calculated in this approximation , is an underestimate of the actual path length , a constant conversion factor ( true - path correction ) of @xmath90 has been used ; there is some arbitrariness concerning this choice , yet it appears to be reasonable .
it needs to be stressed that , in fig .
[ fig : model2 ] , @xmath5 denotes the direction of the beam propagation ; the auxiliary coordinate system introduced in this figure should not be confused with the coordinate system of fig .
[ fig : geometry1 ] , which is the formal one in medical applications . in the generation of the mc events ,
the suggestion of yao ( 2006 ) , page 262 , for the quantities @xmath65 and @xmath89 has been followed .
two independent gaussian random variables ( @xmath91,@xmath92 ) with mean @xmath48 and variance @xmath3 are first created in each event ( track hitting the block ) .
the quantities @xmath65 and @xmath89 are expressed in terms of ( @xmath91,@xmath92 ) as follows : @xmath93 and @xmath94 where @xmath95 is taken from eq .
( [ eq : theta0new ] ) .
the values of the doublet ( @xmath91,@xmath92 ) fix the dependence of @xmath65 and @xmath89 on @xmath72 in each generated event .
it may then be determined ( either analytically or numerically ) whether the particular track leaves the block . finally , for those tracks leaving the block , a simple rotation yields the coordinates of the exit point in the ( @xmath5,@xmath6 ) coordinate system of fig .
[ fig : model2 ] .
the energy of the leaving proton is determined from its residual range ( original range minus the actual path length inside the material of the block ) .
the definitions of the three types of scattered protons have been given in section [ sec : introduction ] ; in fig .
[ fig : model1 ] , one obtains a rough schematic view of the corresponding contributions to the fluence . in this section , we will introduce convenient forms to parameterise the lateral fluence distributions of the scattered protons . as far as these distributions are concerned , the uninteresting offset @xmath96 ( lateral displacement of the block ) in fig .
[ fig : model1 ] will be omitted .
therefore , for the needs of this section , @xmath97 mm at the extension of the block in fig .
[ fig : model1 ] , that is , not at the position where the central beam axis intersects the calculation plane . to obtain the lateral fluence distributions , literally corresponding to fig .
[ fig : model1 ] , one has to take the offset @xmath96 into account . in the formulae below ,
the placement of the half - block is assumed as shown in fig .
[ fig : model1 ] ( i.e. , extending to @xmath98 ) . the empirical formula for the description of the lateral fluence distribution of the ots reads as @xmath99 where @xmath100 .
four conditions for the parameters @xmath101 , @xmath85 , and @xmath102 must be fulfilled : @xmath103 , @xmath104 , @xmath105 , and @xmath106 .
the same empirical formula is used in the parameterisation of the lateral fluence distribution of the bsits ; the four aforementioned conditions also apply .
( the resulting optimal parameter values are , of course , different . ) the optimal description of the ( broader distribution of the ) gtits is achieved on the basis of a lorentzian @xmath107 multiplied by the asymmetry factor @xmath108 to account for the observed skewness of the lateral fluence distribution ( toward positive @xmath6 values ) .
three conditions must be fulfilled : @xmath103 , @xmath109 , and @xmath110 .
the lateral fluence distribution of the pristine tracks is fitted by using the standard formula @xmath111 where @xmath103 and @xmath110 ; @xmath112 is the cumulative gaussian distribution function . only the factor @xmath101
will be retained from the fits to the pristine - beam data , to be used in the _ normalisation _ of the fluence corresponding to each of the three types of scattered protons .
expressing the contributions of the scattered protons as fractions of the pristine - beam fluence enables the efficient application of the corrections at planning time . from all the above ,
it is evident that the description of the contribution to the fluence ( at fixed @xmath7 ) of any of the three types of the scattered protons is achieved on the basis of three parameters ; hence , there are nine parameters in total . at the end of each cycle ( comprising a set of mc runs for a number of @xmath96 values , for a given energy - net combination ) , each of these nine parameters @xmath113 is expanded in terms of @xmath96 and @xmath7 , using the quadratic model @xmath114 the final results are the @xmath115 coefficients @xmath116 , @xmath117 , @xmath118 , obtained at several energy - net combinations for the option of the machine which is under calibration . at planning time , the values of the parameters @xmath113 are reconstructed from the existing results ( via simple 2d interpolations in the incident energy and net ) , the values of @xmath96 ( corresponding to the particular miniblock which is processed ) , and @xmath7 ( corresponding to the calculation plane which is processed ) .
let us assume that one option of a selected machine has been chosen for calibration ( at beam - configuration phase ) .
all half - block data ( which is contained in that option ) is used in the determination of the parameter @xmath58 on the basis of an optimisation scheme ( featuring the c++ implementation of the standard optimisation package minuit of the cern library ) , along with eq .
( [ eq : fluenceopen ] ) with the @xmath66 value of eq .
( [ eq : t0 ] ) , one could also use the measurements of the open - field fluence , along with eq .
( [ eq : fluenceopen ] ) with @xmath63 .
however , what is habitually called ` open - field measurements ' in the field of radiation therapy corresponds to beams which have already been restricted in size by the primary collimator . ] .
representative incident - energy and net values are chosen from the ranges of values associated with the option which is under calibration .
for each acceptable energy - net combination , a number of mc runs are performed , each corresponding to one @xmath96 value . in each of these mc runs , events (
i.e. , ( @xmath6,@xmath33 ) pairs , each corresponding to one proton track ) are generated according to the formalism developed in section [ sec : parameterisation ] .
the value of the parameter @xmath58 , obtained from the half - block data ( for the option in question ) at the previous step , comprises ( i.e. , apart from geometrical characteristics of the machine which is configured ) the only input to these mc runs .
the resulting tracks are followed until they either hit the block or pass through it .
the tracking of the protons inside the material of the block is done according to section [ sec : mcruns ] ; finally , these tracks either vanish ( being absorbed in the material of the block ) or emerge from it ( bore , downstream face ) and deliver dose . the tracks which emerge from the block are properly flagged ( ots , bsits , or gtits ) and their contributions to the fluence at a number of @xmath7 depths are stored ( histogrammed )
. fits to these distributions , using the empirical formulae of section [ sec : latdist ] , lead to the extraction of the parameters achieving the optimal description of the stored data . after the completion of all the runs for all the chosen values of @xmath96 , the entire set of the parameter values
is subjected to fits by using eq .
( [ eq : sr ] ) for each parameter separately .
finally , the coefficients @xmath116 , @xmath117 , @xmath118 , appearing in eq .
( [ eq : sr ] ) , are stored in files along with the values of the incident energy and net corresponding to the particular mc run ; these output files will comprise the only input when the corrections to a particular plan will be derived ( planning time ) .
the procedure above is repeated until all options of the given proton - treatment machine have been calibrated .
the variability in the material and in the thickness of the block is also taken into account in the current implementation ( by looping over those combinations requested by the user ) .
finally , it is worth repeating that this time - consuming part of obtaining the files containing the @xmath116 coefficients ( a question of a few hours per option ) has to be performed only once , when the proton - treatment machine is configured . at planning time
, the pristine - beam fluence is calculated first .
the beam - scattering corrections are then obtained ( i.e. , if they have been requested ) after employing a number of elements developed in the course of the present section : * the concept of miniblocks introduced in section [ sec : miniblocks ] . * the reconstruction of each of the nine parameters , used in the description of the scattered protons , at a few @xmath7 values , on the basis of eq .
( [ eq : sr ] ) from the results pertaining to the option selected in the treatment plan . the appropriate file ( corresponding to the block material and thickness in the plan )
is used as input .
the final results for the various parameters are obtained via simple 2d interpolations in the incident energy and net . * the empirical formulae of section [ sec : latdist ] . * a simple ( linear ) interpolation to obtain the corrections at all depths @xmath7 in the plan .
the final step involves the application of the corrections to the pristine - beam fluence .
the break - up of the task of determining the scattering corrections into two steps , as described in this section of the paper , enables the minimisation of the time consumption during the evaluation and application of these corrections in proton planning ; of central importance , in this respect , is the concept of miniblocks .
the measurements , which are analysed in this paper , have been obtained at the proton therapy centre of the national cancer centre ( ncc ) , south korea .
the first report on the clinical commissioning and the quality assurance for proton treatment at the ncc appeared slightly more than one year ago , see lee ( 2007 ) .
the ncc proton - treatment machine has been manufactured by ion beam applications ( iba ) , louvain - la - neuve , belgium .
its nominal sad is @xmath119 mm and the distance between the first scatterer and the isocentre is @xmath120 mm .
the double - scattering technique currently supports eight options with incident energies ranging from @xmath121 to @xmath122 mev .
all half - block measurements have been taken in air , using @xmath123 mm .
the lateral displacement of the @xmath124mm - thick brass block , used in these measurements , was @xmath125 mm ; the block was positioned opposite to what is shown in fig .
[ fig : model1 ] , i.e. , blocking to beam from @xmath126 mm to ( theoretically ) @xmath127 . to apply eq .
( [ eq : fluenceopen ] ) with the @xmath66 value of eq .
( [ eq : t0 ] ) , the @xmath6 axis was inverted , as a result of which the @xmath96 value of @xmath128 mm was finally used in eqs .
( [ eq : fluenceopen ] ) and ( [ eq : t0 ] ) .
each option of the double - scattering technique at the ncc comprises @xmath129 energy - net combinations ; in each of these combinations , the lateral fluence distributions have been obtained at four @xmath7 positions , namely at @xmath130 , @xmath131 , @xmath48 , and @xmath132 mm .
one example of these profiles is given in fig .
[ fig : halfblockprofiles ] .
it is worth mentioning that each profile had been separately normalised ( during the data taking ) by setting the corresponding average value of the fluence ` close ' to the central beam axis to ( the arbitrary value of ) @xmath133 of the corresponding maximal fluence ( of each set , separately ) be brought to @xmath97 mm ; thus , an important offset is irretrievably lost . ] ; unfortunately , the individual normalisation factors are not available .
the ` ears ' of the distribution for the data set at @xmath134 mm , which are presumably due to block scattering , have been removed via software cuts . to avoid fitting the noise , fluence values below @xmath135 of the maximal value in each data set were not processed .
the block - thickness effects were removed from the data prior to processing . before advancing to the results ,
one important remark is prompt .
a significant deterioration of the description of the data in the last of the options ( option @xmath136 ) was discovered during the analysis ; at present , it is not entirely clear what causes this problem .
it seems that our beamline model leads to a systematic overestimation of the penumbra in option @xmath136 .
it has to be mentioned that the data in that option yields an unusually small spot size ( @xmath4 ) , about @xmath137 times smaller than the typical values extracted in the other seven options .
option @xmath136 is the only one in which the iba second scatterer @xmath138 ( which is admittedly of peculiar design , is smaller than in the cases of the @xmath139 and @xmath140 scatterers which are used in options @xmath3-@xmath141 .
furthermore , the thickness of lexan _ abruptly _ increases close to the rim of this scatterer .
evidently , the amount of material used in the @xmath138 can not provide efficient broadening of the beam ; adding material would imply smaller beam energy at nozzle exit , at a time when the emphasis in this option is obviously put on the high end of the energy spectrum . to be compatible with the requirement for flatness of the resulting field at isocentre , a smaller maximal field diameter ( @xmath142 versus @xmath143 to @xmath144 mm of the other options ) in the clinical application of option @xmath136 had to be imposed . ] ) is used . in all probability ,
the problems seen in option @xmath136 originate from the shape of the second scatterer , namely from the fact that the energy loss in its material is not kept constant radially ( actually , it is ` discontinuous ' at @xmath145 ) . in order to investigate the sensitivity of our conclusions to the treatment of option @xmath136 , we will perform the analysis both including in and excluding from the database the measurements of that option .
the extracted values of the parameter @xmath58 for all the energy - net combinations of all double - scattering options of the ncc machine are shown in fig .
[ fig : lambda ] .
the uncertainties in the case of option @xmath136 are ( on average ) larger than in the other options .
the variability of the values within each option is due to the fact that , as a result of the numerous assumptions made to simplify the problem , @xmath58 ( which should , in principle , characterise only the second scatterer ) was finally turned into an effective parameter . to decrease the ` noise ' seen in fig .
[ fig : lambda ] , the weighted average of the extracted @xmath58 values within each option was finally used in the ensuing mc runs for all energy - net combinations in that option ( see table [ tab : lambda ] ) .
@xmath146 million events have been generated in each energy - net combination , at each of three @xmath96 values ( @xmath147 , @xmath148 , and @xmath149 mm ) .
the lateral fluence distributions have been obtained at @xmath150 positions in depth , from @xmath151 to @xmath152 mm .
the parameters of these distributions for the three types of scattered protons have been extracted using the formulae of section [ sec : latdist ] . in all cases ( i.e. , including option @xmath136 ) , the description of the data was good ;
the reduced @xmath153 values ( @xmath154 , ndf being the number of degrees of freedom in the fit ) came out reasonably close to @xmath3 .
figures [ fig : outer ] and [ fig : inner ] show typical lateral fluence distributions for outer and inner tracks , respectively . figures [ fig : correctionsat100mm]-[fig : correctionsatm100 mm ] , obtained with a lateral block displacement of @xmath155 mm , show the scattering corrections ( to be applied to the lateral fluence distributions of the pristine tracks ) at three @xmath7 positions around the isocentre ( @xmath133 , @xmath48 , and @xmath152 mm , respectively ) .
as expected , the distributions broaden when receding from the block ; the mode of the fluence contribution of the ots moves about @xmath3 mm away from the block extension for every @xmath156 mm of depth , thus indicating an ` average ' exit angle ( to the bore ) of about @xmath157 .
concerning their magnitude , the corrections generally amount to a few percent of the corresponding pristine fluence for the typical distances involved in clinical applications .
it is now time to enter the subject of the energy loss of the scattered protons . to a good approximation
, one may assume that the energy distributions of the scattered protons depend only on the ratio @xmath158 , where @xmath159 is the energy of the scattered proton and @xmath160 denotes the energy of the pristine beam ( i.e. , the energy at nozzle exit ) . in their study
, kimstrand ( 2008 ) made the same observation .
the energy distributions of the scattered protons were investigated in the case of three energy - net combinations of the ncc machine : ( @xmath161 mev , @xmath162 mm ) , ( @xmath163 mev , @xmath164 mm ) , and ( @xmath165 mev , @xmath166 mm ) .
these combinations have been selected from the data sets of options @xmath3 , @xmath167 , and @xmath136 , respectively , in which different second scatterers are employed ; the corresponding energies @xmath160 are : @xmath168 , @xmath169 , and @xmath170 mev .
the results have been obtained using a @xmath124-mm thick brass block .
figures [ fig : energyouter ] and [ fig : energyinner ] show the energy distributions of the scattered protons as functions of the ratio @xmath171 ; instead of referring explicitly to each energy - net combination , we will simply use the option number ( or @xmath160 ) to identify the results .
the following remarks are worth making .
* at low and moderate values of the exit energy @xmath160 , the energy distributions of the two types of inner tracks are similar , following the @xmath172 probability distribution . at the high end of the energies used in the ncc machine ,
the energy distribution of the bsits departs from this simplified picture , attaining a peak close to @xmath173 ; this is due to bsits which almost ` brush ' the surface of the bore .
the energy distribution of the gtits remains unchanged .
nevertheless , to retain simplicity , we will assume that the energy distribution of all inner tracks follows the formula @xmath172 .
* the energy distribution of the ots is strongly smeared toward low @xmath171 values .
the ots lose a significant amount of their energy when traversing the material of the block ; as shown in fig . [
fig : energyouter ] , their energy distributions peak around @xmath174 to @xmath175 . to fit the energy distribution of the ots
, we used the empirical formula @xmath176 where the parameter @xmath177 turns out to be around @xmath178 to @xmath179 ; the constant of proportionality is obtained from the normalisation of the probability distribution .
courant s effective - size corrections , corresponding to the three aforementioned cases , are : @xmath180 , @xmath181 , and @xmath182 mm . at small aperture sizes ,
the dominant contribution to the fluence ( of the scattered protons ) originates from ots ; the inner tracks dominate at high aperture values .
the crossover is energy - dependent , ranging from about @xmath129 ( option-@xmath3 result ) to @xmath183 mm ( option-@xmath136 result ) .
the absolute yields of the different types of the scattered protons are given in table [ tab : yields ] for an aperture size ( i.e. , the diameter , assuming circular shape ) of @xmath133 mm .
we observe that the absolute yield of the ots increases by a factor of about @xmath178 between @xmath168 and @xmath169 mev , and by more than @xmath184 between @xmath169 and @xmath170 mev . at @xmath168 mev ,
the bsits account for more than @xmath79 of the total yield of the scattered protons .
however , the importance of the bsits diminishes with increasing energy , reaching the level of @xmath3 out of @xmath184 ots at @xmath170 mev . on the contrary
, the yield of the gtits flattens out at about @xmath185 per ot .
one might conclude that the bsits seem to be more important at low energies and the ots at high energies .
the ratio of the yields of the gtits and ots is almost constant , varying only from @xmath186 to @xmath187 as the energy increases from @xmath168 to @xmath170 mev .
the verification of the method of the present paper should obviously involve the reproduction of dedicated dose measurements obtained in some material which , as far as the stopping power for protons is concerned , resembles human tissue , e.g. , in water .
the measurements should cover the region around the isocentre , where the tumour is usually placed , and , in order that the approach be validated also in the entrance region , they should extend to small distances from ( the downstream face of ) the block . finally , the method must be validated for a range of depth values ( associated with the energy at nozzle exit ) . at present ,
given the lack of dedicated dose measurements , the only possibility for verification rests on re - using the calibration data ( i.e. , the half - block fluence measurements described in section [ sec : measurements ] ) .
we are aware of the fact that using the same data for configuring a system and for validating its output does not constitute an acceptable practice .
however , since parts of the data ( i.e. , the areas which are obviously contaminated by the block - scattering effects ) had been removed from the database before extracting the @xmath58 values , this approach becomes a valid option .
luckily , as far as the validation of the scattering corrections to the fluence is concerned , our interest lies in ( the reproduction of ) those excluded areas ; naturally , for the rest of the measurements ( i.e. , for those which _ were _ used in the determination of the @xmath58 values ) , it has to be verified that the quality of the description of the experimental data is not impaired by the inclusion of the block - scattering corrections .
a typical reproduction of the measurements is given in figs .
[ fig : rprdctn1 ] and [ fig : rprdctn2 ] ; the data correspond to the first energy - net combination of option @xmath3 of the ncc machine , taken at @xmath134 mm ( i.e. , @xmath133 mm away from the block ) . shown in fig .
[ fig : rprdctn1 ] are the lateral fluence measurements ( continuous line ) along with the mc data corresponding only to the pristine beam ; the effects of the scattered protons are added to the pristine - beam fluence ( resulting in what will be called henceforth ` total fluence ' ) in fig .
[ fig : rprdctn2 ] . on the basis of the visual inspection of these two figures ,
there is no doubt that the quality of the reproduction of the measurements in the latter case ( i.e. , when including the block - scattering effects ) is superior .
we will next investigate the goodness of the reproduction of all measurements on the basis of a commonly - used statistical measure , e.g. , of the standard @xmath153 function .
alternative options have been established ( e.g. , the @xmath102-index approach of low ( 1998 ) ) , but have not been tried in this work .
one has to bear in mind that the block - scattering contributions are larger at small distances from the block and in the area neighbouring its extension ; at large distances , the distributions of the scattered protons broaden as a result of the angular divergence of the scattered beam and , less importantly , of scattering in air ( an effect which has not been included in this paper ) .
evidently , the assessment of the goodness of the reproduction of the data , paying no attention to the characteristics of the effect in terms of the depth @xmath7 , makes little sense .
the measurements in the area corresponding to the penumbra are very sensitive to the ( input ) value used for the lateral displacement of the block ; small inaccuracy in this value affects the description of the data significantly , introducing spurious effects in the @xmath153 function .
this area , albeit very important in the determination of the value of the parameter @xmath58 , was not included in this part of the analysis .
the resulting @xmath153 values are given in table [ tab : chisquare ] , separately for the four @xmath7 positions at which the measurements have been obtained ; it is evident that , for all depth values , the quality of the reproduction of the measurements when including the block - scattering effects is superior to the case that only the pristine - beam contribution is considered .
the importance of the inclusion of these effects decreases with increasing distance to the block .
the improvement for @xmath134 mm when including the block - scattering effects is impressive .
judging from the @xmath153 values for the given degrees of freedom , there can be no doubt that the overall reproduction of the data is satisfactory .
last but not least , despite the fact that the description of the option-@xmath136 measurements is debatable , our conclusions do not depend on the treatment of the data in that option ( i.e. , inclusion in or exclusion from the database ) . out of the available @xmath188 lateral fluence profiles , corresponding to @xmath134 mm
, only five profiles did not show improvement after the scattering effects were included . in two of these profiles ,
namely the energy - net combinations of ( @xmath189 mev , @xmath190 mm ) and ( @xmath191 mev , @xmath192 mm ) , the scattering contributions are present in the measurements , yet at different amounts compared to the mc - generated data ; additionally , a hard - to - explain slope ( i.e. , of different sign to what is expected after including the scattering contributions ) is clearly seen in these measurements around @xmath97 mm . on the other hand ,
no scattering effects ( ` ears ' ) are discernible in the ( @xmath193 mev , @xmath194 mm ) , ( @xmath195 mev , @xmath196 mm ) , and ( @xmath197 mev , @xmath198 mm ) combinations at @xmath134 mm .
it has to be stressed that these five data sets are surrounded by a multitude of measurements showing an impressive agreement between the experimentally - obtained and the mc - generated total - fluence distributions ; due to this reason , we rather consider the absence of improvement ( that is , after the scattering contributions are included ) in the description of the data in these five profiles as indicative of experimental problems . an alternative way of displaying the content of figs .
[ fig : rprdctn1 ] and [ fig : rprdctn2 ] is given in fig .
[ fig : normres ] ; instead of showing separately the measurements and the mc data , shown in fig .
[ fig : normres ] are the normalised residuals , defined as @xmath199 where @xmath200 denotes the @xmath201-th measurement and @xmath202 the corresponding mc - obtained fluence ; @xmath203 and @xmath204 represent their uncertainties .
the advantage of such a plot is evident as direct information on the reproduction may be obtained faster than from figs .
[ fig : rprdctn1 ] and [ fig : rprdctn2 ] ; for instance , not only can one immediately become aware of the failure of the pristine - beam data close to the borders of the block , but also of the severity of this failure .
evidently , the pristine - beam contribution underestimates the fluence by about @xmath3 to @xmath137 standard deviations for @xmath126 mm @xmath205 mm .
it is also interesting to note that , after the scattering effects have been included , the normalised residuals @xmath206 show significantly smaller dependence on the lateral distance @xmath6 ( ideally , no dependence should be seen ) .
the aim of the present paper was to provide the systematic description of a method to be used in the determination and application of the corrections which are due to the presence of bl / bsds in proton planning . despite the fact that no emphasis was meant to be put on a clinical investigation , one simple example ( of the application of the corrections )
may nevertheless be called for . to this end , the current version of the treatment planning system eclipse@xmath0 ( varian medical systems inc . , palo alto , california )
was extensively modified to include the derivation ( in beam configuration ) and the application ( in planning ) of both block - relating corrections ; the application of each of the two corrections may be requested separately in the user interface .
( a review article , providing the details of the dose evaluation in proton therapy , as well as its implementation in eclipse , will appear soon , see ulmer ( 2009 ) . )
a simple rectangular water phantom was created , within which a planning treatment volume ( ptv ) of @xmath207 was arbitrarily outlined .
four one - field treatment plans were subsequently created as follows : * \a ) a plan without block - relating corrections , * \b ) a plan with block - thickness corrections , * \c ) a plan with block - scattering corrections , and * \d ) a plan with both block - relating corrections . in each of these plans ,
a brass block was inserted and fitted to the cross section of the ptv .
subsequently , a dose of @xmath133 gy was delivered to the ptv , using the double - scattering technique of the ncc machine .
the block - relating corrections were estimated on the basis of @xmath208 ; in fact , the results are practically insensitive to the value of @xmath8 , for @xmath209 .
finally , the resulting dose maps were compared ; the dose differences of plans ( b ) , ( c ) , and ( d ) to plan ( a ) were estimated and compared ( figs . [
fig : eclipse1 ] and [ fig : eclipse2 ] ) , leading to the following conclusions . * as expected , the application of the block - thickness corrections results in lower dose values .
this is due to the fact that part of the incident flux is blocked as a result of the nonzero thickness of the block . *
as expected , the application of the block - scattering corrections leads to higher dose values .
this is because some protons , which would otherwise fail to contribute to the fluence ( as they impinge upon the block ) , scatter off the material of the block and ` re - emerge ' at positions in the bore or on the downstream face of the block . *
as far as the delivered dose is concerned , the effects of the block thickness and block scattering ` compete ' one another . in the water phantom used in this section , the block - thickness corrections dominate .
* the presence of the block in the plan of the water phantom used in this section induces effects which amount to a few percent of the prescribed dose ( see fig . [
fig : eclipse2 ] ) .
the largest effects appear in the area neighbouring the border of the block .
it is also worth noticing the characteristic contributions of the scattered protons in the entrance region in the frontal and sagittal views in fig .
[ fig : eclipse2 ] ; the largest part of the dose in the entrance region corresponds to the low - energy component of the scattered beam . as the _ local _ dose , delivered in the entrance region , is significantly smaller than the corresponding value delivered to the target ( which , in fact , is an agrument in favour of the use of protons in radiation therapy ) , the corrections which one has to apply to it , though representing a small fraction of the _ prescribed _ dose , are sizable
the present work deals with corrections which are due to the presence of beam - limiting and beam - shaping devices in proton planning .
the application of these corrections is greatly facilitated by decomposing the effects of two - dimensional objects into one - dimensional , easily - calculable contributions ( miniblocks ) . in the derivation of the thickness corrections , we follow the strategy of slopsema and kooy ( 2006 ) . given the time restrictions during the planning , the derivation of the scattering corrections necessitates the introduction of a two - step approach .
the first step occurs at the beam - configuration phase . at first
, the value of the only parameter of our model ( @xmath58 ) is extracted from the half - block fluence measurements .
a number of monte - carlo runs follow , the output of which consists of the parameters pertaining to convenient parameterisations of the fluence contributions of the scattered protons .
these runs take account of the variability in the block material and thickness , incident energy , and nozzle - equivalent thickness in all the options for which a proton - treatment machine is configured . to enable the easy use of the mc results , the output is put in the form of expansion parameters in two geometrical quantities which are involved in the description of the scattering effects .
the scattering corrections for all the blocks in a particular plan are determined from the results , obtained at beam - configuration phase , via simple interpolations .
the verification of the method should involve the reproduction of dedicated dose measurements . at present , given the lack of such measurements , the only possibility for verification rested on re - using the half - block fluence measurements , formerly analysed to extract the @xmath58 value ; this is a valid option because parts of the input data had been removed from the database to suppress the ( present in the measurements ) block - scattering contributions .
we investigated the goodness of the reproduction of the measurements on the basis of the @xmath153 function and concluded that the inclusion of the scattering effects leads to substantial improvement .
the method presented in this paper was applied to one plan involving a simple water phantom ; the different contributions from the two block - relating effects have been separately presented and compared .
these effects amount to a few percent of the prescribed dose and are significant in the entrance region and in the area neighbouring the border of the block .
the author acknowledges helpful discussions with barbara schaffner concerning the optimal implementation of the method in eclipse .
barbara also modelled and implemented in eclipse the important ( in the entrance region ) dose contributions of the low - energy scattered protons ( i.e. , those with energies below the lowest value used in the particular plan to ` spread out ' the bragg peak ) .
the author is grateful to se byeong lee for providing the original ncc half - block fluence measurements , as well as important information on the data taking .
.[tab : lambda]the weighted averages of the extracted @xmath58 values ( and their statistical uncertainties ) for the eight options of the ncc machine .
the uncertainties are shown only for the sake of completeness ; they have not been taken into account in the results of section [ sec : output ] . [ cols="^,^",options="header " , ] evaluation of the effects of the bl / bsd at the point q of fig .
[ fig : miniblock2 ] ( not shown ) , whose projection onto the bl / bsd plane is the point p , on the basis of four directions ( resulting , in this case , in four miniblocks ) . in this figure ,
the point p lies within the area b.,width=585 ] evaluation of the effects of the bl / bsd at the point q of fig .
[ fig : miniblock2 ] ( not shown ) , whose projection onto the bl / bsd plane is the point p , on the basis of four directions ( resulting , in this case , in five miniblocks ) . in this figure ,
the point p lies outside the area b.,width=585 ] due to one - dimensional nature of the miniblocks , a point lying within the extension ( of a miniblock ) may also receive contributions which are characteristic to points lying in the exterior of the extension.,width=585 ] the description of the kinematics inside the block .
the auxiliary coordinate system introduced in this figure should not be confused with the formal coordinate system of fig .
[ fig : geometry1].,width=585 ] one case of the scattering corrections ( to be applied to the lateral fluence distribution of the pristine tracks ) at @xmath212 mm ( i.e. , at isocentre ) .
the lateral displacement of the block was @xmath155 mm.,width=585 ] the lateral fluence measurements ( continuous line ) corresponding to one energy - net combination of one option of the ncc machine , taken @xmath133 mm away from the downstream face of the block .
the monte - carlo data shown correspond only to the pristine - beam fluence obtained at the same incident - energy , net , and @xmath7 values ; the measurements have been scaled up by a factor which is equal to the ratio of the median values ( of the two distributions ) , estimated over the fluence plateau.,width=585 ] the lateral fluence measurements ( continuous line ) corresponding to one energy - net combination of one option of the ncc machine , taken @xmath133 mm away from the downstream face of the block .
the monte - carlo data shown correspond to the total ( pristine - beam plus scattered - protons ) fluence obtained at the same incident - energy , net , and @xmath7 values ; the measurements have been scaled up by a factor which is equal to the ratio of the median values ( of the two distributions ) , estimated over the fluence plateau.,width=585 ] an alternative way of displaying the contents of figs .
[ fig : rprdctn1 ] and [ fig : rprdctn2 ] ; shown in this figure are the normalised residuals , plotted versus the lateral distance @xmath6 .
evidently , the pristine - beam contribution underestimates the fluence by about @xmath3 to @xmath137 standard deviations for @xmath126 mm @xmath205 mm . after the inclusion of the block - scattering effects , the residuals nicely cluster around @xmath48.,width=585 ] | the present paper pertains to corrections which are due to the presence of beam - limiting and beam - shaping devices in proton planning .
two types of corrections are considered : those originating from the nonzero thickness of such devices ( geometrical effects ) and those relating to the scattering of beam particles off their material .
the application of these two types of corrections is greatly facilitated by decomposing their effects on the fluence into easily - calculable contributions . to expedite the derivation of the scattering corrections ,
a two - step process has been introduced into a commercial product which is widely used in planning ( treatment planning system eclipse@xmath0 ) .
the first step occurs at the beam - configuration phase and comprises the analysis of half - block fluence measurements and the extraction of the one parameter of the model which is used in the description of the beamline characteristics ; subsequently , a number of monte - carlo runs yield the parameters of a convenient parameterisation of the relevant fluence contributions .
the second step involves ( at planning time ) the reconstruction of the parameters ( which are used in the description of the scattering contributions ) via simple interpolations , performed on the results obtained during the beam - configuration phase .
given the lack of dedicated data , the validation of the method is currently based on the reproduction of the parts of the half - block fluence measurements ( i.e. , of the data used as input during the beam configuration ) which had been removed from the database to suppress unwanted ( block - scattering ) contributions ; it is convincingly demonstrated that the inclusion of the scattering effects leads to substantial improvement in the reproduction of the experimental data .
the contributions from the block - thickness and block - scattering effects ( to the fluence ) are presented separately in the case of a simple water phantom ; in this example , the maximal contribution of the two block - relating effects amounts to a few percent of the prescribed dose .
_ keywords _ : particle therapy , proton , block , collimator , slit , corrections , scattering + |
the derangement polynomial of order @xmath6 is an interesting @xmath7-analogue of the number of derangements ( elements without fixed points ) in the symmetric group @xmath8 .
it is defined by the formula @xmath9 where @xmath10 is the number of excedances ( see section [ sec : perms ] for missing definitions ) of @xmath11 and @xmath12 is the set of derangements in @xmath8 .
the polynomial @xmath13 , first studied by brenti @xcite in the context of symmetric functions , has a number of pleasant properties .
for instance , it has symmetric and unimodal coefficients @xcite ( see also ( * ? ? ?
* section 4 ) ( * ? ? ? * section 5 ) @xcite ) and only real roots @xcite .
it can also be expressed as @xmath14 where @xmath15 is the @xmath16-th eulerian polynomial .
we will be concerned with a natural analogue of @xmath13 for the hyperoctahedral group @xmath17 of signed permutations , introduced and studied independently by chen , tang and zhao @xcite and by chow @xcite .
it is defined by the formula @xmath18 where @xmath19 is the number of type @xmath2 excedances of @xmath20 , introduced by brenti @xcite , and @xmath21 is the set of derangements in @xmath17 .
the derangement polynomial @xmath22 shares most of the main properties of @xmath23 .
for instance , it is real - rooted @xcite , hence it has unimodal ( but not symmetric ) coefficients , and satisfies the analogue @xmath24 of ( [ eq : dna ] ) , where @xmath25 is the @xmath16-th eulerian polynomial of type @xmath2 . our first main result is the following combinatorial formula for @xmath22 .
[ thm : main ] we have @xmath26 for @xmath27 , where
@xmath28 , @xmath29 and the sum ranges over all @xmath30 and over all sequences @xmath31 of nonnegative integers which sum to @xmath6 .
chow ( * ? ? ?
* section 4 ) gave an additional proof of the unimodality of @xmath32 by expressing it as a sum of certain nonnegative unimodal polynomials , defined by a symmetric function identity , of a common mode .
theorem [ thm : main ] implies that @xmath3 can be written as a sum of two polynomials with nonnegative , symmetric and unimodal coefficients , whose centers of symmetry differ by a half , and thus provides a new proof of its unimodality , as we now explain .
since @xmath3 has degree @xmath6 and zero constant term , it can be written uniquely in the form @xmath33 where @xmath34 and @xmath35 are polynomials of degrees at most @xmath36 and @xmath6 , respectively , satisfying @xmath37 ( see , for instance , ( * ? ? ? * lemma 2.4 ) for this elementary fact ) .
for the first few values of @xmath6 we have @xmath38 and @xmath39 the following information for the polynomials @xmath34 and @xmath35 and for @xmath3 can be derived from ( [ eq : main ] ) .
[ cor : main ] we have @xmath40 and @xmath41 for @xmath27 , where the sums range over all @xmath30 and over all sequences @xmath42 ( respectively , @xmath43 ) of nonnegative integers which sum to @xmath6
. moreover , @xmath34 and @xmath35 are @xmath44-nonnegative , meaning there exist nonnegative integers @xmath45 and @xmath46 such that @xmath47 and @xmath48 in particular , @xmath34 and @xmath35 are symmetric and unimodal , with center of symmetry @xmath49 and @xmath50 , respectively , and @xmath51 is unimodal with a peak at @xmath52 . much of the motivation behind this paper comes from the theory of subdivisions and local @xmath4-vectors , developed by stanley @xcite , and its extension @xcite .
we recall that the local @xmath4-vector is a fundamental enumerative invariant of a simplicial subdivision ( triangulation ) of the simplex .
an example by stanley ( see ( * ? ? ?
* proposition 2.4 ) ) shows that @xmath13 is equal to the local @xmath4-polynomial of the ( first ) simplicial barycentric subdivision of the @xmath53-dimensional simplex .
this fact gives a geometric interpretation to @xmath23 and another proof of its symmetry and unimodality .
= 3.0 in our second main result provides a type @xmath2 analogue to this interpretation .
to state it , we introduce the following notation .
we denote by @xmath54 the simplicial barycentric subdivision of the cubical barycentric subdivision of the @xmath53-dimensional simplex ( figure [ fig : k3 ] shows this subdivision for @xmath55 ) .
we also introduce the ` half eulerian polynomials ' @xmath56 and @xmath57 for the group @xmath17 , where @xmath58 and @xmath59 are the sets of signed permutations of length @xmath6 with positive and negative , respectively , last entry , and set @xmath60 and @xmath61 ( the set @xmath58 has appeared in the context of major indices for classical weyl groups ; see @xcite ) .
[ thm : localint ] the polynomial @xmath34 is equal to the local @xmath4-polynomial of the simplicial subdivision @xmath54 ( in particular , @xmath34 has nonnegative , symmetric and unimodal coefficients ) .
moreover , we have @xmath62 and @xmath63 for @xmath27
. we should point out that it is theorem [ thm : localint ] and the methods of @xcite which led the authors to suspect that formula ( [ eq : main ] ) holds .
indeed , it follows from the relevant definitions and some more work ( see section [ sec : geom ] ) that the local @xmath4-polynomial of @xmath54 is equal to the right - hand side of ( [ eq : localint+ ] ) . by exploiting the symmetry of this polynomial and certain recurrence relations for that and for @xmath3 ( see section [ sec : half ] ) , one can show that the local @xmath4-polynomial of @xmath54 is equal to @xmath34 , as defined by the decomposition ( [ eq : dnbsymsum ] ) .
a formula for the change in the local @xmath4-vector of a simplicial subdivision of the simplex after further subdivision ( * ? ? ? * proposition 3.6 ) ( see also proposition [ prop : localrelformula ] ) can then be used to produce equation ( [ eq : main+ ] ) .
this suggested that ( [ eq : main- ] ) , and hence ( [ eq : main ] ) , hold as well . the structure and other results of this paper are as follows .
section [ sec : perms ] provides the necessary background on ( signed ) permutations , simplicial complexes and subdivisions .
section [ sec : proof ] proves theorem [ thm : main ] and corollary [ cor : main ] .
a bijective proof of theorem [ thm : main ] , as well as one using generating functions , is given and the exponential generating functions of @xmath34 and @xmath64 are computed .
section [ sec : comb ] gives a combinatorial interpretation to the coefficients of these polynomials .
section [ sec : relative ] proves the main properties of the relative local @xmath4-vector , a generalization of the concept of local @xmath4-vector which was introduced in ( * ? ? ?
* section 3 ) ( and , in a variant form , in @xcite ) and derives a monotonicity property for local @xmath4-vectors .
these results were stated without proof in @xcite . as an example ( used in one of the proofs of theorem [ thm : localint ] ) , the relative local @xmath4-vector of the barycentric subdivision of the simplex is computed .
section [ sec : geom ] gives two proofs of theorem [ thm : localint ] .
a first step towards these proofs is to interpret @xmath65 as the @xmath4-polynomial of the simplicial complex @xmath54 ( proposition [ prop : knhpoly ] ) . given that , one proof uses the theory of ( relative ) local @xmath4-vectors , as discussed earlier , while the other uses recurrences and generating functions . section [ sec : half ] studies the polynomials @xmath65 and @xmath66 .
a simple relation between the two is shown to hold ( lemma [ lem : rec ] ) . using its interpretation as the @xmath4-polynomial of @xmath54 and the theory of local @xmath4-vectors ,
a simple formula for @xmath65 ( hence one for @xmath67 and one for the eulerian polynomial @xmath68 ) in terms of the eulerian polynomial @xmath69 is proven ( proposition [ prop : bn+formula ] ) . using this formula ,
it is shown that @xmath65 and @xmath66 are real - rooted , hence unimodal and log - concave , and a new proof of the unimodality of @xmath68 is deduced .
recurrences and generating functions for @xmath65 and @xmath66 , as well as for @xmath34 and @xmath35 , are also given and a third proof of theorem [ thm : localint ] is deduced .
this section fixes notation and includes background material on ( signed ) permutations , simplicial complexes and their subdivisions . for more information on these topics ,
the reader is referred to @xcite . throughout this paper
, @xmath70 denotes the set of nonnegative integers . for each positive integer @xmath6 we set @xmath71:= \{1 , 2,\dots , n\}$ ] and @xmath72 .
we denote by @xmath73 the cardinality , and by @xmath74 the set of all subsets , of a finite set @xmath75 . a _ permutation _ of a finite set @xmath75 is a bijective map @xmath76 .
we denote by @xmath77 the set of all permutations of @xmath75 and set @xmath78)$ ] .
suppose that @xmath79 has @xmath6 elements , which are totally ordered by @xmath80 .
a permutation @xmath81 can be represented as the sequence @xmath82 , or as the word @xmath83 , or as a disjoint union of cycles ( * ? ? ?
* section 1.3 ) .
the _ standard cycle form _ is defined by requiring that ( a ) each cycle is written with its largest element ( with respect to the total order @xmath84 ) first and ( b ) the cycles are written in increasing order of their largest element @xcite . given @xmath85 , an element @xmath86 is called an _ excedance _ of @xmath87 ( with respect to @xmath84 ) if @xmath88 and an _ inverse excedance _ if @xmath89 . the element @xmath90 is called a _ descent _
( respectively , _ ascent _ ) of @xmath87 if @xmath91 $ ] and @xmath92 ( respectively , @xmath93 ) .
the number of excedances ( respectively , inverse excedances , descents or ascents ) of @xmath87 will be denoted by @xmath94 ( respectively , by @xmath95 , @xmath96 or @xmath97 ) .
the @xmath6th eulerian polynomial ( * ? ? ?
* section 1.4 ) is defined by the formulas @xmath98 clearly , these sums depend only on @xmath6 and not on @xmath75 or the choice of total order @xmath84 .
the previous definitions apply in particular to @xmath8 ( with the standard choice of @xmath84 obtained by setting @xmath99 for @xmath100 ) .
we will denote by @xmath12 the set of all derangements ( permutations without fixed points ) in @xmath8 . for the purposes of this paper
, it will be convenient to define a _ signed permutation _ of @xmath71 $ ] as a choice of a subset @xmath79 of @xmath101 such that @xmath102 for @xmath103 and permutation @xmath104 .
we will represent such a permutation @xmath87 as the sequence @xmath105 , or as the word @xmath83 , or as a disjoint union of cycles
. we will find it convenient to define the standard cycle form of @xmath87 using the total order on @xmath75 which is the reverse of the one inherited from the natural total order on @xmath106 .
thus , cycles of @xmath87 will be written with their _ smallest _ element first and in _ decreasing _ order of their smallest element .
we will say that @xmath87 is a _ derangement _ if there is no @xmath107 $ ] such that @xmath108 .
we will denote the set of all signed permutations of @xmath71 $ ] by @xmath17 and the set of all derangements in @xmath17 by @xmath21 .
given @xmath20 as before , we say that @xmath109 is a _ @xmath2-descent _
( respectively , @xmath2-_ascent _ ) of @xmath87 if @xmath110 ( respectively , @xmath111 ) , where @xmath112 by convention . the @xmath6th eulerian polynomial of type @xmath2 ( * ? ? ?
* section 3 ) can be defined by @xmath113 where @xmath114 stands for the number of @xmath2-descents and @xmath115 for the number of @xmath2-ascents of @xmath20 .
following brenti @xcite , we say that @xmath86 is a _
@xmath2-excedance _ of @xmath87 if @xmath116 , or if @xmath117 $ ] and @xmath108 .
we say that @xmath86 is an _
inverse @xmath2-excedance _ of @xmath87 if @xmath118 , or if @xmath119 $ ] and @xmath108
. the number of @xmath2-excedances of @xmath87 will be denoted by @xmath19 and that of inverse @xmath2-excedances by @xmath120 .
we then have @xmath121 and ( see theorem 3.15 and corollary 3.16 in @xcite ) @xmath122 the @xmath6th derangement polynomial of type @xmath2 is defined by ( [ eq : dnbdef ] ) . since @xmath123 and the map which sends a permutation @xmath124 to its inverse @xmath125 induces an involution on @xmath17 which preserves fixed points , we have @xmath126 for the similar reasons , ( [ eq : dndef ] ) continues to hold if @xmath127 is replaced by @xmath128 and ( [ eq : eulerbdef2 ] ) continues to hold if @xmath129 is replaced by @xmath130 .
let @xmath131 be a polynomial with real coefficients .
we recall that @xmath132 is unimodal ( and has unimodal coefficients ) if there exists an index @xmath133 such that @xmath134 for @xmath135 and @xmath136 for @xmath137 .
such an index is called a _
peak_. the polynomial @xmath132 is said to be log - concave if @xmath138 for @xmath139 and to have internal zeros if there exist indices @xmath140 such that @xmath141 and @xmath142 .
we will say that @xmath132 is symmetric ( and that it has symmetric coefficients ) if there exists an integer @xmath143 such that @xmath144 for @xmath145 .
the _ center of symmetry _ of @xmath132 is then defined to be @xmath49 ( this is well - defined provided @xmath132 is nonzero ) .
we will say that @xmath132 is _ real - rooted _ if all its complex roots are real .
it is well - known ( see , for instance , @xcite ) that if @xmath132 is a real - rooted polynomial with nonnegative coefficients , then @xmath132 is log - concave and unimodal , with no internal zeros .
the following theorem , first proved by edrei @xcite , gives a necessary and sufficient condition for a polynomial with nonnegative real coefficients to be real - rooted .
( @xcite ) [ thm : realroots ] let @xmath146 $ ] be a polynomial with @xmath147 for every @xmath30 and set @xmath148 for all negative integers @xmath16 .
then @xmath132 is real - rooted if and only if every minor of the lower triangular matrix @xmath149 is nonnegative .
a ( nonzero ) symmetric polynomial @xmath150 $ ] can be written ( uniquely ) in the form @xmath151 for some polynomial @xmath152 .
we say that @xmath132 is _ @xmath44-nonnegative _ if @xmath153 for every @xmath154 . clearly , every @xmath44-nonnegative polynomial is unimodal . for classes of @xmath44-nonnegative polynomials which appear in combinatorics
we refer the reader , for instance , to @xcite and references therein . an ( _ abstract _ ) _ simplicial complex
_ @xmath155 on the ground set @xmath156 is a collection @xmath155 of subsets of @xmath156 such that @xmath157 implies @xmath158 ( all simplicial complexes considered in this paper will be assumed to be finite ) .
the elements of @xmath155 are called _
faces_. the _ dimension _ of a face is equal to one less than its cardinality .
the _ dimension _ of @xmath155 is the maximum dimension of its faces .
faces of dimension 0 and 1 are called _ vertices _ and _ edges _ , respectively .
a _ facet _ of @xmath155 is a face which is maximal with respect to inclusion .
the complex @xmath155 is said to be _ pure _ if all its facets have the same dimension .
the _ face poset _
@xmath159 of a simplicial complex @xmath155 is the set of nonempty faces of @xmath155 , partially ordered by inclusion .
the _ open star _
@xmath160 of a face @xmath158 is the collection of all faces of @xmath155 containing @xmath161 .
the _ link _ of a face @xmath161 in @xmath155 is the subcomplex of @xmath155 defined as @xmath162 .
suppose that @xmath163 and @xmath164 are simplicial complexes on disjoint ground sets .
the _ simplicial join _ of @xmath163 and @xmath164 is the simplicial complex @xmath165 whose faces are the sets of the form @xmath166 , where @xmath167 and @xmath168 .
the _ order complex _
* section 9.3 ) ( * ? ? ?
* section 3.8 ) of a ( finite ) partially ordered set @xmath169 is defined as the simplicial complex of chains ( totally ordered subsets ) of @xmath169 .
all topological properties or invariants of @xmath155 mentioned in the sequel will refer to those of its geometric realization @xmath170 ( * ? ? ?
* section 9.1 ) .
for example , @xmath155 is a _ simplicial ball _ if @xmath170 is homeomorphic to a ball . for a simplicial @xmath171-dimensional ball @xmath155 ,
we denote by @xmath172 the subcomplex consisting of all subsets of the @xmath173-dimensional faces which are contained in a unique facet of @xmath155 .
we call @xmath172 the _ boundary _ and @xmath174 the _ interior _ of @xmath155 .
let @xmath155 be a simplicial complex .
a ( _ topological _ ) _ simplicial subdivision _ of @xmath155 ( * ? ? ?
* section 2 ) is a simplicial complex @xmath175 together with a map @xmath176 such that the following hold for every @xmath158 : ( a ) the set @xmath177 is a subcomplex of @xmath175 which is a simplicial ball of dimension @xmath178 ; and ( b ) the interior of @xmath179 is equal to @xmath180 .
the subcomplex @xmath179 is called the _ restriction _ of @xmath175 to @xmath161 . the face @xmath181 is called the _ carrier _ of @xmath182 .
the subdivision @xmath175 is called _ quasi - geometric _
* definition 4.1 ( a ) ) if no face of @xmath175 has the carriers of its vertices contained in a face of @xmath155 of smaller dimension .
moreover , @xmath175 is called _ geometric _ ( * ? ? ?
* definition 4.1 ( b ) ) if there exists a geometric realization of @xmath175 which geometrically subdivides a geometric realization of @xmath155 , in the way prescribed by @xmath183 .
clearly , all geometric subdivisions ( such as the barycentric subdivisions considered in this paper ) are quasi - geometric .
we now describe two common ways to subdivide a simplicial complex @xmath155 .
the order complex of the face poset @xmath184 , denoted by @xmath185 , consists of the chains of nonempty faces of @xmath155 .
this complex is naturally a ( geometric ) simplicial subdivision of @xmath155 , called the _ barycentric subdivision _
, where the carrier of a chain @xmath186 of nonempty faces of @xmath155 is defined as the maximum element of @xmath186
. given a face @xmath158 we set @xmath187 , where @xmath188 is a new vertex added and @xmath189 .
then @xmath175 is a simplicial complex which is a simplicial subdivision of @xmath155 , called the _
stellar subdivision _ of @xmath155 on @xmath161 .
let @xmath155 be a @xmath173-dimensional simplicial complex .
we denote by @xmath190 the number of @xmath154-dimensional faces of @xmath155 .
a fundamental enumerative invariant of @xmath155 is the _
@xmath191-polynomial _ , defined by @xmath192 the _ @xmath4-polynomial _ of @xmath155 is defined by @xmath193 for the importance of @xmath4-polynomials , the reader is referred to ( * ? ? ?
* chapter ii ) . for the simplicial join @xmath165 of two simplicial complexes we have @xmath194 .
let @xmath195 be a simplicial subdivision of a @xmath173-dimensional simplex @xmath196 .
the polynomial @xmath197 defined by @xmath198 is the _
local @xmath4-polynomial _ of @xmath195 ( with respect to @xmath156 ) ( * ? ? ?
* definition 2.1 ) .
the sequence @xmath199 is the
_ local @xmath4-vector _ of @xmath195 ( with respect to @xmath156 ) .
the following theorem summarizes some of the main properties of local @xmath4-vectors ( see theorems 3.2 and 3.3 and corollary 4.7 in @xcite ) . for the definition of regular subdivision we refer the reader to (
* definition 5.1 ) .
( stanley @xcite ) [ thm : stalocal ] * for every simplicial subdivision @xmath175 of a pure simplicial complex @xmath155 we have @xmath200 * the local @xmath4-polynomial @xmath201 is symmetric for every simplicial subdivision @xmath195 of the simplex @xmath196 , i.e. we have @xmath202 for @xmath203 . *
the local @xmath4-polynomial @xmath201 has nonnegative coefficients for every quasi - geometric simplicial subdivision @xmath195 of the simplex @xmath196 . *
the local @xmath4-polynomial @xmath201 has unimodal coefficients for every regular simplicial subdivision @xmath195 of the simplex @xmath196 .
this section gives two proofs of theorem [ thm : main ] , one bijective and one using generating functions , and deduces corollary [ cor : main ] . as a byproduct of the second proof ,
the exponential generating functions of @xmath34 and @xmath35 are computed .
let us denote by @xmath204 the collection of sequences @xmath205 of permutations , where @xmath30 and @xmath206 for @xmath207 , such that @xmath208 is a weak ordered partition of @xmath71 $ ] with @xmath209 nonempty for @xmath210 and @xmath211 is a derangement of @xmath212 .
we will describe a one - to - one correspondence @xmath213 such that @xmath214 for every @xmath215 , where @xmath216 and @xmath217 stands for @xmath218 or @xmath219 , if @xmath154 is even or odd , respectively .
given this , using ( [ eq : dnbidef ] ) and recalling that there are @xmath220 weak ordered partitions @xmath221 of @xmath71 $ ] satisfying @xmath222 for @xmath223 , we get @xmath224 and the proof follows . to define @xmath225 , consider a derangement @xmath215 and let @xmath226 be the standard cycle form of @xmath87 .
then there is an index @xmath227 such that all elements of @xmath228 are positive and the first ( smallest ) element of @xmath229 is negative . we define @xmath211 as the product of @xmath228 and @xmath212 as the set of all elements which appear in these cycles , so that @xmath230 is a derangement .
the remaining cycles @xmath231 form a word @xmath232 whose first element is negative .
this word decomposes uniquely as a product @xmath233 of subwords @xmath234 so that for @xmath210 , all elements of @xmath234 are negative if @xmath154 is odd and positive if @xmath154 is even .
we define @xmath209 as the set of absolute values of the elements of @xmath234 and @xmath206 as the permutation which corresponds to the word @xmath234 . for instance , if @xmath235 and @xmath236 in standard cycle form , then @xmath237 in cycle form , @xmath238 and @xmath239 , @xmath240 , @xmath241 , as sequences .
we set @xmath242 and leave it to the reader to verify that the map @xmath243 is a well defined bijection . to verify ( [ eq : bicondition ] ) we let @xmath215 with @xmath244 and @xmath245 be the word defined in the previous paragraph .
then , by the definitions of standard cycle form and ( inverse ) @xmath2-excedance , @xmath246 is an inverse @xmath2-excedance of @xmath87 if and only if @xmath247 is an inverse excedance of @xmath211 , or @xmath248 for some index @xmath249 with @xmath250 , or @xmath251 .
thus , equation ( [ eq : bicondition ] ) follows . for the second proof of theorem [ thm : main ] we set @xmath252 and ( see ( * ? ? ?
* proposition 5 ) ) @xmath253 where @xmath29 .
we also recall ( see ( * ? ? ?
* theorem 3.3 ) ( * ? ? ? * theorem 3.2 ) ) that @xmath254 where @xmath255 .
we denote by @xmath256 ( respectively , by @xmath257 and @xmath258 ) the right - hand side of ( [ eq : main ] ) ( respectively , of ( [ eq : main+ ] ) and ( [ eq : main- ] ) ) , so that @xmath259 for @xmath27 .
we compute that @xmath260 and similarly that @xmath261 and conclude that @xmath262 combining the previous equation with ( [ eq : anexp ] ) and ( [ eq : dnexp ] ) we get , after some straightforward algebraic manipulations , that @xmath263 and the proof follows . as in the second proof of theorem [ thm :
main ] , we denote by @xmath257 and @xmath258 the right - hand side of ( [ eq : main+ ] ) and ( [ eq : main- ] ) , respectively .
theorem [ thm : main ] shows that @xmath264 for every @xmath27 . from the symmetry properties @xmath265 and @xmath266 of the eulerian and derangement polynomials for @xmath8 it follows that @xmath257 and @xmath258 satisfy ( [ eq : symf+ ] ) and ( [ eq : symf- ] ) , respectively .
the uniqueness of the defining properties of @xmath34 and @xmath35 imply that @xmath267 and @xmath268 for every @xmath27 .
this proves equations ( [ eq : main+ ] ) and ( [ eq : main- ] ) .
the @xmath44-nonnegativity of @xmath34 and @xmath35 follows from equations ( [ eq : main+ ] ) and ( [ eq : main- ] ) and the @xmath44-nonnegativity of @xmath69 and @xmath0 ( see proposition [ prop : xi+-formula ] in the sequel ) .
the last statement in the corollary follows from ( [ eq : main++ ] ) and ( [ eq : main ] ) .
since the polynomials @xmath69 and @xmath0 have nonnegative and symmetric coefficients and only real roots , we can write @xmath269 and @xmath270 for some polynomials @xmath271 and @xmath272 with nonnegative coefficients . explicit combinatorial interpretations to these coefficients are known ( see , for instance , ( * ? ? ?
* theorem 5.6 ) and ( * ? ? ? * section 4 ) ) .
equations ( [ eq : main+ ] ) , ( [ eq : main- ] ) , ( [ eq : angamma ] ) and ( [ eq : dngamma ] ) imply explicit combinatorial formulas for the polynomials @xmath273 and @xmath274 , appearing in corollary [ cor : main ] , which we record in the following proposition .
[ prop : xi+-formula ] we have @xmath275 and @xmath276 where @xmath277 @xmath278 the sums in the previous equations range as in ( [ eq : main+ ] ) and ( [ eq : main- ] ) , respectively , and @xmath279 , @xmath280 .
for the first few values of @xmath6 we have @xmath281 and @xmath282 we are not aware of any combinatorial interpretations for the coefficients of @xmath283 or @xmath284 . the second proof of theorem [ thm : main ] and the proof of corollary [ cor : main ] yield the following explicit formulas for the exponential generating functions of @xmath34 and @xmath35 .
[ prop : f+-exp ] we have @xmath285 and @xmath286 we noticed in the proof of corollary [ cor : main ] that @xmath287 and @xmath268 .
thus , the result follows from the formulas in the second proof of theorem [ thm : main ] and straightforward algebraic manipulations .
this section gives a combinatorial interpretation to the coefficients of @xmath34 and @xmath35 by exploiting the first proof of theorem [ thm : main ] , given in section [ sec : proof ] .
consider a signed permutation @xmath85 , where @xmath288 is as in section [ subsec : signed ] .
we denote by @xmath289 the minimum element of @xmath75 with respect to the natural total order inherited from @xmath106 and set @xmath290 .
[ prop : f+-interpret ] we have @xmath291 and @xmath292 for every @xmath293
. we will follow the setup of the first proof of theorem [ thm : main ] . given @xmath294 with @xmath295 , we observe that @xmath16 is even if and only if the last element in the standard cycle form of @xmath87 is positive
. therefore , equation ( [ eq : main+ ] ) and the argument in the proof of theorem [ thm : main ] show that @xmath296 where the sum ranges over all @xmath215 for which the last element in the standard cycle form is positive .
since this element equals @xmath297 , we get @xmath298 equation ( [ eq : fn - interpret ] ) follows from ( [ eq : fn+interpret ] ) and ( [ eq : dnbdef ] ) , or by a similar argument .
this section reviews the definition of the relative local @xmath4-polynomial of a simplicial subdivision of a simplex , introduced in ( * ? ? ?
* section 3 ) and , independently ( in a different level of generality ) , in @xcite , and establishes some of its main properties ( most of them stated without proof in ( * ? ? ? * section 3 ) ) .
the relative local @xmath4-polynomial of the barycentric subdivision of the simplex is also computed ( example [ ex : barycentrel ] ) .
this computation will be used in section [ sec : geom ] .
we will fix a field @xmath299 in this section and work with the notion of a homology ( rather than topological ) simplicial subdivision over @xmath299 , as in @xcite .
thus , in the definition of a subdivision @xmath176 we require that the subcomplex @xmath300 of @xmath175 is a homology ( rather than topological ) ball over @xmath299 of dimension @xmath178 , for every @xmath301 ; see ( * ? ? ?
* section 2 ) for details .
the following concept was introduced in ( * ? ? ?
* remark 3.7 ) and ( for regular triangulations of polytopes ) in @xcite .
( ( * ? ? ?
* section 3 ) ) [ def : relative ] let @xmath195 be a homology subdivision of a @xmath173-dimensional simplex @xmath196 , with subdivision map @xmath302 , and let @xmath303 .
the polynomial @xmath304 is the _ relative local @xmath4-polynomial _ of @xmath195 ( with respect to @xmath156 ) at @xmath305 .
thus , @xmath306 reduces to the local @xmath4-polynomial @xmath307 for @xmath308 .
[ ex : barycentrel ] let @xmath309 be the barycentric subdivision of an @xmath53-dimensional simplex @xmath196 and @xmath310 be a face of @xmath195 , where @xmath311 are nonempty sets .
we will show that @xmath312 where @xmath313 and @xmath314 for @xmath210 ( with the convention @xmath315 ) .
we recall from section [ subsec : sub ] that the carrier of @xmath305 in @xmath195 is given by @xmath316 .
thus the right - hand side of ( [ eq : deflocalhrel ] ) is a sum over all @xmath317 .
the restriction @xmath318 is the barycentric subdivision of @xmath319 and the link of @xmath305 in this restriction satisfies @xmath320 , where @xmath321 is the simplicial complex of all chains of subsets of @xmath156 which strictly contain @xmath322 and are strictly contained in @xmath209 , for @xmath210 , and @xmath323 is the simplicial complex of all chains of subsets of @xmath156 which strictly contain @xmath324 and are strictly contained in @xmath161 . as a result
, we have @xmath325 multiplying this equation with @xmath326 , summing over all @xmath327 and using ( [ eq : dna ] ) we get ( [ eq : barycentrel ] ) . our motivation for introducing the relative local @xmath4-polynomial comes from the following statement ( for another motivation , see ( * ? ? ?
* section 3 ) ) .
( ( * ? ? ?
* proposition 3.6 ) ) [ prop : localrelformula ] for every homology subdivision @xmath195 of the simplex @xmath196 and every homology subdivision @xmath328 of @xmath195 we have @xmath329 we now confirm that the polynomial @xmath306 shares two of the main properties of @xmath201 and deduce a monotonicity property of local @xmath4-vectors .
these results were stated without proof in ( * ? ? ?
* remark 3.7 ) . here
we will sketch the proof , which follows closely ideas of @xcite and their refinements in @xcite .
for that reason , we will assume familiarity with the corresponding proofs in @xcite .
[ thm : relative ] let @xmath156 be a set with @xmath171 elements .
* the relative local @xmath4-polynomial @xmath306 has symmetric coefficients , in the sense that @xmath330 for every homology subdivision @xmath195 of the simplex @xmath196 and every @xmath331 . *
the relative local @xmath4-polynomial @xmath306 has nonnegative coefficients for every quasi - geometric homology subdivision @xmath195 of the simplex @xmath196 and every @xmath303 .
\(a ) the proof of ( * ? ? ?
* theorem 4.2 ) can be adapted as follows . using the defining equation ( [ eq : deflocalhrel ] ) and ( * ? ? ?
* proposition 2.1 ) , we find that @xmath332 an inclusion - exclusion argument , similar to the one in the proof of ( * ? ? ? * ( 4.3 ) ) , shows that @xmath333 replacing @xmath334 in the first formula by the right - hand side of the previous equation and changing the order of summation , as in the proof of ( * ? ? ?
* theorem 4.2 ) , results in ( [ eq : relsymm ] ) .
\(b ) the special case @xmath308 is equivalent to part ( iii ) of ( * ? ? ?
* theorem 3.3 ) ( essentially , part ( c ) of theorem [ thm : stalocal ] ) .
the general case follows by the argument in the proof of ( * ? ? ?
* theorem 5.1 ) ( generalizing that in the proof of ( * ? ? ?
* theorem 4.6 ) ) , where the role of @xmath155 in that proof is played by @xmath335 , the role of @xmath171 is played by @xmath336 and the role of @xmath337 is played by the rank latexmath:[$d - lattice of subsets of @xmath156 . for polynomials
@xmath339 $ ] we write @xmath340 if the difference @xmath341 has nonnegative coefficients .
[ cor : localmonotone ] for every quasi - geometric homology subdivision @xmath195 of the simplex @xmath196 and every quasi - geometric homology subdivision @xmath328 of @xmath195 , we have @xmath342 .
the right - hand side of ( [ eq : localrelformula ] ) reduces to @xmath343 for @xmath308 .
the other terms in the sum are nonnegative by theorems [ thm : stalocal ] ( c ) and [ thm : relative ] ( b ) and the proof follows .
this section formally defines the simplicial subdivision @xmath54 and gives two proofs of theorem [ thm : localint ] , one using the theory of ( relative ) local @xmath4-vectors ( specifically , proposition [ prop : localrelformula ] ) and another using generating functions .
let @xmath155 be a simplicial complex .
the _ cubical barycentric subdivision _
( see , for instance , ( * ? ? ?
* section 2.3 ) ) of @xmath155 , denoted @xmath344 , is defined as the set of all nonempty closed intervals @xmath345 $ ] in the face poset @xmath184 , partially ordered by inclusion .
it follows from ( * ? ? ?
* theorem 6.1 ( a ) ) and ( * ? ? ?
* equation ( 3.24 ) ) that the order complex , say @xmath175 , of @xmath344 is homeomorphic to @xmath155 .
moreover , @xmath175 is naturally a simplicial subdivision of @xmath155 : the carrier of a face of @xmath175 is the maximum element of the largest of the intervals in the corresponding chain of intervals of @xmath184 .
we will denote by @xmath54 the order complex of @xmath346})$ ] , so that @xmath54 is a simplicial subdivision of the simplex @xmath347}$ ] ( see figure [ fig : k3 ] for the case @xmath55 ) .
we note that @xmath54 is the special case @xmath348 of a subdivision of the simplex considered in @xcite .
the following statement is an essential step for both proofs of theorem [ thm : localint ] which will be given in this section .
[ prop : knhpoly ] we have @xmath349 for @xmath27 .
the poset @xmath350})$ ] consists of all intervals of the form @xmath351 $ ] , where @xmath352 $ ] , partially ordered by inclusion . to describe this poset differently ,
we consider the following poset @xmath353 .
the elements of @xmath354 are the subsets of @xmath101 which contain at least one positive number and at most one number from each set @xmath355 for @xmath356 ; the partial order is reverse inclusion .
we observe that the map @xmath357 } ) \to p_n$ ] defined by @xmath358 ) = \ a \ , \cup \ , ( -([n ] \sm b ) )
$ ] is a poset isomorphism .
thus , we may identify @xmath54 with the order complex of @xmath354 . for @xmath359 $ ] with @xmath360
, we define @xmath361 as the number of chains @xmath362 in @xmath354 such that @xmath363 for @xmath364 .
the map @xmath365 } \rightarrow \nn$ ] is the flag @xmath191-vector of @xmath354 ; see ( * ? ? ?
* section 3.13 ) .
the chains of @xmath354 enumerated by @xmath361 are in one - to - one correspondence with the elements @xmath366 for which @xmath367 .
indeed , given such a chain , the corresponding element of @xmath58 consists of the elements of @xmath71 \sm \{|s| : s \in f_1\}$ ] in increasing order , followed by those of @xmath368 in increasing order and so on , followed at the end by the elements of @xmath369 in increasing order .
recall that the flag @xmath4-vector @xmath370 } \rightarrow \zz$ ] of @xmath354 is defined by @xmath371 for @xmath372 $ ] , or equivalently , by @xmath373 for @xmath374 $ ] .
since @xmath361 enumerates signed permutations @xmath375 for which @xmath376 , by the principle of inclusion - exclusion we get that @xmath377 enumerates signed permutation @xmath378 for which @xmath379 .
the result follows from this interpretation by recalling ( * ? ? ?
* section 3.13 ) that @xmath380 , |s|=k } \beta_{p_n}(s)\ ] ] and switching @xmath75 to @xmath381 in the previous equation .
our first proof of theorem [ thm : localint ] will be based on the fact that @xmath54 can be viewed as a subdivision of the barycentric subdivision @xmath382})$ ] . to explain how
, we consider the following setup .
let @xmath383 be a set totally ordered by @xmath384 .
we recall that @xmath385 denotes the poset of intervals in @xmath156 of the form @xmath386 = \{v_i , v_{i+1},\dots , v_j\}$ ] for @xmath387 , partially ordered by inclusion .
we denote by @xmath195 the order complex of @xmath385 , consisting of all chains of such intervals .
for such a chain @xmath388 , we define @xmath389 as the set of all endpoints of the intervals in @xmath390 .
thus we have a well defined map @xmath302 .
= 2.0 in [ lem : knbary ] under the previous assumptions and notation , the map @xmath302 turns @xmath195 into a geometric subdivision of @xmath196 .
the number of facets of @xmath195 is equal to @xmath391 , where @xmath171 is the number of elements of @xmath156 .
let @xmath392 be a geometric @xmath173-dimensional simplex whose vertices are labeled by the singleton subsets of @xmath393 .
we will construct a geometric simplicial subdivision ( triangulation ) @xmath394 of @xmath392 whose vertices are labeled ( in a one - to - one fashion ) with the closed intervals in the total order @xmath156 , so that : ( a ) the singleton intervals label the vertices of @xmath392 ; ( b ) the point labeled by a non - singleton interval @xmath395 \in \sd_c ( v)$ ] lies in the relative interior of the edge of @xmath392 whose endpoints are labeled by @xmath396 and @xmath397 ; and ( c ) the faces of @xmath394 correspond to the chains of intervals ( see figure [ fig : int3 ] for the case @xmath398 ) .
we proceed by induction on @xmath171 .
the triangulation @xmath394 is a single point for @xmath399 and the triangulation of a line segment with one interior point ( labeled by @xmath400 ) for @xmath401 .
we assume @xmath402 and set @xmath403 and @xmath404 .
we choose the simplices @xmath405 and @xmath406 as the codimension one faces of @xmath392 which correspond to @xmath407 and @xmath156 and , using the inductive hypothesis , triangulations @xmath408 and @xmath409 of these two simplices having properties ( a ) , ( b ) and ( c ) with respect to the totally ordered subsets @xmath407 and @xmath410 of @xmath156 , respectively .
clearly , we may choose these triangulations to have the same restriction on the face @xmath411 of @xmath392 .
we then label by @xmath156 an arbitrary point @xmath412 in the relative interior of the edge of @xmath392 whose endpoints are labeled with @xmath413 and @xmath414 and define @xmath394 as the collection consisting of all simplices in @xmath415 and the cones of these on the vertex @xmath412 .
we leave it to the reader to verify that @xmath394 has properties ( a ) , ( b ) and ( c ) and that it realizes an abstract simplicial subdivision of @xmath196 with the required properties .
we now recall that @xmath54 consists of all chains of intervals of the form @xmath351 $ ] , where @xmath416 $ ] .
we define the carrier of such a chain @xmath186 as the set of all endpoints of the intervals in @xmath186 and note that this set is a chain in the poset @xmath417})$ ] of nonempty subsets of @xmath71 $ ] and hence belongs to the barycentric subdivision @xmath418})$ ] . applying lemma [ lem : knbary ] to an arbitrary chain @xmath419})$ ] we conclude that @xmath54 is a subdivision of @xmath418})$ ] and that the restriction of this subdivision to a nonempty face @xmath420})$ ] of dimension @xmath421 has exactly @xmath391 facets . [ lem:2^n-1facets ]
let @xmath195 be a quasi - geometric simplicial subdivision of a @xmath173-dimensional simplex @xmath196 .
if the restriction @xmath318 has exactly @xmath422 facets for every nonempty face @xmath161 of @xmath196 , then @xmath423 we recall that the number of facets of a simplicial complex @xmath155 is equal to the value of the @xmath4-polynomial @xmath424 at @xmath425 .
thus , setting @xmath425 in the defining equation ( [ eq : deflocalh ] ) and using the assumption on @xmath195 , we find that @xmath426 and the result follows from parts ( b ) and ( c ) of theorem [ thm : stalocal ] .
let us denote by @xmath427 the local @xmath4-polynomial of @xmath54 . to compute this polynomial
, we will apply proposition [ prop : localrelformula ] to @xmath428 and @xmath429})$ ] .
let @xmath310 be a face of @xmath195 with @xmath16 elements , where @xmath430 $ ] are nonempty sets .
we have already noted that the restriction @xmath431 satisfies the assumptions of lemma [ lem:2^n-1facets ] .
thus , by lemma [ lem:2^n-1facets ] we have @xmath432 the relative local @xmath4-vector of @xmath195 was computed in example [ ex : barycentrel ] .
thus , in view of ( [ eq:2^k-1facets ] ) and ( [ eq : barycentrel ] ) , proposition [ prop : localrelformula ] yields that @xmath433 where the sum ranges over all even numbers @xmath30 and over all sequences @xmath31 of nonnegative integers which sum to @xmath6 .
this equation and ( [ eq : main+ ] ) imply that @xmath434 and the first statement of theorem [ thm : localint ] follows .
we leave to the reader to verify that @xmath54 can be obtained from the trivial subdivision of the simplex by successive stellar subdivisions .
this implies that @xmath54 is a regular subdivision .
the claim that @xmath34 has nonnegative , symmetric and unimodal coefficients follows from the main properties of local @xmath4-polynomials @xcite ( see theorem [ thm : stalocal ] ) .
equation ( [ eq : localint+ ] ) follows from the fact that @xmath435 , the defining equation ( [ eq : deflocalh ] ) of local @xmath4-polynomials and proposition [ prop : knhpoly ] .
given that @xmath436 and @xmath437 for every @xmath6 , equation ( [ eq : localint- ] ) is a consequence of ( [ eq : dnb ] ) and ( [ eq : localint+ ] ) .
for the second proof of theorem [ thm : localint ] we will need the exponential generating functions of @xmath65 and @xmath66 .
these will be computed in section [ sec : half ] .
let us denote by @xmath427 and @xmath438 the right - hand side of ( [ eq : localint+ ] ) and ( [ eq : localint- ] ) , respectively .
proposition [ prop : knhpoly ] and ( [ eq : deflocalh ] ) imply that @xmath427 is equal to the local @xmath4-polynomial of @xmath54 .
thus , we need to show that @xmath439 and @xmath440 for every @xmath6 . from the definition of @xmath427 and @xmath438 and
proposition [ prop : b+-nexp ] we get @xmath441 and @xmath442 the result follows from these equations and proposition [ prop : f+-exp ] .
this section studies the decomposition of the eulerian polynomial @xmath68 as a sum of @xmath65 and @xmath66 .
first , it is observed that a simple relation between the two summands holds .
then , using the theory of local @xmath4-vectors and results of section [ sec : geom ] , a simple formula for @xmath443 in terms of the eulerian polynomial @xmath69 is proven ( proposition [ prop : bn+formula ] ) . from this formula , it is deduced that @xmath65 and @xmath66 are real - rooted ( corollary [ cor : b+-nrealroots ] ) , hence unimodal and log - concave , and a new proof of the unimodality of @xmath68 is derived .
finally , recurrences and generating functions for @xmath65 and @xmath66 are given .
these lead to recurrences and generating functions for @xmath34 and @xmath35 and to yet another proof of theorem [ thm : localint ] .
we recall that @xmath65 and @xmath66 are defined by ( [ eq : bn+def ] ) and ( [ eq : bn - def ] ) . for the first few values of @xmath6 we have @xmath444 and @xmath445
the previous data suggest the following statement .
[ lem : rec ] we have @xmath446 for @xmath293 . given a signed permutation @xmath447 , where the notation is as in section [ subsec : signed ] , we set @xmath448 .
then the induced map @xmath449 defined by @xmath450 is a bijection .
moreover , for every @xmath451 , an index @xmath109 is a @xmath2-ascent of @xmath87 if and only if @xmath154 is a @xmath2-descent of @xmath452 and the proof follows . to prove the formula for @xmath65 promised , we will use the construction of the @xmath453th edgewise subdivision @xmath454 of a simplicial complex @xmath155 .
we refer the reader to @xcite for the definition and history of this subdivision and recall the following known facts .
first , the restriction @xmath455 of @xmath454 has exactly @xmath456 facets for every nonempty face @xmath158 .
second , combining ( * ? ? ?
* corollary 6.8 ) with ( * ? ? ?
* corollary 1.2 ) , one gets the explicit formula @xmath457 for the @xmath4-polynomial of @xmath454 , where @xmath421 is the dimension of @xmath155 and @xmath458 is the operator on polynomials ( more generally , on formal power series ) defined by @xmath459 figure [ fig : kk3 ] shows the second edgewise subdivision of the barycentric subdivision of the 2-dimensional simplex .
= 1.5 in [ prop : bn+formula ] we have @xmath460 for every @xmath293 .
we consider the subdivision @xmath54 and the second edgewise subdivision @xmath461 of the barycentric subdivision @xmath418})$ ] ( see figures [ fig : k3 ] and [ fig : kk3 ] for the special case @xmath55 ) . applying ( [ eq : hformula ] ) for @xmath462 or @xmath461 , respectively , and @xmath463})$ ]
we get @xmath464 since both restrictions @xmath465 and @xmath466 have exactly @xmath422 facets for every nonempty face @xmath158 , it follows from the previous formulas and lemma [ lem:2^n-1facets ] that @xmath467 . combining this equality with proposition [ prop : knhpoly ]
we get @xmath468 for every @xmath293 .
formula ( [ eq : hedgewise ] ) implies that @xmath469 for @xmath293 and the proof follows .
[ rem : cry ] we thank mirk visontai for informing us that a formula similar to the one in proposition [ prop : bn+formula ] can be derived from ( * ? ? ? * theorem 4.4 ) , for which a bijective proof was given in @xcite .
we will use the following lemma to deduce the real - rootedness of @xmath65 and @xmath66 .
[ lem : er ] let @xmath132 be a polynomial with real coefficients and let @xmath453 be a positive integer .
* if @xmath132 has unimodal coefficients , then so does @xmath470 .
* if @xmath132 has nonnegative and log - concave coefficients , with no internal zeros , then so does @xmath470 . *
if @xmath132 is real - rooted , then so is @xmath470 .
part ( a ) is trivial and part ( b ) can be left as an excercise .
for part ( c ) we set @xmath471 and note that the matrix @xmath472 is a submatrix of @xmath473 .
therefore , every minor of the former is also a minor of the latter and the result follows from theorem [ thm : realroots ] .
[ cor : b+-nrealroots ] the polynomials @xmath65 and @xmath66 are real - rooted for every @xmath474 .
they are unimodal with peaks at @xmath475 and @xmath476 , respectively , for every @xmath477 .
the first statement follows from lemma [ lem : rec ] , proposition [ prop : bn+formula ] and the fact that the eulerian polynomial @xmath478 is real - rooted for @xmath293 , via part ( c ) of lemma [ lem : er ] .
the second statement follows from lemma [ lem : rec ] , proposition [ prop : bn+formula ] and the fact that @xmath479 is a polynomial of degree @xmath480 with symmetric and unimodal coefficients , via part ( a ) of lemma [ lem : er ] .
[ rem : bnformula ] since @xmath481 , lemma [ lem : rec ] and proposition [ prop : bn+formula ] express the eulerian polynomial @xmath68 as a sum of two unimodal polynomials with peaks which differ by at most one ( see corollary [ cor : b+-nrealroots ] ) .
this decomposition shows that the unimodality of @xmath68 is a consequence of the unimodality of @xmath69 . for a @xmath44-nonnegativity proof of the unimodality of @xmath68 , see ( *
* proposition 4.16 ) . for an equation relating the eulerian polynomials of types @xmath5 , @xmath2 and @xmath482 , see ( * ? ? ?
* lemma 9.1 ) .
we will now give recurrences and generating functions for @xmath65 and @xmath66 .
[ prop : b+-nexp ] we have @xmath483 for every @xmath293 , @xmath484 and @xmath485 let @xmath486 be a signed permutation , represented as a word . for @xmath487 ,
we will denote by @xmath488 ( respectively , @xmath489 ) the signed permutation in @xmath17 obtained from @xmath87 by inserting @xmath6 ( respectively , @xmath490 ) between @xmath491 and @xmath234 .
for @xmath492 we have @xmath493 ( respectively , @xmath494 ) if and only if @xmath495 . on the other hand , @xmath496 and @xmath497 for every @xmath498 .
moreover , for @xmath499 we have @xmath500 and @xmath501 .
thus , we compute that @xmath502 which proves ( [ eq : b+nrec ] ) .
we now claim that @xmath503 given that @xmath504 , equation ( [ eq : b+nexp ] ) then follows by straightforward computations . to prove ( [ eq : ratb+n ] ) , denote by @xmath505 the coefficient of @xmath506 in the expansion of @xmath507 as a formal power series . dividing ( [ eq : b+nrec ] ) by @xmath508 and using the equality @xmath509
we find that @xmath510 comparing the coefficients of @xmath506 in the two sides of the previous equation and using ( * ? ? ?
* theorem 3.4 ( ii ) ) , we get @xmath511 .
the claim then follows by induction on @xmath6 .
equation ( [ eq : b - nexp ] ) follows from ( [ eq : b+nexp ] ) and lemma [ lem : rec ] . alternatively
, it follows from ( [ eq :
b - nexp ] ) and the formula for the exponential generating function of @xmath68 ( * ? ? ?
* theorem 3.4 ( iv ) ) .
we now deduce recurrence relations for @xmath34 and @xmath35 .
[ prop : recf+n ] for @xmath293 we have @xmath512 using equation ( [ eq : localint+ ] ) , we compute that @xmath513 substituting for @xmath514 the right - hand side of ( [ eq : b+nrec ] ) , setting @xmath515 and using ( [ eq : dnb ] ) , we get @xmath516 finally , using equation ( [ eq : localint+ ] ) , we compute that @xmath517 and the proof follows
. we will denote by @xmath518 , @xmath519 and @xmath520 the coefficient of @xmath521 in @xmath34 , @xmath35 and @xmath3 , respectively .
the following recurrence relations can be derived from proposition [ prop : recf+n ] and ( * ? ? ?
* corollary 4.3 ) .
[ cor : ank ] for @xmath477 and @xmath522 we have @xmath523 and @xmath524 equation ( [ eq : a+nk ] ) follows from the formula of proposition [ prop : recf+n ] by comparing the coefficients of @xmath521 .
since @xmath525 , equation ( [ eq : a - nk ] ) follows from ( [ eq : a+nk ] ) and the recurrence relation for @xmath526 given in ( * ? ? ?
* corollary 4.3 ) . as in the second proof , we denote by @xmath427 and @xmath438 the right - hand sides of ( [ eq : localint+ ] ) and ( [ eq : localint- ] ) , respectively , and note that @xmath527 and that @xmath427 is equal to the local @xmath4-polynomial of @xmath54 . in particular , we have @xmath528 by theorem [ thm : stalocal ] ( b ) .
the proofs of proposition [ prop : recf+n ] and corollary [ cor : ank ] show that the coefficients of @xmath427 and @xmath438 satisfy ( [ eq : a+nk ] ) and ( [ eq : a - nk ] ) , respectively .
since @xmath525 , we may rewrite ( [ eq : a+nk ] ) as @xmath529 switching @xmath16 to @xmath530 in this equality and using the symmetry @xmath531 shows that @xmath532 equivalently , we have @xmath533 for all @xmath6 and @xmath16 and hence @xmath534 for every @xmath27 .
the uniqueness of the defining properties of @xmath34 and @xmath35 shows that @xmath434 and @xmath535 for every @xmath536 .
we have verified that @xmath34 and @xmath35 are real - rooted for @xmath537 .
thus , it is natural to conjecture the following statement .
[ conj : f+-n ] the polynomials @xmath34 and @xmath35 are real - rooted for every @xmath538 .
the authors wish to thank ron adin , benjamin nill , yuval roichman , john stembridge , mirk visontai and volkmar welker for useful pointers to the literature .
the second author also thanks francesco brenti and mirk visontai for useful discussions .
the second author was co - financed by the european union ( european social fund - esf ) and greek national funds through the operational program education and lifelong learning " of the national strategic reference framework ( nsrf ) - research funding program : heracleitus ii . investing in knowledge society through the european social fund .
adin , f. brenti and y. roichman , _ descent numbers and major indices for the hyperoctahedral group _ , adv . in appl
* 27 * ( 2001 ) , 210224 .
athanasiadis , _ cubical subdivisions and local @xmath4-vectors _ , ann . comb . * 16 * ( 2012 ) , 421448 . c.a .
athanasiadis , _ flag subdivisions and @xmath44-vectors _ , pacific j. math .
* 259 * ( 2012 ) , 257278 .
athanasiadis and c. savvidou , _ the local @xmath4-vector of the cluster subdivision of a simplex _ , sm .
* 66 * ( 2012 ) , article b66c , 21pp ( electronic ) .
babson , l.j .
billera and c.s .
chan , _ neighborly cubical spheres and a cubical lower bound conjecture _ , israel j. math .
* 102 * ( 1997 ) , 297315 . m. beck and a. stapledon , _ on the log - concavity of hilbert series of veronese subrings and ehrhart series _ , math . z. * 264 * ( 2010 ) , 195207 .
r. biagoli and f. caselli , _ invariant algebras and major indices for classical weyl groups _ , proc . lond .
* 88 * ( 2004 ) , 603631 .
a. bjrner , _ topological methods _ , in _ handbook of combinatorics _ ( r.l .
graham , m. grtschel and l. lovsz , eds . ) , north holland , amsterdam , 1995 , pp . 18191872 .
a. bjrner and f. brenti , combinatorics of coxeter groups , graduate texts in mathematics * 231 * , springer - verlag , new york , 2005 .
f. brenti , _ unimodal polynomials arising from symmetric functions _ , proc .
* 108 * ( 1990 ) , 11331141 . f. brenti , _
@xmath7-eulerian polynomials arising from coxeter groups _ , european j. combin . *
15 * ( 1994 ) , 417441 .
f. brenti and v. welker , _ the veronese construction for formal power series and graded algebras _ , adv . in appl
* 42 * ( 2009 ) , 545556 .
m. brun and t. rmer , _ subdivisions of toric complexes _ , j. algebraic combin .
* 21 * ( 2005 ) , 423448 .
j. cheeger , w. mller and r. schrader , _ on the curvature of piecewise flat spaces _ ,
* 92 * ( 1984 ) , 405454 .
chen , r.l .
tang and a.f.y .
zhao , _ derangement polynomials and excedances of type @xmath2 _ , electron
. j. combin .
* 16 * ( 2 ) ( 2009 ) , research paper 15 , 16pp ( electronic ) .
chow , _ on derangement polynomials of type @xmath2 .
ii _ , j. combin . theory series a * 116 * ( 2009 ) , 816830 .
k. dilks , t.k . petersen and j.r .
stembridge , _ affine descents and the steinberg torus _ , adv . in appl
* 42 * ( 2009 ) , 423444 .
a. edrei , _ proof of a conjecture of schoenberg on the generating function of a totally positive sequence _ , canad . j. math . * 5 * ( 1953 ) , 8694 .
d. foata and m .-
schtzenberger , thorie gometrique des polynmes eulriens , lecture notes in mathematics * 138 * , springer - verlag , 1970 .
lai and t.k .
petersen , _ euler - mahonian distributions of type @xmath17 _ , discrete math .
* 311 * ( 2011 ) , 645650 .
b. nill , _ combinatorial questions related to stringy @xmath305-polynomials of gorenstein polytopes _ , in _ toric geometry _ ( k. altmann et .
al . , eds . ) , oberwolfach reports ( 2012 ) , no .
21 , pp . 6264 .
petersen , _ enriched @xmath539-partitions and peak algebras _ , adv .
* 209 * ( 2007 ) , 561610 .
j. shareshian and m.l .
wachs , _ eulerian quasisymmetric functions _ , adv .
math . * 225 * ( 2010 ) , 29212966 .
stanley , _ log - concave and unimodal sequences in algebra , combinatorics , and geometry _ , in _ graph theory and its applications : east and west _ , annals of the new york academy of sciences * 576 * , new york acad .
, new york , 1989 , pp .
stanley , _ subdivisions and local @xmath4-vectors _ , j. amer .
* 5 * ( 1992 ) , 805851 .
stanley , combinatorics and commutative algebra , second edition , birkhuser , basel , 1996 . r.p .
stanley , enumerative combinatorics , vol . 1
, cambridge studies in advanced mathematics * 49 * , cambridge university press , second edition , cambridge , 2011 . r.p .
stanley , enumerative combinatorics , vol . 2 , cambridge studies in advanced mathematics * 62 * , cambridge university press , cambridge , 1999 .
stembridge , _ eulerian numbers , tableaux , and the betti numbers of a toric variety _ , discrete math .
* 99 * ( 1992 ) , 307320 .
stembridge , _ some permutation representations of weyl groups associated with the cohomology of toric varieties _ , adv . math .
* 106 * ( 1994 ) , 244301 .
j. walker , _ canonical homeomorphisms of posets _ , european j. combin . * 9 * ( 1988 ) , 97107
. x. zhang , _ on @xmath7-derangement polynomials _ , in _ combinatorics and graph theory 95 _ , vol . 1 ( hefei ) , pp . 462465 , world sci .
publishing , river edge , nj , 1995 . | the derangement polynomial @xmath0 for the symmetric group enumerates derangements by the number of excedances .
the derangement polynomial @xmath1 for the hyperoctahedral group is a natural type @xmath2 analogue .
a new combinatorial formula for this polynomial is given in this paper .
this formula implies that @xmath3 decomposes as a sum of two nonnegative , symmetric and unimodal polynomials whose centers of symmetry differ by a half and thus provides a new transparent proof of its unimodality . a geometric interpretation , analogous to stanley s interpretation of @xmath0 as the local @xmath4-polynomial of the barycentric subdivision of the simplex , is given to one of the summands of this decomposition .
this interpretation leads to a unimodal decomposition of the eulerian polynomial of type @xmath2 whose summands can be expressed in terms of the eulerian polynomial of type @xmath5 . the various decomposing polynomials introduced here
are also studied in terms of recurrences , generating functions , combinatorial interpretations , expansions and real - rootedness . |
several theoretical studies have suggested that dynamical interactions among young stars could be important perhaps even essential for the formation of brown dwarfs @xcite . in one of the proposed scenarios , the dynamical evolution of a group of protostars leads to the ejection of one of its members from the natal cloud core .
because its accretion is prematurely halted , the ejected object does not grow to become a star .
alternatively , brown dwarfs might form through the fragmentation of circumstellar disks around protostars .
the disruption of these binary systems by interactions with other stars in the surrounding cluster would then produce free - floating brown dwarfs . according to early versions of the ejection models ,
binary brown dwarfs should have separations that are no larger than @xmath9 au @xcite .
this prediction was consistent with the results from initial multiplicity surveys of brown dwarfs ( * ? ? ?
* references therein ) , but more recent observations have uncovered a small number of wide low - mass binaries in both the field and young clusters . because of their fragile nature , it would seem difficult for a dynamical model to explain the existence of the widest and least massive systems ( @xmath10 au , @xmath11 @xmath4 ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
nevertheless , in a new set of simulations , @xcite were able to produce a wide binary brown dwarf through the simultaneous ejection of two brown dwarfs in similar directions , indicating that the ejection models may remain viable .
however , the feasibility of this formation mechanism for wide binaries depends on the environmental conditions of the star - forming cloud , particularly the stellar density .
in fact , because the ejection models require the presence of a stellar cluster , brown dwarfs should not form in isolation , either as singles or binaries , if dynamical interactions are necessary for their formation .
the taurus complex of molecular clouds is the most promising site in which one might find brown dwarfs that have been born in isolation .
taurus is the prototypical example of a low - density star forming region , is well - populated ( @xmath12 known members ) and nearby ( @xmath13 pc ) , and has been surveyed extensively for both stellar and substellar members @xcite . in this paper , we report the discovery of a wide binary brown dwarf in an isolated dark cloud in taurus .
we present optical and infrared ( ir ) photometry and spectroscopy for the components of the pair ( [ sec : obs ] ) and use these data to measure their spectral types ( [ sec : class ] ) , place them on the hertzsprung - russell ( h - r ) diagram ( [ sec : hr ] ) , and construct their spectral energy distributions ( [ sec : sed ] ) .
we then assess the evidence that these two objects are members of taurus ( [ sec : mem ] ) and that they comprise a binary system ( [ sec : bin ] ) . finally , we discuss the implications of this new binary system for the origin of brown dwarfs .
( [ sec : disc ] ) .
the subject of this study is the star fu tau , which is projected against the center of the barnard 215 dark cloud @xcite .
an optical image of fu tau and barnard 215 is shown in figure [ fig : sdss ] .
the cloud lies in a remote area of taurus that is well - removed from most of the known members of the star - forming region , as illustrated by the map in figure [ fig : map ] .
fu tau was first identified as a possible young star by @xcite through an h@xmath2 objective prism survey ( designated as haro 6 - 7 ) . as indicated by its name ,
fu tau has exhibited significant variability @xcite , which is a common characteristic of young stars .
@xcite identified it as a probable member of taurus based on its proper motion . despite the early evidence of the youth and membership of fu tau
, no other significant work has been done on it .
@xcite used mid - ir images from the _ spitzer space telescope _ @xcite to search for new members of taurus that have circumstellar disks . during a continuation of that survey , we identified fu tau as a candidate based on its mid - ir excess emission .
while performing optical spectroscopy on it , we noticed a fainter nearby object in the spectrometer s acquisition images . because fu tau is projected against the center of a dark cloud , an optically visible background star in close proximity to it was unexpected .
therefore , we elected to obtain a spectrum of this object to determine if it is a member of the cloud and a potential companion to fu tau .
the spectra confirmed that both fu tau and the nearby source are young and have late spectral types .
hereafter in this paper , we refer to these objects as fu tau a and fu tau b. we describe our spectroscopic observations of the pair in [ sec : spec ] .
to obtain photometry and astrometry for fu tau a and b , we have searched the public data archives of various observatories and wide - field surveys for optical and ir images that encompass their location .
the data produced by this search are presented in [ sec : image ] .
we obtained long - slit optical spectra of fu tau a and b on the nights of 2007 december 17 and 18 , respectively , using the low dispersion survey spectrograph ( ldss-3 ) on the magellan ii telescope .
the observations were performed with a 1.1@xmath14 slit , which was rotated to the parallactic angle .
we used the vph all and vph red grisms on the first and second nights , respectively , resulting in spectral resolutions of 10 and 5.5 at 7500 .
we collected one 10 min exposure for the primary and three 20 min exposures for the secondary .
after bias subtraction and flat - fielding , we extracted the spectra and wavelength calibrated them with arc lamp data .
we then corrected the spectra for the sensitivity functions of the detectors , which were measured from observations of a spectrophotometric standard star .
fu tau a was also observed at near - ir wavelengths using spex @xcite at the nasa infrared telescope facility . on 2007
december 3 , we operated spex in the prism mode with a @xmath15 slit , which produced a spectrum that extended from 0.82.5 and that exhibited a resolving power of @xmath16 . on 2007 december 26 , we collected another spectrum across the same wavelength range at higher resolution ( @xmath17 ) by using spex in the sxd mode with a @xmath15 slit .
these data were reduced with the spextool package @xcite and corrected for telluric absorption @xcite .
both components of fu tau are detected in images at @xmath18 , @xmath19 , and @xmath20 from the two - micron all - sky survey ( 2mass ; * ? ? ?
fu tau a and b correspond to 2mass j04233539 + 2503026 and j04233573 + 2502596 , respectively , in the 2mass point source catalog .
the 2mass photometry of the primary should have negligible contamination from its companion given the large flux ratio of the pair at near - ir wavelengths ( @xmath21 ) .
those data are provided in table [ tab : phot ] .
because the secondary is only marginally resolved by 2mass , the photometry reported in the point source catalog may not be reliable .
we do not use the 2mass images for measuring the relative positions of fu tau a and b since they are better resolved by data from other facilities .
although it focused on areas of the sky at high galactic latitude , the sloan digital sky survey ( sdss ; * ? ? ?
* ) did obtain images of additional fields closer to the galactic plane @xcite .
those data encompassed a large portion of the taurus cloud complex , including barnard 215 ( see fig . [ fig : sdss ] ) .
we have retrieved images and photometry of fu tau in the five optical bands of sdss ( @xmath22 ; * ? ? ?
* ) from the sixth data release of the survey @xcite .
the calibration of these images is described by @xcite .
two sets of images are available for fu tau , which were taken on 2002 december 6 and 31 .
the fwhm is @xmath23 for point sources in all bands at both epochs .
the components of fu tau are well - resolved from each other in these data .
the primary was detected in each of the five bands while the secondary appeared in @xmath24 , @xmath25 , and @xmath26 .
a variety of flux measurements are provided by the sixth data release . for fu
tau a and b , we have selected the data measured with an aperture radius of @xmath27 , which is small enough to avoid contamination from the primary in the measurement of the secondary .
using other stars in these images , we measured aperture corrections for each band between radii of @xmath27 and @xmath28 and applied these corrections to the @xmath27 data for fu tau a and b. for the photometric errors , we combine the poisson errors in the source and background emission with uncertainties of @xmath29% in the absolute calibrations .
the sdss photometry and relative astrometry for the components of fu tau are listed in tables [ tab : phot ] and [ tab : astro ] , respectively .
@xcite obtained images at @xmath30 and @xmath31 of the barnard 215 cloud as a part of a wide - field survey of taurus .
these data were collected on the canada - france - hawaii telescope ( cfht ) with the cfh12k camera on 2002 december 29 .
the instrument contained twelve @xmath32 ccds in a @xmath33 mosaic .
the total field of view was @xmath34 .
we retrieved pipeline processed images of barnard 215 and the associated photometric standards from the cfht archive .
one 30 s exposure and six 300 s exposures were available in each filter .
the fwhm of stars in the @xmath30-band images is 1@xmath35 .
one of the long @xmath30-band exposures of fu tau is shown in figure [ fig : image ] .
we measured aperture photometry for the primary and secondary from the short and long @xmath30-band exposures , respectively .
the photometry at @xmath30 and relative astrometry are given in tables [ tab : phot ] and [ tab : astro ] , respectively .
because the photometric standards for these observations lack photometry at @xmath31 @xcite , we have not attempted to measure photometry from the images in this filter .
the united kingdom infrared telescope ( ukirt ) infrared deep sky survey ( ukidss , * ? ? ?
* ) is obtaining near - ir images of large areas of the northern sky . as a part of a survey of young nearby clusters , ukidss is in the process of imaging most of the taurus star - forming region in @xmath31 , @xmath36 , @xmath18 , @xmath19 , and @xmath7 .
the @xmath7-band observations of barnard 215 were completed on 2005 december 31 and the resulting data products are now available to the public .
we retrieved the @xmath7-band image and photometry for fu tau from the first ukidss data release @xcite .
point sources in this image exhibit a fwhm of @xmath37 .
the components of the binary are well - resolved , as shown in figure [ fig : image ] .
for a given object and band , ukidss provides several photometric measurements that are based on a range of aperture radii .
we present in table [ tab : phot ] the @xmath7-band photometry for fu tau b that was measured with an aperture radius of @xmath38 @xcite .
ukidss photometry for the primary is unavailable since it is saturated .
we measured the relative positions of the components of fu tau from the ukidss image , which are given in table [ tab : astro ] .
as described in
[ sec : background ] , our study of fu tau began during a survey for young stars with disks using mid - ir images of taurus from the _ spitzer space telescope_. these images were obtained at 3.6 , 4.5 , 5.8 , and 8.0 with _ spitzer _ s infrared array camera ( irac ; * ? ? ?
fu tau appears within two sets of irac data . on 2005
february 23 , the young star ft tau , which is on the edge of barnard 215 ( see fig . [
fig : sdss ] ) , was observed through the irac guaranteed time observations of g. fazio in _
program 37 . while the 3.6 and 5.8 detectors were centered on ft tau ,
the 4.5 and 8.0 detectors observed an adjacent area of sky that happened to encompass fu tau . the astronomical observation request ( aor ) identification for these data is 3964672 .
fu tau also fell within a wide - field irac mosaic that was obtained through the _
legacy program of d. padgett , which has a program identification of 30816 .
these observations were performed on 2007 march 30 and have aor numbers of 19028480 and 19028224 . the irac images from 2005 and 2007 were processed with the _ spitzer _ science center ( ssc ) s14.0.0 and s15.3.0 pipelines , respectively .
the images produced by the pipelines were then combined using r. gutermuth s wcsmosaic idl package .
the reduced images of fu tau a and b at 3.6 and 8.0 from the 2007 observations are shown in figure [ fig : image ] .
we measured aperture aperture photometry for fu tau a in the manner described by @xcite .
we applied the same methods to the secondary except that we first removed light from the primary by subtracting a scaled irac point spread function @xcite .
in addition to the irac data , images of fu tau at 24 have been obtained with the multiband imaging photometer for _ spitzer _ ( mips ; * ? ? ?
these observations were performed on 2007 february 2 through d. padgett s _ spitzer _ legacy program and have aor numbers of 19026688 and 19027200 .
we measured aperture photometry for fu tau a from the images produced by the ssc s16.1.0 pipeline @xcite .
fu tau b is not resolved from the primary in these data .
the irac and mips measurements for fu tau are presented in table [ tab : phot ] .
to investigate the properties of fu tau a and b , we begin by using our optical and ir spectra to measure their spectral types and reddenings and to assess their ages .
the optical spectra are shown in figure [ fig : spec ] . both objects exhibit strong tio and vo absorption bands , which are characteristic of late - m spectral types .
the weak na i , k i , and feh absorption lines @xcite and , in the case of the primary , triangular @xmath19-band continuum @xcite demonstrate that fu tau a and b have the low surface gravities that are found in pre - main - sequence objects .
the h@xmath2 emission in each spectrum is also stronger than that found among field dwarfs @xcite , providing additional evidence of youth .
thus , the spectra confirm that fu tau a and b are young , low - mass objects .
we measured spectral types from the optical spectra by comparing them to averages of dwarfs and giants @xcite and previously - classified members of taurus and other star - forming regions @xcite , arriving at [email protected] and [email protected] for the primary and secondary , respectively . to illustrate the derivation of these classifications , fu tau a and b
are compared to a selection of standards in figure [ fig : spec ] .
fu tau a matches closely with the average of a dwarf and a giant at m7.25 as well as the primary in the young binary 2mass j11011926@xmath407732383 ( hereafter 2 m 1101@xmath407732 , m7.25 ; * ? ? ?
it is slightly cooler than the composite of the young eclipsing binary 2mass j05352184@xmath400546085 ( hereafter 2 m 0535@xmath400546 ; * ? ? ?
* ) , which has an optical spectral type of m6.75 @xcite . meanwhile , fu tau b is intermediate between an m9 dwarf / giant average and the taurus member kpno 4 ( m9.5 ; * ? ? ?
when we compare the ir spectrum of fu tau a to previous spex data for optically - classified young objects @xcite , we derive an ir spectral type that is consistent with the optical measurement .
the slopes of the optical spectra of fu tau a and b are similar to those of our bluest young standards , indicating they have low extinction ( @xmath41 ) .
for instance , the optical spectrum of the primary has the same slope as oph 1622@xmath402405 a from upper sco ( m7.25 ; * ? ? ? * ) , which probably has little reddening since it is not associated with a molecular cloud .
however , the 12.5 spectrum of fu tau a is redder than that of oph 1622@xmath402405 a by an amount that is equivalent to @xmath42 @xcite . the same is true when we compare fu tau a to other young m7 objects that appear to have negligible extinction , such as mho 4 , kpno 2 , and kpno 5 .
this anomalous difference between the optical and ir reddening estimates for fu tau a is probably not caused by errors in the spectra because the 0.61 slopes of the ldss-3 , low - resolution spex data , and medium - resolution spex data are in good agreement while the 12.5 slopes are the same among the two sets of spex data and the colors from 2mass .
it is possible that the ir spectrum appears too red because it is contaminated by long - wavelength emission from circumstellar dust , or the optical spectrum appears too blue because it is contaminated by scattered light or uv emission from accretion . to test the first hypothesis ,
we have compared the @xmath17 @xmath7-band spectrum of fu tau a to spex data for field dwarfs near the same spectral type @xcite .
we find no evidence for significant continuum emission from dust , which would cause the lines to appear weaker , or veiled " , relative to those of a normal stellar photosphere .
thus , the slope of the ir spectrum of fu tau a should accurately reflect its extinction .
the alternative explanation is supported by the analysis of the spectral energy distribution ( sed ) of fu tau a in [ sec : sed ] , which demonstrates the presence of short - wavelength excess emission . to examine the masses and ages of the components of fu tau , we can compare their effective temperatures and bolometric luminosties to the values predicted by theoretical evolutionary models .
we have converted the spectral types of fu tau a and b to effective temperatures with the temperature scale from @xcite . to estimate the luminosity of the primary , we have combined its @xmath18-band magnitude with the average bolometric correction for dwarfs near its spectral type @xcite , an extinction of @xmath42 ( [ sec : class ] ) , and a distance of 140 pc @xcite . for the secondary
, we use its @xmath7-band magnitude since a reliable @xmath18 measurement is not available . by doing so
, we are assuming that any circumstellar disk that might reside around fu tau b produces negligible emission at @xmath7 compared to the stellar photosphere .
this assumption is likely to be valid since disks around brown dwarfs rarely exhibit significant @xmath7-band excess emission @xcite .
the resulting temperatures and luminosities are listed in table [ tab : prop ] .
the uncertainties in @xmath43 , near - ir magnitudes , and bolometric corrections ( @xmath44 , 0.02 , 0.1 ) correspond to errors of @xmath45 in the relative values of log @xmath46 .
when an uncertainty in the distance modulus is included ( @xmath47 ) , the total uncertainties are @xmath48 .
these uncertainties do not include variability , which is known to be significant for the primary ( table [ tab : phot ] ; * ? ? ?
our temperature and luminosity estimates for fu tau a and b are plotted on the h - r diagram in figure [ fig : hr ] with the evolutionary models from @xcite and @xcite .
although the components of binary systems are expected to be coeval , fu tau a and b do not appear near the same model isochrone .
the models imply an age of 1 myr for the secondary while the primary is overluminous by an order of magnitude relative to that isochrone . the discrepancy is reduced somewhat if fu tau a and b are compared to a fit to the cluster sequence for chamaeleon i @xcite , which acts as an empirical isochrone .
as shown in figure [ fig : hr ] , the primary and secondary appear 1.3 and 0.6 dex above this fit , respectively , resulting in a luminosity ratio that is 0.7 dex larger than expected for a coeval pair .
fu tau a is much brighter than other members of taurus with similar spectral types whereas fu tau b falls within the cluster sequence @xcite , indicating that the difference in isochronal ages for the pair is a reflection of an anomalously high luminosity for the primary rather than a low luminosity for the secondary .
one possible explanation is that fu tau a is an unresolved binary .
if so , it would appear near the upper envelope of the cluster sequence for taurus in the h - r diagram , although it would remain overluminous relative to fu tau b by @xmath49 dex .
the data and models in figure [ fig : hr ] imply a mass of @xmath50 @xmath4 for fu tau b. the model predictions do not encompass the temperature and luminosity of fu tau a , and thus do not provide a direct estimate of its mass . however , because of the vertical nature of the mass tracks of low - mass stars and brown dwarfs at young ages ( @xmath51 myr ) , the spectral type of fu tau a should provide a good indication of its mass regardless of luminosity . as shown in figure [ fig : hr ] , the spectral type of fu tau a combined with the adopted temperature scale and models imply a mass of @xmath52 @xmath4 .
this estimate is consistent with the fact that fu tau a is slightly cooler than 2 m 0535@xmath400546 ( figure [ fig : spec ] ) , whose components have dynamical masses of 0.054 and 0.034 @xmath4 @xcite to compare the masses and ages of fu tau a and b to those of previously known young low - mass binaries , we have included the components of 2 m j1101@xmath407732 , oph 1622@xmath402405 , and usco ctio 108 in figure
[ fig : hr ] . because the optical spectra of these systems and fu tau have been directly compared ( figure [ fig : spec ] ;
* ; * ? ? ?
* ) , their relative spectral types should be quite accurate .
in addition , the spectral types and photometry have been converted to temperatures and luminosities with the same methods that we have applied to fu tau .
as shown in figure [ fig : hr ] , fu tau a probably has a mass that is similar to those of the other three primaries
while fu tau b may be slightly less massive than oph 1622@xmath402405 b and usco ctio 108 b. the most striking aspect of this comparison is that the luminosity ratio of fu tau a and b ( @xmath53 ) is much larger than the ratios in the other systems ( @xmath5416 ) , further illustrating that one of the components of fu tau ( probably the primary ) has an anomalous luminosity .
we selected fu tau a as a candidate member of taurus because it exhibits mid - ir colors that are indicative of a circumstellar disk . to demonstrate the presence of this mid - ir excess emission from fu tau a and to determine whether its companion has a disk ,
we have constructed their seds using the photometry compiled in table [ tab : phot ] .
sdss and irac have provided photometry at multiple epochs for the fu tau system ; we adopt the average of the available measurements for a given sdss band and we use the irac data from 2007 since the camera observed fu tau in only two bands in the other epoch .
in addition to the photometry , we include in the sed for fu tau a the low - resolution ir spectrum obtained with spex , which has been flux - calibrated with the photometry from 2mass .
the resulting seds of fu tau a and b are presented in figure [ fig : sed ] . to determine if the seds contain excess emission at short or long wavelengths , we compare them to the seds of stellar photospheres .
we construct an estimate of the photospheric seds of fu tau a and b from data for the taurus members kpno 5 and kpno 4 , respectively , which have similar spectral types as fu tau a and b , have negligible extinction , and lack mid - ir excess emission and significant accretion @xcite . optical measurements in the sdss and @xmath30 bands for the kpno objects are taken from @xcite and @xcite . because kpno 4 has a very low signal - to - noise ratio at @xmath24 , we compute the flux in this band by combining the @xmath25 photometry for kpno 4 with the typical value of @xmath55 for m9 dwarfs @xcite . for kpno 5 , we include a low - resolution ir spectrum collected with spex @xcite .
for the mid - ir portion of these photospheric seds , we adopt the average colors of diskless stars near the spectral types of fu tau a and b @xcite .
the template for fu tau a is reddened by @xmath42 according to the reddening laws from @xcite and @xcite . the photospheric seds are then normalized to the @xmath18- and @xmath7-band fluxes of fu tau a and b , respectively . relative to these seds , both fu tau a and b exhibit significant excess emission at wavelengths longward of 4 , as shown in figure [ fig : sed ] .
the primary also has excess emission in the bluest optical and uv bands , which probably explains why its optical spectrum implies less extinction than the near - ir data ( [ sec : class ] ) .
to classify the seds of fu tau a and b according to the standard scheme for young stars @xcite , we use the ir spectral slope that is defined as @xmath56 log@xmath57 log@xmath58 @xcite . the extinction - corrected slopes between 3.6 and 8 for each object
are given in table [ tab : prop ] . by applying the thresholds from @xcite to these slopes ,
we classify the seds of fu tau a and b as class ii .
the components of fu tau clearly have ages of @xmath1 myr based on their h@xmath2 emission , mid - ir excess emission , and gravity - sensitive spectral features . given that they are projected against the center of barnard 215 ,
it is likely that they were born in this cloud , which is a part of the taurus star - forming region .
the sdss images of barnard 215 also show faint nebulosity centered on fu tau , which supports its association with the cloud . the proper motion of fu tau represents an additional constraint on its membership .
indeed , @xcite identified fu tau has a likely member of taurus through a measurement of its proper motion .
we now perform a new analysis of the motion of fu tau that makes use of the astrometric data that are currently available . in table
[ tab : pm ] , we list proper motion measurements for fu tau a from the usno - b1.0 catalog @xcite and @xcite .
we also include our measurement of the proper motion , which is based on positions in the usno - a2.0 catalog ( * ? ? ?
* epoch 1950.9 ) , the guide star catalog v2.3.2 ( * ? ? ?
* epoch 1994.8 ) , the 2mass point source catalog ( * ? ? ?
* epoch 1997.9 ) , and the carlsberg meridian catalog vol .
14 ( cds catalog i/304 ; epoch 2001.8 ) .
our estimate is similar to the value from usno - b1.0 .
we adopt our measurement of the proper motion in the following discussion . what is the proper motion of a hypothetical member of taurus at the position of fu tau a ?
because taurus covers a large solid angle of sky , the proper motion for a given velocity vector varies across the association due to geometric projection effects . for this calculation
, we adopt a mean galactic velocity vector of ( u , v , w)@xmath59 km s@xmath60 @xcite and a mean distance of 140 pc for taurus @xcite . at the position of fu
tau a , we predict that a star with the mean velocity vector of taurus would have a proper motion of @xmath61 ( 140 pc/@xmath62 ) mas yr@xmath60 and @xmath63 ( 140 pc/@xmath62 ) mas yr@xmath60 , where @xmath62 is the distance in pc .
the mean velocity vector is constrained to @xmath391 km s@xmath60 in each component , which translates to an uncertainty of @xmath31 mas yr@xmath60 in the predicted proper motion for a given distance .
hence , we find that our predicted values of @xmath64 and @xmath65 for an `` ideal '' taurus member are within 0.3@xmath66 and 1.7@xmath66 of the observed motion , respectively .
the predicted motion is also within 1.1 @xmath66 and 2.4 @xmath66 of the usno - b1.0 @xmath64 and @xmath65 motions , respectively . for comparison , in table [ tab : pm ]
we also include the predicted proper motions for hypothetical members of the hyades and pleiades open clusters at their respective mean distances , which are based on data from @xcite , @xcite , and @xcite .
both groups are known to have members in the vicinity of taurus .
however , the motion of fu tau a is completely inconsistent with the kinematic membership of either of these clusters . based on this analysis
, we conclude that the proper motion of fu tau a is in good agreement with the average motion of the stellar population in taurus . according to the astrometric measurements of fu tau a and b in table [ tab : astro ]
, the pair maintained the same relative positions to within @xmath67 across a period of three years , indicating that the secondary shares the same motion as the primary at a level of @xmath330 mas yr@xmath60 .
it is useful to compare the motion of fu tau a to that of ft tau , which is the only other known young star that is projected against the barnard 215 cloud ( figure [ fig : sdss ] ) .
we have measured the proper motion of ft tau using the same catalogs that were employed for fu tau a , arriving at values of @xmath68 mas yr@xmath60 and @xmath69 mas yr@xmath60 .
the motions of the two objects agree to within [email protected] and [email protected] mas yr@xmath60 in right ascension and declination , respectively , which translates into tangential velocities at 140 pc that agree to within [email protected] and [email protected] km s@xmath60 .
these relative motions indicate that fu tau a has remained within a projected distance of 1.7 pc ( @xmath70 ) from ft tau during its lifetime ( @xmath71 myr , figure [ fig : hr ] ) .
the agreement in their velocities and the presence of nebulosity centered on each source strongly suggests that they are associated with the barnard 215 cloud and were not born elsewhere . finally , we have searched for evidence of additional young stars near barnard 215 . aside from fu tau and ft tau , the only other candidate young star that has been previously identified near the cloud is gt tau @xcite . through an analysis similar to that performed for fu tau and ft tau , we find that the proper motion of gt tau is inconsistent with membership in taurus .
previous surveys for young stars in the vicinity of barnard 215 have been conducted at optical wavelengths , and thus could have missed stars that are embedded within the cloud .
the mid - ir images from _ spitzer _ that were used to uncover fu tau easily penetrate the extinction of barnard 215 , and they do not reveal any additional disk - bearing stars .
we can not rule out the presence of diskless young stars within the cloud given that they would be indistinguishable from field stars in _ spitzer _ data .
however , it is unlikely that an embedded population within the cloud would consist of only diskless stars .
we have presented strong evidence indicating that fu tau a and b are members of the taurus star - forming region .
we have also shown that they have similar proper motions , but the accuracy of these measurements is insufficient to distinguish between a binary system and a pair of unrelated members of taurus that are seen in projection near each other . therefore , as in most studies of visual pairs in star - forming regions , we rely on a statistical analysis to assess whether fu tau a and b are likely to comprise a binary system . because of the low stellar density in taurus , it is unlikely that a given member of the region will appear close to another member on the sky , unless they are in a binary system .
this probability is particularly low for fu tau , which resides in one of the most isolated dark clouds in taurus .
as shown in figure [ fig : map ] , only one young star has been found within a radius of @xmath72 surrounding fu tau .
we can roughly quantify the probability that fu tau is a pair of unrelated low - mass taurus members . among the 51 known members of taurus that have spectral types later than m6 , the median distance to the nearest @xmath73m6 neighbor is @xmath74 . for our probability calculation ,
we adopt a population of 51 objects that are randomly distributed across an area of 32.5 deg@xmath75 , which exhibits a median nearest neighbor distance that is similar to the observed value .
the probability of finding a pair of objects in this population with a projected separation of @xmath6 is @xmath76 .
therefore , it is likely that fu tau a and b comprise a binary system .
with component masses of @xmath30.05 and @xmath30.015 @xmath4 and a projected separation of 800 au , fu tau joins the growing sample of wide low - mass binaries that have been discovered in recent years ( [ sec : intro ] ) .
these systems have provided valuable constraints on theories of the formation of brown dwarfs @xcite . unlike the previously known binaries or free - floating brown dwarfs , for that matter fu tau a and b are the first spectroscopically - confirmed brown dwarfs discovered forming in isolation .
the unique formation environment of fu tau represents a new test of theoretical models for the origin of brown dwarfs . as reviewed in
[ sec : intro ] , dynamical models typically create brown dwarfs through ejection from protostellar clusters @xcite or fragmentation of disks around stars and subsequent stripping of the substellar companion through encounters with other stars @xcite . because both mechanisms require the presence of stellar clusters , they can not account for the existence of brown dwarfs like fu tau a and b that are forming in isolation .
although the disk fragmentation models produce brown dwarfs as companions to stars , and those stars can be isolated , this scenario is not plausible for fu tau because the mass and separation of the secondary are much larger than the ( limited ) measurements of masses and radii for brown dwarf disks ( @xmath77 au , @xmath78 @xmath79 ; * ? ? ? * ; * ? ? ?
* ; * ? ? ?
the discovery of fu tau demonstrates that brown dwarfs can arise in isolation without the involvement of dynamical interactions among stars .
it is likely that the same process that made it possible for substellar objects to form in barnard 215 also occurs in clusters as well .
thus , it is unnecessary to invoke additional mechanisms like ejection for creating brown dwarfs unless observations indicate their presence .
dynamical interactions in young clusters probably do affect the formation of brown dwarfs , just as they affect the formation of stars , but there is no evidence to date that they represent a second dominant mechanism that enables the birth of brown dwarfs @xcite .
we thank lee hartmann for helpful comments on the manuscript .
k. l. was supported by grant ast-0544588 from the national science foundation ( nsf ) and e. m. was supported by a clay postdoctoral fellowship from the smithsonian astrophysical observatory .
this publication makes use of data products from 2mass , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by nasa and the nsf .
the guide star catalogue - ii is a joint project of the space telescope science institute and the osservatorio astronomico di torino .
space telescope science institute is operated by the association of universities for research in astronomy , for nasa under contract nas5 - 26555 .
the participation of the osservatorio astronomico di torino is supported by the italian council for research in astronomy .
additional support is provided by european southern observatory , space telescope european coordinating facility , the international gemini project and the european space agency astrophysics division .
this research used the facilities of the canadian astronomy data centre operated by the national research council of canada with the support of the canadian space agency .
funding for sdss has been provided by the alfred p. sloan foundation , the participating institutions , the nsf , the u.s .
department of energy , nasa , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , university of cambridge , case western reserve university , the university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , the johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences , los alamos national laboratory , the max - planck - institute for astronomy , the max - planck - institute for astrophysics , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington .
burgasser , a. j. , reid , i. n. , siegler , n. , close , l. , allen , p. , lowrance , p. , gizis , j. 2007 , protostars and planets v , b. reipurth , d. jewitt , and k. keil ( eds . ) , university of arizona press , tucson , 427 llll @xmath80 & [email protected] & & 2002 dec 6 + @xmath80 & [email protected] & & 2002 dec 29 + @xmath81 & [email protected] & & 2002 dec 6 + @xmath81 & [email protected] & & 2002 dec 29 + @xmath24 & [email protected] & [email protected] & 2002 dec 6 + @xmath24 & [email protected] & [email protected] & 2002 dec 29 + @xmath25 & [email protected] & [email protected] & 2002 dec 6 + @xmath25 & [email protected] & [email protected] & 2002 dec 29 + @xmath26 & [email protected] & [email protected] & 2002 dec 6 + @xmath26 & [email protected] & [email protected] & 2002 dec 29 + @xmath30 & [email protected] & [email protected] & 2002 dec 29 + @xmath18 & [email protected] & & 1997 nov 29 + @xmath19 & [email protected] & & 1997 nov 29 + @xmath20 & [email protected] & & 1997 nov 29 + @xmath7 & saturated & [email protected] & 2005 dec 31 + @xmath82 $ ] & out & out & 2005 feb 23 + @xmath82 $ ] & [email protected] & [email protected] & 2007 mar 3 + @xmath83 $ ] & [email protected] & [email protected] & 2005 feb 23 + @xmath83 $ ] & [email protected] & [email protected] & 2007 mar 3 + @xmath84 $ ] & out & out & 2005 feb 23 + @xmath84 $ ] & [email protected] & [email protected] & 2007 mar 3 + @xmath85 $ ] & [email protected] & [email protected] & 2005 feb 23 + @xmath85 $ ] & [email protected] & [email protected] & 2007 mar 3 + @xmath86 $ ] & [email protected] & & 2007 feb 2 + llllllll @xmath64 & + 6@xmath392 & + 14@xmath397 & + [email protected] & & + 8.4 & + 109.3 & + 14.5 + @xmath65 & @xmath4018@xmath392 & @xmath4026@xmath397 & @[email protected] & & @xmath4023.5 & @xmath4052.9 & @xmath4047.2 + | during a survey for stars with disks in the taurus star - forming region using the _ spitzer space telescope _ , we have discovered a pair of young brown dwarfs , fu tau a and b , in the barnard 215 dark cloud .
they have a projected angular separation of @xmath0 , corresponding to 800 au at the distance of taurus . to assess the nature of these two objects , we have obtained spectra of them and have constructed their spectral energy distributions .
both sources are young ( @xmath1 myr ) according to their h@xmath2 emission , gravity - sensitive spectral features , and mid - ir excess emission . the proper motion of fu tau a provides additional evidence of its membership in taurus .
we measure spectral types of m7.25 and m9.25 for fu tau a and b , respectively , which correspond to masses of @xmath30.05 and @xmath30.015 @xmath4 according to the evolutionary models of chabrier and baraffe .
fu tau a is significantly overluminous relative to an isochrone passing through fu tau b and relative to other members of taurus near its spectral type , which may indicate that it is an unresolved binary .
fu tau a and b are likely to be components of a binary system based on the low probability ( @xmath5 ) that taurus would produce two unrelated brown dwarfs with a projected separation of @xmath6 .
barnard 215 contains only one other young star and is in a remote area of taurus , making fu tau a and b the first spectroscopically - confirmed brown dwarfs discovered forming in isolation rather than in a stellar cluster or aggregate .
because they were born in isolation and comprise a weakly bound binary , dynamical interactions with stars could not have played a role in their formation , and thus are not essential for the birth of brown dwarfs .
* erratum : * the @xmath7-band magnitude for fu tau b in table 1 is incorrect and should be 13.33 .
the bolometric luminosity of fu tau b in table 3 and figure 5 is incorrect because of that mistake and a separate arithmetic error .
the correct value of the luminosity is 0.0039 @xmath8 .
fu tau a and b exhibited different isochronal ages in the original hertzsprung - russell diagram in figure 5 , which was unexpected for members of a binary system .
this discrepancy is reduced in the corrected version of figure 5 since both objects are now above the isochrone for 1 myr .
given the large uncertainties in model isochrones at such young ages , the positions of fu tau a and b in figure 5 could be roughly consistent with coevality . |
a finite ergodic markov chain is said to exhibit _ cutoff _ in total - variation if its @xmath0-distance from the stationary distribution drops abruptly from near its maximum to near @xmath1 .
in other words , one should run the markov chain until the cutoff point for it to even slightly mix in @xmath0 whereas running it any further is essentially redundant .
let @xmath2 be an aperiodic irreducible discrete - time markov chain on a finite state space @xmath3 with stationary distribution @xmath4 .
the worst - case total - variation distance to stationarity at time @xmath5 is defined as @xmath6 where @xmath7 denotes the probability given @xmath8 and where @xmath9 , the _ total - variation distance _ of two distributions @xmath10 on @xmath3 , is given by @xmath11 define @xmath12 , the total - variation _ mixing - time _ of @xmath2 for @xmath13 , to be @xmath14 let @xmath15 be a family of such chains , each with its total - variation distance from stationarity @xmath16 , its mixing - time @xmath17 , etc .
this family exhibits _ cutoff _ iff the following sharp transition in its convergence to equilibrium occurs : @xmath18 the rate of convergence in is addressed by the following notion of a _ cutoff window _ : for two sequences @xmath19 with @xmath20 we say that @xmath15 has cutoff at time @xmath21 with window @xmath22 if and only if @xmath23 or equivalently , cutoff at time @xmath21 with window @xmath22 occurs if and only if @xmath24 the cutoff phenomenon was first identified for random transpositions on the symmetric group in @xcite and for random walks on the hypercube in @xcite .
the term `` cutoff '' was coined by aldous and diaconis in @xcite , where cutoff was shown for the top - in - at - random card shuffling process . while believed to be widespread ,
there are relatively few examples where the cutoff phenomenon has been rigorously confirmed .
even for fairly simple chains , determining whether there is cutoff often requires the full understanding of their delicate behavior around the mixing threshold .
see @xcite and the references therein for more on the cutoff phenomenon .
a specific markov chain which found numerous applications in a wide range of areas in mathematics over the last quarter of a century is the simple random walk ( ) on a bounded - degree _ expander _ graph .
a finite graph is called an expander if every small subset of the vertices has a relatively large edge boundary .
formally , the cheeger constant of a @xmath25-regular graph @xmath26 on @xmath27 vertices ( also referred to as the edge isoperimetric constant ) is defined as @xmath28 where @xmath29 is the set of edges with exactly one endpoint in @xmath30 .
we say that @xmath26 is a @xmath31-edge - expander for some fixed @xmath32 if it satisfies @xmath33 .
the well - known discrete analogue of cheeger s
inequality @xcite implies that the spectral - gap of the on a family of @xmath31-edge - expander graphs on @xmath27 vertices is uniformly bounded away from @xmath1 , hence these chains rapidly converge to equilibrium within @xmath34 steps .
see the survey @xcite for more on the applications of random walks on expanders . in 2004
, peres@xcite observed that for any family of reversible markov chains , total - variation cutoff can only occur if the inverse spectral - gap has smaller order than the mixing time .
note that this condition clearly holds for the simple random walk on an @xmath27-vertex expander , where the inverse - gap is @xmath35 whereas @xmath36 .
it was shown by chen and saloff - coste @xcite that when measuring convergence in @xmath37-distance for @xmath38 this criterion does ensure cutoff , however the case @xmath39 ( cutoff in total - variation ) has proved to be significantly more complicated .
there are known examples where the above condition does not imply cutoff ( see @xcite*section 6 ) , yet it was conjectured by peres to be sufficient in many natural families of chains ( e.g. @xcite confirming this for birth - and - death chains ) .
in particular , this was conjectured for the lazy random walk on bounded - degree transitive graphs .
the first example of a family of bounded - degree graphs where the random walk exhibits cutoff in total - variation was provided only very recently @xcite , when the authors showed this for a typical random regular graph .
it is well known that for any fixed @xmath40 , a random @xmath25-regular graph is with high probability ( w.h.p . ) a very good expander , hence the simple random walk on almost every @xmath25-regular expander exhibits worst - case total - variation cutoff . however , to this date there were no known examples for an _ explicit _ ( deterministic ) family of expanders with this phenomenon . in section [ sec - explicit ]
we provide what is , to the best of our knowledge , the first explicit construction of a family of bounded - degree expanders where the simple random walk has worst - case total - variation cutoff .
[ thm-1 ] there is an explicit family of @xmath41-regular expanders on which the _ _ from a worst case initial position exhibits total - variation cutoff .
the construction mimics the behavior of the on random regular graphs , whose mixing was shown in @xcite ( as conjectured by durrett @xcite and berestycki @xcite ) to resemble that of a walk started at a root of a @xmath25-regular tree .
two smaller expanders that are embedded into the graph structure allow careful control over the mixing time from all possible initial positions . a straightforward modification of the above construction yields an explicit family of cubic expanders where the from a worst - case initial position does _ not _ exhibit cutoff in total - variation ( despite peres cutoff criterion ) .
note that peres and wilson @xcite had already sketched an example for a family of expanders without total - variation cutoff .
we describe our simple construction achieving this in section [ sec - nocutoff ] for completeness .
a final variant of the construction , presented in section [ sec - prescribed ] , provides cubic graphs with cutoff occurring at essentially any prescribed order of location .
namely , there is an explicit family of cubic graphs where the has cutoff at any specified order between @xmath42 whereas for @xmath43 there can not be cutoff ( it is well - known that on any family of bounded - degree graphs on @xmath27 vertices the has @xmath44 for some fixed @xmath45 ) .
[ thm-2 ] let @xmath21 be a monotone sequence with @xmath46 and @xmath47 .
there is an explicit family @xmath48 of @xmath41-regular graphs with @xmath49 vertices where the _ _ from a worst - case initial position exhibits total - variation cutoff at @xmath50 .
furthermore , for any family of bounded - degree @xmath27-vertex graphs where the _ _ has @xmath43 ( largest possible order of mixing ) there can not be cutoff .
to simplify the exposition , we will first construct a family of 5-regular expanders where the from a worst initial position exhibits cutoff .
subsequently , we will describe how to modify the construction to yield a family of cubic expanders with this property ( as per the statement of theorem [ thm-1 ] ) .
the graph we construct will contain a smaller explicit expander on a fixed proportion of its vertices , connected to what is essentially a product of another expander with a `` stretched '' regular tree ( one where the edges in certain levels are replaced by paths ) .
let @xmath51 and let @xmath52 be two explicit expanders as follows ( cf .
e.g. @xcite for an explicit construction of a @xmath41-regular graph , as well as @xcite and the references therein for additional explicit constructions of constant - degree expander graphs ) : * @xmath53 : an explicit @xmath41-regular expander on @xmath54 vertices .
* @xmath55 : an explicit @xmath56-regular expander on @xmath57 vertices .
let @xmath58 denote the largest absolute - value of any nontrivial eigenvalue of @xmath59 for @xmath60 .
finally , let @xmath61 be some sufficiently large fixed integer whose value will be specified later . our final construction for @xmath62-regular expander will be based on a ( modified ) regular tree , hence it will be convenient to describe its structure according to the various levels of its vertices .
let the vertex @xmath63 denote the root of the tree , and construct the graph @xmath26 as follows : levels @xmath64 : first levels of a @xmath62-regular tree rooted at @xmath63 .
= 1em denote by @xmath65 the vertices comprising level @xmath66 .
[ cons - part-1 ] levels @xmath67 : stretched @xmath56-ary trees rooted at @xmath68 : = 1em for each @xmath69 place an @xmath70-level @xmath56-ary tree @xmath71 rooted at @xmath72 and identify the vertices of @xmath71 and @xmath73 via the trivial isomorphism . [ cons - part-2 ] replace every edge of each @xmath71 by a ( disjoint ) path of length @xmath61 .
connect @xmath74 , the new interior vertices in @xmath75 ( with initial degree @xmath66 ) to their isomorphic counterparts in @xmath76 ( add @xmath56-cliques between identified interior vertices ) and similarly for @xmath77 etc .
let @xmath78 denote the final @xmath79 vertices comprising level @xmath80 , and associate the vertices of @xmath78 with those of @xmath53 .
[ cons - part-3 ] levels @xmath81 : product of @xmath53 and a stretched @xmath56-ary tree .
= 1em for each @xmath82 place an @xmath70-level @xmath61-stretched @xmath56-ary tree @xmath83 .
connect vertices in @xmath84 with their counterparts in @xmath85 for @xmath86 .
let @xmath87 denote the final @xmath88 vertices comprising level @xmath89 .
levels @xmath90 : a forest of @xmath56-ary trees rooted at @xmath87 .
last level :
associate leaves with @xmath55 and interconnect them accordingly . -regular graph on which the random walk exhibits total - variation cutoff.,width=432 ] finally , the aforementioned parameter @xmath61 is chosen as follows : denote by @xmath91 the minimum spectral - gaps in the explicit expanders that were embedded in our construction ( recalling from the introduction that @xmath92 for both @xmath60 by the definition of expanders together with the discrete analogue of cheeger s inequality ) , and define @xmath93 see fig . [ fig - expcons ] for an illustration of the above construction .
we begin by establishing the expansion of the above constructed @xmath26 . throughout the proof
we omit ceilings and floors in order to simplify the exposition .
[ lem - exp ] let @xmath94 for @xmath55 our explicit @xmath56-regular expander .
for any integer @xmath95 , the cheeger constant of the above described @xmath62-regular graph @xmath26 with parameter @xmath61 satisfies @xmath96 . moreover , the induced subgraph @xmath97 on the last @xmath70 levels ( i.e. , levels @xmath98 ) has @xmath99 .
first consider the entire graph @xmath26 .
since we are only interested in a lower bound on @xmath100 , clearly it is valid to omit edges from the graph , in particular we may erase the cross edges between any subtrees @xmath101 described in items [ cons - part-2],[cons - part-3 ] of the construction .
this converts every stretched edge of the @xmath62-regular tree of @xmath26 simply into a @xmath66-path ( one where all interior vertices have degree 2 ) of length @xmath61 .
next , we contract all the above mentioned @xmath66-paths into single edges and denote the resulting graph by @xmath102 .
the next simple claim shows that this decreases the cheeger constant by at most @xmath103 .
let @xmath102 be a connected graph with maximal degree @xmath104 and let @xmath26 be a graph on at most @xmath105 vertices obtained via replacing some of the edges of @xmath102 by @xmath66-paths of length @xmath61
. then @xmath106 .
let @xmath107 be a set of cardinality at most @xmath108 achieving the cheeger constant of @xmath26 .
we may assume that @xmath109 otherwise @xmath102 is a disjoint union of paths and cycles and the result holds trivially .
notice that if @xmath110 contains two endpoints of a @xmath66-path @xmath111 while only containing @xmath112 interior vertices of @xmath113 then we can assume that @xmath114 , i.e. all the interior vertices are adjacent ( this maintains the same cardinality of @xmath110 while not increasing @xmath115 ) . with this in mind ,
modify the set @xmath110 into the set @xmath116 by repeating the following operation : as long as there is a 2-path @xmath113 as above ( with @xmath117 and @xmath118 in @xmath110 for some @xmath112 ) we replace @xmath118 by @xmath119 .
this maintains the cardinality of the set while increasing its edge - boundary by at most @xmath120 ( as @xmath118 formerly contributed at least @xmath121 edge to this boundary due to @xmath122 ) .
altogether , this yields a set @xmath116 where no @xmath66-path @xmath123 has both @xmath124 , while @xmath116 satisfies @xmath125 the obtained subset @xmath116 is possibly disconnected , and we will next argue that its connected components satisfy an appropriate isoperimetric inequality .
consider @xmath126 , the connected component of @xmath116 that minimizes @xmath127 .
if @xmath126 is completely contained in the interior of one of the new @xmath66-paths then the statement of the claim immediately holds since @xmath128 with the last inequality due to the fact @xmath129 .
suppose therefore that this is not the case hence we may now assume that @xmath126 contains at least one endpoint of any @xmath66-path it intersects .
let @xmath130 , i.e. the subset of the vertices of @xmath102 obtained from @xmath126 by excluding any vertex that was created in @xmath26 due to subdivision of edges .
observe that our assumption on @xmath116 implies that @xmath131 since either a @xmath66-path @xmath113 is completely contained in @xmath126 ( not contributing to @xmath132 ) or @xmath133 for some @xmath134 ( contributing the edge @xmath135 to @xmath132 , corresponding to the edge @xmath136 in @xmath137 ) .
it remains to consider @xmath138 .
clearly , @xmath126 can be obtained from @xmath139 by adding at most @xmath104 new 2-paths with @xmath140 new interior points per vertex , hence @xmath141 on the other hand , since @xmath142 and @xmath143 we have @xmath144 which together with the fact that @xmath145 implies that @xmath146 altogether , @xmath147 in light of the above claim we have @xmath148 where the graph @xmath102 is the result of taking a complete @xmath62-regular tree of height @xmath149 levels and connecting its @xmath150 leaves , denoted by @xmath151 , via the @xmath56-regular expander @xmath55 .
it therefore remains to show that @xmath152 .
let @xmath30 be a set of size @xmath153 vertices that achieves @xmath154 .
define its subset of the leaves @xmath155 and set @xmath156 . since @xmath157 we clearly have @xmath158 hence @xmath159 .
we thus have the following two options : 1 .
@xmath160 : in this case @xmath161 2 .
@xmath162 : letting @xmath163 denote the infinite @xmath62-regular tree ( whose cheeger constant equals 3 ) we get @xmath164 altogether we deduce that @xmath165 the second part of the lemma ( the statement on the subgraph @xmath97 ) follows from essentially the same argument given above for @xmath154 , as @xmath97 is precisely a forest of @xmath62-regular trees of height @xmath70 where all the leaves are connected via the expander @xmath55 .
again we get @xmath166 , completing the proof .
consider the started @xmath167 levels above the bottom ( i.e. at level @xmath168 ) of the graph @xmath26 .
the height of the walk is then a one - dimensional biased random walk with positive speed @xmath169 , implying that it would reach the bottom after @xmath170 steps with high probability .
on the other hand , if the is started closer to the root , i.e. at level @xmath171 , then the one - dimensional random walk is delayed by two factors , horizontal ( cross - edges ) and vertical ( stretching the edges into paths ) . until reaching level @xmath89 ( after which the previous analysis applies ) ,
these delays are encountered along @xmath172 stretched edges with the following effect : the former incurs a laziness delay with probability @xmath169 whenever the walk is positioned on an interior vertex of a 2-path .
the latter delays the walk by the passage time of a through an @xmath61-long 2-path , where the walk leaves the origin with probability @xmath121 .
it is well - known ( and easy to derive ) that the expected passage time of the one - dimensional from @xmath1 to @xmath173 is precisely @xmath174 and the expected number of visits to the origin by then ( including the starting position ) is exactly @xmath61 .
it thus follows that the expected delay of the one - dimensional walk representing our height in the tree along a single stretched edge is @xmath175 combining the above cases we arrive at the following conclusion : [ clm - tau - ell ] consider the _
_ on the graph @xmath26 started at a vertex on level @xmath176 .
set @xmath177 and let @xmath178 be the hitting time of the walk to the leaves ( i.e. to level @xmath149 )
. then w.h.p .
@xmath179 h & \mbox{if $ 0 \leq \alpha \leq 2$}\,,\\ \frac53 ( 3-\alpha ) h+o(h ) & \mbox{if $ \alpha \geq 2 $ } \ , .
\end{array}\right.\ ] ] the next lemma relates @xmath178 , the hitting time to the leaves ( addressed by the above claim ) , and the mixing of the on the graph .
[ lem - tmix - upper - bound - from - top ] let @xmath180 , let @xmath181 be some vertex on level @xmath182 and @xmath183 for @xmath184 fixed , where @xmath178 is the hitting time of the _ _ to the leaves .
then @xmath185 for any sufficiently large @xmath70 .
let @xmath186 denote the started at some vertex @xmath181 in level @xmath187 and @xmath4 be the uniform distribution on @xmath188 .
let @xmath189 be a random walk started from the uniform distribution @xmath190 .
write @xmath191 for @xmath192 for the vertices of level @xmath193 in @xmath26 ( accounting for all the vertices except interior ones along the 2-paths of length @xmath61 corresponding to stretched edges ) and let @xmath194 map vertices in the graph to their level ( while mapping interior vertices of @xmath66-paths to the lower of their endpoint levels ) .
further let @xmath195 .
clearly , for large enough @xmath70 we have @xmath196 furthermore , due to the bias of the towards the leaves and the fact that @xmath197 ( recall claim [ clm - tau - ell ] and that @xmath61 is fixed ) we deduce that @xmath198 except with probability exponentially small in @xmath70 , and in particular for any sufficiently large @xmath70 @xmath199 therefore , an elementary calculation shows that if @xmath200 and @xmath201 are close in total - variation and so are @xmath202 given @xmath203 and @xmath204 given @xmath205 for all @xmath206 , then the required statement on @xmath207 would follow .
namely , if we should show that at time @xmath208 we have @xmath209 then we would get that @xmath210 ( with room to spare ) .
examine the period spent by @xmath186 in levels @xmath211 .
the graph in these levels is essentially a product of a @xmath56-ary tree whose edges are stretched into @xmath61-long @xmath66-paths and the expander @xmath53 .
let @xmath212 map the vertices in these levels to their corresponding vertices in @xmath53 , and let @xmath213 be the hitting times of @xmath186 to levels @xmath214 and @xmath89 respectively .
as argued above , the started at level @xmath171 is a one - dimensional biased random walk that w.h.p
. passes through @xmath215 stretched edges until reaching level @xmath89 for the first time .
in particular , between times @xmath213 the walk w.h.p . passes through @xmath216 stretched edges . along each
stretched edge among levels @xmath211 , the walk traverses a cross - edge in @xmath53 ( that is , @xmath217 is uniformly distributed over the neighbors of @xmath218 in @xmath53 ) with probability @xmath169 whenever it is in an interior vertex in the @xmath66-path , for a total expected number of @xmath219 such moves . finally , due to its bias towards the leaves , with high probability
the from level @xmath220 reaches level @xmath89 ( the vertices @xmath87 ) before hitting level @xmath80 . applying clt
we conclude that the w.h.p .
traverses @xmath221 cross - edges ( each corresponding to a single step of the on @xmath53 ) between times @xmath213 , where the last inequality holds for @xmath222 and large enough @xmath70 .
this amounts to at least @xmath223 consecutive steps of a along @xmath53 .
aiming for a bound on the total - variation mixing , we may clearly condition on events that occur with high probability : condition therefore throughout the proof that indeed the above statement holds .
in particular , letting @xmath2 be the on the expander @xmath53 and recalling that @xmath224 and @xmath225 it follows that @xmath226 recalling that @xmath53 is a @xmath41-regular with second largest ( in absolute value ) eigenvalue @xmath227 and writing @xmath228 , @xmath229 < |a|^{-2}\,,\end{aligned}\ ] ] with the last inequality justified for any sufficiently large @xmath70 provided that @xmath230 a fact inferred from the choice of @xmath61 in . in this case , for any @xmath181 @xmath231 by symmetry we now conclude that for any @xmath232 and @xmath233 we have @xmath234 this immediately establishes . to obtain ,
note that @xmath235 for @xmath236 w.h.p .
satisfies @xmath237 . conditioned on this event
we can apply a monotone - coupling to successfully couple @xmath186 and @xmath189 such that @xmath238 , yielding .
the concentration of @xmath178 established in claim [ clm - tau - ell ] carries the above two bounds to time @xmath239 , thus completing the proof .
let @xmath240 denote the total - variation mixing time from a given starting position @xmath241 .
that is , if @xmath2 is an ergodic markov chain on a finite state space @xmath3 with stationary distribution @xmath4 then @xmath242 the above lemma gives an upper bound on this quantity for the started at one of the levels @xmath243 , which we now claim is asymptotically tight : [ cor - mixing - top ] consider the _ _ on @xmath26 started at some vertex @xmath181 on level @xmath182 and let @xmath178 be the hitting time of the walk to the leaves .
then for any fixed @xmath244 we have @xmath245 .
the upper bound on @xmath246 was established in lemma [ lem - tmix - upper - bound - from - top ] . for a matching lower bound on @xmath247 choose some fixed integer @xmath248 such that the bottom @xmath249 levels of the graph comprise at least a @xmath250-fraction of the vertices of @xmath26 , i.e. @xmath251 the lower bound now follows from observing that , by the same arguments that established claim [ clm - tau - ell ] , the hitting time from level @xmath252 to level @xmath253 is w.h.p .
@xmath254 for any sufficiently large @xmath70 .
having established the asymptotic mixing time of the started at the top @xmath70 levels , we next wish to show that from all other vertices the mixing time is faster .
[ clm - mix - bottom ] let @xmath186 be the _ _ started at some vertex @xmath241 in level @xmath255 . for every @xmath13 and
any sufficiently large @xmath70 we have @xmath256 .
let @xmath257 denote the induced subgraph on the bottom @xmath258 levels of the graph ( i.e. , levels @xmath98 ) . since @xmath259 it clearly suffices to show that @xmath186 mixes within total - variation distance @xmath260 on @xmath257 . by claim [ clm - tau - ell ]
( taking the worst case @xmath261 corresponding to @xmath262 ) with high probability we have @xmath263h < 5l^2 h { \stackrel{\scriptscriptstyle\triangle}{=}}t_1\,,\ ] ] where the above strict inequality holds for any sufficiently large @xmath70 ( as @xmath264 ) .
recall that @xmath265 by lemma [ lem - exp ] , where @xmath55 is the explicit @xmath56-regular expander with second largest ( in absolute value ) eigenvalue @xmath266 .
further consider the graph @xmath267 obtained by adding to @xmath257 a perfect matching on level @xmath89 , thus making it @xmath62-regular .
clearly , adding edges can only increase the cheeger constant and so @xmath268 as well .
moreover , the discrete form of cheeger s inequality ( @xcite ) , which for a @xmath25-regular expander @xmath269 with second largest ( in absolute value ) eigenvalue @xmath270 states that @xmath271 here gives the following : @xmath272 in particular we obtain that @xmath273 satisfies @xmath274 while a simple random walk @xmath2 on @xmath267 is well - known to satisfy @xmath275 as @xmath276 we infer that after @xmath277 ( the strict inequality holding for large enough @xmath70 ) steps we have @xmath278 recall that our choice of @xmath279 is such that @xmath280 w.h.p . , i.e. the reaches level @xmath149 by that time , and thereafter ( due to its bias towards the leaves ) it does not revisit level @xmath89 until time @xmath281 except with a probability that is exponentially small in @xmath70 . since @xmath26 and @xmath267 are identical on levels @xmath90 we deduce that w.h.p . the performs at least @xmath282 consecutive steps in @xmath267 following @xmath178 .
altogether , for large enough @xmath70 we have @xmath283 and so @xmath284 . finally , bearing and the choice of @xmath61 in in mind , @xmath285 hence @xmath286 and so @xmath287 and @xmath288 , as required .
we now claim that the worst - case mixing time within any @xmath13 is attained by an initial vertex at distance @xmath289 from the root .
fix @xmath290 , let @xmath241 be the initial vertex maximizing @xmath291 and recall that @xmath292 denotes its level in the graph .
the combination of claim [ clm - tau - ell ] and corollary [ cor - mixing - top ] ensures that if @xmath293 then necessarily @xmath294 , in which case @xmath295 an immediate consequence of the requirement [ eq - l - cond2 ] on @xmath61 is that @xmath296 , hence @xmath297 for any sufficiently large @xmath70 .
therefore , we can not have @xmath298 since by claim [ clm - mix - bottom ] that would imply that @xmath299 contradicting the fact that @xmath241 achieves the worst - case mixing time .
overall we deduce that for any @xmath13 we have @xmath300 thus confirming that the on the above constructed family of @xmath62-regular expanders exhibits total - variation cutoff from a worst starting location .
it remains to describe how our construction can be ( relatively easily ) modified to be @xmath41-regular rather than @xmath62-regular .
the immediate step is to use binary trees instead of @xmath56-ary trees , after which we are left with the problem of embedding the explicit expanders @xmath53 and @xmath55 without increasing the degree .
this will be achieved via the line - graphs of these expanders , hence our explicit expanders will now have slightly different parameters : * @xmath53 : an explicit @xmath41-regular expander on @xmath301 vertices . *
@xmath55 : an explicit @xmath41-regular expander on @xmath302 vertices . recall that given a tree rooted at some vertex @xmath303 , denoted by @xmath304 , its edge - stretched version is obtained by replacing each edge by a @xmath66-path of length @xmath61 , and the collection of all new interior vertices ( due to subdivision of edges ) is denoted by @xmath305 .
the modified construction is as follows : 1 .
levels 0,1,2 : first levels of a binary tree .
* denote by @xmath306 the vertices in level @xmath66 .
levels @xmath67 : stretched binary trees rooted at @xmath68 : * connect vertices from @xmath307 ( interior vertices along @xmath66-paths ) to the corresponding ( isomorphic ) vertices in @xmath308 , i.e. inter - connect the interior vertices via perfect matchings .
* denote by @xmath78 the @xmath309 vertices in level @xmath80 .
levels @xmath81 : edge - stretched binary trees rooted at @xmath78 and inter - connected via the line - graph of @xmath53 using auxiliary vertices : * associate each binary tree @xmath310 rooted at @xmath78 to an _ edge _ of @xmath53 .
* for each @xmath311 ( interior vertex on a @xmath66-path ) we connect it to a new auxiliary vertex @xmath312 and associate @xmath312 with a unique edge of @xmath53 .
* we say that @xmath313 and @xmath314 are isomorphic if the isomorphism from @xmath310 to @xmath315 maps @xmath241 to @xmath316 .
add @xmath317 new auxiliary vertices per equivalence class of @xmath318 such isomorphic vertices , identify them with the vertices of @xmath53 and connect every new vertex @xmath319 to the auxiliary vertices @xmath320 representing the edges incident to it .
levels @xmath90 : a forest of binary trees .
last level : leaves are inter - connected via the line - graph of @xmath55 : * associate the @xmath321 vertices with the edges of @xmath55 . *
add @xmath322 new auxiliary vertices , each connected to the leaves corresponding to edges that are incident to it in @xmath55 .
it is easy to verify that the walk along the cross - edges of the @xmath323 s now corresponds to a lazy ( unbiased ) random walk on the edges of @xmath53 .
similarly , the walk along the cross - edges connecting the leaves corresponds to the on the edges of @xmath55 .
hence , all of the original arguments remain valid in this modified setting for an appropriately chosen fixed @xmath61 .
this completes the proof of theorem [ thm-1 ] .
the explicit cubic expanders with cutoff constructed in the previous section ( illustrated in fig .
[ fig - expcons ] ) can be easily modified so that the on them from a worst starting position would _ not _ exhibit total - variation cutoff .
to do so , recall that in the above - described family of graphs , each vertex of the subset @xmath68 was the root of a regular tree of height @xmath70 whose edges were stretched into @xmath61-long @xmath66-paths ( see item [ cons - part-1 ] of the construction ) .
we now tweak this construction by stretching some of the edges into @xmath66-paths of length @xmath324 .
namely , for subtrees rooted at the _ odd _ vertices in level @xmath325 of these trees we stretch the edges into paths of length @xmath326 . under this modified stretching the trees @xmath71
are clearly still isomorphic , hence the cross edges are inter - connecting @xmath66-paths of the same lengths . by the arguments above , starting from any level @xmath327 the mixing is faster compared to the root , and if @xmath328 is sufficiently small then the root remains the asymptotically worst starting position .
however , starting from the root ( and in fact , starting from any level @xmath329 ) the hitting time to the set @xmath78 is no longer concentrated due to the odd / even choice of subtree at level @xmath325 .
therefore , from the worst starting position we have that the hitting time to the leaves is concentrated on two distinct values ( differing by a fixed multiplicative constant ) , each with probability @xmath330 .
this implies that the ratio @xmath331 is bounded away from @xmath121 and in particular this explicit family of expanders does not have total - variation cutoff .
suppose @xmath269 is an explicit @xmath41-regular expander on @xmath332 vertices provided by theorem [ thm-1 ] , and recall that the on this graph exhibits cutoff at @xmath333 where @xmath334 is some absolute constant .
our graph @xmath26 will be the result of replacing every edge of @xmath269 by the @xmath41-regular analogue of a @xmath66-path , which we refer to as a `` cylinder '' , illustrated in fig . [ fig - expconsgen ] .
the length of each cylinder is set to be @xmath335 satisfying @xmath336 . notice that the total number of vertices in @xmath26 is @xmath337 since @xmath338 and the on @xmath269 , started at a worst - case starting position , traverses @xmath339 edges until mixing , we infer from clt , as well as the fact that the expected passage - time through an @xmath61-long cylinder is @xmath174 , that the analogous random walk on @xmath26 has cutoff at @xmath340 when @xmath341 we have @xmath36 . on the other extreme end , when @xmath61 grows arbitrarily fast as a function of @xmath332 we obtain that it approaches @xmath27 arbitrarily closely but we must still having @xmath342 since @xmath343 . in that case @xmath344
approaches @xmath345 arbitrarily closely while having a strictly smaller order . to complete the construction it remains to observe that we may choose @xmath61 and @xmath332 so that @xmath49 and @xmath50 .
this can be achieved by first selecting @xmath61 so that @xmath346 and then selecting a graph constructed through theorem [ thm-1 ] on @xmath332 vertices for some @xmath347 ( note that in the theorem we construct graphs of size essentially @xmath348 so this is always possible ) . to show that there is no cutoff whenever @xmath43 we will argue that in that case we have @xmath349 , in contrast to the necessary condition for cutoff @xmath350 due to peres ( cf .
@xcite ) , discussed in the introduction .
[ lem - trel - no - cutoff ] let @xmath26 be a graph on @xmath27 vertices with degrees bounded by some @xmath104 fixed on which the _ _ has @xmath43 .
then the spectral - gap of the walk satisfies @xmath351 .
in particular , the _ _ on @xmath26 does not exhibit cutoff .
observe that the above graph must satisfy @xmath352 for some fixed @xmath32 as it is well - known ( cf .
, e.g. , @xcite ) that the lazy walk on any graph @xmath269 has @xmath353 and in our case @xmath354 .
let @xmath355 be two vertices whose distance in @xmath26 is @xmath356 our lower bound on the gap will be derived from its representation via the dirichlet form , according to which @xmath357 ^ 2\pi(x)p(x , y)}{\operatorname{var}_\pi f}\,,\end{aligned}\ ] ] where @xmath4 is the ( uniform ) stationary measure and @xmath358 is the transition kernel of the . as a test - function @xmath359 in the above form
choose @xmath360 clearly we have @xmath361 and a lower bound of order @xmath345 on the variance follows from the fact that two sets of linear size each have a linear discrepancy according to @xmath362 .
namely , @xmath363 as a result of which @xmath364 we conclude that @xmath349 , thus completing the proofs of lemma [ lem - trel - no - cutoff ] and theorem [ thm-2 ] .
= 2em recent results in @xcite showed that almost every regular expander graph has total - variation cutoff ( prior to that there were no known examples for bounded - degree graphs with this phenomenon ) ; here we provided a _ first explicit _
construction for bounded - degree expanders with cutoff .
the expanders constructed in this work are non - transitive
. moreover , our proof exploits their highly asymmetric structure in order to control the mixing time of the random walk from various starting locations .
it would be interesting to obtain an explicit construction of transitive expanders with total - variation cutoff . a slight variant of our construction gives an example of a family of expanders where the does _ not _ exhibit cutoff , thereby disagreeing with peres cutoff - criterion .
both here and in another such example due to peres and wilson @xcite the expanders are non - transitive ( hence the restriction to transitive graphs in peres conjecture stated next ) . while it is conjectured by peres that the random walk on any family of transitive bounded - degree expanders exhibits total - variation cutoff , there is not even a single example of such a transitive family where cutoff was proved ( or disproved ) . for general ( not necessarily expanding ) bounded - degree graphs on @xmath27 vertices
it is well - known that @xmath365 .
here we showed that cutoff can occur essentially anywhere up to @xmath366 by constructing cubic graphs with cutoff at any such prescribed location .
furthermore , this is tight as we prove that if @xmath367 then cutoff can not occur . | the cutoff phenomenon describes a sharp transition in the convergence of an ergodic finite markov chain to equilibrium . of particular interest
is understanding this convergence for the simple random walk on a bounded - degree expander graph .
the first example of a family of bounded - degree graphs where the random walk exhibits cutoff in total - variation was provided only very recently , when the authors showed this for a typical random regular graph .
however , no example was known for an explicit ( deterministic ) family of expanders with this phenomenon .
here we construct a family of cubic expanders where the random walk from a worst case initial position exhibits total - variation cutoff .
variants of this construction give cubic expanders without cutoff , as well as cubic graphs with cutoff at any prescribed time - point . |
electroweak symmetry breaking is a cornerstone for our understanding of particle physics . however , despite the spectacular phenomenological success of the standard model ( sm ) , the fundamental mechanism of electroweak symmetry breaking remains a mystery .
various new models have been proposed to explain this mechanism .
one such class of models is based upon topcolor @xcite which can generate a large top - quark mass .
the topcolor model also predicts a @xmath0 @xcite .
the physics in production and decay of the @xmath0 are discussed in ref .
@xcite under different model assumptions . in this paper
, we consider the @xmath0 from model iv which represents a novel class and has predicted cross sections large enough to be experimentally accessible at hadron colliders at the fermilab tevatron and the large hadron collider ( lhc ) at cern .
such @xmath0 resonances couple strongly only to the first and third generations of quarks , and have no significant couplings to the leptons .
they are , therefore , leptophobic and topophyllic .
this article presents an updated calculation valid for the tevatron and all proposed lhc collision energies .
the leptophobic topcolor @xmath0 decaying to @xmath5 has been searched for at both the tevatron @xcite and the lhc @xcite .
the tevatron searches used a previous calculation of the @xmath0 cross section @xcite . the lhc search conducted by the cms collaboration
have used the cross section calculation presented here to set limits on the @xmath0 mass .
this article is primarily intended to document that @xmath0 cross section used for a @xmath3 collision energy of @xmath6 tev . for current and future lhc searches ,
we present the @xmath0 cross section at the current lhc @xmath3 collision energy of @xmath7 tev and the potential future collision energies of 10 and 14 tev . we also present the cross section in @xmath8 collisions at @xmath9 tev for comparison with tevatron searches .
as discussed in ref .
@xcite , non - standard models can be constructed in which the @xmath10 and the generations are grouped differently .
model iv is the case where quark generations @xmath11 .
the dominant part of the interaction lagrangian for model iv is : l_iv & = & ( g_1_h)z^ ( |t_l_t_l + |b_l_b_l + f_1|t_r_t_r . + & & .
+ f_2|b_r_b_r -|u_l_u_l -|d_l_d_l -f_1|u_r_u_r .
-f_2|d_r_d_r ) , where @xmath12 is the sm coupling constant , and we require the following : @xmath13 ( attractive @xmath14 channel ) and/or @xmath15 ( repulsive @xmath16 channel ) . also , @xmath17 to avoid fine - tuning .
the parton - level subprocess cross section that we use in the next section is obtained from this lagrangian .
the total lowest - order cross section for a @xmath0 produced in hadron collisions and decaying into top - antitop pairs from model iv discussed above , is given by : @xmath18 where @xmath19 , the differential cross section at @xmath5 invariant mass @xmath20 , is given by @xmath21 here @xmath22 is the parton - level subprocess cross section .
the kinematic variable @xmath23 is related to the initial - state parton fractional momenta inside the first hadron @xmath24 and the second hadron @xmath25 by @xmath26 .
the boost of the partonic system @xmath27 is given by @xmath28 .
the partonic `` luminosity function '' is just the product of parton distribution functions : @xmath29 where @xmath30 ( @xmath31 ) is the parton distribution function of a quark ( anti - quark ) evaluated at fractional momenta @xmath32 and renormalization scale @xmath33 .
the parton - level subprocess cross section , @xmath22 , for a leptophobic , b@xmath34-phobic , and topophyllic , @xmath0 is given by : @xmath35 \left [ \frac{s}{(\hat{s}-m_{z'}^2)^2 + \hat{s}\gamma_{z'}^2 } \right]\theta(\hat{s}-4m_t^2 ) , \label{eq_sigma_iv}\end{aligned}\ ] ] where @xmath36 can be obtained from the total decay - width of the resonance @xcite : @xmath37 . } \label{eq_gamma_iv}\ ] ] the subprocess cross section in eq .
[ eq_sigma_iv ] is for spin and color summing on both initial and final state legs , while most parton distributions assume spin and color averaged on the initial - state legs and spin and color summing on the final - state legs .
therefore , the subprocess cross section given by eq .
[ eq_sigma_iv ] must be multiplied by a factor of @xmath38 when used with parton distributions from pdflib @xcite and other standard sources .
we have taken this into account when calculating the cross section .
we have also used @xmath39 gev / c@xmath40 , and @xmath41 .
for a default parton distribution set we have chosen cteq6l @xcite .
this is a modern parton distribution set appropriate for leading - order calculations and is available in pdflib @xcite . for a default renormalization scale
we choose @xmath42 , half the @xmath5 invariant mass .
this scale has the benefit that it reduces to the usual @xmath43 at top production threshold , but also increases with increasing @xmath5 invariant mass . with these choices ,
the total cross section for the production of a leptophobic and topophyllic @xmath0 and its subsequent decay into @xmath44 pairs is presented in sections [ sec : tevatron ] and [ sec : lhc ] for the tevatron and the lhc .
we calculate numerically the lowest order cross section for the process @xmath45 at the tevatron at @xmath9 tev , using eq .
[ eq_tot_xsec ] and taking into account the spin - color factor in eq .
[ eq_spin_color ] .
the only remaining parameter of the topcolor model that affects the cross section is the mixing angle @xmath46 , or equivalently the width @xmath47 which is related to it .
we calculate the cross section for different choices of @xmath47 , equal to @xmath48 and @xmath49 of @xmath50 , both of which qualify as narrow resonances at the tevatron .
the integration in eq .
[ eq_tot_xsec ] is performed using the full available phase space of @xmath51 .
the results are tabulated in table [ tab : zprimexsectev2 ] and displayed in fig .
[ fig : tevatron ] . | we present numerical calculations of the production cross section of a heavy @xmath0 resonance in hadron - hadron collisions with subsequent decay into top - antitop pairs . in particular , we consider the leptophobic topcolor @xmath0 discussed under model iv of hep - ph/9911288 which has predicted cross sections large enough to be experimentally accessible at the fermilab tevatron and the large hadron collider at cern .
this article presents an updated calculation valid for the tevatron and all proposed lhc collision energies .
cross sections are presented for various @xmath0 widths , in @xmath1 collisions at @xmath2 tev , and in @xmath3 collisions at @xmath4 , 8 , 10 and 14 tev .
fermilab - fn-0940-cms - ppd + |
a very interesting empirical phenomenon in the study of weighted networks is the power - law correlation between strength @xmath0 and degree @xmath1 of nodes @xmath2 @xcite . very recently , wang _
et al _ have proposed a mutual selection model to explain the origin of this power - law correlation @xcite .
this model can provide a partial explanation for social weighted networks , that is , although the general people want to make friend with powerful men , these powerful persons may not wish to be friendly to them .
however , this model can not explain the origin of power - law strength - degree correlation in weighted technological networks . in many cases ,
the concepts of edge - weight and node - strength are associated with network dynamics .
for example , the weight in communication networks is often defined by the load along with the edge @xcite , and the strength in epidemic contact networks is defined by the individual infectivity @xcite . on the one hand , although the weight / strength distribution may evolve into a stable form , the individual value is being changed with time by the dynamical process upon network . on the other hand , the weight / strength distribution will greatly affect the corresponding dynamic behaviors , such as the epidemic spreading and synchronization @xcite . inspired by the interplay of weight and network dynamics , barrat , barthlemy , and vespignani proposed an evolution model ( bbv model for short ) for weighted networks @xcite .
although this model can naturally reproduce the power - law distribution of degree , edge - weight , and node - strength , it fails to obtain the power - law correlation between strength and degree . in bbv model ,
the dynamics of weight and network structure are assumed in the same time scale , that is , in each time step , the weight distribution and network topology change simultaneously .
here we argue that the above two time scales are far different .
actually , in many real - life situations , the individual weight varies momently whereas the network topology only slightly changes during a relatively long period .
similar to the traffic dynamics based on the local routing protocol @xcite , we investigate the dynamic behaviors of resource / traffic flow on scale - free networks with given structures , which may give some illuminations about the origin of power - law correlation between strength and degree in weighted scale - free networks .
as mentioned above , strength usually represents resources or substances allocated to each node , such as wealth of individuals of financial contact networks @xcite , the number of passengers in airports of world - wide airport networks @xcite , the throughput of power stations of electric power grids @xcite , and so on .
these resources also flow constantly in networks : money shifts from one person to another by currency , electric power is transmitted to every city from power plants by several power hubs , and passengers travel from one airport to another . further more , resources prefers to flow to larger - degree nodes . in transport networks ,
large nodes imply hubs or centers in traffic system .
so passengers can get a quick arrival to the destinations by choosing larger airports or stations . in financial systems , people also like to buy stocks of larger companies or deposit more capital in the banks with more capital because larger companies and banks generally have more power to make profits and more capacity to avoid losses .
inspired by the above facts , we propose a simple mechanism to describe the resource flow with preferential allocation in networks .
[ 0.5 ] resources in node @xmath3 are divided into several pieces and then flow to its neighbors .
the thicker lines imply there are more resources flowing .
it is worth pointing out that , in order to give a clearer illustration we do not plot the resource flow into node @xmath3 or out of node @xmath4.,title="fig : " ] at each time , as shown in fig .
1 , resources in each node are divided into several pieces and then flow to its neighbors .
the amount of each piece is determined by its neighbors degrees .
we can regulate the extent of preference by a tunable parameter @xmath5 .
the equations of resource flow are @xmath6 where @xmath7 is the amount of resources moving from node @xmath3 to @xmath4 at time @xmath8 , @xmath9 is the amount of resources owned by node @xmath3 at time @xmath8 , @xmath10 is the degree of node @xmath4 and @xmath11 is the set of neighbors of node @xmath3 .
if @xmath4 and @xmath3 are not neighboring , then @xmath12
. meanwhile each node also gets resources from its neighbors , so at time @xmath13 , @xmath14 @xmath4 @xmath15 [ 1 ]
( color online ) the evolution of the strengths of node @xmath16 and @xmath17 , where nodes @xmath16 and @xmath17 are randomly selected for observation .
the three cases are in different initial states which simply satisfy @xmath18 . the exponent @xmath19.,title="fig : " ] [ 0.8 ] scatter plots of @xmath20 vs @xmath10 for all the nodes , title="fig : " ]
the eq . [ strength of node i ] can be expressed in terms of a matrix equation , which reads @xmath21 where the elements of matrix @xmath22 are given by @xmath23 since @xmath24 , @xmath25 , the spectral radius of matrix @xmath22 obeys the equality @xmath26 , according to the gershgrin disk theorem @xcite . here , the spectral radius , @xmath27 , of a matrix @xmath22 , is the largest absolute value of an eigenvalue .
further more , since the considered network is symmetry - free ( that is to say , the network is strongly connected thus for any two nodes @xmath4 and @xmath3 , these exists at least one path from @xmath4 to @xmath3 ) , @xmath28 will converge to a constant matrix for infinite @xmath1 .
that is , if given the initial boundary condition to eq .
[ matrix ] ( e.g. let @xmath29 , where @xmath30 denotes the total number of nodes in network ) , then @xmath31 will converge in the limit of infinite @xmath8 as @xmath32 for each node @xmath4 .
consequently , denote @xmath33 , one can obtain @xmath34 that is , for any @xmath4 , @xmath35 from eq .
[ kinetic equilibrium ] , it is clear that @xmath36 is just the kinetic equilibrium state of the resource flow in our model . since @xmath37 , @xmath36 is determined only by matrix @xmath22 , if given the initial boundary condition with @xmath38 satisfying @xmath39 .
since matrix @xmath22 is determined by the topology only , for each node @xmath4 in the kinetic equilibrium , @xmath40 is completely determined by the network structure .
@xmath20 denotes the amount of resource eventually allocated to node @xmath4 , thus it is reasonable to define @xmath20 as the strength of node @xmath4 .
[ 0.8 ] ( color online ) the correlation between degree and strength with different @xmath5 . in the inset
, the relation between @xmath41 and @xmath5 is given , where the squares come from the simulations and the solid line represents the theoretical result @xmath42.,title="fig : " ] [ 0.8 ] ( color online ) the distribution of strength with different @xmath5 .
the inset exhibits the relation between @xmath43 and @xmath5 , where the squares come from the simulations and the solid line represents the theoretic analysis @xmath44.,title="fig : " ]
the solution of eq . ( 6 ) reads @xmath45 where @xmath46 is a normalized factor . in principle
, this solution gives the analytical relation between @xmath20 and @xmath10 when @xmath47 can be analytically obtained from the degree distribution . for uncorrelated networks
@xcite , statistically speaking we have @xmath48 where @xmath49 denotes the probability a randomly selected node is of degree @xmath1 . since @xmath50 is a constant when given a network structure , one has @xmath51 , thus @xmath52 where @xmath53 denotes the average strength over all the nodes with degree @xmath1 .
this power - law correlation @xmath54 where @xmath42 , observed in many real weighted networks , can be considered as a result of the conjunct effect of the above power - law correlation and the scale - free property . obviously ,
if the degree distribution in a weighted network obeys the form @xmath55 , one can immediately obtain the distribution of the strength @xmath56 where the power - law exponent @xmath44 .
recent empirical studies in network science show that many real - life networks display the scale - free property @xcite , thus we use scale - free networks as the samples .
since the barabsi - albert ( ba ) model @xcite is the mostly studied model and lacks structural - biases such as non - zero degree - degree correlation , we use ba network with size @xmath57 and average degree @xmath58 for simulations .
the dynamics start from a completely random distribution of resource . as is shown in fig .
2 , we randomly pick two nodes @xmath16 and @xmath17 , and record their strengths vs time @xmath59 and @xmath60 for three different initial conditions . clearly , the resource owned by each node will reach a stable state quickly . and no matter how and where the one unit resource flow in , the final state is the same .
[ 0.8 ] ( color online ) the distributions of degree ( left panel ) and strength ( right panel ) with different @xmath5 .
the networks are generated by the strength - pa mechanism , and those shown here are the sampling of size @xmath57.,title="fig : " ] [ 0.8 ] ( color online ) the correlation between degree and strength with different @xmath5 . in the inset ,
the relation between @xmath41 and @xmath5 is given , where the squares come from the simulations and the solid line represents the theoretical result @xmath42 .
the networks are generated by the strength - pa mechanism , and those shown here are the sampling of size @xmath57.,title="fig : " ] similar to the mechanism used to judge the weight of web by google - searching ( see a recent review paper @xcite about the _ pagerank algorithm _ proposed by google ) , the strength of a node is not only determined by its degree , but also by the strengths of its neighbors ( see eq . 7 ) .
although statistically @xmath61 for uncorrelated networks , the strengths of the nodes with the same degree may be far different especially for low - degree nodes as exhibited in fig .
3 . in succession
, we average the strengths of nodes with the same degree and plot fig . 4 to verify our theoretical analysis that there is a power - law correlation @xmath62 between strength and degree , with exponent @xmath42 .
5 shows that the strength also obeys power - law distribution , as observed in many real - life scale - free weighted networks . and the simulations agree well with analytical results .
in this paper , we proposed a model for resource - allocation dynamics on scale - free networks , in which the system can approach to a kinetic equilibrium state with power - law strength - degree correlation . if the resource flow is unbiased ( i.e. @xmath63 ) , similar to the bbv model @xcite , the strength will be linearly correlated with degree as @xmath64 . therefore , the present model suggests that the power - law correlation between degree and strength arises from the mechanism that resources in networks tend to flow to larger nodes rather than smaller ones .
this preferential flow has been observed in some real traffic systems .
for example , very recently , we investigated the empirical data of chinese city - airport network , where each node denote a city , and the edge - weight is defined as the number of passengers travelling along this edge per week @xcite .
we found that the passenger number from one city to its larger - degree neighbor is much larger than that from this city to its smaller - degree neighbor .
in addition , in chinese city - airport network @xcite and us airport network @xcite , the power - law exponents are @xmath65 and @xmath66 , respectively , which is within the range of @xmath41 predicted by the present model .
the readers should be warned that the analytical solution shown in this paper is only valid for static networks without any degree - degree correlation .
however , we have done some further simulations about the cases of growing networks ( see appendix a ) and correlated networks ( see appendix b ) .
the results are quantitatively the same with slight difference in quantity .
[ 0.9 ] ( color online ) the correlation between degree and strength with different assortative coefficients @xmath67 .
the parameter @xmath68 is fixed .
the inset shows the numerically fitting value of @xmath41 vs assortative coefficients .
the networks are generated by the generalized ba algorithm @xcite of size @xmath57 and average degree @xmath58.,title="fig : " ] finally , in this model , the resource flow will approach to a kinetic equilibrium , which is determined only by the topology of the networks , so we can predict the weight of a network just from its topology by the equilibrium state . therefore , our proposed mechanism can well apply to estimate the behaviors in many networks .
when given topology of a traffic network , people can easily predict the traffic load in individual nodes and links by using this model , so that this model may be helpful to a better design of traffic networks . the authors wish to thank miss . ming zhao for writing the c++ programme that can generate the scale - free networks with tunable assortative coefficients . this work has been partially supported by the national natural science foundation of china under grant nos .
70471033 , 10472116 , and 10635040 , the special research founds for theoretical physics frontier problems under grant no .
a0524701 , and specialized program under president funding of chinese academy of science .
since many real networks , such as www and internet , are growing momently .
the performance of the present resource - allocation flow on growing networks is thus of interest .
we have implemented the present dynamical model on the growing scale - free networks following the usual preferential attachment ( pa ) scheme of barabsi - albert @xcite .
since the topological change is independent of the dynamics taking place on it , and the relaxation time before converging to a kinetic equilibrium state is very short ( see fig .
2 ) , if the network size is large enough ( like in this paper @xmath69 ) , then the continued growth of network has only very slight effect on topology and the results is almost the same as those of the ungrowing case shown above . furthermore , we investigate the possible interplay between the growing mechanism and the resource - allocation dynamics . in this case , the initial network is a few fully connected connected nodes , and the resource is distributed to each node randomly .
then , the present resource - allocation process works following eq .
( 2 ) , and simultaneously , the network itself grows following a strength - pa mechanism instead of the degree - pa mechanism proposed by ba model . that is to say , at each time step , one node is added into the network with @xmath70 edges attaching to the existing nodes with probability proportional to their strengths ( in a growing ba network , the corresponding probability is proportional to their degrees ) .
clearly , under this scenarios , there exists strong interplay between network topology and dynamic . when the network becomes sufficient large ( @xmath69 ) , as shown in fig .
6 , the evolution approaches a stable process with both the degree distribution and strength distribution approximately following the power - law forms .
furthermore , we report the relationship between strength and degree in fig . 7 , which indicate that the power - law scaling , @xmath71 with @xmath42 , also holds even for the growing networks with strong interplay with the resource - allocation dynamics .
note that , the eq .
( 8) is valid under the assumption that the underlying network is uncorrelated . however , many real - life networks exhibit degree - degree correlation in some extent . in this section , we will investigate the case of correlated networks .
the model used in this section is a generalized ba model @xcite : starting from @xmath72 fully connected nodes , then , at each time step , a new node is added to the network and @xmath70 ( @xmath73 ) previously existing nodes are chosen to be connected to it with probability @xmath74 where @xmath75 and @xmath10 denote the choosing probability and degree of node @xmath4 , respectively . by varying the free parameter @xmath76 @xmath77 , one can obtain the scale - free networks with different assortative coefficients @xmath67 ( see the ref .
@xcite for the definition of assortative coefficient ) .
the simulation results are shown in fig .
8 , from which one can find that the power - law correlations between strength and degree in the correlated networks are quantitatively the same as that of the uncorrelated networks , however , the power - law exponents , @xmath41 , are slightly different . actually , in the positive correlated networks , the large - degree nodes prefer to connect with some other large - degree nodes rather than those small - degree nodes , thus there may exist a cluster consisting of large - degree nodes that can hold the majority of resource .
that cluster makes the large - degree nodes having even more resource than in the uncorrelated case , thus leading to a larger @xmath41 . in the inset of fig .
8 , one can find that @xmath41 is larger in the positive correlated networks , and smaller in the negative correlated networks .
however , the analytical solution have not yet achieved when taking into account the degree - degree correlation , which needs a in - depth analysis in the future .
li2004 w. li , and x. cai , phys .
e * 69 * , 046106 ( 2004 ) .
a. barrat , m. barthlemy , r. pastor - satorras , and a. vespignani , proc .
101 * , 3747 ( 2004 ) .
wang , b. -h .
wang , b. hu , g. yan , and q. ou , phys . rev . lett . * 94 * , 188702 ( 2005 ) . h. -k .
liu , and t. zhou , acta physica sinica ( to be published ) .
wang , b. hu , t. zhou , b. -h .
wang , and y. -b .
xie , phys .
e * 72 * , 046140 ( 2005 ) .
m. e. j. newman , and m. girvan , phys .
e * 69 * , 026113 ( 2004 ) .
t. zhou , z. -q .
fu , and b. -h .
wang , prog .
* 16 * , 452 ( 2006 ) .
g. yan , t. zhou , j. wang , z. -q .
fu , and b. -h .
wang , chin .
* 22 * , 510 ( 2005 ) .
a. e. motter , c. zhou , and j. kurths , phys .
e * 71 * , 016116 ( 2005 ) .
m. chavez , d. -u .
hwang , a. amann , h. g. e. hentschel , and s. boccaletti , phys .
lett . * 94 * , 218701 ( 2005 ) .
m. zhao , t. zhou , and b. -h .
wang , eur .
j. b ( to be published ) .
a. barrat , m. barthlemy , and a. vespignani , phys .
. lett . * 92 * , 228701 ( 2004 ) .
a. barrat , m. barthlemy , and a. vespignani , phys .
e * 70 * , 066149 ( 2004 ) .
p. holme , adv .
complex syst .
* 6 * , 163 ( 2003 ) b. tadi , s. thurner and g. j. rodgers , phys .
e * 69 * , 036102 ( 2004 ) . c. -y .
yin , b. -h .
wang , w. -x .
wang , t. zhou , and h. -j .
yang , phys .
lett . a * 351 * , 220 ( 2006 ) .
wang , b. -h .
wang , c. -y .
yin , y. -b .
xie , and t. zhou , phys .
e * 73 * , 026111 ( 2006 ) .
xie , b. -h .
wang , b. hu , and t. zhou , phys .
e * 71 * , 046135 ( 2005 ) .
r. guimera , s. mossa , a. turtschi , and l. a. n. amaral , proc .
* 102 * , 7794 ( 2005 ) .
r. albert , i. albert , and g. l. nakarado , phys .
e * 69 * , 025103 ( 2004 ) .
r. a. horn , and c. r. johnson , _ matrix analysis _ ( cambridge university press , cambridge , 1985 ) .
m. e. j. newman , phys .
lett . * 89 * , 208701 ( 2002 ) .
r. albert , and a. -l .
barabsi , rev .
* 74 * , 47 ( 2002 ) .
barabsi , and r. albert , science * 286 * , 509 ( 1999 ) .
k. bryan , and t. leise , siam rev . * 48 * , 569 ( 2006 ) .
s. n. dorogovtsev , j. f. f. mendes , and a. n. samukhin , phys .
lett . * 85 * , 4633 ( 2000 ) .
p. l. krapivsky , and s. redner , phys .
e * 63 * , 066123 ( 2001 ) . | many weighted scale - free networks are known to have a power - law correlation between strength and degree of nodes , which , however , has not been well explained .
we investigate the dynamic behavior of resource / traffic flow on scale - free networks .
the dynamical system will evolve into a kinetic equilibrium state , where the strength , defined by the amount of resource or traffic load , is correlated with the degree in a power - law form with tunable exponent .
the analytical results agree well with simulations . |
in this work , we consider fractional integrals and fractional derivatives with respect to another function @xmath0 ( see @xcite ) .
to fix notation , in the following @xmath1 is a real , @xmath2 $ ] is an increasing function , such that @xmath3 , for all @xmath4 $ ] , and @xmath5\to\mathbb r$ ] is a function with some assumptions , so that the fractional operators that we deal with are well defined .
the left fractional integral of @xmath6 , with respect to @xmath0 of order @xmath7 , is defined as @xmath8 and the right fractional integral of @xmath6 is @xmath9 considering special cases for the kernel , that is , for the function @xmath0 , we recover e.g. the riemann
liouville , the hadamard and the erdlyi kober integrals . for fractional derivatives of riemann
liouville type , the left and right fractional derivatives of @xmath6 are defined as @xmath10 and @xmath11 respectively , where @xmath12 + 1 $ ] . in this paper , we deal mainly with a caputo - type fractional derivative .
the concept is similar to the riemann
liouville derivative , but the order of the dual integral / derivative is switched ( see @xcite ) .
let @xmath13 $ ] be a function .
the left caputo fractional derivative of @xmath6 of order @xmath7 with respect to @xmath0 is given by @xmath14 while the right caputo fractional derivative of @xmath6 is given by @xmath15 where @xmath16 + 1 \ , \mbox { for } \ , \a\notin\mathbb n , \quad n=\a\ , \mbox { for } \ , \a\in\mathbb n.\ ] ] it results that , if @xmath17 is an integer , then @xmath18 on the other hand , if @xmath19 , then @xmath20 and @xmath21 since we are interested in a generalization of the ordinary derivatives , we will consider this second case only .
for example , we have the following formulas . given @xmath22 , @xmath23 and @xmath24 also , given @xmath25 , we have @xmath26 and @xmath27 there exists a relation between the fractional integral and the fractional derivative operators . in a certain sense , they are the inverse operation of each other .
in fact , we have that @xmath28 and @xmath29 for the converse , the relation is the following : @xmath30 one crucial formula , when dealing with variational problems , is a form of integration by parts , with respect to the fractional operators . for these caputo - type fractional derivatives ,
they are as follows .
[ integrationparts ] given @xmath31 $ ] and @xmath32 $ ] , the following holds : @xmath33_{t = a}^{t = b}\end{aligned}\ ] ] and @xmath34_{t = a}^{t = b}.\end{aligned}\ ] ] the paper is organized in the following way . in section [ sec : main ] , we present the main problem , and in theorem [ teo : main ] we prove an euler lagrange type equation . in section [ extension ]
we extend this result , by considering functionals where the lower bound of integration is greater than the lower bound of the fractional derivative .
next , in section [ sec : iso ] , we consider the variational problem subject to an integral constraint , in what is known as an isoperimetric problem , and in section [ sec : legendre ] , we deduce a second - order necessary condition that allows us to verify if the extremals are minimizers or not . in sections
[ sec : infinite ] and [ sec : delay ] , we consider the infinite horizon problem and the case where the lagrange function has a delay , respectively . in section [ sec
: high ] , we consider high - order derivatives in the functional , and derive the respective high - order euler lagrange equation , and in section [ sec : optimal ] , we find a necessary condition that allows us to find the best fractional order to provide a minimum to the functional .
finally , in section [ sec : suff ] , we prove a sufficient condition that guarantees the solutions of the euler
lagrange equations are almost minimizers .
in this section , we study several variational problems , where the dynamic of the trajectories is described by a caputo type fractional derivative .
we consider the initial point to be fixed , @xmath35 ( @xmath36 ) , and the terminal point @xmath37 to be free , and thus it is also a variable of the problem .
we are interested in finding the optimal pair @xmath38 for the objective functionals .
the most important result in the calculus of variations is the so called euler lagrange equation , which is a first order necessary condition every extremizer of the functional must satisfy . for functionals depending on fractional operators ,
we find in the literature numerous works already done for different kinds of fractional derivatives and initial / terminal conditions .
some examples are for the riemann liouville derivative @xcite , for the caputo derivative @xcite , for the riesz derivative @xcite .
we mention the recent books @xcite , where analytical and numerical methods are explained , respectively .
our fractional variation problem with free terminal time is described in the following way .
let @xmath39\times\mathbb r^2\to\mathbb r$ ] be a continuous function , such that there exist and are continuous the functions @xmath40 and @xmath41 .
define the functional @xmath42,\ ] ] where @xmath43 is the set @xmath44 : x(a)\ , \mbox { is a fixed real}\},\ ] ] which we endow with the norm @xmath45}|x(t)|+\max_{t\in[a , b]}\left|\ld x(t)\right|.\ ] ] we say that @xmath46 assumes its minimum value at @xmath47 in @xmath48 $ ] , relative to the norm @xmath49 provided that @xmath50 \ , : \ , \|(x^*,t^*)-(x , t)\|<\epsilon \rightarrow j(x^*,t^ * ) \leq j(x , t).\ ] ] in this case , we say that @xmath47 is a local minimum for @xmath46 . an admissible variation for @xmath47 is a pair @xmath51 , where @xmath52 $ ] and @xmath53 , @xmath54 and @xmath55 .
the next result provides a necessary condition that every local minimum for @xmath46 must satisfy . in order to simplify the notation
, we define @xmath56 $ ] as @xmath57(t):=(t , x(t),\ld x(t ) ) , \quad ( x , t)\in \omega\times[a , b].\ ] ] [ teo : main ] suppose that @xmath47 is a local minimum for @xmath46 as in on the space @xmath48 $ ] .
if there exists and is continuous the function @xmath58(t)/\p'(t)\right)$ ] on @xmath59 $ ] , then @xmath60(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0\ ] ] for each @xmath61 $ ] , and at @xmath62 , the following transversality conditions are satisfied : @xmath63(t)}{\p'(t)}\right)=0\\ l[x^*](t)=0 .
\end{array } \right.\ ] ] consider an admissible variation of the optimal solution of the form @xmath51 .
if we define the function @xmath64 in a neighborhood of zero by the expression @xmath65 we have that @xmath66 .
differentiating @xmath64 at @xmath67 , and using the integration by parts formula as in theorem [ integrationparts ] , we obtain @xmath68(t ) \cdot v(t)+\partial_3 l[x^*](t ) \cdot \ld v(t)\right]dt+\triangle t\cdot l[x^*](t^*)\\ & = \int_a^{t^*}\left[\partial_2 l[x^*](t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)\right]v(t)\,dt\\ & \quad + \left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^ * } + \triangle t\cdot l[x^*](t^*).\\\end{aligned}\ ] ] since @xmath53 , if we consider @xmath69 , by the fundamental lemma of the calculus of variations ( cf .
* lemma 2.2.2 ) ) , we conclude that for all @xmath61 $ ] , @xmath70 satisfies the condition @xmath71(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0.\ ] ] therefore , we have @xmath72(t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^ * } + \triangle t \cdot l[x^*](t^*).\ ] ] since @xmath73 and @xmath74 are free , we obtain the two transversality conditions .
equations like are called euler lagrange equations , and they provide a first - order necessary condition that all minimizers of the problem must satisfy . notice that , although the functional depends on a caputo type fractional derivative , the euler lagrange equation involves a riemann liouville fractional derivative .
we can rewrite it in such a way that the fractional equation depends on the caputo derivative as well . observe that , given a differentiable function @xmath75 and @xmath76 , we have @xmath77 using this new relation , the euler lagrange equation is written in the form @xmath71(t)+{^cd_{t^*-}^{\a,\p}}\left(\frac{\partial_3
l[x^*](t)}{\p'(t)}\right)\p'(t)+ \frac{\partial_3 l[x^*](t^*)}{\p'(t^*)\gamma(1-\a)}(\p(t^*)-\p(t))^{-\a}\p'(t)=0.\ ] ] the variational problem involving several dependent variables is similar , and we omit the proof here .
consider the functional @xmath78(t)\,dt,\ ] ] where @xmath79 , the functions @xmath80 verify the two assumptions @xmath81 $ ] and @xmath82 is a fixed real , for all @xmath83 , the real @xmath84 belongs to the interval @xmath85 $ ] , and @xmath57(t):=(t , x_1(t),\ldots , x_m(t),{^cd_{a+}^{\a_1,\p } } x_1(t),\ldots,{^cd_{a+}^{\a_m,\p } } x_m(t ) ) , \quad \a_1,\ldots,\a_m\in(0,1).\ ] ] suppose that @xmath86 is a local minimum for @xmath46 , and that there exist and are continuous the functions @xmath87(t)/\p'(t)\right)$ ] on @xmath59 $ ] , for all @xmath83 .
then , @xmath88(t)+{d_{t^*-}^{\a_i,\p}}\left(\frac{\partial_{i+1+m } l[x^*](t)}{\p'(t)}\right)\p'(t)=0\ ] ] for all @xmath83 and for all @xmath61 $ ] , and at @xmath62 , the following holds : @xmath89(t)}{\p'(t)}\right)=0 , \quad i=1,\ldots , m\\ l[x^*](t)=0 .
\end{array } \right.\ ] ] in the previous problem , the lower limits of the cost functional and of the fractional derivative were the same , at @xmath90 . in this section
we generalize it , by considering a cost functional starting at a point @xmath91 .
consider the functional @xmath92(t)\,dt,\ ] ] where @xmath93 $ ] ( @xmath94 and @xmath95 may be fixed or not ) and @xmath96 $ ] with @xmath97 .
assume that @xmath47 is a local minimum for @xmath46 , and that there exist and are continuous the functions @xmath58(t)/\p'(t)\right)$ ] on @xmath59 $ ] and @xmath98(t)/\p'(t)\right)$ ] on @xmath99 $ ] .
then , @xmath100(t)}{\p'(t)}\right)-{d_{a-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)=0,\ ] ] for each @xmath101 $ ] , and @xmath71(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0\ ] ] for each @xmath102 $ ] . at @xmath62 ,
the following holds : @xmath103(t)}{\p'(t)}\right)=0\\ l[x^*](t)=0 .
\end{array } \right.\ ] ] moreover , if @xmath94 is free , then at @xmath90 : @xmath104(t)}{\p'(t)}\right)-{i_{a-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)=0,\ ] ] and if @xmath95 is free , then at @xmath105 : @xmath106(t)}{\p'(t)}\right)=0.\ ] ] the first variation of the functional at an extremum must vanish , and so we conclude that @xmath107(t ) \cdot v(t)+\partial_3 l[x^*](t ) \cdot \ld v(t)\right]dt+\triangle t\cdot l[x^*](t^*)\\ & = \int_a^{t^*}\left[\partial_2 l[x^*](t ) \cdot v(t)+\partial_3 l[x^*](t ) \cdot \ld v(t)\right]dt\\ & \quad -\int_a^a\left[\partial_2 l[x^*](t ) \cdot v(t)+\partial_3 l[x^*](t ) \cdot \ld v(t)\right]dt+\triangle t\cdot l[x^*](t^*)\\ & = \int_a^{t^*}\left[\partial_2 l[x^*](t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)\right]v(t)\,dt + \left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^*}\\ & \quad -\int_a^a\left[\partial_2 l[x^*](t)+{d_{a-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)\right]v(t)\,dt - \left[{i_{a-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = a}\\ & \quad + \triangle t\cdot l[x^*](t^*)\\ & = \int_a^a\left[{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)-{d_{a-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\right ] \p'(t)\cdot v(t)\,dt \\ & \quad + \int_a^{t^*}\left[\partial_2 l[x^*](t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)\right]\cdot v(t)\,dt\\ & \quad + \left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^ * } - \left[{i_{a-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = a}\\ & \quad + \triangle t\cdot l[x^*](t^*).\end{aligned}\ ] ] by the arbitrariness of @xmath108 and @xmath74 , we obtain the necessary optimality conditions .
we formulate now the variational problem when in presence of an integral constraint .
we refer to @xcite , where similar problems were solved involving fractional derivatives .
this kind of problems are known in the literature as isoperimetric problems .
the most ancient problem of this type goes back to the ancient greece , with the question of finding out which of all closed planar curves of the same length would enclose the greatest area .
nowadays , an isoperimetric problem is a variational problem , restricted to a subclass of functions satisfying a side condition of the form @xmath109 here , we replace the ordinary derivative by a fractional derivative , and since the terminal time is free , the integral value is not a constant , but a function depending on the terminal time
. let @xmath110\times\mathbb r^2\to\mathbb r$ ] be a continuous function , such that there exist and are continuous the functions @xmath111 and @xmath112 .
[ teo : isop ] suppose that @xmath47 is a local minimum for @xmath46 as in on the space @xmath48 $ ] , subject to the integral constraint @xmath113(t)\,dt=\phi(t),\ ] ] where @xmath114 $ ] .
suppose that @xmath47 is not a solution of the equation @xmath115(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 m[x](t)}{\p'(t)}\right)\p'(t)=0 , \quad
\forall t \in[a , t^*],\ ] ] and that there exist and are continuous the functions @xmath116(t)/\p'(t)\right)$ ] and @xmath117(t)/\p'(t)\right)$ ] on @xmath59 $ ] . then , there exists a real constant @xmath118 , such that if we define the augmented function @xmath119 , then @xmath47 satisfies the equation @xmath120(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 f[x](t)}{\p'(t)}\right)\p'(t)=0 , \quad \forall t \in[a , t^*],\ ] ] and the system @xmath121(t)}{\p'(t)}\right)=0\\ f[x](t)=\lambda \phi'(t ) \end{array}\right .
\quad \mbox{at } \ , \ , t = t^*.\ ] ] consider admissible variations of two parameters of kind @xmath122 , where @xmath123 $ ] with @xmath124 , and @xmath125 with @xmath126 .
define the two functions : @xmath127 since @xmath128(t)+{d_{t^*-}^{\a,\p } } \left(\frac{\partial_3 m[x^*](t)}{\p'(t)}\right)\p'(t)\right]v_1(t)\,dt\\ & \quad+ \left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 m[x^*](t)}{\p'(t)}\right)\cdot v_1(t)\right]_{t = a}^{t = t^ * } , \end{aligned}\ ] ] and @xmath47 is not a solution for eq .
, we deduce that there exists a function @xmath129 $ ] such that @xmath130 .
we can appeal to the implicit function theorem , which asserts that there exists a function @xmath131 , defined on a neighborhood of zero , such that @xmath132 .
thus , there exists a subfamily of admissible variations satisfying the integral constraint . attending that @xmath64 is minimum at @xmath133 subject to the constraint @xmath134 , and since @xmath135 ,
by the lagrange multiplier rule , there exists a real number @xmath118 such that @xmath136 in particular , @xmath137 . repeating the calculations as done before , we arrive at the desired formulas .
we now formulate a second - order necessary condition , usually called legendre condition , which provides us with a necessary condition for minimization . in @xcite , by the first time , a legendre type condition was obtained for fractional variational calculus .
here , we derive a similar condition to a more general form of fractional derivative .
assume now that the lagrange function @xmath138 is such that its second order partial derivatives @xmath139 , with @xmath140 , exist and are continuous .
suppose that @xmath47 is a local minimum for @xmath46 as in on the space @xmath48 $ ] .
then for all @xmath141 $ ] , @xmath142(t)\geq0.\ ] ] let us consider variations over @xmath70 only , that is , we restrict to the case @xmath143 .
so , if we consider @xmath144 , we have @xmath145 , and so we conclude that @xmath146(t)\cdot v^2(t)+2\,\partial^2_{23}l[x^*](t)\cdot v(t)\ld v(t)+\partial^2_{33}l[x^*](t)\cdot \left(\ld v(t)\right)^2\right]\,dt\geq0.\ ] ] suppose that the legendre condition is violated at some some @xmath147 $ ] : @xmath142(t_0)<0.\ ] ] then , there exists a subinterval @xmath148\subseteq [ a , t^*]$ ] and three real constants @xmath149 with @xmath150 such that @xmath151(t)\leq c_1 , \quad \partial^2_{23}l[x^*](t)\leq c_2 ,
\quad \partial^2_{33}l[x^*](t)\leq c_3,\ ] ] for all @xmath152 $ ] .
define the function @xmath153\to\mathbb r$ ] by the formula @xmath154 then , @xmath155 and @xmath156 .
once @xmath157 we also have @xmath158 . besides this , for every @xmath159 $ ] , @xmath160 and @xmath161 define the function @xmath162\to\mathbb r$ ] by the rule @xmath163\\ 0 , & \quad \mbox{if } t \notin [ c , d].\\ \end{array}\right.\ ] ] by the properties of function @xmath164 , we have that @xmath52 $ ] , @xmath53 and @xmath165\\ 0 , & \quad \mbox{if } t \notin[c , d].\\ \end{array}\right.\ ] ] note that , for @xmath166 , @xmath167 . replacing this variation into eq . , we get @xmath168(t)\cdot v^2(t)+2\,\partial^2_{23}l[x^*](t)\cdot v(t)\ld v(t)+\partial^2_{33}l[x^*](t)\cdot(\ld v(t))^2\right]\,dt\\ & \leq \int_a^{t^*}\left [ 14 ^ 2 c_1(\p(d)-\p(c))^{2\a+2}+2\cdot 14\cdot 50 c_2(\p(d)-\p(c))^{\a+2}+50 ^ 2c_3 ( \p(d)-\p(c))^2 \right]\,dt\\ & = ( \p(d)-\p(c))^2 ( t^*-a)\left [ 196 c_1 ( \p(d)-\p(c))^{2\a}+1400c_2(\p(d)-\p(c))^\a+2500c_3 \right]<0\end{aligned}\ ] ] if we assume that @xmath169 , and thus we obtain a contradiction .
we study now a new problem , important when we want to consider the effects at a long term .
this issue is especially pertinent when the planning horizon is assumed to be of infinite length .
the objective functional is given by an improper integral , the initial state @xmath94 is fixed and the terminal state ( at infinity ) is free , that is , no constraints are imposed on the behaviour of the admissible trajectories at large times .
this kind of problems are known as infinite horizon problems , where the objective functional is given by @xmath170 where @xmath171 is the set @xmath172 endowed with the norm @xmath173 we have to be careful when defining a minimal curve for functional , since any admissible function for which the improper integral diverges to @xmath174 would be a minimal path , according to the usual definition of minimum . here
, we follow the one presented in @xcite .
a curve @xmath70 in @xmath171 is a local weakly minimal for @xmath46 as in if there exists some @xmath175 such that , for all @xmath176 , if @xmath177 , then the lower limit @xmath178(t)-l[x](t ) ] \ , dt \leq 0.\ ] ] for the following result , we will need some extra functions .
fixed two functions @xmath179 , and given @xmath180 and @xmath181 , define @xmath182(t)- l[x^*](t)}{\epsilon}dt;\\ v(\epsilon , t ) & : = \inf_{t^ * \geq t}\int_a^{t^ * } [ l[x^*+\epsilon v](t)-l[x^*](t)]dt;\\ w(\epsilon)&:= \lim_{t\to\infty } v(\epsilon , t).\end{aligned}\ ] ] let @xmath70 be a local weakly minimal for @xmath46 as in .
suppose that : 1 .
@xmath183 exists for all @xmath84 ; 2 .
@xmath184 exists uniformly for all @xmath185 ; 3 . for every @xmath186 and @xmath187
, there exists a sequence @xmath188 such that @xmath189 uniformly for @xmath185 . if there exists and is continuous the function @xmath58(t)/\p'(t)\right)$ ] on @xmath59 $ ] , for all @xmath190 , then @xmath71(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0,\ ] ] for all @xmath190
also , we have @xmath191(t)}{\p'(t)}\right)=0 \quad \mbox{at } \ , \ , t = t^*.\ ] ] by the definition of minimum curve for the infinite horizon problem , we have that @xmath192 in a neighborhood of zero , and @xmath193 .
thus , @xmath194 and so we have the following : @xmath195(t)- l[x^*](t)}{\epsilon } \ , dt\\ & = \lim_{t\to\infty } \inf_{t^ *
\geq t}\int_{a}^{t^*}\left[\partial_2 l[x^*](t ) \cdot v(t)+\partial_3 l[x^*](t ) \cdot \ld v(t)\right]dt\\ & = \lim_{t\to\infty } \inf_{t^ * \geq t}\left[\int_a^{t^*}\left[\partial_2 l[x^*](t)+{d_{t^*-}^{\a,\p } } \left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)\right]v(t)\,dt\right.\\ & \quad \left.+\left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^*}\right].\end{aligned}\ ] ] if we assume that @xmath196 , we deduce that @xmath197(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right ) \p'(t)\right]v(t)\,dt=0,\ ] ] and so ( see @xcite ) @xmath71(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0,\ ] ] for all @xmath198 and for all @xmath199
. also , using this last condition , we get @xmath191(t)}{\p'(t)}\right)=0 \quad \mbox{at } \ , \ , t = t^*.\ ] ] in this section we consider time - delay variational problems .
this is an important subject , since in many systems there is almost always a time delay @xcite .
a natural generalization of such theory is to replace ordinary derivatives by fractional derivatives , since fractional operators contain memory , and their present state is determined by all past states .
there exist already some works dealing with fractional operators , for example @xcite .
let @xmath39\times\mathbb r^3\to\mathbb r$ ] be a continuous function such that there exist and are continuous the functions @xmath200 , for @xmath201 .
given @xmath202 such that @xmath203 , define the functional @xmath204,\ ] ] where @xmath205 : x(t)=\theta(t)\ , \mbox { for } \ , t\in[a-\t , a]\},\ ] ] and @xmath206 is a given function .
let @xmath56_\t$ ] denote the vector @xmath57_\t(t):=(t , x(t),x(t-\t),\ld x(t)).\ ] ] [ teo : delay ] let the pair @xmath47 be local minimum for @xmath46 as in .
if there exist and are continuous the functions @xmath207(t)/\p'(t)\right)$ ] and @xmath208(t)/\p'(t)\right)$ ] on @xmath59 $ ] , then for all @xmath209 $ ] , @xmath71_\t(t)+\partial_3 l[x^*]_\t(t+\t)+{d_{(t^*-\t)-}^{\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\p'(t)\ ] ] @xmath210_\t(s)}{\p'(s ) } \ , ds \cdot \p'(t)=0,\ ] ] and for all @xmath211 $ ] , @xmath71_\t(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\p'(t)=0.\ ] ] also , at @xmath62 , it is true that @xmath212_\t(t)}{\p'(t)}\right)=0\\ l[x^*]_\t(t)=0 . \end{array } \right.\ ] ] let @xmath213 $ ] be such that @xmath214 , for all @xmath215,$ ] and consider variations of the form @xmath51 . since the first variation of the functional must vanish at an extremum point , we have @xmath216_\t(t ) \cdot v(t)+\partial_3 l[x^*]_\t(t ) \cdot v(t-\t)+ \partial_4 l[x^*](t ) \cdot \ld v(t)\right]dt+\triangle t\cdot l[x^*](t^*)=0.\ ] ] observe that @xmath217_\t(t ) \cdot v(t-\t)\,dt=\int_a^{t^*-\t } \partial_3 l[x^*]_\t(t+\t ) \cdot v(t)\,dt\ ] ] since @xmath214 , for all @xmath215 $ ] .
also , since @xmath218_\t(t)}{\p'(t)}\right ) & = { d_{(t^*-\t)-}^{\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\\ & \quad-\frac{1}{\gamma(1-\a)}\left(\frac{1}{\p'(t)}\frac{d}{dt}\right)\int_{t^*-\t}^{t^*}\p'(s)(\p(s)-\p(t))^{-\a}\frac{\partial_4 l[x^*]_\t(s)}{\p'(s)}\,ds\end{aligned}\ ] ] we obtain the following @xmath219_\t(t ) \cdot \ld v(t)\,dt=\int_a^{t^ * } { d_{t^*-}^{\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\p'(t ) \cdot v(t)\,dt+\left [ { i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\cdot v(t)\right]_a^{t^*}\ ] ] @xmath220_\t(t)}{\p'(t)}\right)\p'(t ) -\frac{1}{\gamma(1-\a)}\left(\frac{1}{\p'(t)}\frac{d}{dt}\right)\int_{t^*-\t}^{t^*}\p'(s)(\p(s)-\p(t))^{-\a } \frac{\partial_4 l[x^*]_\t(s)}{\p'(s)}\,ds\right.\ ] ] @xmath221v(t)\,dt + \int_{t^*-\t}^{t^ * } { d_{t^*-}^{\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\p'(t ) \cdot v(t)\,dt+\left [ { i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\cdot v(t)\right]_a^{t^*}.\ ] ] finally , combining all the previous formulas , we obtain @xmath222_\t(t)+\partial_3 l[x^*]_\t(t+\t)+{d_{(t^*-\t)-}^{\a,\p } } \left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\p'(t)\right.\ ] ] @xmath223_\t(s)}{\p'(s ) } \ , ds \cdot \p'(t ) \right]v(t)dt\ ] ] @xmath224_\t(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_4 l[x^*]_\t(t)}{\p'(t)}\right)\p'(t)\right]v(t)dt\ ] ] @xmath225_\t(t)}{\p'(t)}\right)\cdot v(t)\right]_a^{t^*}+\triangle t\cdot l[x^*]_\t(t^*)=0.\ ] ] since @xmath108 is arbitrary on the interval @xmath59 $ ] , as well as @xmath74 , we obtain the desired result .
so far we considered a fractional order as a real between 0 and 1 . using similar techniques as the ones presented in the proof of theorem [ teo : main ]
, we can generalize the previous results in order to include high order derivatives .
we show how to do it for the basic problem of the calculus of variations , and we deduce the respective euler lagrange equation .
consider the functional @xmath226,\ ] ] where 1 .
@xmath79 , and for all @xmath227 , we have @xmath228 ; 2 . @xmath39\times\mathbb r^{m+1}\to\mathbb r$ ] is a continuous function ; 3 . there exist and are continuous the functions @xmath40 , @xmath41 , @xmath229 , and @xmath230 ; 4 .
@xmath231 : x(a ) , x^{(1)}(a),\ldots , x^{(m-1)}(a ) \ , \mbox { are fixed reals}\}$ ] .
suppose that @xmath47 is a local minimum for @xmath46 .
if , for all @xmath227 , there exist and are continuous the functions @xmath232(t)/\p'(t)\right)$ ] on @xmath59 $ ] , then @xmath71_m(t)+\sum_{n=1}^m{d_{t^*-}^{\a_n,\p}}\left(\frac{\partial_{n+2 } l[x^*]_m(t)}{\p'(t)}\right)\p'(t)=0,\ ] ] for each @xmath61 $ ] , and at @xmath62 , we have @xmath233_m(t)}{\p'(t)}\right)=0 , \quad \ , k=1,\ldots ,
m,\\ l[x^*]_m(t)=0 , \end{array } \right.\ ] ] where @xmath234_m(t):=(t , x^*(t),{^cd_{a+}^{\a_1,\p } } x^*(t),\ldots , { ^cd_{a+}^{\a_m,\p } } x^*(t)).\ ] ] consider admissible variations of the form @xmath51 , where @xmath235 $ ] and @xmath236 , for all @xmath237 .
since the first variation of the functional must vanish at an extremum point , we deduce the following : @xmath216_m(t ) \cdot v(t)+\sum_{n=1}^m\partial_{n+2 } l[x^*]_m(t)\cdot{^cd_{a+}^{\a_n,\p } } v(t)\right]dt+\triangle t\cdot l[x^*]_m(t^*)=0.\ ] ] integrating by parts , we get @xmath238_m(t)+\sum_{n=1}^m{d_{t^*-}^{\a_n,\p}}\left(\frac{\partial_{n+2 } l[x^*]_m(t)}{\p'(t)}\right)\p'(t)\right]v(t)\,dt+ \triangle t\cdot l[x^*]_m(t^*)\ ] ] @xmath239_m(t)}{\p'(t)}\right ) \cdot v(t)\right.\ ] ] @xmath240_m(t)}{\p'(t)}\right)\cdot \left(\frac{1}{\p'(t)}\frac{d}{dt}\right)v(t)\ ] ] @xmath241_m(t)}{\p'(t)}\right)\cdot \left(\frac{1}{\p'(t)}\frac{d}{dt}\right)^{m-1}v(t ) \right]_{t = a}^{t = t^*}=0.\ ] ] choosing appropriate variations , we deduce the result .
one advantage of considering fractional derivatives in modelling phenomena is that they may describe more efficiently the dynamics than ordinary derivatives .
so , a natural question to pose is what should be the order of the fractional operator , in order to minimize the functional .
the next result provides necessary conditions to determine the fractional order .
[ fracorder ] consider the functional @xmath242(t)\,dt,\ ] ] where @xmath93 $ ] with @xmath94 fixed , @xmath243 $ ] and @xmath76 .
assume that @xmath244 is a local minimum for @xmath46 , and that there exists and is continuous the function @xmath58(t)/\p'(t)\right)$ ] on @xmath59 $ ] .
given @xmath61 $ ] , define a function @xmath245 by the rule @xmath246(t)$ ] .
then , @xmath71(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0\ ] ] for each @xmath61 $ ] , and at @xmath62 , the following holds : @xmath103(t)}{\p'(t)}\right)=0\\ l[x^*](t)=0 .
\end{array } \right.\ ] ] also , we have the following equality : @xmath247(t)\cdot \lambda_t'(\a^*)\,dt=0.\ ] ] define @xmath248 where @xmath52 $ ] with @xmath53 and @xmath249 are two fixed reals .
since @xmath66 , using integration by parts , we arrive at the formula @xmath216(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3
l[x^*](t)}{\p'(t)}\right)\p'(t)\right]\cdot v(t)\,dt + \left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^*}\ ] ] @xmath250(t^*)+\triangle \a\cdot \int_a^{t^*}\partial_3 l[x^*](t)\cdot \lambda_t'(\a^*)\,dt=0.\ ] ] by the arbitrariness of @xmath108 , @xmath74 and @xmath251 , we prove the desired result .
so far we deduced necessary conditions that every minimizer of the functional must verify .
now , we will deal with a sufficient condition , and for that some conditions of convexity over the lagrangian are needed .
assume that @xmath138 is convex in @xmath85\times\mathbb r^2 $ ] , in the sense that @xmath252 for all @xmath253 $ ] and @xmath254 . if @xmath47 satisfies the euler lagrange equation and the transversality conditions , then @xmath255 on the space @xmath48 $ ] , where functional @xmath46 is given by .
observe that @xmath256(t)\,dt-\int_a^{t^*}l[x^*](t)\,dt\\ & = \int_a^{t^*}\left[l[x^*+v](t)-l[x^*](t)\right]\,dt+\int_{t^*}^{t^*+\triangle t}l[x^*+v](t)\,dt\\ & \geq\int_a^{t^*}\left[\partial_2 l[x^*](t ) \cdot
v(t)+\partial_3 l[x^*](t ) \cdot \ld v(t)\right]dt+\triangle t\cdot l[x^*+v](c)=\star,\end{aligned}\ ] ] for some @xmath257 between @xmath258 and @xmath259 .
integrating by parts , we get @xmath260(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)\right]v(t)\,dt\\ & \quad+ \left[{i_{t^*-}^{1-\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\cdot v(t)\right]_{t = a}^{t = t^ * } + \triangle t\cdot l[x^*+v](c)\\ & = \triangle t\cdot l[x^*+v](c).\end{aligned}\ ] ] since @xmath261(t^*)=0 $ ] and the map @xmath262(t)$ ] is continuous , given @xmath263 , there exists some @xmath264 such that @xmath265(c)\right|<\frac{\epsilon}{|\triangle t|}\ ] ] ( the case @xmath143 is obvious ) .
thus , @xmath266 we remark that we can not conclude that @xmath267 under these assumptions .
for example , if we consider @xmath268 with the initial condition @xmath269 , it is easy to verify that @xmath138 is convex and that @xmath270 , where @xmath271 $ ] is an arbitrary function , verifies eqs and .
however , we have the following : @xmath272 which is a negative number when @xmath273 .
consider the functional @xmath274\,dt\ ] ] subject to the restriction @xmath275 . from theorem [ teo : main ]
, the necessary condition that every minimizer of the functional must fulfill is the following @xmath276 for all @xmath277 $ ] , and also the two next transversality conditions @xmath278(t)=0 .
\end{array } \right.\ ] ] must be meet at @xmath279 .
once @xmath280 we see that the pair @xmath281 satisfies all three conditions . also , using the legendre condition , a local minimizer for the functional must verify the condition @xmath282(t)\geq0,\ ] ] which for our example is verified , since @xmath283(t)=2 $ ] .
we remark that , for every curve @xmath6 and for every endpoint @xmath84 , the inequality @xmath284\,dt=-\frac23,\ ] ] holds , and that @xmath285 and so @xmath47 is actually the global minimizer of @xmath46 . for our second example , let @xmath286\,dt\ ] ] subject to the restrictions @xmath275 and @xmath287\,dt=\phi(t),\ ] ] where the function @xmath288 is defined as @xmath289 ^ 2\,dt.\ ] ] considering @xmath290 , the augmented function is given by @xmath291 and as we have seen in the previous example , the function @xmath292 satisfies the two first necessary conditions , as stated in theorem [ teo : isop ] . using the second transversality condition of theorem [ teo : isop ] , we determine the optimal time @xmath258 by solving the equation @xmath293 ^ 2.\ ] ] for our next example , we determine the best fractional order to minimize a given functional , using the necessary conditions given by theorem [ fracorder ] .
let @xmath294dt,\ ] ] subject to the initial condition @xmath295 . if we define @xmath296 , then @xmath297 and so @xmath298(t)=0 $ ] , for all @xmath299 .
thus , the three next conditions @xmath71(t)+{d_{t^*-}^{\a,\p}}\left(\frac{\partial_3 l[x^*](t)}{\p'(t)}\right)\p'(t)=0 , \quad t\in[a , t^*],\ ] ] @xmath104(t)}{\p'(t)}\right)=0 , \quad t = t^*\ ] ] and @xmath247(t)\cdot \lambda_t'(\a^*)\,dt=0,\ ] ] are verified for all @xmath258 and @xmath300 .
it remains to solve the equation @xmath261(t^*)=0 $ ] .
since @xmath301(t)=0 \leftrightarrow \p(t)=\p(a)+40^{1/(\a+2)},\ ] ] we establish a relation between @xmath258 and @xmath300 given by condition .
observe that @xmath302\,dt,\ ] ] and if we differentiate with respect to @xmath7 , and then put the resulting function equal to zero , we get @xmath303\,dt=0,\ ] ] where @xmath304 denotes the digamma function .
for example , for @xmath305 and @xmath306 , with initial point @xmath307 , we have @xmath308 and @xmath309 , respectively . the graphs of the functions @xmath310 are given in figure [ fracorder5 ] . solving numerically eq .
, we obtain an approximation of the fractional order that minimizes the problem : @xmath311
work supported by portuguese funds through the cidma - center for research and development in mathematics and applications , and the portuguese foundation for science and technology ( fct - fundao para a cincia e a tecnologia ) , within project uid / mat/04106/2013 . | this paper provides necessary and sufficient conditions of optimality for a variational problem involving a fractional derivative with respect to another function .
fractional euler
lagrange equations are proven for the fundamental problem and when in presence of an integral constraint .
a legendre condition , which is a second - order necessary condition , is also obtained .
other cases , such as the infinite horizon problem , with delays in the lagrangian , and with high - order derivatives , are considered . a necessary condition that the optimal fractional order must satisfy
is proved .
* mathematics subject classification 2010 : * 26a33 , 49k05 , 34a08 .
* keywords : * fractional calculus , euler lagrange equation , legendre condition , isoperimetric problem . |
let @xmath24 be a graph with the vertex set @xmath25 and edge set @xmath26 .
a _ @xmath27-edge colouring _ of a graph @xmath24 is a function @xmath28 such that @xmath29 and @xmath30 for every two adjacent edges @xmath31 and @xmath32 .
the _ chromatic index _ of @xmath24 , denoted by @xmath33 is the minimum number @xmath27 for which @xmath24 has a @xmath27-edge colouring .
a _ @xmath34-factor _ of a graph @xmath24 is a set of independent edges which covers all vertices of @xmath24 .
given a colouring of @xmath24 , a _ rainbow @xmath34-factor _ is a @xmath34-factor all of whose edges have different colours .
we denote the complete graph of order @xmath3 by @xmath35 . a _ zero - sum flow _ of a graph @xmath24 is an assignment of non - zero real numbers to the edges of @xmath24 such that the sum of the values of all edges incident with any given vertex is zero .
let @xmath3 be a natural number .
zero - sum @xmath3-flow _ is a zero - sum flow with values from the set . for a subset @xmath36
, the _ weight of @xmath37 _ is defined to be the sum of the values of all edges of @xmath37 .
such flows have been studied in @xcite .
zero - sum flows are motivated by nowhere - zero flows , which were first introduced by tutte in 1949 @xcite .
let @xmath3 be a positive integer .
latin square of order @xmath3 _ with entries from @xmath38 is an @xmath39 array @xmath40 such that every row and column of @xmath40 is a permutation of @xmath38 .
suppose that @xmath41 is a latin square of order @xmath3 with entries from @xmath38 and @xmath42 is a latin square of order @xmath3 with entries from @xmath43 .
we say that @xmath41 and @xmath42 are _ orthogonal _ provided that , for every @xmath44 and for every @xmath45 , there is a unique cell @xmath46 such that @xmath47 and @xmath48 .
a latin square , @xmath40 , is called _ indempotent _ if @xmath49 for every @xmath50 .
it is well known that for every positive integer @xmath51 , there exist two orthogonal latin squares of order @xmath3 .
transversal _ of a latin square is a set of cells which between them contain each row , column and entry exactly once . if a square @xmath40 has an orthogonal mate , @xmath52 , it is possible to partition the cells of @xmath40 into transversals , @xmath53 , each of which correspond to the positions of the entry @xmath50 in @xmath52 .
we refer the reader to @xcite for notation and further results on latin squares .
a _ @xmath0-@xmath54 design _ @xmath55 ( briefly , @xmath0-design ) , is a pair @xmath56 , where @xmath38 is a @xmath10-set of points and @xmath57 is a collection of @xmath27-subsets of @xmath38 , called _ blocks _ , with the property that every @xmath0-subset of @xmath38 is contained in exactly @xmath58 blocks .
traditionally , the number of blocks and the frequency of occurrences of points in blocks are denoted by @xmath59 and @xmath60 , respectively .
given an indexing of the points and blocks of a @xmath0-design , the _ incidence matrix _ of @xmath55 is the @xmath61 @xmath62-matrix @xmath63 $ ] , where @xmath64 we refer the reader to @xcite for notation and further results on designs . given a @xmath0-design @xmath65 , we define a zero - sum @xmath3-flow of @xmath55 to be a function @xmath66 such that the sum of the block weights around any point is zero , i.e.@xmath67 this is equivalent to finding a vector in the nullspace of the incidence matrix of the design whose entries are all in the set @xmath68 .
zero - sum @xmath3-flow of designs have been previously been studied in @xcite .
a _ parallel class _ of a @xmath0-design is a collection of disjoint blocks which between them contain every point of @xmath38 exactly once .
a @xmath0-design is called _ resolvable _ if there exists a partition of the set of blocks into parallel classes .
an _ @xmath69-resolution class _ is a collection of blocks @xmath70 which contain every point of @xmath38 exactly @xmath69 times .
if the block set @xmath57 can be partitioned into @xmath69-resolution classes we call the design _
@xmath69-resolvable _ ; in this case we denote the number of @xmath69-resolution classes by @xmath71 .
an _ automorphism _ of a @xmath0-design is a permutation of the point set which maps blocks to blocks .
the set of automorphisms of a @xmath0-design forms a group , called the _ automorphism group _ , @xmath24 , of the design .
a @xmath0-@xmath1 design , @xmath72 , is called _ cyclic _ if its automorphism group contains a cycle of length @xmath10 . in such cases
there exists a set of _ starter blocks _ @xmath73 , such that the full design may be obtained by acting on these blocks by @xmath24 .
the orbit of a starter block which has length less than @xmath10 is called a _ short orbit _ ; if the orbit is of length @xmath10 , it is called a _ full orbit_. a @xmath0-@xmath74 design is called a _ steiner triple system _ of order @xmath10 , denoted @xmath75 , and a resolvable sts@xmath9 is called a _ kirkman triple system _ , and denoted kts@xmath9 .
it is well known that a steiner triple system on @xmath10 points exists if and only if @xmath76 or @xmath77 and that a kts@xmath9 exists if and only if @xmath78 .
more generally , we call a @xmath0-@xmath79 design a _ triple system _ and denote it by @xmath80 .
we refer the reader to @xcite for further results on triple systems .
we have the following existence result . [ @xcite ]
a ts@xmath81 exists if and only if @xmath82 and @xmath83 .
a connection between zero sum flows on steiner triple systems and more traditional geometric flows may be obtained by considering the well known embedding of the points of a ts@xmath84 into the torus , so that the blocks form equilateral triangles which may be coloured black and white , in such a way that no triangles of the same colour share an edge @xcite . giving each black triangle weight @xmath34 and each white triangle weight @xmath85 , it is easy to see that this is equivalent to a zero - sum @xmath0-flow of the ts@xmath84 . in this paper
we consider the application of zero - sum flows to triple systems . in @xcite
it was conjectured that every @xmath86 with @xmath87 admits a zero - sum @xmath12-flow .
we generalise this conjecture to the case of arbitrary @xmath58 as follows .
[ flow conjecture ] every ts@xmath81 admits a zero - sum @xmath12-flow , except when @xmath14 . in support of the original conjecture
, a computer search has shown that every sts@xmath9 , @xmath88 , admits a zero - sum @xmath12-flow ; see @xcite . in this paper , as further evidence in support of both the original conjecture in @xcite and conjecture [ flow conjecture ] , we show that for every admissible order @xmath87 there exists an sts@xmath9 which admits a zero - sum @xmath12-flow .
further , we show in support of conjecture [ flow conjecture ] that for every admissible @xmath89 , there exists a ts@xmath13 that admits a zero - sum @xmath12-flow , except possibly when @xmath90 and @xmath91 . in this latter case
we show that a ts@xmath13 admitting a zero - sum @xmath22-flow exists .
in the rest of this section we give some general results which apply to @xmath0-designs in general . in the next section
we then show the existence result for triple systems mentioned above . in section [ bound ]
we give an @xmath92 bound for a zero - sum flow on a ts@xmath13 .
finally , in section [ recursive ] , we provide some recursive @xmath18-constructions .
we now develop some general tools for determining zero - sum flows applicable to @xmath0-designs .
we begin with the following result on @xmath69-resolvable designs , recalling that we denote the number of @xmath69-resolution classes by @xmath71 .
[ alpha ] an @xmath69-resolvable @xmath0-@xmath54 design has a zero - sum @xmath12-flow if @xmath93 is odd and a zero - sum @xmath0-flow if @xmath94 is even .
if @xmath94 is even , we colour all of the blocks of each of the @xmath69-resolution classes alternately @xmath95 and @xmath85 .
each pair of oppositely signed classes then generates a zero - sum flow .
if @xmath94 is odd we choose three classes and colour the blocks of two of them with @xmath95 and the blocks of the third @xmath96 .
the remaining classes ( of which there are an even number ) have their blocks coloured with @xmath95 and @xmath85 as above .
taking @xmath97 we obtain the following result .
[ resolvable ] a resolvable design with at least two classes has a zero - sum @xmath12-flow .
clearly there can be no resolvable @xmath98 .
a _ hanani triple system _ of order @xmath99 has @xmath100 almost parallel classes @xmath101 ( each of which contain @xmath102 points ) , and one partial parallel class @xmath103 of size @xmath10 , containing @xmath100 points .
thus , hanani triple systems are in some sense as close to resolvable as one can get when the number of points is @xmath104 .
there exists a hanani triple system of order @xmath99 if and only if @xmath105 @xcite .
we now show that these also admit a zero - sum @xmath12-flow .
let @xmath106 be an integer .
any hanani triple system of order @xmath99 admits a zero - sum @xmath12-flow .
let the blocks of @xmath103 be @xmath107 .
for @xmath108 , let @xmath109 be the point which does not appear in a block of @xmath110 .
then the set of vertices which appear in blocks in @xmath103 is @xmath111 .
first suppose that @xmath10 is even .
for @xmath112 , assign @xmath113 the label @xmath114 .
next we colour the blocks of the remaining parallel classes . for each @xmath108 , if @xmath109 is in a block with label @xmath115 , assign label @xmath115 to the blocks of parallel class @xmath110 .
it is easy to check that the resulting labelling is a zero - sum flow . now , suppose that @xmath10 is odd .
for @xmath116 , assign @xmath114 to block @xmath113 .
assign @xmath85 to @xmath117 and @xmath118 and 2 to @xmath119 .
again , for each @xmath108 , if @xmath109 is in a block with label @xmath115 , assign label @xmath115 to parallel class @xmath110 . given a cyclic @xmath0-@xmath1 design ,
we note that the set of blocks developed from a single starter which generates a full orbit forms a @xmath27-resolution class .
thus , if a cyclic design has no short orbits the design is @xmath27-resolvable .
this observation , in conjunction with theorem [ resolvable ] , gives us the following result .
[ cyclic ] a cyclic @xmath0-@xmath1 design with no short orbits has a zero - sum @xmath12-flow .
lemma ii 2.3 of @xcite states that the incidence matrix of a non - trivial symmetric design is non - singular .
we thus have the following result . [ symmetric ]
a symmetric @xmath0-design has no zero - sum flow .
in this section we show that for every pair @xmath13 such that a ts@xmath13 exists , there is one with a zero - sum @xmath12-flow , except when @xmath14 and except possibly when @xmath15 and @xmath16 .
necessary and sufficient conditions for the existence of a ts@xmath13 were settled by hanani @xcite in 1961 .
[ nec ] the necessary conditions for the existence of a ts@xmath13 are : 1 .
@xmath120 and @xmath121 ; 2 .
@xmath122 and @xmath123 ; 3 .
@xmath124 and @xmath125 ; 4 .
@xmath126 and @xmath106 ; the @xmath13 values in theorem [ nec ] are called _
we first give a non - existence result as a direct corollary of theorem [ symmetric ] .
[ non ] a zero - sum flow does not exist for ts@xmath127 , ts@xmath128 and ts@xmath129 . next we give zero - sum flows for some small parameter values . ' '' '' + [ small ] 1 .
there is no ts@xmath130 admitting a zero - sum @xmath12-flow , but there is one with a zero - sum @xmath20-flow .
there exists a ts@xmath131 with a zero - sum @xmath0-flow 3
. there exists a ts@xmath132 with a zero - sum @xmath12-flow . ' '' '' 1 .
a computer search showed that the unique ts@xmath130 has no zero - sum @xmath12-flow .
a zero - sum @xmath20-flow is shown below .
@xmath133 2 .
we create a ts@xmath131 with point set @xmath134 by developing the following blocks @xmath135 , where @xmath136 is a fixed point .
a zero - sum @xmath0-flow is obtained by applying the given value of @xmath7 to each block developed from that starter block .
@xmath137 3 .
similarly , we create a ts@xmath132 with point set @xmath134 by developing the following blocks @xmath135 , where @xmath136 is a fixed point .
a zero - sum @xmath12-flow is obtained by applying the given value of @xmath7 to each block developed from that starter block .
@xmath138 the existence of resolvable triple systems has been determined ( see @xcite ) , giving us the following result .
[ resolvable ts ] a ts@xmath13 with a zero - sum @xmath12-flow exists whenever @xmath139 , or @xmath140 and @xmath58 is even , except when @xmath141 . a resolvable ts@xmath13 exists if and only if @xmath139 , or @xmath140 and @xmath58 is even , @xmath142 , @xcite .
the result then follows from theorem [ resolvable ] , noting that when @xmath143 , theorem [ resolvable ] does not apply as there is only one resolution class .
cyclic triple systems have also been well studied ; their existence was established in @xcite . in @xcite
the structure of the short orbits is considered .
[ cyclic ts ] there exists a ts@xmath13 which admits a zero - sum @xmath12-flow for the following : 1 .
@xmath144 and @xmath145 ; 2 .
@xmath146 and @xmath147 ; 3 .
@xmath124 and @xmath125 ; 4 .
@xmath148 and @xmath149 ; 5 .
@xmath150 and @xmath151 ; 6 .
@xmath152 and @xmath106 ; for the given values of @xmath10 and @xmath58 , there exists a cyclic ts@xmath13 with no short orbits @xcite
. apply theorem [ cyclic ] to get the result .
[ 0mod6 ] there exists a ts@xmath13 with a zero - sum @xmath12-flow for every @xmath153 .
a ts@xmath132 with a zero - sum @xmath12-flow is given in lemma [ small ] .
for @xmath154 , there exists an idempotent latin square @xmath40 of order @xmath10 with an orthogonal mate @xmath52 .
we take the symbol sets of these squares to be @xmath155 .
we form the design by taking the @xmath156 triples @xmath157 the fact that @xmath40 has an orthogonal mate , @xmath52 , means that @xmath40 can be decomposed into transversals , @xmath53 , @xmath158 , each of which corresponds to the occurences of a given entry in @xmath52 ( see @xcite ) .
we note that by reordering rows and columns and permuting symbols if necessary , we can arrange @xmath40 so that @xmath159 consists of the diagonal entries of @xmath40 .
each @xmath53 , @xmath160 , then corresponds to a collection of blocks in the design . taking the transversals in pairs , @xmath53 , @xmath161 , we label the blocks which come from @xmath53 , @xmath95 , and @xmath161 , @xmath85 ; if @xmath10 is even ( so there is an odd number of transversals ) , we label the blocks from @xmath162 , @xmath163 , @xmath164 with @xmath165 , @xmath85 , @xmath85 respectively .
[ ts result ] there exists a ts@xmath13 with a zero - sum @xmath12-flow for every admissible pair @xmath13 except for @xmath14 and except possibly when @xmath15 and @xmath16 .
the non - existence of any zero - sum flow for a ts@xmath127 , ts@xmath128 and ts@xmath166 are given in lemma [ non ] and the non - existence of a zero - sum @xmath12-flow for a ts@xmath130 is in lemma [ small ] .
we note that for a fixed @xmath10 we may take copies of designs to build up @xmath58 .
in particular , the existence of a ts@xmath167 from lemma [ small ] means that often it suffices to consider only the first value @xmath168 in @xmath58 .
we consider the cases in @xmath169 .
@xmath170 * ( @xmath121 ) * for @xmath171 the result follows from theorem [ cyclic ts ] .
for @xmath172 the result follows from theorem [ resolvable ts ] .
@xmath173 * ( @xmath123 ) * when @xmath174 , @xmath175 , the result follows from theorem [ resolvable ts ] . when @xmath176 this is covered by theorem [ cyclic ts ] , except when @xmath15 and @xmath177 .
we now suppose @xmath15 and consider the cases in @xmath58 .
when @xmath178 , we may combine the blocks of a ts@xmath179 design with a zero - sum @xmath12-flow and a ts@xmath180 with a zero - sum @xmath12-flow . note that @xmath181 , so this design exists by lemma [ 0mod6 ]
when @xmath182 , @xmath183 , we may combine the blocks of a ts@xmath179 with a zero - sum @xmath12-flow and a ts@xmath180 design with a zero - sum @xmath12-flow .
note that @xmath184 , so this is the design above .
@xmath185 * ( @xmath125 ) * this is covered by theorem [ cyclic ts ] .
@xmath153 * ( @xmath106 ) * this is covered by lemma [ 0mod6 ] .
finally we give the following theorem which shows that in the case of @xmath15 and @xmath16 , there is a ts@xmath13 with a zero - sum 5-flow .
[ 2mod6 4 ] there exists a ts@xmath13 with a zero - sum 5-flow for every @xmath186 , @xmath187 and @xmath173 .
there exists a ts@xmath13 with a zero - sum @xmath12-flow for every @xmath188 , @xmath189 and @xmath173 .
we use a modification of the bose construction ; see @xcite .
we construct the design on @xmath190 , where @xmath191 , and denote points by @xmath109 , where @xmath192 and @xmath193 . for @xmath194
there exists an idempotent latin square @xmath40 of side @xmath3 with an orthogonal mate @xmath52 @xcite . for each pair
@xmath195 , @xmath196 , we form the design by taking the triples @xmath197 .
we label the blocks which come from the transversals of @xmath40 , @xmath53 , @xmath198 , with @xmath95 , @xmath85 in pairs ; if @xmath10 is odd , we label the blocks from @xmath162 , @xmath163 , @xmath164 with @xmath199 respectively .
this creates a partial design with a zero - sum @xmath12-flow on the points of @xmath200 . for each vertical , @xmath201 , @xmath202
, we place a ts@xmath128 on @xmath203 with the flow values as indicated below .
@xmath204 this gives weight 0 to each @xmath205 and weight @xmath206 to @xmath136 .
if @xmath207 is even we label the designs on the verticals with @xmath208 in pairs , to get a zero - sum @xmath12-flow .
if @xmath3 is odd , we label three designs on the verticals with @xmath209 ( which gives a block of weight 4 ) and the rest with @xmath208 in pairs to get a zero - sum 5-flow ( @xmath210 ) .
in this section we establish an @xmath92 bound on the size of a zero - sum flow on a ts@xmath13 when @xmath211 , @xmath212 .
further , this zero - sum @xmath92-flow takes at most five distinct values . for @xmath211 , @xmath212 ,
every ts@xmath81 admits a zero - sum flow whose entries are in the set @xmath213 consider a ts@xmath81 with incidence matrix @xmath214 .
we know that the number of appearances of each element in blocks is @xmath215 .
we show that the last column of @xmath214 is a linear combination of the other columns with coefficients from the set @xmath216 . by a suitable ordering of the elements of @xmath217
, we can assume that the last column of @xmath214 is @xmath218^t$ ] .
now , we remove the last column of @xmath214 and call the remaining matrix @xmath219 .
we then have the following equality : @xmath220,\ ] ] where @xmath221 is the @xmath222 all-1 matrix . for simplicity
we denote @xmath223 by @xmath224 .
we now show that the system of linear equations @xmath225 has integral solutions such that each component of @xmath43 is in the set @xmath226 .
assume that @xmath227^t$ ] , where @xmath228 and @xmath59 are unknown variables and there are @xmath229 @xmath59s .
we may solve the equations @xmath230 to get @xmath231 , @xmath232 ( @xmath211 ) and @xmath233 now , since @xmath234 and each row of @xmath235 has exactly three @xmath34s , we may conclude that there exists a vector , @xmath38 , in the nullspace of @xmath214 whose components are all of the form @xmath236 , @xmath237 , @xmath238 , @xmath239 or @xmath240 .
finally , we note that when @xmath211 , @xmath212 , all of the components of @xmath38 are non - zero integers which are divisible by 3 .
now , @xmath241 is also in the nullspace of @xmath214 and all of its components are in the set @xmath37 .
in this section we show that the standard @xmath18-construction for steiner triple systems can be adapted to respect zero - sum flows in many cases .
while we work in the case @xmath242 , the generalisation to higher @xmath58 is clear .
we say that a graph @xmath24 has a _
@xmath27-null 1-factorisation _ if @xmath24 has a zero - sum @xmath27-flow and there is a @xmath34-factorisation in which the weight of each @xmath34-factor is zero .
we call each 1-factor of a @xmath27-null 1-factorisation a _ @xmath27-null 1-factor_. [ 2v+1 ] if @xmath243 has a @xmath27-null 1-factorisation and there exists an sts@xmath9 , @xmath4 , with a zero - sum @xmath244-flow , then @xmath4 can be embedded into an sts@xmath18 with a zero - sum @xmath245-flow .
let @xmath72 be an sts@xmath9 , with @xmath246 , and let @xmath43 be a set of order @xmath247 .
we will construct the new design on @xmath248 . construct a @xmath27-null 1-factorisation on @xmath243 with point set @xmath43 .
let @xmath249 be the 1-factors in this @xmath34-factorisation , and suppose that @xmath250 for @xmath251 .
form the triples @xmath252 it is easy to see that @xmath253 is an sts@xmath18 . in order to obtain a zero - sum @xmath245-flow
, we retain the original weighting on @xmath57 , and for each triple @xmath254 , we give it the weight of the edge @xmath255 in @xmath256 .
it is not hard to see that we obtain a zero - sum @xmath245-flow on the sts@xmath18 , @xmath257 .
first assume that @xmath260 and let @xmath41 and @xmath42 be two orthogonal latin squares of order @xmath3 .
let @xmath261 and @xmath262 be the two parts of the complete bipartite graph @xmath258 .
we use @xmath41 to define a colouring , @xmath263 , on the edges of @xmath258 , and @xmath42 to factor @xmath258 into @xmath27-null 1-factors as follows . for each edge
@xmath264 , @xmath265 , we give @xmath266 colour @xmath267 .
we now define a 1-factorisation of @xmath258 by , @xmath268 .
note that since @xmath41 and @xmath42 are orthogonal , each @xmath256 contains exactly one edge of each colour .
if @xmath3 is even , then we assign @xmath34 to every edge whose colour is in the set @xmath269 and assign @xmath85 to the remaining edges . if @xmath3 is odd , then we assign @xmath270 to all edges with colours , @xmath271 and @xmath12 , respectively and assign @xmath272 to all edges with colours @xmath273 , alternately .
it is not hard to see that this assignment is the desired zero - sum @xmath12-flow for @xmath258 .
let @xmath280 , @xmath281 , and consider the complete graph @xmath277 as the join of two complete graphs @xmath282 and @xmath282 . by lemma [ bipart ] , @xmath283 has a @xmath12-null 1-factorisation .
let @xmath284 and @xmath285 be two @xmath34-factorisations for the first and the second @xmath282 , respectively
. then @xmath286 forms a @xmath34-factorisation for the disjoint union of the two @xmath282 .
now assign @xmath0 to all edges of @xmath287 , @xmath96 to all edges of @xmath288 , @xmath85 to all edges of @xmath289 and @xmath34 to all edges of @xmath290 . for each @xmath50 , @xmath291 assign @xmath85 and @xmath34 to all edges of @xmath292 , alternately .
for each @xmath50 , @xmath291 assign @xmath34 and @xmath85 to @xmath293 , alternately .
a similar proof to that of lemma [ 4 ] shows that if @xmath299 then @xmath35 has a zero - sum @xmath0-flow with a @xmath34-factorisation in which the weight of each @xmath34-factor is zero . by lemma [ 2v+1 ] , this shows that when @xmath300 every @xmath296 with a zero - sum @xmath0-flow can be embedded into an @xmath297 with a zero - sum @xmath0-flow .
let the point set of @xmath301 be @xmath303 and let @xmath304 be the @xmath34-factorisation of @xmath301 defined by @xmath305 , @xmath306 , where @xmath307 .
we claim that @xmath301 can be partitioned into @xmath308 cubic graphs each isomorphic to @xmath309 where the number of @xmath276 is @xmath27 .
it is not hard to see that @xmath310 is a disjoint union of a @xmath311 with vertex set @xmath312 and @xmath27 copies of @xmath276 with the vertex sets @xmath313 and partite sets @xmath314 clearly , @xmath315 , for @xmath316 and so it can also be decomposed as required
. given two graphs @xmath24 and @xmath319 , the _ cartesian product _ of @xmath24 and @xmath319 , @xmath320 , is the graph with vertex set @xmath321 , and @xmath322 if and only if either @xmath323 and @xmath324 , or @xmath325 and @xmath326 .
we can deal with the case @xmath327 allowing for a @xmath22-flow
. given two graphs @xmath24 and @xmath319 , the _ wreath product _ of @xmath24 and @xmath319 , @xmath330 , is the graph with vertex set @xmath321 , and @xmath331 if and only if @xmath326 , or @xmath323 and @xmath324 .
@xmath0-factor _ is a collection of cycles that spans all vertices of the graph .
@xmath0-factorisation _ of a graph @xmath24 is an edge decomposition of @xmath24 into @xmath0-factors .
the problem of finding an @xmath332-factorisation of @xmath35 , where @xmath332 is a given @xmath0-factor of order @xmath3 , is the well known oberwolfach problem .
clearly a @xmath0-factorisation of @xmath35 can not exist when @xmath3 is even ; in this case it is common practice to consider factorisations of @xmath333 , the complete graph with the edges of a 1-factor @xmath334 removed .
it is known that the oberwolfach problem has no solution when @xmath335 .
otherwise , a solution is known for every case where @xmath336 @xcite , and if every component of the factor is isomorphic @xcite .
the case where @xmath332 is bipartite is nearly solved @xcite ; the case where @xmath332 consists of exactly two parts is solved @xcite .
rotational solutions have been studied @xcite and many other families are known @xcite , but no general solution is known .
see ( * ? ? ?
* section vi.12 ) for a survey .
hggkvist proved the following very useful result in @xcite .
[ hag ] for any @xmath337 and for each bipartite @xmath0-regular graph @xmath332 of order @xmath338 , there exists a @xmath0-factorisation of @xmath339 , in which each @xmath0-factor is isomorphic to @xmath332 . by theorem [ wr ]
, @xmath341 can be decomposed into @xmath342 , graphs isomorphic to @xmath343 , and one @xmath344 .
now , let @xmath332 be a @xmath0-regular graph which is disjoint union of one @xmath345 and @xmath342 , @xmath346s
. by lemma [ hag ] , @xmath343 , has a @xmath0-factorisation in which each @xmath0-factor is isomorphic to @xmath332 .
assign @xmath96 and @xmath0 to the edges of @xmath345 , alternately .
if the number of @xmath346 in @xmath332 is even , then assign @xmath96 and @xmath0 to the edges of one @xmath346 , alternately and assign @xmath85 and @xmath34 to the edges of each other @xmath346 , alternately .
it is not hard to see that @xmath332 has a @xmath12-null 1-factorisation .
if the number of @xmath346 is odd , then assign @xmath348 and @xmath12 to the edges of one @xmath346 , alternately and assign @xmath85 and @xmath34 to the edges of each other @xmath346 , alternately .
again , it is not hard to see that @xmath332 has a @xmath12-null 1-factorisation .
let @xmath353 be the cubic graph shown in figure [ h_i ] , where there are exactly @xmath50 blocks of @xmath311 minus one edge .
now , consider the @xmath22-regular graph @xmath344 , figure [ decomp ] shows that @xmath344 can be decomposed into a copy of @xmath332 and the cubic graph @xmath354 .
figures [ 5-flow h_1 ] , [ 5-flow h_3 ] and [ 5-flow h_5 ] show a @xmath22-null 1-facorisation of @xmath350 , @xmath351 and @xmath352 , respectively .
figure [ continue ] gives a graph which can be appended to @xmath350 , @xmath351 or @xmath352 to get a zero - sum @xmath22-flow with a @xmath34-factor in which the weight of each @xmath34-factor is zero for any @xmath353 , except @xmath355 .
now , for the case @xmath355 , so @xmath356 , let @xmath40 be the bipartite @xmath0-factor which is the disjoint union of a @xmath345 and a @xmath357 .
it is not hard to see that @xmath358 can be decomposed into @xmath40 and the cubic graph given in figure [ k=3 ] , which has a zero - sum @xmath22-flow with a desired @xmath34-factor .
now , alternately give weight @xmath12 and @xmath348 to all edges of @xmath357 and @xmath20 and @xmath360 to all edges of @xmath345 in @xmath40 , alternately .
this implies that @xmath361 has a zero - sum @xmath22-flow with a @xmath34-factorisation in which the weight of each @xmath34-factor is zero .
we note that the zero - sum flow on the blocks of the embedded sts@xmath9 remains unchanged . + * acknowledgments * + this work was done while s.akbari was visiting the university of toronto and supported by nserc discovery grant 455994 . also , the research of the first author was in part supported by a grant from ipm ( no .
93050212 ) .
the second and third authors are both supported by the nserc discovery grant program . | given a @xmath0-@xmath1 design , @xmath2 , a _
zero - sum @xmath3-flow _ of @xmath4 is a map @xmath5 such that for any point @xmath6 , the sum of @xmath7 around all the blocks incident with @xmath8 is zero
. it has been conjectured that every steiner triple system , sts@xmath9 , on @xmath10 points @xmath11 admits a zero - sum @xmath12-flow .
we show that for every pair @xmath13 , for which a triple system , ts@xmath13 exists , there exists one which has a zero - sum @xmath12-flow , except when @xmath14 and except possibly when @xmath15 and @xmath16 .
we also give a @xmath17 bound on @xmath3 and a recursive result which shows that every sts@xmath9 with a zero - sum @xmath12-flow can be embedded in an sts@xmath18 with a zero - sum @xmath12-flow if @xmath19 , a zero - sum @xmath20-flow if @xmath21 and with a zero - sum @xmath22-flow if @xmath23 .
keywords : zero - sum flow , steiner triple system , steiner system , chromatic index of steiner systems . + ams 2010 subject classification : 05b05 , 05b20 , 05c15 , 05c21 . |
single molecule magnets ( smm s ) , defined as compounds where a magnetic domain can , in principle , be reduced to a single molecule @xcite , have recently been of high theoretical and experimental interest due to their novel properties and potential applications .
these properties and potential applications include quantum tunneling of their magnetization ( qtm ) @xcite , whose detailed mechanism is still not fully understood , molecular memory devices @xcite and elements of quantum computers @xcite .
one of the best characterized smm s is [ ( c@xmath9h@xmath10n@xmath11)@xmath9fe@xmath4(@xmath12-o)@xmath13(@xmath14-oh)@xmath15br@xmath16(h@xmath13o)br@xmath178h@xmath13o , abbreviated fe@xmath18br@xmath18 @xcite , whose main spin - bearing skeleton is shown in fig .
[ struc ] br@xmath4 @xcite .
the arrows represent spin orientations of the fe@xmath19 ions in the @xmath5 ground state .
the organic ligands and the eight br@xmath20 anions have been omitted for clarity .
each fe@xmath19 has @xmath21 , thus the ground state can be seen as @xmath22 . ]
studies by magnetization @xcite , neutron scattering @xcite , electron paramagnetic resonance ( epr ) @xcite , and nmr @xcite techniques have established that the ground state has a spin value @xmath5 .
figure [ struc ] shows a schematic of the spin configuration of the fe@xmath19 ( @xmath23 ) ions . at temperatures below 1 k
, the magnetization relaxation takes places via qtm @xcite .
while a great deal of progress has been made in understanding the nature of qtm in fe@xmath18br@xmath18 , many questions still remain unclear .
for example , the magnitudes of calculated tunneling rates are much lower than the observed values @xcite .
secondly , there is a lack of data on the nature and magnitude of the spin - exchange constants , j s , between the eight fe@xmath19 ions in the fe@xmath4br@xmath4 core .
the best estimates come from the temperature dependence of the _ dc _ magnetic susceptibility , @xmath24@xmath25 @xcite .
the @xmath24@xmath25 @xcite could be fitted by several sets of j s .
one of the criteria of such a procedure is the prediction of the proper spin @xmath26 value of the ground state , together with the location of the excited states , in particular the @xmath3 state .
the @xmath27 fit yielded at least two sets of j s , but the set providing the better fit yielded the position of the @xmath3 state to be less than 0.5 @xmath28 above the ground state @xcite .
the other set predicted the @xmath3 state at greater than 25 @xmath28 ( 36 k ) above the ground state .
clearly , from the point of view of understanding the magnetic structure of fe@xmath4br@xmath4 , it is important to experimentally determine the location of the excited states , in particular the @xmath3 manifold .
a peak around 30 @xmath28 has been reported in the far infrared spectrum @xcite , but it was not established whether it was a normal molecular vibration involving metal ions , or a transition to the @xmath3 level from the @xmath5 ground state .
moreover this peak was not observed for the closely similar compound fe@xmath4br@xmath29(clo@xmath29)@xmath29 , whose spin hamiltonian ( eq . [ ham ] ) , parameters are known to be close to those of fe@xmath18br@xmath18 @xcite .
the magnetic properties of the ground state of fe@xmath4br@xmath4 have been well described by the following spin hamiltonian , with @xmath30 @xcite : @xmath31 + where the first term is the zeeman interaction , @xmath32 represents the usual zero - field , uniaxial anisotropy parameter , and @xmath33 the second order rhombic anisotropy .
the fourth order terms are given by : @xmath34 = 35 @xmath35 - [ 30 @xmath26(@xmath26 + 1)- 25]@xmath36 - 6@xmath26(@xmath26 + 1 ) + 3 @xmath37(@xmath26 + 1)@xmath38 , @xmath39 = 1/4[7@xmath36 - @xmath26(@xmath26 + 1 ) - 5](@xmath40 + @xmath41)+ ( @xmath40 + @xmath41)[7@xmath36 - @xmath26(@xmath26 +
1 ) - 5 ] , and @xmath42 = 1/2(@xmath43 + @xmath44 ) . here
, we report on an epr detection of the @xmath3 state of fe@xmath18br@xmath18 , and preliminary evaluation of its spin hamiltonian parameters as defined in eq .
( [ ham ] ) .
we observe a series of peaks , called @xmath45 transitions , which we assign to an @xmath3 excited state of fe@xmath4br@xmath4 . from the temperature variation of the line intensities , we establish that the @xmath3 manifold lies at 24 @xmath2 2 k above the @xmath5 ground state , in contrast to the suggestion of @xmath46 36 k from susceptibility analysis @xcite .
section ii below describes the single crystal preparation and the epr instrumentation .
the results obtained , and their analysis , are presented in section iii , with the discussion presented in section iv , and the main conclusions presented in section v.
[ ( c@xmath9h@xmath10n@xmath11)@xmath9fe@xmath4(@xmath12-o)@xmath13(@xmath14-oh)@xmath15br@xmath16(h@xmath13o)br@xmath178h@xmath13o ( fe@xmath18br@xmath18 ) was synthesized following the procedure of weighardt @xcite . relatively large ( 2 mm x 2 mm x 0.5 mm ) ,
optical quality single crystals were prepared by slow evaporation .
the crystals were aligned with the zeeman field applied along the easy axis of magnetization by sight , to within a few degrees @xcite .
the final orientation was confirmed by epr splittings , being the extremum for the canonical orientations .
the epr measurements were made using a variable frequency ( 44 - 200 ghz ) , cavity based , high sensitivity spectrometer described earlier @xcite .
the main component of the spectrometer is a millimeterwave vector network analyzer ( mvna ) , a phase - sensitive , fully sweepable , superheterodyne source - detection system .
a variable flow cryostat situated within the bore of an 17 t superconducting solenoid allows for temperatures down to 1.5 k , with an accuracy of @xmath2 0.01 k. the high sensitivity of the mvna technique ( 10@xmath47 spins g@xmath48 s@xmath48 ) allows for observation of the low level transitions of the fe@xmath4br@xmath4 ground and excited states in a single crystal , and for angular variation studies , as described earlier @xcite .
br@xmath4 at 131 ghz and 35 k with the zeeman field applied along the easy axis ( bottom panel ) , along with the energy level diagram corresponding to the @xmath5 spin system ( top panel ) .
the @xmath5 energy level diagram has been constructed through the spin hamiltonian parameters of caciuffo _ et al .
the single - headed arrow at 0.8 tesla designates the expected position where a @xmath45@xmath49 would appear if the @xmath45 transitions originated from an @xmath5 state . ]
figure [ exp ] ( bottom panel ) shows a typical epr spectrum of fe@xmath4br@xmath4 at 131 ghz with the zeeman field applied along the easy axis of an fe@xmath4br@xmath4 single crystal at 35 k. the spectrum consists of a series of strong peaks @xmath50 , @xmath51 , @xmath52 , etc ; the subscripts represent the spin projection quantum number , m@xmath1 , corresponding to the level from which the epr absorption transition originates , in the @xmath30 ground state following the convention introduced earlier @xcite .
the @xmath53 transitions have been very well analyzed earlier @xcite , and have been shown to arise from the magnetic dipole ( @xmath0m@xmath1 = @xmath2 1 ) transitions within the 21 m@xmath1 levels of the @xmath5 multiplet .
the specific peak assignment is indicated in the top panel for the @xmath53 transitions .
in addition to the @xmath53 resonances , there are additional peaks present in fig . [ exp ] that are labelled as the @xmath45 transitions and are the focus of the current investigation . our analysis procedure consisted of three steps : ( a ) to ascertain that the @xmath45 transitions are from an excited state , ( b ) to determine the spin multiplicity of this excited state , and ( c ) to deduce the spin hamiltonian parameters for the @xmath45 spin system and its energy position relative to the ground state ( @xmath5 ) .
direct evidence that the @xmath45 transitions originate from a thermally populated excited state is provided by the temperature dependence of their intensities .
figure [ extemp ] shows spectra at 5 , 15 , and 35 k , respectively , giving clear experimental evidence that the @xmath45 transitions arise from a thermally populated excited state , because their intensities rapidly decrease as the temperature is lowered . in order to quantitatively
measure the intensities of the @xmath45 peaks as a function of temperature , the spectra of the ground and excited states needed to be separated .
the separation was accomplished by using a gaussian fit for each individual peak .
the validity of gaussian fits , especially for the low field transitions , has been previously established @xcite .
the criterion for the goodness of the fit was that the sum of the @xmath53 and @xmath45 transitions mirrors the experimental data quite well by visual inspection and also by minimizing the remaining intensity obtained by subtracting the gaussian fits from the experimental spectra .
this remaining intensity was within the experimental noise in our separation procedure .
each gaussian fit was then summed to yield a separated @xmath5 spectrum .
the spectra were then separated by subtracting the gaussian fits of the @xmath53 peaks from the experimental spectra .
these gaussian fits were then summed , giving a separated @xmath5 spectrum .
the remaining @xmath45 peaks were then fit with gaussian functions , and summed , yielding a separated @xmath3 spectrum .
we verified that these separated spectra agreed well with the experimental data by taking the sum of the @xmath53 and @xmath45 spectra and comparing it with experiment , as shown in fig .
[ tempdep](a ) .
figure [ tempdep ] also shows the relative decrease in the intensity of the separated excited state ( @xmath45 ) spectra in relation to the @xmath5 spectra as the temperature is decreased . the spectral envelope present in the @xmath5 spectra at 5 , 15 , and 35
k is generally evident in the corresponding excited state spectra , though the relative intensities are modified slightly due to differing matrix elements in the transition probabilities , p@xmath54 @xmath55 @xmath56 , of the two spin systems , given by eq .
( [ element ] ) : @xmath57 . \label{element}\ ] ] br@xmath4 , for b @xmath58 z at 131 ghz and 5 , 15 , and 35 k. ] and @xmath45 , separated @xmath3 , and separated @xmath5 spectra at 35 k and 131 ghz for b @xmath58 z. ( b ) separated spectra at 15 k , ( c ) and at 5 k. ] in order to analyze the intensity of the @xmath45 peaks we normalized their intensities .
the normalization process involved dividing the intensity of a @xmath45 peak , with a given m@xmath1 value , by the corresponding @xmath53 peak with the same m@xmath1 value .
this enabled us to ignore instrumental effects and spectral envelope changes with temperature . at 5
k , the @xmath45 peaks are barely discernable from the noise level of the spectrum . at 10 k , however , they increase in intensity enough to emerge from the @xmath53 transitions .
the decreasing intensity of the @xmath45 peaks , upon lowering temperature , unambiguously designate the @xmath45 peaks as being due to a thermally populated excited state .
once it was concluded that the @xmath45 peaks originated from an excited state , the spin multiplicity of the excited state needed to be determined .
the presence of an excited state ( with perhaps @xmath3 ) close to the ground state of fe@xmath4br@xmath4 has been previously inferred from magnetic susceptibility @xcite , and mentioned in subsequent @xmath59sr @xcite , epr @xcite , and neutron diffraction studies @xcite , without any evidence for its location or multiplicity .
herein , the spin multiplicity of the excited state has been determined by two independent methods .
first , by the location of the leading peak ( @xmath60 ) in the set of peaks assigned to the @xmath3 state .
figure [ exp ] shows an experimental spectrum taken at 131 ghz and 35 k. as is evident in all spectra taken , there is no @xmath45 peak between the @xmath50 and @xmath51 peaks .
furthermore , the @xmath50 transition is symmetric and shows a clear gaussian shape , as would be expected at this temperature for a well separated , individual peak @xcite .
this lack of a @xmath61 transition , barring spin hamiltonian parameters being very different from those for the ground state , is strong evidence that the excited state is @xmath3 .
the location of the first @xmath45 peak is consistent with slightly modified spin hamiltonian parameters and a spin multiplicity of @xmath3 for the excited state , as anticipated theoretically @xcite .
.experimental and simulated peak positions at 131 ghz for the @xmath3 state , for b @xmath58 z , using d = -0.27 k , e = @xmath20.05 ( @xmath20.015 ) k , b@xmath29@xmath62 = -1.3 @xmath7 10@xmath8 k , b@xmath29@xmath38 = 0 k , b@xmath29@xmath63 = 0 k [ cols= " < , < , < " , ] br@xmath4 , for b @xmath58 z at 35 k and 131 ghz . ] additional , more quantitative support , that the spin @xmath26 of this excited state is indeed @xmath3 was provided by computer simulations which were run using sim @xcite . the procedure was first checked for the @xmath5 state for which the parameters are known @xcite .
the simulations for the @xmath30 state were performed with the spin hamiltonian parameters previously determined by caciuffo _
_ @xcite using neutron scattering .
a typical comparison is shown in fig .
the simulated spectra for the @xmath64 ground state were in close agreement with our experimental results , thereby validating the simulation procedure .
transitions in fe@xmath4br@xmath4 with b @xmath58 z. the solid lines portray the frequency dependence given by the spin hamiltonian parameters determined in this work , whereas the solid squares are the observed field positions . ]
state , i@xmath65 , normalized to the -8 to -7 transition in the @xmath5 state , i@xmath49 .
the curve joining the experimental points is a guide to the eye .
( b ) boltzmann analysis of the normalized intensities of the @xmath3 spin state .
the slope yields the excitation energy as 18 @xmath2 2 k between level with the same m@xmath1 values in the @xmath5 and @xmath3 manifolds .
addition of the zero field splitting between the m@xmath66 and m@xmath67 levels ( 5.5 k ) leads to the location of the @xmath68 state at 24 @xmath2 2 k above the ground state . ]
the @xmath45 transitions were then accurately simulated for the three frequencies utilized ( 110 , 131 , and 155 ghz ) , using the spin hamiltonian parameters obtained in the present study . a field - frequency plot
is shown in fig .
quite good agreement can be seen between the simulated curves ( continuous lines ) and the observed peak positions .
a more explicit and quantitative comparison is shown in table i , using the measured data for 131 ghz .
the d parameter decreases in magnitude by 8% , to -0.27 k , while the e term increases by 7% , to @xmath2 -0.05 k ( @xmath2
0.015 k ) .
surprisingly , the b@xmath6 term is similar in magnitude , -1.3 @xmath7 10@xmath8 k , but opposite in sign to the @xmath5 parameter ( 1.01 @xmath7 10@xmath8 k ) . due to the fact that the experimental spectra were taken along the easy axis of magnetization
, the b@xmath69 and b@xmath70 terms were not included in the simulations . a simulated spectrum at 35 k , with the spin hamiltonian parameters determined for both the @xmath5 and @xmath71 states is shown in fig .
linewidths were not an optimized parameter in the simulations , though the relative intensities matched the experimental data well .
therefore , the simulated spectrum presented in this work is the sum of the @xmath30 and @xmath3 separated spectra using peak positions generated by sim @xcite .
the agreement between the experimental and simulated spectra , as shown in fig .
[ sim ] , can be seen to be quite satisfactory , thereby supporting the parameter assignment .
we are thus able to assign the full experimental spectrum to transitions in the @xmath5 multiplet ( @xmath53 s ) and the @xmath3 multiplet ( @xmath45 s ) , as shown in figure [ exp ] .
once the spin hamiltonian parameters had been determined , the relative intensities of the @xmath53 and @xmath45 transitions were used , at temperatures from 5 to 35 k , to determine the location of the @xmath3 state above the ground state .
the intensity of a specific peak between given m@xmath1 and m@xmath72 states , is proportional to the population difference between the m@xmath1 and m@xmath72 states and the transition probability , @xmath73 , as given in eq .
( [ popdiff ] ) : @xmath74 therefore , assuming very similar partition functions and d values for the two states , the intensity ratio between two transitions of the same m@xmath1 states in the @xmath3 and @xmath5 manifolds , respectively , is given by eq .
( [ ratio ] ) : @xmath75 where k is the boltzmann constant and @xmath0e@xmath76 is the energy difference between given m@xmath1 states in the @xmath5 and @xmath3 manifolds .
the areas of the gaussian fits , for a given m@xmath1 to m@xmath72 transition , were factored by their transition probabilities , in both the @xmath3 and @xmath5 manifolds , in order to determine their ratios . due to the fact that each @xmath45 transition is normalized to its corresponding @xmath53 transition
, the ratio of intensities should be constant , regardless of the specific m@xmath1 pair , for any given temperature .
a boltzmann analysis of the intensity ratios is shown in fig .
[ arrhenius](b ) .
the ratios of intensities of the @xmath45 to @xmath53 transitions were compared at 5 , 15 , 20 , 25 , 30 and 35 k , and plotted vs. inverse temperature ( t@xmath77 ) .
the slope yields the energy difference ( 18 k ) between a given m@xmath1(9 ) state and the corresponding m@xmath1(10 ) state .
therefore , the energy difference between the m@xmath1 = 10 and m@xmath1 = 9 in zero field ( 5.46 k ) must be added to the energy obtained from the boltzmann analysis , yielding an energy difference @xmath0 of 24 @xmath2 2 k ( 16.7 @xmath2 1.5 @xmath28 ) .
figure [ 10 - 9eld ] shows a schematic of the energy levels in zero field for the @xmath5 and @xmath3 manifolds based on the present study
. clearly , the higher m@xmath1 levels of the @xmath3 state overlap with the lower m@xmath1 levels of the @xmath5 state , indicating at least a partial breakdown of the single spin model .
our determination that the @xmath3 manifold in fe@xmath4br@xmath4 is located 24 @xmath2 2 k above the @xmath5 ground state is in contrast to with earlier suggestions based on magnetic susceptibility ( @xmath46 36 k ) @xcite .
the four central fe@xmath19 ions in the fe@xmath4br@xmath4 core ( fe@xmath78@xmath19 , fe@xmath13@xmath19 , fe@xmath11@xmath19 , and fe@xmath29@xmath19 ( as shown in fig .
1 ) ) can be described as that of a butterfly configuration .
all magnetic coupling interactions , j , between fe@xmath19 ions in fe@xmath4br@xmath4 have been determined to be antiferromagnetic @xcite .
therefore , the magnitude of these coupling constants dictates both the location and spin topology of the @xmath3 excited state . as mentioned in the introduction , a detailed analysis of the _ dc _ magnetic susceptibility data led delfs _
@xcite to two reasonable sets of exchange parameters ; ( a ) j@xmath79 = 20 @xmath28 , j@xmath80 = 120 @xmath28 , j@xmath81 = 15 @xmath28 , and j@xmath82 = 35 @xmath28 , and ( b ) j@xmath79 = 102 @xmath28 , j@xmath80 = 120 @xmath28 , j@xmath81 = 15 @xmath28 , and j@xmath82 = 35 @xmath28 .
while set ( b ) provided a much better fit to the experimental data , it predicted the position of the first excited state , @xmath3 , at less than 0.5 @xmath28 above the @xmath5 ground state .
we do note , however , that delfs @xmath83 did not include any zero field splitting terms in their susceptibility analysis .
nevertheless , this same basic configuration of exchange constants has been recently supported by detailed symmetry - based calculations by raghu @xmath83 @xcite .
though the magnitudes of the coupling constants calculated by these authors @xcite are different from those of delfs @xcite , the dominance of the j@xmath80 interaction over other magnetic couplings remains consistent .
the present study supports the essential correctness of set ( a ) .
the coupling set ( a ) of delfs _ et al .
_ @xcite and the best set of raghu _ et al . _
@xcite , both show that j@xmath80 dominates the exchange interactions .
the perturbation leading to the @xmath3 excited state must result from the smallest difference in j s acting on the same ion or symmetrically equivalent set of fe@xmath19 ions .
thus it seems reasonable to deduce that this perturbation does not involve j@xmath80 , hence the butterfly core ( fe@xmath78 , fe@xmath13 , fe@xmath11 , fe@xmath29 ) , but rather the fe s on the corners of the cluster ( fe@xmath84 , fe@xmath9 , fe@xmath16 , fe@xmath4 ) via some linear combination of their wavefunctions . and the @xmath3 states in zero magnetic field .
@xmath3 is located at an energy @xmath0 = 24 @xmath2 2 k as marked . ]
the spin hamiltonian parameters determined for the @xmath3 state provide some insight to the origin of the anisotropy of the cluster .
the @xmath3 parameters are slightly different from those of the @xmath5 ground state .
the 7@xmath85 larger e value for @xmath3 is in accord with increased transverse distortion in the fe@xmath4 structure .
the decrease in d with decreasing magnetic moment indicates that the anisotropy present in the fe@xmath4br@xmath4 core has some dipolar contribution rather than arising purely from a spin - orbit interaction .
similarly , the change in sign of the b@xmath6 term indicates that b@xmath6 originates from many - body interactions between fe@xmath19 ions , and not from a collective sum of individual b@xmath6 terms .
this argument is in line with the fact that a b@xmath29 term needs an effective interaction involving at least four spins .
alternatively , this significant change in b@xmath6 may be a result of the breakdown of the single spin model , as has been proposed by katsnelson _
_ in connection with mn@xmath86-acetate @xcite .
additional , detailed angular variation studies are underway for precise measurement and understanding of these questions . the anisotropy barrier , estimated from d and e values for the @xmath71 manifold is 22 k , as compared to that for the @xmath5 ground state ( 29 k ) .
using variable frequency , high - field , epr measurements on single crystals of fe@xmath18br@xmath18 , we have detected a set of transitions , labelled as @xmath87 , which have been conclusively assigned to the @xmath3 spin multiplet , located at an energy @xmath0 = 24 @xmath2 2 k ( 17 @xmath88 @xmath28 ) above the ground ( @xmath5 ) state .
the spin hamiltonian parameters have been determined to a good accuracy , and differ from those of the @xmath30 state : d is smaller by 8% , while e is larger by 7% .
these parameters yield the anisotropy barrier ( @xmath89 ds@xmath90@xmath38 @xmath89 22 k ) , about 25% smaller than for @xmath5 .
b@xmath29@xmath62 for @xmath3 also shows a dramatic change , the sign is opposite to that for the @xmath5 state . although electronic structure calculations have been reported for fe@xmath18br@xmath18 @xcite , there has been little definitive data on excited states .
the results of the present study should serve as sensitive basis for more refined theoretical modelling of the bonding and magnetic properties of these materials .
we would like to thank the national science foundation ( nirt grant no .
dmr 0103290 ) for financial support . | high frequency electron paramagnetic resonance has been used to observe the magnetic dipole , @xmath0 m@xmath1 = @xmath2 1 , transitions in the @xmath3 excited state of the single molecule magnet fe@xmath4br@xmath4 .
a boltzmann analysis of the measured intensities locates it at 24 @xmath2 2 k above the @xmath5 ground state , while the line positions yield its magnetic parameters d = -0.27 k , e = @xmath20.05 k , and b@xmath6 = -1.3@xmath7 10@xmath8 k. d is thus smaller by 8% and e larger by 7% than for @xmath5 . the anisotropy barrier for @xmath3 is estimated as 22 k , which is 25% smaller than that for @xmath5 ( 29 k ) .
these data also help assign the spin exchange constants(j s ) and thus provide a basis for improved electronic structure calculations of fe@xmath4br@xmath4 .
[ 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 |
one of the famous conjectures in graph theory is erds - faber - lovsz conjecture .
it states that if @xmath0 is a linear hypergraph consisting of @xmath1 edges of cardinality @xmath1 , then it is possible to color the vertices of @xmath0 with @xmath1 colors so that no two vertices with the same color are in the same edge @xcite .
erds , in 1975 , offered 50 usd @xcite and in 1981 , offered 500 usd @xcite for the proof or disproof of the conjecture .
kahn @xcite showed that the chromatic number of @xmath0 is at most @xmath15 .
jakson et al .
@xcite proved that the conjecture is true when the partial hypergraph @xmath16 of @xmath0 determined by the edges of size at least three can be @xmath17-edge - colored and satisfies @xmath18 .
in particular , the conjecture holds when @xmath16 is unimodular and @xmath18 .
sanhez - arrayo @xcite proved the conjecture for dense hypergraphs .
faber @xcite proves that for fixed degree , there can be only finitely many counterexamples to efl on this class ( both regular and uniform ) of hypergraphs .
if @xmath0 is a linear hypergraph consisting of @xmath1 edges of cardinality @xmath1 , then it is possible to color the vertices of @xmath0 with @xmath1 colors so that no two vertices with the same color are in the same edge .
we consider the equivalent version of the conjecture for graphs given by deza , erds and frankl in 1978 @xcite .
[ efl ] let @xmath2 denote a graph with @xmath1 complete graphs @xmath19 @xmath20 , each having exactly @xmath1 vertices and have the property that every pair of complete graphs has at most one common vertex , then the chromatic number of @xmath5 is @xmath1 .
[ d1 ] let @xmath2 denote a graph with @xmath1 complete graphs @xmath3 @xmath4 , each having exactly @xmath1 vertices and the property that every pair of complete graphs has at most one common vertex .
the clique degree @xmath21 of a vertex @xmath7 in @xmath5 is given by @xmath8 .
the maximum clique degree @xmath22 of the graph @xmath5 is given by @xmath23 . from the above definition
, one can observe that degree of a vertex in hypergraph is same as the clique degree of a vertex in a graph .
let @xmath5 be the graph satisfying the hypothesis of conjecture [ efl ] .
let @xmath24 be the graph obtained by removing the vertices of clique degree one from graph @xmath5 .
i.e. @xmath24 is the induced subgraph of @xmath5 having all the common vertices of @xmath5 .
[ sy1 ] if @xmath5 is a graph satisfying the hypothesis of the conjecture [ efl ] and every @xmath9 ( @xmath10 ) has at most @xmath11 vertices of clique degree greater than 1 , then @xmath5 is @xmath1-colorable .
let @xmath5 be a graph satisfying the hypothesis of the conjecture [ efl ] and every @xmath9 ( @xmath10 ) has at most @xmath11 vertices of clique degree greater than 1 .
let @xmath24 be the induced subgraph of @xmath5 consisting of the vertices of clique degree greater than one in @xmath5 .
define @xmath25 , @xmath26 for @xmath27 . from @xcite ,
it is true that the vertices of clique degree greater than or equal to @xmath11 can be assigned with at most @xmath1 colors .
assign the colors to the vertices of clique degree in non increasing order .
assume we next color a vertex @xmath7 of clique degree @xmath28 .
at this point only vertices of clique degree @xmath29 have been assigned colors . by assumption
every @xmath9 ( @xmath10 ) has at most @xmath11 vertices of clique degree greater than 1 and clique degree of @xmath7 is @xmath30 ( @xmath31 ) , then for these @xmath30 @xmath32 there are at most @xmath33 vertices have been colored .
therefore , there is an unused color from the set of @xmath1 colors , then that color can be assigned to the vertex @xmath7 .
[ sy2 ] if @xmath5 is a graph satisfying the hypothesis of the conjecture [ efl ] and every @xmath9 ( @xmath10 ) has at most @xmath12 vertices of clique degree greater than or equal to @xmath13 ( @xmath14 ) , then @xmath5 is @xmath1-colorable .
let @xmath5 be a graph satisfying the hypothesis of the conjecture [ efl ] and every @xmath9 ( @xmath10 ) has at most @xmath34 vertices of clique degree greater than or equal to @xmath13 ( @xmath14 ) .
let @xmath24 be the induced subgraph of @xmath5 consisting of the vertices of clique degree greater than one in @xmath5 .
define @xmath25 , @xmath26 for @xmath35 .
assign the colors to the vertices of clique degree in non increasing order .
assume we next color a vertex @xmath7 of clique degree @xmath30 .
at this point only vertices of clique degree @xmath29 have been assigned colors . by assumption
every @xmath9 ( @xmath10 ) has at most @xmath11 vertices of clique degree greater than 1 and clique degree of @xmath7 is @xmath30 ( @xmath31 ) , then for these @xmath30 @xmath9 s there are at most @xmath36 vertices have been colored .
therefore , there is an unused color from the set of @xmath1 colors , then that color can be assigned to the vertex @xmath7 .
let @xmath5 be a graph satisfying the hypothesis of the conjecture [ efl ] , the intersection matrix is a square @xmath37 matrix @xmath38 such that its element @xmath39 is @xmath40 when @xmath41 otherwise zero for @xmath42 and the diagonal elements of the matrix are all zero .
given below is a method to color graph @xmath5 satisfying the hypothesis of the conjecture [ efl ] using intersection matrix ( color matrix ) of the cliques @xmath9 s of @xmath5 and clique degrees of the vertices of @xmath5 .
let @xmath5 be a graph satisfying the hypothesis of conjecture [ efl ] .
let @xmath24 be the induced subgraph of @xmath5 consisting of the vertices of clique degree greater than @xmath43 in @xmath5 .
relable the vertex @xmath7 of clique degree greater than @xmath43 in @xmath5 by @xmath44 , where @xmath45 ; vertex @xmath7 is in @xmath46 .
define @xmath25 , @xmath26 for @xmath27 . if @xmath56 , then the graph @xmath5 has no vertex of clique degree greater than one , which implies @xmath5 has exactly @xmath57(maximum number ) vertices .
i.e. , @xmath5 is @xmath1 components of @xmath58 . otherwise , consider the intersection matrix @xmath38 as the color matrix and follow the steps .
step 1 : : : if @xmath59 , stop the process .
otherwise , let @xmath60 , for some @xmath61 .
then consider the sets @xmath62 and @xmath63 , go to step 2 .
step 2 : : : if @xmath64 , go to step 1 . otherwise , choose a vertex @xmath65 from @xmath62 , where @xmath66 and go to step 3 .
step 3 : : : let @xmath67 color @xmath68 appears atleast @xmath69 times in the @xmath70 row of the color matrix @xmath71 , @xmath72 . if @xmath73 , let @xmath74 , @xmath75 and go to step 4 . otherwise , construct a new color matrix @xmath76 by putting least @xmath77 in @xmath39 , where @xmath78 , @xmath49 , @xmath79 . then add the vertex @xmath65 to @xmath63 and remove it from @xmath62 , go to step 2 .
step 4 : : : choose a vertex @xmath7 from @xmath80 such that @xmath81 , for some @xmath82 , @xmath83 .
let @xmath84 and go to step 5 .
step 5 : : : let @xmath67 color @xmath68 appears atleast @xmath69 times in the @xmath70 row of the color matrix@xmath71 , for every @xmath85 . if @xmath86 add the vertex @xmath7 to @xmath87 and remove it from @xmath80 , go to step 4 . otherwise construct a new color matrix @xmath88 by putting @xmath77 in @xmath39 , where @xmath89 , @xmath49 , @xmath90 .
go to step 3 .
thus , we get the modified color matrix @xmath91 .
then , color the vertex @xmath7 of @xmath24 by @xmath39 of @xmath91 , whenever @xmath92 .
then , extend the coloring of @xmath24 to @xmath5 .
thus @xmath5 is @xmath1-colorable .
relabel the vertices of clique degree greater than one in @xmath5 by @xmath99 where @xmath100 .
the labeled graph is shown in figure [ n6lg ] .
figure [ n6lgh ] is the graph @xmath24 , where @xmath24 is obtained by removing the vertices of clique degree 1 from @xmath5 .
* step 3 : * since @xmath122 , @xmath123 , @xmath124 , @xmath125 and @xmath126 , choose the minimum color from the set @xmath127 and construct a new color matrix @xmath76 by putting @xmath43 in @xmath39 , @xmath49 , @xmath128 .
add the vertex @xmath120 to @xmath129 and remove it from @xmath121 .
then * step 3 : * since @xmath139 , @xmath140 , @xmath141 , and @xmath142 , choose the minimum color from the set @xmath143 and construct a new color matrix @xmath88 by putting @xmath144 in @xmath39 , @xmath49 , @xmath145 .
add the vertex @xmath137 to @xmath146 and remove it from @xmath138 . then * step 3 : * since @xmath156 , @xmath157 and @xmath158 , choose the minimum color from the set @xmath159 and construct a new color matrix @xmath160 by putting @xmath161 in @xmath39 , @xmath49 , @xmath162 .
add the vertex @xmath154 to @xmath163 and remove it from @xmath155 . then * step 3 : * since @xmath168 , @xmath169 and @xmath170 , choose the minimum color from the set @xmath171 and construct a new color matrix @xmath172 by putting @xmath173 in @xmath39 , @xmath49 , @xmath174 . add the vertex @xmath167 to @xmath163 and remove it from @xmath155 . then * step 3 : * since @xmath179 , @xmath180 and @xmath181 , choose the minimum color from the set @xmath182 and construct a new color matrix @xmath183 by putting @xmath161 in @xmath39 , @xmath49 , @xmath184 .
add the vertex @xmath178 to @xmath163 and remove it from @xmath155 .
then * step 3 : * since @xmath189 , @xmath190 and @xmath191 , choose the minimum color from the set @xmath192 and construct a new color matrix @xmath193 by putting @xmath173 in @xmath39 , @xmath49 , @xmath194 . add the vertex @xmath188 to @xmath163 and remove it from @xmath155 . then assign the colors to the graph @xmath24 using the matrix @xmath199 , i.e. , color the vertex @xmath7 by the @xmath200-th entry @xmath39 of the matrix @xmath91 , whenever @xmath41 ( see figure [ n6clgh ] ) , where the numbers 1 , 2 , 3 , 4 , 5 , 6 corresponds to the colors maroon , tan , green , red , blue , cyan respectively . extend the coloring of @xmath24 to @xmath5 by assigning the remaining colors which are not used for @xmath9 from the set of @xmath201-colors to the vertices of clique degree one in each @xmath9 , @xmath202 . the colored graph @xmath5 is shown in figure [ n6clg ] .
if @xmath5 is isomorphic to the graph @xmath205 for some @xmath1 , then the result is obvious .
if not there exists at least one vertex @xmath7 of clique degree greater than @xmath144 . define @xmath206 then @xmath207 . for every unordered pair of elements @xmath200 of @xmath208
there is a vertex @xmath209 in @xmath205 . therefore corresponding to the elements of @xmath208
there are @xmath210 vertices in @xmath205 .
since @xmath5 satisfies the hypothesis of conjecture [ efl ] , there is no vertex @xmath211 different from @xmath7 in @xmath5 such that @xmath212 where @xmath213 .
therefore for every vertex @xmath7 of clique degree greater than @xmath43 in @xmath5 , there are @xmath214 vertices of clique degree greater than @xmath43 in @xmath205 . as there are @xmath215 vertices of clique degree greater than @xmath43 in @xmath205 , @xmath216 where @xmath204 is the set of all vertices of clique degree greater than @xmath43 in @xmath5 .
let @xmath220 be the set of vertices of clique degree greater than @xmath43 in @xmath5 and @xmath221 .
we have to prove that @xmath5 has at most @xmath222 vertices of clique degree @xmath218 .
suppose @xmath5 has @xmath223 vertices of clique degree @xmath218 .
then by the definition of @xmath224 , it follows that , @xmath225 vertices are in @xmath224 .
let those vertices be @xmath226 . by theorem [ t2 ] | in 1972 , erds - faber - lovsz ( efl ) conjectured that , if @xmath0 is a linear hypergraph consisting of @xmath1 edges of cardinality @xmath1 , then it is possible to color the vertices with @xmath1 colors so that no two vertices with the same color are in the same edge . in 1978 ,
deza , erds and frankl had given an equivalent version of the same for graphs : let @xmath2 denote a graph with @xmath1 complete graphs @xmath3 @xmath4 , each having exactly @xmath1 vertices and have the property that every pair of complete graphs has at most one common vertex , then the chromatic number of @xmath5 is @xmath1 .
the clique degree @xmath6 of a vertex @xmath7 in @xmath5 is given by @xmath8 .
in this paper we give a method for assigning colors to the graphs satisfying the hypothesis of the erds - faber - lovsz conjecture using intersection matrix of the cliques @xmath9 s of @xmath5 and clique degrees of the vertices of @xmath5 .
also , we give theoretical proof of the conjecture for some class of graphs . in particular we show that : 1 .
if @xmath5 is a graph satisfying the hypothesis of the conjecture [ efl ] and every @xmath9 ( @xmath10 ) has at most @xmath11 vertices of clique degree greater than 1 , then @xmath5 is @xmath1-colorable .
2 . if @xmath5 is a graph satisfying the hypothesis of the conjecture [ efl ] and every @xmath9 ( @xmath10 ) has at most @xmath12 vertices of clique degree greater than or equal to @xmath13 ( @xmath14 ) , then @xmath5 is @xmath1-colorable .
department of mathematical and computational sciences , + national institute of technology karnataka , + surathkal , mangalore-575025 + _ e - mail : _
[email protected] , [email protected] * keyword * _ chromatic number , erds - faber - lovsz conjecture
_ * msc[2010 ] * _ 05a05 , 05b15 , 05c15 _ |
the era of large spectroscopic surveys of massive stars has already begun , providing us with a huge amount of high - quality spectra of galactic and extragalactic o - type stars .
the _ iacob spectroscopic survey of northern galactic ob - stars _ is one of them .
this long - term observational project is aimed at building a multi - epoch , homogeneous spectroscopic database of high - resolution , high signal - to - noise ratio spectra of galactic bright ob - stars and [ 2 ] . ] . associated with this spectroscopic dataset
are several working packages aimed at its scientific exploitation . within the framework of the working package _
wp3 : quantitative spectroscopic analysis _
, we have developed a powerful tool for the automatic analysis of optical spectra of o - type stars .
the tool provides a fast and objective way to determine the stellar parameters and the associated uncertainties of large samples of o - stars within a reasonable computational time .
initially developed to be used for the analysis of spectra of o - type stars from the _ iacob spectroscopic database _ , the tool is now also being applied in the context of the _ vlt - flames tarantula survey _
( vfts ) project [ 3 ] , and other studies of stars of this type .
apart from the already mentioned characteristics ( automatic , objective , and fast ) , we also aimed at a tool that is portable , versatile , adaptable , extensible , and easy to use .
as shown throughout the following text , this philosophy has guided the whole development of the tool .
+ one of the key - stones of any automatic quantitative spectroscopic analysis is the computation of large samples of synthetic spectra ( using a stellar atmosphere code ) , to be compared with the observed spectrum .
in contrast to other possible alternatives ( the _ genetic algorithm _ ga employed by [ 4 ] , or the _ principal component analysis _
pca approach followed by [ 5 ] ) , we decided to base our automatic tool on an extensive , precomputed grid of stellar atmosphere models and a line - profile fitting technique ( i.e. , a grid - based approach for the analysis of a sample of 12 low resolution spectra of b supergiants in ngc55 . ] ) this option was possible thanks to the fast performance of the fastwind code is the fastest . ] , and the availability of a cluster of @xmath0300 computers at the instituto de astrofsica de canarias ( iac ) connected through the condor workload management system . + the grid is complemented by a variety of programs implemented in idl , to handle the observations , perform the automatic analysis , and visualize the results .
the idl - package has been built in a modular way , allowing the user , e.g. , to easily modify how the mean values and uncertainties of the considered parameters are computed , or which evolutionary tracks will be used to estimate the evolutionary masses . in the following we outline
the main characteristics of the fastwind grid used for the analysis of galactic o - type stars that is presently incorporated within the iacob grid - based tool .
+ * * effective temperature ( @xmath1 ) and gravity ( log_g _ ) : * table [ grid_table ] indicates the ranges of @xmath1 and log_g _ considered in the grid .
basically , the grid points were selected to cover the region of the log@xmath1log_g _ diagram where the o - type stars are located . *
* helium abundance ( @xmath2 ) and microturbulence ( @xmath3 ) * : the grid includes six values of helium abundance @xmath2=n@xmath4/(n@xmath5+n@xmath4 ) , indicated in table [ grid_table ] . for all models , a microturbulence @xmath3@xmath6= 15kms@xmath7 was adopted in the computation of the atmospheric structure , and four values of the microturbulence were considered in the formal solution . * * radius ( @xmath8 ) : * computing a fastwind model requires an input value for the radius .
this radius has to be close to the actual one ( which will be derived in the final step of the analysis , and hence is not know from the beginning on ) .
we assumed a radius for each ( @xmath1 , log_g _ ) pair following the calibration by [ 8 ] .
the grid is hence divided in 20 regions in which a different radius is considered . for example , for the case of log_g_=4.0 and 3.5 dex , the radii range from 7 to 12 @xmath9 and from 19 to 22 @xmath9 , respectively . *
* wind parameters ( @xmath10 , @xmath11 , @xmath12 ) : * as conventional for grid computations of stellar atmosphere models for the optical analysis of o - stars , the mass loss rate ( @xmath10 ) and terminal velocity ( @xmath11 ) of the wind have been compressed , together with the radius , into the wind strength parameter ( or optical depth invariant ) , @xmath13 ( see [ 7 ] ) .
ten log@xmath14-planes were considered for the grid . for each fastwind model , @xmath10 and @xmath11
need to be specified .
first , a terminal velocity @xmath11=2.65@xmath15 was adopted , following [ 9 ] ; then a mass loss rate was computed for the given log@xmath14 , @xmath11 , and @xmath8 .
finally , the exponent of the velocity law , @xmath12 , was assumed as a free parameter ranging from 0.8 to 1.8 . * * metallicity * : a solar metallicity ( following [ 10 ] ) was assumed for the whole grid=0.5 , 0.4 , and 0.2 @xmath16 ) . ] .
+ _ synthetic lines : _ the following ( optical ) lines where synthesized in the formal solution : h@xmath17 , hei@xmath184026 , 4120 , 4143 , 4387 , 4471 , 4713 , 4922 , 5015 , 5048 , 5875 , 6678 , and heii@xmath184026 , 4200 , 4541 , 4686 , 5411 , 6406 , 6527 , 6683 . location of ( @xmath1 , log_g_)-pairs considered in the fastwind grid incorporated into the present version of the iacob grid - based tool .
evolutionary tracks from [ 11 ] . ]
r l log_g _ : & [ 2.64.3]dex ( with step size 0.1 dex ) + @xmath1 : & @xmath1925000k ( step size 500k ) , upper limit defined by the 120 m@xmath20 track ( see fig . [ grid_figure ] ) + @xmath3 : & model : 15 kms@xmath7 , formal solution : 5 , 10 , 15 , 20 kms@xmath7 + @xmath2 : & 0.06 , 0.09 , 0.13 , 0.17 , 0.20 , 0.23 + log@xmath14 : & -15.0 , -14.0 , -13.5 , -13.0 , -12.7 , -12.5 , -12.3 , -12.1 , -11.9 , -11.7 + @xmath12 : & 0.8 , 1.0 , 1.2 , 1.5 , 1.8 + in order to optimize the size and the read - out time of the grid , only part of the output from the fastwind models is kept and stored in idl xdr - files .
this includes the input parameters , the h / he line profiles and equivalent widths , the information about the stellar atmosphere structure and the emergent flux distribution , and the synthetic photometry resulting from the computed emergent flux distribution . using these xdr - files , the size of the grid can be reduced to a 10% of the original one .
idl can restore each xdr - file and compute the corresponding @xmath21 quantity ( see sect .
[ idl ] ) in 0.020.1 s per model , i.e. , the tool can pass through 80,000 models in 30 min1 hour .
following the main philosophy of the iacob grid - tool , only the reduced grid is used by the tool , whilst the original grid is safely kept on hard - disk at the iac .
this allows for an easy transfer of the grid to an external disk , hence satisfying our constraint for portability . +
the fastwind grid presently incorporated into the iacob grid - based tool consists of @xmath0192,000 models ( @xmath032,000 models per he - plane ) .
the reduced size is @xmath034 gb . this predefined grid can be easily updated and/or extended if necessary using appropriate scripts implemented in idl , condor , and linux .
for example , a new he - plane can be computed and prepared for use in @xmath03 days .
the guideline of the automatic analysis is based on standard techniques applied within the quantitative spectroscopic analysis of o - stars using optical h / he lines that have been described elsewhere ( e.g. [ 12 ] , [ 13 ] ) .
the whole spectroscopic analysis is performed by means of a variety of idl programs following the steps indicated below . in brief , once the observed spectrum is processed , the tool obtains the quantity @xmath22 ( i.e. an estimation of the goodness of fit ) for each model within a subgrid of models selected from the global grid , and determines the stellar parameters and their associated uncertainties by interpreting the obtained @xmath22 distributions .
+ + in this first step , the user has to provide the observed spectrum , to indicate the corresponding resolving power ( @xmath8 ) , and to give some pre - determined information about the star the projected rotational velocity ( @xmath23sin@xmath24 ) , the size of the extra line - broadening ( @xmath25 ) , and the absolute visual magnitude ( @xmath26 ) .
+ concerning the grid , the user must select the appropriate metallicity , and indicate the range of values for the various free parameters ( defining the subgrid of models to be considered within the analysis ) .
this latter option allows the tool to be faster ( using optimized ranges for the various free parameters ) , or the user to perform a preliminary quick analysis by fixing some of the parameters .
for example , one can obtain a quick estimate on @xmath1 , log_g _ , and log@xmath14 in less than 5 min , by fixing the other three parameters ( @xmath3 @xmath2 , and/or @xmath12 ) .
+ finally , the h / he lines which shall be considered in the analysis and the corresponding weights need to be specified .
+ _ : processing of the observed spectrum _ + in many cases , the observed spectra need to be processed before launching the automatic analysis , because of , e.g. , nebular contamination affecting the cores of the h and hei lines , the need to improve the normalization of the continuum adjacent to the line , and the presence of cosmic rays .
+ the possibility to correct the observed spectrum for these effects has been incorporated into corresponding idl procedures .
the implemented options include ( for each of the considered lines ) : local renormalization , selection of the wavelength range of the line , clipping / restoring of certain parts of the line , and computation of the signal - to - noise ratio ( s / n ) in the adjacent continuum .
+ once finalized , the processed spectrum is stored .
this spectrum and all associated information is used in _
step 3 _ , and also each time the analysis of the same star / spectrum is ( re-)launched . this way the user can easily reconsider his / her decisions after a first analysis .
+ in this step , the user can select a model from the grid to help him / her with the processing of the observed spectrum .
one can , for example , make a quick preliminary analysis ( see above ) fixing some parameters and including only few lines , and then use the resulting best model to check the processed lines and find constraints for the processing of the others . + + this step , together with _
step 4 _ , are the most important ones , constituting the core of the program . in the present version of the iacob grid - based tool we have considered a procedure as described below ; however , we are aware that this procedure can be subject to discussion / improvements . having this in mind , and following the philosophy of the tool ( regarding versatility and adaptability ) ,
the corresponding idl modules have been implemented with the possibility to be easily modified .
the fast performance of the iacob grid - based tool makes it very powerful to investigate various alternative strategies .
+ in our present version , the tool computes , for every model in the subgrid , the quantity @xmath27 for each considered line @xmath28 where @xmath29 and @xmath30 are the normalized fluxes corresponding to the synthetic and observed spectrum , respectively ; @xmath31=(s / n)@xmath7 accounts for the s / n of the line ; and @xmath32 is the number of frequency points in the line .
under ideal conditions ( e.g. , for a perfect model , but see below ) , @xmath27 corresponds to a reduced @xmath33 .
+ in a second step , the @xmath27 values for each model are corrected for possible deficiencies in the synthetic lines ( due , for example , to deficiencies in the model , an incorrect characterization of the noise of the line , bad placement of the continuum , or bad characterization of the line - broadening ) . to this aim
we compute , for each line , the standard deviation of the residuals @xmath34=@xmath35(@xmath29@xmath30 ) from that model that results in the minimum @xmath27 for the given line .
then , the following correction is applied : @xmath36 using these individual @xmath37 values and the weights assumed for each of the considered lines ( @xmath38 , e.g. , [ 4 ] ) , a global @xmath21 is obtained : @xmath39 where @xmath40 is the number of lines . for a large number of frequency points per line , @xmath21 should be normally distributed .
+ thus , _ step 3 _ provides the values of the quantity @xmath22 associated with each of the models included in the considered subgrid . as indicated in section [ grid ] , this step can last from a few seconds to less than about 1 hour , depending on the number of models in the subgrid .
an example of @xmath22-distributions ( actually , @xmath41=e@xmath42 ) with respect to the various stellar parameters is presented in figure [ fig2 ] .
+ = 46,000 ) .
@xmath23sin@xmath24 and @xmath25 were previously determined using a combined fourier transform + goodness of fit technique ( see e.g. [ 14 ] ) . in this example , only 5 h / he lines were used for the actual analysis . note
that also all other synthetic lines fit the observations perfectly at the derived parameters .
the analysis was performed in @xmath02 min .
see [ 2 ] for details on the distance , apparent visual magnitude and visual extinction used to determine the absolute visual magnitude ( @xmath43 ) .
[ fig2 ] ] _ : computation of mean values and uncertainties _
+ the previous step provides the @xmath21distributions for each of the six parameters derived through the spectroscopic analysis ( @xmath1 , log_g _ , @xmath2 , @xmath3 , log@xmath14 , and @xmath12 ) .
if the absolute visual magnitude is provided , the tool automatically determines @xmath8 , log@xmath44 , and @xmath45 ( spectroscopic mass ) for each model , and hence the corresponding @xmath21distributions for these parameters are available as well .
if also the terminal velocity is provided , a similar computation is performed for the mass - loss rate ( using @xmath14 , @xmath8 , and @xmath11 ) .
finally , one of the idl modules computes the evolutionary masses ( @xmath46 ) of the models from an interpolation in the ( log@xmath1 , log_g_)-plane using the tracks provided by a stellar evolution code . since the @xmath21distributions for @xmath1 and log_g _ are computed and stored in _
step 3 _ , the computation of the corresponding distributions for @xmath8 , log@xmath44 , @xmath45 , @xmath46 can be easily repeated in a few seconds , in case a different @xmath43 and/or evolutionary tracks need to be considered . + these distributions are then used to compute mean values and uncertainties for each parameter ( taking into account that models above a given threshold in @xmath21 can be discarded ) .
+ the tool also allows to easily investigate possible degeneracies ( for example , for stars with weak winds the analysis of the optical h / he lines only provides an upper limit for the @xmath14 parameter , and does not allow to constrain @xmath12 ) , and the contribution of the various parameters to the final uncertainty ( by fixing one of the free parameters to a certain value and recomputing the statistics for the other parameters ) .
+ + the last step is the creation of a summary plot for better visualization of the results ( see figure [ fig2 ] as an example ) .
the present version of the tool includes : * @xmath41distributions for the various parameters involved in the analysis ( upper - left panels ) . in those panels ,
vertical lines indicate the limiting values adopted for the six free parameters ( i.e. , the sub - grid of fastwind models for which a @xmath22 is computed ) , and horizontal lines indicate the @xmath41 value corresponding to @xmath21(threshold ) .
the user can select three of the models that will be indicated as red , blue , and pink dots ( the same colors are also used in other parts of the figure ) . * a summary of the input parameters ( @xmath23sin@xmath24 , @xmath25 , and @xmath43 ) and the result from the statistics for each parameter resulting from the analysis ( upper - right part of the figure ) .
* a log@xmath1log_g _ diagram including evolutionary tracks ( e.g. , from [ 11 ] , without rotation , in the figure ) and the position of all models with @xmath22@xmath47@xmath21(threshold ) . *
a set of panels ( lower part of the figure ) were the observed and synthetic h / he lines indicated by the user are compared .
the spectral regions that are used to compute the @xmath22 for each line are indicated in black , while the clipped ( or not used ) regions are presented in green .
note that the plotted lines are not limited to those used in the analysis process ( the latter are marked with an x ) . the fast performance and versatility of the iacob grid - based tool not only allows to analyze large samples of o - type stars in a reasonable time , but also to easily investigate the effects of * _ the assumed line - broadening _ : the common strategy in previous spectroscopic analyses of o - stars was to assume pure rotational line - broadening ( in addition to natural , thermal , stark / collisional and microturbulent broadening included in the synthetic spectra ) .
however , recent studies have shown that an important extra - broadening contribution ( commonly quoted as _ macroturbulent broadening _ ) affects the shape of the line - profiles of this type of stars ( see e.g. [ 15 ] , and references therein ) .
how much are the derived parameters affected when this extra - broadening is neglected ? can both broadening contributions be represented by one fake rotational profile ( @xmath48sin@xmath24 ) , without disturbing the resulting parameters ? what is the effect of the uncertainty in the derived / assumed broadening ? some results from a preliminary investigation of these questions can be found in [ 16 ] . * _ the placement of the continuum _ : the normalization of the spectra of o - type stars is sometimes complicated , especially in the case of echelle spectra .
it is commonly argued that the derived gravity can be severely affected by the assumed normalization , but , to which extent ? to investigate this effect one could use the iacob grid - based tool , applying small modifications to the continuum placement .
an example of this type of analysis for the case of low resolution spectra of b - supergiants can be found in [ 6 ] . * _ neglecting / including a variety of different commonly used diagnostic lines _ : before computing the global @xmath21 ( see equation 3 ) , the resulting @xmath27 for each line are stored . this way
the user can easily recompute @xmath21 , discarding some of the initially considered lines .
this option allows , for example , to investigate the change in the wind - strength parameter when both h@xmath49 and heii4686 lines or only one of them are included in the analysis . *
_ clipping part of the lines _ : most of the o - stars are associated to hii regions . in some cases , the stellar spectra can be heavily contaminated by nebular emission lines ( mainly the cores of the hydrogen lines , but also the hei lines ) .
this contamination must be identified and eliminated from the stellar spectrum to obtain meaningful results from the spectroscopic analysis .
when the resolution of the spectra is high , the clipped region is small in comparison with the total line width ; however , even for a moderate resolution an important region of the line needs to be eliminated .
what is the effect of such a strong clipping in the h@xmath49 line on the determination of the mass - loss rate ?
how large is the effect of clipping the core of the hei lines in the @xmath1 determination ?
* _ the way the statistics is derived from the @xmath21distributions _ : as commented in section [ idl ] , this is a critical point in the determination of final best values and uncertainties .
these are only some examples of things can be investigated with the iacob grid - based tool .
many other tests can be easily performed .
in addition , we expect the tool to be of great benefit for the analysis of the o - star samples included in , e.g. , the eso - gaia , vfts , own , iacob and other similar large surveys
. financial support by the spanish ministerio de ciencia e innovacin under projects aya2008 - 06166-c03 - 01 , aya2010 - 21697-c05 - 04 , and by the gobierno
de canarias under project pid2010119 .
this work has also been partially funded by the spanish micinn under the consolider - ingenio 2010 program grant csd2006 - 00070 : first science with the gtc ( http://www.iac.es/consolider-ingenio-gtc ) .
16 simn - daz s , castro n , garcia m , herrero a and markova n 2011a , _ bsrs liege _ ,
* 80 * , 514 simn - daz s , garcia m , herrero a , maz - apellniz j , and negueruela i 2011b , _ proc .
`` star clusters and associations : a ria workshop on gaia '' _ preprint _ arxiv1109.2665s evans c j et al . 2011 , _
a&a _ , * 530 * , a108 mokiem m r , de koter a , puls j , herrero a , najarro f and villamariz m r 2005 , _ a&a _ , * 441 * , 711 urbaneja m a , kudritzki r p , bresolin f , przybilla n , gieren w and pietrzyski g 2008 , _ apj _ , * 684 * , 118 castro , n. , herrero , a. , urbaneja , m. a , et al . (
_ a&a _ , submitted ) puls j , urbaneja m a , venero r , repolust t , springmann u , jokuthy a and mokiem m r 2005 , _
a&a _ , * 435 * , 669 martins f , schaerer d and hillier , d. j. 2005 , _ a&a _ , * 436 * , 1049 kudritzki r p and puls j 2000 , _ a&a _ , * 38 * , 613 asplund m , grevesse n , sauval a j and scott p 2009 , _
ara&a _ , * 47 * , 481 schaller g , schaerer d , meynet g and maeder a 1992 , _ a&as _ , * 96 * , 269 herrero a , puls j. and najarro f 2002 , _ a&a _ , * 396 * , 949 repolust t , puls j and herrero a 2004 , _
a&a _ , * 415 * , 349 simn - daz s , herrero a , uytterhoeven k , castro n , aerts c and puls j 2010 , _ apjl _ , * 720 * , l174 simn - daz s 2011 , _ brsr liege _ ,
* 80 * , 86 sabn - sanjulian c , simn - daz s , garcia m , herrero a , puls j and castro n 2011 , _ proc .
_ in honour of a. moffat | we present the iacob grid - based automatic tool for the quantitative spectroscopic analysis of o - stars .
the tool consists of an extensive grid of fastwind models , and a variety of programs implemented in idl to handle the observations , perform the automatic analysis , and visualize the results .
the tool provides a fast and objective way to determine the stellar parameters and the associated uncertainties of large samples of o - type stars within a reasonable computational time . |
besides its theoretical interest , the role of quantum entanglement has been acknowledged for its wide range of applications in quantum information protocols @xcite and quantum communication @xcite . on the theoretical side ,
one of the key results is the positive partial transposed ( ppt ) separability criterion @xcite , which provides a necessary and , in some cases , a sufficient condition for distinguishing between separable and entangled states in discrete quantum systems .
this criterion was extended to continuous variable systems @xcite by implementing the partial transpose operation as a mirror reflection in the wigner phase - space .
an important application of the `` continuous '' ppt criterion is in the theory of quantum information of gaussian states @xcite , which is at the core of the test beds for estimating quantum correlations @xcite . in this case , the ppt criterion yields both a necessary and sufficient separability condition @xcite . as will be seen , gaussian states are also quite useful for investigating the entanglement exclusively induced by a noncommutative ( nc ) deformation of phase - space . noncommutativity is believed to be an essential feature of quantum gravity and string theory @xcite . in the context of quantum mechanics ( qm ) and quantum cosmology , phase - space noncommutativity has been shown to exhibit striking features with applications for the black hole singularity @xcite and the equivalence principle @xcite .
furthermore , the quantum mechanical aspects of nc theories have focused on studies of the quantum hall effect @xcite , the gravitational quantum well for ultra - cold neutrons @xcite , the landau/@xmath0-oscillator problem in the phase - space @xcite , and as a probe of quantum beating and missing information effects @xcite .
the phase - space nc generalizations of qm are based on extensions of the heisenberg - weyl algebra @xcite . in this letter ,
the ppt entanglement criterion for continuous variable systems is generalized to the case of a phase - space with canonical momentum - momentum and position - position noncommutativity .
two alternative formulations of the new criterion are presented .
the second formulation consists on a condition on the smallest symplectic eigenvalue of the partially transposed covariance matrix of the state , and is very efficient from the computational point of view .
we will use it to study in detail the effect of noncommutativity on the quantum nature and separability properties of gaussian states .
most interesting is the conclusion that noncommutativity alone can induce the entanglement of gaussian states . besides its theoretical interest , this result may become important for the experimental tests of nc qm . for the purpose of quantifying how noncommutativity affects the separability of quantum states , let us consider a bipartite quantum system described in terms of a @xmath1-dimensional subsystem a ( alice ) and a @xmath2-dimensional subsystem b ( bob ) with @xmath3 .
one may write the collective degrees of freedom of the composite system @xmath4 , where @xmath5 and @xmath6 are the corresponding generalized variables of the two subsystems , which obey the commutation relations @xmath7 = i \omega_{ij } , \hspace{1 cm } i , j = 1 , \cdots , 2n , \label{eq1}\ ] ] where the associated matrix is given by @xmath8 \equiv
diag\left[\bf \omega^a,\,\bf \omega^b\right]$ ] , with @xmath9 a real skew - symmetric non - singular @xmath10 matrix @xmath11 of the form @xmath12 where @xmath13 $ ] and @xmath14 $ ] measure the noncommutativity of the position - position and momentum - momentum sectors of the subsystem @xmath15 , @xmath16 is the @xmath17 identity matrix , and we have set @xmath18 .
the nc structure can be formulated in terms of commuting variables by considering a linear darboux transformation ( dt ) , @xmath19 , where @xmath20 , with @xmath21 and @xmath22 , satisfy the usual commutation relations : @xmath23 = i j_{ij } , \hspace{1 cm } i , j=1 , \cdots , 2n .
\label{eq3}\ ] ] here @xmath24=diag \left[{\bf j^a},{\bf j^b } \right ] $ ] where @xmath25 is the @xmath26 standard symplectic matrix for @xmath27 .
the linear transformation @xmath28 is such that @xmath29 $ ] , and @xmath30 , or equivalently @xmath31 for @xmath32 .
notice that the map @xmath33 is not uniquely defined .
if we compose @xmath33 with block - diagonal canonical transformations we obtain an equally valid dt .
we will see , however , that our main results are independent of the particular choice of the map @xmath34 $ ] .
let us assume that the composite quantum system is described by the density matrix @xmath35 , function of the nc variables @xmath36 .
the dt yields the density matrix @xmath37 , which is associated with the wigner function : w ( ) = _
^n_a dy^a _ ^n_b dy^b
e^- i ( y^a k^a + y^b k^b ) q^a + y^a 2,q^b + y^b 2 q^a -y^a 2,q^b -y^b 2 . [ eq5 ] upon inversion of the dt , we obtain the nc wigner function @xcite : @xmath38 if @xmath39 denotes the covariance matrix of @xmath40 and @xmath41 that of @xmath42 , then the two are related by : @xmath43 a necessary condition for the phase - space function @xmath42 with covariance matrix @xmath44 to be an admissible ( commutative ) wigner function is that it satisfies the robertson - schrdinger uncertainty principle ( rsup ) : @xmath45 .
that is , the matrix @xmath46 has to be a @xmath47 positive matrix in @xmath48 . from this condition and eq .
( [ eq7 ] ) we conclude that for @xmath40 to be a _ bona fide _ nc wigner function , it has to satisfy the nc rsup @xcite : @xmath49 for gaussian states this condition is also sufficient @xcite .
a composite quantum system is separable if its density matrix takes the form @xmath50 , where @xmath51 , for all @xmath52 , @xmath53 and @xmath54 ( resp .
@xmath55 ) is a density matrix which involves only alice s ( resp .
bob s ) coordinates @xmath56 ( resp .
@xmath57 ) .
the associated wigner function is : @xmath58 using the dt we obtain the commutative counterpart in terms of the commutative variables @xmath59 : @xmath60 to simplify the manipulation of the above results , let us define @xmath61 to be the @xmath62 matrix @xmath63 $ ] , with @xmath64 $ ] .
thus , the transformation @xmath65 amounts to a mirror reflection of bob s momenta . according to the ppt criterion ,
if a wigner function @xmath66 is that of a separable state , then under the transformation @xmath67 one obtains an equally admissible wigner function . hence , if the state @xmath68 is separable then @xmath69 where @xmath70 is the covariance matrix of @xmath71 . the transformation , eq .
( [ eq11 ] ) , can be rewritten in terms of the nc variables as follows : @xmath72 where @xmath73 and @xmath74 are defined accordingly to eq .
( [ eq6 ] ) and @xmath75 $ ] .
it follows from eq .
( [ eq7 ] ) that the covariance matrices @xmath76 and @xmath77 of @xmath78 and @xmath79 are related by @xmath80 .
hence , the separability condition eq .
( [ eq12 ] ) can then be written exclusively in terms of the nc objects @xmath81 in addition , notice that @xmath82 where @xmath39 is the covariance matrix of @xmath83 .
defining @xmath84 we obtain @xmath85 we point out that the matrix @xmath86 is simply given by @xmath87 , \label{eq15.1}\ ] ] where one has used the definition of @xmath88 and the fact that @xmath89 is an anti - symplectic transformation , i. e. @xmath90 . by itself , this result is an elegant generalization of the result from ref .
@xcite stating that @xmath91 is replaced by @xmath92 $ ] under ppt .
we stress the fact that eq .
( [ eq15.1 ] ) is valid assuming that the dts take the block - diagonal form @xmath93 $ ] .
finally , for gaussian functions our results yield the following characterization : a real normalizable phase - space gaussian function @xmath94 with covariance matrix @xmath39 is the nc wigner function of a quantum separable state if and only if @xmath39 satisfies the eqs .
( [ eq8 ] ) and ( [ eq15 ] ) with respect to quantum nature and separability , respectively .
to proceed we reexpress the nc quantumness and separability criteria in terms of the so - called nc symplectic spectrum .
first recall that the _ symplectic spectrum _ of a @xmath47 real symmetric positive - definite matrix @xmath41 is the set of eigenvalues of the matrix @xmath95 and these eigenvalues are called the williamson invariants of @xmath41 .
they are all positive and we denote by @xmath96 the smallest williamson invariant . by williamson s theorem @xcite , one can show that @xmath41 complies with the rsup ( @xmath45 ) if and only if @xmath97 .
by the same token , we call the set of eigenvalues of @xmath98 the _ nc symplectic spectrum _ of @xmath39 and the eigenvalues are called the nc williamson invariants .
we denote by @xmath99 the smallest nc williamson invariant of @xmath39 . now consider a covariance matrix @xmath39 and let @xmath33 be any dt .
suppose that @xmath100 is related to @xmath39 by eq .
( [ eq7 ] ) . then we claim that the nc symplectic spectrum of @xmath39 coincides with the symplectic spectrum of @xmath100 . indeed : @xmath101 we then have the following sequence of equivalent statements : @xmath102 in other words , @xmath39 satisfies the nc rsup if and only if its smallest nc williamson invariant satisfies @xmath103 .
now we turn to the separability criterion eq .
( [ eq15 ] ) .
let @xmath104 denote the smallest nc williamson invariant of @xmath105 . if @xmath39 is the covariance matrix associated with a nc separable state , then @xmath106
. to summarize , a necessary condition for @xmath39 to be the covariance matrix of a nc quantum state is given by @xmath103 .
likewise , @xmath106 is a necessary condition for @xmath39 to be the covariance matrix of a separable nc quantum state . again ,
if the state is a gaussian both are sufficient criteria .
the nc symplectic spectrum of @xmath39 depends only on @xmath39 and on @xmath88 and hence it is manifestly independent of the dt @xmath33 . consequently , the nc quantum nature condition , @xmath103 , is also independent of @xmath33 . on the other hand ,
the nc symplectic spectrum of @xmath107 is also independent of @xmath33 , provided it is block - diagonal .
this is easily seen , if we consider the fact that the spectrum of @xmath108 is the same as that of @xmath109 and @xmath86 takes the form ( [ eq15.1 ] ) for all block - diagonal dts .
in order to evince the entanglement induced by the noncommutative deformation , let us consider a gaussian state , @xmath111 living on a @xmath112-dimensional nc phase - space with @xmath113 , @xmath114 , @xmath115 and @xmath116 , with @xmath117 , and where the covariance matrix is given by , = b2 ( c c * i*_4&*^t * + * * & * i*_4 ) , * * = ( c c c c n & 0 & m & 0 + 0 & n & 0 & -m + m & 0 & -n & 0 + 0 & -m & 0 & -n ) , [ eq20 ] where @xmath118 and @xmath119 are real parameters which we assume , for simplicity , to be constrained by the relations @xmath120 , @xmath121 . in this case , the dt corresponds to the map @xmath122 $ ] , with = * s^b * = ( c c c c & 0 & 0 & -/2 + 0 & & /2 & 0 + 0 & /2 & & 0 + /2 & 0 & 0 & ) , [ eq21 ] subject to @xmath123 , which ensures the map invertibility .
the nc parameters , @xmath124 and @xmath125 , are real positive constants satisfying the condition @xmath126 .
the smallest nc williamson invariants of @xmath39 and @xmath107 are given by _ - = 1 1+r , _
-^ = 1 1+r , respectively , where @xmath127 , which allows for examining the role of the nc parameters on the gaussian entanglement . the nc quantum nature and the separability of gaussian states
are ensured for @xmath128 and @xmath129 , respectively .
entangled quantum states are found in the range @xmath130 ( cf .
[ fig : results2 ] ) . setting @xmath131 ,
the obtained nc williamson invariants , @xmath99 and @xmath104 , depend exclusively on @xmath125 , as depicted in fig .
[ fig : results1 ] .
the standard ( commutative ) qm limit is obtained by setting @xmath132 , which implies that @xmath133 and @xmath134 .
this means that for @xmath135 , all states are quantum ( @xmath136 ) , and separable ( @xmath129 ) .
[ fig : results2 ] shows how the nc phase - space ( @xmath137 ) induces the entanglement of the corresponding gaussian states .
the influence of the nc parameters , @xmath124 and @xmath125 , on the quantum nature , separability and entanglement of states can be depicted for several values of @xmath138 ( with a degeneracy associated to @xmath139 ) .
in this work the ppt criterion for entanglement - separability was extended to nc qm in phase - space , and formulated exclusively in terms of the nc framework .
moreover , we studied the impact of changing the nc structure on the quantum nature and separability properties of gaussian states and showed that the entanglement can be exclusively induced by the nc structure of phase space . finally , we point out that these results may become increasingly important in view of the recently observed violation of the uncertainty principle @xcite .
_ acknowledgements - _ the work of cb is supported by the fct ( portugal ) grant sfrh / bpd/62861/2009 .
the aeb work is supported by the fapesp ( brazil ) grant 2012/03561 - 0 .
the work of ob is partially supported by the fct project ptdc / fis/111362/2009 .
ncd and jnp have been supported by the fct grant ptdc / mat/099880/2008 .
m. horodecki , j. oppenheim , and a. winter , nature * 436 * , 6 ( 2005 ) ; comm .
268 , 107 ( 2007 ) ; i. devetak , a. w. harrow , and a. winter , phys . rev . lett . * 93 * , 230503 ( 2004 ) . i. bengtsson and k. zyczkowski , _ geometry of quantum states : an introduction to quantum entanglement _ ( cambridge university press , cambridge , 2006 ) ; s. l. braunstein and a. k. pati , _ quantum information theory with continuous variables _ ( kluwer , dordrecht , 2002 ) ; m.a .
nielsen , i.l .
chuang , _ quantum computation and quantum information _
( cambridge university press , 2007 ) .
a. peres , phys .
77 * , 1413 ( 1996 ) ; p. horodecki , phys .
a * 232 * , 333 ( 1997 ) .
r. simon , e. c. g. sudarshan and n. mukunda , phys . rev . * a 36 * , 3868 ( 1987 ) .
g. adesso and f. illuminati , j. phys . a : math .
theor . * 40 * , 7821 ( 2007 ) ; n. schuch , j. i. cirac , and m. m. wolf , commun .
phys . * 267 * , 65 ( 2006 ) ; m. m. wolf , g. giedke , and j. i. cirac , phys .
96 , 080502 ( 2006 ) .
g. adesso , a. serafini and f. illuminati , phys .
lett . * 92 * , 087901 ( 2004 ) ; g. adesso , a. datta , phys
105 , 030501 ( 2010 ) ; g. adesso , d. girolami , a. serafini , phys .
lett . * 109 * , 190502 ( 2012 ) .
c. weedbrook , s. pirandola , r. garcia - patron , n. j. cerf , t. c. ralph , j. h. shapiro , and s. lloyd , rev .
phys . * 84 * , 671 ( 2012 ) .
a. connes , m. r. douglas and a. schwarz , jhep * 02 * , 003 ( 1998 ) ; m. r. douglas and c. hull , jhep * 02 * , 008 ( 1998 ) ; v. schomerus , jhep * 9906 * , 030 ( 1999 ) . n. seiberg and e. witten , jhep * 9909 * , 032 ( 1999 ) .
c.bastos , o. bertolami , n. c. dias and j. prata , phys . rev . *
d78 * , 023516 ( 2008 ) ; phys . rev . * d80 * , 124038 ( 2009 ) ; phys . rev .
* d82 * , 041502 ( 2010 ) ; phys . rev . * d84 * , 024005 ( 2011 ) .
c.bastos , o. bertolami , n. c. dias and j. prata , class .
* 28 * , 125007 ( 2011 ) .
r. prange and s. girvin , _ the quantum hall effect _ , ( springer , new york , 1987 ) .
j. bellissard , a. van elst and h. schulz - baldes , arxiv : cond - mat/9301005 .
o. bertolami , j. g. rosa , c. arago , p. castorina and d. zappal , phys . rev . *
d72 * , 025010 ( 2005 ) .
m. r. douglas and n. a. nekrasov , rev .
mod . phys . * 73 * , 977 ( 2001 ) .
m. rosenbaum and j. david vergara , gen .
rel . grav . * 38 * , 607 ( 2006 ) ; j. gamboa , m. loewe and j. c. rojas , phys . rev .
* d64 * , 067901 ( 2001 ) .
a. e. bernardini and o. bertolami , arxiv:1303.0685 [ quant - ph ] . c. bastos , o. bertolami , n. c. dias , j. n. prata , j. math
. phys . * 49 * , 072101 ( 2008 ) ; c. bastos , n. c. dias , j. n. prata , commun . math . phys . * 299 * , 709 ( 2010 ) ; n. c. dias , m. de gosson , f. luef , j. n. prata , j. math . phys . * 51 * , 072101 ( 2010 ) ; j. math .
pures appl . *
96 * , 423 ( 2011 ) ; s. adhikari , b. chakraborty and a. s. majumdar , phys . rev . * a 79 * , 042109 , ( 2009 ) . n. c. dias and j. n. prata , annals phys . * 324 * , 73 ( 2009 ) . c. bastos , o. bertolami , n.c .
dias and j.n .
prata , phys . rev . *
d 86 * , 105030 ( 2012 ) .
l. a. rozema , a. darabi , d. h. mahler , a. hayat , y. soudagar , and a. m. steinberg , phys .
lett . * 109 * , 100404 ( 2012 ) .
( dashed lines ) and @xmath140 ( solid lines ) for @xmath141 , and @xmath142 as function of @xmath125 in the range @xmath143 $ ] .
black lines correspond to the respective smallest eigenvalues , @xmath99 and @xmath144 .
notice that entanglement ( @xmath145 ) coexists with quantum behavior ( @xmath128 ) for @xmath146 . ] | the entanglement criterion for continuous variable systems and the conditions under which the uncertainty relations are fulfilled are generalized to the case of a noncommutative ( nc ) phase - space .
the quantum nature and the separability of nc two - mode gaussian states are examined .
it is shown that the entanglement of gaussian states may be exclusively induced by switching on the noncommutative deformation . |
investigating the spatial distribution of galaxies in the sky has taught us much about how the universe evolves and structures form .
however , galaxies contain only a few percent of the total baryonic mass in the universe ( e.g. * ? ? ?
* ) and their spatial distribution is strongly biased with respect to the dark matter distribution ( e.g. * ? ? ?
* ; * ? ? ?
mapping the distribution of the intergalactic medium ( igm ) could provide one of the most powerful tools in modern astrophysical cosmology as it contains most of the baryons and is only weakly biased with respect to the dark matter . while absorption lines in the spectra of background quasars provide 1-dimensional information along the line of sight , emission lines map the full 3-dimensional distribution of the gas , thus providing a more realistic picture of the physical state of the gas and of its evolution with time .
since the igm is the main reservoir of fresh material to fuel star formation in galaxies and is in return affected by feedback ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , understanding the properties of cosmic gas ultimately implies improving our understanding of the process of galaxy formation .
predictions for the emission from the low - redshift igm in the rest - frame ultraviolet ( uv ) have been made for both h ly@xmath1 ( @xcite ; @xcite ) and metal lines ( @xcite ; @xcite , paper ii hereafter ) . at @xmath6 ,
the emission is redshifted in to the optical band and accessible to ground - based telescopes , but predictions have only been made for h ly@xmath1 and he emission ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite and @xcite investigated fluorescent ly@xmath1 emission induced by the uv background and by the ionising radiation of bright quasars .
@xcite claim that while the detection of the glow from the diffuse igm is beyond the sensitivity of current telescopes , quasar - induced fluorescence in high - density regions should be detectable around bright quasars ( see also * ? ? ?
* ; * ? ? ?
indeed , such emission may already have been detected @xcite .
@xcite discussed the relative importance of fluorescence and collisional excitation and claim , however , that away from bright sources the latter is the dominant channel for the production of bright ly@xmath1 emission .
@xcite predict that he emission at @xmath7 is detectable by current facilities , and that the he 1640 line in particular , could be a useful tool to investigate the cooling mechanisms in the so - called `` ly@xmath1 blobs '' .
recent observations by @xcite and @xcite may have confirmed this prediction . as of today , ly@xmath1 blobs remain the main form of diffuse emission detected at high redshift ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) but their nature is still controversial . the ly@xmath1 emission in the blobs might be due to cooling radiation from gas infalling on to haloes ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but it is difficult to rule out other mechanisms such as galactic winds driven by supernova explosions ( e.g. @xcite ; @xcite ; @xcite ) and particularly recombination radiation powered by internal sources ( e.g. * ? ? ?
* ; * ? ? ?
the detection ( or non - detection ) of metal - line emission from the same gas phase that produces ly@xmath1 radiation would help reveal the nature of the blobs .
in the next few years , a number of optical integral field unit spectrographs will come online with the special purpose of detecting diffuse emission from the igm at @xmath6 .
the cosmic web imager ( cwi , @xcite ) started operating on the 200 inch telescope at palomar in 2009 , while the multi unit spectroscopic explorer ( muse , @xcite ) for the very large telescope ( vlt ) and the keck cosmic web imager ( kcwi , @xcite ) will begin operations in 2012/2013 .
these will hopefully be followed by the proposed antarctic cosmic web imager ( acwi , @xcite ) .
in addition , balloon experiments in the near uv , such as the faint intergalactic redshifted emission balloon ( fireball , @xcite ) , are set to explore the universe at lower redshifts .
these instruments have the potential to revolutionise our understanding of galaxy formation . the ability to map the surface brightness and velocity profiles of the circumgalactic gas in one or more emission lines would provide invaluable information about two crucial , but poorly understood processes : gas accretion onto galaxies and galactic winds driven by feedback from star formation and/or accreting supermassive black holes . in this paper
we investigate the rest - frame uv emission from the igm at redshifts @xmath8 , and explore the possibility of detecting it with ground - based , optical telescopes .
we will focus on metal - line emission , but will provide some estimates for ly@xmath1 as a point of reference . using a subset of simulations from the overwhelmingly large simulations project ( owls , @xcite ) , we will show that upcoming instruments such as muse and kcwi have the potential to detect diffuse emission from multiple transitions arising in the gas around galaxies .
this paper is organised as follows .
section [ method ] briefly describes the owls runs and our method to calculate and map the gas emission .
section [ results ] presents detailed results for emission in the reference ( _ ref _ , hereafter ) model . in section [ physics ]
we compare the predictions of the reference model with results from a selection of other owls runs with different implementations of a number of physical prescriptions . finally , we draw our conclusions in section [ conclusion ] and present convergence tests in the appendix . throughout this paper
we assume a wmap3 @xmath9cdm cosmology @xcite with parameters @xmath10 , @xmath11 , @xmath12 , @xmath13 , and @xmath14 .
the hubble constant is parametrised as @xmath15 @xmath16 km s@xmath3 mpc@xmath3 , with @xmath17 . for reference
, the cosmic mean density corresponds to a hydrogen number density of @xmath18 ^ 3 $ ] .
we adopt a solar abundance of @xmath19 , corresponding to the value obtained using the default abundance set of cloudy @xcite .
this abundance set , listed in table [ table_abund ] , combines the abundances of @xcite , @xcite and @xcite and assumes @xmath20 .
.adopted solar abundances , from @xcite , @xcite and @xcite .
[ cols="<,^,<,^",options="header " , ] [ exper ] a number of optical instruments will come online within the next few years with the special purpose of detecting diffuse emission from the igm at @xmath6 . in 2009 cwi
@xcite started operating on the 200 inch telescope at palomar .
muse @xcite will be commissioned at the vlt in 2012 and kcwi @xcite will hopefully be installed at keck soon after that . in the more distant future these instruments may be followed by the proposed acwi @xcite , which would operate from antarctica . in addition , the exploration of the low - redshift universe has begun with the near uv balloon experiment fireball @xcite .
the main characteristics of cwi , muse , kcwi and fireball are listed in table [ exper ] .
fireball only covers a narrow wavelength range in the uv , @xmath21 , but has a large fov of 160 @xmath22 160 ( angular resolution 10 ) .
cwi has a 60@xmath23 fov sampled by 24 2.5@xmath23 reflective slits .
although the wavelength range is large @xcite . ] , the instantaneous bandwidth is limited to 150 .
the kcwi will have various modes .
the highest spectral resolution mode , @xmath24 , will only be available with a small fov of 20@xmath25
( angular resolution 0.35 ) and for a narrow , instantaneous band pass of 300 . on the other hand , at the coarsest spectral resolution , @xmath26
, the fov is 20@xmath27 ( angular resolution of 1.4 ) and the band width 2500 .
there will be a number of intermediate settings as well . while muse has a much bigger fov ( 60@xmath22 60 ) , a higher angular resolution ( 0.2 , note that muse will make use of adaptive optics ) and a larger instantaneous band width ( @xmath28 ) , its maximum resolving power is limited to @xmath29 and it will not cover the blue . for a spectrally unresolved emission line , km s@xmath3 .
using a slice thickness that matches the spectral resolution does not change our conclusions for metal lines . for h
ly@xmath1 the situation may , however , be different because resonant scattering , which our simulations do not model , may strongly broaden the lines . ] an 80 hour observation with muse will reach a limiting surface brightness of @xmath30erg s@xmath3@xmath4arcsec@xmath31 for a s / n of five and a 1 arcsec feature at @xmath32 .
we expect kcwi to reach similar depths . using equation ( [ eq : unitconv ] )
, we see that muse and kcwi should be able to detect intensities of @xmath33photon s@xmath3@xmath4sr@xmath3 from the ground for 1 features and a s / n of five . for more extended emission fainter levels
could be reached , of course all provided that the background can be subtracted to the required accuracy . as some of the emission will indeed be extended ( see fig .
[ angle_figure ] ) , we thus expect deep observations with upcoming instruments to be able to directly detect diffuse emission down to @xmath34 photon s@xmath3@xmath4sr@xmath3 .
stacking galaxies may allow fainter emission to be detected statistically .
indeed , @xcite have already detected diffuse ly@xmath1 emission down to @xmath35erg s@xmath3@xmath4arcsec@xmath31 by stacking @xmath36 star - forming galaxies .
table [ eltable ] lists the minimum and maximum redshifts for which the various lines that we investigate fall within the range of muse and kcwi , respectively .
the redshift range for cwi is only slightly smaller than that of kcwi ( see table [ exper ] ) .
according to our predictions , a number of emission lines should be detectable by upcoming instruments . as a benchmark
, we use the instrument sensitivity of 200 photon s@xmath3@xmath4sr@xmath3 , reported by @xcite for cwi for 10 arcsec features .
we note , however , that @xcite quote a much worse limiting sensitivity of @xmath37erg s@xmath3@xmath4arcsec@xmath31 for features of 600 arcsec@xmath38 and a 16 hour integration , which corresponds to @xmath39photon s@xmath3@xmath4sr@xmath3 .
they note , though , that the sensitivity can be improved by over an order of magnitude if the background subtraction is improved using the nod - and - shuffle technique . in any case , 200 photon s@xmath3@xmath4sr@xmath3 will for example be reachable with muse for @xmath40 , spectroscopically unresolved features ( at 5@xmath41 in an 80 hour integration ) . h ly@xmath1 emission from regions with @xmath42 should be detectable up to at least @xmath43 ( taking our optically thin calculations at face value ) , with the glow of less dense regions ( @xmath44 ) visible only at lower redshifts .
the he h@xmath1 line may be visible in dense ( but still @xmath45 ) gas at the lowest redshift we consider ( @xmath46 ) .
c , o and si will all be detectable up to @xmath47 , while c and si should be detectable up to @xmath48 or higher .
the n and ne lines should also be within the capabilities of muse and kcwi and with some luck they might even be able to detect weak o and o emission at @xmath49 . in practice , however , detection of o , o and ne , which have rest - frame wavelengths shorter than the hydrogen lyman limit ( 912 ) , may not be feasible due to the high optical depth provided by intervening absorption systems
. the same may be true for c and si , which have rest - frame wavelengths of 977.0 and 1206.5 and are thus susceptible to intervening ly@xmath1 absorption . as we noted earlier , statistical detections , e.g. by stacking h ly@xmath1 emitters , should enable the detection of fainter lines .
recall that since our focus is the igm , we excluded all gas with densities @xmath50 .
as the emissivity scales as density squared , this high - density gas may be more easily detected , particularly for instruments with high angular resolution such as muse and kcwi .
on the other hand , for observations with instruments with relatively poor angular resolution , such as cwi , it may be difficult to tell whether apparently diffuse emission is in fact due to a few unresolved , dense knots , thus potentially mistaking emission from the ism of small galaxies for emission from diffuse , intergalactic gas .
it will be necessary to reach fluxes well below the optical background light ( obl , hereafter ) to detect emission from the diffuse igm .
most of the obl luminosity is due to zodiacal light , with only relatively minor contributions from the extragalactic background light and the diffuse galactic emission .
the intensity of the obl has been measured to be @xmath51 , @xmath52 , and @xmath53 erg s@xmath3@xmath4sr@xmath3@xmath3 at 3000 , 5500 , and 8000 , respectively ( @xcite ; see also @xcite and references therein ) .
assuming a resolving power of @xmath54 , these correspond to @xmath55 , @xmath56 and @xmath56 photon s@xmath3@xmath4sr@xmath3 at 3500 , 5500 and 8000 , respectively .
as discussed in paper ii , the fireball balloon experiment @xcite provides a good chance to detect igm emission in the uv band at lower redshifts and shorter wavelengths than cwi .
the instrument targets the wavelength window 1970 - 2130 , where atmospheric absorption is minimal .
the surface brightness limit is about 2000 photon s@xmath3@xmath4sr@xmath3 for an angular resolution of 10 " .
the instrument aims to detect h ly@xmath1 , c and o emission at @xmath57 . according to our predictions , in this energy range it might be possible to detect the o line at 630 at @xmath58 , if intervening absorption is not very strong . for this emission line at this redshift
we predict a maximum intensity @xmath59 photon s@xmath3@xmath4sr@xmath3 , which might be within reach of the instrument capabilities .
fireball might also be able to detect c emission at @xmath60 , whose surface brightness should be about an order of magnitude larger than its detection limit .
unfortunately , both the o and the c lines are singlets and might thus be difficult to identify in spectra with limited redshift and wavelength coverage .
as such , it is likely that both lines will simply be treated as unknown contaminants .
however , cross - correlation with stronger lines , such as h ly@xmath1 , may allow a statistical detection of singlets .
a positive detection would provide proof that c and o , and lines from beryllium - like atoms in general , are indeed an efficient cooling channel and would encourage further studies of emission in the far uv band . finally , we caution the reader that the discussion presented in this section should only be taken as a rough guide to what may be possible with upcoming facilities .
more robust conclusions would require a proper study of detectability , which would involve analysing virtual observations targeted to specific instruments , instrumental modes , and observing strategies and using realistic analysis techniques .
while such a study is clearly beyond the scope of the present work , the results presented here suggest that such investigations are worthwhile for several transitions and redshifts . in this section
we compare the intensity of pairs of lines .
results for a subsample of emission lines at @xmath5 are presented in fig .
[ flux ] for the _ ref _ model .
the chosen lines are representative of all interesting transitions . fig .
[ flux ] focuses only on the high surface brightness end of the distribution , with @xmath61 photon s@xmath3@xmath4sr@xmath3 .
as we will discuss further in the next sections , the intensity of the emission lines depends on the gas temperature , density and metallicity . at a fixed temperature ,
sufficiently high for collisional ionisation to dominate over photo - ionisation , the relative intensity of one line with respect to another is independent of the density and proportional to the ratio of the elemental abundances .
since the shape and values of the emissivity curves as a function of temperature are characteristic for each line , the line ratios vary line by line .
this is seen in fig .
the lines of different ions of the same element are strongly correlated , but the line ratios of ions from different elements show a larger scatter .
this is because the gas that produces strong emission in one line is often different from gas that produces emission in a different line .
indeed , the scatter is largest for ratios between lines whose emissivities peak at very different temperatures .
h ly@xmath1 emission is almost always stronger than any other line ( but note that this changes if we ignore h ly@xmath1 from self - shielded gas ; see e.g. fig .
[ lines_pdf ] ) , with the exception of a few pixels in which carbon and oxygen lines dominate .
these likely trace highly metal enriched gas in the haloes of galaxies and groups .
the ratio between the intensity of pairs of emission lines ultimately provides information about the ionisation state and the temperature of the gas .
the intensity of a line is highest when the gas has a temperature close to the peak temperature of its emissivity curve .
a higher value of the line ratio indicates that the gas temperature is closer to the peak temperature of the line in the numerator . .
results are for the _ ref _ simulation .
the contours are spaced logarithmically by 1.5 dex .
the vertical line indicates the threshold at @xmath62 @xmath63 above which we impose an effective equation of state for star - forming gas @xcite .
the horizontal lines define the range of temperatures of whim gas , @xmath64k @xmath65k . ] in this section we investigate the physical properties of the gas that produces most of the emission .
in particular , we want to understand what the typical densities and temperatures are of the gas that produces the bulk of the emission for each line .
this will for example tell us whether the gas is bound to galaxies , resides in groups , or traces lower density regions like filaments and voids .
[ twomass ] shows the mass - weighted ( left panel ) and the metal mass - weighted ( right panel ) gas distributions in the temperature - density plane at @xmath5 .
most of the mass resides in intergalactic gas with low density and temperature .
compared with the gas mass distribution , the distribution of metals in the igm is biased towards higher densities and temperatures . as a consequence , the bulk of the metals do not trace the bulk of the mass .
most of the metals are concentrated in two separate areas of the distribution : i ) in the dense ( @xmath66 @xmath63 ) star - forming gas , i.e. the ism , for which we impose an effective equation of state and which we exclude from the analysis , and ii ) in warm - hot gas in overdense regions with @xmath67 @xmath63 @xmath68 @xmath63 .
this result is qualitatively similar to the findings of paper i for @xmath69 ( see also @xcite ) .
[ denstemp ] shows the emission - weighted gas distributions in the temperature - density plane at @xmath5 . from left to right , the panels show the distributions weighted by h ly@xmath1 , h ly@xmath1ss , he h@xmath1 , c and c ( upper panels ) , n , o , ne , si and si ( bottom panels ) . + a comparison of figs .
[ twomass ] and [ denstemp ] demonstrates that the bulk of the emission from metal lines traces neither the igm , nor the metal mass .
this is consistent with what we found in papers i and ii at lower redshifts for x - ray and uv emission lines , respectively .
instead , each metal line traces a special region of the temperature - density plane centered on the temperature for which the line emissivity peaks ( see fig .
[ z2uv ] ) .
the width of the distribution in temperature corresponds roughly to the width of the emissivity curve as a function of temperature . in terms of density and metallicity , all the emission - weighted distributions peak where the density is high , but also where both the metallicity and the density are high at the same time
the situation is different for the hydrogen and helium lines , which on average trace the gas mass well , once the quadratic dependence of the emissivity on the density is taken into account .
there are some notable differences between the different emission lines .
carbon and silicon lines trace gas with lower temperatures ( @xmath70k ) than other lines .
o traces diffuse gas with temperatures in the range @xmath64k @xmath71k .
n lines arise from gas with temperatures that fall between those of carbon and oxygen lines , while the ne lines trace hotter gas than any other line we consider here , with @xmath72k .
[ physcut ] and [ physcut2 ] show explicitly how the surface brightness pdf varies with the gas density ( fig .
[ physcut ] ) and temperature ( fig .
[ physcut2 ] ) for h ly@xmath1 ( left columns ) , c ( middle columns ) , and o ( right columns ) .
we created these figures from emission maps that were computed by only including particles with the physical properties indicated in the legends .
while h ly@xmath1 is highly biased towards high - density gas , with the highest surface brightness due to gas with overdensities @xmath73 , the brightest metal - line emission arises from gas with a wide range of overdensities , though still @xmath74 .
the gas responsible for the strongest c and o emission has temperatures that are close to the peaks of the emissivity curves shown in fig .
[ z2uv ] , @xmath75k and @xmath76k , respectively .
in this section we discuss the influence of different physical processes on the results .
we focus our analysis at @xmath5 , but all our conclusions are valid for results at higher redshifts .
we consider five different sets of owls runs , in which the physical modules have been changed one at a time .
the models are as follows ( see @xcite for more detailed descriptions ) : 1 .
_ ref _ : our reference model , as described in section [ owls ] .
all results presented in the preceding sections were obtained from this simulation .
this model uses the star formation prescription of @xcite , the supernova - driven wind model of @xcite with initial wind mass loading @xmath77 and initial wind velocity @xmath78 km s@xmath3 , the metal - dependent cooling rates of @xcite , the chemical evolution model described in @xcite and the wmap3 cosmology .
agn feedback is not included .
nosn_nozcool _ : as _ ref _ , but without supernova feedback and with radiative cooling rates that assume primordial abundances .
the metals produced by stars are thus only transported by gas mixing and no energy is transferred to the igm and ism when sne explode .
_ nozcool _ : as _ ref _ , but with radiative cooling rates calculated assuming primordial abundances . neglecting the contributions of metal lines implies slower gas cooling than in the _ ref
_ model .
wvcirc _ : as _ ref _ , but with `` momentum driven '' galactic winds following the prescription of @xcite , with the difference that , as in the _ ref _ model , winds are `` local '' to the star formation event and are fully hydrodynamically coupled ( see @xcite for a discussion of the importance of these effects ) .
the wind velocity and mass loading factor are functions of the galaxy velocity dispersion @xmath41 : @xmath79 and @xmath80 , with @xmath81 km s@xmath3 .
the velocity dispersion @xmath82 is estimated using an on - the - fly friends - of - friends halo finder .
this model was motivated by the idea that galactic winds may be driven by radiation pressure on dust grains rather than sne .
the energy injected into the wind becomes much higher than for _ ref _ for halo masses exceeding @xmath83 .
_ agn _ : as _ ref _ , but with the addition of agn feedback , using the prescription of @xcite .
we chose this set of simulations because these are the models that produced the largest variations in papers i and ii . in these papers , which focused on @xmath69
, we found that for most models the variations in the results are small relative to the _ ref _ model , especially for uv lines . in terms of the bright - end tail of the surface brightness pdf , the largest discrepancies with the _ ref _
results were found for model _
nozcool_. the inclusion of agn feedback could decrease the x - ray emission at the centres of groups and clusters by up to a factor of @xmath84 , but barely affected the emission from diffuse igm gas .
paper ii also showed that varying the feedback parameters that control the intensity of the winds ( e.g. the wind mass loss rate and velocity ) introduced variations in the uv results of a factor of a few or less .
[ four1 ] shows maps of the surface brightness of h ly@xmath1 , c and o lines in the different simulations .
the pdfs of the surface brightness are shown in fig .
[ both_pdf ] .
all maps assume an angular resolution of 2 " and show results at @xmath5 .
most models predict comparable maximum surface brightness values , as can be seen by comparing the high surface brightness tails of the pdfs ( @xmath85 photon s@xmath3@xmath4sr@xmath3 ) , but the variations can be very large for fainter metal - line emission . for h
ly@xmath1 , on the other hand , the differences are always very small . in the _ nosn_nozcool
_ model the emission is concentrated close to galaxy haloes , while the surface brightness of low - density regions is extremely low for metal lines .
this produces the flattening observed in the pdfs of metal lines for @xmath86 photon s@xmath3@xmath4sr@xmath3 and is due to the inefficient transport of metals from galaxies to the diffuse igm in the absence of sn feedback . model _ nozcool _ predicts somewhat stronger emission from metals than the _ ref _ model , but comparable h ly@xmath1 emission . in dense regions ,
the o emission exceeds the predictions of the _ ref _
model by up to an order of magnitude .
this indicates that metal - line cooling is important in the gas for which the o emission peaks , i.e. highly enriched gas with @xmath87k .
note that ignoring metal - line cooling is not self - consistent , as it allows the gas to linger too long at the temperatures for which the emissivity is high . for model _ wvcirc _ the peaks of the metal - line pdfs
are shifted to higher surface brightness values . in other words ,
compared with model _ ref _ , a larger fraction of the sky , and a larger fraction of the mass in the universe , shines bright in the uv .
the outflows in model _ wvcirc _ are more efficient at transporting metals into the igm , mainly as a result of the higher mass loading in low - mass galaxies . as such , the different spatial distribution of the uv emission we see in fig .
[ four1 ] between the _ ref _ and _ wvcirc _ models is mostly due to differences in the metallicity of the gas , and to a lesser extent to differences in the temperature and density . in principle , the detection of diffuse metal - line emission could thus help decide which one of the descriptions of the physics of galactic winds is more `` realistic '' .
a larger spatial extent and a more filamentary structure of the emission on large scales would weigh in favour of highly mass - loaded galactic winds . finally , the predictions of model _ agn _
are comparable to those of the _ ref _
model in regions with low surface brightness and low density .
however , at the bright tail of the distribution model _ agn
_ predicts surface brightnesses that are up to 0.5 dex lower .
this is a consequence of the fact that agn feedback decreases the density and metallicity of the gas near massive galaxies .
the relatively small differences between the metal - line emission predicted by models that include very different physical processes gives us confidence that our predictions for the brighter emission are robust to factors of a few .
on the other hand , the fact that the differences in the predictions for the diffuse metal - line emission are not negligible , implies that observations of this faint glow have the potential to discriminate between different models for feedback from star formation and agn .
finally , we recall that while the different simulations predict very similar h ly@xmath1 emission , the predictions for this line are uncertain due to the importance of self - shielding from ionising radiation , resonant scattering , and absorption by dust , which are all ignored here .
the potential impact is illustrated by the blue dashed curve labelled ` ref ss ' in the left panel of fig .
[ both_pdf ] , which indicates the surface brightness pdf if we ignore all emission from gas with @xmath88 , an assumption that is very conservative for our purposes .
we have investigated the nature and detectability of redshifted rest - frame uv line emission from the igm ( i.e. @xmath89 ) at @xmath8 . while more strongly biased towards high - density gas than absorption lines , emission lines offer the possibility to directly map the distribution of the gas and metals in three dimensions . following the procedures of @xcite
, we used a set of large , cosmological , hydrodynamical simulations taken from the owls project @xcite to create simulated maps of a set of strong rest - frame uv emission lines .
we focused on c , c , n , o , ne , si and si lines ( see table [ eltable ] ) , which are the strongest metal lines that trace the diffuse igm and that can potentially be detected by current and upcoming optical instruments . for reference , we also showed results for h ly@xmath1 and he h@xmath1 and for the former we illustrated the potential impact of self - shielding and dust extinction by considering only the flux emitted by gas with @xmath90 , which provides a conservative estimate of the minimum signal that can be expected .
our main results can be summarised as follows : 1 .
depending on the line and the redshift , we predict a maximum surface brightnesses of order @xmath91 photon s@xmath3@xmath4sr@xmath3 in the redshift interval @xmath0 for 2 pixels ( see fig .
[ redshift_pdf ] ) .
the strongest igm emission comes from h ly@xmath1 , provided we include emission from gas with @xmath92 .
the highest surface brightness metal - line emission comes from c and is about an order of magnitude fainter .
the highest surface brightness regions for c , si , si and o lines are fainter than the brightest c by factors of a few .
the n and ne lines , as well as he h@xmath1 , are substantially fainter but their maximum surface brightnesses still exceed @xmath2 photon s@xmath3@xmath4sr@xmath3 for @xmath5 ( and 2 pixels ) .
2 . lines from different ions trace gas in different temperature regimes .
the gas responsible for the highest surface brightness metal - line emission typically has temperatures close to the peak of the corresponding emissivity curve .
h ly@xmath1 , he ly@xmath1 , c , si and si lines prevalently trace gas with @xmath93k near galaxies , making them excellent probes of cold accretion flows and the colder parts of galactic winds
. c ( @xmath94k ) , n ( @xmath95k ) , o ( @xmath96k ) , and ne ( @xmath97k ) trace the warmer gas in haloes and filaments and thus constitute powerful tools to study the whim and the warmer parts of outflows .
while the brightest h ly@xmath1 emission arises exclusively in highly overdense gas ( @xmath73 ) , at least if we assume the gas to be optically thin , a much wider range of overdensities contributes to the brightest pixels of he ly@xmath1 and high - ionisation metal lines such as c , o and ne .
the spatial distribution of the emission is related to the ionisation state of the emitting gas .
lower ionisation gas produces clumpier and more clustered emission , while higher ionisation gas produces smoother and more diffuse emission profiles .
the degree of ionisation decreases with increasing density but increases with the temperature .
as the emissivity increases with the gas density , emission from lower - ionisation gas tends to be brighter .
indeed , we find for example that the beryllium - like atoms n and o ( which were not shown here because they only shift into the optical waveband for very high redshifts ) are stronger than their higher - ionisation lithium - like counterparts .
we have compared our fiducial simulation with models that ignore sn feedback and/or metal - line cooling , with a model that includes highly mass - loaded `` momentum - driven '' galactic winds , and with a simulation that takes feedback from agn into account . while the faint metal - line emission is very sensitive to the inclusion and implementation of galactic winds ( because these are responsible for enriching the low - density igm with metals , where the weak emission is produced ) , the results for bright metal - line emission , and even more so for both faint and bright h ly@xmath1 emission , are remarkably robust to changes in these physical processes .
accurate cooling rates are required to predict the bright metal - line emission with a precision better than a factor of a few , particularly for lines that peak at temperatures for which the cooling time is short , such as o. agn feedback primarily affects the emission from the cores of haloes , reducing the brightest metal - line emission by up to factors of a few .
highly mass - loaded galactic winds , on the other hand , can boost the brightest metal - line emission by similar factors .
5 . according to our predictions
, muse and kcwi should be able to detect h ly@xmath1 emission from regions with @xmath42 up to at least @xmath43 ( provided our optically thin calculation is appropriate in this regime ) , while h ly@xmath1 from less dense regions and the he h@xmath1 line should be visible at lower redshifts .
c , o and si will all be detectable up to @xmath7 , while c and si should be detectable up to @xmath48 .
the n and ne lines should also be within the capabilities of these instruments .
we note , however , that lines with rest - frame wavelengths shorter than that of ly@xmath1 ( i.e. c , o , ne , and si ) may suffer substantial foreground absorption , which we have not accounted for here .
o lines at @xmath58 and c lines at @xmath60 are roughly within the sensitivity of the fireball experiment , and should be considered as possible contaminants of ly@xmath1 emission .
+ we conclude that several rest - frame uv emission lines from the high - redshift igm will become detectable in the near future , possibly starting with the cwi which is already operating on palomar .
observations of these lines have the potential to revolutionise studies of the interactions between high - redshift galaxies and their environment .
lower ionisation lines provide us with tools to image cold accretion flows as well as cold , outflowing clouds .
higher ionisation lines , on the other hand , will allow us to image the metal - enriched whim and galactic winds .
such observations , particularly the simultaneous detection of multiple lines , will provide strong constraints on the density , temperature , and chemical composition of the gas .
we are grateful to all members of the owls team for help during the development of this project and to tom theuns and freeke van de voort also for a careful reading of the manuscript .
we would like to thank xavier prochaska and christopher martin for stimulating discussions and we are grateful to the anonymous referee for a helpful report .
the simulations presented here were run on stella , the lofar bluegene / l system in groningen , on the cosmology machine at the institute for computational cosmology in durham as part of the virgo consortium research programme , and on darwin in cambridge .
this work was sponsored by the dutch national computing facilities foundation ( ncf ) for the use of supercomputer facilities , with financial support from the netherlands organization for scientific research ( nwo ) .
sb acknowledges support by nsf grants ast-0507117 and ast-0908910 .
this work was supported by an nwo vidi grant and by the marie curie initial training network cosmocomp ( pitn - ga-2009 - 238356 ) .
99 adelberger k.l . ,
steidel c.c . ,
kollmeier j.a . ,
reddy n.a . , 2006 ,
apj , 637 , 74 allen p.d . ,
moustakas l.a . , dalton g. , macdonald e. , blake c. , clewley l. , heymans c. , wegner g. , 2005 , mnras , 360 , 1244 allende prieto c. , lambert d.l . , asplund m. , 2001 , apjl , 556 , 63 allende prieto c. , lambert d.l . , asplund m. , 2002 , apjl , 573 , 137 altay g. , theuns t. , schaye j. , crighton n. h. m. , dalla vecchia c. , 2011 , apj , 737 , l37 bacon r. , bauer s. , brau - nogu s. , et al .
, 2009 , in `` science with the vlt in the elt era '' , ed .
moorwood a. , 331 bernstein , r.a . , 2007 , apj , 666 , 663 bernstein r.a . , freedman w.l . ,
madore b.f . , 2002 , apj , 571 , 56 bertone s. , stoehr f. , white s.d.m . , 2005 , mnras , 359 , 1201 bertone s. , white s.d.m . , 2006 , mnras , 367 , 247 bertone s. , schaye j. , dolag k. , 2008 , ssr , 134 , 295 bertone s. , schaye j. , dalla vecchia c. , booth c.m . , theuns t. , wiersma r.p.c . ,
2010a , mnras , 407 , 544 ( paper i ) bertone s. , schaye j. , booth c.m .
, dalla vecchia c.,theuns t. , wiersma r.p.c .
, 2010b , mnras , 408 , 1120 ( paper ii ) blanton m.r .
, eisenstein d. , hogg d.w .
, zehavi i. , 2006 , apj , 645 , 977 booth c.m . ,
schaye j. , 2009 , mnras , 398 , 53 bower r. g. , morris s.l . , bacon r. , et al . , 2004 , mnras , 351 , 63 cantalupo s. , porciani c. , lilly s.j . , miniati f. , 2005 , apj , 628 , 61 cantalupo s. , lilly s. j. , porciani c. , 2007 , apj , 657 , 135 cantalupo s. 2010 , mnras , 403 , l16 cen r. , & fang t. , 2006 , apj , 650 , 573 cen r. , chisari n. e. , 2010 , eprint arxiv:1005.1451 chabrier g. , 2003 , pasp , 115 , 763 dalla vecchia c. , & schaye j. , 2008 , mnras , 387 , 1431 dijkstra m. , loeb a. , 2009 , mnras , 400 , 1109 dijkstra m. , haiman z. , spaans m. , 2006 , apj , 649 , 37 fardal m.a . , katz n. , gardner j.p . , hernquist l. , weinberg d.h . , dav r. , 2001 , apj , 562 , 605 faucher - gigure c. , keres d. , dijkstra m. , hernquist l. , zaldarriaga m. , 2010 , apj , 725 , 633 ferland g.j .
, korista k.t . ,
verner d.a .
, ferguson j.w . , kingdon j.b .
, verner e.m . , 1998 , pasp , 110 , 761 francis p. j. , bland - hawthorn j. , 2004 , mnras , 353 , 301 francis p. j. , mcdonnell s. , 2006 , mnras , 370 , 1372 fukugita m. , & peebles p.j.e .
, 2004 , apj , 616 , 643 furlanetto s.r .
, schaye j. , springel v. , & hernquist l. , 2003 , apjl , 599 , 1 furlanetto s.r . , schaye j. , springel v. , & hernquist l. , 2004 , apj , 606 , 221 furlanetto s.r . , schaye j. , springel v. , & hernquist l. , 2005 , apj , 622 , 7 geach j.e . ,
alexander d.m . ,
lehmer b.d . , et al . , 2009 , apj , 700 , 1 gnat o. , sternberg a. , 2007 , apjs , 168 , 213 gnat o. , sternberg a. , 2009 , apj , 693 , 1514 goerdt t. , dekel a. , sternberg a. , ceverino d. , teyssier r. , primack j. r. , 2010 , mnras , 407 , 613 gould a. , weinberg d. h. , 1996 , apj , 468 , 462 haardt f. , madau p. , 2001
, in neumann d.m . , tran j.t.v
. , eds . , clusters of galaxies and the high redshift universe observed in x - rays .
hogan c. j. , weymann r. j. , 1987 , mnras , 225 , 1p holweger h. , 2001 , in wimmer - schweingruber r.f . , ed . ,
joint soho / ace workshop solar and galactic composition vol .
598 of american institute of physics conference series , photospheric abundances : problems , updates , implications , 23 kennicutt jr .
, 1998 , apj , 498 , 541 kollmeier j.a .
, zheng z. , dav r. , gould a. , katz n. , miralda - escud j. , weinberg d.h .
, 2010 , apj , 708 , 1048 komatsu e. , dunkley j. , nolta m.r . ,
2009 , apjs , 180 , 330 lodders k. , 2003 , apj , 591 , 1220 martin c. , bertone s. , bryan g. , cen r. , prochaska j.x . , schaye j. , schiminovich d. , steidel c. , 2009 , the astronomy and astrophysics decadal survey , science white papers , no
. 195 matsuda y. , yamada , t. , hayashino t. , et al .
, 2004 , aj , 128 , 569 matsuda y. , yamada t. , hayashino t. , yamauchi r. , nakamura y. , 2006 , apjl , 640 , 123 matuszewski m. , chang d. , crabill r. m. , martin d. c. , moore a. m. , morrissey p. , rahman s. , 2010 , spie , 7735 , moore a.m. , martin c. , maitless n.c . , travouillon t. , 2008 , spie , 7012 , 80 mori m. , umemura m. , ferrara a. , 2004 , apjl , 613 , 97 nilsson k. , fynbo j.p.u . , moller p. , sommer - larsen j. , ledoux c. , 2006 , a&a _ lett _ , 452 , 23 oppenheimer b.d . , dav r. , 2008 , mnras , 387 , 577 ohyama y. , taniguchi y. , kawabata k. s. , et al , 2003 , apjl , 591 , 9 prescott m. k. m. , dey a. , jannuzi b. t. , 2009 , apj , 702 , 554 rahman s. , martin c. , mclean r. , matuszewski m. , chang d. , 2006 , spie , 6269 , 124 rauch m. , haehnelt m. , bunker a. , et al . , 2008 , apj , 681 , 856 saito t. , shimasaku k. , okamura s. , ouchi m. , akiyama m. , yoshida m. , 2006 , apj , 648 , 54 scarlata c. , colbert j. , teplitz h.i .
, et al . , 2009 , apj , 706 , 1241 schaye j. 2001 , apj , 559 , 507 schaye j. 2006 , apj , 643 , 59 schaye j. , dalla vecchia c. , 2008 , mnras , 383 , 1210 schaye j. , dalla vecchia c. , booth c.m .
, et al . , 2010 , mnras , 402 , 1536 smith d.j.b .
, jarvis m.j . , 2007 , mnras _ lett _ , 378 , 49 spergel d.n . , et al . , 2007 , apjs , 170 , 377 springel v. , 2005 , mnras , 364 , 1105 steidel c.c . ,
adelberger k.l . ,
shapley a.e . , pettini m. , dickinson m. , giavalisco m. , 2000 , apj , 532 , 170 steidel c. c. , bogosavljevi m. , shapley a. e. , kollmeier j. a. , reddy n. a. , erb d. k. , pettini m. , 2011 , apj , 736 , 160 tescari e. , viel m. , dodorico v. , cristiani s. , calura f. , borgani s. , tornatore l. , 2011 , mnras , 411 , 826 theuns t. , viel m. , kay s. , schaye j. , carswell r.f .
, tzanavaris p. , 2002 ,
apjl , 578 , 5 tuttle s.e . , schiminovich d. , milliard b. , grange r. , martin c. , rahman s. , deharveng j .- m . , mclean r. , tajiri g. , matuszewski m. , 2008 , spie , 7014 , 61 weidinger m. , mller p. , fynbo j.p.u .
, thomsen b. , 2005 , a&a , 436 , 825 weijmans a .-
, bower r. g. , geach j. e. , swinbank a. m. , wilman r. j. , de zeeuw p. t. , morris s. l. , 2010 , mnras , 402 , 2245 wiersma r.p.c .
, schaye j. , smith b.d .
, 2009a , mnras , 393 , 99 wiersma r.p.c . , schaye j. , theuns t. , dalla vecchia c. , tornatore l. , 2009b , mnras , 399 , 574 yang y. , zabludoff a.i . , dav r. , eisenstein d.j . , pinto p.a . , katz n. , weinberg d.h . , barton e.j . , 2006 , apj , 640 , 539 yang y. , zabludoff a. , tremonti c. , eisenstein d. , dav r. , 2009 , apj , 693 , 1579 yoshida n. , furlanetto s.r . , &
hernquist l. , 2005 , apjl , 618 , l91 yoshikawa k. , & sasaki s. , 2006 , pasj , 58 , 641
in this appendix we test the effect of varying the angular resolution used to make the emission maps . fig .
[ angle_figure ] compares the pdfs of the surface brightness of , from left to right , h ly@xmath1 , c , and o at @xmath5 for maps with resolutions @xmath98 , and @xmath99 , which correspond to physical scales varying from 3.1 @xmath16 kpc to 199 @xmath16 kpc .
the distributions converge well for all but the highest surface brightness values .
the increase in the maximum brightness with decreasing pixel size reflects the fact that large pixel sizes smooth the strong emission coming from structures with an angular size smaller than that used to observe them .
the differences between the pdfs corresponding to different pixel sizes are much smaller for o than for c and h. this is because o emission arises in gas of lower densities , which will have larger coherence lengths ( because the jeans scale increases with decreasing density ; see e.g. @xcite ) .
in this appendix we present a number of tests aimed to verify how resolution affects the numerical predictions .
we consider the effect of varying the numerical resolution in appendix [ number ] , the box size in appendix [ box ] and the slice thickness in appendix [ thick ] .
all tests use the _ ref _ model and results are shown only for the o 1032 line . to test the effect of varying the numerical resolution
, we compare emission maps for three different realisations of the same simulation with particle numbers @xmath100 , @xmath101 and @xmath102 . the corresponding dark matter particle masses are @xmath103 , @xmath104 , and @xmath105 @xmath16 m@xmath106 , respectively .
[ number_figure ] shows the pdfs of the @xmath5 surface brightness for , from left to right , h ly@xmath1 , c , and o for maps with 2 " angular resolution .
all distributions are similar at high and intermediate surface brightnesses .
only the lowest resolution run with @xmath100 deviates strongly from the other two and only for @xmath107 photon s@xmath3@xmath4sr@xmath3 .
we conclude that our conclusions are not significantly affected by the finite resolution of our simulations .
we test the effect of varying the size of the simulated box , while keeping the mass resolution constant .
the runs necessarily assume different initial conditions and thus allow us to test how the large - scale structure affects the flux statistics . in fig .
[ box_figure ] we compare results for h ly@xmath1 , c , and o at @xmath5 using simulations with dark matter particle mass @xmath108 @xmath16 m@xmath106 and box sizes of 25 , 12.5 , and 6.25 @xmath16 mpc ( comoving ) , respectively .
the maps used to calculate the pdf have 2 " angular resolution and 6.25 @xmath16 mpc thickness , consistent with the box size of the smallest simulation .
the results are converged with respect to the size of the simulation volume , although larger boxes sample the flux pdf to higher values . in this section
we describe how varying the thickness of the slices we cut through the simulation box affects the shape of the flux pdf . in fig .
[ thick_pdf ] we compare the @xmath5 surface brightness pdfs of , from left to right , h ly@xmath1 , c , and o for three different , evenly - spaced slice thicknesses : 25 , 5 and 1 @xmath16 mpc . to facilitate the comparison
, we have normalised the surface brightness pdf by the slice thickness @xmath109 . throughout this work
, we have assumed a slice thickness of 25 @xmath16 mpc . except for c ,
the results appear to be independent of the slice thickness ( within limits and in the detectable regime ) when the pdf is normalised by the slice thickness .
in other words , the normalisation of the pdf is proportional to the slice thickness .
this happens when the slice is thick enough to fully contain collapsed objects , but sufficiently small to avoid the superposition of multiple structures , which would strongly increase the flux in those pixels where more than one structure could be found along the line of sight . for c
there is a small difference between 5 and 25 @xmath16 mpc , but not between 1 and 5 @xmath16 mpc , indicating that projection effects slightly enhance the flux for 25 @xmath16 mpc slices .
the shape of the pdf at the faint end , on the other hand , depends on the level of the flux in low - density regions and on the number of low - density structures in each slice , which is dependent on the slice thickness . as a consequence , the fraction of pixels at the lowest fluxes increases for decreasing slice thickness | rest - frame ultraviolet emission lines offer the exciting possibility to directly image the gas around high - redshift galaxies with upcoming optical instruments .
we use a suite of large , hydrodynamical simulations to predict the nature and detectability of emission lines from the intergalactic medium at @xmath0 .
the brightest emission comes from h ly@xmath1 ( 1216 ) and the strongest metal line , c ( 977 ) , is about an order of magnitude fainter , although h ly@xmath1 may be fainter if the gas is self - shielded to the uv background or if dust is important .
the highest surface brightness regions for c ( 1548,1551 ) , si ( 1207 ) , si ( 1393,1403 ) and o ( 1032,1038 ) are fainter than the brightest c by factors of a few .
the n ( 1239,1243 ) and ne ( 770,780 ) lines , as well as he h@xmath1 ( 1640 ) , are substantially weaker but their maximum surface brightnesses still exceed @xmath2 photon s@xmath3@xmath4sr@xmath3 at @xmath5 ( for 2 pixels ) .
lower ionisation lines typically arise in denser and colder gas that produces clumpier emission .
the brightest h ly@xmath1 emission arises exclusively in highly overdense gas , but the highest surface brightness emission from high - ionisation metal lines traces a much wider range of overdensities .
bright metal - line emission traces gas with temperatures close to the peak of the corresponding emissivity curve .
while h ly@xmath1 , he h@xmath1 , c , si , and si are excellent probes of cold accretion flows and the colder parts of outflows , c , n , o , and ne are powerful tracers of the diffuse warm - hot intergalactic medium and galactic winds
. a comparison of results from simulations with varying physical prescriptions demonstrates that the predictions for the brighter metal - line emission are robust to within factors of a few .
several rest - frame uv emission lines from the high - redshift igm will become detectable in the near future , possibly starting with the cosmic web imager , which is already operating on palomar .
the multi unit spectroscopic explorer , which will be commissioned in 2012 on the very large telescope , and the proposed keck cosmic web imager have the potential to revolutionise studies of the interactions between high - redshift galaxies and their environment . [ firstpage ] intergalactic medium diffuse radiation
radiation mechanisms : thermal galaxies : formation cosmology : theory |
mesh patterns were introduced in @xcite by brndn and claesson , and they were studied in a series of papers ( e.g. see @xcite by kitaev and liese , and references therein ) .
a particular class of mesh patterns is _ boxed patterns _ introduced in @xcite by avgustinovich et al .
, who later suggested to call this type of patterns _ frame patterns_. the simplest frame pattern which is called the @xmath0 pattern is defined as follows .
let @xmath12 denote the set of all permutations of @xmath23 .
given @xmath24 , we say that a pair @xmath25 is an occurrence of the @xmath0 pattern in @xmath26 if @xmath27 , @xmath28 , and there is no @xmath8 such that @xmath29 ( the last condition is indicated by the shaded area in figure [ fig:48261357 ] ) .
the @xmath0 pattern is shown in figure [ fig : meshpattern ] using the notation in @xcite .
.,scaledwidth=15.0% ] similarly , we say that the pair @xmath25 is an occurrence of the @xmath30 pattern in @xmath26 if @xmath27 , @xmath31 , and there is no @xmath8 such that @xmath32 .
for example , if @xmath33 , then the occurrences of @xmath0 in @xmath26 are @xmath34 and the occurrences of @xmath30 in @xmath26 are @xmath35 we let @xmath36 ( resp . , @xmath37 ) denote the number of occurrences of the @xmath0 ( resp . , @xmath30 ) in @xmath26 .
the _ reverse _ of @xmath38 , @xmath39 , is the permutation @xmath40 , and the _ complement _ of @xmath26 , @xmath41 , is the permutation @xmath42 .
it is easy to see that @xmath43 and thus , since the reverse and complement are trivial bijections from @xmath12 to itself , studying the distribution of @xmath0-matches in @xmath12 is equivalent to studying the distribution of @xmath30=mathches in @xmath12 .
highlighted.,scaledwidth=20.0% ] if we graph a given permutation @xmath26 as dots on a grid as in figure [ fig:48261357 ] , then one can see that each occurrence of @xmath0 is a pair of increasing dots such that there are no dots within the rectangle created by the original dots .
the occurrence @xmath44 is highlighted in figure [ fig:48261357 ] ; in particular , there are no dots within the shaded rectangle .
jones and remmel studied the distribution of cycle - occurrences of classical consecutive patterns in @xcite .
see @xcite for a comprehensive introduction to the theory of permutation patterns . in this paper
, we shall study the distribution of the _ cycle - occurrences _ of @xmath0 in the cycle structure of permutations in the symmetric group @xmath12 .
that is , suppose that we are given a @xmath7-cycle @xmath2 in a permutation @xmath26 in the symmetric group @xmath12 .
then we will always write @xmath45 where @xmath46 is the smallest element of the cycle
. we will always draw such a cycle with @xmath46 at the top and assume that we traverse around the cycle in a clockwise direction .
for example , if @xmath47 , then we would picture @xmath2 as in figure [ fig : cycle ] .
we say that @xmath48 is _ cyclically between _ @xmath49 and @xmath50 in @xmath2 , if starting at @xmath49 , we encounter @xmath48 before we encounter @xmath50 as we traverse around the cycle in a clockwise direction .
alternatively , we can identify @xmath2 with the permutation @xmath51 . in this notation
, @xmath48 is cyclically between @xmath49 and @xmath50 if either @xmath52 , or @xmath53 , or @xmath54 .
for example , in the cycle @xmath47 , 8 is cyclically between 2 and 3 .
then we say that the pair @xmath55 is a _ cycle - occurrence of @xmath0 in @xmath2 _ if @xmath56 and there is no @xmath48 such that @xmath57 and @xmath48 is cyclically between @xmath49 and @xmath50 .
similarly , we say that the pair @xmath55 is a cycle - occurrence of @xmath30 in @xmath2 , if @xmath58 and there is no @xmath48 such that @xmath59 and @xmath48 is cyclically between @xmath49 and @xmath50 . as is the case with permutations ,
the study of the number of cycle - occurrences of @xmath0 in the cycle structures of permutations is equivalent to the study of the number of cycle - occurrences of @xmath0 in the cycle structures of permutations .
that is , given the cycle structure @xmath60 of a permutation @xmath61 , the cycle complement of @xmath26 , @xmath62 , is the permutation whose cycle structure arises from the cycle structure of @xmath26 by replacing each number @xmath5 by @xmath63 .
for example , if @xmath26 consists of the cycle @xmath64 , then @xmath62 consists of the cycles @xmath65 .
it is then easy to see that for all @xmath61 , @xmath26 has @xmath7 cycle - occurrences of @xmath0 if and only if @xmath62 has @xmath7 cycle - occurrences of @xmath30 .
let @xmath66 be the set of @xmath1-cycles in @xmath12 .
if @xmath67 , then it is clear that @xmath68 is always a cycle - occurrence of @xmath0 in @xmath2 .
we shall call such cycle - occurrences _ trivial occurrences of @xmath0 _ or _ trivial @xmath0-matches _ in @xmath2 and all other cycle - occurrences of @xmath0 in @xmath2 will be called _ nontrivial occurrences of @xmath0 _ or _
nontrivial @xmath0-matches _ in @xmath2 .
we let @xmath69 denote the number of occurrences of @xmath0 in @xmath2 and @xmath70 denote the number of nontrivial occurrences of @xmath0 in @xmath2 .
for example , if @xmath47 , the nontrivial occurrences of @xmath0 in @xmath2 are the pairs @xmath71 , @xmath72 , @xmath73 , @xmath74 , and @xmath75 , so that @xmath76 . clearly , if @xmath77 is an @xmath1-cycle in @xmath66 , then @xmath78 , since for all @xmath79 , @xmath80 will be a trivial occurrence of @xmath0 in @xmath2 . if @xmath2 is a 1-cycle , then @xmath81 .
.,scaledwidth=20.0% ] if @xmath82 is any sequence of distinct integers , we let @xmath83 denote the permutation of @xmath12 that is obtained by replacing the @xmath5th largest element of @xmath84 by @xmath5 for @xmath85 . for example , @xmath86 . similarly ,
if @xmath45 is a @xmath7-cycle in some permutation @xmath61 , we let @xmath87 be the @xmath7-cycle in @xmath88 which is obtained by replacing the @xmath5th largest element of @xmath2 by @xmath5 . for example , if @xmath89 , then @xmath90 . in such a situation
, we let @xmath91 .
finally , if @xmath26 consists of cycles @xmath92 , then we let @xmath93 .
we note that if one wishes to study the generating function @xmath94 where @xmath95 denotes the number of cycles in @xmath26 , then it is enough to study the generating function @xmath96 where @xmath97 .
that is , it easily follows from the exponential formula that if @xmath98 is a weight function on @xmath1-cycles @xmath99 and for any permutation @xmath26 whose cycle structure is @xmath60 , we define @xmath100 , then @xmath101 hence @xmath102 the main focus of this paper is to study the polynomials @xmath103 . in table [ tab : incont0 ] we provide polynomials @xmath103 for @xmath104 that were calculated using mathematica .
c||*2l @xmath105 & @xmath106 + @xmath107 & @xmath106 + @xmath108 & @xmath109 + @xmath110 & @xmath111 + @xmath112 & @xmath113 + @xmath114 & @xmath115 + @xmath116 & @xmath117 + & @xmath118 + @xmath119 & @xmath120 + & @xmath121 + & @xmath122 + @xmath123 & @xmath124 + & @xmath125 + & @xmath126 + & @xmath127 + @xmath128 & @xmath129 + & @xmath130 + & @xmath131 + & @xmath132 + & @xmath133 + & @xmath134 + given a polynomial @xmath135 , we let @xmath136 denote the coefficient of @xmath137 in @xmath138 .
it is easy to explain several of the coefficients that appear in @xmath103 .
for example , for @xmath139 , @xmath140 because @xmath141 is the only @xmath1-cycle that has no nontrivial occurrences of @xmath0 .
similarly , @xmath142 since the only @xmath1-cycle with @xmath143 nontrivial occurrences of @xmath0 is @xmath144.[maxcycle ] we computed that the sequence @xmath145 starts out @xmath146 which leads to a conjecture that @xmath147 for @xmath148 .
similarly , we computed that the sequence @xmath149 starts out @xmath150 which is the initial terms of sequence a002415 in the _ on - line encyclopedia of integer sequences _ ( _ oeis _ ) whose @xmath1th term is given by the formula @xmath151 .
we will verify this by proving that for @xmath152 , @xmath153 .
in fact , we shall show that for any fixed @xmath154 , there are constants @xmath155 such that @xmath156 for all @xmath157 . to prove this result
, we will need to study what we call incontractible @xmath1-cycles which are @xmath1-cycles @xmath2 for which there are no integers @xmath5 such that @xmath11 immediately follows @xmath5 in @xmath2 .
we let @xmath158 denote the set of incontractible @xmath1-cycles in @xmath66 .
we will show that @xmath159 is @xmath13 where @xmath14 is the number of derangements of @xmath12 , i.e. the number of @xmath61 such that @xmath26 has no fixed points .
we will also study the polynomials @xmath160 we have computed table [ tab : poly ] for the polynomials @xmath17 using mathematica . c||*2l @xmath161 & @xmath106 + @xmath162 & @xmath163 + @xmath164 & @xmath18 + @xmath165 & @xmath166 + @xmath167 & @xmath168 + @xmath169 & @xmath170 + @xmath171 & @xmath172 + & @xmath173 + @xmath174 & @xmath175 + & @xmath176 + @xmath177 & @xmath178 + & @xmath179 + & @xmath180 + @xmath181 & @xmath182 + & @xmath183 + & @xmath184 + & @xmath185 + & @xmath186 + we shall show that @xmath21 and @xmath22 are the number of partitions of @xmath7 for sufficiently large @xmath1 .
we show this by plotting the _ non-@xmath0-matches _ of an @xmath1-cycle @xmath99 , i.e. the pairs of integers @xmath3 with @xmath4 which are not occurrences of the @xmath0 pattern in @xmath2 , on an @xmath187 grid and shading a cell in the @xmath6th row and @xmath5th column if and only if @xmath188 is a non-@xmath0-match of @xmath2 .
for example , the non-@xmath0-matches of the cycle @xmath189 are @xmath190 and @xmath191 .
these pairs can be plotted on the @xmath192 grid as shown in figure [ fig : fdex1 ] .
plotted on an @xmath193 grid , scaledwidth=30.0% ] the shaded plot in figure [ fig : fdex1 ] is of the form of the _ ferrers diagram _ of the integer partition @xmath194 .
we will show that if @xmath99 and there are fewer than @xmath195 non - matches in @xmath2 , then shaded squares in the plot of the non-@xmath0-matches in @xmath2 will be of the form of a ferrers diagram of an integer partition .
moreover , we shall show that if @xmath99 has fewer than @xmath195 non - matches , then @xmath2 must be incontractible .
hence , it follows that @xmath21 and @xmath22 are the number of partitions of @xmath7 for sufficiently large @xmath1 .
we will show that @xmath196 for @xmath157 and that the lowest power of @xmath18 which appears in either @xmath197 or @xmath19 is @xmath198 .
we computed that the sequence @xmath199 starts out @xmath200 which led us to conjecture that @xmath201 where @xmath202 is the @xmath7th _ catalan number_. we shall prove this conjecture and our proof of this conjecture led to a surprising connection between the _ charge statistic on permutations _ as defined by lascoux and schtzenberger @xcite and our problem .
that is , given a permutation @xmath203 , one defines the index , @xmath204 , of @xmath205 in @xmath26 as follows .
first , @xmath206 . then , inductively , @xmath207 if @xmath11 appears to the right of @xmath5 in @xmath26 and @xmath208 if @xmath11 appears to the left @xmath5 in @xmath26 . for example
, if @xmath209 and we use subscripts to indicate the index of @xmath5 in @xmath26 , then we see that the indices associated with @xmath26 are @xmath210 the charge of @xmath26 , @xmath211 , is equal to @xmath212 .
thus , for example , if @xmath209 , then @xmath213 . we will associate a path which we call the _ charge path _ of @xmath26 whose vertices are elements @xmath214 and whose edges are @xmath215 for @xmath216 . for example , if @xmath209 as above , then charge path of @xmath26 , which we denote by @xmath217 , is pictured in figure [ fig : chgr ] . in our particular example , the charge graph of @xmath26 is a _ dyck path _ ( a lattice path with steps ( 1,1 ) and ( 1,-1 ) from the origin ( 0,0 ) to @xmath218 that never goes below the @xmath219-axis ) and if @xmath220 is the @xmath1-cycle induced by @xmath26 , then @xmath221 .
this is no accident .
that is , we shall show that if @xmath222 is @xmath223-cycle in @xmath224 , then @xmath225 if and only if the charge path of @xmath226 is a dyck path of length @xmath227 .
the outline of this paper is as follows . in section [ cont ]
, we shall introduce the notion of the contraction of an @xmath1-cycle and use it to show that for all @xmath228 , there are positive integers @xmath229 such that @xmath230 . in section [ incont ] , we shall study incontractible cycles and the polynomials @xmath17 .
we shall show that for all @xmath228 , @xmath231 where @xmath232 is @xmath1th catalan number .
in section [ intpart ] , we study the non-@xmath0-matches in @xmath1-cycles of @xmath66 and use them to show that @xmath21 and @xmath22 are the number of partitions of @xmath7 for sufficiently large @xmath1 . in section [ concl ] , we will state our conclusions and some open problems .
define a _ bond _ of an @xmath1-cycle @xmath99 to be a pair of consecutive integers @xmath233 such that @xmath234 is immediately followed by @xmath235 in the cycle @xmath2 .
for example , the bonds of the cycle @xmath236 are the pairs @xmath237 , @xmath238 , and @xmath239 .
if @xmath240 is an @xmath1-cylce where @xmath241 , define the sets @xmath242 recursively as follows .
first , let @xmath243 .
then inductively , if @xmath244 , then @xmath245 if @xmath246 , and @xmath247 otherwise .
for example , for the cycle @xmath248 these sets are @xmath249 .
we will call these sets the _ consecutive runs of @xmath2_. define the _ contraction _ of a cycle @xmath2 with consecutive runs @xmath242 to be @xmath250 where @xmath251 denotes the maximum element in a run @xmath252 . for example , the contraction of @xmath248 is @xmath253 we say that @xmath2 _ contracts _ to @xmath254 if @xmath255 . clearly , a cycle @xmath2 is _ incontractible _ if @xmath256 .
we claim that the number of non - trivial @xmath0 matches does not change as we pass from @xmath2 to @xmath257 .
that is , we have the following theorem . [ ntmu ] for any @xmath1-cycle @xmath99 , @xmath258 we proceed by induction on @xmath1 .
the theorem is clearly true for @xmath259 . now
, suppose that we are given an @xmath1-cycle @xmath99 and @xmath260 are the consecutive runs .
if @xmath261 , then , clearly , @xmath2 is incontractible . otherwise , suppose that @xmath262 is the least @xmath5 such that @xmath263 .
let @xmath264 .
then it is easy to see that in @xmath2 , there are no pairs @xmath265 which are nontrivial occurrences of @xmath0 in @xmath2 for @xmath266 since @xmath235 immediately follows @xmath234 in @xmath2 .
similarly , there are no pairs @xmath267 which are nontrivial occurrences of @xmath0 in @xmath2 for @xmath268 since @xmath269 immediately precedes @xmath270 in @xmath2 . now
, suppose that @xmath271 arises from @xmath2 by removing @xmath272 and replacing each @xmath273 by @xmath274 .
then it is easy to see that 1 .
if @xmath275 , @xmath276 is a nontrivial occurrence of @xmath277 in @xmath2 if and only if @xmath276 is a nontrivial occurrence of @xmath277 in @xmath271 , 2 . if @xmath278 , then @xmath276 is a nontrivial occurrence of @xmath277 in @xmath2 if and only if @xmath279 is a nontrivial occurrence of @xmath277 in @xmath271 , and 3 . if @xmath280 , then @xmath276 is a nontrivial occurrence of @xmath277 in @xmath2 if and only if + @xmath281 is a nontrivial occurrence of @xmath277 in @xmath271 .
it follows that @xmath282 .
moreover , it is easy to see that @xmath283 . by induction @xmath284 . hence , @xmath285 .
it is easy to count the number of @xmath1-cycles @xmath99 for which @xmath286 .
that is , we have the following theorem . [
th : incont ] let @xmath287 be an incontractible cycle of length @xmath288 .
the number of @xmath1-cycles @xmath99 such that @xmath255 is @xmath289 . if @xmath255 , then there are @xmath288 consecutive runs of @xmath2 , namely @xmath290 such that @xmath291 .
since the maximum of one of the consecutive runs must be @xmath1 , there are @xmath292 ways to choose the rest of the maxima .
this will determine the consecutive runs and then we order them so that @xmath293 . for example , suppose @xmath294 and @xmath295 . the @xmath296-cycle that contracts to @xmath254 corresponding to the choice @xmath297 out of the @xmath298 choices is obtained by letting the maxima of the consecutive runs be @xmath299 .
therefore , the consecutive runs are @xmath300 .
then we arrange them so that @xmath291 and you get @xmath301 , and so the cycle is @xmath302 .
theorems [ ntmu ] and [ th : incont ] imply the following theorem .
[ thm : computec ] for any @xmath154 and @xmath228 , @xmath303 it is easy to see that ( [ eq : computec ] ) follows by partitioning the @xmath1-cycles @xmath99 such that @xmath304 by their contractions .
we shall prove in the next section that the smallest power of @xmath18 that occurs in either @xmath305 or @xmath306 is @xmath307 .
hence it follows that @xmath308 furthermore , the maximum number of nontrivial occurrences of @xmath0 that one can have for a cycle of length @xmath262 is @xmath309 , ( see page ) .
this means that for a cycle of length @xmath262 to have @xmath7 nontrivial occurrences , we must have that @xmath310 or , equivalently , @xmath311 .
it follows that it must be the case that @xmath312 .
hence , @xmath313 thus we have the following theorem . for any @xmath154 and @xmath228 , @xmath314 where @xmath315 .
it follows from our tables for @xmath17 and the fact that @xmath316 which we will prove in corollary [ cor : cat ] that for all @xmath228 @xmath317
let @xmath318 denote the number of incontractible @xmath1-cycles in @xmath66 .
clearly , there is only 1 incontractible 3-cycle , namely , @xmath319 and there are only 2 incontractible 4-cycles , namely , @xmath320 and @xmath321 .
our next theorem shows the numbers @xmath318 satisfies the recursion of the number of derangements .
[ thm : icrec ] for all @xmath152 , @xmath322 let @xmath323 = \{1 , \ldots , n\}$ ] .
suppose that @xmath152 and @xmath324 is an @xmath1-cycle in @xmath158 .
let @xmath325 } = ( 1,c_2^\prime , \ldots , c_{n-1}^\prime)$ ] be the cycle obtained from @xmath2 by removing @xmath1 from @xmath2 .
for example if @xmath326 then @xmath327}=(1,3,2,4,5)$ ] .
we then have two cases depending on whether @xmath325}$ ] is incontractible or not . + + * case 1 .
* @xmath328 } \in \mathcal{ic}_{n-1}$ ] .
+ in this situation , it is easy to see that if @xmath329}$ ] , there are exactly @xmath195 cycles @xmath330 such that @xmath331}$ ] .
these cycles are the result of inserting @xmath1 immediately after @xmath5 in @xmath332 for @xmath333 .
that is , let @xmath334 be the result of inserting @xmath1 immediately after @xmath5 in the cycle structure of @xmath332 where @xmath335 .
then it is easy to see that @xmath336 and @xmath337}$ ] .
for example , if @xmath338 is the 5-cycle pictured at the top of figure [ fig : insertn ] , then @xmath339 are pictured on the second row of figure [ fig : insertn ] , reading from left to right .
thus , there are @xmath340 @xmath341 @xmath1-cycles @xmath342 such that @xmath325}$ ] is an element of @xmath343 .
( note that @xmath344 is not incontractible because it is obtained by inserting @xmath1 directly after @xmath345 . )
we note that if @xmath346 is an @xmath347-cycle in @xmath343 , then @xmath348 that is , if @xmath349 is a nontrivial occurrence of @xmath0 in @xmath332 , then the insertion of @xmath1 does not effect whether @xmath349 is a nontrivial occurrence of @xmath0 in @xmath334 . moreover , we are always guaranteed that @xmath350 is a nontrivial occurrence of @xmath0 in @xmath334 since @xmath351 .
it is possible that @xmath352 is greater than or equal to 1 .
for example , in figure [ fig : insertn ] , it is easy to see that if @xmath338 , then there is only two pairs which are nontrivial occurrences of @xmath0 in @xmath2 , namely @xmath353 and @xmath354 , while in @xmath355 , the insertion of @xmath356 after 1 created three new nontrivial occurrences of @xmath0 in @xmath355 namely , the pairs @xmath357 , @xmath358 , and @xmath359 .
thus , @xmath360 . on the other hand
, it is easy to see that if @xmath361 , then @xmath362 that is , if we insert @xmath1 immediately after @xmath195 in @xmath332 , then for @xmath363 , @xmath350 can not be a nontrivial occurrence of @xmath0 in @xmath364 because @xmath195 will be between @xmath5 and @xmath1 in @xmath332 .
thus , we will create exactly one more pair which is a nontrivial occurrence of @xmath0 in @xmath364 that was not in @xmath332 , namely , @xmath365 .
+ + * case 2 * @xmath325 } \not \in \mathcal{ic}_n$ ] .
+ in this case , it must be that in @xmath2 , @xmath1 is the only element between @xmath5 and @xmath11 in @xmath2 for some @xmath5 so that in @xmath325}$ ] , @xmath11 immediately follows @xmath5 .
clearly , @xmath5 is the only @xmath6 in @xmath325}$ ] such that @xmath366 immediately follows @xmath6 .
hence , we can construct an element of @xmath367 from @xmath325}$ ] by removing @xmath11 and then replacing @xmath6 by @xmath368 for @xmath369 .
vice versa , given @xmath367 and @xmath335 , let @xmath370}$ ] be the cycle that results from @xmath332 by first replacing elements @xmath371 , by @xmath366 , then replacing @xmath5 by a pair @xmath5 immediately followed by @xmath11 , and finally inserting @xmath1 between @xmath5 and @xmath11 .
this process is pictured in figure [ fig : insertn-2 ] for the cycle @xmath372 .
in such a situation , we shall say that @xmath370}$ ] arises by expanding @xmath332 at @xmath5 .
we should note that if @xmath373 is an @xmath340-cycle in @xmath374 , then @xmath375 } ) \geq 1 + nt_{\mu}(d ) \ \mbox{for } \ i = 1 , \ldots , n-2.\ ] ] in the first step of the expansion of @xmath332 at @xmath5 , we replace each @xmath371 by @xmath366 and then replace @xmath5 by a consecutive pair @xmath5 followed by @xmath11 to get a cycle @xmath376 .
it is clear that this keeps the number of nontrivial occurrences of @xmath0 the same .
that is , it is easy to see that there will be no pair @xmath377 which is a nontrivial occurrence of @xmath0 in @xmath376 since @xmath5 is immediately followed by @xmath11 in @xmath376 .
moreover , it is easy to check that 1 . if @xmath275 , then @xmath378 is a nontrivial occurrence of @xmath0 in @xmath332 if and only if @xmath378 is a nontrivial occurrence of @xmath0 in @xmath376 , 2 . if @xmath379 , then @xmath380 is a nontrivial occurrence of @xmath0 in @xmath332 if and only if @xmath381 is a nontrivial occurrence of @xmath0 in @xmath376 .
3 . @xmath382 is a nontrivial occurrence of @xmath0 in @xmath332 if and only if @xmath383 is a nontrivial occurrence of @xmath0 in @xmath376 , and 4 . if @xmath384 , then @xmath380 is a nontrivial occurrence of @xmath0 in @xmath332 if and only if @xmath385 is a nontrivial occurrence of @xmath0 in @xmath376 . then for every pair @xmath386 which is a nontrivial occurrence of @xmath0 in @xmath376 , there exists a nontrivial occurrence of @xmath0 in @xmath370}$ ] after inserting @xmath1 between @xmath5 and @xmath11 .
finally we will create at least one new nontrivial occurrence of @xmath0 in @xmath332 , namely @xmath350 .
again it is possible that @xmath387 } ) -nt_{\mu}(d)$ ] is greater than or equal to 1 .
for example , in figure [ fig : insertn-2 ] , it is easy to see that if @xmath338 , there are only two pairs that are nontrivial occurrences of @xmath0 in @xmath332 , namely @xmath353 and @xmath388 , which correspond to the pairs @xmath389 and @xmath354 that are nontrivial occurrences of @xmath0 in @xmath390}$ ] . however , the insertion of 6 between 2 and 3 in @xmath391 created two new nontrivial occurrences of @xmath0 in @xmath390}$ ] , namely , @xmath392 and @xmath393 .
thus , @xmath394 } ) - nt_{\mu}(d ) = 4 - 2 = 2 $ ] . on the other hand
, it is easy to see that if @xmath367 , then @xmath395 } ) = 1 + nt_{\mu}(d).\ ] ] that is , if we insert @xmath1 immediately after @xmath195 in @xmath396 , then for @xmath363 , @xmath350 can not be a nontrivial occurrences of @xmath0 in @xmath397}$ ] because @xmath195 will be between @xmath5 and @xmath1 in @xmath397}$ ] .
thus , we will create exactly one occurrence of @xmath0 in @xmath397}$ ] that was not in @xmath396 , namely , @xmath365 .
+ we have two important corollaries to our proof of theorem [ thm : icrec ] . for all @xmath398 , @xmath399 where @xmath14 is the number of derangements of @xmath12 .
it is well - known that @xmath400 , @xmath401 , and that for @xmath398 , + @xmath402 .
thus , the corollary follows by recursion . for all @xmath398 , the lowest power of @xmath18 appearing in @xmath305 and
+ @xmath306 is @xmath307 .
we have shown by direct calculation that the lowest power of @xmath18 appearing in @xmath165 and @xmath167 is @xmath403 .
thus , the corollary holds for @xmath404 .
now , assume the corollary holds for @xmath398 .
then we shall show the corollary holds for @xmath405 .
we have shown that each @xmath406-cycle @xmath2 in @xmath407 is either of the form @xmath334 for some @xmath408 and @xmath409 in which case @xmath410 , or of the form @xmath411}$ ] for some @xmath412 and @xmath409 in which case @xmath413 } ) \geq 1 + nt_{\mu}(e)$ ] .
but by induction , we know that @xmath414 and @xmath415 so that @xmath416 .
thus , the smallest possible power of @xmath18 that can appear in @xmath417 is @xmath405 . on the other hand
, we can assume by induction that there is a @xmath408 such that @xmath418 in which case we know that @xmath419 .
hence , the coefficient of @xmath420 in @xmath421 is non - zero .
we have shown that each @xmath422-cycle @xmath2 in @xmath423 is either of the form @xmath334 for some @xmath424 and @xmath425 in which case @xmath410 , or of the form @xmath411}$ ] for some @xmath426 and @xmath425 in which case @xmath413 } ) \geq 1 + nt_{\mu}(e)$ ] . but
by induction , we know that @xmath427 and @xmath415 so that @xmath416 . thus , the smallest possible power of @xmath18 that can appear in @xmath428 is @xmath405 . on the other hand
, we can assume by induction that there is a @xmath408 such that @xmath418 in which case we know that @xmath429 } ) = 1 + nt_{\mu}(d ) = n+1 $ ] .
hence , the coefficient of @xmath420 in @xmath430 is non - zero .
we have not been able to find a recursion for the polynomials @xmath17 .
the problem with the recursion implicit in the proof of theorem [ thm : icrec ] is that for @xmath1-cycles @xmath431 , the contributions of @xmath432 are not uniform .
for example , if @xmath338 , then @xmath433 , @xmath434 , @xmath435 , @xmath436 , and @xmath437 so that @xmath438 .
however , @xmath439 , then @xmath440 , @xmath441 , @xmath442 , @xmath443 , and @xmath444 so that @xmath445 . a similar phenomenon occurs for @xmath446})$ ] .
for example , it is easy to see from figure 6 that if @xmath447 , then @xmath433 , @xmath448 } ) = 3 $ ] , @xmath394 } ) = 4 $ ] , @xmath449 } ) = 3 $ ] , and @xmath450 } ) = 3 $ ] , so that @xmath451 } ) } = ( 3q+q^2)q^{nt_{\mu}(d)}$ ] .
however , as one can see from figure [ fig:2insertn-2 ] below that if @xmath452 , then @xmath453 , @xmath448 } ) = 6 $ ] , @xmath394 } ) = 5 $ ] , @xmath449 } ) = 4 $ ] , and @xmath450 } ) = 4 $ ] , so that @xmath451 } ) } = ( 2q+q^2+q^3)q^{nt_{\mu}(d)}$ ] .
our next goal is to show that @xmath454 where @xmath232 is the @xmath1th catalan number .
[ thm : chpath ] suppose that @xmath455 and @xmath456 is a dyck path of length @xmath227
. then the @xmath1-cycle @xmath457 is incontractible and @xmath458 .
since the path @xmath459 starts at @xmath460 , it follows that @xmath461 .
this can only happen if @xmath462 .
thus , @xmath463 has the standard form of an @xmath1-cycle .
note that the only way @xmath11 immediately follows @xmath5 in @xmath464 is if there is a @xmath6 such that @xmath465 and @xmath466 .
but then @xmath467 which would imply that the charge path of @xmath26 has a level which means that there is an edge drawn horizontally . since we are assuming that the charge path of @xmath464 is a dyck path , there can be no such @xmath5 and , hence , @xmath464 is incontractible .
next we claim that if the @xmath5th step of @xmath459 is an up - step , then @xmath468 is a nontrivial occurrence of @xmath0 in @xmath469 .
that is , if the @xmath5th step of @xmath459 is an up - step , then @xmath470 and @xmath471 for some @xmath154 .
since @xmath459 is a dyck path , there must be some @xmath472 such that @xmath473 because we must pass through level @xmath7 from the point @xmath474 in @xmath217 to get back to the point @xmath218 which is the end point of the path .
let @xmath475 denote the smallest @xmath476 such that @xmath477 .
it follows from the definition of the function @xmath478 that it must be the case that @xmath479 and @xmath480 .
moreover , it must be the case that @xmath481 so that @xmath468 is a nontrivial occurrence of @xmath0 in @xmath469 .
we claim that there can be no @xmath482 such that @xmath483 is a nontrivial occurrence of @xmath0 in @xmath464 .
that is , since @xmath459 is a dyck path , we know that all the vertices on @xmath459 between @xmath484 and @xmath485 lie on levels strictly greater than @xmath7 .
it is easy to see from our inductive definition of @xmath486 that it must be the case that @xmath487 is the least element of @xmath488 and hence the fact that @xmath487 immediately follows @xmath205 in @xmath464 implies that none of the pairs @xmath489 is a nontrivial occurrence of @xmath0 in @xmath464 for @xmath490 .
but then as we traverse around the cycle @xmath464 , the fact that @xmath479 and @xmath480 implies that none of the pairs @xmath491 where either @xmath492 or @xmath493 are nontrivial occurrences of @xmath0 in @xmath464 .
thus , if the @xmath5th step of @xmath459 is an up - step , then @xmath494 is an occurrence of @xmath0 in @xmath464 for exactly one @xmath495 , namely @xmath496 .
next suppose that the @xmath5th step of @xmath459 is a down - step .
then , we claim that there is no @xmath495 such that @xmath494 is an occurrence of @xmath0 in @xmath464 .
we then have two cases .
+ + * case 1 . *
@xmath497 and there is a @xmath476 such that @xmath477 .
+ in this case , let @xmath475 be the least @xmath498 such that @xmath477 .
it then follows that if @xmath479 , then @xmath480 .
because @xmath459 is a dyck path , it must be the case that @xmath499 for all @xmath500 so that @xmath501 for all @xmath500 .
but this means that as we traverse around the cycle @xmath464 , the first element that we encounter after @xmath479 which is bigger than @xmath288 is @xmath502 and hence there is no @xmath503 such that @xmath504 is a nontrivial occurrence of @xmath0 in @xmath464 . + + * case 2 . * @xmath497 and there is no @xmath476 such that @xmath477 .
+ in this case since @xmath459 is a dyck path , all the elements @xmath505 such that @xmath506 must have @xmath499 and hence @xmath507 since , in a charge graph , all the elements whose index in @xmath26 is less than @xmath7 are smaller than all the elements whose index is equal to @xmath7 in @xmath26 .
we now have two subcases . + + * subcase 2.1 .
* there is an @xmath508 such that @xmath509 .
+ let @xmath510 be the smallest @xmath508 such that @xmath509 .
since @xmath205 was the rightmost element whose index relative to @xmath26 equals @xmath7 , it follows that if @xmath511 , then @xmath512 since it is the leftmost element whose index relative to @xmath26 equals @xmath513 .
moreover , for all @xmath514 , either @xmath515 has index less than @xmath7 in @xmath26 or @xmath515 has index @xmath7 in @xmath26 . in either case
, our definition of @xmath486 ensures that @xmath516 .
but this means that as we traverse around the cycle @xmath464 , the first element that we encounter after @xmath479 which is bigger than @xmath288 is @xmath512 and hence there is no @xmath503 such that @xmath504 is a nontrivial occurrence of @xmath0 in @xmath464 . + + * subcase 2.2 .
* there is no @xmath508 such that @xmath509 .
+ in this case , @xmath205 has the highest possible index in @xmath26 and it is the rightmost element whose index in @xmath26 is @xmath7 which means that @xmath517 .
hence , there is no @xmath503 such that @xmath504 is an occurrence of @xmath0 in @xmath464 .
+ + the only other possibility is that we have not considered is that @xmath518 . since @xmath459 is a dyck path , we know that @xmath519 , @xmath520 , and @xmath521 .
but then it follows that if @xmath522 , then @xmath523 and @xmath524 . but
this means that as we traverse around the cycle @xmath464 , the first element that we encounter after @xmath522 which is bigger than @xmath288 is @xmath525 and hence there is no @xmath503 such that @xmath526 is a nontrivial occurrence of @xmath0 in @xmath464 .
thus , we have shown that the nontrivial occurrences of @xmath0 in @xmath464 correspond to the up - steps of @xmath217 and hence @xmath527 .
our next theorem will show that if @xmath528 is a @xmath529-cycle in @xmath224 where @xmath530 , then the charge path of @xmath531 must be a dyck path .
[ thm : minchpath ] let @xmath532 and @xmath531 .
then , if @xmath530 , the charge path of @xmath533 must be a dyck path .
our proof proceeds by induction on @xmath1 . for @xmath259 ,
the only element of @xmath534 is @xmath535 and it is easy to see that @xmath536 is a dyck path .
now assume by induction that if @xmath537 is a @xmath529-cycle in @xmath224 such that @xmath538 , then the charge path of @xmath539 is a dyck path .
now suppose that @xmath540 is a @xmath422-cycle in @xmath423 such that @xmath541 .
let @xmath542 .
then , we know by our proof of theorem [ thm : icrec ] that either @xmath543 for some @xmath424 or @xmath544}$ ] for some @xmath408
. we claim that @xmath2 can not be of the form @xmath545 for some @xmath424 .
that is , we know that for all @xmath424 we have @xmath546 and , hence , @xmath547 .
thus , there must be some @xmath408 such that @xmath548}$ ] for some @xmath6 .
but then we know that @xmath549 is a dyck path .
let @xmath550 and @xmath551 .
note that it follows from our arguments in theorem [ thm : icrec ] that @xmath552 } ) - nt_{\mu}(d)$ ] is equal to the number of pairs @xmath553 which are nontrivial occurrences of @xmath0 in @xmath554}$ ] .
it is also the case that the charge path of @xmath26 is easily constructed from the charge path of @xmath555 .
that is , suppose that @xmath556 . then @xmath557 are in the same order in @xmath26 and @xmath555 so that @xmath558 for @xmath559 . in going from @xmath555 to @xmath26 , we first increased the value of any @xmath560 by one and then inserted @xmath366 immediately after @xmath6 to get a permutation @xmath561 . thus @xmath562 . moreover , @xmath563 are in the same relative order in @xmath561 as @xmath564 in @xmath555 so that for @xmath565 , @xmath566
. finally , inserting @xmath567 in @xmath561 between @xmath6 and @xmath366 has no effect on the indices assigned to @xmath568 .
thus for @xmath559 , @xmath569 and for @xmath565 , @xmath570 .
this means that one can construct the charge path of @xmath26 by essentially starting with the charge graph of @xmath555 , then replacing the vertex @xmath571 by a horizontal edge , and finally replacing that horizontal edge by a pair of edges @xmath572 and @xmath573 . in particular , this means the charge paths from @xmath106 up to @xmath5 are identical for both @xmath26 and @xmath555 and the charge path from @xmath574 to @xmath567 in @xmath26 is identical to the charge path from @xmath5 to @xmath223 in @xmath555 .
we now consider several cases depending on which @xmath5 is such that @xmath556 and the value @xmath575 in the dyck path @xmath576 .
+ * case 1 . *
@xmath577 where @xmath578 .
+ in this case , we will have @xmath579 , @xmath580 , @xmath581 , and @xmath582 .
since @xmath576 is a dyck path , there will be some @xmath7 such that @xmath583 , and @xmath584 so that we will be in the situation pictured in figure [ fig : case1a ] . because @xmath585 , it must be the case that @xmath586 .
it is easy to see that @xmath587 is an occurrence of @xmath0 in @xmath2 .
we then have two subcases .
+ + * subcase 1a .
* there is an @xmath508 such that @xmath588 .
+ in this case , let @xmath589 be the largest @xmath262 such that @xmath508 and @xmath588 . because @xmath576 is a dyck path , it must be the case that for all @xmath590 , @xmath591 because on the path from @xmath592 to @xmath484 , we can not have an @xmath593 such that @xmath594 without having some @xmath595 such that @xmath596 which would violate our choice of @xmath589 .
but this means that @xmath597 for all @xmath598 and hence , @xmath599 is a nontrivial occurrence of @xmath0 in @xmath2 .
thus , @xmath600 . + + * subcase 1b .
* there is no @xmath508 such that @xmath588 .
+ in this case , let @xmath589 be the largest @xmath262 such that @xmath601 and @xmath602 . because @xmath576 is a dyck path such an @xmath262 must exist since @xmath585 and a dyck path that reaches level @xmath513 must subsequently descend to level @xmath7 .
but this means that for all @xmath603 , @xmath604 and hence , @xmath605 .
moreover , for all @xmath606 , @xmath607 and hence , @xmath608 .
but it then follows that since @xmath609 , @xmath599 is a nontrivial occurrence of @xmath0 in @xmath2 .
thus , @xmath610 .
* case 2 . * @xmath611 .
+ in this case , we will have @xmath612 , @xmath613 and @xmath614 .
thus , the charge path of @xmath615}$ ] looks like one of the two situations pictured in figure [ fig : case2 ] .
that is , we have two subcases . + + * subcase 2a . * @xmath616 .
+ in this case , @xmath617 will be a dyck path .
this case can only happen if @xmath618 for all @xmath619 . + + * subcase 2b . * @xmath620 .
+ in this case , @xmath217 will not be a dyck path .
this case can only happen if there is an @xmath262 such that @xmath621 .
but then since @xmath622 , it must be the case that @xmath623 .
hence both @xmath624 and @xmath625 are nontrivial occurrences of @xmath0 in @xmath2 so that @xmath626 .
+ * case 3 . * @xmath577 where @xmath627 .
+ in this case , we will have @xmath628 , @xmath579 , @xmath580 , @xmath581 , and @xmath629 .
there are now two cases . + + * case 3a .
* @xmath630 .
+ in this case , we will have the situation pictured in figure [ fig : case4a ] .
that is , since @xmath631 , it must be the case the either @xmath632 is to the left of @xmath633 in which case @xmath634 or @xmath632 is the right of @xmath635 in which case @xmath636 and @xmath567 is to the left of @xmath632 in @xmath26 . in either case , it must be that @xmath637 .
it is then easy to see that both @xmath638 and @xmath639 are nontrivial occurrences of @xmath0 in @xmath2 so that @xmath626 .
+ * case 3b .
* @xmath640 .
+ in this case , we will have the situation pictured in figure [ fig : case4b ] .
in this situation , the charge path of @xmath641}$ ] will be a dyck path .
+ * case 4 . *
@xmath642 where @xmath643 .
+ in this case , we will have @xmath644 , @xmath579 , @xmath580 , @xmath581 , and @xmath629 .
there are now two subcases . + + * subcase 4a .
* @xmath637 .
+ in this case , we have the situation pictured in figure [ fig : case6a ] .
it will always be the case that @xmath645 is a nontrivial occurrence of @xmath0 in @xmath2 .
we then have two more subcases . + + * subcase 4a1 .
* there is an @xmath508 such that @xmath646 .
+ in this case , let @xmath589 be the largest @xmath262 such that @xmath508 and @xmath588 . because @xmath576 is a dyck path , we can argue as in case 1a that it must be the case that for all @xmath590 , @xmath647 .
but this means that @xmath597 for all @xmath598 and hence , @xmath599 is also a nontrivial occurrence of @xmath0 in @xmath2 .
thus , @xmath648 . + + * subcase 4a2 .
* there is no @xmath508 such that @xmath646 .
+ in this case , let @xmath589 be the largest @xmath262 such that @xmath601 and @xmath649 . note that @xmath589 exists since @xmath650 . but
this means that for all @xmath603 , @xmath604 and , hence , @xmath597 .
moreover , for all @xmath606 , @xmath651 and hence , @xmath652 .
but it then follows that since @xmath609 , @xmath599 is also a nontrivial occurrence of @xmath0 in @xmath2 .
thus , @xmath610 . + *
subcase 4b . * @xmath653 .
+ in this case , we have the situation pictured in figure [ fig : case6b ] and , hence , @xmath217 will be a dyck path .
thus , we have shown that if @xmath408 is such that @xmath538 and @xmath370 } = c$ ] , then either @xmath217 is a dyck path or @xmath387 } ) \geq n+2 $ ] . hence ,
if @xmath541 , then @xmath217 is a dyck path .
theorems [ thm : chpath ] and [ thm : minchpath ] yield the following corollary . [ cor : cat ] @xmath454 where @xmath654 is the @xmath1th catalan number .
in this section we will show that @xmath21 and @xmath22 are equal to the number of partitions of @xmath7 for sufficiently large @xmath1 by plotting the _ non-@xmath0-matches _ on a grid and mapping them to integer partitions .
given an @xmath1-cycle @xmath99 , we say that a pair @xmath655 with @xmath27 is a non-@xmath0-match of @xmath2 if @xmath655 is not an occurrence of @xmath0 in @xmath2 .
in other words , @xmath655 is a non-@xmath0-match if there exists an integer @xmath219 such that @xmath656 and @xmath219 is cyclically between @xmath5 and @xmath6 in @xmath2 .
furthermore , let @xmath657 be the set of non-@xmath0-matches in @xmath2 and let @xmath658 .
let @xmath659 note that for any @xmath99 , @xmath660 so that + @xmath661
. table [ tab : nm ] shows the polynomials @xmath662 for @xmath663 .
our data suggests the following theorem .
c||*2l @xmath664 & @xmath106 + @xmath665 & @xmath106 + @xmath666 & @xmath109 + @xmath667 & @xmath668 + @xmath669 & @xmath670 + @xmath356 & @xmath671 + @xmath672 & @xmath673 + & @xmath674 + @xmath296 & @xmath675 + & @xmath676 + & @xmath677 + @xmath678 & @xmath679 + & @xmath680 + & @xmath681 + & @xmath682 + @xmath683 & @xmath684 + & @xmath685 + & @xmath686 + & @xmath687 + & @xmath688 + [ nm ] for @xmath689 , @xmath690 where @xmath691 is the number of integer partitions of @xmath7 . to prove theorem [ nm ]
, we need to define what we call the plot of non-@xmath0-matches in @xmath99 .
we consider an @xmath187 grid where the rows are labeled with @xmath692 , reading from bottom to top , and the columns are labeled with @xmath693 , reading from left to right .
the cell @xmath694 is the cell which lies in the @xmath5th row and @xmath6th column .
then , given a set @xmath695 of ordered pairs @xmath655 with @xmath696 , we let @xmath697 denote the diagram that arises by shading a cell @xmath698 on the @xmath187 grid if and only if @xmath699 . given an @xmath1-cycle @xmath99 ,
we let @xmath700 .
for example , if @xmath701 , then @xmath702 and @xmath703 is pictured in figure [ fig : nmplot ] .
.,scaledwidth=25.0% ] first , we observe that if @xmath99 , then @xmath657 completely determines @xmath2 .
that is , we have the following theorem . if @xmath695 is a set of ordered pairs such that @xmath704 for some @xmath1-cycle @xmath99 , then @xmath2 is the only @xmath1-cycle such that @xmath704 .
our proof proceeds by induction on @xmath1 . clearly , the theorem holds for @xmath259 and @xmath404 .
now , assume that the theorem holds for @xmath1 .
let @xmath705 and @xmath706 .
let @xmath707 be the @xmath1-cycle that is obtained from @xmath2 by removing @xmath405 .
then it is easy to see that @xmath708 .
hence , @xmath2 is the unique @xmath1-cycle in @xmath66 such that @xmath708 .
let @xmath709 denote the cycle that results by inserting @xmath405 immediately after @xmath5 in @xmath2 .
since @xmath2 is unique , it follows that @xmath271 must equal @xmath709 for some @xmath710 .
however , it is easy to see that if @xmath711 , then @xmath712 for @xmath713 and @xmath714 . if @xmath715 , then @xmath712 for @xmath716 .
thus , it follows that @xmath717 are pairwise distinct .
hence , there is exactly one cycle @xmath271 such that @xmath718 .
an integer sequence @xmath719 is a _ partition of @xmath1 _ if @xmath720 and @xmath721 .
if @xmath719 is a partition of @xmath1 , we will write @xmath722 , and let @xmath723 denote the size of @xmath724 and @xmath725 denote the length of @xmath724 . the _
ferrers diagram _ of a partition @xmath724 on the @xmath726 grid , denoted @xmath727 , is the diagram that results by shading the squares @xmath728 in the top row , shading the squares @xmath729 in the second row from the top , and , in general , shading the squares @xmath730 in the @xmath5th row from the top .
for example , the ferrers diagram of @xmath731 on a @xmath732 grid is pictured in figure [ fig : ferrers ] . , scaledwidth=25.0% ] it is easy to see that the following three properties characterize the ferrers diagrams of partitions @xmath724 in an @xmath733 grid . 1 .
if @xmath724 is not empty , then the square @xmath734 is shaded in @xmath727 .
if @xmath735 and the square @xmath736 is shaded in @xmath727 , then the square @xmath737 is shaded in @xmath727 .
if @xmath738 and the square @xmath736 is shaded in @xmath727 , then the square @xmath739 is shaded in @xmath727 .
this means that if @xmath695 is a set of pairs @xmath694 such that @xmath697 is a ferrers diagram , then it must be the case that 1 . if @xmath695 is not empty , then the square @xmath740 .
2 . if @xmath735 and the square @xmath741 , then @xmath742 .
3 . if @xmath738 and the square @xmath741 , then @xmath743 .
our next lemma and corollary will show that if @xmath99 and @xmath744 , then @xmath745 has the same three properties .
[ lem : n-2 ] assume that @xmath99 . 1 . if @xmath746 and @xmath747 , then @xmath748 .
2 . if @xmath749 , @xmath750 , and @xmath751 , then @xmath748 .
3 . if @xmath752 , @xmath750 , and @xmath753 , then @xmath748 . let @xmath754 be an @xmath1-cycle where @xmath755
. for part ( 1 ) , suppose that @xmath746 and @xmath747
. then , there are no integers that are cyclically in between @xmath106 and @xmath1 in @xmath2 so that @xmath756 . since @xmath757 , @xmath758
thus , @xmath2 must be of the form @xmath759 where @xmath760 . that is , the sequence @xmath761 consists of the decreasing interval from @xmath1 to @xmath762 and then @xmath763 . it follows that the pairs @xmath764 are all _
non-@xmath0-matches _ in @xmath2 which accounts for @xmath765 non-@xmath0-matches in @xmath2 .
let @xmath254 be the set of integers cyclically between @xmath270 and @xmath766 in @xmath2 and let @xmath767 be the set of integers cyclically between @xmath766 and @xmath106 in @xmath2 .
note that 1 . if @xmath768 and @xmath769 , then @xmath770 , 2 . if @xmath768 and @xmath771 , then @xmath772 , 3 .
if @xmath773 and @xmath769 , then @xmath774 , and 4 . if @xmath773 and @xmath771 , then @xmath772 .
it follows that each element in @xmath254 and @xmath767 is part of at least one non-@xmath0-match in @xmath2 so that we know that @xmath775 .
thus , since @xmath776 , we have @xmath777 .
+ for part ( 2 ) , suppose that @xmath749 , @xmath750 , and @xmath751
. then @xmath219 can not be cyclically between @xmath778 and @xmath779 in @xmath2 .
also , there exists an integer @xmath780 with @xmath781 such that @xmath780 is cyclically between @xmath219 and @xmath779 in @xmath2 , but @xmath780 is not cyclically between @xmath778 and @xmath779 in @xmath2 .
it follows that @xmath2 is of the form @xmath782 where @xmath783 is the set of integers cyclically between @xmath219 and @xmath778 in @xmath2 that are not equal to @xmath780 , @xmath784 is the set of integers cyclically between @xmath778 and @xmath779 in @xmath2 , and @xmath785 is the set of integers between @xmath779 and @xmath219 in @xmath2 .
note that @xmath786 and @xmath787 are in @xmath657 .
moreover , 1 . if @xmath788 and @xmath789 , then @xmath790 , 2 . if @xmath788 and @xmath791 , then @xmath792 , 3 . if @xmath793 and @xmath789 , then @xmath794 , 4 .
there are no @xmath793 with @xmath795 since @xmath751 , 5 . if @xmath793 and @xmath796 , then @xmath797 , 6 . if @xmath798 and @xmath799 , then @xmath790 , and 7 . if @xmath798 and @xmath800 , then @xmath801 . therefore
, any integer @xmath802 is part of a distinct non-@xmath0-match in @xmath2 . since + @xmath803
, it follows that @xmath804 . for part ( 3 ) , suppose that @xmath752 , @xmath750 , and @xmath753
. then @xmath779 can not be cyclically between @xmath219 and @xmath805 in @xmath2 .
also , there exists an integer @xmath780 with @xmath781 such that @xmath780 is cyclically between @xmath219 and @xmath779 in @xmath2 , but @xmath780 is not cyclically between @xmath219 and @xmath805 in @xmath2 .
thus , @xmath2 must be of the form @xmath806 where @xmath807 is the set of integers cyclically between @xmath219 and @xmath805 in @xmath2 , @xmath808 is the set of integers cyclically between @xmath805 and @xmath779 in @xmath2 that are not equal to @xmath780 , and @xmath809 is the set of integers cyclically between @xmath779 and @xmath219 in @xmath2 .
note that @xmath786 and @xmath810 are in @xmath657 .
moreover , 1 . if @xmath811 and @xmath812 , then
@xmath813 , 2 .
there is no @xmath811 with @xmath814 since @xmath753 , 3 . if @xmath811 and @xmath815 , then @xmath816 , 4 .
if @xmath817 and @xmath818 , then @xmath819 , 5 . if @xmath817 and @xmath815 , then @xmath820 , 6 . if @xmath821 and @xmath822 , then @xmath813 , and 7 . if @xmath821 and @xmath823 then @xmath820 .
therefore , any integer @xmath824 is part of a distinct non-@xmath0-match in @xmath2 .
since @xmath825 , it follows that @xmath826 .
[ nmplot ] if @xmath827 , then @xmath703 is a ferrers diagram .
this corollary follows directly from lemma [ lem : n-2 ] . that is ,
suppose that @xmath99 and @xmath744 .
first , if @xmath828 , then @xmath829 in which case @xmath830 has no shaded squares which correspond to the empty partition .
thus , assume that @xmath831 .
then , it follows from part ( 1 ) of lemma [ lem : n-2 ] that @xmath832 .
next it follows from part ( 2 ) of lemma [ lem : n-2 ] that if @xmath749 and @xmath750 , then @xmath833 . finally , it follows from part ( 3 ) of lemma [ lem : n-2 ] that if @xmath752 and @xmath750 , then @xmath834 .
hence , the shaded cells @xmath703 must be a ferrers diagram of a partition @xmath724 of @xmath835 . if @xmath148 , we let @xmath836 be the plot of the ferrers diagram in the @xmath187 grid corresponding the to partition @xmath837 . thus ,
for example , @xmath838 is pictured in figure [ fig : t7 ] .
one can see that the sets that @xmath836 have the property that @xmath839 where @xmath232 is the @xmath1th catalan number since the lower boundaries of the plots of @xmath840 for such @xmath724 correspond to dyck paths .
+ on the @xmath841 grid.,scaledwidth=25.0% ] our next theorem will show that for any @xmath148 and any partition @xmath842 , we can construct an @xmath1-cycle @xmath99 such that @xmath843 .
[ thm : const ] suppose that @xmath148 and @xmath724 is a partition such that @xmath844 .
then , there is an @xmath1-cycle @xmath99 such that @xmath843 .
we proceed by induction on @xmath1 . for @xmath845 ,
it is easy to see that if @xmath846 , the @xmath847 is empty and if @xmath848 , then @xmath849 .
note that @xmath850 and @xmath851 have the property that if the largest part of the corresponding partition is of size @xmath5 , then @xmath666 immediately follows @xmath11 in the cycle .
thus , our theorem holds for @xmath845 . now , suppose that @xmath852 and @xmath853 is a partition which is contained in @xmath836 .
it follows that @xmath854 is contained in @xmath855 .
assume by induction that there is an @xmath856-cycle @xmath271 such that @xmath857 and @xmath345 immediately follows @xmath858 in @xmath271 .
thus , in @xmath271 , the pairs @xmath859 must match @xmath0 in @xmath271 .
this means that if @xmath860 , then @xmath861 must lie between @xmath195 and @xmath345 in @xmath271 .
next if @xmath862 , then @xmath863 must lie between @xmath861 and @xmath345 in @xmath271 .
in general , if @xmath864 , then @xmath865 must lie between @xmath866 and @xmath345 in @xmath271 . thus , @xmath271 must be of the following form @xmath867 where @xmath868 is the set of elements cyclically between @xmath869 and @xmath870 in @xmath271 for @xmath871
. now let @xmath2 be the cycle that results from @xmath271 by inserting @xmath1 immediately after @xmath872 in @xmath271 .
inserting @xmath1 into @xmath271 does not effect on whether pairs @xmath655 with @xmath873 are @xmath0-matches in @xmath2 .
that is , for such pairs @xmath874 if and only if @xmath875 . thus , the diagram of @xmath876 and @xmath703 are the same up to row @xmath345 . now , in row @xmath1 , we know that the cells @xmath877 are shaded since the fact that @xmath1 immediately follows @xmath872 in @xmath2 means that @xmath878 for @xmath879 . however , it is easy to see from the form of @xmath271 above that @xmath880 are @xmath0-matches in @xmath2 since @xmath881 appear in decreasing order as we traverse clockwise around the cycle @xmath2 .
thus , @xmath843 .
note the proof of theorem [ thm : const ] gives a simple algorithm to construct an @xmath1-cycle @xmath882 such that @xmath883 for any @xmath844 .
for example , suppose that @xmath884 and @xmath885 so that the ferrers diagram in the @xmath886 grid is pictured in figure [ fig : ex ] .
.,scaledwidth=30.0% ] since there are no non-@xmath0-matches in the first eight rows , we must start with the cycle @xmath887 .
then , our proof of theorem [ thm : const ] tells us that we should build up the cycle structure by first creating a cycle @xmath888 by inserting @xmath678 immediately after 2 in @xmath889 , since the number of non-@xmath0-matches in row 9 is 1 .
then we create a cycle @xmath890 by inserting @xmath683 immediately after 3 in @xmath891 since the number of non-@xmath0-matches in row 10 is 2 .
then , we create a cycle @xmath892 by inserting @xmath893 immediately after 4 in @xmath894 since the number of non-@xmath0-matches in row 11 is 3 .
finally , we create a cycle @xmath895 by inserting @xmath896 immediately after 4 in @xmath897 since the number of non-@xmath0-matches in row 12 is 3 .
thus , @xmath898 then we have that @xmath899 now we can prove theorem [ nm ] .
suppose that @xmath900 .
let @xmath901 theorem [ nmplot ] shows that @xmath902 and theorem [ thm : const ] shows that @xmath903 .
thus , @xmath904 .
hence , @xmath905 equals the number of partitions of @xmath7 .
now , we will show that for @xmath689 , @xmath906 , where @xmath691 is the number of partitions of @xmath7 .
first we shall show that if @xmath99 has fewer than @xmath195 non - matches , then @xmath2 must be incontractible .
[ n-2 in ] if @xmath99 has fewer than @xmath195 non - matches , then @xmath2 must be incontractible .
if @xmath11 immediately follows @xmath5 in an @xmath1-cycle @xmath99 , then , clearly , @xmath907 are non-@xmath0-matches for @xmath908 and @xmath909 is a non-@xmath0-match for @xmath910 .
this gives @xmath195 non-@xmath0-matches .
for @xmath689 , @xmath906 where @xmath691 is the number of partitions of @xmath7 . by definition , @xmath911 .
thus , by theorem [ nm ] , for @xmath689 , @xmath912 . by lemma [ n-2 in ] , we have that if a cycle @xmath2 has @xmath7 non-@xmath0-matches where @xmath689 , then @xmath2 is incontractible .
it follows that if a cycle has @xmath913 non - trivial @xmath0-matches , then it is incontractible .
thus , for @xmath689 , @xmath914 . and so @xmath915 for @xmath689 .
in this paper , we studied the polynomials @xmath916 and @xmath917 .
we showed that @xmath918 is the number of derangements of @xmath919 .
thus , the polynomial @xmath17 is a @xmath18-analogue of the derangement number @xmath13 .
there are several @xmath18-analogues of the derangement numbers that have been studied in the literature , see the papers by garsia and remmel @xcite and wachs@xcite .
our @xmath18-analogue of @xmath13 is different from either the garsia - remmel @xmath18-analogue or the wachs @xmath18-analogue of the derangement numbers .
moreover , we proved that @xmath920 so that the coefficients of the polynomial @xmath103 can be expressed in terms of the coefficients of the polynomials @xmath921 .
we also showed that @xmath922 equals the number of partition of @xmath7 for @xmath923 .
the main open question is to find some sort of recursion or generating function that would allow us to compute @xmath17 .
note that several of the sequences @xmath924 appear in the oeis @xcite .
for example , we showed that @xmath925 so that the sequence @xmath149 starts out @xmath926 .
the @xmath1th term of this sequence has several combinatorial interpretations including being the number of @xmath61 which are @xmath927-avoiding and have exactly two descents , the number of dyck paths on length @xmath632 with @xmath345 peaks , and the number of squares with corners on the @xmath187 grid .
thus , it would be interesting to find bijections from these objects to our @xmath928-cycles @xmath99 such that @xmath929 .
we also showed that @xmath930 so that the sequence @xmath931 starts out @xmath932 this sequences does not appear in the oeis .
similarly , the sequence @xmath933 starts out @xmath934 and it does not appear in the oeis . finally , it would be interesting to characterize the charge graphs of those cycles @xmath935 for which @xmath936 .
one can see from our tables that sequence @xmath937 starts out @xmath938 .
moreover , we have computed the @xmath939 .
this suggests that this sequence is sequence a008549 in the oeis .
if so , this would mean that @xmath940 for @xmath398 . | in this paper , we study the distribution of the number of occurrences of the simplest frame pattern , called the @xmath0 pattern , in @xmath1-cycles . given an @xmath1-cycle @xmath2 , we say that a pair @xmath3 matches the @xmath0 pattern if @xmath4 and as we traverse around @xmath2 in a clockwise direction starting at @xmath5 and ending at @xmath6 , we never encounter a @xmath7 with @xmath8 .
we say that @xmath9 is a nontrivial @xmath0-match if @xmath10 .
also , an @xmath1-cycle @xmath2 is incontractible if there is no @xmath5 such that @xmath11 immediately follows @xmath5 in @xmath2 .
we show that the number of incontractible @xmath1-cycles in the symmetric group @xmath12 is @xmath13 , where @xmath14 is the number of derangements in @xmath12 .
further , we prove that the number of @xmath1-cycles in @xmath12 with exactly @xmath7 @xmath0-matches can be expressed as a linear combination of binomial coefficients of the form @xmath15 where @xmath16 .
we also show that the generating function @xmath17 of @xmath18 raised to the number of nontrivial @xmath0-matches in @xmath2 over all incontractible @xmath1-cycles in @xmath12 is a new @xmath18-analogue of @xmath13 , which is different from the @xmath18-analogues of the derangement numbers that have been studied by garsia and remmel and by wachs .
we show that there is a rather surprising connection between the charge statistic on permutations due to lascoux and schzenberger and our polynomials in that the coefficient of the smallest power of @xmath18 in @xmath19 is the number of permutations in @xmath20 whose charge path is a dyck path .
finally , we show that @xmath21 and @xmath22 are the number of partitions of @xmath7 for sufficiently large @xmath1 . |
cmes are large - scale magnetized plasma structures that are ejected from the sun into the heliosphere over the course of minutes to hours , with a wide range of sizes , masses , and velocities .
cmes can be associated with the eruption of filaments and/or solar flares or without any apparent trigger ( see , e.g. , * ? ? ?
* ; * ? ? ?
the heliospheric counterparts of cmes are often referred to as interplanetary coronal mass ejections ( icmes ) , and they are one of the main drivers of space weather events on earth @xcite .
in fact , icmes can produce severe magnetic storms when interacting with the earth s magnetosphere ( see , e.g. , * ? ? ? * ; * ? ? ? * ) .
many aspects related to the initial phases and subsequent evolution of cmes are , at present , only poorly understood . to deepen as much as possible
the knowledge of the processes involved in cmes initiation and propagation is crucial to improve the magnetohydrodynamics ( mhd ) simulation techniques and , in turn , the space weather forecasting techniques .
significant insights on these problems can be gained by studying the three - dimensional ( 3d ) structure of cmes .
for instance , it has been proved that the geo - effectiveness of a cme ( i.e. , its capability to interact with the earth s magnetosphere and trigger geomagnetic storms ) strongly depends on the magnetic configuration of the associated icme , in particular on the relative orientation of the magnetic field embedded in the magnetic cloud associated with the icme with respect to the earth s magnetic field ( * ? ? ? * ; * ? ? ? * ; * ? ? ? * e.g , ) .
therefore , a better determination of the 3d structure of cmes may lead to a better understanding of the possible correlations between the initial evolution of cmes , the subsequent propagation and magnetic configuration of icmes , and their eventual impact on the earth .
cmes are typically observed in white - light ( wl ) images provided by coronagraphs , such as the large angle and spectrometric coronagraph ( lasco ; * ? ? ?
* ) on board the solar and heliospheric observatory ( soho ) , and the cor coronagraphs on board the solar terrestrial relations observatory ( stereo ) twin spacecraft @xcite .
white - light continuum emission arising from the coronal plasma is produced by thomson scattering of photospheric light by free electrons @xcite .
since the corona is optically thin , white - light images actually show the projection on the plane of the sky ( pos ) of the 3d cme structure , integrated along the instrumental lines of sight ( loss ) .
moreover , cmes are not observed in isolation , but in the presence of the fine structural detail of the surrounding corona @xcite .
projection and los integration effects can be rather severe and can significantly affect the determination of cme properties ( e.g. , * ? ? ?
* ) . for halo and partial - halo cmes ,
the projected altitude of cme features inferred from coronagraphic images can notably differ from the actual one ; for this reason , real position and speed of generic plasma elements can not be obtained from such observations unless a geometry of the cme is assumed . however , even in that case limitations and approximations are unavoidable ( see , e.g. , * ? ? ?
the main challenge is therefore to determine the location and the geometry of cmes in the 3d space using one or more bi - dimensional images , corresponding to different viewpoints of the same cme feature .
the stereo mission was conceived to properly address these issues ; since its launch in 2006 , a wealth of methods have been developed and tested in order to determine the full 3d structure of cmes .
we refer the interested reader to @xcite , who compared the results of different 3d reconstruction techniques applied to stereo images , _ in primis _ the triangulation or tie - pointing technique ( see , e.g. , * ? ? ?
* ) and the polarization ratio technique , originally developed by @xcite . a different approach to estimate the 3d cme structure uses forward modeling , that consists in making strong assumptions on the morphology of the feature that is intended to be reconstructed ( e.g. , the graduated cylindrical shell flux - rope model of * ? ? ?
* ; * ? ? ?
. further investigation method of cmes is based on ultraviolet ( uv ) spectroscopic observations , such as those provided by the ultraviolet coronagraph spectrometer ( uvcs ; * ? ? ?
* ) on board soho .
these have the potential to give information on both kinematics ( e.g. , los and radial component of the plasma speed from line doppler shifts and doppler dimming ) and thermodynamics ( e.g. , kinetic temperature of the emitting ions from line widths and plasma electron density from line intensity ratios ) of the cmes , that are complementary to those coming from stereoscopic analyses .
in fact , the derivation of some physical plasma parameters from uv spectra , such as the kinetic temperature , is affected by uncertainties related to the unknown 3d distribution of plasma along the los .
these uncertainties can be sensibly reduced , or even eliminated , if the 3d structure of the emitting plasma is determined , for instance , by the stereoscopic images .
therefore , the study of cmes through a combined analysis of stereoscopic data and spectroscopic uv observations allows ( 1 ) to derive new significant information , such as the contributions to the line broadening due to both the bulk expansion and the heating , whose determination would not be otherwise possible without the combination of the techniques of analyses , and ( 2 ) to place important constraints to mhd numerical simulations attempting to investigate the initiation processes and evolution of these events .
although several works use the above mentioned techniques of 3d reconstruction ( i.e. , triangulation and polarization ratio ) to determine the spatial and geometrical configuration of cmes and measure their kinematic parameters ( e.g. , the initial speed or the propagation direction of the cme front with respect to the los ) , no case to our knowledge of the literature presents a direct combination of stereoscopic analysis applied to stereo images with the results of the 3d geometry of cmes obtained by uv spectroscopic observations . in the present work
, we used the stereo / cor1 and soho / uvcs images of the partial - halo cme occurred on 2007 , may 20 .
in particular , we applied the polarization ratio technique to the cor1 data in order to determine the 3d structure of the cme front along the los .
then , for the first time ever , we exploited the geometric information obtained by the 3d reconstruction to synthesize spectral profiles of the o vi @xmath3 line and compare the simulated profiles with those measured by uvcs in order to provide constraints on the coronal plasma heating associated with the transit of the cme .
the paper is organized as follows : in 2 we illustrate the approach we adopted to select the cme event analyzed here ; in 3 we describe the coronagraphic ( from lasco and stereo ) and spectroscopic ( from uvcs ) observations acquired on 2007 may 20 ; then we summarize results obtained in previous works related to the same event ( 4.1 ) , we describe how stereo data have been analyzed to derive the 3d reconstruction of the erupting filament ( 4.2 ) and of the cme front ( 4.3 ) and how these results have been combined with the spectroscopic observations from uvcs ( 4.4 ) .
a discussion of our results concludes the paper ( 5 ) .
the main target of the present work is to demonstrate the possible advantages for the reconstruction of a cme event when both stereoscopic and spectroscopic data are available .
the only instrument that was able to provide daily observations of uv spectra emitted by the intermediate corona is the uvcs spectrometer onboard soho ( see * ? ? ? * ) , now switched off , while other spectrometers on soho ( cds , sumer ) and subsequent solar missions ( hinode / eis ) focused on euv spectra emitted by the lower corona .
data acquired by the latter instruments can not be used for our purposes , because when the cme falls in the field of view of the spectrometer is out of the field of view of the coronagraphs on board soho and stereo , and vice - versa , making contemporary observations not possible .
uvcs data are available in the catalog until december 2012 , but starting from early 2006 some dark columns began to appear and to broad over the uvcs detectors , thus significantly reducing the region available for scientific analysis on the data and in turn the number of detected cmes .
for this reason , the uvcs cme catalog recently released stops at the end of year 2005 ( see * ? ? ? * ) . on the other hand ,
the stereo mission was launched in october 2006 , when the uvcs detector already had some problems , thus making very difficult a contemporary observation of the same cme .
for this reason there are no published works dealing with stereo and soho / uvcs data relevant to the same event , as done here . in order to identify an event observed by uvcs in the stereo era , we performed an automated cross search of the lasco cme catalog @xcite and the uvcs mission log catalog available online from the soho archive .
we restricted our research to all the events listed in the cme catalog that have occurred since october 2006 till december 2010 .
for each event , we verified if the following conditions were satisfied : ( 1 ) uvcs observations were available on the same day ; ( 2 ) the start ( end ) of uvcs observations preceded ( followed ) the time when the cme front reached the altitude of the uvcs field of view ; ( 3 ) at least 20% of the uvcs slit overlapped the area spanned by the cme front during its expansion .
information on the cme time and linear velocity reported in the lasco cme catalog have been used to calculate the time of the transit of the cme front at the heliocentric distance of the uvcs slit reported in the uvcs mission logs . the cme central position angle and the cme average angular width
have been used to compute the width of the area spanned by the expanding cme front .
the final result is that only one cme event , occurred on 2007 may 20 , fulfills these three criteria . owing to the more and more degraded state and limited field of view of uvcs detectors since year 2006 , it is very difficult that data acquired after december 2010 can be used to study large scale eruptive events like cmes . for this reason
the event of 2007 may 20 could be the unique possibility we have to perform a detailed combination of stereoscopic and spectroscopic analysis of a cme .
in addition , this event gave origin to an icme which was observed _ in situ _ by stereo and the wind satellite ( see 4.1 ) and which appeared to have interacted strongly with the ambient solar wind , triggering minor geomagnetic storms since may 22 , as reported in the may 24 bulletin on solar and geomagnetic activity released by the solar influences data analysis center ( sidc ; http://sidc.oma.be ) .
therefore , the analysis of this cme can also provide further insight into space weather and its applications .
the 2007 may 20 cme , classified as partial halo in the lasco cme catalog , occurred after the eruption of a filament from active region noaa 10596 , starting at 04:30 ut .
the event was observed simultaneously by the euvi and cor1 imagers operating on board the stereo spacecraft and the uvcs spectrometer on board soho . at that time
the separation angle between stereo spacecraft was quite small ( about 8.7@xmath4 ) , hence stereo and soho euv imagers viewed approximately the same portion of the solar disc .
the source active region ( ar ) was located close to the disc center and exhibited a quite complex quadrupolar magnetic field configuration , as observed by soho / mdi ; the eruption was probably triggered by the emergence of a fifth , compact negative flux in the wake of the active region starting about 14 hours before the cme .
the ejection of filament material was observed in the he ii 304 line ( sensitive to plasma with temperature of @xmath5 mk ) with the extreme ultraviolet imager ( euvi ) telescopes on board stereo , between 04:30 ut and 06:00 ut ( see figure [ euvi ] ) .
before the eruption , the filament appeared to be mostly aligned with the north - south direction . during the eruption ,
an x - ray event of class b9 from the source active region was detected by the goes instrument , which measured an enhancement of the soft x - ray flux approximately from 05:00 to 08:00 ut , with peak at @xmath6:56 ut .
the flare was visible in the fe xii line ( 195 , corresponding to a temperature of about 1.5 mk ) with euvi as a rapid brightening of the ar loop system , starting at 04:52 ut .
the cme leading edge was first detected in white - light at 06:20 ut in stereo / cor1-a images and at 06:48 ut in soho / lasco - c2 images ( see figure [ lasco ] ) .
subsequent images show the characteristic three - part structure observed in many cmes , with a quite faint , hemispherical front expanding southward at @xmath7 km s@xmath8 ( as reported in the online lasco cme catalog ) , followed by the void and a more dispersed core plasma .
uvcs observations started at 00:38 ut of may 20 and lasted until 17:10 ut . the slit field of view ( fov ) was centered at a heliocentric distance of 2.4 @xmath9 and at a central latitude of 55@xmath4 s , as shown in figure [ lasco ] .
spectral profiles were acquired in the wavelength ranges of 1017 - 1047 , 997 - 999 , and 953 - 981 with a spatial binning of 8 pixels ( 56 ) , spectral binning of 2 pixels ( 0.198 ) , and exposure time of 120 s. owing to the poor statistics , only the o vi @xmath1 , 1037.61 doublet lines had a significant signal - to - noise ratio to be identified .
as stated above , several dark columns are affecting the uvcs detector .
in particular , in our case no signal was recorded in three spatial strips at bins 26 , 19 - 21 , and 11 - 14 ( corresponding to latitudes of @xmath10s , 48-@xmath11s , and 37-@xmath12s , respectively ) .
the total counts relevant to each dark strip are actually accumulated by the detector in one of the adjacent bins ( 25 , 17 , and 9 , respectively ) ; therefore we were able to recover the signal of each dark column in a strip by replacing that column with the array containing the total counts of the strip divided by the number of bins in that strip .
the result of this procedure , however , is that every column within the same strip will have the same signal ( i.e. , the same spectrum ) .
nevertheless , it is worth noting that just the narrowest of the three dark strips , located at bin 26 , falls in the region of the detector crossed by the cme ( see below ) , while the strip at bins 19 - 21 only marginally affects the outer part of the cme front .
thus the loss of information caused by this average process is not significant for the purposes of our work .
the spectral profiles of the o vi lines have been integrated after usual standard calibration and background subtraction .
a pre - cme coronal spectrum averaged over the pre - cme exposures has also been subtracted in order to exclude any possible contributions coming from the ambient quiet corona .
figure [ uvcs_int ] reports a 2d map of the total o vi @xmath3 line intensity observed at different latitudes along the uvcs slit ( @xmath13-axis ) and at different times ( @xmath14-axis ) over the slit region crossed by the cme . as it is well known , the emergent intensity in the o vi doublet lines is a combination of collisionally excited and resonantly scattered components .
the collisional component is due to ion excitation by collisions with thermal coronal electrons and depends on the los distribution of electron density and temperature .
the radiative component is produced by absorption of chromospheric radiation and depends , in addition , on the plasma flow velocity ( see , e.g. , * ? ? ?
* ) : when plasma flows are not negligible , the absorption profile of o vi ions is doppler - shifted with respect to the disk exciting profile and the scattering is less efficient , resulting in a reduction in intensity of the scattered radiation this effect is known as doppler dimming ( see * ? ? ?
* ; * ? ? ?
if the outflow velocity is sufficiently high to produce significant doppler - dimming , then the ratio between the intensities of o vi @xmath01031.91 , 1037.61 line intensities can be used to determine the ion outflow velocity , as originally pointed out by @xcite . an approach based on these arguments and widely adopted in the analysis of uvcs data ( see , e.g. , reviews by * ? ? ?
* ; * ? ? ?
* ) has been used in our previous paper ( * ? ? ?
* hereafter paper i ) to estimate the radial component of the outflow velocity of the plasma embedded in the front and in the core of the cme ( see later ) .
the o vi intensity map reveals the typical three - part structure that is often observed for most of cmes ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ) : the leading loop - like front , which crossed the uvcs slit at 06:54 ut , i.e. , about 34 minutes after it appeared in the fov of cor1 , the dark void , corresponding to a sudden drop in the observed o vi emission , and then the core .
the 2007 may 20 event has already been studied in earlier works by different authors , with different goals .
@xcite used a method based on height - time measurements of coronal moving features applied to stereo / cor1 data to measure the direction of propagation as well as the real speed of this cme .
they estimated an initial unprojected speed of 548 km s@xmath8 , whereas the projected velocities measured in stereo - a and -b spacecraft were 242 and 253 km s@xmath8 , respectively .
the interplanetary counterpart of the may 20 cme was studied in detail by @xcite using _ in situ _ data from stereo and the wind satellite .
the magnetic cloud ( mc ) associated with this event was observed on may 23 at a distance from the sun of 0.96 au ( astronomical unit , @xmath15 km ) .
the magnetic field configuration of the icme was derived by solving numerically the grad - shafranov equation @xcite on the stereo - a data ; the orientation of the streamer - belt neutral line and the orientations of the filament axis and the post - eruptive arcades appear to correlate with the orientation of the mc axis .
the estimated leading - edge speed of the mc was 535 km s@xmath8 .
an accurate approximation of the icme transit time made by applying existing models for cmes propagation through interplanetary space gave the result that if the initial speed of the 20 may cme determined by @xcite is used , the arrival time of the icme predicted at the stereo - a location differs by about 8 hours from the measured one .
lcc + & front & core + @xmath16 ( km s@xmath8 ) & 289 & 231 + @xmath17 ( km s@xmath8 ) & 500 & 535 + @xmath18 ( km s@xmath8 ) & 578 & 583 + @xmath19 ( km s@xmath8 ) & 75 & 85 + @xmath20 ( @xmath4 ) & 30 & 23 + in our previous paper ( paper i ) , we derived from uvcs data the temporal evolution of the o vi line intensities and widths along the direction of propagation of the cme , and measured the velocity components on the pos , along the los , and in the radial direction , as well as the o vi kinetic temperature , for both the front and the core of the cme .
these results are summarized in table [ table1 ] .
we were also able to estimate the direction of propagation of the cme front with respect to the los , using the measured los and pos components of the outflow velocity , which turned out to be @xmath21 .
these measurements were then compared with first results from the 3d reconstruction of the prominence kinematics using triangulation technique applied to stereo / euvi data . in
what follows , we better explain this method and report more complete results on the reconstruction of the erupting prominence .
the three - part structure of the 2007 may 20 cme clearly visible in the o vi intensity map derived from uvcs data , is typical of many cmes .
the leading edge , the dark cavity , and the bright core , are generally associated with the compressed solar plasma ahead of the ejecta , the erupting magnetic flux rope , and the cool and dense prominence plasma , respectively ( e.g. , * ? ? ?
* ; * ? ? ?
. however , a convincing identification of the white - light core of the cme with the cool prominence material is relatively rare and , although recent studies have demonstrated a strong connection between prominence eruptions and cmes , the correlation is not always one - to - one ( see * ? ? ?
in order to investigate possible correlations between the erupting prominence and the cme structure , we applied one of the available stereoscopic techniques to reconstruct the trajectory and kinematics of the filament and compared the results with those derived from uvcs data and already presented in paper i. according to @xcite , if a cme feature can be identified in both images of a simultaneous pair , such as those recorded by stereo , then the position of the feature in a three - dimensional heliocentric coordinate system can be determined by triangulation , provided that the location and separation of the two stereo spacecraft are known . for a constant error in locating the feature on the 2d images , @xmath22 , the error in the depth estimate of the 3d reconstruction , @xmath23 , increases as the separation angle @xmath24 between
the two stereo spacecraft decreases : @xmath25 ( see * ? ? ?
* ; * ? ? ?
however , a reliable identification of the same feature in the two stereo images can be achieved only if the separation angle is quite small ( @xmath26 ) . for 2007
may 20 , the separation angle was @xmath27 , leading to an error in the height determination of @xmath28% , if the feature could be located to within one pixel .
the reconstruction results presented in this section were obtained by manually locating the same feature in two simultaneous stereo images ; this is often referred to as `` tie - pointing '' technique .
as explained in paper i , the erupting filament was reconstructed applying this technique for 7 different times on 20 may to study its evolution , using running differences of stereo / euvi 304 image pairs in order to enhance the visibility of the prominence .
generally , the filament is seen in euvi 304 , responsive to a temperature of @xmath5 k , as a dark absorbing feature above the surrounding chromospheric plasma .
figure [ prominence ] shows a 3d representation of the reconstructed trajectory of the filament identified in several euvi frames , as observed from three different prespectives .
detailed latitude vs. altitude ( i.e. , as seen side - on ) and altitude vs. longitude ( i.e. , as seen face - on ) diagrams of the filament trajectory are reported in figure [ tie_pointing1 ] . at the beginning of the eruption ( 05:31 ut )
, the filament was laying at a heliocentric distance of @xmath29 @xmath9 above the solar surface and was mostly aligned with the north - south direction . in the following hours , the filament expanded southward and westward , reaching
the maximum height of about 2.1@xmath9 at 07:11 ut . in the last phase of the eruption , between 06:51 ut and 07:01 ut , it underwent a rapid counter - clockwise rotation of about 30@xmath4 around a vertical axis parallel to the los ( see right panel of figure [ tie_pointing1 ] ) .
note that there have been other recent observations of similar rotational dynamics in the low corona ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
figure [ tie_pointing2 ] shows a polynomial 2nd - order extrapolation of the projected altitude and los velocity of the filament at the time of uvcs observations .
the extrapolated quantities match very well the measured ones , in particular the velocity component along the los of the reconstructed filament is in very good agreement with the value of @xmath30 km s@xmath8 derived from the doppler shift of o vi lines .
this result suggests that the bright emission structure identified as the cme core in uvcs images ( see figure [ uvcs_int ] ) actually originates from the prominence plasma embedded in the expanding flux rope .
note that a convincing identification of the core of the cme with the cool prominence material is relatively rare @xcite . as already outlined , in the present work we carried out the first ever investigation of the cme properties , by combining the results of the uvcs spectroscopic observations with the geometric information that can be derived from the stereo images .
to perform the 3d reconstruction of the cme expanding front , we applied the polarization ratio technique ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ) to the stereo / cor1 data .
it is known that the white - light emission in the solar k - corona is originated by thomson scattering of photospheric light produced by free electrons .
detailed description of the thomson scattering theory can be found in various papers ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in particular , it has been demonstrated that the degree of polarization of thomson - scattered white - light is a function of the scattering angle between the incident light direction and the direction towards the observer @xcite , that , in turn , can be related through simple geometric considerations to an effective distance of the scattering electron from the pos ( see * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) .
therefore , the ratio between polarized brightness ( pb ) and unpolarized brightness ( ub ) can be used to estimate pixel - by - pixel the location of the scattering coronal plasma along the los .
we used total and polarized brightness images obtained by the stereo / cor1 ahead and behind observations of the event .
we considered three different times during the expansion of the cme , 06:50 ut , 07:20 ut , and 07:50 ut , which cover the whole transit of the cme through the uvcs field of view .
figure [ cor1 ] reports the sequence of base - difference images from stereo / cor1 at the three considered times , showing the progressive expansion of the cme from the two viewpoints of the ahead and behind spacecrafts .
as it can be noted from a comparison with figure [ lasco ] , where the uvcs fov is superimposed on the images to highlight its position with respect to the cme features , at 07:20 ut the intersection between the cme front and the uvcs field of view was larger , hence , for our discussion , we apply the polarization ratio technique to this time frame .
recently , @xcite have shown that the contribution of the h@xmath31 radiation to the total white - light emission in cor1 coronagraphic images ( whose bandpass , at variance with lasco , also includes the h@xmath31 @xmath32 line ) can produce artificial 3d features ( called `` h@xmath31 horns '' ) in the 3d reconstruction by the polarization ratio technique .
the h@xmath31 horns are inconsistent with the real cme structure and they usually appear to be located far from the pos of the coronagraph .
the reason is that the h@xmath31 emission coming from the cme core is not thomson - scattered radiation and results in a polarization decrease of the white - light ( see , e.g. , * ? ? ?
we needed to exclude this unwanted contribution to determine an accurate 3d cme structure . in particular , to prevent the h@xmath31 contamination , we selected in the cor1 images the area corresponding to the cme front and restricted our analysis to that region .
the ratio pb / ub is calculated pixel - by - pixel and converted into the effective angular distance from the pos along the los ( @xmath33 coordinate ) by using both stereo a and b images . owing to the forward / backward symmetry of thomson scattering
, the brightness ratio does not indicate whether the scatterer is in front or behind the pos ( see * ? ? ? * ) .
however , this ambiguity can be resolved in this case by taking into account that the cme was cataloged as a partial - halo and its source ar was located in the visible hemisphere of the sun , so that the cme was earth - directed ; therefore , the scattering electron turns out to be located in front of the pos ( @xmath34 ) .
in figure [ pb_ratio ] we display the results of the 3d reconstruction , as viewed from three different perspectives .
the blue and red dots represent the 3d distribution of scattering locations for the cme front derived from cor1/stereo a and b pixels , respectively . the analysis of each cor1 image provided about 18000 points in the 3d
coordinate system having the @xmath33 axis directed towards the earth , and the @xmath14 and @xmath13 axes along the east and north directions , respectively .
it is particularly evident from the top panel in the figure that each of the two point distributions is mainly concentrated around the direction of the line of sight of the corresponding spacecraft ( as already found , e.g. , by * ? ? ?
in order to obtain more detailed information on the 3d structure of the cme front , we applied a statistical analysis to the whole distribution of identified scattering locations .
first , we divided the 3d coordinate system into cubic cells with 0.1 @xmath9 side and measured the number of scattering points included in each cell , @xmath35 ( the indexes @xmath36 , @xmath37 , and @xmath38 denote the position of each cell in the 3d space ) ; then , for each column in the @xmath33 direction , defined by a particular @xmath39 pair , we calculated the average of the resulting @xmath40 distribution , @xmath41 as well as the standard deviation of the distribution , @xmath42 .
quantities @xmath43 and @xmath42 have been computed by averaging over all the 3d points obtained by both stereo - a and -b spacecraft images .
the results are reported in figure [ front_geom1 ] .
the left panel shows a topographical 2d map of the cme front displaying the average distance of the scattering locations from the @xmath44 plane ( the quantity @xmath43 ) for each column in the los direction .
the color scale indicates different distances from the pos in solar radii .
the identified points are distributed between 0 and 12 @xmath9 .
the front turns out to be similar to a `` curtain '' of plasma or a loop arcade system expanding towards the earth ( similar to the results reported by * ? ? ?
it is worth noting that the greatest distances ( @xmath45 @xmath9 ) are achieved at the edges of the front .
the right panel reports the corresponding standard deviation of the distribution , that we interpret as the depth along the los of the cme front .
the front depth is up to @xmath46 @xmath9 , and the maximum value is achieved in the same regions that appear to be more distant from the pos . in order to use these geometric parameters in correlation with uvcs data
, we extrapolated the mean cme front position and the mean front depth at the location of the uvcs field of view .
figure [ front_geom2 ] reports these quantities as functions of the position along the uvcs slit .
note that the cme front is located at @xmath46 @xmath9from the pos at the left edge of the slit , which is located near the center of the front ( see figure [ front_geom2 ] ) , and progressively at lower distances from the pos towards the right .
correspondingly , the los depth of the front is maximum at left edge of the slit ( @xmath47 @xmath9 ) , then it decreases .
typically , one of the signatures of cme fronts in uv and euv spectra is the broadening of line profiles that are observed during their transit ( see , e.g. , * ? ? ?
* ; * ? ? ?
both expansion of the emitting volume and increasing kinetic temperature can contribute to broaden the line profiles .
the contribution of plasma bulk expansion to the line width at the front can be significant in particular for halo and partial - halo cmes , due to geometrical effects related to the relative angle between the los and the flow velocity .
hence , in order to correctly interpret the measured line widths at the cme fronts , one needs to disentangle bulk expansion and thermal broadening .
@xcite , for instance , attempted to discriminate between the two components for several observed halo cmes by assuming a spherical shell model of the expanding front , by estimating the component of the broadening due to bulk expansion using information on the los and pos speeds , and by comparing the result with line widths of the o vi spectral profiles measured with uvcs .
although the authors were able to attribute line broadening to plasma heating in some cases and to the los component of the cme expansion in other cases , this approach has some limitations , most related to the assumption of the cme front shape . in this section
we address this issues by combining for the first time the results derived with uvcs spectroscopic data and those derived with stereo 3d reconstructions . in particular , the information on the position and depth of the cme front along the los derived with the pb - ratio technique applied to stereo / cor1 images are used for the synthesis of spectral profiles of the o vi @xmath481031.91 line to be compared with the measured line profiles
. it is well know that , since coronal plasma is optically thin , uv spectra in the solar corona originate from the superimposition along the line of sight of radiation emitted locally by plasma elements characterized by different electron density , electron and kinetic temperatures , and flow velocity . therefore , in order to build up a synthetic profile of a particular spectral line , one has to make assumptions on the distribution along the los of those plasma quantities . in our case
, we performed a synthesis of the o vi @xmath481031.91 line considering in particular the lines of sight defined by the spatial bins of the uvcs field of view , limiting the calculation to the portion of the uvcs slit that has been effectively crossed by the front of the cme at 07:20 ut of may 20 ( i.e. , about 50% of the slit ) . for a given bin ( i.e. , for a given los ) , we made the simplifying assumption that all plasma elements contributing to the o vi total line profile for that los emitted a purely gaussian spectral profile given by the following expression : @xmath49 ^ 2\right\},\ ] ] where the index @xmath50 denotes the @xmath50-th bin of the uvcs slit , @xmath51 is the rest wavelength of the o vi line transition
, @xmath52 is the light speed , @xmath53 and @xmath54 are the assumed functions giving the variation along the los of the o vi emitted intensity and flow velocity , and @xmath55 is the assumed width of the contributing profile , corresponding to an assumption on the kinetic temperature of the o@xmath56 ions .
the functions @xmath53 and @xmath54 were set up _
ad hoc _ and iteratively adjusted until the best agreement between the full width at half maximum ( fwhm ) of the synthetic profiles and that of the observed profiles was found , within the uncertainties .
the functions giving the best result have the following expressions : @xmath57,\ ] ] @xmath58 where @xmath59 and @xmath60 are the average position and the width of the cme front derived as explained above for the @xmath50-th bin along the uvcs slit , @xmath61 km s@xmath8 is a scale factor , @xmath62 is a parameter that is introduced to take into account asymmetries of the observed spectral profiles , and @xmath63 is the front velocity measured from the doppler shift of the o vi @xmath64 line using uvcs data .
note that the peak intensity distribution along the los is a gaussian centered at the position of the cme front and having a @xmath65 full width that is 10% of the front width ; this indicates that the los depth of the cme front in uv can be narrower than that derived from white light data , probably because o vi emission is efficient only in the very interior of the front due to plasma density and temperature distributions .
the velocity function is such that plasma elements located ahead ( behind ) the cme front have flow speeds larger ( smaller ) than that of the front itself , i.e. , the front is not expanding rigidly .
few considerations must be done about the assumption on the line width of the contributing profiles @xmath55 , i.e. , on the kinetic temperature of the emitting plasma elements . for the sake of simplicity
, we assumed that the distribution of line widths is uniform both along the los and the uvcs slit , i.e. , @xmath66 , corresponding to an effective o@xmath56 temperature of @xmath67 k. this temperature is quite higher than typical values measured from uv data in the front of cmes ( @xmath68 mk ; see , e.g. , * ? ? ?
* ) , but it is @xmath69% lower than the temperature measured from data analyzed here in the undisturbed pre - cme corona at the same altitude ( @xmath70 k , as reported in paper i ) .
it is worth noting that the assumption of uniform temperature along the los could not be representative of the real solar corona due to the possible superimposition along the los of different structures characterized by different temperatures ; in our case , however , the bulk of the o vi emission comes from a relatively narrow region ( @xmath71 @xmath9wide ) centered around the cme front , as explained above , so this assumption can be considered fairly valid .
the total synthetic profile for a given bin of the uvcs field of view is obtained by integration along the los of the elemental contributing profile : @xmath72 this integral is turned into a sum considering that the @xmath33 coordinate is a discretized variable .
the fwhm of the total profile is then compared as function of the position along the uvcs slit , with the measured fwhm of the observed o vi profiles . in order to enhance the statistics and increase the signal - to - noise ratio , we accumulated the observed profiles over 8 exposures ( corresponding to a 16 mins average ) around 07:20 ut .
line profiles were integrated and the measured fwhms were corrected for instrumental broadening as usual ( see , e.g. , * ? ? ?
the resulting observed and synthetic fwhms are shown in the plot of figure [ fwhm ] as function of the heliocentric distance along the uvcs slit .
note that points at higher altitude are located at the front center , while points at lower altitude are located in the peripheral part of the front ( see , for comparison , figure [ front_geom1 ] ) .
the uncertainties in the observed fwhms were computed with the standard error propagation formulae and are plotted as error bars in the figure .
the observed line widths are well reproduced ( within the uncertainties ) by the synthetic ones only in the peripheral part of the cme front . at the front center
there is no agreement , and the observed fwhms are @xmath2% higher than the synthetic ones . any other intensity and velocity distribution along the los able to reproduce
the fwhms observed at the front center was unable to reproduce at the same time the fwhms measured at other locations along the slit .
when the fwhm is converted into effective temperature of the o@xmath56 ions , it turns out that the observed line widths imply a temperature that is @xmath73 times higher than the temperature corresponding to the synthetic line widths . the inability of reproducing the observed line widths at the front center through the synthesis of spectral profiles that takes into account both thermal and non - thermal components of the line widths suggests that , at least at the front center , physical mechanisms other than bulk expansion are responsible for line broadening and provide the additional heating that is necessary to increase the plasma temperature up to the observed values .
one of the possible mechanisms is represented by plasma compression that most likely occurs in the central part of the cme front due to the relatively rapid expansion of the eruption in the pre - cme medium .
it is worth noting that observed line widths at the peripheral part of the cme front were reproduced by assuming a 70% lower kinetic temperature with respect to the pre - cme corona .
this can indicate that the cme front is made of plasma dragged by the expansion of the flux rope and coming from lower layers of the solar corona , characterized by lower temperatures .
this association between the cme front plasma and the ambient coronal plasma dragged by the outward expansion of the cme core is a common characteristic of several cme models ( e.g. , the flux - rope models or the breakout models ; see forbes , 2006 ) , and our results can provide clear evidence for that .
in this work we carried out the first ever analysis of a partial - halo cme by combining the results from the uvcs spectroscopic observations with the 3d geometric information that can be derived from the stereo data .
we studied the first cme observed simultaneously by stereo and soho / uvcs : this event occurred on 2007 , may 20 and was visible from about 04:30 ut to 09:00 ut . we performed 3d reconstructions of both the erupting prominence associated with the cme and the cme front using euvi and cor1 data .
then , we exploited the geometric information derived from the reconstructed cme structure to synthesize spectral profiles of the o vi @xmath481031.91 line and compared them with spectral profiles observed by uvcs during the transit of the cme . 1 .
we applied the tie - pointing technique to the stereo / euvi 304 images of the cme source active region in order to reconstruct the 3d kinematics of the erupting filament . during the eruption ,
between 06:51 ut and 07:01 ut of may 20 , the filament , which was initially oriented in the north - south direction , underwent a rapid counter - clockwise rotation of about 30@xmath4 around a vertical axis parallel to the los .
a comparison of a polynomial extrapolation of the projected altitude and los velocity of the filament at the time of uvcs observations with the altitude of the uvcs field of view and the los velocity component derived for the cme core suggests that the bright emission structure identified as the cme core in white - light images and uv data actually originates from the prominence plasma embedded in the expanding flux rope .
we applied the polarization ratio technique to the stereo / cor1-a and -b data at 07:20 ut of may 20 in order to perform the 3d reconstruction of the cme expanding front .
the scattering locations identified with this method are distributed between 0 and 12 @xmath9from the pos .
the cme front turns out to be similar to a `` curtain '' of plasma or a loop arcade system expanding towards the earth with the greatest distances achieved at the edges of the front .
the los depth of the front is up to @xmath46 @xmath9 ; the front appears wider at the center , where it is also more distant from the pos .
we used the geometric information derived by the polarization - ratio analysis to synthesize spectral profiles of the o vi @xmath481031.91 line expected at different positions along the uvcs field of view and taking into account bulk expansion of the emitting volume by assuming an intensity and velocity distribution of the plasma elements along the los and a constant kinetic temperature of the o@xmath56 ions .
the synthetic line widths well reproduce ( within the uncertainties ) the observed ones only in the peripheral part of the cme front , while at the front center , where the distance from the pos is greater , they turn out to be @xmath2% lower than the others . by converting the synthetic line widths into effective temperatures ,
the front center is characterized by a temperature @xmath73 times higher than that of the peripheral regions .
this indicates the presence of a significant heating mechanism responsible for line broadening other than bulk expansion and los integration .
the occurrence of plasma heating is , together with the detection of type - ii radio bursts , a good indication of cme - driven shock propagation .
in particular , plasma heating detected here is an increase by a factor @xmath73 in the o@xmath56 ion kinetic temperatures at the center of the cme front . on the other hand ,
this event was not associated with any type - ii radio burst , and this indicates that no shock was excited by the expansion of the cme front .
this is also in agreement with the quite low front velocity we derived : when the front was at the unprojected heliocentric distance of ( @xmath74 @xmath9)@xmath75 @xmath9 , its velocity was 578 km s@xmath8 , hence smaller than local alfvn velocities ( @xmath76 km s@xmath8 ) expected at these altitudes ( see , e.g. * ? ? ?
hence , we may argue that plasma heating inferred here is simply due to plasma compression : for instance , under the assumption of adiabatic compression of a perfect gas , a temperature increase by a factor @xmath77 corresponds to a density increase by a factor @xmath78 , which is quite a reasonable value .
interestingly , the lack of significant plasma heating we detect at the peripheral region of the front indicates that stronger plasma compression is occurring only at the front center .
this behaviour is also similar to what recently found by @xcite for the adiabatic heating occurring across a cme - driven shock .
overall , this work demonstrates the potential of combining 3d kinematical information from stereoscopic and polarized data with spectroscopic observations providing a new capability for a better determination of the cme 3d structure , unprojected expansion velocity and associated plasma heating .
unfortunately , the lack of uv spectrometers of the extended corona ( like soho / uvcs ) in the payload of the next accepted solar missions will prevent the possibility to apply the analysis described here to future data .
alternative possibilities could be offered by future coronagraphs whose fields of view will extend well below 2 @xmath9 , thus overlapping with spectroscopic observations provided by future solar missions ( like solar orbiter / spice ) .
this work has received funding from the european commission s seventh framework programme ( fp7/2007 - 2013 ) under the grant agreement swiff ( project no .
263340 , www.swiff.eu ) and from the agenzia spaziale italiana through the contract asi / inaf no .
i/013/12/0 .
1031.91 intensity evolution along the uvcs slit ( @xmath13-axis ) at different times ( @xmath14-axis ) as observed at 2.4 @xmath9during the transit of the cme .
the typical three - part structure of the cme is clearly visible.[uvcs_int ] ] ( horizontal solid line ) .
start and peak time of the solar flare associated with the cme are also shown for comparison purposes .
right : same as left panel , for the real and los projected velocity of the filament.[tie_pointing2 ] ] plane ; left panel ) and average depth along the los ( @xmath33 axis ; right panel ) of the cme front , as derived from a statistical analysis applied to the point distribution shown in figure [ pb_ratio ] ( see text).[front_geom1 ] ] | a three - dimensional ( 3d ) reconstruction of the 2007 , may 20 partial - halo coronal mass ejection ( cme ) has been made using stereo / euvi and stereo / cor1 coronagraphic images .
the trajectory and kinematics of the erupting filament have been derived from euvi image pairs with the `` tie - pointing '' triangulation technique , while the polarization ratio technique has been applied to cor1 data to determine the average position and depth of the cme front along the line of sight .
these 3d geometrical information have been combined for the first time with spectroscopic measurements of the o vi @xmath01031.91 , 1037.61 line profiles made with the _ ultraviolet coronagraph spectrometer _ ( uvcs ) on board
soho .
comparison between the prominence trajectory extrapolated at the altitude of uvcs observations and the core transit time measured from uvcs data made possible a firm identification of the cme core observed in white light and uv with the prominence plasma expelled during the cme .
results on the 3d structure of the cme front have been used to calculate synthetic spectral profiles of the o vi @xmath1 line expected along the uvcs slit , in an attempt to reproduce the measured line widths . observed line widths can be reproduced within the uncertainties only in the peripheral part of the cme front ; at the front center , where the distance of the emitting plasma from the plane of the sky is greater , synthetic widths turn out to be @xmath2% lower than the measured ones .
this provides strong evidence of line broadening due to plasma heating mechanisms in addition to bulk expansion of the emitting volume . |
recently , the next - to - leading order ( nlo ) power correction has been shown an important role in understanding the exclusive processes @xmath11 @xcite and @xmath12 @xcite . the method for calculating the nlo power correction
is called the collinear expansion @xcite .
this power expansion method is compatible with pqcd factorization @xcite , which demonstrates the amplitude as convolutions of perturbatively calculable hard scattering amplitudes ( the hard function ) and nonperturbative hadronic wave functions ( the soft function ) .
furthermore , it is a feynman diagram approach such that the partonic interpretation for the nlo power correction can be preserved .
the asymptotic limit of the @xmath13 transition form factor to be @xmath14 with @xmath15 mev , is about @xmath16 higher than the upper end of the cleo data @xcite .
this deviation can be successfully explained by nlo power correction .
in addition , nlo power correction can well describe the low energy portion of the cleo data .
our purpose in this paper is to generalize the approach for the @xmath13 form factor to the @xmath1 and @xmath2 form factors .
the first problem we shall face is that there are many independent degrees of freedom associated with the @xmath17 and @xmath18 lowest valence fock states .
the @xmath17 and @xmath18 mesons are admixtures of the @xmath19 octet and singlet states , this counts eight quantities : four wave functions and four related decay constants .
the @xmath17 and @xmath18 mesons can also receive contributions from the @xmath20 anomaly to have intrinsic heavy quark and gluon contents . to reduce the number of independent degrees of freedom , we invoke phenomenological constraints and physical assumptions over the wave functions and decay constants . for the phenomenological constraints
, we shall employ the large momentum transfer data for @xmath1 and @xmath21 transition form factor @xcite , the two - photon decay widths of the @xmath17 and @xmath18 mesons and the ratio @xmath22 of the @xmath23 and @xmath24 decay widths .
as for the physical assumptions we shall invoke the @xmath19 octet - singlet mixing scheme . in this mixing scheme , both @xmath17 and @xmath18 are linear combinations of @xmath25 and @xmath26 , the octet and singlet states in the @xmath19 representation .
the mixing is controlled by the mixing angle . in the one mixing angle scheme , in which the octet and singlet meson states have different decay constants @xmath27 and @xmath28 and share a common mixing angle @xmath29 , the mixing angle @xmath30 is in range of @xmath31 @xcite . in recent years , many evidences @xcite have indicated that a two mixing angle scheme , in which the octet and singlet mixing angles @xmath32 and @xmath33 can take different values , is more general than the one mixing angle scheme .
the @xmath1 and @xmath2 form factors have been investigated by either one mixing angle scheme @xcite or two mixing angle scheme@xcite .
we would like to make a more refined analysis for these form factors by including nlo power correction . in this paper
, we shall investigate the @xmath1 and @xmath2 form factors and the mixing pattern of @xmath3 system by employing pqcd formula with nlo power correction in the standard hard scattering approach .
we shall employ the collinear expansion to derive nlo ( @xmath34 ) power correction for @xmath1 and @xmath2 form factors in sec .
ii . in sec .
iii we shall analyze the high momentum transfer data for form factors , the two photon decay widths and the ratio @xmath22 .
the values of the mixing parameters are determined from a @xmath5 analysis for the data .
iv is devoted to conclusions .
our strategy in calculating the power corrections to the @xmath17 and @xmath18 meson - photon transition form factor is to invoke the collinear expansion @xcite . for simplicity
, we shall first ignore the meson mass effects . that is we choose the momentum of the initial state meson , @xmath35 , and that of the final state photon , @xmath36 , as p_1^&=&(q,,0_)p^+n^p^ , p_2^&=&(0,,0 _ ) n^ , such that the virtual photon has momentum @xmath37 with virtuality @xmath38 to make pqcd applicable .
vectors @xmath39 and @xmath40 are in the @xmath41 and @xmath42 directions in the light - cone reference frame and have properties @xmath43 and @xmath44 .
@xmath45 denotes the mass of the initial state meson . for the feynman diagrams displayed in figs .
1(a ) and ( b ) , the amplitudes are written as a(p_1,p_2)=where the trace is taken over the color and spin indices and the meson da @xmath46 has expression for @xmath47 meson ( k , p_1,q^2)=d^4 z e^ikz0||q(0)q(z)|p .
we assign the loop momentum @xmath48 for the valence antiquark and let it flow into the hard function . the hard function @xmath49 contains two parton photon interaction vertices and one virtual internal parton propagator .
the amplitude @xmath50 contains leading , next - to - leading and higher twist contributions .
the quantity twist is understood as an effective twist for nonlocal operators and is not exactly the same with the usual twist defined for local operators .
the twist has different meanings for the hard and soft functions . for the hard function , the twist is defined as the power of the inverse of the photon virtuality @xmath51 , and , for the soft function , the twist represents the power of the small scale @xmath52 with magnitude of order @xmath53 . by employing collinear expansion , we can systematically separate the leading twist ( lt ) contributions from the next - to - leading twist ( nlt ) contributions .
the lt contributions are from collinear loop momentum @xmath54 .
it is therefore convenient to parameterize the loop momentum @xmath48 into k^=x p^+ n^+ k_^ , where @xmath48 contains on - shell part k^_l = x p^+ n^+ k_^ and off - shell part k^_s = n^ . in the first step , we expand the hard function @xmath55 with respect to @xmath56 as [ hexp ] h(k)=h(k=)+|_k=(k-)^+ with the help of @xmath57 and @xmath58 , we can factorize the loop parton propagator @xmath59 into its long distance part , @xmath60 , and short distance part ( the special propagator defined in @xcite ) , @xmath61 , which take expressions f_l(k)&= & , f_s(k)= . the propagators @xmath60 and @xmath61 have different physical meanings . to see this , it is amount to consider their propagations on the light cone .
the integrals of @xmath60 and @xmath61 over @xmath62 give f_l(,)&=&e^ikn(-)f_l(k)(- ) , f_s(,)&=&e^ikn(-)f_s(k)(- ) , where @xmath17 and @xmath63 mean the light - cone distances in the @xmath42 direction .
it is obvious that @xmath61 is not propagating on the light cone .
this means that @xmath61 should be included into the hard function . by dimensional
counting , @xmath61 is of order @xmath64 .
therefore , including one @xmath61 into the hard function then increases one twist order for the hard function .
there are different effects as @xmath60 and @xmath61 act on the spin structures of hard function , the terms proportional to @xmath65 or @xmath66 . as @xmath60 acts on @xmath65 ,
its collinear part vanishes and non - collinear parts are retained [ cl1 ] f_l(k)=-f(k)(k-)^(i_)f_s(k ) where minus sign comes from the anti - particle propagator .
the vertex @xmath67 and short distance propagator @xmath61 are then absorbed into the hard function .
the factor @xmath68 is included into the soft function to become a coordinate derivative on the quark fields . as @xmath60 acts on @xmath66 ,
its collinear part contributes to leading order .
the short distance propagator @xmath61 only serves to introduce the interaction term @xmath69 for @xmath65 vertex , where @xmath70 denote the gluon fields .
the total effects of @xmath60 and @xmath61 acting on @xmath65 are to include one @xmath67 and one @xmath61 into the hard function and to absorb the factor @xmath68 and gauge fields @xmath71 into the soft function to become a covariant derivative , @xmath72 with @xmath73 the strong coupling . the contributions from the second term of eq .
( [ hexp ] ) and from figs .
1.(c ) and ( d ) are of twist-6 or higher twist and will not be considered in below discussions .
the reason is that the possible non - vanishing components of @xmath74 in @xmath75 are @xmath76 or @xmath42 , but both vanish as @xmath75 contract with @xmath68 or @xmath77 .
we substitute the first term of eq .
( [ hexp ] ) into the integral with the soft function and apply the identity dx ( x - kn)=1 to convert the loop momentum integral into the fraction variable integral .
the amplitude then becomes , approximately , a(p_1,p_2)dx which contains lt and nlt contributions .
the meson da @xmath78 has the expression ( x)=_0^e^ix0||q(0)q(n)|p .
we now discuss how to separate the lt from the nlt contributions for amplitude @xmath50 . due to the fact that the final state photon is real and has transverse polarization , the hard function @xmath79 can have spin structures : @xmath80 , @xmath81 @xmath82 and @xmath83 where @xmath84 with @xmath85 .
the first spin structure leads to lt contribution , while the second and third ones result in the nlt contributions .
the last spin structure would lead to next - next - to - leading twist contribution and will not be considered below . to calculate the nlt contributions , we need to apply eq .
( [ cl1 ] ) to extract the contributions from non - collinear loop momentum . as a result
, we get the amplitude up to nlt as [ nloa ] adx + dx where the first term of the right hand side of eq .
( [ nloa ] ) comes from the feynman diagrams shown in fig .
1(a ) and ( b ) and the second term from those diagrams shown in fig .
the tensor @xmath86 is defined as @xmath87 .
the nlt hard function @xmath88 is defined as h_(x)&=&(i_)h(x)+h(x)(i _ ) and the nlt meson da @xmath89 has expression ^(x)=_0 ^ 1 dx_1_0^_0^e^i(x_1-x)e^ix0||q(0)d^(n)q(n)|p .
the factorization of momentum integral is finished . to complete the factorization
, we still need to perform the factorizations of the color and spin indices . to separate the color indices , we take the convention that the color factors of the hard function are extracted and absorbed into the soft function . as for the spin indices , we employ the expansion of the soft functions into their spin components & = & _
^ , ^&=&_^ ; , where @xmath90 denotes gamma matrix and @xmath91 is the related spin component of the distribution amplitude . for a given order of @xmath92 , we choose the component @xmath91 with lowest twist .
the determination of the lowest twist @xmath91 can be done as follows .
firstly , we notice that the tensor structure of @xmath91 can be expressed in terms of @xmath39 , @xmath40 , @xmath93 and @xmath94 . the vectors @xmath39 and @xmath40 have dimensions @xmath95=1 $ ] and @xmath96=-1 $ ] with respect to the hard scale @xmath51 .
secondly , note that the matrix element for the soft function is written as ~0|q|p , ^~0|d^ q|p . by the above facts , we can derive a power counting rule as follows .
consider the @xmath97 has the fermion index @xmath98 and the boson index @xmath99 .
the fermion index @xmath98 arise from the spin index factorization for @xmath100 fermion lines connecting the soft function and the hard function and the boson index @xmath99 denotes the @xmath101 power of momenta in previous collinear expansion and the @xmath102 gluon lines as @xmath103 .
we may write [ pc1 ] ^_1_f;_1_b=_i ^_i-1e^_1_f;_1_b_i^i where @xmath52 denotes a small scale associated with da .
spin polarizers @xmath104 denote the combination of vectors @xmath105 , @xmath106 and @xmath107 .
variable @xmath108 represents the twist of da @xmath109 . the restrictions over polarizers
@xmath110 are [ pc2 ] n__je^_1_f;_1_j_b_i=0 which are due to the fact that polarizers @xmath104 are always projected by @xmath111 .
the dimension of @xmath97 is determined by dimensional analysis [ pc3 ] d()=3f+b-1 by equating the dimensions of both sides of eq.([pc1 ] ) , one can derive the minimum of @xmath108 [ pc4 ] _
i^ = 2f+b+[1-(-1)^b ] .
it is obvious from eq.([pc4 ] ) that there are only finite numbers of fermion lines , gluon lines and derivatives contributes to a given power of @xmath92 .
we now demonstrate that the collinear expansion is compatible with the conventional approach for proving the pqcd factorization at one loop order of radiative correction . to show this
, we consider the radiative correction for @xmath79 as displayed in fig .
if the radiative gluon in fig .
3(a ) is collinear with momentum @xmath112 , where @xmath113 , the lower virtual antiquark has the momentum @xmath114 with @xmath115 .
it is obvious that the virtual antiquark in the collinear region behaves similarly to the loop antiquark in the tree amplitude . the collinear expansion for fig .
3(a ) in the collinear region is the same as the expansion for the tree diagram fig .
1(a ) in the leading configuration .
to demonstrate this , we write the integrand for fig .
3(a ) as i_3a&= & & & where we have defined @xmath116 , @xmath117 , h_3a(l)&=&-ie^2[f(l_3)^ ] , _
3a(l)&=&[_f(l_1 ) f(l_2 ) _ ] , and @xmath118 .
we , firstly , expand @xmath119 h_3a(l)=h_3a(l=)+|_l=(l - l)^with @xmath120 .
repeating the same considerations for the expansion of the tree amplitude , we can recast @xmath121 into i_3a = d+d+ , where _
3a()&=&e^i(+x- ln)[_f(l_1 ) f(l_2 ) _ ] , _ 3a^^()&=&e^i(+x - ln ) [ _ f(l_1 ) l^^f(l_2 ) _ ] .
it is obvious that both @xmath122 and @xmath123 are collinear divergent for collinear @xmath124 .
we introduce corresponding soft functions @xmath125 and @xmath126 to absorbe @xmath122 and @xmath123 .
the corresponding tree level hard functions are denoted as @xmath127 and @xmath128 , respectively . if the radiative gluons in fig .
3(a ) are soft , i.e. the gluons have momentum @xmath129 , there are no effects on power expansion .
this is because the eikonal approximation up to @xmath130 can be applied to factorize the soft gluons from the valence quark propagator .
the other two particle reducible diagrams figs .
3(b ) and ( c ) can be dealt with similarly . it is noted that the double logarithms in figs .
3(a ) to ( c ) arising from the mixing contributions from the soft and collinear divergences are cancelled each other . in light - cone gauge @xmath131 ,
3(d ) and ( e ) in collinear region are more suppressed than figs .
3(a ) to ( c ) in collinear region by , at least , @xmath132 .
after subtracting the soft contributions ( the soft and collinear divergences ) from the one loop radiative correction diagrams figs .
3(a ) to ( c ) , we can obtain the one loop corrected hard functions ( lo and nlo ) @xmath133 and @xmath134 , separately .
the analysis for the radiative corrections to @xmath88 is simple , since it only involves radiative correction and has no need to consider collinear expansion .
the diagrams for the radiative corrections to @xmath88 are shown in fig .
4 . as a result ,
up to first order in radiative and power corrections , we can arrive at the factorized amplitudes as a^(0)+a^(1)(h^(0)_0+h_0^(1))(_0^(0)+_0^(1))+(h_1^(0)+h_1^(1))(_1^(0)+_1^(1 ) ) where the superscript indices @xmath135 denote the order of radiative correction and the subscript indices @xmath136 mean the order of power correction .
the notation @xmath137 represents the convolution integral and the trace over the color and spin indices . to prove pqcd factorization
, we need to generalize the one loop factorization to arbitrary orders .
it can be done straightforwardly @xcite . for convenience , we may write the amplitude as @xmath138 where @xmath139 denotes the polarization vector of the final state photon .
the form factors are expressed in terms of the octet and singlet components [ nloff ] f_p(q^2)=_i=8,1a^p_i_pf_p_i(q^2 ) where the expansion coefficients @xmath140 depend on the mixing scheme ( see next section ) . due to the @xmath3 mixing
, we take the octet and singlet states as the basis for our investigation of the form factors . the superscript @xmath141 or
@xmath142 denote the contributions from octet or singlet current ( see below notation ) .
the leading order of @xmath143 is calculated from figs.1(a ) and ( b ) and takes expression f_p_i^(q^2)=4c_i_0 ^ 1 dx , where the charge factors are defined as @xmath144 and @xmath145 .
the nlo of @xmath143 is evaluated from fig.2 @xmath146}{q^4x(1-x)}\ .\end{aligned}\ ] ] we have taken the symmetry between the exchange of @xmath147 for @xmath148 , @xmath149 and @xmath150 .
the relevant das are expressed explicitly as follows @xmath151 where nonlocal currents are defined as j^8_5(0,n)&=&[|u(0)_5u(n)+|d(0)_5d(n)-2|s(0)_5s(n ) ] , j^1_5(0,n)&=&[|u(0)_5u(n)+|d(0)_5d(n)+|s(0)_5s(n ) ] , j^8_(0,n , n)&=&[|u(0)_d_(n)u(n)+|d(0)_d_(n)d(n)-2|s(0)_d_(n)s(n ) ] , j^1_(0,n , n)&=&[|u(0)_d_(n)u(n)+|d(0)_d_(n)d(n)+|s(0)_d_(n)s(n ) ] , j^8_5(0,n , n)&=&[|u(0)_5_d_(n)u(n)+|d(0)_5_d_(n)d(n)-2|s(0)_5_d_(n)s(n ) ] , j^1_5(0,n , n)&=&[|u(0)_5_d_(n)u(n)+|d(0)_5_d_(n)d(n)+|s(0)_5_d_(n)s(n ) ] . due to the factor @xmath152 for @xmath153 , @xmath154
become dominate .
the normalizations of @xmath155 and @xmath156 are determined from the leptonic weak decay and the axial anomaly for @xmath157 meson , respectively .
this is similar to the pion case @xcite .
we employ the @xmath19 octet and singlet states to describe the @xmath3 system .
the @xmath17 and @xmath18 meson states can be described by means of the octet and singlet states @xmath158 and @xmath159 through the one mixing angle scheme [ stmix ] |&=&|_8-|_1|^&=&|_8+|_1where the mixing angle @xmath30 controls the relative strength . with the mixing ,
the @xmath1 and @xmath2 form factors take expressions [ ffmix ] f_(q^2)&=&f__8(q^2)-f__1(q^2 ) , f_^(q^2)&=&f__8(q^2)+f__1(q^2 ) . to proceed
, we also assume that the octet and singlet das take asymptotical form .
therefore , we have @xmath160 and @xmath161 . the form factors @xmath162 ( i=8,1 ) are simplified by substituting @xmath163 and @xmath164 [ ffi ] f__i(q^2)&=&4c_i dx [ -4]&=&6c_i[1- ] .
using eq .
( [ ffi ] ) into eq .
( [ ffmix ] ) , we may derive the coefficients @xmath165 in eq .
( [ nloff ] ) . to compare the form factors with the data , we extrapolate the form factors to all orders
f__i(q^2)= .
this formula gives a theoretical support to the approach using the interpolating formula for the @xmath1 and @xmath2 form factors @xcite .
the decay constants @xmath166 and the mixing angle @xmath30 will be determined by a least @xmath5 fit to the transition form factor data above @xmath142 gev@xmath167 and the two photon decay widths @xcite = ( 0.460.04 ) , = ( 4.280.19 ) .
the decay rates have theoretical expressions [ dw ] & = & m_^3 ^ 2 , & = & m_^^3 ^ 2 , where decay constants are defined as f_^8&=&f__1 , f_^1=-f__1 , f_^^8&=&f__1 , f_^^1=f__1 .
the @xmath5 fit results are shown as fit i in figs . 5 and 6 and in table .
i. it is seen that the fit i is in good agreement with the data for the form factors . to test the fit parameters
, we employ the ratio of the decay rates for @xmath4 into @xmath2 and @xmath1 r_j/==5.00.6 .
it is usually assumed that the radiative @xmath168 decays are dominated by nonperturbative gluon matrix elements @xmath169 and @xmath170 such that the ratio takes expression @xcite r_j/=()^2()^3 with @xmath171 being the three momentum of the @xmath172-meson .
our fit result is close to that one obtained from chiral perturbation theory ( @xmath173pt ) , except the octet decay constant @xmath174 , [ chipt1 ] : -20^~-10^ ; f__8=1.28f _ ; f__11.1 f _
. the octet decay constant @xmath27 is calculable by @xmath173pt up to one loop approximation f__8f_=1.28 f _ .
it is noted that the predict results for @xmath175 , @xmath176 and @xmath22 shown in table i are close to the experimental values within @xmath177 accuracy .
recently , it has been proposed @xcite that @xmath178 and @xmath179 can mix through a two mixing angle scheme as [ 2stmix ] |&=&_8 mixing angles . using this mixing scheme ,
the @xmath1 and @xmath2 form factors can be expressed , correspondingly , as [ 2ffmix ] f_(q^2)&=&_8 f__8(q^2)-_1 f__1(q^2 ) ,
f_^(q^2)&=&_8 f__8(q^2)+_1 f__1(q^2 ) . the form factors @xmath162 are the same as those in the one mixing angle scheme .
we first change the values of @xmath180 and @xmath181 to fit the data .
the @xmath5 fit result is shown as fit ii in figs . 5 and 6 and table .
i. from fit i and fit ii in table i , it is found that the two mixing angle scheme is better in accuracy than the one mixing angle scheme by 100@xmath182 .
this is close to the investigations @xcite .
a large value of @xmath183 @xcite , which is responsible for the intrinsic charm content of the @xmath18 meson , has been proposed to resolve the large branching ratios br@xmath184 and br@xmath185 .
we may explore this within our approach by adding intrinsic charm content into our formalism .
the effects of the intrinsic charm content are similar to that of the singlet component .
that means one can replace @xmath28 with @xmath186 , the decay constant for intrinsic charm for the corresponding singlet terms .
that is the part of the form factors from the intrinsic charm taking expression f__c(q^2)=4e_c^2where @xmath187 and the related das are @xmath188 and @xmath189 .
the effects from the large value of the charm quark mass has been absorbed into the twist-4 da @xmath190@xcite . after including the contribution of the intrinsic charm ,
the @xmath1 and @xmath2 form factors then become [ 2ffcmix ] f_(q^2)&=&_8 f__8(q^2)-_1 ( f__1(q^2)+f__c(q^2 ) ) , f_^(q^2)&=&_8 f__8(q^2)+_1 ( f__1(q^2)+f__c(q^2 ) ) . as the case of the octet and singlet form factors , the extrapolation of @xmath191 to all orders
is implied . by observing eq .
( [ 2ffcmix ] ) , form factor @xmath192 has a larger dependence of @xmath193 than @xmath194 .
we make a least @xmath5 fit to the form factor data to determine possible values of @xmath193 by keeping other parameters fixed . from table ii
, one may see that including the intrinsic charm content can indeed improve the accuracy .
this shows that our formalism is consistent in perturbation theory that the higher fock state can be reasonably added in . as shown below that the allowed value for @xmath193 is less than @xmath195 . in literature @xcite
, @xmath193 is proposed in the range
@xmath196 mev @xmath197 15 mev . to test this , we plot in fig .
7 the @xmath5 distribution for each set of mixing parameters list in table ii over a wide range of @xmath196 mev @xmath198 mev .
it is seen that the range of @xmath193 : @xmath199 mev @xmath197 -4 mev are allowed by the data . because the value of @xmath200 is close to unity , @xmath201 is almost equal to @xmath193 . from the above analysis , one may see that combining the high energy data and the low energy experiment can result in constraints on the mixing parameters in a very efficient way
. we can give a general analysis for the two mixing angle scheme .
we shall investigate the @xmath5 distributions of the mixing parameters .
the procedure of analysis is as followed .
we firstly separate the data into two groups . the data for the form factors , the two photon decay rates and the ratio for @xmath4 radiative decays are denoted as set i while the latter two data are chosen as set ii .
we then determine the least @xmath5 values for sets i and ii , respectively .
the reason for separating the data into set i and set ii is that neither set i nor set ii can completely constraint the parameters .
the parameters locate in @xmath177 accuracy of the set i still have large uncertainties and require further restrictions , which can be obtained by the data set ii . to be more explicit ,
we plot the allowable regions for the mixing parameters within @xmath177 error with respect to those values associated with the minimal @xmath5 points . as shown in figs .
8 and 9 , both allowable regions for data set i and ii are large while their intersections are rather restricted .
the reason for this fact is easily understood . within the data set i
, the experimental errors are shared for high and low energy data . because the @xmath5 distribution can only measure the correlations between the mixing parameters , the hope for constraining each parameters in a independent way
can not be obtained and only partial restrictions over the correlations of the parameters can be derived .
this can be seen from figs .
8 , in which the @xmath202 parameter is not contrained in a reasonable way . to compensate this flaw
, we note that the @xmath5 distribution of the data set ii can intersect with that @xmath5 distribution of the data set i. the intersections between two @xmath5 distributions can give better constraints on parameters than the data set i or ii .
one should note that although the data set ii is a subset of the data set i , the @xmath5 distribution of the data set ii is not necessary to be a subset of the @xmath5 distribution of the data set i , since the effects of correlations between the parameters are different for two data sets . from the overlapped regions of the data set i and ii in @xmath177 error , we may extract from figs .
8 and 9 the allowable regions for the mixing parameters . in fig .
8 , we plot the possible allowable ranges for @xmath27 and @xmath28 . it can be observed from fig .
8 that the overlapped region for @xmath27 and @xmath28 is quite stringent : @xmath203 and @xmath204 .
figure 9 shows the allowable region for @xmath205 and @xmath32 from both data set i and set ii .
the overlapped region indicates that : @xmath206 and @xmath207 . combining figs .
8 and 9 , we may derive the @xmath208 limits of the scaled @xmath1 and @xmath2 form factors q^2f_(q^2)|_q^2= [ f__8_8 - 2f__1_1]=(18946 ) , q^2f_^(q^2)|_q^2= [ f__8_8 + 2f__1_1]=(29535 ) .
the error in the scaled @xmath1 form factor being larger than that of the scaled @xmath2 form factor is due to the fact that the errors in the data for @xmath1 and @xmath2 form factors are shared in our analysis .
this is consistent with the concept of @xmath3 mixing .
we have shown that the collinear expansion for @xmath209 can be systematically performed in compatibility with pqcd factorization .
the @xmath34 power corrections for @xmath194 and @xmath192 have been evaluated .
the magnitudes of nlo power corrections are determined .
we have made a general analysis for the allowed values for the mixing parameters by combining the high and low energy data .
except of @xmath27 , the other three parameters @xmath210 can be constrained in a reasonable region .
the large error for the fit @xmath27 is mainly from the experimental error and can be improved by future experiment with higher accuracy . at present
, we would invoke the chiral perturbation theory calculation for @xmath27 .
we have also shown that the intrinsic charm content of @xmath18 meson has little contributions .
any sizable contribution from the intrinsic charm would lead to large @xmath5 value as shown in fig .
7 . of course
, there require further investigations for this point , if more accurate form factors data are available .
so far , we have not considered the finite meson mass effects for the @xmath1 and @xmath2 form factors . at first glance
, the meson mass effects can not be safely ignored . because the values of @xmath17 and @xmath18 meson masses are large as @xmath211 mev and @xmath212 mev , power corrections from the mass effects of order @xmath213 and @xmath214 are also important .
the type of corrections @xmath213 are kinematics .
this is similar to the nachtmann s correction for deep inelastic scattering @xcite and can be negligible .
the second type of corrections @xmath214 are dynamics and can be argued that they are of twist-6 , at least .
this is because the associated spin projector is @xmath215 ( cf . the leading spin projector @xmath216 ) , which will introduce two additional @xmath217 propagators into the related hard function . as a result ,
the dynamical type meson mass corrections are of order @xmath218 .
of course , a complete analysis for these meson mass effects is important .
m. acciarri _ et al . _
[ l3 collaboration ] , phys .
b * 418 * , 399 ( 1998 ) .
h. j. behrend _ et al .
_ [ cello collaboration ] , z. phys .
c * 49 * , 401 ( 1991 ) .
h. aihara _ et al .
_ [ tpc / two gamma collaboration ] , phys .
. lett . * 64 * , 172 ( 1990 ) . c. berger _
[ pluto collaboration ] , phys .
b * 142 * , 125 ( 1984 ) . | the next - to - leading - order ( nlo ) @xmath0 power correction for @xmath1 and @xmath2 form factors are evaluated and employed to explore the @xmath3 mixing .
the parameters of the two mixing angle scheme are extracted from the data for form factors , two photon decay widths and radiative @xmath4 decays .
the @xmath5 analysis gives the result : @xmath6 , where @xmath7 and @xmath8 are the decay constants and the mixing angles for the singlet ( octet ) state .
in addition , we arrive at a stringent range for @xmath9 mev@xmath10 mev .
# 1/ # 1#2 |
due to the ever growing quantity of high resolution data , in the last few years great interest has been focused on the central region of galaxies where various phenomena co - exist .
thanks to the high resolution images provided by the hubble space telescope , it is clear , nowadays , that the nuclei of the majority of both elliptical and early type spiral galaxies ( @xmath2m@xmath1 ) harbour massive or supermassive black holes ( smbhs ) , whose masses , @xmath3 , range in the @xmath4m@xmath1 interval and may be up to @xmath5 m@xmath1 as in the case of the smbh in ngc1277 @xcite . in some cases ,
the central smbh is surrounded by a massive , very compact , star cluster commonly referred as nuclear star cluster ( nsc ) .
nscs are observed in galaxies of every type of the hubble sequence @xcite and their modes of formation and evolution are still under debate . in the case of elliptical galaxy hosts ,
the nuclear clusters are also referred to as resolved stellar nuclei. for the sake of this paper we will refer to nscs or resolved stellar nuclei indifferently .
nscs are sited at the photometric and kinematic centre of the host galaxy , i.e. at the bottom of the potential well @xcite .
this is likely connected to a peculiar formation history . as a matter of fact
, all nscs contain an old stellar population ( @xmath6 gyr ) and most of them show , also , the presence of a young population , with ages below @xmath7 myr @xcite .
nscs are bright ( about 4 mag brighter than ordinary globular clusters ) , massive objects ( @xmath8 m@xmath1 ) , very dense and with a half - light radius of @xmath9 pc .
their small sizes and large masses make them the densest stellar systems in the universe @xcite .
the relation between nscs and smbhs is poorly known ; they seem to be two faces of the same coin , constituting central massive objects ( cmos ) whose actual presence depends on the host mass : galaxies with mass above @xmath10 m@xmath1 usually host an smbh while lighter galaxies have , instead , a well resolved central star cluster ( an nsc ) .
moreover , a transition region exists for galaxies with mass between @xmath11 and @xmath10 m@xmath1 in which both the objects co - exist @xcite . with regard to the lack of evidence of nscs in high mass ( @xmath12 m@xmath1 ) galaxies
, one possible explanation is the formation of giant ellipticals through merging of smaller galaxies @xcite .
quantitatively speaking , @xcite simulations showed that if the two colliding galaxies host mbhs , a black hole binary ( bhb ) could form which heats up the resulting stellar nucleus causing its progressive evaporation .
this process can destroy the super cluster , shaping significantly the density profile of the merger product and leaving behind a bhb that shrinks due to gravitational wave emission leading eventually to a smbh .
another possibility is that in the early phase of the galaxy life , an initial nsc could be the seed for the bh birth as suggested first by @xcite and later by @xcite and @xcite .
recently , a number of researchers studied the existence of scaling relations between nscs and their galactic hosts ; similar studies have already been done , seeking for scaling relations between smbhs and their hosts .
however , it is still unclear the robustness of the nsc - galaxy relations and whether these relations are linked to those between smbhs and the hosts . for instance
, @xcite showed that the nsc mass ( @xmath13 ) vs galaxy velocity dispersion ( @xmath14 ) relation is roughly the same of that observed for smbhs .
on the other hand , more recent studies @xcite claim that the @xmath15 relation is shallower than for smbhs , @xmath16 . moreover , it has been shown that while smbh masses correlate with the galaxy mass , the nsc masses correlate better with the bulge mass @xcite .
at present , two are the most credited frameworks for the nsc formation .
one scenario refers to the so called ( dissipational ) in - situ model @xcite . according to this model , an injection of gas in the central region of a galaxy hosting a seed black hole could lead to the formation of a nsc if the typical crossing time of the parental galaxy is shorter than the so - called salpeter time , which is the time - scale over which the central bh can grow by accretion @xcite .
another ( dissipationless ) scenario invokes the action of the dynamical friction process which makes massive globular clusters ( gcs ) sink toward the centre of the host galaxy @xcite .
their subsequent merging leads to a super star cluster with characteristics indistinguishable from those of an nsc @xcite .
this scenario is often referred to as infall - merger scenario or migratory - merger model .
both the theories above encounter some troubles in explaining completely the nsc formation . according to some qualitative considerations
, the in situ model would predict too massive nscs , while a possible problem for the gcs infall model is that it would give lighter nscs than observed @xcite .
@xcite say that mergers of star clusters are able to produce a wide variety of observed properties , including densities , structural scaling relations , shapes ( including the presence of young discs ) and even rapid rotation , nonetheless claim that some kinematical properties of observed nscs are hardly compatible with merger models .
they suggest the need of a 50@xmath17 of gas in the overall scheme .
@xcite referring to their fornax acs survey and to some speculative considerations conclude that , for galaxies and nuclei in their sample , the infall formation mechanism is the more likely for low to intermediate mass galaxies while for more massive ones accretion triggered by mergers , accretions , and tidal torques is likely to dominate .
the two mechanisms smoothly vanish their efficiency on the intermediate mass galactic range , and they indeed provide some evidence of `` hybrid nuclei '' which could be the result of parallelly acting formation mechanisms . while the in - situ model has remained , so far , almost speculative and difficult to constrain to available observations , several authors provided detailed numerical tests for the gc merger scenario starting from the original idea in @xcite .
the first simulations were done by @xcite and @xcite in galaxy models without massive black holes and stellar discs ; @xcite studied the role of stellar discs and @xcite the role of a central galactic mbh .
in particular , @xcite made a full @xmath18-body simulation of the decay and merging of 12 gcs in a milky way model accounting for the presence of the sgr a * @xmath19 m@xmath1 central black hole , obtaining an nsc that has global properties fully consistent with those observed in the nucleus of our galaxy .
one recent work by @xcite presents the same merger simulations performed in @xcite with the inclusion of different stellar populations in the various infalling globular clusters , and shows that infalling clusters can produce thick flattened structures with varied orientations , possibly related to disky structures that are observed in galactic nuclei and clusters ( see @xcite for a discussion of the evolution of such discs ) . on another side , @xcite ,
by mean of a semyanalytical model , made some comparisons among the expexcted result of a merger scenario for the nsc formation and some scaling laws .
the aim of this paper is to check in a more complete and extensive way the reliability of the infall - merger model for the nsc formation .
to reach this aim we build theoretical scaling laws connecting nsc properties with those of the galactic hosts in a synthetic modelization of the global evolution of a globular cluster system ( gcs ) in a galaxy , considering the dynamical friction and tidal disruption as evolutionary engines .
these scaling laws are to be compared with those observationally obtained .
the paper is organized as follows : in section [ dynfri ] the role of dynamical friction is discussed as well as the way we modeled galaxies and their gcs ; in the same section the sample of data used for the comparison with observation is presented ; section [ appro ] presents two different theoretical modelizations of the nsc growth in galaxies ; in section [ laws ] we provide a set of theoretical scaling laws which connect nscs with their hosts and compare them with the observed laws ; in section [ tid ] , instead , we take into account the effect of tidal disruption of gcs on the resulting nsc mass .
finally , section [ end ] is devoted to a summary of the main results , providing some general remarks such to draw conclusions .
the formation of a compact nucleus in the centre of a galaxy through the orbital decay of globular clusters has been discussed , first , by @xcite . working on a model of the m31 galaxy
, they demonstrated that the efficiency of the dynamical friction mechanism could provide an amount of matter sufficient to form a compact nucleus of @xmath20m@xmath1 in the centre of this galaxy .
@xcite turned out the importance of considering the tidal disruption of the clusters as a competitive process that tunes the effect of dynamical friction . before
approaching in a deeper way the theme , we now give a relevant analytical support to the idea that a nsc can grow in the centre of a galaxy by mean of gc decay in the innermost region via dynamical friction braking . a direct and easy way to obtain a scaling between the mass accumulated to the galactic centre and the background velocity dispersion
is based on the assumption that the galaxy has the mass density of a singular isothermal sphere ( r)= , where @xmath21 is the gravitational constant , @xmath22 is the galactocentric distance and @xmath23 is the circular velocity , constant with radius and related to the velocity dispersion , @xmath14 , by @xmath24 .
approximating the motion of the test mass , @xmath25 , as a decreasing energy sequence of circular motions , the evolution equation of the modulus of the orbital angular momentum per unit mass , @xmath26 , is @xmath27 where @xmath28 is the absolute value of the dynamical friction force exerted by the galaxy on the test object , which , using the chandrasekhar s formula in its local approximation , is given by @xmath29 with @xmath30 where @xmath31 ( @xmath32 for a singular isothermal sphere ) , @xmath33 is the usual error function and @xmath34 is the coulomb logarithm .
the time evolution of the radius @xmath35 of the nearly circular orbit of the test mass under the previous assumptions is thus governed by the differential equation @xmath36 where @xmath37 , which , with the initial condition @xmath38 , is easily integrated to give @xmath39 and leads to @xmath40 as fully decay time ( @xmath41 ) of the object of mass @xmath25 initially moving on the circular orbit of radius @xmath42 .. given , for the gc population , a density distribution in the form of a power - law @xmath43 where @xmath44 is an a priori free parameter and @xmath45 is a normalization constant constrained to give the total mass of the gcs , @xmath46 : @xmath47 assuming the total gcs mass to be a fraction @xmath48 of the galactic mass @xmath49 , @xmath50 @xmath45 turns out to be a function of the galactic velocity dispersion and radius @xmath51 if the gcs dynamically decayed to the galactic centre go to grow a nucleus therein , the value of the nucleus mass at any age , @xmath52 , of the galaxy can be obtained by mean of @xmath53 , which is the maximum radius of the gc circular orbit decayed to the centre within time @xmath52 .
consequently , @xmath54 , is simply @xmath55 for @xmath56 , while @xmath54 saturates to @xmath46 at @xmath57 . note
that equation [ mnt ] is independent of the galactic radius @xmath58 and reduces to the @xmath59 relation obtained by @xcite in the case of @xmath60 , i.e. for gc distributed the same way as the galactic isothermal background .
this is the only case where the dependence on @xmath58 cancels out . if , instead , @xmath61 the dependence of @xmath54 on @xmath14 becomes , in the assumption of a virial relation between galactic @xmath58 and @xmath49 ( @xmath62 ) @xmath63 which corresponds , assuming a constant @xmath64 , to a slope in the range from @xmath65 of the steeper ( @xmath66 ) gcs radial distribution to @xmath67 of the flat ( @xmath68 ) distribution .
the relevant result here is that the slope of the @xmath69 relation in the regime of dynamical friction dominated infall process is expected to have an upper bound in any case smaller than that of the @xmath70 relation .
the aim of this work is to show that the dry merger scenario can reproduce the correlations of observed nscs with their hosts in a wide range of galaxy masses . to reach this aim ,
we need three important ingredients : i ) a robust data base to compare our results with real observations of nscs and their hosts , ii ) a reliable treatment of the dynamical friction and tidal disruption processes and iii ) a detailed model for the host galaxies to reproduce the environment in which gcs evolve .
the data base for the purposes of this work has been extracted from three different papers .
the first ( @xcite , hereafter eg12 ) , combines data coming from different works covering galaxies of the hubble types s0-sm ; on another side , @xcite ( hereafter lkb12 ) provide data for @xmath71 early type galaxies in the advanced camera virgo cluster survey @xcite ; finally , we considered data given in @xcite ( hereafter sg13 ) which is a collection of data from earlier works . at the end
, we gathered a total sample of @xmath72 galaxies covering a wide range of hubble types which contains several structural parameters of each galaxy such as mass , effective radius , velocity dispersion , and of the nsc masses . to evaluate reliably the dynamical friction braking of gcs in their host galaxies
, we have to assume galactic density profiles . as first approximation , we modeled galaxies as spherically symmetric distributions in the form of dehnen s spheres
whose density is : @xmath73 where @xmath74 and @xmath75 is linked by @xmath76 to the total mass of the galaxy , @xmath49 and to its length scale , @xmath77 . generally , galaxies fainter than @xmath78 show steep surface luminosity profiles @xmath79 with slope @xmath80 ( power - law galaxies ) , while brighter galaxies show less pronounced cusps ( core galaxies with @xmath81 ) @xcite .
the two slopes , @xmath82 and @xmath83 , are linked by the relation @xmath84 in the case @xmath85 @xcite .
consequently , for each galaxy mass @xmath49 , the @xmath83 exponent is randomly chosen in the range @xmath86 for @xmath87 m@xmath1 , and in the range @xmath88 for @xmath89 m@xmath1 .
on the other hand , since the slope of the @xmath90 depends critically on the surface brightness profile and may vary from different kind of galaxies , we allowed also the @xmath83 parameter to vary in a more general way , i.e. extracting it randomly between @xmath91 for each galaxy model , finding not significant changes in our results .
for this reason , we decide to choose @xmath83 in the two ranges explained above , in agreement with the fact that brighter galaxies seems to have flatter surface brightness profiles .
it is relevant noting that the validity of true cuspidal density profile models to describe the matter distribution of galaxies has been questioned by many authors that claim that the core - srsic profiles are better suited to describe the innermost ( @xmath92 to @xmath93 arcsecs ) regions of early type galaxies @xcite .
however , @xcite showed that power - law and core galaxies show more or less the same steepness of the density profiles in their outer regions , with small changes of the slope of the surface brightness profiles .
since we are interested in the study of the dynamics of stellar clusters on a relatively large scale of the galaxy and not in its innermost region , where the investigation required a specific , more accurate modelling which is out of our scopes , we consider the simple @xmath83 density profiles as appropriate for our purposes .
this choice allows us to use the results on dynamical friction recently obtained for cuspy dehnen s profile @xcite .
actually , dehnen s profiles have a central cusp , and it has been demonstrated that the in density cusps the classical chandrasekhar dynamical friction formula fails ( see for instance @xcite ) . in particular , both @xcite and @xcite found that dynamical friction braking is reduced in the vicinity of a mbh . to overcome this problem , @xcite provided a formulation for the dynamical friction process which is valid in cuspy galaxies , giving a useful fitting expression for the dynamical friction timescale ( @xcite , equation 21 ) : @xmath94 where @xmath95 is the orbital eccentricity of the text object of mass @xmath25 and @xmath96 is a normalization factor whose value is given by : @xmath97 where @xmath77 ( in @xmath98 ) is the scale radius and @xmath99 the total mass of the @xmath83 model galaxy in unit of @xmath100 . equation [ tdf ]
gives the dynamical friction decay time for a gc of mass @xmath25 initially moving on an orbit of eccentricity @xmath101 in a spherical galaxy of mass @xmath49 and length scale @xmath77 .
since @xmath102 depends , other than on @xmath95 , on @xmath49 , @xmath77 and @xmath83 , to a good estimate of the dynamical friction time reliable values of these three parameters are needed . in other words , it is important providing reliable models of the parent galaxy to ensure that the environment where dynamical friction acts is well reproduced .
the simplest way to produce such model environments is via linking the parameters needed to establish the theoretical model to the observable quantities .
as example , the lkb12 data sample contains the mass , @xmath49 , and effective radius , @xmath103 , which is the radius containing half of the total light .
this parameter is important because it can be connected with the scale radius of the theoretical @xmath83 model , @xmath77 .
however , the data sample contains a limited number of galaxies in the range @xmath104m@xmath1 and does not provide data for heavier galaxies , hence only a small fraction of the total range of galaxy masses could be investigated and , moreover , not always both the @xmath64 and @xmath105 values are available .
to overcome these limitations , we need to extend with some proper extrapolations to a range of galaxies covering a wider , @xmath106m@xmath1 , mass interval . to this aim , we used data in lkb12 to correlate the galaxy mass with the effective radius and found a good fitting formula linking these two quantities as : @xmath107 where @xmath49 is in units of @xmath108m@xmath1 .
then , given the galaxy mass we can evaluate its own effective radius , which is an astrophysical observable , and , finally , we get the scale length @xmath109 by mean of these two relations @xcite : @xmath110 @xmath111}. \label{rh}\ ] ] the first one being valid in the range of @xmath83 considered in this paper .
finally , the scale length @xmath109 is related to @xmath83 and @xmath64 by this relation @xmath112 in figure [ mre ] the above @xmath113 curve is drawn together with observed data taken from lbk12 . estimated with equation [ effec].,width=321 ] on the other hand , to give an estimate of the total radius of the galaxy we developed the following relation : @xmath114 which allows us to obtain total radii for our galaxy models going from few @xmath98 for dwarf galaxies to several @xmath98 for giant ellipticals , and give us an estimate of the maximum distance from the galactic centre allowed as initial position for the clusters .
as example , for a galaxy mass @xmath115m@xmath1 we obtain a radius @xmath116 , that is a reasonable value for such galaxies ( the supergiant elliptical galaxy m87 has a radius @xmath117 , comparable to this value ) .
the other important parameter that we used to compare our galaxy models with real galaxies is the velocity dispersion , @xmath118 . following lbk12 , to evaluate @xmath118 we used the formula given in @xcite : @xmath119 where @xmath120 is the baryonic mass fraction assumed @xmath121 as in lbk12 .
again , for any given galaxy mass , we selected randomly the parameter @xmath83 in the ranges explained above .
the comparison between @xmath118 vs @xmath49 in lbk12 with our estimate is shown in figure [ msig ] .
figures [ mre ] and [ msig ] convince us that we modeled the hosts sufficiently well to obtain reliable estimation of df times and , as a consequence , reliable values of the nsc masses .
several authors had pointed out that the correlation between the bulge and the nsc mass is more dispersed than the @xmath122 relation @xcite .
this is also related to the fact that many galaxies are actually bulgeless systems .
since the sample of galaxies we used as reference contains also early and late - type spirals , we decided to evaluate the bulge mass for our systems by using the correlation bulge - host given in eg12 : @xmath123 this allows us to sample bulges in good agreement with the observed values for spiral galaxies , as it is shown in figure [ bulga ] , and it is still coherent to describe the mass of the stellar spheroid in ellipticals and dwarf spheroidal galaxies , where the spheroid mass is indistinguishable from the whole galaxy mass .
another fundamental ingredient in this framework is the globular cluster system ( gcs ) total mass .
a lower limit to the gcs mass can be obtained considering that the ratio between the gcs mass and the galaxy mass goes from @xmath124 for small galaxies ( @xmath125 m@xmath1 ) , up to @xmath126 for the largest ( @xmath49 up to @xmath127 m@xmath1 ) .
this suggests a weak correlation between the gcs initial mass and the galaxy mass ; a good fit formula to this correlation is : @xmath128 @xcite pointed out that the number of gcs in a given galaxy correlates with the galaxy mass through the relation : @xmath129 this relation can be rewritten properly to describe the relation between the galaxy mass and the gcs mass : @xmath130 where @xmath131 is the mean value of the gc mass in the galaxy . allowing a mean value of @xmath132m@xmath1 for the gc masses , we recover a similar expression to equation [ mgcs1 ] .
due to that smaller galaxies host light globulars ( @xmath133 m@xmath1 ) , while in heavier galaxies the globulars masses range in the @xmath134 m@xmath1 interval @xcite , we set the minimum and maximum value of the gc mass as a function of the galaxy mass : @xmath135 as we will show in section [ stat ] , this choice gives mean gc masses in good agreement with observations .
a sufficiently accurate estimate of the nsc mass accumulated to the centre of the galaxy in the merger scenario may be given by means of the following considerations .
letting @xmath136 be the ( infinitesimal ) number of gcs with mass in the @xmath137 $ ] range and in the volume @xmath138 centred at @xmath139 , the total mass of the gcs is : @xmath140 where @xmath141 and @xmath142 indicate , respectively , the lower and upper value for the gc mass and @xmath143 is the volume occupied by all the gcs . keeping a sufficient level of generality
, we can assume @xmath144 in the form of the product of a function of @xmath25 and a function of @xmath22 : @xmath145 where @xmath146 is a normalization constant given by @xmath147 with @xmath148 the total number of clusters in the galaxy .
a suitable expression for the mass function , @xmath149 , is a ( truncated ) power - law ( see for instance @xcite ) : @xmath150 on the other hand , the distribution of radial positions is , in principle , arbitrary .
the simple inversion of equation [ tdf ] yields the maximum radius , @xmath151 , which contains all the clusters with mass @xmath152 and with initial eccentricity @xmath153 that have been confined around the galactic centre in a time @xmath154 : @xmath155 with @xmath156 .
consequently , an estimate of the nsc mass as a result of the accumulation of gcs to the galactic centre , caused by dynamical friction , is : @xmath157 with @xmath158 given by : @xmath159 let us now consider two different radial distributions for the gc population : a generic power - law distribution , @xmath160 , and a @xmath83 model density law ( equation [ dhn ] )
. in the first case , equation [ mgcsmax ] reduces to : @xmath161 with @xmath18 the total number of clusters in the galaxy . by substitution of this relation into equation [ mnsct ]
we obtain : @xmath162 with @xmath163 . where @xmath164 and @xmath165 is a function of @xmath166 and @xmath52 whose explicit expression is : @xmath167 after integration , equation [ mnsc ] yields @xmath168 looking at equation [ tan ] , it is evident now the dependence of the nsc mass from the @xmath169 parameter , i.e. from the steepness of the density profile .
the more general case in which we consider a @xmath83 density law , instead , will be discussed in the appendix . allowing @xmath170 to vary in equation [ tan ]
we estimate the mass of the nscs for different values of the slope of the gcs mass function @xmath171 at varying the galaxy mass in the range @xmath172$]m@xmath1 .
equation [ tan ] allows us to see how the nsc mass increase as a function of time .
figure [ growth ] shows the nsc growth as a function of time in the case @xmath173 for two extreme values of the galaxy mass ( @xmath174 and @xmath127 m@xmath1 ) and three values of @xmath171 considering the mass function in equation [ imf ] , i.e. @xmath175 with @xmath141 and @xmath142 as defined by equation [ m2 ] .
m@xmath1 ( upper panel ) , and @xmath176m@xmath1 ( bottom panel).,title="fig:",width=321 ] + m@xmath1 ( upper panel ) , and @xmath176m@xmath1 ( bottom panel).,title="fig:",width=321 ] the figure shows that the nsc mass increases rapidly in an early phase ( @xmath177 gyr ) to slow down later its growth .
the slower increase in the case of larger values of @xmath171 depends on the smaller fraction of heavy , and fast decaying , gcs for steeper mass functions .
using the case @xmath178 , @xmath173 as reference , we show in figure [ growthcmp1 ] and [ growthcmp1 ] the ratio between the nsc mass evaluated letting @xmath179 , @xmath180 and that obtained with @xmath173 at fixed @xmath171 and @xmath64 .
m@xmath1 evaluate with @xmath181 ( straight line ) and @xmath179 ( dotted line ) and @xmath182 obtained lettin @xmath173 . from top to bottom , we set @xmath183.,title="fig:",width=302 ] m@xmath1 evaluate with @xmath181 ( straight line ) and @xmath179 ( dotted line ) and @xmath182 obtained lettin @xmath173 . from top to bottom , we set @xmath183.,title="fig:",width=302 ] m@xmath1 evaluate with @xmath181 ( straight line ) and @xmath179 ( dotted line ) and @xmath182 obtained lettin @xmath173 . from top to bottom , we set @xmath183.,title="fig:",width=302 ] , but for a galaxy with @xmath184m@xmath1.,title="fig:",width=302 ] , but for a galaxy with @xmath184m@xmath1.,title="fig:",width=302 ] , but for a galaxy with @xmath184m@xmath1.,title="fig:",width=302 ] considering the case @xmath185m@xmath1 , it is evident that in an early phase ( @xmath186 ) the smaller @xmath170 the faster the nsc mass growth ; however as the time increase is evident that the final mass of the cluster is slightly small with respect to the reference case @xmath173 if @xmath187 , while considering @xmath188 the smaller the @xmath170 the greater the final mass of the nsc . considering instead more massive galaxies ( @xmath184m@xmath1 ) we found that the smaller the @xmath170 the greater the final nsc mass .
figure [ mnsccmp ] shows the mass of nscs as a function of the host mass for @xmath189 and @xmath190 . at any fixed value of @xmath171 ,
there is not a significant difference between nsc masses estimate with different values of @xmath170 . and @xmath170 .
from top to bottom , we set @xmath183 . in each panel , are shown results for @xmath191.,title="fig:",width=302 ] and @xmath170 . from top to bottom , we set @xmath183 . in each panel , are shown results for @xmath191.,title="fig:",width=302 ] and @xmath170 . from top to bottom
, we set @xmath183 . in each panel , are shown results for @xmath191.,title="fig:",width=302 ] in figure [ nscm ] we compare the theoretical nsc mass ( that evaluated at @xmath192 gyr ) with the observational values from eg12 and lbk12 .
the best agreement is achieved choosing @xmath193 , as it will be more deeply discussed in section [ laws ] .
moreover , we found that a good correlation is achieved in the case @xmath194 , as it is shown in figure [ nscm2 ] , but this extreme case in which both the density profile and the mass function are very steeps , is really unlikely to reproduce real galaxies .
for this reason , we limited the analysis to the case @xmath193 . in the following ,
we refer to the model @xmath178 as the combination ( @xmath193 ) . in this context , it is relevant noting that the mass distribution of young luminous clusters ( often referred to as ylcs ) in many galaxies is a power - law with a spectral index ranging between @xmath195 and @xmath178 @xcite .
assuming such power - law as initial mass function for a gcs , @xcite showed that the evolution of such a system in a model of our galaxy , leads to a final mass function for the gcs in agreement with the actual mass function of the milky way gcs , giving us some confirmation about the choices made to model gcss .
with @xmath178 , @xmath173 ; compared with nsc masses from data in lkb12 ( stars ) , eg12 ( squares ) and sg13 ( filled circles ) . from top to bottom , the various panels give the correlation with hosts masses , velocity dispersion and bulge mass.,title="fig:",width=321 ] + with @xmath178 , @xmath173 ; compared with nsc masses from data in lkb12 ( stars ) , eg12 ( squares ) and sg13 ( filled circles ) . from top to bottom , the various panels give the correlation with hosts masses , velocity dispersion and bulge mass.,title="fig:",width=321 ] + with @xmath178 , @xmath173 ; compared with nsc masses from data in lkb12 ( stars ) , eg12 ( squares ) and sg13 ( filled circles ) . from top to bottom , the various panels give the correlation with hosts masses , velocity dispersion and bulge mass.,title="fig:",width=321 ] + , but in the case @xmath188 , @xmath181.,title="fig:",width=321 ] + , but in the case @xmath188 , @xmath181.,title="fig:",width=321 ] + , but in the case @xmath188 , @xmath181.,title="fig:",width=321 ] + beside the analytical method illustrated above to estimate the nscs masses , we investigate the infall scenario also from a statistical point of view , in order to obtain information on the radial distribution of gcs in the host galaxy , the number of globulars centrally decayed and that of gc survivors .
the idea behind this statistical approach is sampling the initial gcs of a given galaxy , and evaluate how many gcs sink toward the galactic centre within a hubble time . a statistical estimate of the expected nsc mass is thus obtained by making @xmath196 realizations of the gcs of a galaxy , in order to give constraints to the error .
each galaxy was modeled as explained in section [ model ] , while , for each cluster , we sampled its initial radial position , @xmath42 , and orbital eccentricity , @xmath95 , from a random , flat distribution . from this gcs sampling
we infer two relevant parameters to compare with observations .
the first is the mean value of the gc mass for any given host mass , that can be compared with data given in lkb12 ; the second parameter is the number of survived clusters , which goes from few ( @xmath197 ) gcs for small galaxies ( @xmath198m@xmath1 ) , to few hundreds in intermediate mass galaxies , and up to @xmath199 in giant ellipticals .
the relatively strong dependence of the df braking time on the individual gc mass ( see equation [ tdf ] ) deserves a careful treatment in the gc sampling , as explained in detail here below .
as mentioned above , in this section we give estimate of the expected nsc mass for a given galaxy mass , sampling the whole gcs of the galaxy and looking at which clusters can sink toward the galactic centre within a hubble time . since initial position , eccentricity of the orbits and mass of each cluster are fundamental parameters in the evaluation of the decay time ( see equation [ tdf ] ) , we vary the sampling method for the gcs , changing spatial distribution and mass function of the clusters as explained in the following .
this model ( referred to as pls model ) is characterized by a flat spatial distribution of gcs within the radial range @xmath200 $ ] , with @xmath58 the maximum distance defined in equation [ rtot ] ; their eccentricities are sampled randomly between @xmath65 and @xmath201 .
gc masses are distributed according to a power - law distribution , @xmath202 .
pls is , actually , the statistical version of the analytical treatment ( see section [ ana ] ) . when @xmath203 ( flat mass distribution ) we refer to as the random sampling model , called rnd . in the gss model
, the spatial location of gcs is the same as above , while the mass sampling is made by means of a gaussian generator , given a mean value @xmath204 $ ] , and a fixed dispersion , @xmath205 . to exclude unrealistic , too light or too massive objects in the mass distribution
, we truncated the gaussian at a low mass @xmath206m@xmath1 and at a high mass @xmath207m@xmath1 . there is no compelling evidence that the gcs and stars of the parent galaxy followed , initially , different density profiles , so we found worth examining the case where the initial gcs density profile is the same @xmath209 profile of the parent galaxy .
the quality of our gc sampling can be tested by a comparison of the gcs mean masses obtained , @xmath210 , with data given in lkb12 .
looking at figures [ mgc ] , we see that gss and rho models give a decent agreement with observations in the whole range of masses covered by the data available ; on the other hand , the rnd model seems to overestimate the mean gc mass , while the pls model gives an underestimate for host masses above @xmath211 .
+ another relevant quantity obtainable with this approach is the number of survived " clusters , which can be compared with the actually observed clusters . in figure [ srv ] the fraction of decayed clusters as a function of the hosting galaxy mass
moreover , figure [ srv2 ] shows also the number of decayed clusters .
it should not surprise that for sufficiently massive galaxies ( @xmath89m@xmath1 ) the number of survived gcs can exceed @xmath199 .
many massive galaxies actually host such large populations of clusters .
as example , the giant elliptical galaxy m87 ( also known as virgo a , with a mass @xmath212 m@xmath1 @xcite hosts about @xmath213 clusters , in agreement with our prediction . on the other hand , the small number of gc expected in galaxies with @xmath214m@xmath1 , could provide a possible explanation for the lack of nucleated region in dwarf spheroidal galaxies in the virgo cluster @xcite .
in fact in such small galaxies , only few clusters are expected to form , this would imply that the formation of a nsc , at least in the framework of the dry - merger scenario , depends strongly on the initial conditions of each cluster of the galaxy .
fluctuations in this low number statistics , together with the low mass of gcs in these environments ( and so long dynamical friction times ) make the probability of finding nscs in the center of these small hosts very low .
figures [ tdfdist ] show the ratio between the df decay time and the hubble time for gcs belonging to a galaxy of mass @xmath215m@xmath1 whose gcs is sampled with the pls , rnd , gss and rho models .
all clusters with @xmath216 are decayed , therefore , the expected nsc mass is evaluated by summing the masses of all the decayed clusters , and in figure [ mnscmg ] the resulting nsc mass vs. the host mass is reported .
is shown for a sampled galaxy @xmath215m@xmath1 in pls , rnd , rho and gss models , respectively .
decayed clusters ( triangles ) lie all in a region whose radius is roughly @xmath217.,title="fig:",width=302 ] is shown for a sampled galaxy @xmath215m@xmath1 in pls , rnd , rho and gss models , respectively .
decayed clusters ( triangles ) lie all in a region whose radius is roughly @xmath217.,title="fig:",width=302 ] is shown for a sampled galaxy @xmath215m@xmath1 in pls , rnd , rho and gss models , respectively .
decayed clusters ( triangles ) lie all in a region whose radius is roughly @xmath217.,title="fig:",width=302 ] is shown for a sampled galaxy @xmath215m@xmath1 in pls , rnd , rho and gss models , respectively .
decayed clusters ( triangles ) lie all in a region whose radius is roughly @xmath217.,title="fig:",width=302 ] however , it is not easy to understand which model fits better the observations . a more quantitative analysis is needed to reveal the real agreement , as , for instance , that of drawing scaling laws which connect the nsc mass with some of the host properties in the following section , we deepen the study of the comparison of our theoretical and observed nsc , drawing the theoretical scaling laws mentioned above to compare them with those actually observed . and actually observed scaling laws .
as we said in the introduction , the existence of correlations and scaling relations between the central compact object and the galactic host parameters may be an important clue to the understanding of the actual mechanisms of cmo formation .
it is well known that smbh masses show a tight correlation with the host galaxy bulge velocity dispersion , @xmath118 , @xcite and with the galactic bulge mass , @xmath218 , ( see for example @xcite and @xcite ) .
the implication claimed is that similar processes drove both smbh and galaxy growth .
in particular , @xcite suggested that a feedback exists between the early stage of life of a galaxy and its central bh . in the last years
, many studies were devoted to derive , also , scaling relations among nscs and their host galaxies , finding that they follow relations in part similar as smbhs do @xcite .
however , it is still unclear what , if any , the two different types of cmos have in common , so to imply an intimate link between central galactic bhs and nscs growth and evolution . actually , differences in scaling relations of bhs respect to nscs are being presently debated . as an example
, @xcite claimed that nscs follow the same mass - sigma relation of massive central bhs , which is a power law with an exponent between @xmath219 and @xmath220 . on the other side , @xcite and @xcite
find a significantly shallower relation for the mass-@xmath14 relation of nscs , with the exponent in the range @xmath221 to @xmath92 . at this regard
, while the in situ `` model is compatible with the steeper relation found in @xcite ( see for example @xcite ) , the dry merger '' scenario , instead , fits well with the @xcite and @xcite relations , as we have seen in section [ teo ] of this work ( see also @xcite ) . by means of both the statistical and the analytical approaches presented above we can draw various correlations , including @xmath122 as well as @xmath222 and @xmath223 relations . at this scope , in figure [ mnscsig ] and figure [ mnscmb ] we report , respectively , the nsc masses as functions of the host galaxy velocity dispersion and bulge mass of our models compared with observed data .
note that due to the somewhat ill definition of the bulge , the relation @xmath13 vs @xmath224 is not very reliable .
it has been shown that the nsc mass correlates better with the total galaxy mass , while the correlation with the bulge is not statistically very significant @xcite .
we obtained a power law best - fitting for our sampled correlation between @xmath13 and @xmath49 in the form @xmath226 where the coefficients @xmath227 and @xmath228 ( see table [ tabm ] ) are computed by using the marquardt - levenberg nonlinear - regression algorithm . for the sake of comparison we report table [ tabmobs ] the slope @xmath228 of the best - fittings to the @xmath122 relations in the eg12 , lkb12 and sg13 data samples .
the comparison between values in table [ tabm ] and [ tabmobs ] indicates that both the analytical ( @xmath178 ) model and the statistical approaches ( pls , gss , rnd and rho models ) give a slope in good agreement with the observed relation , within the errors . .@xmath122 scaling relation parameters for various models . [
cols="^,^,^,^,^",options="header " , ] column 1 : sample name .
column 2 - 3 : slope @xmath228 and and relative error . as final remark of this section
, we note that the fig .
16 of @xcite paper shows a relevant feature that nsc formation model should interpret .
on one side it gives evidence that much of the mass of nsc is in old stars ( thing straightforwardly compatible with the merger model ) ; on the other side , it seems to indicate the presence of older star population in more massive nscs .
this is not at odd with the merger model .
actually , there is evidence of the presence of a certain fraction of young stars in nscs ( see for instance the milky way nsc ) and this implies that some star formation occurred there in relatively recent imes from some gas there present . given this ,
if we assume that the quantity of newly born ( in situ ) stars is the same in different , increasing in mass , galaxies hosting more massive nscs , it comes back naturally that these more massive nscs are , indeed , more massive because grown by a larger quantity of mass in decayed globular clusters which , consequently , have an increasingly old stellar population inside , due to the increased number fraction of old to young stars .
anyway , this is only a speculation that deserves a deeper investigation .
in the previous sections we showed that the dry - merger scenario provides scaling relations connecting the nsc masses with global parameters of their hosts . however , there are at least two effects which could prevent the formation of nscs , acting in competition with the dynamical friction process : the two - body relaxation mechanism and the tidal heating process . in this section
we study their effects on the formation of nscs and show that the scaling laws derived in this case still agree with observations in the whole range of galaxy masses . in the last section
, we neglected the effect of the disruption of cluster since , as we will show in this section , in small galaxies ( @xmath229m@xmath1 ) the dominant process in the formation of the galactic nucleus is the dynamical friction process , while in heavier galaxies tidal processes could prevent its formation . during its lifetime ,
a gc undergoes internal dynamical evolution experiencing two - body relaxation and suffering of external tidal perturbations that , in some cases , can lead to its total , or partial , dissolution .
actually , it is well known that two - body encounters between stars may bring some of them beyond the gc tidal boundary after few hundred times the typical two - body relaxation time @xcite .
@xcite , using results by @xcite , gave the following formula for the evaluation of the dissolution time of a cluster due to the effects of two - body encounters : @xmath230 where @xmath25 is the gc mass , @xmath22 is the distance from the galactic centre , @xmath23 the circular velocity at @xmath22 , and @xmath95 the eccentricity of the orbit .
moreover , gravitational encounters between the stellar system and a perturber ( which could be a black hole ) , the disc , or the nucleus , of the galaxy , could lead to the destruction of the system over a time comparable to the dynamical friction decay time .
this implies that the phenomenon of cluster destruction can not be neglected . in the case
in which the perturber is a point mass , a black hole , @xcite studied the effect of such perturbation on a stellar system of mass @xmath231 , in the hypothesis that the duration of the encounter is short compared to the internal crossing time of the cluster .
this is the impulse approximation , which assumes that as a consequence of the encounter , stars in the perturbed system suffer only a change in their velocities , but not in the positions .
moreover , due to the slow duration of the perturbation , the cluster trajectory can be approximated with a straight line . under this hypothesis ,
it is possible to show that the cluster , as a consequence of the gravitational encounter with the perturber of mass @xmath232 , gains an energy per unit mass : @xmath233 where @xmath143 is the relative velocity between the two objects , @xmath228 is the impact parameter and @xmath234 is the mean dimension of the perturbed cluster . a number of studies have been devoted to generalize this result to an extended spherical perturber with an arbitrary mass distribution @xcite , and to the case where the perturber is a spherical nucleus of stars embedded in a triaxial ellipsoid @xcite .
defining @xmath235 the ratio between the impulsive energy change due to a perturber of half mass radius @xmath236 and that caused by a point of same mass , @xmath232 , the total change in energy per unit mass caused by a mass distribution is given by : @xmath237 where the function @xmath235 drops rapidly to @xmath65 when @xmath238 approaches zero , while tends to @xmath201 for large values of @xmath238 and should be evaluated numerically .
if the energy change exceeds the internal gravitational energy of the system per unit mass @xcite : @xmath239 the cluster is disrupted .
the typical time over which this disruption occurs is @xmath240 where @xmath241 is the orbital period of the cluster and @xmath169 is the number of encounters within a period .
further , the encounters are charcterized by two extreme regimes : the catastrophic regime , if a single encounter could disrupt completely the system , and the diffusive regime , when the cumulative effect of encounters leads to the disruption of the system over a longer time . defining as @xmath242 the impact parameter that corresponds to an energy enhancement equal to the internal gravitational energy of the system , i.e. @xmath243 , it is possible to determine the duration of the encounter , @xmath244 , that is the typical time - scale which discriminates between the two regimes : hence a catastrophic collision occurs if the duration of the encounter is short compared to the crossing time of the cluster , @xmath245 ; on the other hand , slower encounter leads to the diffusive regime .
therefore , in the case of the catastrophic regime ( @xmath246 ) , the tipical disruption time is given by : @xmath247 where @xmath248 is the perturber density , and @xmath249 a constant . in the diffusive regime @xmath250 , instead , it is possible to show that the disruption time is given by : @xmath251 where @xmath252 is defined as : @xmath253 where @xmath254 and @xmath255 is defined above , and should be computed numerically .
tidal effects are accounted for in our calculations in the above framework , so to investigate their role on the expected value of the nsc mass . in figure
[ times ] we compare the dynamical friction time , @xmath256 , evaluated using equation [ tdf ] , the dissolution time , @xmath257 , given in equation [ tdis ] , and the tidal disruption time in the catastrophic regime @xmath258 for three values of the galaxy mass ( @xmath259m@xmath1 ) as a function of the distance , @xmath22 , from the centre of the host galaxy . on the other hand , the disruption time in the diffusive regime , @xmath260 ,
is not reported in the graph since it is sistematically greater than the other time - scales .
we performed the estimation setting the gc mass to @xmath261m@xmath1 and selecting a circular orbit .
looking at figure [ times ] , it is clear that while in small galaxies ( @xmath262m@xmath1 ) , the dynamical friction time is smaller than the disruption times over all length scales , in more massive galaxies it dominates only in a region around the centre of the galaxy , while in the range @xmath263 dominates the tidal effect due to the interaction between the cluster and the galactic nucleus , suppressing the role of dynamical friction process and , then , the consequent formation of a nsc .
m@xmath1 ) .
times are given in units of @xmath264 while positions are given in units of @xmath265 . in smaller galaxies ,
friction process dominates on all length scales , while in heavier systems the tidal disruption is much more rapid in a region @xmath266 around , suppressing the friction mechanism.,title="fig:",width=321 ] + m@xmath1 ) .
times are given in units of @xmath264 while positions are given in units of @xmath265 . in smaller galaxies ,
friction process dominates on all length scales , while in heavier systems the tidal disruption is much more rapid in a region @xmath266 around , suppressing the friction mechanism.,title="fig:",width=321 ] + m@xmath1 ) .
times are given in units of @xmath264 while positions are given in units of @xmath265 .
in smaller galaxies , friction process dominates on all length scales , while in heavier systems the tidal disruption is much more rapid in a region @xmath266 around , suppressing the friction mechanism.,title="fig:",width=321 ] the two competitive processes , make that the number of decayed clusters depends strongly on their space distribution . in the rnd , pls and gss models
, we set @xmath267pc as minimum distance from the galaxy centre of the gc sample .
clusters lying in the central @xmath268pc , which are massive enough ( @xmath269m@xmath1 ) or on eccentric orbits are likely to decay and give a large contribute to the final nsc mass .
figure [ rndn ] shows the fraction of the number of decayed clusters to the total number for the rnd model , considering or not the tidal effect .
when tidal disruption is considered , changing by roughly the same factor the nsc final mass.,width=321 ] while in small galaxies almost all the clusters have time to decay , even if the tidal disruption mechanism is active , in massive galaxies the perturbations induced by the external field corresponds to a reduction of the nsc mass .
however , the action of the tidal disruption mechanism does not change dramatically the final mass of the nsc ; in fact , the decrease in mass of the final nuclear cluster is reduced of only the @xmath270 respect to the case when the tidal action of the galaxy on the clusters is not considered .
hence , the effect of tidal processes in such models is not too important in the determination of the final nsc masses . on the other hand , since in the rho model
the minimum distance from the galaxy centre allowed for gc formation is given by the constraint that the gcs mass profile follows a dehnen profile , it could exceed @xmath271pc .
this implies that many clusters lie in the region in which the tidal effects dominate , affecting strongly the final mass of the nsc .
in this case , as we can see in figure [ rhon ] , the number of decayed clusters in very massive galaxies drops to zero avoiding the formation of nscs . , but for the rho model .
in this case , the number of decayed clusters for massive galaxies drops to 0 and there is no nsc formation.,width=321 ] the difference between the rho model and the others , where the gcs positions are sampled randomly , puts in evidence two interesting things : the mass distibution of the clusters is not very important in deriving the nsc mass , but instead , what care is how clusters are distributed within the galaxy .
this because at intermediate radial scales the disruption time is smaller than the decay time , while it is longer in the central region .
hence , a concentrated spatial distribution allows the formation of the nsc because clusters in the innermost region of the galaxy decay rapidly . in figure [ rhotid ]
we report the values of nsc masses with respect to host masses for all the models considered taking into account the tidal disruption process .
it is seen that in the case of the rho model the tidal interaction ( tidal heating ) inibhits the nsc formation for galaxies masses above @xmath272m@xmath1 , putting in evidence how important the spatial distribution of the clusters is .
it is interesting noting that in this case , the scaling relations found are again in good agreement with the observations , if we restrict the comparison to the actually observed range of masses ( @xmath273m@xmath1 ) .
moreover , it is relevant noting that the flattening observed in figure [ rhotid ] for the rho model has an observative `` counterpart '' : in fact it seems that galaxies more massive than @xmath108m@xmath1 do not host nscs .
our results suggest that in heavy galaxies , in which tidal processes act against the dynamical friction , should form a nsc lighter than expected .
as example , extrapolating from observation the mass of a nsc in a galaxy of @xmath127m@xmath1 , we should expect a nsc with @xmath274m@xmath1 ; however , the range of galaxy masses in which nscs are observed is dominated mainly by dynamical friction process .
considering instead that in more massive galaxies disrupting processes dominate , it is possible to have nscs in galaxies heavier than @xmath108m@xmath1 with masses few times @xmath275m@xmath1 . in this picture , nscs may form in heavy ( but not _ too _ heavy ) galaxies , but are too small to emerge from the galactic background . in a forthcoming paper
, we will investigate such matter in a more complete way .
we found that the dry - merger scenario predicts masses for nscs and scaling laws between them and their host galaxies in excellent agreement with observations .
the summary of our work is : 1 .
an analytical treatment to estimate the formation and growth of nscs masses has been developed ; 2 .
reliable galaxy models have been provided , as it has been shown comparing theoretical and observational global properties ; 3 . assuming for the gc system in a galaxy a power - law mass function and a uniform spatial distribution , the analytical predictions fit very well observations ( sect .
[ res ] ) ; 4 .
the consequences of different initial mass distributions of the set of gcs in the host galaxies on the nsc final mass have been investigated from a statistical point of view , by sampling , for each galaxy , its gcs and considering how many clusters were able to sink to the galactic centre within a hubble time ; 5 . by means of the statistical approach , we obtained some useful parameters , such as the gc mean mass and the number of survived clusters , which result in good agreement with observations ( sect.[stat ] and [ statmod ] ) ; 6 .
scaling laws which connect the nsc parameters with total mass , velocity dispersion and bulge mass of the host have been deduced ; the agreement found between all the models considered and observations indicates that the gc mass distribution does not play a crucial role in determining the final nsc mass ( sect .
[ laws ] ) ; 7 .
the role of tidal disruption mechanism has been investigated under different assumptions for the spatial and mass distributions of gcs in their host galaxies : in rnd , pls and gss models the tidal heating causes a decrease of predicted nsc masses from few percent in small galaxies ( down to @xmath10m@xmath1 ) to @xmath270 in heavier galaxies . on the other hand
, tidal disruption strongly affects nsc formation in the rho model , where the predicted nsc masses are almost constant in the range @xmath276m@xmath1 .
an important result is that the best comparison for the nsc mass versus galactic host mass correlation is obtained when gcs have an initial spatial distribution equal to that of the galaxy , assumed to be in the dehnen s form .
this because the relation shows the same flattening at high galactic masses than observed , while in the case of a different initial density profile for gcs , the @xmath277 relation keeps raising at high masses ( sect . [ tid ] ) ; 8 . finally , our results suggest that in galaxies with masses above few times @xmath108m@xmath1 , hosting central black holes more massive than @xmath11 m@xmath1 , tidal processes dominate over dynamical friction , leading to nscs too `` small '' to emerge from the galactic background and be detected .
this agrees with both ancient , general , results by @xcite , @xcite and @xcite and the more specific recent results by @xcite .
the overall conclusion is that the migratory - merger model for the formation of dense stellar agglomerates in galactic centers seems to be valid for a large range of types and masses of galaxies giving scaling relations in good agreement with observations and providing a possible explanation for the lack of nscs in bright galaxies .
an important topic which remains to be investigated thoroughly is what fraction of young to old stars actually reside in nsc , thing which can constitute an important test for the nsc formation models .
we thank the anonymous referee , whose comments and suggestions allowed us to improve the paper .
considering as density law the @xmath83 profile in equation [ mnsct ] leads to : with @xmath282 , @xmath283 and @xmath284{f}{_1}(a , b;c;z)$ ] the gauss hypergeometric function , defined as : @xmath285{f}{_1}(a , b;c;z)=\frac{\gamma(c)}{\gamma(b)\gamma(b - c)}\int_0 ^ 1 \frac{t^{b-1}(1-t)^{c - b-1}}{(1-tz)^a}dt,\ ] ] | nuclear star clusters ( nscs ) are often present in spiral galaxies as well as resolved stellar nuclei ( sni ) in elliptical galaxies centres . ever growing observational data indicate the existence of correlations between the properties of these very dense central star aggregates and those of host galaxies , which constitute a significant constraint for the validity of theoretical models of their origin and formation . in the framework of the well known migratory and merger model for nsc and sn formation , in this paper
we obtain , first , by a simple argument the expected scaling of the nsc / sn mass with both time and parent galaxy velocity dispersion in the case of dynamical friction as dominant effect on the globular cluster system evolution .
this generalizes previous results by @xcite and is in good agreement with available observational data showing a shallow correlation between nsc / sn mass and galactic bulge velocity dispersion .
moreover , we give statistical relevance to predictions of this formation model , obtaining a set of parameters to correlate with the galactic host parameters .
we find that the correlations between the masses of nscs in the migratory model and the global properties of the hosts reproduce quite well the observed correlations , supporting the validity of the migratory - merger model .
in particular , one important result is the flattening or even decrease of the value of the nsc / sn mass obtained by the merger model as function of the galaxy mass for high values of the galactic mass , i.e. @xmath0m@xmath1 , in agreement with some growing observational evidence .
[ firstpage ] galaxies : nuclei , galaxies : star clusters ; methods : numerical . |
galactic winds are one of the main mechanisms by which galaxies expel metals , dust , cosmic rays , magnetic fields , etc , to the intergalactic medium .
this feedback makes winds an important phenomenon in our global understanding of galaxies .
these winds are usually associated to galaxies with strong nuclear activity or with starbursts ( see * ? ? ?
* and references therein ) which are powerful enough to feed the winds .
but very little is known of this phenomenon in quiescent , inactive galaxies . in this paper
, we explore this phenomenon in a relatively quiet galaxy .
m100 is a nearby ( d=16.1 mpc ) galaxy with low nuclear activity ( cataloged as transition liner / hii by * ? ? ?
* ) and a moderate star formation rate of a few solar masses per year ( according to its ir luminosity of @xmath2 , * ? ? ?
despite this apparent quiescence , we will show that this galaxy presents an unexpected fast wind emanating from its nuclear region .
winds have been studied in highly inclined galaxies where the extraplanar gas can be directly seen through imaging of the emission lines of the ionized gas ( see * ? ? ?
this technique has the inconvenient of lacking information on the kinematics of the outflow .
alternatively , winds have also been studied through detection in the spectra of a blueshifted component in some interstellar absorption lines ( see for example * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this latter approach lacks spatial resolution , and consequently does not provide information on the wind geometry .
we have taken a new approach by using integral field spectroscopy , which combines spectroscopic and imaging techniques .
we are therefore able to detect absorbing material throughout the field of view and to produce a map of the distribution and kinematics of the absorbers .
in addition , we have been able to discriminate kinematically the ionized component of the outflowing material and to produce maps for the distribution and kinematics of this hotter phase of the wind .
this approach provides a unique simultaneous global view of these two phases of a galactic wind .
the data analyzed in this paper were obtained on 2002 march 16 at the observatorio del roque de los muchachos on la palma with the fibre system integral @xcite at the william herschel telescope .
the observational setup and basic data reduction are detailed in @xcite .
we only mention here the aspects more relevant to the present work .
the observed spectral range ( 5600 - 6850 @xmath3 ) contains a few strong interstellar emission lines ( @xmath4 , [ nii]@xmath5 , [ sii]@xmath6 ) as well as several stellar absorption lines ( nad doublet and weaker , mostly blended fei and cai lines ) .
a first inspection of the spectra in the fibres with the highest s / n reveals an unusual ratio of the nad absorption doublet with respect to the other stellar lines , with the nad lines being deeper than expected ( see fig . 1 in * ? ? ?
the most straightforward explanation for this is contamination by interstellar nad absorption .
we therefore proceed to separate the stellar and interstellar components in the fibre spectra .
we use a synthetic stellar population of 1 gyr with solar metallicity to subtract the stellar component of the spectra ( see * ? ? ?
* for a detailed explanation ) .
this synthetic spectrum is used as a template to fit the spectrum of each fibre .
the free parameters in the fit are the velocity , velocity dispersion and ( global ) line strength .
it is important to remove from the fitting the wavelength range containing the nad doublet , as it is likely to be contaminated with interstellar absorption .
the result of this procedure reproduces very well the stellar features in the spectra except for the nad lines .
the stellar template fits the redmost part of the nad lines , leaving a strong interstellar residual which is clearly blueshifted with respect to the stellar component ( see fig . 1 in * ? ? ?
the presence of blueshifted interstellar absorption lines is an unambiguous signature of outflowing material ( see for example * ? ? ?
* and references therein ) .
after subtraction of the stellar component , a spectrum of the interstellar material remains ( with both , emission and absorption lines ) .
we analyze these interstellar spectra throughout the rest of this paper .
most works on the study of outflows or galactic winds by means of the nad absorption ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) do either select samples of objects where the nad doublet is dominated by the interstellar component , or try to make a global rough correction for the stellar contribution .
we present here a spatially resolved separation of these two components . with this aim
, we apply the analysis described in section [ sectobs ] to the spectra of the 2d collection obtained from the central region of m100 , deriving full 2d maps of the distribution and kinematics of the absorbing material .
a first result of this study is that regions where the nad is dominated by the interstellar component coexist in the same galaxy with regions where the stellar component is dominant and there is very little , if any , cold interstellar absorbing gas .
we would like to point out at this point that 2d spectroscopy has been critical in detecting the wind thanks to its ability to isolate the locations where interstellar absorption is relevant .
indeed , the wind might have been unnoticed in a spectrum taken in a large aperture where stars dominate the nad absorption . in fig .
[ nadew ] we present the equivalent width map for the interstellar nad doublet , which ranges between 800 and 2100 @xmath7 throughout the field of view ( lower values of the ew have been masked out ) .
the distribution of nai in the map is clearly elongated along the bar position angle @xcite with a central maximum and two knots .
the central maximum coincides with the nuclear starburst and the location of the two knots matches that of the nir knots corresponding to active star forming regions .
these knots are also related to two strong concentrations of absorbing material at these locations that are clearly seen in the co map ( from * ? ? ?
* ) . with the exception of these knots ,
most of the gas responsible for the nad absorption is concentrated in the region interior to the star formation ring .
the determination of the column density of nai from the spectra is , in general , not straightforward ( see * ? ? ?
* and references therein ) .
in the case under study , the situation is worsened because the stellar component must be removed , which can seriously affect the shape of the final spectra .
thus , the fitting of optical depth and covering factor as performed by @xcite might not be fully reliable for all spectra ( particularly but not exclusively for those with relatively low s / n ratio ) .
the equivalent width is much less affected by the above - mentioned effects and is therefore a more * reliable * estimator of the column density .
we then estimate the column density of nai from the equivalent width of the nad doublet using a curve of growth technique .
we have taken a hybrid approach for this purpose . in first place
, we use a direct fitting as in @xcite on a selected subset of high s / n spectra to estimate some average parameters of the redder , @xmath8 , line : covering fraction , @xmath9 , doppler parameter , @xmath10 , and optical depth , @xmath11 . in a second step
, we use these average values and the curve of growth to find the following relation between the nai column density and the equivalent width of the nad doublet : @xmath12 . finally , to convert n(nai ) to hydrogen column density
, we adopt the relation : @xmath13=\beta\log[n(h)]+\gamma}$ ] with @xmath14 and @xmath15 from @xcite as @xcite .
the resulting map of hydrogen column density is shown in fig .
[ colhi ] . from this map
we have calculated the total hydrogen mass in the wind to be @xmath16 .
@xmath17.,width=294 ] the velocity at each fibre is calculated by fitting two gaussians to the nad doublet in the ( stellar removed ) spectra .
errors in the velocity determination range from 20 km / s to 70 km / s , with the highest errors occurring at the locations with the lowest ew .
the velocity map of the absorbing gas ( fig . [ nadvel ] ) presents a rotation pattern with a major kinematic axis orientation and velocity amplitude that clearly resembles that of the ionized gas in the disk ( see * ? ? ?
* ) . however , this rotation pattern is globally blueshifted with respect to the underlying ionized gas in the galactic disk ( @xmath18 1567 km / s ) by @xmath0 115 km / s .
the detection of this rotation pattern in the absorbing gas indicates that this neutral phase is very likely disk material that has been dragged upwards by the galactic wind , but that still keeps most of its angular momentum .
it is probably material near the galactic disk , on the interface between the fast , hot gas filling the cavity , and the underlying quiescent gas in the galactic disk . in order to measure the motion of the absorbing gas with respect to the underlying galaxy
, we have subtracted the velocity field of the ionized gas in the disk ( see * ? ? ?
* ) from the nad velocity map .
the resulting projected velocities range from @xmath0 -210 km / s to roughly -55 km / s with an average of -115 km / s .
the distribution and kinematics of the ionized gas in the circumnuclear region of m100 has been extensively studied by several authors ( see for example * ? ? ? * and references therein ) . however , after our finding of a strongly blueshifted component in the neutral gas traced by the nad lines , it is reasonable to wonder whether this outflowing gas has an ionized counterpart , and if this is the case , whether we can detect it .
this detection has a major difficulty with respect to the neutral gas , namely , the fact that we see the outflowing material in front of the bright background galaxy . fortunately , integral field spectroscopic data provides kinematical information , and we may try to kinematically discriminate these two ionized components ( see for example * ? ? ?
* ; * ? ? ?
visual inspection shows indeed a blue shoulder in the emission lines in many spectra taken with the sb3 and sb2 bundles ( see , for instance , fig .
[ twocompspec ] ) .
unfortunately , the spectral resolution with the sb3 fibre bundle is not high enough ( 4.8 @xmath19 ) , and this blue shoulder is only visible in a few fibres . the sb2 bundle offers a much better ( 2.8 @xmath19 ) spectral resolution , and the line asymmetry is visible in many more fibres .
we therefore perform this kinematical discrimination in the sb2 bundle only . in order to study this blueward component
we try to fit two kinematical components to the emission lines of each spectra . to do this
we simultaneously fit two sets of five spectral lines ( @xmath4 , [ nii]@xmath5 , [ sii]@xmath6 ) .
we assume that , for each set , all these five lines have the same velocity and velocity dispersion .
the theoretical 3:1 line ratio is fixed for the [ nii ] doublet .
[ twocompspec ] shows the fit to a spectrum where the double component is clearly detected . a kinematical double component is said to be detected only if the five spectral lines are detected in the two sets and the reduced @xmath20 of the fit shows a significant improvement of at least 50% with respect to the fit to a single kinematical component . as a final check ,
each fit is inspected visually and in case of doubt a single component is kept .
when this procedure is finished , there is still some ambiguity at some positions on which of the two components belongs to the background galaxy and which to the outflow .
continuity in the produced maps ( particularly in the velocity field of the background galaxy ) , and the higher velocity dispersion of the secondary component are used to discriminate between the two .
intensity of the wind component ( red ) , hst / acs f555w ( green ) and @xmath21 intensity of the underlying galaxy disk ( blue ) .
intensities are arbitrarily scaled.,width=294 ] .,width=294 ] in figs .
[ haint2 ] , [ havel2 ] we present the intensity and velocity maps corresponding to the wind component of the ionized gas phase .
the spatial distribution of the ionized gas wind component is concentrated around the nuclear starburst ( although somewhat elongated along the major axis ) with a bright ridge that extends towards the sw and a few weaker patches .
the central and sw ridge may be enhanced emission from the walls of the expanding shell seen in projection .
the ionized component of the wind fills the region interior to the ring . if there is indeed an expanding shell in this region it could be possible that the star forming ring is somehow related to the wind .
the velocity field of the ionized gas shows no rotation pattern .
nevertheless , the velocity field of the underlying ionized gas in the galaxy disk has been subtracted to view the motion of the wind with respect to it .
it is blueshifted in most locations if we make exception of a redshifted region in the n. the maximum blueshifted velocity is located in the nucleus and on a ridge extending from the s to the nw , which can be interpreted in terms of the projection of the walls of an outflowing cone or shell .
these results are compatible with a bi - polar outflow ( in which case we could identify the region of receding velocities with the dimmed wind on the far side of the galaxy ) .
the wind component has [ nii]/h@xmath1 ratios in the 1.2 to 2.7 range and [ sii]/h@xmath1 in the 0.4 to 1.5 range , typical of shock ionization .
the galactic component shows lower values , with [ nii]/h@xmath1 of 0.9 for the nucleus , and @xmath22/h\alpha<1.4 $ ] ( @xmath23/h\alpha<0.7 $ ] ) for the region inside the star forming ring and of about 0.2 in this ring ( see fig .
[ liner ] ) .
this indicates that even for the galactic component , shock ionization is not negligible in the nucleus , and is even more important in the region inside the star forming ring .
this region also shows the highest velocity dispersion ( up to 100 km / s ) of the region . whether a hidden wind component sharing the disc kinematics is responsible for the high line ratio and velocity dispersion in the galactic component in this region around the nucleus , remains unclear . on the contrary
, the ionization in the stellar formation ring appears not to be directly related to the wind .
the velocity dispersion of the ionized gas in the wind is in the range between 150 and 230 km / s , while for the galactic component this value is much lower , ranging between 50 and 100 km / s .
in previous sections we have reported the discovery of an important wind in the central region of m100 .
this is certainly an unexpected finding , particularly because its extent ( @xmath24 ) and velocity ( @xmath25 ) are quite large for a galaxy with a moderate nuclear and/or star forming activity .
in fact , the velocity of this wind is comparable to the values found by @xcite for ulirgs . we have estimated the total mass of gas in the wind to be of about @xmath26 .
although this value is about an order of magnitude lower than the values found by @xcite , we must keep in mind that ulirgs are forming stars one hundred times faster .
this comparison makes the wind in m100 even more spectacular and surprising .
although the geometry of the wind is uncertain , our findings indicate that the galactic wind is most likely driven uniquely by the nuclear starburst .
the presence of rotation in the kinematics of the absorbing neutral gas indicates that it is likely disk ism near the plane that has been entrained by the wind at the interface with the hot expanding shell .
the fact that we have not detected a rotation pattern in the wind component of the ionized gas , together with the high [ nii]/h@xmath1 and [ sii]/h@xmath1 ratios , indicates that this is shocked gas at the top and/or walls of the expanding shell .
this scenario is also fully compatible with the irregular geometry of the extended continuum radio emission found by @xcite , and the diffuse x ray emission reported by @xcite .
this new scenario also opens the possibility that the nuclear star forming ring in m100 is related to this wind , either by direct compression of the ism at the edges of the shell , or by material of the wind falling down to the disk .
in fact , @xcite found evidence for sequentially triggered star formation in the ring of m100 , compatible with this scenario . taking into account that we find projected velocities of up to @xmath0 210 km / s and velocity dispersions of up to 220 km / s for the ionized gas , there is a non negligible fraction of the gas ( with @xmath27 ) which might be able to escape the gravitational potential of the galaxy ( taken from a singular isothermal sphere as in @xcite ) .
although this rough estimate should be taken with caution , it indicates that winds of this type in the abundant galaxies with low nuclear activity may be very important in polluting the igm with dust and metals .
the relevant question then is : how representative is the wind in m100 of similar phenomena in other normal , relatively unactive galaxies ? .
this question can only be answered by observing a significant sample of low to moderate activity galaxies with the techniques used in this paper .
this paper has been supported by the `` secretara de estado de poltica cientfica y tecnolgica '' ( aya2004 - 08251-c02 - 02,esp2004 - 06870-c02 - 02,aya2006 - 02358 ) .
this research has made use of ned which is operated by the jpl , caltech , under contract with nasa .
we have used observations made with the nasa / esa hst , obtained from the data archive at the stsci . j.j.v . acknowledges support from the consejera de educacin y ciencia de la junta de andaluca . a.c.m . acknowledges the support from the ministerio de educacin y ciencia .
we also acknowledge the support of the rtn euro3d : `` promoting 3d spectroscopy in europe '' .
we are grateful to the anonymous referee for valuable comments that have helped to improve the quality of the paper . | we report the discovery of a galactic wind in the central region of the galaxy m100 . this result is based on a careful 2d spectroscopic study performed on observations made with the fibre system integral on the wht .
the primary evidence of the wind is the presence of blueshifted interstellar nad absorption lines .
the velocity field of the absorbers show a clear rotation pattern but globally blueshifted ( @xmath0 -115 km / s ) with respect to the systemic velocity of the galaxy .
the emission lines also present a blueward component arising from the ionized gas phase of the galactic wind .
the velocity field of the ionized gas wind component shows no evidences of rotation but exhibits a pattern that can be interpreted in terms of the projection of an outflowing cone or shell .
the wind component has [ nii]/h@xmath1 ratios of about 1.8 , typical of shock ionization .
the ionized component of the wind can be identified with an expanding shell of shocked gas , and the neutral component with disk gas entrained in the wind at the interface of the expanding shell with the galactic ism .
the galactic wind seems to be driven uniquely by the nuclear starburst .
our analysis indicates that a non negligible fraction of the wind material might escape to the igm . in this case , if the wind detected in m100 were representative of similar phenomena in other galaxies with low to moderate activity , the current estimates of metal and dust content of the igm might be drastically underestimated .
[ firstpage ] galaxies : individual : m100 galaxies : ism ism : jets and outflows . |
in the following we use the clauser - horne - shimony - holt ( chsh ) form of bell inequality @xcite , which is the most suitable for an experimental test of nonlocality in bipartite quantum systems composed by spin-@xmath0 objects . let the operator @xmath7 be a spin observable with eigenvalues @xmath8 associated to the spin @xmath9 defined by @xmath10 , where @xmath11 is the unit vector indicating an arbitrary direction in the spin space and @xmath12 is the pauli matrices vector .
the chsh - bell inequality associated to a two - spin-@xmath0 state @xmath13 is then @xmath14 @xcite , where @xmath15 is the bell function defined as @xmath16 where @xmath17 is the correlation function of observables
@xmath18 , @xmath19 and @xmath20 . if , given the state @xmath13 , it is possible to find a set of angles @xmath21 and @xmath22 such that the chsh - bell inequality is violated ( @xmath23 ) the correlations are nonlocal and can not be reproduced by any classical local model . a necessary and sufficient condition for violating the chsh - bell inequality by an arbitrary two - spin-@xmath0 mixed state is known @xcite
. once introduced the @xmath24 real matrix @xmath25 , defined by the matrix elements @xmath26 , where the @xmath27 s are the pauli matrices , one must build the matrix @xmath28 , a symmetric matrix which can be therefore diagonalized ( @xmath29 indicates the transposed matrix of @xmath25 ) .
the maximum of the bell function @xmath15 results then to be given by @xmath30 , where @xmath31 and @xmath32 s are the three eigenvalues of the matrix @xmath33 . a procedure to obtain the angles giving the maximum value of the bell function
is also provided @xcite .
in this section we give the maximum value of the bell function for the class of states whose density matrix @xmath34 , in the standard computational basis @xmath35 , has a x structure of the kind @xmath36 this class of states is sufficiently general to include the two - qubit states most considered both theoretically and experimentally , like bell states ( pure two - qubit maximally entangled states ) and werner states ( mixture of bell states with white noise ) @xcite .
moreover , a x structure density matrix arises in a wide variety of physical situations @xcite .
a further remarkable aspect of the x states is that , under various kinds of dynamics , the initial x structure is maintained during the evolution @xcite .
using the criterion previously described to obtain @xmath37 , the three eigenvalues @xmath38 s , in terms of the density matrix elements , are found to be @xmath39 as also already reported @xcite .
being @xmath40 always larger than @xmath41 , the maximum of bell function for x states results to be @xmath42
here we give the obps , that is the angles @xmath43 s and @xmath44 s maximizing the bell function @xmath15 of eq .
, for the x states class of eq . . following a known procedure @xcite we find the explicit obps , which result in two different sets of angles in correspondence of the two regions @xmath45 ( region 1 ) and @xmath46 ( region 2 )
* set 1 : @xmath45 . * the obps in this case
are given by @xmath47/2,\quad\phi'_1=0,\nonumber\\ \phi_2&=&\phi'_2=\left[\arg(\rho_{23})-\arg(\rho_{14})\right]/2,\end{aligned}\ ] ] where the the function @xmath48 is @xmath49 if @xmath50 and @xmath51 if @xmath52 .
* set 2 : @xmath46 . *
the obps in this case are @xmath53/2,\nonumber\\ \phi'_1&=&\phi_1+\mathrm{sign}(|\rho_{23}|-|\rho_{14}|)\pi/2,\nonumber\\ \phi_2&=&\frac{\arg(\rho_{23})-\arg(\rho_{14})}{2}+\arctan\sqrt{\frac{u_3}{u_1}},\nonumber\\ \phi'_2&=&\frac{\arg(\rho_{23})-\arg(\rho_{14})}{2}-\arctan\sqrt{\frac{u_3}{u_1}}.\end{aligned}\ ] ] let us observe that the angles of the two sets are not equal when @xmath54 .
this means that , when this condition is satisfied , two different obps and so experimental settings do give the same value of @xmath3 .
moreover , considering the obps as function of the quantum state , to a slight continuous change of the quantum state crossing the point @xmath54 , it corresponds an unexpected finite `` jump '' of the suitable experimental settings .
it is worth to emphasize the importance to have an analytical result for the two - qubit configuration here considered .
in fact , if an analytical criterion to find obps was absent , numerical procedures would be necessary .
this is the case , for example , in systems of more than two qubits where mermin inequalities are considered to test nonlocality @xcite . on the basis on the two - qubit result , one may expect that a finite jump of the obps take place even for the multiqubit configuration . however , if this jump was present , starting from one of the obps sets , numerical procedures could give rise to serious problems in identifying the point of this possible obps jump because of the finite distance between the two sets of parameters .
it is clear that the angles of eqs . and , which determine
the orientations of the measurement apparatus in real space giving @xmath3 for two spin-@xmath0 objects , do correspond to different experimental settings for a general two - qubit system . in cavity
qed , for example , qubits are represented by either two - level rydberg atoms or cavities having one or zero photons ; ramsey zones followed by field ionization detectors are then the devices suitable for measuring the qubit along an arbitrary pseudospin direction , whose angles @xmath55 are controlled by the classical microwave field amplitude and the atom - field interaction time inside the ramsey zone @xcite . in the context of traveling polarized photons , where the chsh - bell inequality was originally applied @xcite ,
the spin levels are usually coded in the horizontal ( @xmath56 ) and vertical ( @xmath57 ) polarizations of the photon , the angles @xmath44 s are usually set at zero and the angles @xmath43 s are related to the angles @xmath58 s of photon polarizers by @xmath59 @xcite ; the measurement is finally performed by photodetectors . in solid - state physics with josephson phase qubits , coded by the two first levels of a well potential ,
pseudospin directions are selected by rotating the qubit states using microwave pulses before performing measurements by a pulse of current along the @xmath60 axis of the bloch sphere ( basis states of the qubit ) @xcite .
so far we have found the optimized parameters ( obps ) that give the best condition for the bell test on a given two - qubit x state . from an experimental point of view ,
one firstly needs to know the density matrix of the state on which to measure @xmath3 and confirm if there is a violation of the chsh - bell inequality .
this can be done by quantum state tomography @xcite .
therefore , in presence of a time - evolving quantum state , one may establish the form of the state with its density matrix elements at the time of the measurement .
the experimental settings of the devices for the bell test can be then appropriately adjusted according to our indications . in this section
we give a dynamical example where the obps not only are crucial to reveal nonlocality , but they also present an unexpected behavior . consider a system of two identical independent qubits , namely 1 and 2 , each embedded in a zero - temperature bosonic environment .
each qubit has ground and excited states @xmath61 , @xmath62 separated by a transition frequency @xmath63 . for our study
we need the two - qubit reduced density matrix at the time @xmath5 .
this can be accomplished by exploiting a procedure based on the knowledge of the single - qubit dynamics @xcite .
the dynamics of the single part `` qubit - environment '' is described , in the case of bosonic reservoir under rotating wave and dipole approximations , by the hamiltonian @xcite @xmath64,\ ] ] where @xmath65 , @xmath66 are the qubit raising and lowering operators , @xmath67 , @xmath68 are the photon creation and annihilation operators and @xmath69 is the coupling constant of the mode @xmath70 with frequency @xmath71 . it is found that , for a zero - temperature environment , the single - qubit evolution under the hamiltonian of eq .
( [ hamiltonian ] ) can be described in terms of a single function @xmath72 , whose square modulus @xmath73 can be linked to the single - qubit excited state population . in the basis @xmath74 , the single - qubit reduced density matrix @xmath75 can be then written as @xcite @xmath76 the fact that the density matrix @xmath75 can be expressed in terms of a single function @xmath72 is general , the information on the environment spectral density and coupling constants being contained in the explicit time dependence of the function @xmath72 itself .
the two - qubit density matrix @xmath77 is then obtained @xcite , its elements depending only on their initial values and on the function @xmath72 , regardless of the environment structure .
hereafter , for the sake of simplicity , we shall indicate @xmath78 .
we choose as two - qubit initial state the extended werner - like state defined as @xcite @xmath79 where @xmath80 is the purity parameter , @xmath81 the @xmath82 identity matrix and @xmath83 is the one - excitation bell - like state with @xmath84 positive real and @xmath85 .
@xmath86 reduces to a werner state for @xmath87 and @xmath88 , that is when its pure part becomes a bell state . for @xmath89
the state above becomes a totally mixed state , while for @xmath90 it reduces to the bell - like ( pure ) state @xmath91 .
this state has an x form which is maintained during the dynamics due to the spin - boson hamiltonian of eq . .
we can analyze the dependence of the dynamics of @xmath3 on the initial mixedness and degree of entanglement .
starting from the two - qubit state defined above , the expressions of @xmath40 , @xmath92 and @xmath41 of eq .
are found to be in terms of @xmath93 as @xmath94 ^ 2.\ ] ] the condition @xmath54 determines a closed surface in the 3d space ( @xmath95 ) given in fig . [
fig : qsurface ] , with two values @xmath96 , @xmath97 for a fixed couple @xmath98 .
* surface of the quantum state condition @xmath54 in the @xmath99 space . * given @xmath98 two values of @xmath93 ( @xmath96 and @xmath97 ) satisfy @xmath54 .
the points outside the surface represent the region 1 where @xmath100 , while those inside the surface represent the region 2 where @xmath101.,width=5,height=4 ] the points enclosed by this surface represent the region 2 ( @xmath101 ) , while the points outside represent the region 1 ( @xmath100 ) . dynamically @xmath93 changes with time and fig .
[ fig : qsurface ] clearly shows that , given @xmath102 and @xmath80 , when @xmath93 crosses the surface at time @xmath5 , the obps immediately before and after this time are given respectively by the two different sets 1 or 2 .
this in turn gives place to a discontinuous change with time of the experimental settings giving the maximum of bell function .
* dynamical behavior of @xmath103 and @xmath104 with abrupt `` jump '' of the obps . * * a * , plot of @xmath105 ( solid curve ) and @xmath106 ( dashed curve ) as a function of @xmath93 , for @xmath107 and @xmath90 .
the inset evidences the violation region where @xmath108 .
the value @xmath109 indicates the change from branch 1 to branch 2 ( going from right to left ) , while between the values @xmath110 and @xmath111 only @xmath112 is larger than two . *
b * , the spin-@xmath0 measurement vectors determined by the obps are illustrated ( for @xmath113 ) in correspondence of @xmath109 .
the abrupt jump of the obps is well evidenced while passing from @xmath114 to @xmath115 .
immediately before and after the crossing point @xmath109 , the spin measurement vectors lie on the @xmath116-@xmath60 plane and @xmath116-@xmath117 plane , respectively.,width=8,height=8 ] in order to examine this surprising aspect , we plot in fig .
[ fig : b1b2measdirections ] both @xmath105 and @xmath106 as a function of @xmath93 for @xmath107 and @xmath90 . in this case
the initial state of eq . is a pure non - maximally entangled state and @xmath118 . in the plot
one moves from right to left with time ( @xmath93 going from 1 to 0 ) and @xmath3 is given by @xmath103 ( branch 1 ; solid curve ) before @xmath109 and by @xmath104 ( branch 2 ; dashed curve ) after it .
the region of violation of chsh - bell inequality ( @xmath119 ) is zoomed in the inset . at value
@xmath109 there is the passage from branch 1 to branch 2 and the corresponding abrupt discontinuous `` jump '' of the obps ( see part * b * of fig .
[ fig : b1b2measdirections ] ) which pass from set 1 to set 2 .
moreover , for values @xmath120 , using the parameters of set 1 ( giving the maximal violation when @xmath121 ) one finds no violation of bell inequality while using those of set 2 one finds a violation .
the environment determines the explicit time evolution of @xmath122 . for markovian environments
@xmath73 decays exponentially @xcite and the values @xmath109 and @xmath110 are typically reached after a short time . differently for structured environments inhibiting the population decay , such as photonic crystals ,
@xmath73 decay can be slow @xcite and the crossing value @xmath109 could be reached after a rather long time .
another case is that of non - markovian environments where collapses and revivals of @xmath73 may be present , such as in high-@xmath123 cavities @xcite , and the surface of fig .
[ fig : qsurface ] could be crossed several times during the evolution .
therefore , when one needs to detect the presence of nonlocal correlations without ambiguity , the striking behavior of the obps here shown highlights the importance of our analysis to fix the most appropriate experimental settings to reveal nonlocality .
in this paper we have found the optimized parameters determining the experimental settings suitable to reveal at best the presence of nonlocality in a bipartite spin-@xmath0-like system .
we have reported the explicit dependence on the state of these optimized bell parameters ( obps ) for a general class of quantum states .
we have found that this class of states can be divided in two parts each of them corresponding to a different set of obps .
we have also found that even if the state of the system changes continuously , as when it evolves with time , a surprising abrupt jump of these obps may occur in proximity of some states . in particular , we have shown that these jumps of the experimental settings that maximize the bell function occur during the evolution of a paradigmatic two - qubit open system .
this phenomenon must therefore be taken into account in any experimental test aiming at obtaining the best confirmation of the presence of nonlocality in a quantum system . | nonlocality lies at the core of quantum mechanics from both a fundamental and applicative point of view .
it is typically revealed by a bell test , that is by violation of a bell inequality , whose success depends both on the state of the system and on parameters linked to experimental settings . this leads to find , given the state , optimized parameters for a successful test .
here we provide , for a quite general class of quantum states , the explicit expressions of these optimized parameters and point out that , for a continuous change of the state , the corresponding suitable experimental settings may unexpectedly vary discontinuously .
we finally show in a paradigmatic open quantum system that this abrupt `` jump '' of the experimental settings may even occur during the time evolution of the system .
these jumps must be taken into account in order not to compromise the correct detection of nonlocality in the system .
violation of a bell inequality indicates , in two - qubit ( spin-@xmath0-like ) systems , the presence of quantum correlations nonreproducible by any classical local model ( nonlocal correlations ) @xcite .
a violation is identified by values of the bell function @xmath1 , consisting of a combination of correlations averages , larger than a maximum classical threshold @xmath2 . for a given system
the bell function does depend both on the state of the system and on parameters linked to spin - like measurement directions that correspondingly fix the experimental settings .
many experimental tests have proven the occurrence of bell inequality violations in different contexts as entangled - polarized photons @xcite , photon - atom systems @xcite , atom - atom systems @xcite and recently also superconducting josephson phase qubits @xcite .
bell inequalities have also been extended to characterize the presence of nonlocality in multiqubit systems ( mermin inequalities ) @xcite .
nonlocal correlations also play a crucial role in quantum information science , as in the realization of device - independent and security - proof quantum key distribution protocols @xcite . for pure states of any degree of entanglement
it is possible to show that , by appropriately choosing the bell parameters , violation of bell inequality always occurs @xcite .
this is not the case for mixed states which may present entanglement without any corresponding direct violation of a bell inequality @xcite . indeed , mixed states take place as a consequence of unavoidable environmental noise which causes decoherence and a nonlocal correlations decay during the evolution @xcite . in order to test at best the presence of nonlocality , it is essential to adopt optimized bell parameters ( obps ) so that the maximum value of the bell function @xmath3 is measured .
these obps in turn determine the appropriate experimental settings which allow to perform the optimal bell test in order to state that , if @xmath4 , quantum nonlocality is surely nondetectable directly . because of the dependence of the obps on the state of the system , these are expected to change in time .
if one wants to exploit nonlocal features of the system at time @xmath5 , it is required to detect at best the presence of quantum nonlocal correlations at that time .
therefore , one needs to know how these obps vary in time so to correspondingly adjust the experimental settings and measure @xmath6 . a procedure to obtain the obps and the corresponding maximum value of the bell function for an arbitrary two - qubit state
is known @xcite .
however , the explicit expressions of these obps have never been reported before but for some very limited class of pure quantum states @xcite .
here we firstly give the explicit expressions of the optimized bell parameters for a quite general class of pure and mixed quantum states showing moreover that , changing the quantum state in a continuous way , the obps may unexpectedly undergo a discontinuous change .
these expressions are given for a two - spin-@xmath0 system where the parameters are the angles fixing directions in real space .
nevertheless , these parameters can be translated into the suitable experimental settings for any two - qubit system realization in different physical frameworks .
we finally consider the inescapable environmental effects on the evolution of the two - qubit system and their influence on the bell test . in a paradigmatic open quantum system
we show that the abrupt discontinuous change of the obps may occur during the time evolution and this must be considered in order not to compromise the correct detection of nonlocality in the system . |
phase locking or synchronization effects can arise in coupled oscillators when the different frequencies lock together over a certain range of parameter space , an effect that was first reported by huygens for the synchronization of pendulum clocks @xcite .
phase locking has been extensively studied for numerous dynamical systems ranging from a pair of coupled oscillators to an entire coupled oscillator array @xcite .
a single particle moving over a tilted one - dimensional washboard potential can also experience phase locking when an additional ac driving force is applied .
the substrate periodicity produces intrinsic periodic modulations of the particle velocity in the absence of an ac drive which increase in frequency as the magnitude of the tilt or dc drive increases .
addition of an external fixed - frequency ac drive produces locking regimes in which the average dc velocity remains constant even as the magnitude of the dc drive is increased .
the same picture can be applied to josephson junctions , where the analog of a velocity - force curve is the voltage - current curve , which exhibits a series of phase locking regions called shapiro steps under an applied ac drive for single junctions @xcite and coupled arrays of junctions @xcite .
one of the hallmarks of shapiro steps is that the step width oscillates as a function of the ac drive amplitude @xcite .
shapiro step phenomena also arise in dc and ac driven charge density waves @xcite , spin density waves @xcite , and frenkel - kontorova models consisting of commensurate or incommensurate arrangements of particles moving over ordered or disordered substrates @xcite . in the case of vortex motion in type - ii superconductors , martinoli
_ et al . _ reported the first observation of shapiro steps for dc and ac driven vortices interacting with a periodic one - dimensional ( 1d ) substrate created by periodic thickness modulations of the sample @xcite , while similar effects were observed for vortices driven over 1d @xcite or two - dimensional ( 2d ) @xcite periodic substrates .
more recently , shapiro steps have been found for ac and dc driven colloidal particles moving over a quasi-1d periodic substrate @xcite .
shapiro steps can also occur when a lattice of collectively interacting particles moves over a _
substrate under combined dc and ac drives . here , the effective elastic coupling between the particles comprising the lattice generates an intrinsic washboard frequency that can lock to the applied ac driving frequency .
such steps have been studied for vortices moving over random disorder @xcite or through confined channel geometries @xcite . for particles confined to 2d and moving over a quasi-1d substrate , both
the ac and dc drives must be applied in the same direction to produce shapiro steps ; however , for vortices moving over 2d periodic or egg - carton substrates , it is possible to obtain what are called transverse phase locking steps when the ac drive is perpendicular to the direction of the dc drive @xcite .
these phase locking steps are distinct from shapiro steps , and their widths grow quadratically with increasing ac amplitude rather than showing the oscillatory behavior associated with shapiro steps .
phase locking effects can also occur for overdamped particle motion in 2d periodic systems under combinations of two perpendicular ac drives , producing localized and delocalized motion as well as rectification effects @xcite . in systems such as vortices and colloidal particles ,
an overdamped description of the equations of motion is appropriate .
in contrast , the skyrmions that were recently discovered in chiral magnets have particle - like properties and many similarities to superconducting vortices , but have the important distinction that there is a strong non - dissipative magnus force in their motion @xcite .
the skyrmions can be set into motion by an applied current and are observed to have a very small depinning threshold @xcite , in part because the effectiveness of the magnus force can be up to ten times stronger than the dissipative force component .
the magnus force introduces a velocity component of the skyrmion that is perpendicular to the direction of an imposed external force , so a skyrmion deflects from or spirals around an attractive pinning site rather than moving directly toward the potential minimum as would occur for systems governed by overdamped dynamics @xcite . since skyrmions are particle - like objects , many of their dynamical properties can be captured using a point particle model based on a modified theile s equation that takes into account repulsive skyrmion - skyrmion interactions , the magnus force , damping , and substrate interactions @xcite . such as approach
has been shown to match well with micromagnetic modeling @xcite of the depinning of skyrmions in periodic @xcite and random pinning arrays @xcite .
particle - based skyrmion models were used to describe the motion of skyrmions interacting with single pinning sites @xcite as well as skyrmion motion in confined regions @xcite . since
skyrmions can easily be driven with an applied external drive they potentially open a new class of experimentally accessible dynamical systems where the magnus force has a dominant effect .
it should be possible to create various types of potential energy landscapes for skyrmions through techniques such as thickness modulations , periodic applied stain , controlled irradiation , or spatially periodic doping .
an open question is how known phase locking phenomena would be affected by the inclusion of a magnus force , and whether new types of phase locking effects might appear that are absent in overdamped systems .
skyrmions also have potential for various spintronic applications @xcite which would require the skyrmions to move in a controlled manner , so an understanding of skyrmion phase locking dynamics could be useful for producing new methods for precision control of skyrmion motion . on a periodic quasi-1d substrate with @xmath0 .
the darker regions are potential maxima and the lighter regions are potential minima , while lines indicate the skyrmion trajectories .
( a ) for an ac drive @xmath1 applied in the longitudinal or @xmath2-direction at @xmath3 and @xmath4 , the skyrmions oscillate in 1d paths at a @xmath5 angle to the @xmath2 axis .
( b ) for an ac drive @xmath6 applied in the transverse or @xmath7-direction with @xmath4 at @xmath8 or the overdamped limit , the skyrmions move in 1d paths along the @xmath7 direction .
( c ) the same as in ( b ) with @xmath3 , where the skyrmions form elliptical 2d counterclockwise orbits .
, width=336 ] in this work we examine shapiro steps for skyrmions moving over a quasi-1d periodic washboard substrate similar to the geometry considered by martinoli _
et al . _ for vortices moving over substrates with a periodic thickness modulation @xcite .
we consider the motion of single skyrmions and collectively interacting skyrmions over a periodic substrate , as illustrated in fig . [ fig:1 ] . here
the longitudinal or @xmath2-direction is aligned with the direction of the periodicity of the substrate , while the @xmath7-direction corresponds to the transverse direction . in this geometry in the overdamped limit ,
shapiro steps occur when the dc ( @xmath9 ) and ac ( @xmath10 ) driving forces are both applied along the @xmath2-direction .
when a finite magnus force is present , we find that phase locking continues to occur ; however , the net skyrmion motion is rotated at an angle with respect to the substrate periodicity direction . as the magnitude of the magus force prefactor
is increased , the width and number of steps for a fixed dc drive gradually decreases , and the steps shift to higher values of @xmath9 . in the overdamped limit ,
if the ac drive is applied in the transverse or @xmath7-direction , no shapiro steps occur , but for a finite magnus force a new set of shapiro steps can arise , with the width and the number of resolvable steps in the velocity - force curve increasing with increasing magnus force prefactor . on a phase - locked step ,
the skyrmion motion takes the form of intricate periodic 2d orbits .
as a function of the ac amplitude , we show that the widths of the phase locking steps for both longitudinal and transverse ac driving exhibit the oscillatory behavior associated with shapiro steps .
we note that the shapiro steps we observe are different than the previously studied phase locking motion for purely dc - driven skyrmions moving over a 2d periodic substrate @xcite . in the latter case ,
the phase locking was associated with a directional locking effect in which the skyrmion motion locks to certain symmetry directions of the substrate potential . in the present study
there is no phase locking without an ac drive .
we consider a 2d system of size @xmath11 with periodic boundary conditions in the @xmath2 and @xmath7 directions containing @xmath12 skyrmions at a density of @xmath13 .
single ( @xmath14 ) or multiple skyrmions interact with a quasi-1d periodic sinusoidal potential with a periodicity direction running along the @xmath2 direction , as illustrated in fig .
[ fig:1 ] .
the equation of motion for a single skyrmion @xmath15 with velocity @xmath16 moving in the @xmath17 plane is @xmath18 here @xmath19 is the location of the skyrmion and @xmath20 is the prefactor of the damping force that aligns the skyrmion velocity in the direction of the net external forces .
the second term is the magnus force with prefactor @xmath21 , which rotates the velocity into the direction _
perpendicular _ to the net external forces . in order to maintain a constant magnitude of the skyrmion velocity
we impose the constraint @xmath22 and vary the relative importance of the magnus force to the damping force by changing the ratio @xmath23 . in the overdamped limit @xmath24 , while for skyrmions @xmath23 can be ten or larger @xcite .
the skyrmion - skyrmion interaction force is @xmath25 where @xmath26 , @xmath27 , and @xmath28 is the modified bessel function .
this interaction is repulsive and falls off exponentially for large @xmath29 . for most of this work
we remain in the limit where skyrmion - skyrmion interactions are weak so that we can consider the dynamics of a single skyrmion ; however , we show that most of our results are robust under the inclusion of skyrmion - skyrmion interactions .
the substrate force @xmath30 arises from a washboard potential @xmath31 where @xmath32 , @xmath33 is the periodicity of the substrate , and we define the substrate strength to be @xmath34 .
unless otherwise noted , we take @xmath0 .
the dc driving term @xmath35 is slowly increased in magnitude to avoid any transient effects .
the ac driving term is either @xmath36 for transverse driving or @xmath37 for perpendicular driving .
we measure the time - averaged skyrmion velocities in the @xmath2 direction @xmath38 and @xmath7-direction @xmath39 . here , due to the periodicity of the substrate , phase locked steps occur when the skyrmions travel integer multiples of the substrate periodicity @xmath40 during each ac drive cycle , allowing us to label the steps @xmath41 for the pinned phase and @xmath42 for the higher order steps .
we focus on the two ac frequencies @xmath43 inverse simulation time steps for the longitudinal ac driving and @xmath44 inverse simulation time steps for the transverse ac driving , and use a substrate lattice constant of @xmath45 .
we use two different driving protocols as illustrated in fig .
[ fig:1 ] . for longitudinal driving ,
we have @xmath46 corresponding to the conditions under which shapiro steps arise for an overdamped system . for transverse driving
, we have @xmath47 which would produce no shapiro steps in the overdamped limit . in fig .
[ fig:1](a ) we show the skyrmion trajectories for @xmath48 , @xmath49 , @xmath3 , and a skyrmion density of @xmath50 . in this case
the skyrmions are pinned and form a triangular lattice that is commensurate with the substrate .
the ac drive causes the skyrmions to oscillate in the potential minima ; however , their motion is not strictly in the @xmath2-direction but is tilted at an angle of @xmath51 with respect to the @xmath2 direction due to the magnus force , which induces a velocity component perpendicular to the ac driving direction . in the absence of a substrate , a dc or ac drive applied in the @xmath2-direction causes the skyrmions to move at an angle @xmath52 with respect to the driving direction , so that in the overdamped limit of @xmath24 the skyrmion moves parallel to the direction of the net external driving force . in fig .
[ fig:1](b ) , we rotate the direction of the ac drive to be in the transverse direction with @xmath6 and @xmath4 for a sample with @xmath8 . in this case
the skyrmion motion follows strictly 1d paths aligned with the @xmath7-direction that pass through the potential minima of the substrate . for @xmath3 , as shown in fig .
[ fig:1](c ) , the skyrmions rotate in counterclockwise elliptical patterns , showing that the magnus force can induce @xmath2-direction motion even when the drive is applied only in the @xmath7-direction . in the absence of the substrate the ac drive would produce only 1d trajectories at an angle with respect to the @xmath7-axis .
this highlights the fact that the magnus force affects how the skyrmions move when interacting with forces induced by the substrate .
( upper blue curves ) and @xmath53 ( lower red curves ) vs @xmath9 for the system in fig .
[ fig:1](a ) in the single skyrmion limit at @xmath48 . ( a ) in the overdamped limit of @xmath54 , @xmath55 and a series of steps appear in @xmath56 indicating phase locking .
( b ) at @xmath57 , @xmath53 is finite .
( c ) @xmath58 and ( d ) @xmath59 show the increase of skyrmion motion in the direction transverse to the substrate and the shift in the locking phases.,width=336 ] we first consider the case illustrated in fig . [ fig:1](a ) of ac driving in the longitudinal direction .
we conduct a series of simulations for increasing @xmath23 and focus on the single skyrmion limit .
in general we find that the shapiro steps we observe remain robust when finite skyrmion - skyrmion interactions are included ; however , additional features can arise for varied fillings when the skyrmion structure is incommensurate with the substrate , as we discuss in section v. in fig .
[ fig:2](a ) we plot @xmath60 and @xmath53 versus @xmath9 for the system in fig .
[ fig:1](a ) at @xmath48 in the overdamped limit of @xmath54 . here
@xmath61 while @xmath60 shows a series of steps indicative of the phase locking .
these features are similar to those observed for other overdamped systems moving over quasi-1d periodic substrates such as vortices in type - ii superconductors moving over quasi-1d substrate modulations . in fig .
[ fig:2](b ) , when @xmath57 , both @xmath53 and @xmath60 are finite and have a ratio of @xmath62 . here
the phase locking is still occurring , but the intervals of @xmath9 in which the phase locking steps appear are shifted .
figure [ fig:2](c ) shows that at @xmath58 , both @xmath63 and some of the step widths have increased in size , and there are no clear regions between the steps where no phase locking is occurring . in fig .
[ fig:2](d ) , at @xmath59 , there is only a single phase locking step .
vs @xmath9 at @xmath64 for @xmath8 ( brown ) , 0.577 ( light blue ) , 0.98 ( dark purple ) , 1.33 ( light purple ) , 2.06 ( dark orange ) , 3.042 ( light orange ) , 4.0 ( dark red ) , 4.92 ( light red ) , 7.0 ( dark green ) , 8.407 ( light green ) , 9.962 ( dark blue ) , and @xmath65 ( black ) , from left to right . here @xmath60 exhibits quantized values corresponding to specific steps .
( b ) the corresponding values of @xmath53 vs @xmath9 , which contains steps that are not quantized . ,
width=336 ] to @xmath66 steps as a function of @xmath9 and @xmath23 .
the width of the steps is reduced and the steps shift to higher values of @xmath9 with increasing @xmath23 .
, width=336 ] to more clearly demonstrate the behavior of the steps for varied @xmath23 , in fig .
[ fig:3](a ) we plot @xmath60 versus @xmath9 for @xmath23 ranging from @xmath67 to @xmath65 , with the evolution of the first three locking steps @xmath68 to 3 highlighted . for a given value of @xmath69 , the step in @xmath56 has a fixed value regardless of the choice of @xmath70 , and each step shifts to higher values of @xmath9 with increasing @xmath23 .
the corresponding @xmath53 versus @xmath9 plot in fig .
[ fig:3](b ) shows that the steps in @xmath71 are not quantized in integer multiples of @xmath72 .
the quantization of the @xmath60 arises from the periodicity of the substrate in the @xmath2-direction , and since the @xmath7-direction has no periodicity , there is no quantization of @xmath71 . in fig .
[ fig:4 ] we highlight the evolution of the widths of the @xmath41 through @xmath73 steps as a function of @xmath9 and @xmath23 at @xmath48 . at @xmath54 , the largest number of phase locking steps can be resolved .
we observe two general trends as @xmath23 increases .
first , for @xmath74 , the widths of the locking regions decrease and the intervals of @xmath9 over which the locking occurs shift to higher values of @xmath9 , with the magnitude of this shift increasing with increasing @xmath69 .
second , the width of the @xmath75 , 2 , and @xmath76 steps initially increases for increasing @xmath23 before reaching a maximum and then decreasing again .
the width of the @xmath41 step reaches a maximum with increasing @xmath70 and then saturates .
the shift in the locations of the phase locking regions arises because the angle at which the skyrmions move with respect to the @xmath2-axis increases with increasing @xmath23 , causing the skyrmions to spend larger intervals of time interacting with the repulsive portion of the substrate potential . as a result ,
higher values of @xmath9 must be applied to cause the skyrmion to translate in the @xmath2-direction at larger @xmath70 .
vs @xmath9 for @xmath59 .
( a ) @xmath77 .
( b ) @xmath78 .
( c ) @xmath79 .
, width=336 ] of the @xmath80 step vs @xmath81 for the system in fig .
[ fig:5 ] at @xmath59 .
the solid line is a fit to the @xmath82 bessel function .
( b ) the width @xmath83 of the @xmath68 step vs @xmath84 for the same system .
the solid line is a fit to the @xmath85 bessel function . in each case
the width of step @xmath69 shows an oscillation of the form of the bessel function @xmath86 , which is characteristic of shapiro step phase locking .
, width=336 ] we next determine if the phase locking steps at high magnus force prefactor are of the shapiro type . in fig .
[ fig:5 ] we plot @xmath60 versus @xmath9 for @xmath59 at @xmath77 , 4.2 , and @xmath87 to show the variation in the widths of the @xmath41 , @xmath68 , and @xmath88 steps . in fig .
[ fig:6 ] we plot the widths @xmath89 and @xmath90 of the @xmath91 and @xmath68 steps , respectively , versus @xmath84 .
each step shows the characteristic oscillation expected for shapiro steps , where the width of step @xmath69 is proportional to @xmath92 , where @xmath93 is the @xmath69th - order bessel function @xcite .
the solid lines in fig .
[ fig:6](a , b ) are fits to @xmath82 and @xmath85 , respectively .
the higher order steps obey similar fits .
this indicates that in the magnus - dominated limit , shapiro step phase locking is occurring .
vs @xmath9 for a system with dc driving in the @xmath2-direction and ac driving @xmath94 in the @xmath7-direction .
( a ) at @xmath8 , there are no steps in @xmath56 .
( b ) at @xmath95 , steps are present .
( c ) @xmath96 .
( d ) @xmath97 .
( e ) @xmath98 .
( f ) @xmath59 .
, width=336 ] we next consider the case illustrated in fig . [ fig:1](b , c ) , where the ac drive is applied transverse to the direction of the substrate potential . in the overdamped limit of @xmath54 ,
such a drive causes the skyrmion to oscillate in the @xmath7-direction as shown in fig .
[ fig:1](b ) , and when a finite dc drive is applied in the longitudinal direction , a single washboard oscillation frequency in the @xmath2-direction is generated by the motion of the skyrmion over the periodic substrate .
since only one frequency is present , there is no coupling between two frequencies , so mode locking does not occur .
when the magnus force is finite , the transverse ac drive induces an oscillating velocity component in the longitudinal or @xmath2-direction as well as in the @xmath7-direction , as illustrated in fig .
[ fig:1](c ) , so that it is possible for the dc - induced washboard frequency to couple to the transverse ac frequency and generate a transverse shapiro step . in fig .
[ fig:7 ] we plot @xmath60 vs @xmath9 for a single skyrmion moving with @xmath94 . at @xmath99 , shown in fig . [ fig:7](a ) , there are no steps in @xmath60 , indicating the lack of phase locking , while the corresponding @xmath100 depinning occurs at the threshold value @xmath101 of @xmath102 .
figure [ fig:7](b ) shows that at @xmath95 , the depinning threshold has dropped substantially to @xmath103 and a series of steps are now visible for @xmath104 , indicating that phase locking is occurring . for @xmath105
, @xmath71 is finite and the @xmath63 versus @xmath9 curve has exactly the same form as @xmath60 versus @xmath9 , but the magnitude of @xmath106 is multiplied by @xmath23 . in fig .
[ fig:7](c , d ) we plot @xmath60 versus @xmath9 for samples with @xmath96 and @xmath107 , respectively . here , the widths of the locking steps increase with increasing @xmath70 and the step locations are shifted to higher values of @xmath9 . in samples with @xmath98 and @xmath108 , as shown in fig .
[ fig:7](e , d ) , respectively , the steps extend out to larger values of @xmath9 , and the non - phase locking regions between the steps are also extended .
the steps in @xmath109 once again occur at quantized values of @xmath110 due to the periodicity in the @xmath2-direction , while the steps in @xmath71 do not have quantized values .
step as a function of @xmath9 and @xmath23 for the system in fig .
[ fig:6 ] with a transverse ac drive of @xmath94 . here
there are several local minima and maxima that are associated with changes in the skyrmion orbits , as shown in fig .
[ fig:9 ] at the points marked a - d .
, width=336 ] in fig .
[ fig:8 ] we plot the location of the upper edge of the @xmath91 step as a function of @xmath9 and @xmath70 for the system shown in fig .
[ fig:6 ] with @xmath0 .
this is equivalent to the threshold depinning force @xmath111 . here @xmath112 at @xmath8 and it decreases to zero at @xmath113 .
there is a local maximum in @xmath114 at @xmath115 , followed by another minimum near @xmath116 and a broad plateau for higher values of @xmath23 .
this oscillatory behavior in the @xmath41 step width is absent for longitudinal ac driving , as shown in fig .
[ fig:5 ] where @xmath101 exhibits only monotonic behavior .
the dips and maxima in @xmath101 for the transverse ac driving are associated with transitions in the shape of the skyrmion orbits during a single ac drive cycle for increasing @xmath23 . along the @xmath41 step at the points labeled
( a - d ) in fig .
[ fig:8 ] .
( a ) at @xmath8 for @xmath117 , there is 1d motion in the @xmath7-direction .
( b ) @xmath118 at @xmath119 .
( c ) at @xmath120 and @xmath117 , the skyrmion moves between two potential minima .
( d ) at @xmath121 and @xmath122 , the skyrmion moves between three potential minima .
, width=336 ] in fig .
[ fig:9 ] we illustrate the skyrmion trajectories in a subsection of the system on the @xmath41 step at the points labeled ( a - d ) in fig .
[ fig:8 ] .
figure [ fig:9](a ) shows that for @xmath54 and @xmath117 , the skyrmion moves in a 1d path in the @xmath7-direction along the potential minimum . at @xmath118 and @xmath119 , in fig .
[ fig:9](b ) , the skyrmion forms an elliptical orbit that is confined within a single potential trough . on the local maximum in the @xmath91 step marked point c in fig .
[ fig:8 ] , at @xmath120 and @xmath117 , the skyrmion forms a more complicated 2d orbit that has three lobes . in a single ac drive cycle
the skyrmion translates back and forth by two substrate lattice constants .
the dip in @xmath101 at @xmath113 shown in fig .
[ fig:8 ] corresponds to the point at which the skyrmion orbit transitions from being confined in one potential minimum to traversing two potential minima . above the second local minimum at @xmath116 in fig .
[ fig:8 ] , the skyrmion orbit becomes even more complex , as illustrated in fig .
[ fig:9](d ) for @xmath121 and @xmath119 .
the skyrmion now moves between three substrate potential minima in a single ac drive cycle .
the local minimum in the @xmath41 step width at @xmath116 then corresponds to the transition in the skyrmion motion from traversing two substrate minima to traversing three substrate minima . for higher values of @xmath23 , additional minima in @xmath101 could occur
that would be correlated with orbits traversing four or more substrate minima .
we expect that additional substrate minima would be resolvable in samples with a smaller substrate lattice constant @xmath33 . , 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , and @xmath123 steps ( from bottom to top ) appear as a function of @xmath9 and @xmath23 for @xmath94 . increasing the magnus force produces enhanced phase locking .
( b ) a blowup of panel ( a ) in the region of small @xmath23 showing that the steps vanish as @xmath23 goes to zero . , width=336 ] . ( a ) @xmath75 orbit at @xmath118 and @xmath124 . ( b ) @xmath125 orbit at @xmath118 and @xmath126 . ( c )
@xmath125 orbit at @xmath115 and @xmath127 .
( d ) @xmath75 orbit at @xmath96 and @xmath128 .
here the skyrmion translates by 2@xmath33 in the positive @xmath2 direction followed by @xmath33 in the negative @xmath2 direction for a net transport by a distance @xmath33 in the @xmath2-direction during each ac cycle .
( e ) @xmath129 orbit at @xmath96 and @xmath130 .
( f ) @xmath75 orbit at @xmath131 and @xmath130 .
, width=336 ] in fig .
[ fig:10](a ) we highlight the regions of phase locking as a function of @xmath9 and @xmath23 for steps @xmath91 through @xmath132 for the system in fig .
[ fig:7 ] .
when @xmath54 , all the steps with @xmath133 vanish , as illustrated in fig .
[ fig:10](b ) where we plot the regime @xmath134 . as the magnus force increases , a larger number of steps can be resolved . in general
, the step widths increase with increasing @xmath23 ; however , certain steps such as @xmath75 , 2 , and @xmath76 show step width oscillations . in the case of longitudinal ac driving , the skyrmion orbits along the different locking steps are always 1d in nature . in contrast , the orbits are much more complicated for transverse ac driving . in fig .
[ fig:11](a ) we show the @xmath75 skyrmion orbit from fig .
[ fig:10 ] at @xmath118 and @xmath124 .
the skyrmion translates in the positive @xmath2-direction and negative @xmath7-direction , making an angle close to @xmath135 with the @xmath2-axis . during a single orbit the skyrmion passes through a loop and
translates by one lattice constant in the @xmath2-direction .
figure [ fig:11](b ) illustrates the @xmath88 orbit for @xmath118 at @xmath126 , where the skyrmion translates two lattice constants in the @xmath2-direction per ac cycle . on the @xmath88 step for @xmath115 , and @xmath127 , shown in fig .
[ fig:11](c ) , the skyrmion moves at a steeper angle of @xmath136 from the @xmath2-axis . in fig .
[ fig:11](d ) , which shows the @xmath75 orbit at @xmath96 and @xmath128 , during a single ac drive cycle the skyrmion initially moves @xmath137 in the positive @xmath2-direction before moving @xmath33 in the negative @xmath2-direction , producing a net translation in the @xmath2-direction of a distance @xmath33 per ac cycle .
figure [ fig:11](e ) shows the @xmath96 system in the @xmath129 orbit at @xmath138 , where the skyrmion translates by @xmath139 in a single ac cycle . on the @xmath68 step at @xmath121 and @xmath130 , plotted in fig .
[ fig:11](f ) , the skyrmion moves @xmath139 in the positive @xmath2-direction during the first portion of the ac drive cycle followed by @xmath137 in the negative @xmath2-direction during the second portion of the ac drive cycle , producing a net translation of @xmath33 in the @xmath2 direction during a single ac cycle .
we observe similar orbits for the other values of @xmath69 , and find that the net angle of the skyrmion motion with respect to the @xmath2 axis increases with increasing @xmath23 .
vs @xmath9 for @xmath94 , @xmath59 , and @xmath140 ( black ) , @xmath141 ( green ) , 4.0 ( blue ) , and @xmath142 ( red ) .
( b ) the evolution of the @xmath91 , 1 , 2 , and 3 step widths as a function of @xmath9 and @xmath143 for the system in panel ( a ) . , width=336 ] we next consider the effect of the substrate strength on the transverse locking steps at @xmath59 and @xmath94 . in fig .
[ fig:12](a ) we plot @xmath144 versus @xmath9 for @xmath140 , 2.0 , 4.0 , and @xmath142 . at the lower values of @xmath143 , the phase locking steps decrease in width , and the steps completely vanish when @xmath145 .
this is highlighted in fig .
[ fig:12](b ) where we plot the widths of the @xmath41 , 1 , 2 , and @xmath76 steps as a function of @xmath9 and @xmath143 .
the width of the locking regions oscillates with increasing @xmath143 , and for @xmath146 all the locking phases shift linearly to higher values of @xmath9 with increasing @xmath143 .
the step width oscillations arise due to variations in the number of potential minima through which the skyrmion orbit passes during a single ac drive cycle , similar to what was observed for fixed @xmath143 and varied @xmath23 .
this result shows that the transverse phase locking is a generic feature that appears in both the strong and weak substrate regimes , and that it is more pronounced for stronger substrates .
of the @xmath80 step vs @xmath147 for the system in fig .
[ fig:12 ] at @xmath148 and @xmath0 .
the solid line is a fit to the @xmath149 bessel function .
( b ) the width @xmath83 of the @xmath68 step vs @xmath147 for the same system .
the solid line is a fit to the @xmath85 bessel function .
, width=336 ] we next examine the dependence of the step widths at a fixed @xmath143 on the ac driving amplitudes , as shown in fig .
[ fig:13 ] where we plot @xmath150 and @xmath83 versus @xmath147 for @xmath64 and @xmath59 .
the solid lines are fits to @xmath151 , indicating that the transverse phase locking steps are also of the shapiro step type , similar to the longitudinal phase locking steps .
vs @xmath9 at @xmath152 . ( a ) ac driving in the @xmath2-direction with @xmath48 for a single skyrmion ( dark blue line ) and a sample containing multiple skyrmions at a density of @xmath153 ( light orange line ) , showing that fractional phase locking steps can arise .
( b ) the same for ac driving in the @xmath7-direction at @xmath154 .
, width=336 ] we next consider assemblies of interacting skyrmions for the system shown in fig .
[ fig:1 ] . in general ,
when the skyrmion density is commensurate with the substrate and the skyrmions can form a triangular lattice , skyrmion - skyrmion interactions cancel and we find the same types of phase locking observed in the single skyrmion systems .
for incommensurate fillings where dislocations are present or when the skyrmion structure becomes distorted or anisotropic in the pinned phase , we find that it is possible for additional fractional phase locking to occur between the integer phase locking steps .
these fractional locking steps occur when a portion of the skyrmions are locked to step @xmath69 and the remainder of the skyrmions are locked to step @xmath155 . in fig .
[ fig:14](a ) we plot @xmath60 versus @xmath9 for a system with ac driving in the @xmath2-direction at @xmath96 and @xmath48 to compare the results for a single skyrmion with a system at a skyrmion density of @xmath153 .
there are no fractional steps in the single skyrmion system ; however , when interacting skyrmions are present we find fractional steps @xmath156 , where @xmath69 and @xmath157 are integers .
figure [ fig:14](b ) shows the same system for ac driving in the @xmath7-direction , where the same types of fractional steps arise .
the fractional steps appear at incommensurate fields when it is possible to have two effective particle species in the system .
one species is commensurate and the other is associated with interstitials , dislocations , or vacancies . in overdamped systems such as superconducting vortices moving over 2d periodic substrates , similar integer steps for individual or non
- interacting vortices appear at commensurate matching fillings while additional fractional locking steps arise at non - matching fields @xcite . at much higher skyrmion densities and for sufficiently strong substrate strengths , the pinned skyrmion structures become highly anisotropic due to the confinement in the 1d pinning rows . in the moving phase just above depinning ,
the effectiveness of the pinning is partially reduced and the repulsive skyrmion - skyrmion interactions favor a more uniform structure .
the competition between skyrmion - skyrmion and skyrmion - substrate interactions produces a series of order - disorder transitions in the moving state . on the phase - locked steps ,
the skyrmions form an ordered moving anisotropic lattice and travel in a synchronized fashion , while between the phase locking steps the skyrmions adopt a more isotropic or liquid like configuration .
vs @xmath9 at @xmath94 and @xmath96 for a sample with a skyrmion density of @xmath158 .
( b ) the fraction of six - fold coordinated particles @xmath159 vs @xmath9 for the same system showing that along the phase - locked steps the skyrmions form a much more ordered state .
, width=336 ] and ( b , d , f ) the corresponding structure factors @xmath160 . ( a , b ) the @xmath75 phase locked step at @xmath161 from the point labeled * a * in fig . [
fig:15](a ) shows a partially ordered anisotropic structure .
( c , d ) on the non - step region at @xmath162 labeled * b * in fig .
[ fig:15](a ) , the skyrmions form a disordered liquid like structure . ( d , e ) on a non - step region at @xmath163 , the system forms a moving lattice . , width=336 ] in fig .
[ fig:15](a ) we plot @xmath60 versus @xmath9 for a sample with @xmath94 , @xmath96 , and a skyrmion density of @xmath158 , showing the @xmath68 , 3 , and 4 phase locking steps .
figure [ fig:15](b ) illustrates the corresponding fraction of six - fold coordinated skyrmions @xmath164 , where @xmath165 is the coordination number of skyrmion @xmath15 obtained from a voronoi construction . on the phase locking steps
, @xmath166 increases to @xmath167 , while between the steps @xmath168 on average and shows strong fluctuations . in fig .
[ fig:16](a , b ) we show the real space locations of the skyrmions and the corresponding structure factor @xmath160 on the @xmath75 step at @xmath161 from fig .
[ fig:15 ] .
the skyrmions are all moving together and form a partially ordered but anisotropic lattice . even though the system is anisotropic , most of the skyrmions have six neighbors , so
that @xmath169 .
figure [ fig:16](c , d ) shows the same sample at @xmath124 , corresponding to the non - phase locking region labeled * b * in fig .
[ fig:15 ] . here
the skyrmions form a disordered structure that is less anisotropic than the phase locked state .
we observe similar sets of dynamical order - disorder transitions between step and non - step regions for increasing @xmath9 and find similar effects for ac driving in the @xmath2-direction .
studies in overdamped systems of collections of interacting vortices also show that the vortices are more ordered and exhibit suppressed noise fluctuations in a phase locked region @xcite . at higher @xmath9 , the effectiveness of the substrate gradually diminishes , the phase locking steps disappear , and the skyrmions can reorder into a more uniform moving crystal state as shown in fig .
[ fig:16](e , f ) at @xmath163 .
similar dynamical reordering to a triangular lattice for high drives has been observed for skyrmions interacting with random pinning @xcite as well as for vortices driven over random pinning arrays @xcite .
these results show that shapiro steps for skyrmions interacting with a periodic substrate are a robust feature that occurs for a variety of skyrmion densities and substrate strengths .
the change in the skyrmion lattice structure as the system passes in and out of phase locked states as a driving current is swept could be observed using neutron scattering or noise measurements .
we have analyzed shapiro steps for skyrmions interacting with periodic quasi - one - dimensional substrates in the presence of combined dc and ac drives , with a specific focus on the role of the magnus force in the dynamics . when the dc and ac drives are both applied in the longitudinal direction , which is aligned with the substrate periodicity ,
phase locking occurs , and as the role of the magnus force increases , the phase locking steps gradually reduce in width and shift to higher values of the driving force .
the skyrmions move at an angle to the direction of the external dc drive that increases as the contribution of the magnus force increases .
when the ac drive is applied perpendicular to the dc drive and the substrate periodicity direction , there is no phase locking in the overdamped limit ; however , when there is a finite magnus force , phase locking can occur . on the phase locked steps the skyrmions move in intricate two - dimensional periodic orbits .
we map out the evolution of the phase locked regions for the transverse and longitudinal ac driving for varied contribution of the magnus force , ac driving amplitudes , and substrate strength .
when collective interactions between skyrmions are introduced , fractional shapiro steps can appear .
for strong substrate strengths and higher skyrmion densities , both longitudinal and transverse phase locking steps occur that are associated with dynamically induced transitions between an ordered anisotropic solid on the steps to a fluctuating liquid state in the non - phase locked regimes .
such transitions could be observed with neutron scattering .
s. p. benz , m. s. rzchowski , m. tinkham , and c. j. lobb , fractional giant shapiro steps and spatially correlated phase motion in 2d josephson arrays , phys .
lett . * 64 * , 693 ( 1990 ) ; d. dom ' inguez and j. v. jos ' e , giant shapiro steps with screening currents , _ phys .
lett . _ * 69 * , 514 ( 1992 ) .
r. e. thorne , j. s. hubacek , w. g. lyons , j. w. lyding , and j. r. tucker , ac - dc interference , complete mode locking , and origin of coherent oscillations in sliding charge - density - wave systems , _ phys .
b _ * 37 * , 10055 ( 1988 ) . b. hu and j. teki ' c , amplitude and frequency dependence of the shapiro steps in the dc- and ac - driven overdamped frenkel - kontorova model , _ phys
e _ * 75 * , 056608 ( 2007 ) ; j. teki ' c and b. hu , noise - induced bessel - like oscillations of shapiro steps with the period of the ac force , _ phys .
b _ * 78 * , 104305 ( 2008 ) .
j. harris , n. ong , r. gagnon , and l. taillefer , washboard frequency of the moving vortex lattice in yba@xmath170cu@xmath171o@xmath172 detected by ac - dc interference , _ phys .
lett . _ * 74 * , 3684 ( 1995 ) .
marconi , a.b .
kolton , d. dom ' inguez , and n. grnbech - jensen , transverse phase locking in fully frustrated josephson junction arrays : a different type of fractional giant steps , _ phys .
b _ * 68 * , 104521 ( 2003 ) . t. schulz , r. ritz , a. bauer , m. halder , m. wagner , c. franz , c. pfleiderer , k. everschor , m. garst , and a. rosch , emergent electrodynamics of skyrmions in a chiral magnet , _ nature phys . _ * 8 * , 301 ( 2012 ) . x.z .
yu , n. kanazawa , w.z .
zhang , t. nagai , t. hara , k. kimoto , y. matsui , y. onose , and y. tokura , skyrmion flow near room temperature in an ultralow current density , _ nature commun . _ * 3 * , 988 ( 2012 ) .
s. woo , k. litzius , b. kr " uger , m .- y .
i m , l. caretta , k. richter , m. mann , a. krone , r. reeve , m. weigand , p. agrawal , p. fischer , m. kl " aui , g.s.d .
beach , observation of room temperature magnetic skyrmions and their current - driven dynamics in ultrathin co films , arxiv:1502.07376 ( unpublished ) .
c. moreau - luchaire , c. moutafis , n. reyren , j. sampaio , n. van horne , c.a.f .
vaz , k. bouzehouane , k. garcia , c. deranlot , p. warnicke , p. wohlh " uter , j.m .
george , j. raabe , v. cros , and a. fert , skyrmions at room temperature : from magnetic thin films to magnetic multilayers , arxiv:1502.07853 ( unpublished ) .
w. jiang , p. upadhyaya , w. zhang , g. yu , m.b .
jungfleisch , f.y .
fradin , j.e .
pearson , y. tserkovnyak , k.l .
wang , o. heinonen , s.g.e . te velthuis , and a. hoffmann , blowing magnetic skyrmion bubbles , _ science _ , in press ( 2015 ) .
y. tokunaga , x.z .
yu , j. s. white , h. m. rnnow , d. morikawa , y. taguchi , and y. tokura , a new class of chiral materials hosting magnetic skyrmions beyond room temperature , _ nature commun . _ * 6 * , 7638 ( 2015 ) . f. b " uttner , c. moutafis , m. schneider , b. kr " uger , c.m .
g " unther , j. geilhufe , c. v. korff schmising , j. mohanty , b. pfau , s. schaffert , a. bisig , m. foerster , t. schulz , c.a.f .
vaz , j.h .
franken , h.j.m .
swagten , m. kl " aui , and s. eisebitt , dynamics and inertia of skyrmionic spin structures , nature phys . * 11 * , 225 ( 2015 ) . | we numerically study the behavior of two - dimensional skyrmions in the presence of a quasi - one - dimensional sinusoidal substrate under the influence of externally applied dc and ac drives . in the overdamped limit ,
when both dc and ac drives are aligned in the longitudinal direction parallel to the direction of the substrate modulation , the velocity - force curves exhibit classic shapiro step features when the frequency of the ac drive matches the washboard frequency that is dynamically generated by the motion of the skyrmions over the substrate , similar to previous observations in superconducting vortex systems . in the case of skyrmions , the additional contribution to the skyrmion motion from a non - dissipative magnus force shifts the location of the locking steps to higher dc drives , and we find that the skyrmions move at an angle with respect to the direction of the dc drive . for a longitudinal dc drive and a perpendicular or transverse ac drive , the overdamped
system exhibits no shapiro steps ; however , when a finite magnus force is present we find pronounced transverse shapiro steps along with complex two - dimensional periodic orbits of the skyrmions in the phase - locked regimes .
both the longitudinal and transverse ac drives produce locking steps whose widths oscillate with increasing ac drive amplitude .
we examine the role of collective skyrmion interactions and find that additional fractional locking steps occur for both longitudinal and transverse ac drives . at higher skyrmion densities ,
the system undergoes a series of dynamical order - disorder transitions , with the skyrmions forming a moving solid on the phase locking steps and a fluctuating dynamical liquid in regimes between the steps . |
exponentially harmonic maps were introduced by james eells and studied by j. eells and l. lemaire @xcite .
these maps generalize usual harmonic maps @xcite in the following sense .
let @xmath3 and @xmath4 be two riemannian manifolds and @xmath5 , a smooth map .
an exponentially harmonic map is then an extremal of the following functional : @xmath6 where @xmath7 is the riemannian volume element and @xmath8 is the so called `` energy density '' of the map @xmath9 . in local coordinates @xmath10 and @xmath11 , the equation for exponentially harmonic maps , which derives from the variation of the functional ( [ functional ] ) , is @xmath12 in the sequel we will study somme particular cases of ( [ variation of the functional ] ) even when @xmath15 is noncompact and ( [ functional ] ) unbounded . in this case ,
equation ( [ variation of the functional ] ) is taken as the definition of what we call an exponentially harmonic map .
following eells and lemaire , we will consider the energy - momentum tensor asociated to @xmath9 : @xmath24 which is conserved ( @xmath25 ) when @xmath9 is a solution of ( [ variation of the functional ] ) .
it is worth noting that some particular cases of ( [ variation of the functional ] ) were studied by mathematicians in the context of the theory of elliptic partial differential equations .
gilbarg and trudinger @xcite quote , for example , the equation : @xmath26 deriving from the variational principle : @xmath27 lut us remark finally that , if we want to make some contact with physics , we have to modify ( [ functional ] ) as follows : @xmath28d\mu(\phi)\end{aligned}\ ] ] which , if we interpret @xmath29 as a spin vector , looks like the hamiltonian for the continuum limit of a heisenberg model .
when @xmath30 is small enough , the variational principle @xmath31 leads to equations which approximate those of the usual harmonic maps ( sigma models ) .
the equations derived from @xmath32 can be obtained directly from ( [ variation of the functional ] ) via the shift @xmath33 .
in order to get some feeling about the solutions of ( [ variation of the functional ] ) , we study here scalar exponentially harmonic maps on two - dimensional euclidean and minkowskian manifolds : @xmath34 let @xmath36 , @xmath37 and @xmath38 .
then , equation ( [ variation of the functional ] ) becomes @xmath39 we are searching solutions of the following form : @xmath40 .
we are thus led to the equations : @xmath41f_{xx}=-[1+(g_y)^2]g_{yy}=l\end{aligned}\ ] ] where @xmath42 happens to be an arbitrary positive real constant .
@xmath43 leads , of course , to trivial solutions @xmath44 .
if @xmath45 , let @xmath46 . this leads to the system : @xmath47 then ( [ system1eq2 ] ) gives @xmath48 let us derive ( [ solution1 du syteme1 ] ) with respect to @xmath49 .
this gives ( with @xmath50=constant ) : @xmath51 the elimination of @xmath52 from ( [ solution1 du syteme1 ] ) and ( [ solution2 du syteme1 ] ) gives : @xmath53 ^ 4\right.\nonumber\\ & & \left.+2[s_+(x;l;c)+s_-(x;l;c)]^2\right\rbrace -k\\ p & = & s_+(x;l;c)+s_-(x;l;c).\end{aligned}\ ] ] where @xmath54 .
similary we get the solution @xmath55 .
a solution of ( [ variation of the functional when e = r2 ] ) can thus be written , with the above notations as , @xmath56 ^ 4 \right.\\ & & \left.+2[s_+(x ;- l;c)+s_-(x ;- l;c)]^2\right\rbrace -k\end{aligned}\ ] ] which leads to the solution we look for
. if we write @xmath57 , then we can represent @xmath58 in the following parametric form : @xmath59 it is also possible to solve ( [ variation of the functional when e = r2 ] ) by the so - called hodograph method @xcite .
let @xmath60 and @xmath61 .
if @xmath62 , we get immediatly : @xmath63 we know that if @xmath64 then @xmath65 .
thus , there exists a function @xmath66 such that @xmath67 and @xmath68 .
equation ( [ variation of the functional when e = r2 ] ) can be written @xmath69 or , using polar coordinates , @xmath70 : @xmath71 which can be solved by factorizing : @xmath72 .
we have now to solve the two equations ( with an arbitrary constant @xmath73 ) : @xmath74 if we are interested in periodic solutions , we set @xmath75 thus ( [ 1.14 ] ) can be reduced to the normal form by the following functional change : @xmath76 equation ( [ 1.14 ] ) then becomes : @xmath77 this equation can be reduced to the whittaker equation @xcite . writting + @xmath78 ,
then the function @xmath79 sastifies the whittaker equation : @xmath80 with @xmath81 .
this solution can be expressed in terms of the confluent hypergeometric function @xmath82 which leads to the solution of ( [ 1.14 ] ) written as @xmath83 if @xmath84 , this solution is regular at the origin .
we know that @xmath85 .
then , for @xmath86 , we get @xmath87 , which is a solution of ( [ 1.14 ] ) .
nevertheless this trivial solution ( and related to the case @xmath88 ) leads to some problems because , if we set @xmath89 , then @xmath90 , @xmath91 .
the computations are similar to the preceding case .
the equation ( [ variation of the functional ] ) leads to @xmath93 the solutions of the form @xmath94 can be written with a parametric representation : @xmath95
let m be a friedmann - lema^ itre ( fl ) universe endowed with the following metric : @xmath96 we consider exponentially harmonic maps : @xmath97 which are , for physical reasons explained above , extremals of the functional @xmath32 .
they satisfy a modified version of equation ( [ variation of the functional ] ) which is written in this case : @xmath98 if we restrict ourselves to @xmath99 .
this gives : @xmath100 let us take , for example , @xmath101 i.e. , @xmath15 is euclidean fl - universe ( without cosmological constant ) and @xmath102 where @xmath103 is the present value of the hubble constant ( @xmath104 ) .
we get the field equation @xmath105 i ) when @xmath106 : this gives @xmath107 constant .
+ ii ) when @xmath108 and @xmath30 small : @xmath109 which is not regular at @xmath110 .
see figure1 when @xmath111 we recover the well - know theory of harmonic maps coupled with gravity studied by t. hirschmann , r. schimming @xcite , t. hughes , t. kato , j. marden @xcite and lemaire and a. j. vanderwinden @xcite .
let us start now with the following action , describing a gravitational field coupled with an exponentially scalar field : @xmath112 @xmath113 is a constant , @xmath114 is the cosmological constant , @xmath115 is the lagrangian density for matter and @xmath116 is a scalar field defined on a @xmath117-dimensional spacetime .
the variation of @xmath118 leads to einstein s equations : @xmath119 where @xmath120 is the usual energy - momentum tensor for matter .
the variation with respect to @xmath9 gives a field equation which is very similar to ( [ variation of the functional ] ) .
let us assume that @xmath99 and let @xmath15 be a friedmann - lema^ itre universe .
then the field equations can be written : @xmath121 let us consider some particular cases .
@xmath122 let us define @xmath123 and let @xmath124 be the hubble constant .
we check that + @xmath125 , then the equations above define a dynamical system : @xmath126 submitted to the constraint : @xmath127 let us derive ( [ 4.7 ] ) and , using ( [ 4.6 ] ) , let us compare the result with ( [ 4.5 ] ) .
we see that the preceding equations are compatible .
we are thus led to solve the following equation : @xmath128 for very small values of @xmath30 , this becomes : @xmath129 this gives solutions : @xmath130 where @xmath131 are constants which the same sign , @xmath132 a positive constant and @xmath133 is an arbitrary constant .
@xmath134 ( @xmath135 is a first integral when @xmath136 ) .
the equation ( [ 4.7 ] ) allows to write @xmath137 .
see figure 2 it is interesting to note here that the coupling of @xmath9 with the gravitational field can make @xmath9 regular at @xmath110 in this euclidean case , which was not the case in the uncoupled situation ( see section3 ) .
let us return to the case where @xmath30 is an arbitrary constant .
we define @xmath138 , equation ( [ 4.8 ] ) leads to @xmath139 which is a dynamical system with a fixed point @xmath140 .
if we put @xmath141 we get : @xmath142 the hubble constant can be deduced from the numerical integration of this dynamical system . from ( [ 4.7 ] ) we get : @xmath143 ^{1/2}.\end{aligned}\ ] ] around @xmath144 ( i.e. @xmath145 constant ) we can neglect terms of order greater or equal to @xmath146 . using ( [ 4.10 ] ) - ( [ 4.11 ] ) we have at first order : @xmath147 as @xmath141 , we have to suppose that ( for @xmath148 ) : @xmath149 when @xmath150 .
let us derive ( [ 4.4 ] ) . using ( [ 4.2 ] ) and
( [ 4.3 ] ) with @xmath151,we find the following equation : @xmath152 equation ( [ 4.12 ] ) can be solved numerically using the variable change : @xmath153 .
then ( [ 4.12 ] ) becomes : @xmath154 making the substitution @xmath155 , ( [ 4.13 ] ) can be written as .
@xmath156 \frac{1}{(\lambda+1)}\end{aligned}\ ] ] wich can be studied numerically . using ( [ 4.4 ] ) , we can determine @xmath157 @xmath158 if @xmath111 , equation ( [ 4.12 ] ) leads to ( [ 4.9 ] ) with @xmath159 if we assume that @xmath160 .
if @xmath161 , the case @xmath111 can be studied from : @xmath162 with the above notation this gives @xmath163 the case @xmath164 is obvious .
when @xmath165 , we get : @xmath166\\ \ ] ] or @xmath167\right\rbrace^{-3}+c\end{aligned}\ ] ] where @xmath168 is a constant .
we have thus : @xmath169 } { 2\cos^2\left[\frac{1}{3 } \left(\frac{3k}{2}(\lambda+1)\right)^{1/2}(t - t_o)\right ] } \right\rbrace } \nonumber\\ & \times & \left [ \tan \left ( \frac{\pi}{4}+\frac{1}{6 } \left(\frac{3k}{2}(t - t_o)\right)^{1/2}\right ) \right ] ^{\frac{3}{2}\left(\frac{3k}{2}(\lambda+1)\right)^{-1/2}}\exp{(ct)}.\end{aligned}\ ] ] when @xmath170 we have : @xmath171 } { 2\cosh^2\left[\frac{-1}{3 } \left(\frac{3k}{2}(\lambda+1)\right)^{1/2}(t - t_o)\right ] } \right.\nonumber\\ & + & \left .
\frac{1}{2 } \tan^{-1 } \left [ \sinh\frac{-1}{3 } \left(\frac{3k}{2}(-1-\lambda)\right)^{1/2}(t - t_o ) \right ] \right\rbrace \exp{(ct)}.\end{aligned}\ ] ] let us multiply ( [ 4.2 ] ) by @xmath173 and let us add ( [ 4.2 ] ) and ( [ 4.3 ] ) .
we get : @xmath174 using ( [ 4.4 ] ) , we find : @xmath175 then , by ( [ 4.2 ] ) and using the preceding notation , we can write the the equation : @xmath176 which is nothing but ( [ 4.12 ] ) when @xmath177 and @xmath178 ; @xmath178 implies @xmath179 ( see ( [ 4.14 ] ) ) . using the conservation law @xmath180 and equation([4.4 ] ) , we find : @xmath181 an example of numerical solutions is given in fig.3 , 4 and 5 . in fig.3 , we can see the typical behaviour of a friedmann - lema^ itre universe whose expansion is driven by exponentially harmonic maps and matter or radiation .
this evolution know three different ages : first , the universe behaves in the usual way under the presence of matter or radiation , then its characteristic deceleration stops and finally , when the exponentially harmonic map @xmath9 becomes too small , the expansion accelerates ( see eq.([4.14 ] ) ) .
but this is also the typical bahaviour of a universe that is dominated eventually by a positive cosmological constant or something that acts like it , such as a potential related to a scalar field ( quintessence ) .
nevertheless , the careful reader has already noticed that this particular behaviour is only due to the exponentially harmonic feature of the scalar field @xmath9 and its particular initial conditions @xmath182 and @xmath183 .
this proves again that these interesting features , suggested among others by the observations of type ia supernovae , arise naturally in the context of exponentially harmonic maps without requiring an adhoc mechanism of inflation . in fig.4 , we have represented the evolution of the energy densities of matter , radiation and the scalar field coupled to it . in the very early ages of the universe , the dominant contribution to the total energy is due to @xmath9 .
but the universe expands just as if there was only usual matter and radiation .
as this expansion goes on , the energy density of the scalar field @xmath184 goes below the density of matter and radiation , making the expansion accelerating when the field @xmath9 become tiny ( see ( [ 4.14 ] ) ) .
this would correspond to the well known _
`` cosmic coincidence''_. note that we have chosen the arbitrary constants arising in the integration of eq.([divt0 ] ) in order to get this peculiar model of expansion .
the present study is purely qualitative , more work is under way to construct a quantitative model that could be compared , for example , with those of quintessence . in fig.5 , we have represented the evolution of the scalar field @xmath9 in order to satisfy eq.([4.17 ] ) both in the presence of matter ( @xmath177 ) and radiation ( @xmath185 ) .
the main interest of the use of exponentially harmonic maps is the fact that the lagrangian ( [ 4.1 ] ) is a generalisation of bosonic string lagrangian ( with only a dilatonic field ) written in einstein frame @xcite . indeed , if @xmath186 , the lagrangian ( [ 4.1 ] ) tends to this bosonic string lagrangian , if @xmath187 , where @xmath188 is the small cosmological constant .
but mathematically , this limit @xmath189 is highly problematic due to the fact that unlike usual harmonic maps , exponentially harmonic maps have no good invariance property under homothetic changes of the field , as it was shown in the introduction .
20 eells j and lemaire l some properties of exponentially harmonic maps , _ proc .
banach center semester on p.d.e .
1990 - 1991 _ eells j and lemaire l a report on harmonic maps 1978 _ bull _ _ london math .
* 10 * 1 - 68 ; 1988 another report on harmonc maps _ bull _ _ london math . soc . _
* 20 * 385 - 524 ; 1983 _ selected topics in harmonic maps _
( regional conference series in mathematics @xmath190 50 , american mathematical society ) ; hlein f 1996 _ applications harmoniques , lois de conservation et reperes mobiles _ ( paris : diderot editeur ) .
guilbarg d and trudinger n s 1977 _ elliptic partial differential equations of _ _ second order _
( berlin : springer ) courant r and hilbert d 1962 _ methods of mathematical physics ii , partial _ _ differential equations _
( new york : interscience publishers ) bucholz h 1969 _ the confluent hypergeometric function _
( berlin : springer ) ; abramovitz m and stegun i a 1965 _ handbook on mathematical functions _
( new york : dover ) hirschman t and schimming r 1988 harmonic maps from spacetime and their coupling to gravitation _ astro .
nachrichten _ * 309 * 5 hughes t , kato t and marsden j 1977 well posed quasi - linear second - order hyperbolic systems , with applications to nonlinear electrodynamics an general relativity _ arch .
* 63 * 273 - 294 vanderwinden a j 1992 _ exemples dapplications harmoniques _ ( unpublished phd thesis , ulb ) 55 - 74 see for example : lidsey j e , wands d and copeland e j superstring cosmology _ hep - th 9909061 _ | the aim of this paper is to study some examples of exponentially harmonic maps .
we study such maps firstly on flat euclidean and minkowski spaces and secondly on friedmann - lema^ itre universes .
we also consider some new models of exponentially harmonic maps which are coupled with gravity which happen to be based on a generalization of the lagrangian for bosonic strings coupled with dilatonic field .
@xmath0universit dabomey - calavi , unit de recherche en physique thorique , bp 2628 , porto - novo , bnin @xmath1universit catholique de louvain - la - neuve , unit fyma , chemin du cyclotron , 2 , b-1348 louvain - la - neuve , belgium @xmath2facults universitaires n .- d .
de la paix , unit de systemes dynamiques , rue de bruxelles , 61 , b-5000 namur , belgium + e - mail : [email protected] , [email protected] and [email protected] |
( tmtsf)@xmath0@xmath5 are quasi one dimensional molecular conductors , known as bechgaard salts , where tmtsf denotes tetramethyltetraselenafulvalene and @xmath5 is an inorganic anion with various possible symmetries : spherical ( octahedral ) ( @xmath5@xmath6pf@xmath7 , asf@xmath7 ) , tetrahedral ( @xmath5@xmath6clo@xmath8 , reo@xmath8 ) , or triangular ( no@xmath1 ) .
the well known are their very complex ( pressure , magnetic field , temperature ) phase diagrams with a variety of electronic ground states : ( conventional ) spin density wave ( sdw ) , field induced sdw ( fisdw ) , superconductivity ( triplet , unconventional ) , unconventional spin density wave ( usdw).@xcite the observation of superconductivity in the ( tmtsf)@xmath0@xmath5 series requires the use of high pressure , with the exception of ( tmtsf)@xmath0clo@xmath8 , which is superconducting under ambient pressure , and ( tmtsf)@xmath0no@xmath1 , which never becomes a superconductor even under pressure .
no@xmath1 anions are in an orientational disorder at ambient pressure and for @xmath9k .
the anion ordering ( ao ) transition takes place at @xmath10k , with wave vector @xmath11 .
contrary to ao in most other salts , @xmath12 has the nonzero component parallel to the most conducting direction .
the sdw state develops below @xmath2k . from the resistivity data very small activation energy was obtained , of order of @xmath13ev , but the curvature of the @xmath14 vs. @xmath15 plot indicated that the ground state should be considered as semimetalic , rather than semiconducting.@xcite the phase diagram of bechgaard salts under pressure is interpreted in terms of the standard model , where the approximate nesting of the quasi - one dimensional fermi surface ( i.e. the imperfect nesting ) , and the repulsive coulomb interaction between electrons are the crucial ingredients.@xcite the applied pressure increases the 2-dimensionality of bechgaard salts through the increase of the imperfect nesting term.@xcite however , the standard model does not yet describe neither the triplet@xcite superconductivity nor usdw .
udw is a density wave , whose gap function depends on the wavevector , vanishes on certain subsets of the fermi surface , allowing for low energy excitations .
the average of the gap function over the fermi surface is zero , causing the lack of periodic modulation of the charge and/or spin density . as noted by nersesyan et al .
( refs . and ) , the quasiparticle spectrum in udw is quantized in a magnetic field .
this landau quantization gives rise to the spectacular angle dependent magnetoresistance ( admr ) and giant nernst effect.@xcite as we shall see later both the angle dependent magnetoresistance@xcite and the nonlinear hall resistance@xcite of ( tmtsf)@xmath0no@xmath1 are described nicely in terms of usdw .
we note that an earlier attempt to describe the magnetoresistance of ( tmtsf)@xmath0no@xmath1 in terms of conventional sdw with a _ large imperfect nesting _ might not be the most appropriate model,@xcite since it can not describe the details of the resistance quantitatively .
we also propose the revision of the generally accepted phase diagram , taking into account the identification of usdw state in several bechgaard compounds.@xcite
here we summarize briefly what is known about unconventional density wave.@xcite the unconventional density wave is a kind of density wave , where the quasiparticle energy gap vanishes along lines on the fermi surface . in the present instance
we can assume @xmath16 or @xmath17 as in earlier analysis of ucdw in @xmath18(et)@xmath0khg(scn)@xmath8.@xcite then the quasiparticle green function is given by @xmath19 where @xmath20 s are the pauli matrices and @xmath21 operates on the nambu spinor space.@xcite the quasiparticle energy in udw is formed from the pole of @xmath22 as @xmath23 where @xmath24 is the kinetic energy of electrons measured from the fermi energy in the normal state , @xmath25 , @xmath26 is the imperfect nesting term and @xmath27 . here , @xmath28 denotes the fermi velocity in the chain ( @xmath29 ) direction , @xmath30 is the order parameter for unconventional sdw , and @xmath31 is the lattice constant . then in a magnetic field @xmath32 in the @xmath33 plane with angle @xmath34 from the @xmath35 axis the quasiparticle energy changes into @xmath36 with @xmath37 .
here we have neglected @xmath26 for simplicity .
also in the following we assume @xmath38 and @xmath39m / s.@xcite eq .
( [ eq:3 ] ) is the consequence of the landau quantization of the quasiparticle spectrum in udw , or the _
nersesyan effect_.@xcite then the conductivity tensor is constructed as @xmath40 @xmath41 @xmath42 with @xmath43\ ] ] where @xmath44 and we have assumed that @xmath45 .
also , we have assumed that @xmath46 , @xmath47 , @xmath48 , @xmath49 and @xmath50 are weakly dependent on @xmath51 and @xmath52 .
then from eqs.([eq : sxx])([eq : sxy ] ) we can construct the resistivity tensor as @xmath53 @xmath54 in fig .
[ fig : admr ] we show our fitting of the angle dependent magnetoresistance data for ( tmtsf)@xmath0no@xmath1 at @xmath55k for a variety of magnetic field.@xcite ( color online ) the angular dependence of the normalized resistance @xmath56 at @xmath57k ( full lines : experimental data ; dashed lines : fits to the theory ) .
magnetic field is rotated in @xmath33 plane , and @xmath58 corresponds to @xmath59 .
data are from ref .
( @xmath60 t ) and ref .
( @xmath61t).,width=325 ] from this fitting we obtain usdw order parameter @xmath62k and @xmath63 .
as is readily seen the fitting is excellent except for the bumpy structures .
these should come from the imperfect nesting term as discussed in refs . , and .
also we note @xmath64 indicating that @xmath65 ; therefore @xmath66 is dominated by the @xmath67 excitations . in fig .
[ fig : rxy ] we show @xmath68 fitted by eq.([eq : rxy ] ) ; again we obtain reasonable fitting with @xmath69 t .
( color online ) the magnetic field @xmath52 dependence of hall resistance @xmath68 at several temperatures ( points : experimental data ; dashed lines : fit to the theory ) .
data are from ref .. ,width=325 ] figure [ fig : fitparams ] shows temperature dependence of the parameter @xmath70 , along with the temperature dependence of the fitting parameters @xmath70 ( @xmath71 ) see text .
dashed line shows temperature dependence of the resistance @xmath72 of ( tmtsf)@xmath0no@xmath1 for @xmath73.,width=325 ] temperature dependence of the resistance @xmath72 .
there appears to be a slight change of the parameter @xmath70 across @xmath74k : for @xmath75k it follows temperature dependence of resistance @xmath72 , while for lower temperature it is nearly constant .
it signals the possible occurrence of yet another phase transition at @xmath76k as in ( tmtsf)@xmath0pf@xmath7 , in agreement with several other suggestions.@xcite
recently , one of us proposed phase diagram for bechgaard salts with octahedral ( centrosymmetric ) anion like pf@xmath7 which exhibit metallic behaviour down to the sdw transition at @xmath77k ( see ref . ) .
the salts with non - centrosymmetric anions undergo the ao transition and become insulating at ambient pressure , except for @xmath5@xmath6clo@xmath8 and no@xmath1 . here
, we propose an extension / revision of the phase diagram ( see fig . [
fig : phasediag ] ) . as indicated in fig .
[ fig : phasediag ] , the schematic pressure temperature phase diagram for bechgaard salts .
arrows denote position of ( tmtsf)@xmath0pf@xmath7 , ( tmtsf)@xmath0clo@xmath8 and ( tmtsf)@xmath0no@xmath1 in the phase diagram at ambient pressure.,width=302 ] ( tmtsf)@xmath0pf@xmath7 at ambient pressure undergoes yet another transition around @xmath78k . the drastic change in the quasi - particle spectrum through @xmath79 has been interpreted as appearance of sdw+usdw.@xcite further , from the angle dependent magnetoresistance of ( tmtsf)@xmath0pf@xmath7 and ( tmtsf)@xmath0reo@xmath8 for @xmath80kbar the existence of usdw in the high pressure range is inferred.@xcite then , it is customarily to put ( tmtsf)@xmath0clo@xmath8 at ambient pressure around @xmath81kbar in fig .
[ fig : phasediag ] , where transition from metallic state to superconducting one takes place . in
this way we may understand the superconductivity at ambient pressure . similarly , we may put ( tmtsf)@xmath0no@xmath1 at ambient pressure around @xmath82kbar , since the transition from metallic state to density wave state takes place at @xmath2k .
the further behaviour of @xmath83 vs. pressure is based on the experiments , which have shown that @xmath83 is gradually suppressed under increasing pressure.@xcite then is the absence of superconductivity , and appearance of fisdw only at high pressure and high magnetic fields ( @xmath84kbar , @xmath85t),@xcite very surprising .
we think that the lack of inversion symmetry in no@xmath1 is at the heart of the absence of superconductivity and fisdw ( for low pressure , @xmath86kbar ) in ( tmtsf)@xmath0no@xmath1 .
for example p.w.anderson@xcite speculated that the triplet superconductor can not exist in a crystal without inversion symmetry .
also the nature of superconductivity in cept@xmath1si , the crystal without inversion symmetry , is hotly discussed in the current literature.@xcite the inversion symmetry breaking is usually characterized by @xmath87 the chiral symmetry breaking term or the rashba term,@xcite both the absence of the triplet superconductivity ( for @xmath88k ) and the appearance of fisdw for @xmath4 t suggest @xmath89k . also this @xmath87 appears to be consistent with @xmath10k ( @xmath90).@xcite we believe that further study of the electronic properties of ( tmtsf)@xmath0no@xmath1 is of great interest .
we have shown that the anomalous low temperature behaviour of admr and hall resistance of the bechgaard salt ( tmtsf)@xmath0no@xmath1 could be interpreted in terms of unconventional spin density wave , indicating that the possible ground state below @xmath91k is usdw .
this is consistent with the new phase diagram of bechgaard salts proposed recently,@xcite which we in adition revised and extended in this paper . therefore , it will be of great interest to study how the usdw order parameter @xmath30 changes as the pressure is applied .
this will provide a first step to explore the wider phase diagram .
another question is if there are other candidates in bechgaard salts which exhibit usdw under ambient pressure .
we have also proposed possible explanation about the absence of superconductivity .
both the absence of superconductivity and partial suppression of fisdw ( i.e. the absence of fisdw for @xmath92kbar , @xmath93 t ) are due to the inversion symmetry breaking associated with the no@xmath1 anion ordering . the details on this will be published elsewhere.@xcite we thank balazs dra , peter thalmeier and hyekyung won for useful suggestions .
thanks the hospitality of max - planck institut fr physik komplexer systeme at dresden , where the most of this work is done . | among many bechgaard salts , ( tmtsf)@xmath0no@xmath1 exhibits very anomalous low temperature properties . unlike conventional spin density wave ( sdw ) , ( tmtsf)@xmath0no@xmath1 undergoes the sdw transition at @xmath2k and the low temperature quasiparticle excitations are gapless .
also , it is known that ( tmtsf)@xmath0no@xmath1 does not exhibit superconductivity even under pressure , while fisdw is found in ( tmtsf)@xmath0no@xmath1 only for @xmath3kbar and @xmath4 t . here
we shall show that both the angle dependent magnetoresistance data and the nonlinear hall resistance of ( tmtsf)@xmath0no@xmath1 at ambient pressure are interpreted satisfactory in terms of unconventional spin density wave ( usdw ) .
based on these facts , we propose a new phase diagram for bechgaards salts . |
the nucleus @xmath0th is located in a region where strong octupole correlations are important in the properties already of the low - lying excitations .
besides the interplay of collective and single - particle excitations , which takes place in deformed rare earth nuclei , the reflection asymmetry additionally complicates the picture of excitations . already in an earlier publication @xcite
, a conclusion was made that the nature of the first excited 0@xmath4 states in the actinide nuclei is different from that in the rare earth region , where they are due to the quadrupole vibration .
the strong excitations in the ( p , t)-reaction suggest that these states represent a collective excitation different from the @xmath5-vibration .
decay modes of the levels of the band on the first excited 0@xmath4 state in @xmath0th have led to the suggestion that this band might predominantly have an octupole two - phonon structure @xcite .
one has to expect a complicated picture at higher excitations : residual interactions could mix the one - phonon and multiphonon vibrations of quadrupole and octupole character with each other and with quasiparticle excitations .
detailed experimental information on the properties of such excitations is needed . on the experimental side ,
two - neutron transfer reactions are very useful . on the theoretical side , a test of the advanced interacting boson model ( ibm ) and a microscopic approach , such as the quasiparticle - phonon model ( qpm )
, would be very interesting .
after the first observation of a large number of excitations with the @xmath6 transfer in the ( p , t ) reaction seen in the odd nucleus @xmath3pa @xcite , it was logical to investigate such excitations in the even - even nucleus @xmath0th , since @xmath3pa corresponds to @xmath0th + p , as well as in other actinide nuclei .
such measurements were carried out for the nuclei @xmath0th , @xmath7th and @xmath8u , and the results of a limited analysis have been published in @xcite ( besides the earlier preliminary study of @xmath0th and @xmath9u in @xcite ) .
the paper @xcite concentrated only on the energies of the excited @xmath1 states in these actinide nuclei and the ( p , t ) transfer strengths to these states .
the ( p , t ) reaction , however , gives much more extensive information on specific excitations in these nuclei , which was not analyzed previously .
such information was obtained for @xmath7th in our paper @xcite after detailed analysis of the experimental data from the ( p , t ) reaction . for the @xmath1 excitation
, we were able to derive additional information on the moments of inertia , which can be useful in clarifying the structure of these excitations . in this paper
we present the results of a careful and detailed analysis of the experimental data from the high - resolution study of the @xmath7th(p , t)@xmath0th reaction carried out to obtain deeper insight into all excitations in @xmath0th .
the total picture for @xmath0th has to differ from the one for @xmath7th , since the first one is considered as an octupole soft and the latter as a vibration - like nucleus
. it would be interesting to compare the @xmath1 excitations in the even nucleus @xmath0th and the odd nucleus @xmath3pa , the data for which in the low - energy part of excitations are known from the publication @xcite .
information on excited states of @xmath0th prior to this study was obtained mainly from the @xmath10-decay of @xmath8u , the @xmath5- and ec - decay of @xmath0ac and @xmath0pa , as well as from the ( @xmath10 , xn@xmath11)-reaction .
the most complete information was obtained from the @xmath5-decay of @xmath0ac reported by dalmasso et al .
@xcite and from the ec - decay study of @xmath0pa by weber et al . @xcite .
the lowest collective bands in @xmath0th were studied in the ( @xmath10 , xn@xmath11)-reaction @xcite . a total of 58 levels were reported in @xcite and 80 levels were observed in @xcite below 2.1 mev , connected by more than 240 @xmath11-rays that were established in these studies .
present results , derived from the @xmath7th(p , t)@xmath0th reaction , lead to about 163 levels in the energy range up to 3.25 mev .
unfortunately , during the experiment the radioactive target was destroyed and assignments were made only for 106 levels in the range up to 2.5 mev .
energies and cross sections for one angle were obtained additionally for 57 levels . besides 0@xmath4 excitations , where the number of reliable assignments could be increased for five states in comparison with the preliminary analysis in publication @xcite , information on the spins for many other states was obtained .
this information was essentially complementary to what was known from publications @xcite .
some levels are grouped into rotational bands , thus allowing to derive the moment of inertia for some @xmath1 , @xmath12 and @xmath13 , @xmath14 , @xmath15 , @xmath16 bands .
one of the results of ref .
@xcite was the establishment of the one - phonon octupole - quadruplet with @xmath17 states .
in this paper we suggested the two - phonon octupole - quadruplet with @xmath18 states .
a radioactive target of 100 @xmath19g/@xmath20 @xmath7th with half - life t@xmath21 = 8@xmath22 years , evaporated on a 22 @xmath19g/@xmath20 thick carbon backing , was bombarded with 25 mev protons at an intensity of 1 - 2 @xmath19a from the tandem accelerator of the maier - leibnitz - labor of the ludwig - maximilians - universitt and technische universitt mnchen .
the isotopic purity of the target was about 99% .
the tritons were analyzed with the q3d magnetic spectrograph and then detected in a focal plane detector .
the focal plane detector is a multiwire proportional chamber with readout of a cathode foil structure for position determination and de / e particle identification @xcite .
the acceptance of the spectrograph was 11 msr , except for the most forward angle of 5@xmath23 with an acceptance of 6 msr .
the resulting triton spectra have a resolution of 47 kev ( fwhm ) and are background - free .
the angular distributions of the cross sections were obtained from the triton spectra at ten laboratory angles from 5@xmath23 to 40@xmath23 for energies up to 1800 kev , but only at five angles from 7.5@xmath23 to 30@xmath23 for energies from 1800 to 2500 kev .
the energies and cross sections for the states from 2500 kev to 3250 kev were measured only for 10@xmath23 . a triton energy spectrum measured at a detection angle of 10@xmath23 is shown in fig .
[ fig : specth228_10deg ] .
the analysis of the triton spectra was performed with the program gaspan @xcite .
measurements were carried out with two magnetic settings : one for excitations up to 1.75 mev , and another one for the energy region from 1.6 mev to 3.3 mev , respectively .
for the calibration of the energy scale , the triton spectra from the reactions @xmath24w(p , t)@xmath25w , @xmath26w(p , t)@xmath24w and @xmath27u(p , t)@xmath8u were measured at the same magnetic settings .
the levels in @xmath7th known from the study @xcite were also included in the calibration . from 106 levels identified in the spectra
, 60 levels were identified for all ten angles and 46 levels only for 5 angles .
they are listed in table [ tab : expei ] .
the energies and spins of the levels as derived from this study are compared to known energies and spins , mainly from the published data @xcite .
they are given in the first four columns .
the ratios of cross sections at angles 7.5@xmath23 and 26@xmath28 to the one at 16@xmath28 , given in the next two columns , help to highlight the @xmath1 excitations ( large values ) . the column labelled @xmath29 gives the cross section integrated in the region from 7.5@xmath28 to 30@xmath28 , where the cross sections are measured for all level energies .
the column titled @xmath30 gives the ratio of the integrated cross sections , from experimental values and calculations in the dwba approximation ( see sec .
[ sec : dwba ] ) .
the last column lists the notations of the schemes used in the dwba calculations : sw.jj means one - step direct transfer of the @xmath31 neutrons in the ( p , t ) reaction ; notations of the multi - step transfers used in the dwba calculations are displayed in fig .
[ fig : schemes ] . additionally , energies of 57 levels seen only in the spectrum measured at 10@xmath23 and corresponding cross sections are listed in table [ tab : expesigma ] . + ll l ccc cc r c c + + & & & & & ratio&way of + & & [ 2,7 ] & this work&&(7.5@xmath32/16@xmath32 ) & ( 26@xmath32/16@xmath32)&[@xmath19b ] & @xmath33&fitting + + + + & & & & & ratio&way of + & & [ 2,7 ] & this work&&(7.5@xmath32/16@xmath32 ) & ( 26@xmath32/16@xmath32)&[@xmath19b ] & @xmath33&fitting + + 0.0 _ 2 & & 0.00 & @xmath1 & @xmath1 & & 5.83 & 5.61 & 165.56 & 6.20 & sw.gg + 57.8 _ 2 & & 57.76 & @xmath12 & @xmath12 & & 1.59 & 0.68 & 37.07 & 8.30&m1a.gg + 186.8 _ 2 & & 186.83 & @xmath34 & @xmath34 & & 0.74 & 0.38 & 9.07 & 1.90&m1a.gg + 328.0 _ 2 & & 328.00 & @xmath14 & @xmath14 & & 0.45 & 0.66 & 0.82 & 0.50&m2a.gg + 378.2 _ 2 & & 378.18 & @xmath2 & @xmath2 & & 0.58 & 0.71 & 4.48 & 1.60&m2a.gg + 396.9 _ 2 & & 396.08 & @xmath16 & @xmath16 & & 0.54 & 0.33 & 2.89 & 0.56&m3a.gg + 519.2 _ 3 & & 519.20 & @xmath35 & ( @xmath35 ) & & 1.23 & 1.33 & 0.43 & 0.90&sw.gg + 622.5 _ 4 & & 622.50 & @xmath36 & @xmath37 & & 0.20 & 0.94 & 0.26 & & + 695.6 _ 3 & & 695.50 & @xmath38 & @xmath39 & & 0.15 & 0.41 & 0.37 & & + 831.9 _ 2 & & 831.83 & @xmath1 & @xmath1 & & 12.06 & 7.50 & 39.10 & 360 & sw.ii + 874.4 _ 2 & & 874.42 & @xmath12 & @xmath12 & & 1.22 & 0.58 & 9.57 & 160 & m1a.ii + 911.6 _ 5 & & 911.80 & @xmath40 & & & & & & & + 920.6 _ 5 & & 920.80 & @xmath41 & & & & & & & + 938.7 _ 2 & & 938.55 & @xmath1 & @xmath1 & & 18.38 & 7.21 & 6.83 & 8.20 & sw.ii + 943.8 _ 4 & & 944.19 & @xmath14 & @xmath14 & & 0.12 & 0.67 & 0.37 & 1.00 & sw.gg + & & 968.33 & @xmath15 & & & & & & & + & & 968.43 & @xmath34 & & & & & & & + 968.8 _ 2 & & 968.97 & @xmath12 & @xmath12 & & 0.67 & 0.47 & 20.0 & 132&sw.ig + 979.4 _ 2 & & 979.50 & @xmath12 & @xmath12 & & 0.78 & 0.59 & 9.25 & 55.6&sw.ig + 1016.4 _ 2 & & 1016.43 & @xmath42 & @xmath16 & & 0.80 & 0.47 & 5.37 & 1.10&m3a.gg + & & 1022.53 & @xmath43 & & & & & & & + & & 1059.93 & @xmath44 & @xmath45 & & & & & & + 1074.8 _ 3 & & 1074.8 & @xmath34 & @xmath34 & & 0.74 & 1.32 & 1.62 & 0.26 & m1a.gg + 1091.0 _ 3 & & 1091.01 & @xmath34 & @xmath34 & & 0.74 & 0.44 & 0.42 & 0.10 & m1a.gg + 1105.5 _ 3 & & & & @xmath2 & & 0.61 & 0.56 & 0.77 & 21.0 & sw.ii + 1120.1 _ 3 & & 1120.1 & @xmath1 & @xmath1 & & 2.63 & 3.71 & 1.24 & 0.04 & sw.gg + & & 1122.95 & @xmath15 & & & & & & & + 1142.8 _ 3 & & 1143.2 & @xmath35 & @xmath35 & & 0.80 & 0.98 & 1.10 & 26.0 & sw.jj + 1153.3 _ 3 & & 1153.48 & @xmath12 & @xmath12 & & 0.65 & 0.49 & 23.89 & 140 & sw.ig + 1168.0 _ 4 & & 1168.37 & @xmath16 & @xmath16 & & 0.36 & 0.58 & 0.68 & 1.00 & w.gg + & & 1174.52 & @xmath46 & & & & & & & + 1175.2 _ 4 & & 1175.40 & @xmath12 & @xmath12 & & 1.05 & 0.91 & 2.09 & 13.0 & sw.ig + 1201.0 _ 9 & & 1200.54 & @xmath43 & @xmath43 & & 0.31 & 1.10 & 0.40 & 0.56 & m2a.gg + 1225.7 _ 6 & & & & @xmath34 & & 1.00 & 0.64 & 0.25 & 1.75 & sw.jj + & & 1226.55 & @xmath45 & & & & & & & + 1261.6 _ 3 & & 1261.5 & @xmath34 & @xmath34 & & 1.33 & 1.12 & 3.64 & 67.0 & sw.ii + 1270.2 _ 6 & & 1270.0 & & @xmath2 & & 0.40 & 0.97 & 0.31 & 0.15 & sw.gg + 1290.4 _ 3 & & 1290.2 & @xmath34 & @xmath34 & & 1.14 & 0.88 & 3.59 & 67.0 & sw.ii + 1296.0 _ 5 & & 1297.34 & @xmath35 & @xmath47 & & 1.33 & 1.23 & 0.50 & 1.00 & sw.gg + 1319.2 _ 4 & & & & @xmath48 & & 0.74 & 1.08 & 0.24 & 1.50 & sw.ig + 1343.9 _ 5 & & 1344.03 & @xmath16 & @xmath16 & & 0.77 & 0.33 & 0.31 & 0.08 & m3a.gg
+ & & 1393.4 & @xmath49 & @xmath50 & & & & & & + 1415.8 _ 6 & & 1415.92 & @xmath42 & @xmath51 & & 1.15 & 1.20 & 0.05 & 2.80 & sw.jj + 1423.8 _ 5 & & & & @xmath48 & & 2.20 & 1.33 & 0.16 & 0.03 & m1a.gg + 1432.1 _ 5 & & 1431.98 & @xmath52 & @xmath34 & & 1.61 & 1.17 & 0.21 & 6.80 & sw.ii + & & 1448.80 & @xmath53 & & & & & & & + & & 1450.29 & @xmath54 & & & & & & & + 1453.5 _ 5 & & & & @xmath51 & & 0.61 & 0.63 & 1.34 & 1.80 & sw.jj + 1470.0 _ 5 & & & & @xmath55 & & 0.94 & 1.81 & 0.19 & 0.01 & m3a.gg + 1497.4 _ 4 & & 1497.7 & @xmath56 & @xmath47 & & 1.07 & 0.91 & 0.37 & 0.56 & sw.gg + 1511.2 _ 3 & & & & @xmath1 & & 7.96 & 6.96 & 2.13 & 1.10 & sw.ig + 1531.7 _ 3 & & & & @xmath1 & & 2.21 & 0.83 & 0.47 & 2.60 & sw.ii + plus & & 1531.48 & @xmath43 & @xmath43 & & & & & 0.02 & m2a.gg + & & 1539.13 & @xmath57 & & & & & & & + 1544.4 _ 3 & & & & @xmath12 & & 1.27 & 0.65 & 1.61 & 1.53 & m1a.gg + & & 1581.0 & @xmath58 & & & & & & & + 1586.9 _ 4 & & & & @xmath12 & & 0.98 & 0.71 & 0.31 & 1.00 & sw.jj + & & 1588.33 & @xmath45 & & & & & & & + 1613.0 _ 5 & & & & @xmath34 & & 1.06 & 1.26 & 0.54 & 12.0 & sw.ii + 1618.3 _ 5 & & 1617.74 & ( @xmath59 & @xmath34 & & 0.88 & 0.76 & 1.22 & 0.16 & m2a.ii + 1627.9 _ 3 & & & & @xmath1 & & 7.44 & 5.21 & 9.66 & 10.0 & sw.ig + 1638.4 _ 4 & & 1638.25 & @xmath12 & @xmath12 & & 0.59 & 0.37 & 1.45 & 23.5 & sw.ii + & & 1643.18 & @xmath60 & & & & & & & + 1643.8 _ 3 & & 1643.8 & @xmath61 & @xmath34 & & 1.58 & 1.08 & 8.54 & 160 & sw.ii + & & 1645.89 & @xmath43 & & & & & & & + 1651.4 _ 3 & & & & @xmath51 & & 0.08 & 0.79 & 0.86 & 1.20 & sw.gg + 1667.3 _ 5 & & 1667.3 & & @xmath12 & & 0.71 & 0.61 & 3.17 & 46.0 & sw.ii + 1672.3 _ 5 & & & & @xmath12 & & 0.91 & 0.53 & 1.85 & 3.80 & sw.jj + 1678.4 _ 5 & & 1678.4 & @xmath62 & @xmath12 & & 0.88 & 0.75 & 1.48 & 19.5 & sw.ii + & & 1682.70 & @xmath63 & & & & & & & + & & 1683.74 & @xmath64 & & & & & & & + & & 1688.39 & @xmath43 & & & & & & & + 1691.3 _ 4 & & & & @xmath1 & & 2.66 & 2.06 & 1.26 & 0.75 & sw.ig + & & 1707.2 & @xmath65 & & & & & & & + 1710.7 _ 6 & & & & @xmath1 & & 1.38 & 1.86 & 0.54 & 0.02 & sw.gg + 1724.6 _ 4 & & 1724.29 & @xmath12 & @xmath12 & & 1.06 & 0.66 & 2.73 & 5.50 & sw.ii + 1733.8 _ 4 & & 1735.62 & @xmath34 & @xmath34 & & 1.13 & 0.86 & 2.28 & 3.50 & sw.ij + 1742.8 _ 4 & & 1743.86 & @xmath66 & @xmath34 & & 0.81 & 0.56 & 1.36 & 0.16 & m2a.gg + 1750.7 _ 3 & & & & @xmath1 & & 1.27 & 1.84 & 1.75 & 0.70 & sw.jj + 1758.1 _ 3 & & 1757.9 & @xmath67 & @xmath12 & & 0.85 & 0.75 & 4.35 & 26.0 & sw.ig + & & 1758.24 & @xmath63 & & & & & & & + & & 1760.25 & @xmath61 & & & & & & & + & & 1795.9 & @xmath68 & & & & & & & + 1796.8 _ 3 & & 1796.4 & @xmath69 & @xmath34 & & 1.40 & 0.83 & 6.47 & 89.0 & sw.ig + & & 1797.65 & @xmath70 & & & & & & & + 1803.0 _ 4 & & 1802.9 & @xmath71 & @xmath12 & & 0.65 & 0.49 & 15.34 & 90.0 & sw.ig + & & 1804.60 & @xmath72 & & & & & & & + & & 1811.5 & @xmath71 & & & & & & & + 1812.7 _ 4 & & & & @xmath55 & & 1.35 & 1.63 & 0.62 & 0.04 & sw.ig + & & 1817.43 & @xmath45 & & & & & & & + & & 1823.4 & @xmath73 & & & & & & & + 1826.2 _ 4 & & & & @xmath72 & & 1.16 & 0.83 & 1.91 & 7.50 & sw.ij + 1840.3 _ 8 & & 1842.2 & @xmath42 & & & 1.41 & 0.33 & 0.21 & & + 1858.6 _ 5 & & & & @xmath55 & & 0.65 & 1.19 & 1.28 & 0.06 & sw.ig + 1863.9 _ 5 & & 1864.8 & @xmath71 & @xmath48 & & 0.75 & 0.79 & 1.47 & 8.10 & sw.ig + & & 1876.5 & @xmath74 & & & & & & & + 1878.9 _ 5 & & 1879.0 & @xmath74 & @xmath51 & & 1.05 & 0.91 & 1.93 & 110 & sw.ii + & & 1892.98 & @xmath43 & & & & & & & + 1898.2 _ 4 & & 1899.98 & @xmath12 & @xmath48 & & 0.84 & 0.81 & 2.55 & 140 & sw.ii + & & 1901.90 & @xmath34 & & & & & & & + 1903.9 _ 4 & & & & ( @xmath2 ) & & 0.69 & 1.58 & 1.54 & 0.07 & sw.gg + & & 1906.78 & @xmath70 & & & & & & & + & & 1908.4 & @xmath16 & & & & & & & + 1908.9 _ 7 & & & & @xmath1 & & 2.17 & 1.91 & 4.56 & 1.30 & sw.jj + & & 1924.1 & @xmath75 & & & & & & & + & & 1924.6 & @xmath76 & & & & & & & + 1925.4 _ 4 & & 1925.20 & @xmath34 & @xmath56 & & 0.61 & 1.73 & 0.54 & 21.0 & sw.ii + & & 1928.54 & @xmath43 & & & & & & & + & & 1937.16 & @xmath63 & & & & & & & + 1938.3 _ 4 & & 1938.9 & @xmath34 & ( @xmath34 ) & & 1.06 & 0.76 & 1.99 & 0.67 & m2a.gg + & & 1944.85 & @xmath43 & & & & & & & + & & 1945.8 & @xmath56 & & & & & & & + 1947.8 _ 7 & & & & ( @xmath12 ) & & 1.02 & 0.75 & 0.77 & 3.50 & sw.ig + & & 1949.7 & @xmath77 & & & & & & & + 1959.7 _ 6 & & 1958.5 & @xmath12 & @xmath48 & & 0.10 & 1.69 & 0.43 & 1.50 & sw.ig + & & 1964.90 & @xmath48 & & & & & & & + 1971.7 _ 4 & & & & @xmath78 & & 0.66 & 0.81 & 0.79 & 3.10 & sw.ig + & & 1974.20 & @xmath34 & & & & & & & + 1981.9 _ 4 & & 1981.97 & @xmath66 & @xmath51 & & 1.68 & 0.77 & 1.70 & 2.60 & sw.gg + & & 1987.46 & @xmath34 & & & & & & & + 1993.9 _ 5 & & & & ( @xmath16 ) & & 0.97 & 0.72 & 1.80 & 2.80 & sw.gg + 2010.4 _ 6 & & 2010.15 & @xmath79 & @xmath48 & & 0.46 & 0.43 & 0.76 & 13.0 & sw.ig + & & 2013.6 & @xmath63 & & & & & & & + & & 2016.75 & @xmath56 & & & & & & & + & & 2022.73 & @xmath12 & & & & & & & + 2030.3 _ 4 & & 2029.6 & @xmath48 & @xmath12 & & 0.54 & 0.23 & 0.84 & 16.0 & sw.ig + & & 2037.0 & @xmath63 & & & & & & & + 2044.7 _ 5 & & & & @xmath1 & & 9.22 & 4.56 & 0.57 & 3.10 & sw.ii + 2052.1 _ 4 & & & & ( @xmath2 ) & & 0.72 & 1.30 & 3.70 & 180 & sw.ii + 2069.6 _ 5 & & & & @xmath12 & & 0.76 & 0.56 & 1.38 & 6.10 & sw.ig + 2079.9 _ 5 & & & & @xmath1 & & 17.08 & 13.13 & 4.62 & 25.9 & sw.ii + 2091.2 _ 7 & & & & @xmath55 & & 0.62 & 0.82 & 1.20 & 35.0 & sw.ii + 2111.6 _ 5 & & & & @xmath48 & & 0.70 & 0.71 & 2.57 & 11.0 & sw.ig + & & 2123.1 & @xmath78 & & & & & & & + 2131.3 _ 6 & & & & @xmath1 & & 6.84 & 4.53 & 24.80 & 120 & sw.ii + 2152.8 _ 4 & & & & @xmath72 & & 1.30 & 0.90 & 4.13 & 98.0 & sw.ii + 2159.4 _ 6 & & & & @xmath1 & & 3.78 & 1.54 & 1.18 & 8.10 & sw.ii + 2170.3 _ 4 & & & & @xmath48 & & 1.00 & 0.85 & 5.61 & 26.0 & sw.ig + 2198,2 _ 4 & & & & @xmath12 & & 0.59 & 0.62 & 3.81 & 19.5 & sw.ig + 2215.9 _ 4 & & & & ( @xmath34 ) & & 1.40 & 1.15 & 6.00 & 130 & sw.ii + 2235.2 _ 7 & & & & ( @xmath34 ) & & 0.98 & 0.86 & 2.82 & 61.0 & sw.ii + 2290.0 _ 7 & & & & @xmath1 & & 9.96 & 5.75 & 11.00 & 61.0 & sw.ii + 2302.9 _ 5 & & & & @xmath72 & & 1.09 & 0.84 & 2.75 & 62.0 & sw.ii + 2323.2 _ 5 & & & & @xmath12 & & 0.41 & 0.62 & 2.24 & 16.0 & sw.ig + 2335.9 _ 5 & & & & ( @xmath34 ) & & 2.13 & 1.65 & 17.10 & 370 & sw.ig + & & & & ( @xmath1 ) & & & & 4.50@xmath80 & 25.0 & sw.gg+14 + 2344.2 _ 5 & & & & ( @xmath16 ) & & 0.77 & 0.58 & 6.65 & 10.0 & sw.gg + 2356.2 _ 5 & & & & ( @xmath12 ) & & 0.63 & 0.61 & 4.61 & 21.5 & sw.ig + 2375.5 _ 8 & & & & ( @xmath12 ) & & 0.78 & 0.60 & 4.87 & 22.0 & sw.ig + 2398.3 _ 9 & & & & ( @xmath16 ) & & 0.76 & 0.75 & 7.36 & 11.0 & sw.gg + 2408.8 _ 9 & & & & ( @xmath34 ) & & 1.87 & 1.28 & 2.34 & 60.0 & sw.ii + 2441.7 _ 5 & & & & ( @xmath12 ) & & 0.71 & 0.51 & 10.32 & 47.0 & sw.ig + 2456.8 _ 5 & & & & @xmath1 & & 16.18 & 1.27 & 0.53 & 5.20 & sw.ii + 2476.7 _ 5 & & & & ( @xmath12 ) & & 0.62 & 0.52 & 10.38 & 48.0 & sw.ig + 2494.1 _ 5 & & & & ( @xmath12 ) & & 0.65 & 0.47 & 12.74 & 63.5 & sw.ig + + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath80 the value after subtracting a constant of 14 @xmath19b ( see text in sec .
[ sec : dwba ] for explanation . + lclclc + e [ kev ] & @xmath81 & e [ kev ] & @xmath81 & e [ kev ] & @xmath81 + + + 2513.5 _ 7 & 2.00 & 2742.3 _ 4 & 5.50 & 3035.6 _ 9 & 0.95 + 2531.5 _ 7 & 6.60 & 2763.7 _ 4 & 8.60 & 3046.4 _ 6 & 2.10 + 2536.8 _ 9 & 3.20 & 2781.4 _ 5 & 1.75 & 3059.2 _ 5 & 2.15 + 2542.4 _ 9 & 1.85 & 2798.6 _ 8 & 1.55 & 3075.2 _ 5 & 2.20 + 2554.5 _ 5 & 6.00 & 2805.6 _ 7 & 2.00 & 3085.2 _ 8 & 1.25 + 2566.3 _ 6 & 2.20 & 2821.0 _ 5 & 2.90 & 3097.0 _ 6 & 3.10 + 2595.4 _ 5 & 5.40 & 2839.3 _ 6 & 1.30 & 3104.7 _ 6 & 3.40 + 2606.1 _ 5 & 23.5 & 2853.7 _ 5 & 2.75 & 3112.7 _ 11 & 1.70 + 2615.1 _ 9 & 0.15 & 2868.1 _ 5 & 3.20 & 3119.9 _ 9 & 2.30 + 2634.8 _ 5 & 1.60 & 2877.5 _ 8 & 1.80 & 3128.2 _ 10 & 1.25 + 2644.0 _ 3 & 9.20 & 2883.7 _ 9 & 1.60 & 3158.8 _ 8 & 1.50 + 2657.1 _ 4 & 5.20 & 2918.8 _ 6 & 1.85 & 3165.7 _ 6 & 2.00 + 2660.1 _ 5 & 6.00 & 2927.4 _ 5 & 3.25 & 3186.0 _ 6 & 2.00 + 2667.1 _ 5 & 3.30 & 2936.8 _ 9 & 1.40 & 3195.2 _ 6 & 2.60 + 2676.0 _ 6 & 67.2 & 2945.3 _ 9 & 1.35 & 3209.6 _ 12&1.40 + 2688.4 _ 4 & 2.10 & 2955.1 _ 8 & 1.25 & 3214.8 _ 9 & 2.20 + 2695.6 _ 7 & 1.10 & 2993.1 _ 12 & 1.00 & 3225.0 _ 20 & 0.50 + 2705.5 _ 5 & 1.35 & 2999.5 _ 10 & 1.50 & 3232.9 _ 13 & 1.20 + 2718.4 _ 5 & 2.10 & 3014.3 _ 11 & 0.80 & 3239.9
_ 8 & 3.40 + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the spins of the excited states in the final nucleus @xmath0th were assigned via an analysis of the angular distributions of tritons from the ( p , t ) reaction .
the angular distributions for @xmath1 excitations have a steeply rising cross section at very small reaction angles , and a sharp minimum at a detection angle of about 14@xmath23 .
this pronounced feature helped to identify most of these states in complicated and dense spectra , even without fitting experimental angular distributions .
no complication of the angular distributions was expected , since the excitation of @xmath1 states predominantly proceeds via a one - step process .
this is not the case for the excitation of states with other spins , where multi - step processes could play a very important role .
the identification of these states is possible by fitting the experimental angular distributions with those calculated in the distorted - wave born approximation ( dwba ) .
the potential parameters suggested by becchetti and greenlees @xcite for protons and by flynn _
@xcite for tritons were used in the calculations .
these parameters have been tested via their description of angular distributions for the ground states of @xmath0th , @xmath7th and @xmath8u @xcite .
minor changes of the parameters for tritons were needed only for some @xmath16 states , particularly for the states at 396.9 and 1016.4 kev . for these states ,
the triton potential parameters suggested by becchetti and greenlees @xcite were used . for each state the binding energies of the two neutrons are calculated to match the outgoing triton energies .
the corrections to the reaction energy are introduced depending on the excitation energy . for more details
see @xcite .
a problem arising in such calculations is the lack of prior knowledge of the microscopic structure of these states .
we can assume , however , that the overall shape of the angular distribution of the cross section is rather independent of the specific structure of the individual states , since the wave function of the outgoing tritons is restricted to the nuclear exterior and therefore to the tails of the triton form factors . to verify this assumption ,
dwba calculations of angular distributions for different @xmath31 transfer configurations to states with different spins were carried out in our previous paper @xcite .
the coupled - channel approximation ( chuck3 code of kunz @xcite ) was used in these calculations .
indeed , the calculated angular distributions are very similar .
nevertheless , they depend to some degree on the transfer configuration , the most pronounced being found for the 0@xmath4 states , what is confirmed by the experimental angular distributions .
the best reproduction of the angular distribution for the ground state was obtained for the transfer of the @xmath82 configuration in the one - step process .
this orbital is close to the fermi surface and was considered as the most probable one in the transfer process .
other transfer configurations that might be of importance are @xmath83 and @xmath84 , also near the fermi surface .
better reproduction of the angular distribution for some @xmath1 states is obtained just for these configurations .
the main features of the angular distribution shapes for 2@xmath4 and 4@xmath4 states are even more weakly dependent on the transfer configurations .
nevertheless the @xmath82 , @xmath83 and @xmath84 configuration , alone or in combination , were used in the calculations for these states too .
results of fitting the angular distributions for the states assigned as @xmath1 excitations are shown in fig .
[ fig : angl_distr_0 + ] .
the agreement between the fit and the data is excellent for most of the levels .
remarks are needed only for the levels at 1531.7 and 2335.9 kev .
the spin @xmath43 was assigned to the level at 1531.47 kev in @xcite .
a level at the close - lying energy of 1531.7 kev has been observed also in the ( p , t ) reaction , but the angular distribution of tritons can not be fitted by calculations for transition to the @xmath43 state .
the maximum cross section for forward angles suggests the presence of a @xmath1 excitation , though the angular distribution fitted by a calculation for transfer of the @xmath83 configuration to the @xmath1 state is not perfect .
a satisfactory fit of the experimental angular distribution was obtained assuming overlapping states with spins @xmath1 and @xmath43 ( fig .
[ fig : angl_distr_0 + ] ) , thus confirming the assignment for both states .
an ambiguous picture is observed for the 2335.9 kev state , where the angular distribution is measured for a limited range of angles .
the fitting agreement is perfect for a transition to the 4@xmath4 state , but the cross section is surprisingly large for a 4@xmath4 state , twice larger than for the @xmath34 member of the g.s . band .
therefore , the possibility of a 0@xmath4 excitation can not be excluded , but the experimental angular distribution is fitted for a @xmath1 state only with adding a constant of 14 @xmath19b .
this ambiguity can be resolved by measurements in wider angular regions .
thus we can make firm @xmath1 assignments for 17 states for energies excitations below 2.5 mev , in comparison with 12 states found in the preliminary analysis of the experimental data @xcite .
of course , some higher lying 0@xmath4 levels are lost because of the cutoff of the investigated energy region .
but as follows from a similar study for @xmath7th , only a few @xmath1 states are observed above 2500 kev , where the density of @xmath1 excitations decreases for higher energies ( or else that the cross section of such excitations is very low and they are hidden in very dense and complicated spectra ) .
therefore , we can compare 24 @xmath1 states in @xmath7th with only 17 @xmath1 states in @xmath0th in the same energy region .
similar to @xmath1 excitations , the one - step transfer calculations give a satisfactory fit of angular distributions for about 80% of the states with spins different from @xmath1 , but about 20% of these states need the inclusion of multi - step excitations .
multi - step excitations have to be included to fit the angular distributions already for the @xmath12 , @xmath34 and @xmath2 states of the g.s . band .
[ fig : schemes ] shows the schemes of the multi - step excitations , tested for every state in those cases , where one - step transfer did not provide a successful fit .
[ fig : angl_distr_natur ] demonstrates the quality of the fit of some different - shaped angular distributions for the excitation of states with spin @xmath12 by calculations assuming one - step and one - step plus two - step excitations , respectively .
results of similar fits for the states assigned as @xmath34 , @xmath2 and @xmath14 , @xmath16 , @xmath35 excitations are shown in fig .
[ fig : angl_distr_unnatur ] .
the assignments of the spins resulting from such fits are presented in table [ tab : expei ] , together with other experimental data .
special comments are needed for the column displaying the ratio @xmath85 . calculated
cross sections for the specific transfer configurations differ very strongly .
since we have no a priori knowledge of the microscopic structure of the excited states , and thus do not know the relative contributions of the specific @xmath31 transfer configurations to each of these states , these ratios can not be considered as spectroscopic factors .
nevertheless , a very large ratio , such as in the case of the @xmath83 transfer configurations used in the calculation for some @xmath1 and even @xmath12 and @xmath34 states , is unexpected .
surprisingly , the shape just for this neutron configuration gives the best agreement with experiment .
some additional comments on table [ tab : expei ] are needed . in all cases , where the firm assignment were known from the previous studies @xcite ,
they are confirmed by the ( p , t ) angular distribution analysis . in those cases , where two or three possible spin assignments were proposed earlier , the ( p , t ) angular distribution analysis allows to select only one assignment almost in all cases . for the energies above 2030 kev only the assignments from the ( p , t ) reaction are possible at present .
the following remarks are needed in those cases , where the assignments from different publications are in contradiction , which can be removed using the data from the ( p , t ) reaction . _
938.7 kev .
_ spin @xmath15 was assigned for this level in @xcite .
our fit of the angular distribution gives reliably spin @xmath1 in agreement with @xcite . _
943.8 kev .
_ spin @xmath12 was assigned for this level in @xcite .
the angular distribution rejects this value and agrees with the assignment of spin @xmath14 accepted in @xcite .
_ 968 - 969 kev .
_ three levels around 968 kev with spins @xmath86 were identified in @xcite and two levels with spins @xmath87 in @xcite .
there is a discrepancy in assignment for the 968.33 kev level as @xmath16 in @xcite and as @xmath15 in @xcite .
this line is masked in the ( p , t ) spectrum by a strong line from the transition to the 969 kev level .
but since the angular distribution is very well described by a calculation leading to a @xmath12 state and is not disturbed by a transition to spin @xmath16 , the assignment @xmath15 is preferable ( transition is weak ) . _
1016.4 kev . _
the discrepancy in assignment for the 1016.4 kev level as @xmath12 in @xcite and as @xmath16 in @xcite can not be removed by the ( p , t ) angular distribution , since it can be fitted by a transition sw.ig to @xmath12 and m3a.gg to @xmath16 , respectively .
however , transitions seen in the decay of @xmath0pa @xcite from this state to the @xmath35 and @xmath34 states leads to the assignment of @xmath16 .
we accepted this spin also due to strong arguments in @xcite , including the assignment of spins @xmath15 and @xmath14 to the levels at 968.3 kev and 943.8 kev as members of the @xmath88 band .
_ 1059.9 kev .
_ from the tentative assignments @xmath89 in @xcite and the firm assignment @xmath45 in @xcite for this level , the latter has to be additionally supported by the fact , that the corresponding line in the ( p , t ) spectrum was not seen ( transition to the state of unnatural parity ) . _
1225.7 and 1226.56 kev .
_ spin @xmath45 was assigned for the level 1226.56 kev in both @xcite and @xcite .
the level with the close energy 1225.7 kev is seen in the ( p , t ) reaction , but the angular distribution agrees with the assignment of spin @xmath34
. therefore both levels are present in table [ tab : expei ] . _
1393.4 kev_. this level was observed in the decay of @xmath0pa with restriction of the spin - parity to @xmath90 , @xmath91 and @xmath16 by its population and depopulation @xcite .
additional restriction from the @xmath92 angular distribution ratio indicates that this level has most likely @xmath93 and that the @xmath12 and @xmath16 assignments are nearly excluded .
this level was not observed in the ( p , t ) reaction , thus supporting such conclusion .
_ 1415.8 kev .
_ spin @xmath12 was assigned to this level in @xcite and spins @xmath12 or @xmath16 were allowed in @xcite .
the angular distribution of tritons gives preference to spin @xmath16 .
_ 1432.1 kev . _
the discrepancy between @xmath43 in @xcite and @xmath34 in @xcite for the 1432.1 kev level is removed already by the fact of the excitation of this state in the ( p , t ) reaction , and additionally by the angular distribution leading to the @xmath34 assignment .
also additional lines observed in the decay of @xmath0pa @xcite , leading to the 6@xmath4 level , confirm this assignment .
_ 1450.29 kev .
_ spin @xmath16 was assigned to this level in @xcite and spin @xmath45 in @xcite .
the fact that this level is not observed in the ( p , t ) reaction gives preference for an assignment of spin @xmath45 not excluding spin @xmath16 .
_ 1531.7 kev .
_ spin @xmath43 was assigned for the level at 1531.47 kev both in @xcite and @xcite . however , the angular distribution of tritons for the level with the close - lying energy 1531.7 kev indicates another spin value .
it has a steeply rising cross section at small angles as for a @xmath1 excitation , however , the minimum at a detection angle about @xmath94 is not sharp . therefore we assumed an overlapping of peaks of two levels , one of which is a @xmath1 level . _ 1643.8 kev .
_ there are two close - lying levels at 1643.18 kev with spin 2@xmath95 or 3@xmath95 identified both in @xcite and @xcite and 1643.8 kev , respectively , as identified in @xcite with an assignment of possible spins ( 2,3,4)@xmath4 . only the level at 1643.8 kev
is observed in the ( p , t ) reaction with clear assignment of spin 4@xmath4 .
_ 1733.8 kev .
_ we assumed that the level at 1735.6 kev , identified as a 4@xmath4 state in @xcite and as 2@xmath4,3,4@xmath4 state in @xcite , and the level at 1733.8 kev seen in the ( p , t ) reaction with an assignment of spin 4@xmath4 are identical , though the energy difference is larger than the energy error .
_ 1742.8 kev .
_ spin 3 was assigned to the level at 1743.86 kev in @xcite and spin 4@xmath4 in @xcite .
the angular distribution from the ( p , t ) reaction prefers the assignment of spin 4@xmath4 .
_ 1758 - 1760 kev .
_ several close - lying levels were identified at 1757.9 kev with spin @xmath71 @xcite , at 1758.24 kev with spin ( 3,4)@xmath4 @xcite , at 1760.25 kev with spin 4@xmath4 @xcite and with spin @xmath96 @xcite .
different @xmath11-lines were used in the identification of the levels at 1757.9 and 1758.24 kev : 741.9 , 1361.4 , 1430.0 kev for the first one and 1571.52 and 1700.59 kev for the latter . at the same time , in @xcite the line at 1430.0 kev was used for the identification of another level at 1617.74 kev , and the important line at 1758.24 kev was used for the identification of the level at 1944.85 kev .
the line at 1758.11 kev can be attributed to the decay of the level at 1758.24 kev , then the spin of this state distinctly has to be @xmath12 .
the ambiguity can not be solved with the ( p , t ) data .
therefore we put the level at 1758.1 kev with an assignment of spin @xmath12 from the ( p , t ) study in correspondence with the level at 1757.9 kev in @xcite , but for the level at 1758.24 kev in @xcite we do not exclude the spin @xmath12 , too . as far as the level at 1760.25 kev
is concerned , a spin @xmath12 can be nearly excluded , since this line is not observed in the ( p , t ) reaction .
_ 1796.8 kev .
_ two close - lying levels were identified : 1795.9 kev with an assignment @xmath89 in @xcite and 1796.4 kev with an assignment as @xmath69 in @xcite .
the level at 1796.8 kev with spin @xmath34 is observed in the ( p , t ) reaction .
it is problematic to put this level in correspondence with one of the observed ones in decay , considering the assignments .
therefore only energetic proximity was taken into account .
_ 1908.9 kev .
_ the level at 1908.4 kev , @xmath51 was identified in @xcite , however , a level with almost the same energy of 1908.9 kev as observed in the ( p , t ) reaction was clearly identified as a @xmath1 state , they must be considered as a different levels . _
2010.4 kev .
_ there is discrepancy in the assignment of spin to the level at 2010.15 kev : @xmath97 in @xcite and @xmath34 in @xcite .
the angular distribution from the ( p , t ) reaction prefers spin @xmath48 .
+ cccccccccc + @xmath1 & @xmath90 & @xmath12 & @xmath43 & @xmath34 & @xmath46 & @xmath2 & @xmath98 & @xmath99 & + + 0.0 & & 57.8 & & 186.8 & & 378.2 & & 622.5 & + 831.9 & & 874.4 & & 968.4 & & 1105.5 & & 1281 & + & & 968.8 & 1022.5 & 1091.0 & 1174.5 & 1270.2 & 1380 & 1497 & + & & 1153.5 & 1200.5 & 1261.5 & & & & & + 938.6 & & 979.5 & & 1074.8 & & & & & + 1120.1 & & 1175.4 & & 1290.2 & & 1470.0 & & & + & & 1319.2 & & 1432.0 & & & & & + 1511.2 & & 1544.4 & & 1618.3 & & & & & + & & 1638.3 & 1688.4 & 1760.2 & & & & & + & & & & 1643.8 & & ( 1812.7 ) & & & + ( 1531.7 ) & & 1586.9 & & & & & & & + 1627.9 & & 1667.3 & & 1742.8 & & 1858.6 & & & + & & & & 1796.8 & & 1903.9 & & & + & & 1724.6 & 1760.3 & 1826.2 & & & & & + & & 1803.0 & & 1938.3 & & & & & + & & 1898.2 & 1937.2 & 1987.5 & & & & & + & & 1899.98 & 1944.85 & ( 2010.17 ) & & & & & + 1908.9 & & 1947.8 & & 2037.0 & & & & & + 2079.9 & & 2111.6 & & & & & & & + 2131.3 & & 2170.3 & & & & & & & + 2290.0 & & 2323.2 & & & & & & & + ( 2335.9 ) & & 2375.5 & & & & & & & + + @xmath100 & @xmath14 & @xmath15 & @xmath16 & @xmath45 & @xmath35 & @xmath101 & @xmath38 & @xmath102 & @xmath41 + + @xmath13 & 328.0 & & 396.9 & & 519.2 & & 695.6 & & 920.6 + @xmath14 & 943.8 & 968.3 & 1016.4 & 1059.2 & 1143.2 & & & & + @xmath15 & & 1122.9 & 1168.4 & 1226.6 & 1297.3 & & & & + @xmath16 & & & 1344.0 & & 1497.7 & & & & +
after the assignment of spins to all excited states , those sequences of states can be identified , which show the characteristics of a rotational band structure .
an identification of the states attributed to rotational bands was made on the basis of the following conditions : \a ) the angular distribution for a band member candidate state is fitted by dwba calculations for the spin necessary to put this state in the band ; \b ) the transfer cross section in the ( p , t ) reaction to states in the potential band has to decrease with increasing spin ; \c ) the energies of the states in the band can be fitted approximately by the expression for a rotational band @xmath103 with a small and smooth variation of the inertial parameter @xmath104 .
collective bands identified in such a way are shown in fig .
[ fig : bands ] and are listed in table [ tab : bands ] ( for a calculation of the moments of inertia ) .
the procedure can be justified in that some sequences meeting the above criteria are already known from gamma - ray spectroscopy to be rotational bands @xcite , so similar sequences are rotational bands , too .
the straight lines in fig .
[ fig : bands ] strengthen the argument for these assignments .
it is worth mentioning , that the assignments of @xmath1 even for the states at 1531.7 and 2335.9 kev are supported by one @xmath12 state on top of them .
the bands built on states of one - phonon octupole - quadruplet ( the band @xmath88 was not correctly identified in @xcite ) , the band with @xmath105 , 831.9 kev @xcite , @xmath105 , 938.6 , 1120.1 kev @xcite and @xmath106 , 968.8 and 1153.5 kev @xcite were identified earlier .
additional levels are added to these bands from the ( p , t ) study ( table [ tab : bands ] ) .
two bands with @xmath106 are added in table [ tab : bands ] , based only on the analysis of the decay of @xmath0ac @xcite : at 1638.23 kev and at 1899.98 kev .
there is only contradiction in the spin assignment for the 2010.17 kev level .
the ( p , t ) data do not support spin @xmath34 and prefer a @xmath12 assignment instead . in fig .
[ fig : moments ] we present moments of inertia ( moi ) obtained by fitting the level energies of the bands displayed in fig .
[ fig : bands ] by the expression @xmath107 for close - lying levels , i.e. they were determined for band members using the ratio of @xmath108 and @xmath109 $ ] , thus saving the spin dependence of the moi .
this procedure is valid for all bands except the 943.8 kev , @xmath14 band .
the usual procedure leads to strongly staggering values . in the case of the @xmath110 and @xmath111 bands
, the coriolis interaction mixes the band members only for @xmath112 odd . the @xmath112 even members of the @xmath111 band remain unperturbed . in a simple two level model ( @xmath110 and @xmath111 bands )
the following expression can be obtained for the band energies @xmath113 where @xmath114 and @xmath115 , @xmath116 and @xmath117 are the intrinsic bandhead energies , @xmath118 is the inertial parameter and @xmath119 is the strength of the coriolis interaction , which is believed to be small .
an effective parameters of inertia behaves then as @xmath120 + fitting these expressions to the experimental data gives the smoothly changing values of the moment of inertia between 76.9 and 78.6 as shown in fig.[fig : moments ] with parameter @xmath121 ( thus staggering is removed ) . in most cases
moi slightly increase with the increasing spin .
there are few cases of @xmath1 bands where moi are sloping down .
this can be explained as an effect of the coriolis interaction of a 2qp @xmath122 band with a nearby lying 2qp @xmath123 band ( not seen in the ( p , t ) reaction ) .
a similar effect was observed e.g. in @xmath124er for the @xmath16 and @xmath45 2qp bands @xcite . in @xmath0th
an additional moi even - odd spin staggering is expected for the @xmath123 band similar to that for the octupole @xmath88 band since the odd spin states of the @xmath123 band have no counterparts in the @xmath105 band .
the obtained moi cover a broad range , from @xmath12550 mev@xmath126 to @xmath125100 mev@xmath126 .
the negative parity bands based on the states with spin 1@xmath95 , interpreted as the octupole - vibrational bands @xcite , have high moi .
the @xmath1 band at 1120.1 kev , considered as @xmath5 - vibrational band , has the smallest moi , close to the one of the ground - state band . at this stage , it is difficult to state a complete correlation between the intrinsic structure of the bands and the magnitude of their moi .
nevertheless , one can assume for the @xmath1 bands that some of the larger moi could be related to the two - phonon octupole structure and the smallest moi could be related to the one - phonon quadrupole structure .
the bands with intermediate values of the moi could be based on the two - phonon quadrupole excitations .
if the moments of inertia do indeed carry information on the inner structure of the bands , then the number of excitations with a structure as in the g.s . or @xmath5-vibrational states in @xmath0th is small . the lowest negative - parity excitations with @xmath17 are generally interpreted as octupole vibrational .
they are one - phonon octupole excitations .
the corresponding energies in @xmath0th as 328.0 , 944.2 , 1123.0 , and 1344.0 kev were established already in @xcite .
here we confirmed these assignments after the removal of some ambiguities .
they are the bandheads of the rotational bands which are displayed in fig .
[ fig : quadruplet ] together with the ( p , t ) cross sections and their parameters @xmath104 . in the case of @xmath7th ,
the assumption was made that the strongly excited first 0@xmath4 state , together with also strongly excited states with spins 2@xmath4 and 4@xmath4 , accompanied by somewhat weaker excited state with spin 6@xmath4 , belong to the two - phonon octupole quadruplet @xcite .
strong excitations and close rotational parameters were the arguments for assigning the same structure to these states .
as one can see in fig .
[ fig : strength ] , the first 0@xmath4 state in @xmath0th at 831.9 kev is also strongly excited . taking into account the decay properties of the band on this 0@xmath4 state , the suggestion was made that this band has an octupole two - phonon structure @xcite .
the picture for other states is not so transparent .
there is no prominent excitation strength of the @xmath12 and @xmath34 states just above this 0@xmath4 state .
the first excited 2@xmath4 state , which is not a member of the 0@xmath4 band , is the state at 968.8 kev .
but the de - excitation of the band built on this state demonstrates the properties expected for a @xmath11-vibrational band .
moreover , its moment of inertia is much smaller than the one derived for the 0@xmath4 band at 831.9 kev . for the band built on the @xmath12 state at 1153.3 kev ,
the moment of inertia is close to the one of the band built on the @xmath1 state at 831.9 kev and the state at 1153.3 kev is relatively strong excited in the ( p , t ) reaction .
the @xmath34 and @xmath2 states , which do not belong to rotational bands , which are strongly excited in the ( p , t ) reaction and could be members of the two - phonon octupole quadruplet , are the states at 1643.8 and 1905.8 kev .
the level at 1812.7 kev , tentatively assigned as a @xmath2 state , can be attributed to the band based on the @xmath34 state , the corresponding inertial parameter again is very close to the one for @xmath1 and @xmath12 bands .
no members of a rotational band can be related to the @xmath2 band head at 1905.8 kev . in the interacting boson model ( ibm ) ,
the positive - parity states are described by introducing @xmath127 and @xmath128 bosons , while for the negative parity states one has to introduce additional bosons with odd values of angular momentum ( at least one @xmath129 boson ) . in the region of transitional actinides , where octupole deformation might develop , the ibm-@xmath130 ( which uses @xmath131 and @xmath129 bosons ) was applied with success in refs .
@xcite . in the present paper
, we adopt the ibm-@xmath130 framework for calculating the low - lying positive and negative parity states in @xmath0th . in ref .
@xcite , the ibm calculations for this nucleus have been already performed .
however , these calculations used only a simplified hamiltonian to describe the existing ( up to that date ) electromagnetic data .
more recent calculations ( which also used a simplified hamiltonian ) @xcite indicated that ibm fails completely to reproduce the ( p , t ) spectroscopic factors .
the calculated first excited states were found with a transfer strength of @xmath1321@xmath133 of that of the ground state and the higher states were even weaker , whereas experimentally the first excited state is seen with @xmath13230@xmath133 of the ground - state intensity . in order to treat these spectroscopic observables in a reasonable approach ,
we used the method suggested in ref .
@xcite , where it was shown that the addition of the second - order o(5 ) casimir operator in the hamiltonian can account for the observed ( p , t ) spectroscopic factors .
the hamiltonian employed in the present paper is similar to the one used in refs .
@xcite and is able to describe simultaneously the positive and negative parity states : @xmath134^{(0)}\label{eq1},\end{aligned}\ ] ] where @xmath135 , @xmath136 , and @xmath137 are the boson energies and @xmath138 , @xmath139 , and @xmath140 are the boson number operators , @xmath141 is the quadrupole - quadrupole interaction strength and @xmath142 is the strength of the o(5 ) second order casimir operator . in the @xmath130 model ,
the quadrupole operator is considered as being @xcite : @xmath143^{(2)}\nonumber\\ + \chi^{(2)}_{pf } \left\ { \frac{9\sqrt{3}}{10}(p^{\dagger}\tilde{p})^{(2)}+\frac{3\sqrt{42}}{10}(f^{\dagger}\tilde{f})^{(2)}\right\}\label{eq2}\end{aligned}\ ] ] the quadrupole electromagnetic transition operator is : @xmath144 where @xmath145 represents the boson effective charge . to ensure no - vanishing e2 transitions between the states containing no @xmath146 bosons and those having @xmath147 components we follow the approach described in refs .
@xcite , where the mixing of different positive parity - states with different @xmath146 components is achieved by introducing in the hamiltonian a dipole - dipole interaction term of the form : @xmath148 where @xmath149^{(1 ) } + \sqrt{5}[{s}^{\dagger}\tilde{p}+{p}^{\dagger}\tilde{s}]^{(1)}\\\nonumber + \sqrt{7}[{d}^{\dagger}\tilde{f}+{f}^{\dagger}\tilde{d}]^{(1)}\label{eq7}\end{aligned}\ ] ] is the dipole operator arising from the @xmath150(4 ) dynamical symmetry limit , which does not conserve separately the number of positive and negative parity bosons @xcite .
this term will also be important later in the calculations of the two - neutron transfer intensities .
the interaction strength is given by the @xmath10 parameter and is chosen to have a very small value , @xmath10=0.0005 mev , similar to refs .
@xcite . for the @xmath151 transitions ,
a linear combination of the three allowed one - body interactions was taken : @xmath152\label{eq4},\end{aligned}\ ] ] where @xmath153 is the effective charge for the @xmath151 transitions and @xmath154 and @xmath155 are model parameters .
the goal of the present paper is to describe simultaneously both the existing electromagnetic and the hadronic ( transfer strength ) data . to achieve this goal , two - neutron transfer intensities between the ground state of the target nucleus and the excited states of the residual nucleus
were also calculated .
the l=0 transfer operator may contain various terms , but we shall restrict our operator to the following form : @xmath156 where @xmath157 is the pair degeneracy of neutron shell , @xmath158 is the number of neutron pairs , @xmath159 is the total number of bosons , and @xmath160 , @xmath161 , and @xmath162 are constant parameters .
the l=0 transfer operator contains two additional terms beside the leading order term , proportional to the bosonic @xmath163 operator @xcite .
details about the contributions of different terms in calculating the ( p , t ) spectroscopic factors will be given in a forthcoming paper @xcite .
the calculations were performed using the computer code octupole @xcite .
the hamiltonian is diagonalized in a hilbert space with a total number of bosons @xmath164 .
for the present calculations we used an extended basis allowing up to three negative parity bosons ( @xmath165=3 ) .
the vibrational strengths used in the calculations are @xmath135=0.2 mev , @xmath136=1.0 mev , and @xmath137=1.1 mev , while the quadrupole - quadrupole interaction strength has a value of @xmath141=-21 kev .
the strength of the o(5 ) second order casimir operator is set to @xmath166=0.053 mev , while the quadrupole operator parameters are ( @xmath167=-1.09 , @xmath168=-1 ) . the full spectrum of excited 0@xmath169 states obtained in the present experiment is displayed in fig .
[ fig10 ] in comparison to the corresponding calculated values . in the energy range
covered experimentally ( up to 2.5 mev ) , the ibm-@xmath130 calculations predict 10 excited 0@xmath169 states in comparison to the 17 experimentally observed 0@xmath169 excitations ( firm spin assignment ) .
given that there was no attempt to fine tune the calculations to the empirical 0@xmath169 states , there is no point in invoking a precise energy cutoff for the ibm calculations .
therefore , it is appropriate to look also above 2.5 mev , where there is a continuous spectrum of 0@xmath169 states consisting of 20 states up to 3.3 mev and as many as 30 up to 4 mev .
the ibm predicts that some of these states are having 2@xmath146 bosons in their structure and are related ( according to ref .
@xcite ) to the presence of double dipole / octupole excitations . for example , the boson admixtures for the first excited @xmath1 state are @xmath170=4.2 , @xmath171=1.4 , @xmath172=0.6 in comparison with those for the @xmath5-vibrational state as @xmath170=4.5 , @xmath171=0.006 , @xmath172=0.001 .
however , the present data can not allow for a final decision on the nature of the 0@xmath169 states .
additional experimental information is needed to measure the branching ratios and the absolute transition probabilities stemming from these states . in fig .
[ fig10 ] , the 2@xmath169 and 4@xmath169 levels revealed in the present experiment are also compared to the predictions of the ibm .
the experiment revealed 33 excited 2@xmath169 and 25 excited 4@xmath169 states up to 2.5 mev . in the same energy range , the calculations produced only 16 excited 2@xmath169 states and 15 4@xmath169 excitations .
if one looks above this limit , the ibm predicts 32 excited 2@xmath169 states and 33 excited 4@xmath169 states up to 3.3 mev .
.[be_ibm ] experimental and calculated b(e1)/b(e2 ) transition ratios in @xmath0th .
the parameters of the @xmath151 operator are fitted to the experimental data available . [ cols="^,^,^,^,^,^,^",options="header " , ] in @xmath0th
there are no lifetimes measured for the negative - parity states , hence no absolute transition probabilities could be extracted .
therefore we would restrict the present discussion to reproducing the b(e1)/b(e2 ) ratios .
a detailed comparison between the experimental data and the present calculations is presented in table [ be_ibm ] .
the agreement is obtained by using @xmath153=0.005 @xmath173fm and @xmath145=0.19 @xmath173b as the effective charges in eq.([eq4 ] ) and ( [ eq3 ] ) , respectively .
the remaining @xmath151 parameters are @xmath174=0.4 and @xmath175=-1.4 .
the b(e1)/b(e2 ) ratios discussed in table [ be_ibm ] belong to the @xmath176=0@xmath177 ( the predicted double - octupole phonon band ) and @xmath176=0@xmath178 ( @xmath5-vibrational ) bands .
the comparison between them is important , because it can be used as a tool for providing additional information about the nature of the @xmath176=0@xmath177 band .
all the states belonging to this band are having 2 @xmath146 bosons in their structure in the ibm calculations and are supposed to have a double - octupole phonon character .
further information confirming this hypothesis comes from the analysis of the @xmath151 and @xmath179 branching ratios .
if the picture proposed by the ibm is correct , the states belonging to the @xmath176=0@xmath177 band will show strong transitions into the negative - parity states ( if they have a double - octupole character ) , while the levels stemming from the @xmath176=0@xmath178 ( @xmath5-band ) will show very weak @xmath151 transitions to these states .
the experimental values in table [ be_ibm ] fully confirm this hypothesis , showing that the b(e1)/b(e2 ) ratios are at least one order of magnitude larger for the @xmath176=0@xmath177 band . in fig .
[ fig11 ] , we display the calculated two - neutron intensities for @xmath0th in comparison to the integrated experimental cross sections normalized to that of the ground state .
the calculations reproduce the strong excitation of the first 0@xmath4 state at 832 kev in good agreement with the experiment .
the experimental spectrum of 0@xmath169 states is dominated also by a single state located at an energy of 2.1 mev , showing a high cross section of about 15@xmath133 of that of the ground state . in the ibm ,
there is predicted a state located at 2.1 mev , which have the transfer intensities of about 18@xmath133 .
this state has a double - octupole phonon structure .
another state at 2.29 mev with a relative cross section of about 7@xmath133 can be put in correspondence to the predicted in the ibm state at 2.2 mev also with a double - octupole phonon structure . the running sum in fig .
[ fig11 ] is taken up to 3.25 mev , where another group of states with significant transfer strength is predicted by the ibm .
the parameters from eq.([eq5 ] ) were estimated from the fit of the known two - neutron transfer intensities ( integrated cross sections ) in table [ tab : expei ] .
the values employed in the present paper are @xmath160=1.3 mb / sr , @xmath161=-0.4 mb / sr , and @xmath162=0.03 mb / sr .
the location and transfer intensity of the strongest states is very well reproduced by the calculations .
because the calculated energy distribution of the 0@xmath169 states is underestimating the experimental data , this also affects the fragmentation of the transferred strength . however , the main characteristics are well reproduced by the present calculations .
( color online ) comparison between the experimental two - neutron transfer intensities ( panel ( a ) ) for the 0@xmath169 states and the ibm predictions ( panel ( b ) ) . in panel
( c ) the experimental versus computed running sum of the ( p , t ) strengths is given.,width=8 ] the ibm is a phenomenological approach . to gain a detailed information on the properties of the states excited in the ( p , t ) reaction , a microscopic approach is necessary .
the ability of the qpm to describe multiple @xmath1 states ( energies , @xmath179 and @xmath180 strengths , two - nucleon spectroscopic factors ) was demonstrated for @xmath181gd @xcite .
an extension of the qpm to describe the @xmath1 states in the actinides @xcite was made after our publication on the results of a preliminary analysis of the experimental data @xcite .
these calculations are used to compare to the present detailed analysis of the experimental data for @xmath0th . as for the theoretical basis of the calculations , we refer to the publications @xcite .
the experimental spectrum of the @xmath182 relative level reaction strength for the ( p , t ) transfer ( the ratios of the ( p , t ) strength for every state to those for the ground state ) is compared to the results of the qpm calculations in fig .
[ fig : compil - zeroqpm ] .
the ( p , t ) normalized transfer spectroscopic strengths in the qpm are expressed also as ratios @xmath183 ^ 2.\ ] ] the amplitude @xmath184 refers to the transitions to the @xmath112 members of the ground - state rotational band , i.e. to the ground state at an analysis of the @xmath1 excitations .
the amplitude @xmath185 includes the transitions between the ground state and the one- and two - phonon components of the wave function .
the numerical results of the calculations obtained according to the qpm investigation in the publication @xcite are provided to us by a. v. sushkov @xcite .
the qpm generates 15 @xmath1 states below 2.5 mev , in fair agreement with the 17 firmly assigned states .
the calculations reproduce the strong excitation of the first 0@xmath4 state in accordance with the experiment . in fig .
[ fig : compil - zeroqpm ] , we present also the increments of the ( p , t ) strength ratios in comparison to those of the qpm calculated normalized spectroscopic strengths
. as one can see , the calculations are in fair agreement with the experiment .
the ( p , t ) strength for the questionable 0@xmath4 state at 2335.9 kev does not influence considerably the results of comparison ( hence it is not included in the comparison ) .
visible deviation of calculated strength from experiment is seen above 2 mev .
theory predicts many @xmath1 excitations at higher energies ( more than 80 states in the energy range below 4 mev ) , but with small strength . at the same time
, two strong excitations are observed in the experiment at 2.13 and 2.29 mev ( see table [ qpm_structure ] ) , respectively .
it is interesting to note that both the ibm and the qpm predict two strongly excited states and therefore a jump in the increments of the ( p , t ) strength in the vicinity of 2 mev , thus reproducing partly the sharp increase in the experimental increment . in these qpm calculations ,
the dominant phonon structure of the 0@xmath4 states in low part of energies is the one - phonon quadrupole nature . for higher energies admixtures of two quadrupole and two octupole phonons
are present in the structure of these states , and for some of the states they became dominant .
the relatively modest role of the octupole phonons in the structure of the low energy @xmath1 states is explained in @xcite by the enforcement of the pauli principle , leading to a spreading of the lowest two octupole phonon components among several qpm @xmath1 states and pushing them to higher energies .
ccccccl @xmath186 & @xmath187 & @xmath188 & @xmath189 & @xmath190 & @xmath191 & structure from @xcite structure from@xcite + + 0@xmath192 & 0.832 & 0.8 & 0.724 & 0.236 & 0.281 & @xmath193 @xmath1940.3 $ ] + 0@xmath195 & 0.939 & 1.0 & 1.496 & 0.041 & 0.002 & @xmath196 @xmath1970.7 $ ] + 0@xmath198 & 1.120 & 1.2 & 1.570 & 0.008 & 0.001 & @xmath199 @xmath200 + 0@xmath201 & 1.511 & 1.4 & 1.831 & 0.013 & 0.021 & @xmath20221 $ ] @xmath203 + 0@xmath204 & 1.532 & & & 0.003 & & + 0@xmath205 & 1.628 & 1.6 & 1.950 & 0.058 & 0.089 & @[email protected] $ ] + 0@xmath208 & 1.691 & 1.7 & 1.962 & 0.008 & 0.046 & @xmath2096$]@xmath2104 $ ] + 0@xmath211 & 1.710 & 1.8 & 2.138 & 0.003 & 0.001 & @xmath2124$]@xmath2130.2 $ ] + 0@xmath214 & 1.750 & & & 0.011 & & + 0@xmath215 & 1.909 & 1.9 & 2.162 & 0.028 & 0.016 & @xmath2164;[(30)_1(30)_1]23$]@xmath217 + 0@xmath218 & 2.045 & & 2.190 & 0.003 & 0.008 & @xmath21987 $ ] + 0@xmath220 & 2.080 & & 2.270 & 0.028 & 0.013 & @xmath22132 $ ] + 0@xmath222 & 2.131 & & 2.290 & 0.150 & 0.008 & @xmath2236;[(31)_1(31)_1]7 $ ] + 0@xmath224 & 2.159 & & 2.334 & 0.007 & 0.001 & @xmath22587 $ ] + 0@xmath226 & 2.290 & & 2.350 & 0.067 & 0.003 & @xmath22765 $ ] + 0@xmath228 & 2.456 & & 2.359 & 0.003 & 0.001 & @xmath2293;[(30)_1(30)_3]3 $ ] + 0@xmath230 & 0.328 & 0.5 & & 0.005 & & @xmath231 + 1@xmath230 & 0.944 & 1.0 & & 0.002 & & @xmath232 + 3@xmath230 & 1.344 & 1.4 & & 0.002 & & @xmath2333 $ ] + 2@xmath192 & 0.969 & 1.0 & & 0.121 & & @xmath234 + 2@xmath195 & 1.153 & 1.3 & & 0.145 & & @xmath235 + 4@xmath192 & 1.432 & 1.5 & & 0.001 & & @xmath236 + @xmath237 data are taken from @xcite .
@xmath238 data are taken from @xcite + crccrccrcccc k@xmath239 & e@xmath240 & i@xmath240 & & e@xmath241 & i@xmath241 & & e@xmath242 & i@xmath242 & & exp .
( 10@xmath243 b@xmath126 ) & qpm ( 10@xmath243 b@xmath126 ) + + 0@xmath169 & 831.8 & 0@xmath169 & & 328.0 & 1@xmath244 & & 57.8 & 2@xmath169 & & 5.1(4 ) & 2.25 + 0@xmath169 & 938.6 & 0@xmath169 & & 328.0 & 1@xmath244 & & 57.8 & 2@xmath169 & & 6.7(6 ) & 54.8 + 0@xmath169 & 1175.5 & 2@xmath169 & & 328.0 & 1@xmath244 & & 186.8 & 4@xmath169 & & 0.06(3 ) & 1.45 + besides the publication @xcite , other calculations of the same dubna group were carried out for a microscopic description of the level structure , and transition rates between excited states in @xmath0th , observed in the decay of @xmath0pa @xcite .
the wave functions , the level energies from two publication in correspondence to the experimental ones and to the transfer factors are given in table [ qpm_structure ] .
the transfer factors and also the moments of inertia are taken into account to put in correspondence the experimental and calculated levels . the large difference in the transfer factors
is seen only for two levels at 2131 and 2290 kev .
there is an essential difference in the energies of the lowest @xmath1 states in two publications , as they are more close to the experimental ones in @xcite .
this is caused by the choice of the isoscalar quadrupole - quadrupole interaction strength stronger than the critical value in ref .
@xcite . as
a consequence the energy of the lowest collective @xmath1 state becomes imaginary and its properties such as the structure , e2 reduced transition probabilities and the transfer factor are partially transferred to the next @xmath1 collective state . among the transition properties , the calculated b(e1)/b(e2 ) ratios at the decay of some 0@xmath4 states are of special interest in this publication .
we present here some data on the b(e1)/b(e2 ) ratios , in order to note the difference in the explanation of the experimental data by the qpm and the ibm .
the calculated ratios for transitions from the @xmath245 , @xmath246 , and @xmath247 states to the octupole @xmath248 state and the ground state band are compared to experimental ratios in table [ be_qpm ] .
as one can see from table [ qpm_structure ] and table [ be_qpm ] , the small admixtures of the octupole two - phonon components to the wave functions of the @xmath245 and @xmath246 states are responsible for the fast @xmath151 transitions . in the ibm ,
similar results are obtained for the @xmath245 state , having 2 @xmath146 bosons in their structure and which are supposed to have mainly a double - octupole phonon character ( see table [ be_ibm ] and corresponding discussion ) . at the same time , the b(e1)/b(e2 ) ratio is considerably smaller for the decay of the @xmath247 state , which is a @xmath5-vibrational state , again in agreement with the experiment .
the same result is obtained in the ibm . generally , the qpm is quite accurate in nuclei with small ground - state correlations .
these correlations increase with the collectivity of the first one - phonon states , which is exactly the case of the @xmath249 octupole phonon state in @xmath0th . to decrease the correlations the value of the octupole - octupole isoscalar interaction strength was diminished so that the calculated energy of the @xmath249 state was almost 200 kev higher than the experimental value @xcite .
in addition , the effect of multi - phonon admixtures ( three and more phonons ) that pushes two - octupole phonon poles and consequently two - octupole phonon energies to lower values was then underestimated . in summary , the accuracy of the calculations of @xmath0th as stated in ref .
@xcite is worse due to the increased ground - state correlations and the shift of two - phonon poles towards smaller energies . in future qpm studies
one also has to take into account the spin - quadrupole interaction that is known to increase the density of low - lying @xmath1 states @xcite . at a microscopic approach
there can be a few situations of structure of the @xmath182 states .
a @xmath5-vibrational mode can be characterized by the relatively small two - nucleon transfer strength and a relatively large b(e2 ) value with a moment of inertia close to the one of the ground state .
the large ratio b(e1)/b(e2 ) and the increase of the moment of inertia indicate the presence of the octupole two - phonon component .
if a state has a relatively weak b(e2 ) value and also a weak two - nucleon transfer strength , but exhibits an increase of the moment of inertia , it should be a state with one dominant 2qp configuration .
the pairing vibrational excitations can be characterized by their large two - nucleon transfer strengths and relatively small b(e2 ) values .
it is clear that the first and second @xmath1 excited states can not be the @xmath5-vibrational states as usually is observed in deformed rare - earth nuclei since their moments of inertia are much larger than those of the ground state and also their ( p , t ) strengths are large . the actual @xmath5-vibrational state is observed at 1120 kev and it is excited very weakly in the ( p , t ) reaction .
as we have seen , both the ibm and the qpm reproduce the @xmath182 relative level reaction strength for the ( p , t ) reaction and the b(e1)/b(e2 ) ratios of the decay of the lowest @xmath1 states reasonably well . at the same time
, the nature of @xmath1 excitations in the qpm differs significantly from the one in the _ spdf_-ibm . in all low - lying states of the qpm calculations ,
quadrupole phonons are dominant and the octupole phonons are predicted to play a relatively modest role , whereas the ibm predicts the lowest @xmath1 state as having mainly 2@xmath146 bosons in their structure @xcite .
the analysis of the lowest quadrupole phonon wave function in the qpm reveals that the backward rpa amplitudes @xmath250 contribute considerably to the relative ( p , t ) reaction strength thus indicating that the lowest excited @xmath1 state describe pairing vibrations arising from ground state fluctuations @xcite . for an additional hint to the nature of the lowest @xmath1 states ,
we include fig .
[ fig : b(e1)-b(e2 ) ] with the b(e1)/b(e2 ) ratio stemming from the lowest excited states in @xmath0th and @xmath7th .
intuitively , one would expect that a large b(e1)/b(e2 ) ratio might be characteristic for a two - octupole - phonon excitation , whereas a small ratio might indicate a shape oscillation .
such a picture is observed for @xmath0th : large b(e1)/b(e2 ) ratios for the @xmath245 and @xmath246 states , and vanishing values for @xmath5- and @xmath11-vibrational states .
the assumption in @xcite that just the two - phonon structure is a reason of the very strong excitation of the @xmath245 state in the ( p , t ) reaction has to be rejected , since the b(e1)/b(e2 ) ratio for the @xmath245 state in @xmath7th is small at strong excitation in the ( p , t ) reaction ( or the b(e1)/b(e2 ) ratio is no reliable manifestation of the two - phonon nature of a state ) . .
in this aspect a comparison of the relative transfer strengths in the ( p , t ) reaction leading to the @xmath1 states in @xmath0th and the @xmath251 states in @xmath3pa ( data are taken from @xcite ) has to be considered additionally ( see fig . [
fig : b(e1)-b(e2 ) ] ) .
since the ground - state spin in the target nucleus @xmath252pa is @xmath251 , just these spins are excited in a two - neutron @xmath253 transfer .
@xmath3pa can be regarded as @xmath0th plus a strongly coupled proton . the rotational band built on the first @xmath251 excited state in @xmath3pa at 11.6 kev @xcite corresponds to the g.s .
band in @xmath0th and is excited very strongly .
the main component in the structure of the ground state in the target nucleus @xmath252pa is 1/2[530 ] @xcite .
therefore the levels excited in the @xmath253 transfer are @xmath254 states and they are members of collective bands based on the state originating from coupling the @xmath255 state to the @xmath0th core - excited states . such bands , as identified in @xcite , can be used to derive the moments of inertia for at least three @xmath251 states .
values of @xmath256 in mev@xmath126 are given below ( energies in @xmath3pa are relative to the lowest @xmath251 state ) + @xmath3pa : * 79.2*(0.0 kev ) , * 78.4*(703 kev ) , * 127.5*(830 kev ) + ( 1524 kev ) .
+ @xmath0th : * 51.9*(0.0 kev ) , * 70.3*(832 kev ) , * 73.2*(939 kev ) . + the moments of inertia in @xmath3pa are larger than those in @xmath0th which is a manifestation of the contribution of the odd proton
. the large moment of inertia for the 830.5 kev state can probably be explained ( at least partly ) by neglecting the coriolis coupling when fitting the energies of the corresponding band .
nevertheless the increment of the moment of inertia for the state at 830.5 kev relative to other states in @xmath3pa can be put in correspondence to similar increments for the states at 831.8 and 938.6 kev relative to the g.s . in @xmath0th . but this state , as well as other low - lying states in @xmath3pa , are only weakly excited in contrast to the strong excitation of the state 831.8 kev in @xmath0th . at the same time
, the @xmath251 state at 1523.7 kev is excited strongly , but it can not be put in correspondence to the first excited state in @xmath0th : there is practically no increment of its moment of inertia relative to the lowest 3/2@xmath95 state .
besides that , its energy is almost twice larger than the first excited @xmath1 state in @xmath0th . from the insufficient information
we can only conclude that the 831.8 kev state in @xmath0th has the largest pairing vibrational component and in @xmath3pa the additional proton has the effect that the largest pairing vibrational component is moved to the 1523.7 kev state .
no theoretical explanation was undertaken since the publication of the experimental results on @xmath3pa @xcite .
it would be interesting to undertake the theoretical analysis of the excitations with the @xmath6 transfer in odd nuclei and first of all in the @xmath3pa nucleus .
experimental study of excitations in other odd nuclei , as we have seen , may promise unexpected phenomena .
as for the experimental evidence of the nature of other @xmath1 states , we have only the moments of inertia derived from the sequences of states treated as rotational bands and thus only tentative conclusions can be drawn about their structure .
in contrast to @xmath7th @xcite , for which they are distributed almost uniformly over the region from 47 to 98 mev@xmath126 , the moments of inertia in @xmath0th have values close to 50 mev@xmath126 only for the g.s . , @xmath5-vibrational @xmath1 states and for the state at 1531.7 kev , all other @xmath1 states have larger values from 70 to 95 mev@xmath126 .
this fact can indicate that corresponding states are possibly of two - phonon nature too , or two quasi - particle states with an admixture of the pairing vibrations .
excited states in @xmath0th have been studied in ( p , t ) transfer reactions .
106 levels were assigned , using a dwba fit procedure , additionally only the energies are determined for 57 states . among them , 17 excited @xmath1 states have been found in this nucleus up to an energy 2.5 mev , most of them have not been experimentally observed before .
their accumulated strength makes up for more than 70% of the ground - state strength .
firm assignments have been made for most of the 2@xmath4 and 4@xmath4 states and for some of the 6@xmath4 states .
these assignments allowed to suggest multiplets of states , which can be treated as one- and two - phonon octupole quadruplets , and to identify the sequences of states , which have the features of rotational bands with definite inertial parameters . moments of inertia are derived from these sequences . only for the g.s . and @xmath5-vibrational states and additionally for the state at 1531.7 kev ( for which the shape of the angular distribution is different from most other ones ) , the moments of inertia are about 50 mev@xmath126 . for all other states they are larger than 70 mev@xmath126 , i.e. the value for the first excited @xmath1 state .
this information , together with the spectroscopic information on some @xmath257-transitions , were used for conclusions on the nature of the @xmath1 states .
the experimental data have been compared to _ spdf_-ibm and qpm calculations .
spectroscopic factors from the ( p , t ) reaction , and the trend in their change with excitation energy , are approximately reproduced by both the ibm and qpm for the 0@xmath4 states .
a remarkable feature of the ibm and qpm is the prediction of strong first vibrational excitations , close in magnitude and position to the experimental ones . giving also an approximately correct number of @xmath1 states ,
these models provide different predictions for the structure of these states .
the lack of additional information does not allow for final conclusions on the validity of the theoretical approaches . challenging experiments on gamma spectroscopy following
( p , t ) reactions would give much needed information .
the work was supported by the dfg ( c4-gr894/2 - 3 , gu179/3 , jo391/2 - 3 , zl510/4 - 2 ) , mll , and us - doe , contract number de - fg02 - 91er-40609 , and cz.1.05/2.1.00/03.0082 .
wirth , g. graw , s. christen , d. cutoiu , y. eisermann , c. gnther , r. hertenberger , j. jolie , a. i. levon , o. mller , g. thiamova , p. thirolf , d. tonev , and n. v. zamfir , phys .
c * 69 * , 044310 ( 2004 ) .
p. schuler , ch .
lauterbach , y. k. agarwal , j. de boer , k. p. blume , p. a. butler , k. euler , ch .
fleischmann , c. gnther , e. hauber , h. j. maier , m. marten - tolle , ch .
schandera , r. s. simon , r. tolle , and p. zeyen , phys .
b174 * , 241 ( 1986 ) .
f. d. becchetti and g. w. greenlees , proc .
third int . symp . on polarization phenomena in nuclear reactions ,
medison , 1970 , ed .
h. h. barshall and w. haeberli ( university of wisconsin press , medison , 1971 ) p.68 g. graw , y. eisermann , r. hertenberger , h .- f.wirth , s. christen , o. moeller , d. tonev , j. jolie , c. guenther , a. i. levon , and n. v. zamfir , int .
conf . on nucl .
reaction mechanisms , varenna , italy , june 9 - 13 , 2003 . | the excitation spectra in the deformed nucleus @xmath0th have been studied by means of the ( p , t ) reaction , using the q3d spectrograph facility at the munich tandem accelerator .
the angular distributions of tritons were measured for about 110 excitations seen in the triton spectra up to 2.5 mev .
firm @xmath1 assignments are made for 17 excited states by comparison of experimental angular distributions with the calculated ones using the chuck3 code .
assignments up to spin @xmath2 are made for other states .
sequences of states are selected which can be treated as rotational bands and as multiplets of excitations .
moments of inertia have been derived from these sequences , whose values may be considered as evidence of the two - phonon nature of most @xmath1 excitations .
experimental data are compared with interacting boson model ( ibm ) and quasiparticle - phonon model ( qpm ) calculations and with experimental data for @xmath3pa . |
statistical inference is of fundamental importance to science .
inference enables the testing of theoretical models against observations , and provides a rational means of quantifying uncertainty in existing models .
modern approaches to statistical inference , based on monte carlo sampling techniques , provide insight into many complex phenomena @xcite . inference can be described as follows : suppose we have a set of observations , @xmath0 ; a method of determining the likelihood of these observations , @xmath1 , under the assumption of some model characterized by parameter vector , @xmath2 ; and a prior probability density , @xmath3 .
the posterior probability density , @xmath4 , can be computed using bayes theorem , @xmath5 explicit expressions for likelihood functions are rarely available @xcite ; motivating the development of likelihood - free methods , such as approximate bayesian computation ( abc ) @xcite .
abc methods approximate the likelihood through evaluating the discrepancy between data generated by a simulation of the model of interest and the observations , yielding an approximate form of bayes theorem , @xmath6 here , @xmath7 is data generated by the model simulation process , @xmath8 , @xmath9 is a discrepancy metric , and @xmath10 is the acceptance threshold . due to this approximation ,
monte carlo estimators based on equation ( [ eqn : approxbayes ] ) are biased @xcite .
the most simple implementation of abc is abc rejection @xcite , see algorithm [ alg : abc - rej ] .
[ line : sample1]sample prior , @xmath11 .
generate data , @xmath12 .
set @xmath13 .
[ alg : abc - rej ] this method generates @xmath14 samples @xmath15 by accepting proposals , @xmath11 , when the data generated by the model simulation process @xmath16 is within @xmath17 of the observed data , @xmath0 .
while abc rejection is simple to implement , it can be computationally prohibitive in practice @xcite . to improve the efficiency of abc
, one can consider a likelihood - free modification of markov chain monte carlo ( mcmc ) @xcite in which a markov chain is constructed with a stationary distribution identical to the desired posterior .
given the markov chain is in state @xmath18 , a state transition is proposed via a proposal kernel , @xmath19 . the classical metropolis - hastings @xcite state transition probability , @xmath20 , can be modified within an abc framework to yield @xmath21 if @xmath22 , or @xmath23 otherwise @xcite .
the stationary distribution of such a markov chain is the desired approximate posterior , @xmath24 @xcite .
algorithm [ alg : mcmc - abc ] provides a method for computing @xmath25 iterations of this markov chain .
sample transition kernel , @xmath26 .
generate data , @xmath27 .
set @xmath28 . sample uniform distribution , @xmath29 .
set @xmath13 .
set @xmath30 .
set @xmath30 .
[ alg : mcmc - abc ] while mcmc - abc sampling can be highly efficient @xcite , the samples in the sequence , @xmath31 , are not independent .
this can be problematic as it is possible for the markov chain to take long excursions into regions of low posterior probability ; thus incurring additional , and potentially significant , bias @xcite .
a poor choice in the proposal kernel can also have considerable impact upon the efficiency of mcmc - abc @xcite .
the question of how to choose the proposal kernel is non - trivial , and typically determined heuristically .
sequential monte carlo ( smc ) sampling was introduced to address these potential inefficiencies @xcite and later extended within an abc context @xcite .
a set of samples , referred to as particles , is evolved through a sequence of abc posteriors defined through a sequence of @xmath32 acceptance thresholds , @xmath33 @xcite . at each step ,
$ ] , the current abc posterior , @xmath35 , is approximated by a discrete distribution constructed from a set of @xmath36 particles @xmath37 with importance weights , @xmath38 , that is , @xmath39 for @xmath40 .
the collection is updated to represent the abc posterior of the next time step , @xmath41 , through application of rejection sampling on particles perturbed by a proposal kernel , @xmath42 .
if @xmath35 is similar to @xmath41 , the acceptance rate should be high .
the importance weights of the new family of particles are updated using an optimal backwards kernel @xcite .
algorithm [ alg : abc - smc ] outlines the process .
initialize @xmath43 and @xmath44 , for @xmath45 .
set @xmath46 with probability @xmath47 .
sample transition kernel , @xmath48 .
generate data , @xmath49 .
set @xmath50 .
set @xmath51 .
normalize weights so that @xmath52 .
[ alg : abc - smc ] through the use of independent weighted particles , smc - abc avoids long excursions into the distribution tails that are possible with mcmc - abc .
however , the efficiency of each step still depends on the non - trivial problem of selecting a proposal kernel @xcite .
multilevel monte carlo ( mlmc ) is a recent development that can significantly reduce the computational burden in the estimation of expectations @xcite . to demonstrate the basic idea of mlmc ,
consider computing the expectation of a continuous - time stochastic process @xmath53 at time @xmath32 .
let @xmath54 denote a discrete - time approximation to @xmath53 with time step @xmath55 : the expectations of @xmath56 and @xmath57 are related according to @xmath58}~=~{\mathbb{e}\left[z_t^\tau\right]}~+~{\mathbb{e}\left[x_t - z_t^\tau\right]}$ ] .
that is , an estimate of @xmath59}$ ] can be treated as a biased estimate of @xmath60}$ ] . by taking a sequence of time steps @xmath61 ,
the indices of which are referred to as _ levels _ , we can arrive at a telescoping sum , @xmath62 } = { \mathbb{e}\left[z_t^{\tau_1}\right ] } + \sum_{\ell = 2}^l { \mathbb{e}\left[z_t^{\tau_\ell } - z_t^{\tau_{\ell-1}}\right]}.\ ] ] while computing this form of the expectation returns the same bias as when computing @xmath63}$ ] directly , giles @xcite demonstrates that a monte carlo estimator for the telescoping sum can be computed more efficiently than directly estimating @xmath63}$ ] in the context of stochastic differential equations ( sdes ) .
this efficiency comes from exploiting the fact that the bias correction terms , @xmath64}$ ] , only measure the expected difference between levels @xmath65 and @xmath66 , hence samples paths need not be independent . in the case of sdes , samples may be generated in pairs driven by the same underlying brownian motion , this introduces a strong coupling that reduces the variance of the monte carlo estimator .
recently , the concept of mlmc @xcite has been extended to statistical applications with known likelihoods @xcite .
our work considers mlmc in the context of bayesian inference with intractable likelihoods .
we describe a new algorithm for likelihood - free inference , based on mlmc , that is as general as mcmc - abc and smc - abc . here
we describe the algorithm , its implementation , and we compare the performance of the new method against mcmc - abc and smc - abc for a standard benchmark problem .
the key attraction of our method is that its performance does not depend on any user - specified proposal kernel .
we show that the new method significantly outperforms mcmc - abc and smc - abc in the most practical scenario where parameter proposal kernels are independent .
this is important because choosing an optimal proposal density is non - trivial .
therefore , developing new methods that avoid the need for specifying a proposal density kernel can lead to significant computational savings for practical applications .
in this section , we demonstrate our application of mlmc ideas to the likelihood - free inference problem given in equation ( [ eqn : approxbayes ] ) .
the aim is to compute an accurate approximation to the joint posterior cumulative distribution function ( cdf ) of @xmath67 using as few data generation steps as possible .
we define a data generation step to be a simulation of the model of interest given a proposed parameter vector .
we apply mlmc methods to likelihood - free bayesian inference by reposing the problem of computing the posterior cdf as an expectation calculation .
this allows the mlmc telescoping sum idea , as in equation ( [ eqn : mlmc ] ) , to be applied . in this context
, the levels of the mlmc estimator are parameterized by a sequence of @xmath68 acceptance thresholds @xmath69 with @xmath70 for all @xmath71 .
the efficiency of mlmc relies upon computing the terms of the telescoping sum with low variance .
variance reduction is achieved through exploiting features of the telescoping sum for cdf approximation , and further computational gains are achieved by using properties of nested abc rejection samplers .
we first reformulate the bayesian inference problem ( equation ( [ eqn : bayes ] ) ) as an expectation . to this end , note that , given a @xmath72-dimensional parameter space , @xmath73 , and the parameter random vector , @xmath67 , the joint posterior cdf , @xmath74 , is defined as @xmath75 .
if @xmath74 is differentiable , that is , its probability density function ( pdf ) exists , then @xmath76 where @xmath77 .
this can be expressed as an expectation by noting @xmath78},\end{aligned}\ ] ] where @xmath79 is the indicator function : @xmath80 if @xmath81 and @xmath82 otherwise .
now , consider the abc approximation given in equation ( [ eqn : approxbayes ] ) with discrepancy metric @xmath83 and acceptance threshold @xmath17 .
the abc posterior cdf , denoted by @xmath84 , will be @xmath85}.\end{aligned}\ ] ] the marginal abc posterior cdfs , @xmath86 , are @xmath87 .
we now introduce some notation that will simplify further derivations .
we define @xmath88 to be a random vector distributed according to the abc posterior cdf @xmath84 with acceptance threshold @xmath17 as given in equation ( [ eqn : cdf ] ) .
this provides us with simplification in notation for the abc posterior pdf , @xmath89 , and the conditional expectation , @xmath90 } = { \mathbb{e}\left[{\mathds{1}_{a_{\mathbf{s}}}({\boldsymbol{\theta } } ) } \mid { d ( { \mathcal{d } } , { \mathcal{d}_s } ) } < \epsilon\right]}$ ] .
we also use @xmath91 to denote the monte carlo estimator of the expectation , @xmath92 , obtained using @xmath14 samples .
the standard monte carlo integration approach is to generate @xmath14 samples @xmath93 from the abc posterior , @xmath94 , then evaluate the empirical cdf ( ecdf ) , @xmath95 for @xmath96 , where @xmath97 a discretization of the parameter space @xmath73 . for simplicity , we will consider @xmath97 to be a @xmath72-dimensional regular lattice .
the ecdf is not , however , the only monte carlo approximation to the cdf one may consider . in particular , giles et al .
@xcite demonstrate the application of mlmc to a univariate cdf approximation .
we now present a multivariate equivalent of the mlmc cdf of giles et al .
@xcite in the context of abc posterior cdf estimation . given a strictly decreasing sequence of @xmath68 acceptance thresholds , @xmath98
, we can represent the cdf ( equation ( [ eqn : cdf ] ) ) using the telescoping sum @xmath99 } = \sum_{\ell=1}^{l } y_\ell({\mathbf{s}}),\ ] ] where @xmath100 } & \text{if } \ell = 1 , \\ { \mathbb{e}\left[{\mathds{1}_{a_{\mathbf{s}}}({\boldsymbol{\theta}}_{\epsilon_\ell } ) } - { \mathds{1}_{a_{\mathbf{s}}}({\boldsymbol{\theta}}_{\epsilon_{\ell -1}})}\right ] } & \text{if } \ell > 1 .
\end{cases}\ ] ] using our notation , the mlmc estimator for equation ( [ eqn : mlmccdf ] ) and equation ( [ eqn : mlmccdfterms ] ) is given by @xmath101 where @xmath102 & \text{if } \ell > 1 , \end{cases}\ ] ] and @xmath103 is a lipschitz continuous approximation to the indicator function ; this approximation is computed using a tensor product of cubic splines .
such a smoothing is necessary to avoid convergence issues with mlmc caused by the discontinuity of the indicator function @xcite . to compute the @xmath104 term ( equation ( [ eqn : bmcest ] ) )
, we generate @xmath105 samples @xmath106 from @xmath107 ; this represents a biased estimate for @xmath108 . to compensate for this bias ,
correction terms ( equation ( [ eqn : bmcest ] ) ) , @xmath109 , are evaluated for @xmath110 , each requiring the generation of @xmath111 samples @xmath112 from @xmath113 and @xmath111 samples @xmath114 from @xmath115 .
the goal is to introduce a coupling between levels that controls the variance of the bias correction terms . with an effective coupling ,
the result is an estimator with lower variance , hence the number of samples required to obtain an accurate estimate is reduced .
denote @xmath116 as the variance of the estimator @xmath117 . for @xmath118
this can be expressed as @xmath119 } \\
= & \ , { \text{var}\left[g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_\ell})\right ] } + { \text{var}\left[g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_{\ell-1}})\right ] } - { \text{cov}\left[g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_\ell}),g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_{\ell-1}})\right]}.\end{aligned}\ ] ] introducing a positive correlation between the random variables @xmath120 and @xmath121 will have the desired effect of reducing the variance of @xmath109 . in many applications of mlmc , a positive correlation is introduced through driving samplers at both the @xmath65 and @xmath122 level with the same _ randomness_. properties of brownian motion or poisson processes are typically used for the estimation of expectations involving sdes or markov processes @xcite . in the context of abc methods , however , simulation of the quantity of interest is necessarily based on rejection sampling .
the reliance on rejection sampling makes a strong coupling , in the true sense of mlmc , a difficult , if not impossible task .
rather , here we introduce a weaker form of coupling through exploiting the fact that our mlmc estimator is performing the task of computing an estimate of the abc posterior cdf .
we combine this with a property of nested abc rejection samplers to arrive at an efficient algorithm for computing @xmath123 .
we proceed to establish a correlation between levels as follows .
assume we have computed , for some @xmath124 , the terms @xmath125 in equation ( [ eqn : mlmcest ] ) .
that is , we have an estimator to the cdf at level @xmath65 by taking the sum @xmath126 with marginal distributions @xmath127 for @xmath128 .
we can use this to determine a coupling based on matching marginal probabilities when computing @xmath129 . after generating @xmath130 samples @xmath131 from @xmath132 , we compute the ecdf @xmath133 using equation ( [ eqn : ecdf ] ) and obtain the marginal distributions @xmath134 for @xmath135 .
we can thus generate @xmath130 coupled pairs @xmath136 by choosing the @xmath137 with the same marginal probabilities as the empirical probability of @xmath138 .
that is , the @xmath139th component of @xmath137 is given by @xmath140 , where @xmath141 is the @xmath139th component of @xmath138 and @xmath142 is the inverse of @xmath139th marginal distribution of @xmath143 .
this introduces a positive correlation between the sample pairs , @xmath144 , since an increase in any of the components of @xmath145 will cause an increase in the same component @xmath146 .
this correlation reduces the variance in the bias correction estimator @xmath147 computed according to equation ( [ eqn : bmcest ] ) .
we can then update the mlmc cdf using @xmath148 and apply an adjustment that ensures monotonicity .
we continue this process iteratively to obtain @xmath149 .
initialize @xmath150 , @xmath151 and prior @xmath3 .
sample @xmath152 restricted to @xmath153 .
generate data , @xmath12 .
set @xmath154 . set @xmath155 .
set @xmath156 . set @xmath157/n_\ell$ ] .
set @xmath158 .
[ alg : mlmc - abc ] by computing the bias correction terms in an iterative fashion , we can exploit a further computational gain .
let @xmath159 denote the support of a function @xmath160 , and note that , for any @xmath161 $ ] , @xmath162 .
this follows from the fact that if @xmath163 , @xmath164 since @xmath165 .
therefore , we can truncate the prior to the support of @xmath115 when computing @xmath109 , thus increasing the acceptance rate of level @xmath65 samples . in practice , we approximate the support by the smallest bounding box that contains all generated samples .
we now require the sample numbers @xmath166 that are the optimal trade - off between accuracy and efficiency .
denote @xmath167 as the number of data generation steps required during the computation of @xmath109 and let @xmath168 be the average number of data generation steps per accepted abc posterior sample using acceptance threshold @xmath169 .
given @xmath170}$ ] and @xmath171}$ ] , for @xmath172 , one can construct the optimal @xmath111 using a lagrange multiplier method under the constraint @xmath173}~=~\mathcal{o}(h^2)$ ] , where @xmath174 is the target variance of the mlmc cdf estimator . with details left to appendix [ app : optn ] ,
the optimal @xmath175 are given by @xmath176 in practice , the values for @xmath177 and @xmath178 will not have analytic expressions available ; rather , we perform a low accuracy trial simulation with all @xmath179 , for some comparatively small constant , @xmath180 , to obtain the relative scaling of variances and data generation requirements .
once @xmath175 have been estimated , the computation of the mlmc - abc posterior cdf @xmath123 proceeds according to algorithm [ alg : mlmc - abc ] .
we evaluate the performance of mlmc - abc using the model developed by tanaka et al .
@xcite in the study of tuberculosis transmission rate parameters using dna fingerprint data @xcite .
this model has been selected due to the availability of published comparative performance evaluations of mcmc - abc and smc - abc @xcite . the model proposed by tanaka et al .
@xcite describes the occurrence of tuberculosis infections over time and the mutation of the bacterium responsible , _
myobacterium tuberculosis_. the number of infections , @xmath181 , at time @xmath182 is @xmath183 , where @xmath184 is the number of distinct genotypes and @xmath185 is the number of infections caused by the @xmath186th genotype at time @xmath182 . for each genotype
, new infections occur with rate @xmath187 , infections terminate with rate @xmath188 , and mutation occurs with rate @xmath189 ; causing an increase in the number of genotypes .
this process can be described by a discrete - state continuous - time markov process and simulated using the gillespie algorithm @xcite . after a realization of the model is completed , either by extinction or when a maximum infection count is reached , a sub - sample of @xmath190 cases is collected and compared against the is@xmath191 dna fingerprint data of tuberculosis bacteria samples @xcite .
the dataset consists of @xmath192 distinct genotypes ; the infection cases are clustered according to the genotype responsible for the infection .
the collection of clusters can summarized succinctly as @xmath193 where @xmath194 states there are @xmath139 clusters of size @xmath195 .
the discrepancy metric used is @xmath196 where @xmath195 is the size of the population sub - sample ( e.g. , @xmath197 ) , @xmath198 denotes the number of distinct genotypes in the dataset ( e.g. , @xmath199 ) and the genetic diversity is @xmath200/n^2 $ ] , where @xmath201 is the cluster size of the @xmath186th genotype in the dataset .
we perform likelihood - free inference on the tuberculosis model for the parameters @xmath202 with the goal of evaluating the efficiency of mlmc - abc , mcmc - abc and smc - abc .
we use a target posterior distribution of @xmath24 with @xmath83 as defined in equation ( [ eqn : tb_disc ] ) and @xmath203 .
the improper prior is @xmath204 , @xmath205 and @xmath206 @xcite . for the mcmc - abc and smc - abc algorithms
we apply a typical gaussian proposal kernel , @xmath207 , with covariance matrix @xmath208 such a proposal kernel is reasonable to characterize the initial explorations of an abc posterior as no correlations between parameters are assumed . , ( b ) @xmath188 and ( c ) @xmath189for the tuberculosis transmission stochastic model .
estimate computed using using mlmc - abc with @xmath209 ( yellow lines ) , mcmc - abc over @xmath210 iterations ( blue lines ) , smc - abc with @xmath211 particles ( red lines ) and high precision solution ( dashed lines ) . ]
.comparison of mlmc - abc against mcmc - abc and smc - abc using a naive proposal kernel . [ cols="^,^,^,^,^,^,^,^,^ " , ]
our results indicate that while smc - abc and mcmc - abc can be heuristically optimized to be highly efficient , an accurate estimation of the parameter posterior can be obtained using our mlmc - abc in a reasonably automatic fashion .
furthermore , the efficiency of mlmc - abc is comparable or improved over mcmc - abc and smc - abc , even in the case of good proposal densities that have been heuristically determined .
mlmc - abc shares some concepts with smc - abc .
for example , both methods consider a sequence of abc posteriors .
furthermore , both use information gained at each step to improve the acceptance rates of the next step while maintaining independence in the set of samples .
mlmc - abc and smc - abc are still , however , distinct methods .
for example , in smc - abc the prior distribution is only sampled at the initial step , whereas in mlmc - abc the prior must be sampled at all levels to ensure the telescoping sum does not introduce bias .
another distinct difference is that mlmc does not evolve particles through the abc posterior sequence , rather each abc posterior is sampled independently to compute the bias correction terms .
the need to estimate the variances of each bias correction term could be considered a limitation of the mlmc - abc approach .
however , we find in practice that these need not be computed to high accuracy , since the relative variance between the levels is the main requirement and this can often be estimated with a relatively small number of samples . there could be examples of bayesian inference problems where mlmc - abc is inefficient on account of this fact , in which case the variance estimation would have to be considered as a more heuristic process .
we have so far , however , failed to find an example for which @xmath212 samples of each bias correction term is insufficient to obtain a good mlmc - abc estimator .
there are many modifications one could consider to further improve mlmc - abc .
currently , we depend on the sequence of acceptance thresholds to be specified in advance and we acknowledge that , in practice , many smc - abc implementations determine these adaptively ; modification of mlmc - abc to allow for adaptive acceptance thresholds would make mlmc - abc even more practical .
other improvements could focus on the discretization used for the ecdf calculations , for example , removing the requirement of a regular lattice would enable mlmc - abc to scale to much higher dimensional parameter spaces .
coupling strategies are also possible improvement areas , our current coupling depends only on the computation of the full posterior cdf and assumes nothing about the underlying model ; can model specific features improve the coupling and obtain further variance reduction ?
the rejection sampling method certainly makes this difficult , but there may still be opportunities for improvement .
we have shown , in a practical way , how mlmc techniques can be applied to abc inference , and demonstrated that superior performance over modern advanced abc methods can be achieved .
furthermore , the performance of our mlmc - abc method is not dependent on user specified functions such as proposal kernels .
this work is supported by the australian research council ( ft130100148 ) .
computational resources were provided by the high performance computing and research support group , qut .
in this appendix , we demonstrate how the sequence of sample numbers @xmath166 is obtained , as stated in the main text .
first , we assume the variances @xmath213}$ ] and @xmath214}$ ] , for @xmath215 , are known quantities .
we also assume that , on average , @xmath216 data generation steps are required to compute the estimator @xmath217 .
the sequence @xmath166 is considered optimal if @xmath218 is minimised subject to the constraint @xmath222 for some constant @xmath223 and target monte carlo error @xmath174 . to determine optimal @xmath175 , we consider the lagrangian @xmath224 and note that solutions to the optimisation problem exist at @xmath225 .
that is , @xmath226{\mathbf{e}}_\ell + \left[e(n_1,\ldots , n_l ) - kh^2\right]{\mathbf{e}}_{l+1 } & = { \mathbf{0}},\end{aligned}\ ] ] where @xmath227 are the standard orthonormal basis vectors of @xmath228-dimensional euclidean space .
we obtain the following system of equations , @xmath229 first we consider the forms of @xmath218 and @xmath221 . by definition of the mlmc estimator given in main text
, we have @xmath230 } = { \text{var}\left[\sum_{\ell=1}^{l}\hat{y}^{n_\ell}_\ell({\mathbf{s}})\right]}. \end{aligned}\ ] ] furthermore , since @xmath231 are independent , @xmath232 } = & \sum_{\ell=1}^{l}{\text{var}\left[\hat{y}^{n_\ell}_\ell({\mathbf{s}})\right ] } \\ = & { \text{var}\left[\frac{1}{n_1}\sum_{i=1}^{n_1}g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_1}^{i})\right ] } + \sum_{\ell=2}^{l } { \text{var}\left[\frac{1}{n_\ell}\sum_{i=1}^{n_\ell } g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_\ell}^{i } ) -g_{\mathbf{s}}({\boldsymbol{\theta}}_{\epsilon_{\ell-1}}^{i})\right ] } \\ = & \sum_{\ell=1}^{l } \frac{1}{n_\ell^2 } \sum_{i=1}^{n_\ell}v_\ell.\end{aligned}\ ] ] that is , @xmath233 the total number of data generation steps is simply @xmath234 substitution of equation ( [ eqn : const3 ] ) and equation ( [ eqn : const4 ] ) into equation ( [ eqn : const1 ] ) yields @xmath235 substitution of equation ( [ eqn : const3 ] ) and equation ( [ eqn : const5 ] ) into equation ( [ eqn : const2 ] ) allows us to obtain @xmath236 , @xmath237 finally , through substitution of equation ( [ eqn : lambda ] ) back into equation ( [ eqn : const5 ] ) we find that the optimal @xmath238 is given by , @xmath239 as required .
dodwell tj , ketelsen c , scheichl r , teckentrup al ( 2015 ) a hierarchical multilevel markov chain monte carlo algorithm with applications to uncertainty quantification in subsurface flow .
3(1):10751108 .
small pm , et al .
( 1994 ) the epidemiology of tuberculosis in san francisco a population - based study using conventional and molecular methods .
gillespie dt ( 1977 ) exact stochastic simulation of coupled chemical reactions . | likelihood - free methods , such as approximate bayesian computation , are powerful tools for practical inference problems with intractable likelihood functions .
markov chain monte carlo and sequential monte carlo variants of approximate bayesian computation can be effective techniques for sampling posterior distributions without likelihoods .
however , the efficiency of these methods depends crucially on the proposal kernel used to generate proposal posterior samples , and a poor choice can lead to extremely low efficiency .
we propose a new method for likelihood - free bayesian inference based upon ideas from multilevel monte carlo .
our method is accurate and does not require proposal kernels , thereby overcoming a key obstacle in the use of likelihood - free approaches in real - world situations . |
in recent years there has been considerable discussion of the properties of gases adsorbed on bundles of carbon nanotubes .
measurements of thermal properties ( isotherms , isosteric heats , and thermal desorption behavior ) have provided information about the energies of these gases , while scattering and spectral data have provided information about their structures and dynamical properties.@xcite in parallel with these experiments have been various theoretical and simulation studies , predicting virtually every property that is amenable to such measurements . varying assumptions underlying
the calculations pertain to the geometries and the adsorption potentials .
our group has undertaken a series of computer simulation studies of simple classical gases adsorbed on the outside of large bundles.@xcite since these bundle are large , the calculations actually assumed the bundle surface to be flat , which would be precisely correct if the bundle were actually infinite in radius or else if the bundle consists of a layer of nanotubes deposited on a surface .
the potential experienced by an adsorbed molecule was evaluated by summing contributions from a line of parallel tubes , each of radius @xmath0 ; this approach neglects contributions from the interior of the bundle , an assumption that is plausible only for the first or second adlayer .
[ fig:1 ] shows the potential energy that results for a ne atom in such an environment , based on the potential model used in our group .
that model ignores the chirality of the nanotubes by smearing out the carbon atoms into a cylinder made of continuous carbon matter .
these assumptions are semiquantitatively accurate , at best .
our simulations employ either canonical or grand canonical monte carlo methods . in early work ,
the familiar metropolis algorithm was used , but our most recent study@xcite of ar ( denoted i ) employed the efficient wang - landau algorithm@xcite , discussed below .
the adsorption behavior derived from the model consists of a sequence of `` striped '' phases , with the number of stripes increasing as the gas pressure @xmath1 is increased .
the first of these , appearing at low @xmath1 because of the very attractive potential , is the so - called `` groove '' phase , in which the gas is confined to a quasi - one dimensional vicinity of the potential minimum , nestled between two adjacent tubes . as @xmath1 increases , a transition to a `` three stripe '' phase occurs , followed by the appearance of a well defined monolayer .
an intriguing `` second layer groove '' phase has been seen in both simulations and experiments .
multilayer films can occur , but these have not been investigated because desorption precludes their observation , except at very low temperature .
this paper focuses on the thermal properties of ne and xe , extending the ar study , i. the next section summarizes the methodology and reports results for a very low density phase , called @xmath2 .
the number @xmath3 is the number of adparticles confined within one unit cell of the simulation ( which is periodically replicated across the surface ) .
the extent of this cell in the lateral ( @xmath4 ) direction is 1.7 nm , taken to be the spacing between ( 10,10 ) tubes . in the @xmath5 direction ( parallel to the tubes axes )
, the extent is @xmath6 , where @xmath7 is the lennard - jones diameter of the particles ; the other gas parameter entering the calculation is @xmath8 , the well depth of the adatom - carbon pair interaction . in the @xmath9 direction , perpendicular to the bundle surface , the simulation cell extends a distance @xmath10 nm , where a hard wall terminates the simulation volume .
since this boundary underestimates the ( macroscopic ) experimental gas volume , its presence tends to suppress desorption relative to what is seen in experiments .
sections [ sec : c ] and [ sec : d ] describe simulation results obtained for ne and xe , respectively .
section [ sec : e ] summarizes our principal results and draws conclusions .
contours of constant potential energy for a ne atom ( a ) above the nanotube surface compared with ( b ) energy contours generated by the simple model described in the text . shown are isopotential lines for @xmath11 , @xmath12 , @xmath13 , @xmath14 , @xmath15 , @xmath16 , @xmath17 , and @xmath18 k , starting from the groove .
the rapid change in area between curves for @xmath14 k and @xmath15 k ( dashed lines ) manifests a high density of states , giving rise to a heat capacity peak.,title="fig:",width=302 ] contours of constant potential energy for a ne atom ( a ) above the nanotube surface compared with ( b ) energy contours generated by the simple model described in the text . shown
are isopotential lines for @xmath11 , @xmath12 , @xmath13 , @xmath14 , @xmath15 , @xmath16 , @xmath17 , and @xmath18 k , starting from the groove .
the rapid change in area between curves for @xmath14 k and @xmath15 k ( dashed lines ) manifests a high density of states , giving rise to a heat capacity peak.,title="fig:",width=302 ]
the wang - landau algorithm is a simulation technique that generates the density of states @xmath19 , the number of states per unit energy at energy @xmath20 , for a specified number ( @xmath3 ) of particles ; we omit a descriptor @xmath3 from @xmath19 for notational simplicity . the procedure used to compute @xmath19
is discussed in i and in greater detail in a recent thesis reporting this work.@xcite from this function , thermal properties may be computed at any temperature @xmath21 . since classical systems have a specified kinetic energy per particle , @xmath22 , we focus on the potential energy and omit that contribution except when showing numerical results for the heat capacity @xmath23 . for example
, the ( mean ) potential energy @xmath24 of the system is @xmath25\,.\label{eq:1}\ ] ] the specific heat ( in boltzmann units ) is then computed from @xmath26_n\label{eq:2}\,.\ ] ] results for @xmath19 for @xmath2 are shown in fig . [ fig:2].@xcite this function exhibits a striking variation with @xmath20 , which can be understood with simplifying models , as follows . for classical systems
, @xmath27 is strictly proportional to the volume of phase space confined between surfaces of energy @xmath20 and @xmath28 .
this quantity , for n=1 particle , was called the `` volume density of states '' in i , where it was evaluated from histograms of the potential energy distribution in real space ; the result agrees with that found directly with the wang - landau simulation method .
the lowest energy `` states '' originate from the vicinity of the groove region .
the resulting low @xmath20 behavior , @xmath29 , was derived in i by assuming that the constant energy contours are circles , centered on the groove ; ellipses yield the same analytical form .
the two most striking features of @xmath19 are a sharp peak , near a potential energy @xmath30 k of and a divergence near the desorption threshold , @xmath31 .
the latter divergence is of the form @xmath32 , where the power 4/3 was derived in i from the asymptotic van der waals interaction , @xmath33 . the divergent behavior just above the peak at @xmath34 , @xmath35 was derived in i , by analyzing an assumed potential energy near the cylindrical surface at which the monolayer forms : @xmath36 . here
, the value @xmath34 is the holding potential of a particle near a single cylindrical tube ; the quadratic functional form is analogous to that found near the equilibrium position on a flat surface . to understand the divergent behavior of @xmath19 just below the peak ( @xmath37 ) ,
we have invented a simple model potential that roughly characterizes the spatial dependence near the groove and monolayer : @xmath38 here @xmath39 and @xmath40 k / @xmath41 are fitting parameters determined from the potential .
this form is shown in fig .
[ fig:1 ] to resemble that of the `` true '' potential near the divergence for @xmath42 .
even for this simple model , we can not determine @xmath19 analytically , unfortunately , but we can evaluate it numerically . for the `` true '' potential the density of states
was calculated directly using wang - landau monte carlo , and for the simple model it was calculated by discretizing space and binning the volume elements with potentials between @xmath20 and @xmath28 to produce a histogram . as seen in fig .
[ fig:2 ] , there is a qualitative agreement between the two curves for @xmath19 in the region of the divergence , just below @xmath43 .
the divergence originates because there is a large volume , near the outermost ellipse surrounding the groove , for which the potential energy is close to this value .
density of states @xmath19 for @xmath2 ne atom ( full curve ) .
this is compared with the prediction ( dashed curve ) derived from the simple model , eq .
[ eq:4].,width=453 ] the specific heat of ne and xe , for @xmath2 , is analogous to that found for ar . at low @xmath44 , the value 5/2 is expected for the specific heat @xmath45 because particles have two degrees of freedom due to potential energy of the confining potential within the groove , plus three more degrees of freedom from the kinetic energy .
the peak in @xmath23 near @xmath46 k for ne , or @xmath47 k for xe , is due to excitation out of the groove , on to the cylindrical surface of the nanotube .
this is confirmed by model calculations , as described in i , in which particles are located either in the groove region or on a nominally flat surface .
the high @xmath44 limit , @xmath48 , is attributable to the potential energy contribution from one degree of freedom , associated with motion perpendicular to the surface .
this behavior is depicted , along with higher coverage results , in figs .
[ fig:3 ] and [ fig:4 ] .
the wang - landau method was used to calculate the heat capacity @xmath23 for higher coverages ( fig .
[ fig:3 ] ) . , but ultimately failed to converge completely : the density of states at low energies did not achieve sufficient flatness even after the run time was extended .
convergence studies suggest that the small `` shoulder '' next to the first heat capacity peak may be a computational artifact , and the second peak may be slightly larger than is depicted . ] for ne , the groove phase occurs at a density of @xmath49 adatoms in the simulation cell ; the three stripe , monolayer , and second groove phases occur at @xmath50 , 72 , and 81 , respectively .
specific heat of ne for various coverages @xmath3 ( explained in the text ) , as a function of @xmath44.,width=453 ] the low density @xmath2 results show a heat capacity peak near @xmath46 k , due to excitation of ne atoms out of the groove .
the higher coverage results have a similar peak ( referred to here as the second peak ) , as well as an additional low temperature peak ( first peak ) , i.e. , a characteristic double peak structure in @xmath23 .
monte carlo configurational samples indicate that the second peak is due to substantial excitation of adatoms , as in the @xmath2 case .
the excitation is primarily from the groove into the stripe or monolayer regions .
the second peak occurs at somewhat lower temperature than in the @xmath2 case , because it is easier to excite atoms out of a densely packed groove than a nearly empty one .
this is a result of repulsive mutual interactions that predominate when the groove nears completion . the lower temperature , first peak in the heat capacity , in contrast , is not associated with promotion of adatoms into different regions of the substrate .
rather , it is due to `` energy broadening '' , in which the adatoms remain at or near the same sites , but are thermally excited over a broader range of energies .
this phenomenon is exhibited in fig .
[ fig:4 ] for the three stripe phase .
the distribution of ne atoms indicates that at the first peak ( @xmath51 k ) , most of the atoms remain in the stripes , albeit with some promotion into the full monolayer .
the atoms remaining in the stripes spread out , confined less strongly in the stripe potential minima , as a result of energy broadening . at the second peak , atoms in the stripes have been fully promoted into the monolayer and the stripes disappear .
ne density distribution in the three stripe ( @xmath50 ) phase at various temperatures ( cross sectional view of adjacent nanotubes ) .
the @xmath51 and 75 k panels show the ne distribution at the first and second peaks in the heat capacity.,width=453 ]
the xe behavior , shown in figs . [ fig:5 ] and [ fig:6 ] , is qualitatively similar to that of ne , as might be expected .
there are some differences of detail , however .
for example , the peaks are smaller and occur at much higher temperature in the xe case .
the higher temperature is attributable to the higher energy scale associated with the interactions strengths ; the gas - gas well depth for xe is a factor of seven higher than that of ne and the adsorption potential is nearly a factor of three more attractive .
( the behavior of @xmath23 of ar , explored in i , is intermediate between these two . )
we surmise that the smaller peak height is due to the fact that the importance of the adhesive interaction , relative to the cohesive interaction , is reduced in the xe case , so the peak in @xmath23 , attributed primarily to the interaction with the substrate , is less narrow for xe . for the same reason ,
we believe , `` stronger '' cohesive interactions , arising from subtle intra - film interactions , produce wiggles for @xmath23 of xe , not found for ne .
( we have confirmed that these are not artifacts of the calculation by refining our simulations . )
we conjecture that the intermediate `` wiggle '' in @xmath23 for the monolayer film , near @xmath52 k , is due to the two non - groove stripes melting into the monolayer ( compare the @xmath53 and 150 k density distributions in fig . [ fig:6 ] ) ; the lower - energy groove stripe itself remains intact until the groove - to - monolayer promotion at the higher temperature peak near @xmath54 k. specific heat of xe for various coverages @xmath3 , as a function of @xmath44.,width=453 ] xe density distribution in the monolayer ( @xmath55 ) phase at various temperatures , shown as cross sectional view of adjacent nanotubes .
the @xmath53 and 250 k panels show the xe distribution near the first and second peaks in the heat capacity.,width=453 ]
we have derived the evolution of density with t and the heat capacity for ne and xe gases adsorbed on an array of nanotubes .
the behavior is qualitatively consistent with that reported previously in i for the case of ar .
these predictions , along with those of the isotherms,@xcite are testable experimentally and we look forward to such results , from which conclusions may be drawn .
the principal assumptions that may or may not be validated by such studies pertain to the geometry and the interactions .
the latter type of assumption would seem more reliable , since the interactions are semi - empirical , derived from experiments involving adsorption on graphite .
however , extrapolation to the nanotube case , as done here , represents a further , simplified assumption : that the chemistry of the atoms is not sensitive to their structure .
while this might seem plausible for c , it remains to be verified .
more severe , presumably , is the assumption that the interaction arises from a smeared - out array of c atoms . while clearly open to question , this assumption might be justified by the fact that since the assumed pair potential is already a rough approximation , the continuum approximation is not much more drastic
of course , possible commensurate structures are not found in the continuum approach .
concerning the geometry , the assumption of a unique radius for all tubes is another oversimplification,@xcite as is the presumption that the bundle radius is essentially infinite . while more detailed studies are feasible , they require significantly greater computational effort .
99 m. r. johnson , s. rols , p. wass , m. muris , m. bienfait , p. zeppenfeld , and n. dupont - pavlovsky , chem .
phys . * 293 * , 217 ( 2003 ) ; b. k. pradhan , _ et al .
_ , j. mater
17 * , 2209 ( 2002 ) .
a. d. migone and s. talapatra , in _ encyclopedia of nanoscience and nanotechnology _ , vol . 4 , edited by h. s. nalwa ( american scientific publishers , los angeles , 2004 ) , p. 749 . s. ramachandran _
et al . _ ,
j. low temp .
* 134 * , 115 ( 2004 ) .
y. h. kahng , r. b. hallock , and e. dujardin , physica b * 329 * , 280 ( 2003 ) .
t. wilson , a. tyburski , m. r. depies , and o. e. vilches , j. low temp . phys . * 126 * , 403 ( 2002 ) ; f. r. hung , k. e. gubbins , r. radhakrishnan , k. szostak , f. beguin , g. dudziak and m. sliwinska - bartkowiak , applied phys .
lett . * 86 * , 103110 ( 2005 ) .
h. chen , j. k. johnson , and d. s. sholl , j. phys . chem .
b * 110 * , 1971 ( 2006 ) ; c. matranga , l. chen , b. bockrath , and j. k. johnson , phys
b * 70 * , 165416 ( 2004 ) ; a. kuznetsova , j. t. yates , jr . , v. v. simonyan , j. k. johnson , c. b. huffman , and r. e. smalley , j. chem . phys . * 115 * , 6691 ( 2001 ) . _ adsorption by carbons _ , edited by j. m. d. tascon ( elsevier , amsterdam , 2005 ) .
d. g. narehood _
et al . _ ,
b * 67 * , 205409 ( 2003 ) .
s. m. gatica , m. j. bojan , m. w. cole , m. m. calbi , and j. k. johnson , in ref . .
s. m. gatica , m. j. bojan , g. stan , and m. w. cole , j. chem . phys . * 114 * , 3765 ( 2001 ) ; m. m. calbi , s. m. gatica , m. j. bojan and m. w. cole , j. chem . phys . * 115 * , 9975 ( 2001 ) .
n. m. urban , s. m. gatica , m. w. cole , and j. l. riccardo , phys .
b * 71 * , 245410 ( 2005 ) .
f. wang and d. p. landau , phys .
e * 64 * , 056101 ( 2001 ) ; d. p. landau , s .- h .
tsai , and m. exler , am . j. phys .
* 72 * , 10 ( 2004 ) .
the density of states appearing in i for @xmath2 is mistaken .
see n. m. urban , ph.d .
thesis , pennsylvania state university ( 2006 ) , unpublished .
w. shi and j. k. johnson , phys .
lett . * 91 * , 015504 ( 2003 ) . | the structural and thermal properties of ne and xe gases adsorbed on the outer surface of a large nanotube bundle have been evaluated with computer simulation .
the potential energy model and numerical techniques were used previously to study ar [ n. m. urban , s. m. gatica , m. w. cole , and j. l. riccardo , `` correlation functions and thermal properties of ar adsorbed on the external surface of a bundle of carbon nanotubes '' , phys .
rev .
b * 71 * , 245410 ( 2005 ) ] . heat capacity results for ne and xe exhibit peaks associated with reordering and `` stripe '' melting transitions for these gases . |
in this paper we analyse mutations between fake weighted projective spaces ; equivalently we analyse mutations between lattice simplices .
mutations arise naturally when considering mirror symmetry for fano manifolds .
a fano manifold @xmath0 is expected to correspond under mirror symmetry to a laurent polynomial @xcite . in general
there will be many different laurent polynomials which correspond to a given fano manifold , and it is expected that these laurent polynomials are related via birational transformations analogous to cluster transformations @xcite .
these cluster - style transformations act on newton polytopes via mutations ; see definition [ defn : mutation ] below .
a mutation between newton polytopes can be thought of as the tropicalisation of the corresponding cluster - type transformation .
ilten @xcite has shown that if two lattice polytopes @xmath1 and @xmath2 are related by mutation then the corresponding toric varieties @xmath3 and @xmath4 are deformation equivalent : there exists a flat family @xmath5 such that @xmath6 and @xmath7 .
mutations are thus expected to form the one - skeleton of the tropicalisation of the hilbert scheme .
our understanding of this one - skeleton is rudimentary but improving @xcite ; in this paper we conduct the first systematic analysis in higher dimensions .
the notion of mutation raises many new and interesting combinatorial questions : for example , how can polytopes be classified up to mutation , and what properties of polytopes are mutation - invariant ? here
we begin to address these questions by analysing the behaviour of lattice simplicies under mutation . in two dimensions ,
akhtar and kasprzyk determined how the weights of a fake weighted projective plane , i.e. the weights of a lattice triangle , change under mutation , and showed that mutations between fake weighted projective planes are multiplicity - preserving @xcite .
we show below that the situation in higher dimensions is different .
we give an explicit formula ( theorem [ mutationformula ] ) for how the weights of a fake weighted projective space , i.e. the weights of a lattice simplex , change under mutation , and derive a strong necessary condition ( theorem [ onlyedges ] ) for a mutation to preserve multiplicity . in [ sec : canonical_sings][sec : terminal_sings ] we apply our results to the study of fake weighted projective spaces of high degree with canonical and terminal singularities . the authors thank mohammad akhtar and alessio corti for useful conversations .
this research is supported by the royal society , erc starting investigator grant number 240123 , the leverhulme trust , and epsrc grant ep / i008128/1 .
we begin by recalling some standard definitions . throughout let @xmath8 denote a lattice of rank @xmath9 with dual lattice @xmath10 . from the toric viewpoint @xmath11
corresponds to the lattice of one - parameter subgroups and @xmath12 corresponds to the lattice of characters . for an introduction to toric geometry see @xcite .
[ defn : fano_polytope ] a convex lattice polytope @xmath13 is said to be _ fano _ if : 1 .
[ item : nondegenerate ] @xmath1 is of maximum dimension , that is @xmath14 ; 2 .
[ item : fano_orig ] the origin is contained in the strict interior of @xmath1 , that is @xmath15 ; 3 . [
item : fano_primitive ] the vertices @xmath16 of @xmath1 are primitive lattice points .
in addition to being compelling combinatorial objects , fano polytopes are in bijective correspondence with toric fano varieties : see @xcite for an overview .
the complete fan in @xmath11 generated by the faces of @xmath1 , which we call the _ spanning fan _ of @xmath1 , corresponds to a toric fano variety @xmath3 , that is , to a ( possibly singular ) projective toric variety with ample anticanonical divisor @xmath17 .
two fano polytopes @xmath1 and @xmath18 give isomorphic varieties @xmath19 if and only if there exists a change of basis of the underlying lattice @xmath11 sending @xmath1 to @xmath18 . thus we regard @xmath1 as being defined only up to @xmath20-equivalence .
[ defn : weights ] let @xmath21 be a fano @xmath9-simplex .
by definition [ defn : fano_polytope ] there exists a unique choice of @xmath22 coprime positive integers @xmath23 such that @xmath24 .
these are called the _ ( reduced ) weights _ of @xmath1 .
definition [ defn : fano_polytope ] implies that any @xmath9 of the weights are also coprime , or in other words that the weights are _ well - formed_. see @xcite for details of the natural role that reduced and well - formed weights play in the study of weighted projective space .
[ defn : fake_wps ] let @xmath25 be a fano @xmath9-simplex and let @xmath26 be the sublattice in @xmath11 generated by the vertices of @xmath1 . the rank - one @xmath27-factorial toric fano variety @xmath3 given by the spanning fan of @xmath1 is @xmath28 , where the group @xmath29 acts freely in codimension one .
we call @xmath3 a _ fake weighted projective space_. fake weighted projective spaces have been studied in @xcite . one important invariant is the _ multiplicity _ : the index of the sublattice @xmath30 in @xmath11 , denoted by @xmath31 $ ] .
a fake weighted projective space is a weighted projective space if and only if @xmath32 ( * ? ? ?
* proposition 2 ) .
we recall the definition of mutation , following @xcite . a primitive element
@xmath33 determines a surjective linear map @xmath34 which extends naturally to a map @xmath35 .
a point @xmath36 is said to be at _ height _ @xmath37 . given a subset @xmath38 ,
if @xmath39 for all @xmath40 we say that @xmath41 lies at height @xmath42 and write @xmath43 .
the hyperplane @xmath44 is defined to be the set of all points in @xmath45 at height @xmath42 . for a convex lattice polytope
@xmath25 we define @xmath46 to be the ( possibly empty ) convex hull of all lattice points in @xmath1 at height @xmath42 .
we set @xmath47 to be the minimum height occurring amongst the the points of @xmath1 , and @xmath48 to be the maximum height .
since @xmath1 is a lattice polytope , both @xmath49 and @xmath48 are integers .
if @xmath15 , and in particular if @xmath1 is fano , then @xmath50 and @xmath51 .
[ defn : factor ] a _ factor _ of @xmath25 with respect to a primitive height function @xmath33 is a lattice polytope @xmath52 such that : 1 . @xmath53 ; 2 . for every integer @xmath42 with @xmath54
, there exists a ( possibly empty ) lattice polytope @xmath55 such that @xmath56 .
[ defn : mutation ] the _ ( combinatorial ) mutation _ of @xmath25 with respect to the primitive height function @xmath33 and a factor @xmath52 is the convex lattice polytope @xmath57 although this is not obvious from the definition , the mutation @xmath58 is independent of the choice of the @xmath59 ( * ? ? ?
* proposition 1 ) .
furthermore , @xmath58 is a fano polytope if and only if @xmath1 is a fano polytope ( * ? ? ?
* proposition 2 ) .
more obviously , mutations are always invertible ( if @xmath60 then @xmath61 ) ( * ? ? ?
* lemma 2 ) , and translating the factor results in an isomorphic mutation ( i.e. @xmath62 for any @xmath63 with @xmath64 ) .
mutations have a natural description as transformations of the dual polytope @xmath65 a mutation induces a piecewise @xmath20-transformation @xmath66 of @xmath67 such that @xmath68 ; here @xmath69 .
this is analogous to a cluster transformation . as a consequence @xmath70 for any dilation @xmath71 , where @xmath60 ( * ? ? ?
* proposition 4 ) ; hence @xmath72 and @xmath3 and @xmath4 have the same anticanonical degree .
[ exa : p1113p1146 ] the weighted projective spaces @xmath73 and @xmath74 have the largest degree amongst all canonical below for a discussion of canonical singularities . ] _ toric _ fano threefolds @xcite and amongst all _ gorenstein _ canonical fano threefolds @xcite .
they are related by a mutation ( * ? ? ?
* example 7 ) .
the simplex associated to @xmath73 is @xmath75 .
setting @xmath76 gives @xmath77 and @xmath78 , with @xmath79 equal to the edge @xmath80 and @xmath81 given by the vertex @xmath82 .
the factor @xmath83 gives : @xmath84 and this is the simplex associated with @xmath74 .
this mutation is illustrated in figure [ fig : maximal_degree ] .
, depicted on the left , to the polytope corresponding to @xmath74 .
this mutation is described in example [ exa : p1113p1146].,title="fig : " ] , depicted on the left , to the polytope corresponding to @xmath74 .
this mutation is described in example [ exa : p1113p1146].,title="fig : " ]
we begin by establishing some basic properties of mutations between simplices . throughout this section
we assume that @xmath1 is a fano @xmath9-simplex , and that @xmath85 and @xmath86 are ( respectively ) a primitive height function and a factor such that @xmath87 is a simplex .
in other words , we assume that there is a mutation from the fake weighted projective space @xmath0 associated with @xmath1 to the fake weighted projective space @xmath88 associated with @xmath2 .
[ ncasesprop ] let @xmath1 be a fano @xmath9-simplex , let @xmath85 be a primitive height function , and let @xmath86 be a factor such that @xmath89 is a simplex .
suppose that the mutation from @xmath1 to @xmath2 is non - trivial , so that @xmath90
. then @xmath91 is a vertex of @xmath1 and @xmath92 is a translation of @xmath93 .
we first consider @xmath91 .
suppose for a contradiction that @xmath91 is not a vertex of @xmath1 .
then in particular it contains an edge @xmath94 of @xmath1 .
let us pick an edge @xmath95 contained in @xmath96 ( if such an edge does not exist then @xmath96 is a vertex and the mutation is trivial ) .
note that @xmath94 and @xmath95 can not be parallel : if they were then the four endpoints of @xmath94 and @xmath95 would lie in a common two - dimensional affine subspace and thus would be affinely dependent ; this contradicts the fact that @xmath1 is a simplex .
it is an immediate consequence of the definition of mutation that @xmath97 . recall that @xmath92 is a minkowski factor of @xmath96 .
because @xmath1 is a simplex , @xmath96 is also a simplex and so by ( * ? ? ?
* result 13 ) @xmath92 is a dilation and translation of @xmath96 .
it follows that @xmath98 is also a dilation and translation of @xmath96 .
let @xmath99 denote the edge of @xmath98 corresponding to the edge @xmath95 of @xmath96
. then @xmath94 and @xmath99 are not parallel , because @xmath94 and @xmath95 are not parallel .
thus the face @xmath100 of @xmath97 is a quadrilateral .
since @xmath2 is by assumption a simplex , and therefore all faces of @xmath2 are simplices , this gives a contradiction .
we conclude that @xmath91 consists of a single vertex .
now consider the factor @xmath92 . by the definition of factor there exists a lattice polytope @xmath101 such that @xmath102 we claim that @xmath101 is a point .
after mutation we have @xmath103 .
as discussed above , we can mutate @xmath2 back to @xmath1 by taking the height function @xmath104 and the factor @xmath92 .
but @xmath105 and we conclude from the first part of the proposition that @xmath101 is a point : @xmath106 .
it follows that @xmath107 , and so @xmath92 is a translation of @xmath93 as required .
if @xmath96 has @xmath108 vertices then we say that the corresponding mutation is a _ mutation over a @xmath109-face_. cases of particular interest are @xmath110 and @xmath111 , which we call _ mutations over edges _ and _ mutations over facets _
, respectively . [ zeroheight ]
let @xmath1 be a fano @xmath9-simplex , let @xmath85 be a primitive height function , and let @xmath86 be a factor such that @xmath89 is a simplex .
let @xmath112 be such that @xmath113 and @xmath114
. then @xmath64 .
let us write @xmath115 and @xmath116 .
if @xmath117 then the statement holds vacuously , so let us assume that @xmath118 .
without loss of generality we may assume that @xmath119 .
in view of lemma [ ncasesprop ] we may assume further that @xmath120 , that @xmath121 , and that @xmath122 suppose for a contradiction that there exists some @xmath112 such that @xmath123 , @xmath114 , and @xmath124 . without loss of generality we can take @xmath125 .
suppose first that @xmath126 .
let @xmath127 be a primitive lattice point such that @xmath128 for @xmath129 .
we can choose @xmath130 so that @xmath131 and @xmath132 .
let @xmath133 and @xmath134 .
we see that @xmath135 and @xmath136 .
note that by the definition of @xmath92 , @xmath137 and so @xmath138 for all @xmath139 ( with @xmath140 ) . then @xmath141 and @xmath142 .
thus @xmath143 and @xmath144 are two faces of @xmath2 , with vertices @xmath145 and @xmath146 for @xmath147 .
let @xmath148 be any two distinct vertices of @xmath92 ; these exist since otherwise @xmath92 is a point and the mutation is trivial . as @xmath2 is
a simplex @xmath149 is a two - dimensional face of @xmath2 .
but @xmath150 has four vertices and thus is not a simplex .
this gives a contradiction .
suppose instead that @xmath151 . by (
* lemma 3.7 ) there exists a @xmath152 and @xmath153 such that @xmath154 .
applying @xmath155 to this equation yields @xmath156 .
we now have @xmath157 but @xmath158 which , by the preceding argument , gives a contradiction .
it follows that @xmath159 .
combining the previous two results gives a necessary and sufficient combinatorial condition for the existence of a mutation between simplices .
[ iffmut ] let @xmath1 be a fano @xmath9-simplex and let @xmath85 be a primitive height function .
let the vertices of @xmath1 be @xmath160 , ordered such that @xmath161 .
there exists a factor @xmath86 such that @xmath89 is a simplex if and only if the following hold : 1 .
[ item : iffmut_1 ] @xmath162 is a vertex ; 2 .
[ item : iffmut_2 ] @xmath163 for @xmath164 ; 3 .
[ item : iffmut_3 ] if @xmath165 is a vertex of @xmath1 such that @xmath123 and @xmath114 , then @xmath166 .
the `` only if '' direction is lemmas [ ncasesprop ] and [ zeroheight ] . on the other hand , it is clear that if conditions are satisfied , then defining @xmath167 gives a factor with respect to @xmath155 . let us label the vertices of @xmath168 so that : @xmath169 from ( * ? ? ?
* lemma 3.7 ) we have that @xmath170 , where : @xmath171 furthermore by ( * ? ? ?
* proposition 3.11 ) we have that @xmath2 is fano , and hence is of maximal dimension in the @xmath9-dimensional lattice .
it follows that @xmath2 must have at least @xmath22 vertices .
thus @xmath172 , and so @xmath2 is a simplex .
[ exa : p1148p1114 ] consider the @xmath173-simplex @xmath174 .
the weighted projective space associated to @xmath1 is @xmath175 .
let @xmath176 .
lemma [ iffmut ] implies that there is a factor @xmath92 such that @xmath177 is the simplex : @xmath178 the weighted projective space associated to @xmath2 is @xmath179 .
the main result of this paper is : [ mutationformula ] let @xmath0 and @xmath88 be fake weighted projective spaces related by a mutation .
suppose that @xmath180 is the simplex corresponding to @xmath0 , that @xmath181 is the simplex corresponding to @xmath88 , and that the weights of @xmath0 are @xmath182 .
let @xmath33 be the primitive height function and @xmath86 be the factor such that @xmath183 .
then we may relabel the vertices of @xmath1 such that @xmath184 , @xmath185 , and the weights of @xmath88 are : @xmath186 where @xmath187 is a positive integer satisfying : @xmath188 by lemma [ iffmut ] we have @xmath172 where : @xmath189 since @xmath1 is fano we may assume that the weights @xmath182 are well - formed .
we normalise by setting @xmath190 and @xmath191
. then @xmath192 , @xmath193 , and @xmath194 for all @xmath195 .
the sequence @xmath196 is unique with these properties ; these are the _ normalised barycentric co - ordinates _ for @xmath1 .
since @xmath2 is fano there exist @xmath197 such that @xmath198 , @xmath199 , and @xmath200 for all @xmath195 . from @xmath198 and we have that : @xmath201 let @xmath202 denote the coefficient of @xmath203 in the expression above .
we claim that the @xmath202 are normalised barycentric co - ordinates .
it is clear that @xmath204 , and that @xmath205 .
it remains to check that @xmath206 .
suppose for a contradiction that @xmath207 .
then we have : @xmath208 and @xmath209 , so a point on the ray from @xmath210 through @xmath211 lies in the cone over @xmath212 .
this contradicts the fact that @xmath210 lies in the strict interior of @xmath1 .
hence the @xmath202
are normalised barycentric co - ordinates for @xmath1 , and so by uniqueness we have @xmath213 for all @xmath195 . solving these equations for @xmath214 yields :
@xmath215 applying @xmath155 to both sides of the equation @xmath216 , we find that @xmath217 and thus : @xmath218 now we can form integer weights by defining @xmath219 ; this gives : @xmath220 these weights are integers , but they may not be well - formed . however since @xmath2 is fano , we know that the weights are well - formed if and only if they are reduced . therefore it remains to divide through by their greatest common divisor , which we denote by @xmath187 .
thus the weights of @xmath2 are : @xmath221 consider now the degrees of @xmath0 and @xmath88 : @xmath222 since degree is preserved by mutation , we conclude that @xmath223 as claimed .
[ rem : multiplicity_preserving ] when @xmath224 , and in particular if @xmath0 and @xmath88 are weighted projective spaces , we have : @xmath225 this gives an explicit expression for the weights after mutation in terms of the weights before mutation . in the case of a mutation over a facet
we see that the new weights are : @xmath226 note that @xmath227 divides @xmath187 , because @xmath228 for @xmath229 and the weights are well - formed .
on the other hand , after dividing through by @xmath227 we obtain well - formed weights , and so in fact @xmath230 . in this case , therefore , we obtain an explicit formula for how the multiplicity changes : @xmath231
the following result places a strong restriction on which mutations of fake weighted projective spaces can preserve multiplicity .
[ onlyedges ] any non - trivial multiplicity - preserving mutation between fake weighted projective spaces @xmath0 and @xmath88 is a mutation over an edge .
let @xmath232 be the simplex associated to @xmath0 , and let @xmath182 be the corresponding weights .
let @xmath233 be the simplex associated to @xmath88 , and let @xmath196 be the corresponding weights .
suppose for a contradiction that @xmath1 and @xmath2 are related by a non - trivial mutation over a @xmath109-face , for some @xmath234 . by remark [ rem : multiplicity_preserving ] , after reordering weights if necessary , we have : @xmath235 where : @xmath225 we recall from that @xmath236 and write @xmath237 with @xmath238 and @xmath239 coprime integers .
so , for @xmath240 , we have : @xmath241 since @xmath238 and @xmath239 are coprime and @xmath242 , we have that @xmath243 .
similarly for @xmath244 we have : @xmath245 and so in this case too @xmath243 . however @xmath246 are coprime , because the weights of @xmath0 are well - formed , and therefore @xmath247 .
since @xmath248 is an integer , we have that @xmath249 .
it follows that @xmath250 for all @xmath195 , and since the weights @xmath251 are reduced we conclude that @xmath252 .
then @xmath253 is an integer . taking the @xmath254th root ( recall that @xmath234 ) we see that @xmath255 is an integer , and hence that @xmath256 .
thus @xmath257 .
substituting this into our expression for the @xmath251 shows that the mutation is trivial , which is a contradiction .
[ cor : wps_weight_divisibility ] suppose that @xmath0 is a weighted projective space that admits a non - trivial mutation to another weighted projective space .
let @xmath227 be the weight corresponding to the vertex @xmath91 , and let @xmath258 , @xmath259 be the weights corresponding to the vertices of the edge @xmath96 . then : 1 . [ item : wps_weight_divisibility_1 ] @xmath260 2 . [ item : wps_weight_divisibility_2 ] @xmath261 the mutation is an edge mutation , and so theorem [ mutationformula ] implies both that @xmath230 and that @xmath262 .
this proves . looking again at theorem [ mutationformula ] and using well - formedness of weights
, we see that @xmath263 .
this proves .
let @xmath0 be a fake weighted projective plane .
akhtar and kasprzyk characterise mutations from @xmath0 to other fake weighted projective planes in terms of solutions to an associated diophantine equation ( * ? ? ?
* proposition 3.12 ) .
their argument relies on the fact that , for lattice triangles , the square - free parts of the weights are preserved ( up to reordering ) under mutation .
this phenomenon does not persist in higher dimensions : 1 .
example [ exa : p1113p1146 ] above shows that , in general , neither the square - free parts of the weights nor the @xmath109th - power - free parts of the weights nor the square parts of the weights nor the @xmath109th - power parts of the weights are preserved .
2 . example [ exa : p1148p1114 ] above shows that , in general , neither the @xmath9th - power - free parts nor the @xmath9th - power parts of the weights are preserved .
terminal and canonical singularities were introduced by miles reid ; they play a fundamental role in birational geometry @xcite .
terminal singularities form the smallest class of singularities that must be allowed if one wishes to construct minimal models in dimensions three or more .
canonical singularities can be regarded as the limit of terminal singularities ; they arise naturally as the singularities occurring on canonical models of varieties of general type . from the toric viewpoint ,
terminal and canonical singularities have a particularly elegant combinatorial description .
a toric singularity corresponds to a strictly convex rational polyhedral cone @xmath264 @xcite .
the cone @xmath265 is _ terminal _ if and only if : 1 .
[ item : q_gorenstein ] the lattice points @xmath266 corresponding to the primitive generators of the rays of @xmath265 are contained in an affine hyperplane @xmath267 for some @xmath268 ; 2 .
[ item : terminal_hyper ] with the exception of the origin @xmath210 and the generators @xmath269 of the rays , no other lattice points of @xmath11 are contained in the part of @xmath265 on or under @xmath270 , i.e. @xmath271 the cone @xmath265 is _ canonical _ if and only if holds and 1 . [ item : canonical_hyper ] the origin @xmath210 is the only lattice point contained in the part of @xmath265 under @xmath270 , i.e. @xmath272 when the hyperplane @xmath270 in condition corresponds to a lattice point @xmath273 , the singularity is called _
we now apply our results to the study of weighted projective spaces with canonical singularities , giving some geometric context , in terms of mutations , for a recent combinatorial result that characterises the fake weighted projective spaces of maximal degree with at worst canonical singularities @xcite .
we say that a ( fake ) weighted projective space with at worst canonical singularities is a _ canonical _ ( fake ) weighted projective space .
if @xmath274 is a canonical fake weighted projective space then @xmath275 has canonical singularities too .
thus a canonical fake weighted projective space of maximal degree is necessarily a weighted projective space .
[ def : sylvester ] the _ sylvester numbers _
@xmath276 are defined by : @xmath277 we set @xmath278 .
[ lem : sylv ] 1 .
[ item : sylv_1 ] the sylvester numbers are pairwise coprime ; 2 . [
item : sylv_2 ] if @xmath279 and @xmath280 then @xmath281 is an integer . furthermore if @xmath282 is a prime dividing @xmath283 then @xmath284 for @xmath285 , @xmath286 , and @xmath287 ; 3 .
[ item : sylv_3 ] we have : @xmath288 define : @xmath289 @xmath290 is an @xmath9-dimensional canonical weighted projective space .
averkov , krmpelmann , and nill @xcite have proved that , if @xmath291 , then @xmath290 is the unique @xmath9-dimensional canonical weighted projective space of maximum degree ; this generalizes the corresponding result for gorenstein weighted projective spaces , which is due to nill @xcite . in dimension three
there are precisely two canonical weighted projective spaces of maximum degree , and these are connected by a mutation : see example [ exa : p1113p1146 ] . we next determine all weighted projective spaces @xmath292 that might be connected to @xmath290 by a sequence of mutations through weighted projective spaces , and show that none of the @xmath292 are canonical .
[ canonicalwpsmutationgraph ] define integers @xmath293 , where @xmath294 and @xmath295 , by : @xmath296 for @xmath297 .
set : @xmath298 then : 1 .
[ item : canonicalwpsmutationgraph_1 ] @xmath299 for all @xmath300 ; 2 .
[ item : canonicalwpsmutationgraph_2 ] if @xmath301 and if @xmath0 is a weighted projective space with a non - trivial mutation to @xmath292 then either @xmath302 or @xmath303 ; 3 .
[ item : canonicalwpsmutationgraph_3 ] @xmath292 is not canonical for any @xmath297 . in other words ,
the graph of mutations between weighted projective spaces starting at @xmath290 is a subtree of the following graph : ] [ .@xmath290 [ .@xmath304 [ .@xmath305 [ .@xmath306 ] ] ] [ .@xmath307 [ .@xmath307 [ .@xmath307 ] ] ] [ .@xmath308 [ .@xmath309 [ .@xmath310 ] ] ] [ .@xmath311 [ .@xmath312 [ .@xmath313 ] ] ] ] [ fig : canonical_graph ] and the only canonical weighted projective space in this graph is @xmath290 .
statement is trivial . for , recall that a weighted projective space @xmath314 is canonical if and only if @xmath315 for every @xmath316 , where @xmath317 denotes the fractional part of @xmath318 and @xmath42 is the sum of the @xmath251 ( * ? ? ?
* proposition 2.5 ) .
this fails for @xmath297 and @xmath319 , where @xmath320 is the sum of the weights of @xmath292 : see lemma [ lem : not_canonical ] .
it remains to prove .
suppose first that there is a non - trivial mutation from @xmath290 to a weighted projective space @xmath0 . by theorem [ onlyedges ] , this must be an edge mutation .
let @xmath1 be the simplex associated to @xmath290 , let @xmath321 denote the weight associated to the vertex @xmath91 , and let @xmath322 , @xmath323 be the weights associated to the vertices of @xmath96 .
we consider the possible values of @xmath321 , @xmath322 , @xmath323 .
then @xmath325 for some @xmath326 .
@xmath327 , @xmath328 . since @xmath329 , there exist @xmath330 , @xmath331 distinct such that @xmath332 and @xmath333
. at least one of @xmath334 or @xmath335 is not equal to @xmath336 , and hence is a divisor of @xmath321 not equal to 2 or 4 . but
@xmath337 , and this contradicts corollary [ cor : wps_weight_divisibility ] .
@xmath338 , @xmath328 .
then @xmath339 for some @xmath340 .
choose @xmath341 not equal to @xmath342 or @xmath330 , and let @xmath282 be a prime dividing @xmath335 .
lemma [ lem : sylv ] imples that @xmath282 divides both @xmath343 and @xmath344 .
so @xmath282 does not divide @xmath345 , and this contradicts corollary [ cor : wps_weight_divisibility ] .
@xmath346 , @xmath347 . then @xmath339 and @xmath348 for @xmath330 , @xmath341 distinct and not equal to @xmath342
. suppose first that @xmath349 , and let @xmath282 be a prime dividing @xmath350 .
lemma [ lem : sylv ] implies that @xmath351 and @xmath352 , contradicting corollary [ cor : wps_weight_divisibility ] . on the other hand if @xmath353 then the same argument with @xmath354 shows that @xmath355 and @xmath356 , again contradicting corollary [ cor : wps_weight_divisibility ] @xmath357 .
@xmath346 , @xmath328 .
then @xmath358 for some @xmath359 , and theorem [ mutationformula ] implies that @xmath360 .
@xmath346 , @xmath361 .
then @xmath362,@xmath363 for @xmath330 , @xmath341 distinct .
choose @xmath342 not equal to @xmath330 or @xmath341 , and let @xmath282 be a prime dividing @xmath350 .
lemma [ lem : sylv ] implies that @xmath351 and @xmath352 , contradicting corollary [ cor : wps_weight_divisibility ] .
this completes the proof in the case where @xmath364 .
suppose now that @xmath365 , and that there is a non - trivial mutation from @xmath292 to a weighted projective space @xmath0 .
theorem [ onlyedges ] implies that this is an edge mutation .
let @xmath1 be the simplex corresponding to @xmath292 , let @xmath321 denote the weight associated to the vertex @xmath91 , and let @xmath322 , @xmath323 denote the weights associated to the vertices of @xmath96 . to declutter the notation , write @xmath366 for the weight @xmath293 of @xmath292 .
we consider the possible values of @xmath321 , @xmath322 , @xmath323 in turn .
@xmath367 with @xmath368 , @xmath369 , @xmath370 . by lemma [ claim3 ]
we have that @xmath321 and @xmath323 share a common factor that does not divide @xmath322 , and so does not divide @xmath371 .
thus @xmath372 , contradicting corollary [ cor : wps_weight_divisibility ] .
the remaining cases are entirely analogous .
we arrive at a contradiction in all but two cases , as summarized in table [ tab : summary_mgt1 ] below ; here @xmath366 , @xmath373 , @xmath374 denote distinct elements of @xmath375 .
the case @xmath376 , @xmath369 , @xmath370 yields @xmath377 , and the case @xmath378 , @xmath379 , @xmath380 yields @xmath381 .
this completes the proof in the case where @xmath365 .
.a summary of the argument in the case where @xmath365 .
[ cols="^,^,^,<,<,^,^,^,^ , < , < " , ] suppose now that @xmath382 .
we argue exactly as for @xmath365 , except in those cases where lemma [ claim7 ] is used .
we consider these three cases separately , once again writing @xmath366 for @xmath293 .
when @xmath383 , @xmath384 , @xmath385 with @xmath386 , we argue as in proposition [ canonicalwpsmutationgraph ] case [ canonical_wps_case_3 ] .
when @xmath387 , @xmath388 , @xmath389 with @xmath386 , we argue as in proposition [ canonicalwpsmutationgraph ] case [ canonical_wps_case_4 ] .
when @xmath390 with @xmath391 , @xmath392 , @xmath393 with @xmath394 , we have @xmath395 , @xmath396 and @xmath397 . since @xmath398 we can find @xmath399 .
let @xmath282 be a prime dividing @xmath400 .
then @xmath282 divides @xmath321 and @xmath323 but @xmath282 does not divide @xmath322 , contradicting corollary [ cor : wps_weight_divisibility ] . [ .@xmath401 [ .@xmath402 [ .@xmath402 [ .@xmath402 ] ] ] [ .@xmath402 [ .@xmath402 [ .@xmath402 ] ] [ .@xmath402 [ .@xmath402 ] [ .@xmath402 ] ] ] [ .@xmath402 [ .@xmath402 [ .@xmath402 ] [ .@xmath402 ] [ .@xmath402 ] ] [ .@xmath402 [ .@xmath402 ] ] [ .@xmath402 [ .@xmath402 ] [ .@xmath402 ] [ .@xmath402 ] [ .@xmath402 ] ] [ .@xmath402 [ .@xmath402 ] [ .@xmath402 ] ] ] ] the requirement in proposition [ terminalwpsmutationgraph ] that @xmath403 is necessary .
figure [ fig:4dwpsmutationgraph ] shows the graph of mutations between four - dimensional weighted projective spaces starting from @xmath404 to a depth of three , where @xmath402 denotes some weighted projective space .
the rightmost branch of the tree above corresponds to a mutation with @xmath405 and @xmath406 .
it can be shown using the methods of [ sec : canonical_sings ] that this branch continues as a chain .
we do not know what happens in the other branches at greater depth .
recall the definition of sylvester numbers from definition [ def : sylvester ] . recall also
that @xmath407 .
we proceed by induction on @xmath9 .
the base case @xmath411 holds trivially .
note that : @xmath412 if @xmath413 then @xmath414 and , by the induction hypothesis , @xmath415 . thus @xmath409 .
if @xmath416 then @xmath417 , which is certainly divisible by @xmath283 .
this completes the induction step , proving that @xmath418 .
it follows that @xmath419 , and so @xmath420 .
we begin by showing that @xmath426 and @xmath427 are coprime , proceeding by induction .
the base case @xmath364 is trivial .
suppose that @xmath428 and @xmath429 are coprime , and suppose that there exists a prime @xmath282 dividing both @xmath426 and @xmath427 .
then : @xmath430 thus @xmath431 , contradicting the hypothesis that @xmath428 and @xmath429 are coprime .
this completes the induction step , showing that @xmath426 and @xmath427 are coprime for all @xmath423 .
it follows immediately that @xmath432 and @xmath433 are also coprime for all @xmath423 . [ claim2][claim4][claim6 ]
let @xmath293 be as in proposition [ canonicalwpsmutationgraph ] , let @xmath438 , and let @xmath439 .
there exists a prime @xmath282 such that @xmath440 , @xmath441 , and @xmath442 for all @xmath443 .
lemma [ claim1 ] implies that , for @xmath444 , @xmath445 .
thus @xmath446 for all @xmath297 , and so there is some prime @xmath282 such that @xmath447 but @xmath448 .
thus @xmath449 for all @xmath450 .
lemma [ claim5 ] now implies that @xmath451 and @xmath452 .
[ claim7][claim9 ] let @xmath293 be as in proposition [ canonicalwpsmutationgraph ] , let @xmath453 , and let @xmath439 .
there exists a prime @xmath282 such that @xmath454 , @xmath455 , and @xmath442 for @xmath456 and all @xmath386 . by lemma [ claim6 ] there
exists a prime @xmath282 such that @xmath282 divides @xmath457 for @xmath450 but @xmath282 does not divide @xmath458 or @xmath428 .
thus @xmath459 for all @xmath460 and , as @xmath461 , we have by lemma [ claim5 ] that @xmath282 does not divide @xmath426 or @xmath427 .
[ claim8 ] let @xmath293 be as in proposition [ canonicalwpsmutationgraph ] , let @xmath467 , and let @xmath460
. then there exists @xmath468 such that for all @xmath469 , @xmath470 for @xmath471 and @xmath472 for @xmath473 .
we proceed by induction on @xmath423 . for @xmath364
we have @xmath474 for some @xmath475 .
if @xmath476 then let @xmath109 be a prime dividing @xmath477 ; otherwise , let @xmath478
. in either case @xmath109 has the desired properties and the claim holds .
suppose now that there exists @xmath468 such that @xmath479 for @xmath471 and @xmath480 for @xmath473 .
lemma [ claim3 ] implies that @xmath109 and @xmath481 are coprime .
thus as @xmath482 we see that @xmath483 for @xmath484 . since @xmath485 , by lemma [ claim3 ] again
we have that @xmath486 and @xmath487 . finally as @xmath109 does not divide @xmath429 but does divide @xmath428 , it does not divide @xmath488 ; since @xmath109 also does not divide @xmath457 , by the recursion formula it can not divide @xmath293 either . for the case
@xmath364 , take @xmath282 to be an odd prime dividing @xmath491 .
suppose now that there exists a prime @xmath282 such that @xmath492 and @xmath493 .
then @xmath494 and and @xmath495 .
now : @xmath496 and by lemma [ claim3 ] , @xmath497 .
thus @xmath498 .
the result follows by induction on @xmath423 .
the cases @xmath364 and @xmath382 are straightforward .
suppose now that @xmath502 and @xmath503 .
we have that : @xmath504 since @xmath505 , we have by lemma [ claim3 ] that @xmath506 and @xmath458 are coprime
. it thus suffices to show that @xmath506 divides the numerator of the above expression . but
this holds by assumption .
the lemma follows by induction on @xmath423 .
we proceed by induction on @xmath423 .
the base case @xmath364 is trivial .
suppose now that @xmath515 .
then , since @xmath516 , it suffices to show that @xmath517 divides @xmath518 . but : @xmath519 now @xmath517 divides @xmath520 and @xmath521 , so @xmath517 divides the numerator here .
lemma [ claim3 ] implies that @xmath517 is coprime to the denominator .
thus @xmath517 divides @xmath522 .
[ lem : not_canonical ] let @xmath293 be as in proposition [ canonicalwpsmutationgraph ] , let @xmath467 , let @xmath297 , and let @xmath523 .
then @xmath524 , and : @xmath525 where @xmath317 denotes the fractional part of @xmath318 .
the first statement follows immediately from lemma [ claim1 ] .
we claim that : @xmath526 note that : @xmath527 we claim that the last term here is an integer .
it is equal to : @xmath528 now : @xmath529 hence is equal to : @xmath530 recall that @xmath531 .
we will show that each @xmath283 divides the numerator of this expression . by lemma [ canonlemma5 ]
we have that @xmath532 , and by lemma [ canonlemma7 ] we have that , for all @xmath533 , @xmath534 .
hence is an integer .
thus : @xmath535 since holds for @xmath382 and @xmath536 , by induction it holds for all @xmath423 .
[ claim11term ] let @xmath293 be as in proposition [ terminalwpsmutationgraph ] , let @xmath541 , let @xmath467 , and let @xmath195 , @xmath542 be distinct .
there exists a prime @xmath282 such that , for all @xmath469 , @xmath282 divides @xmath426 , @xmath293 , and @xmath543 but @xmath282 does not divide @xmath425 or @xmath427 .
we have @xmath544 , @xmath545 , and @xmath546 .
since @xmath398 we can find @xmath547 .
let @xmath282 be a prime dividing @xmath548 .
then @xmath282 divides @xmath549 , @xmath550 , and @xmath551 and thus , by lemma [ claim3 ] , @xmath282 divides neither @xmath552 nor @xmath553 .
we now proceed by induction on @xmath423 : the induction step is straightforward .
[ lem : not_terminal ] let @xmath293 be as in proposition [ terminalwpsmutationgraph ] , let @xmath467 , let @xmath297 , and let @xmath554
. then @xmath524 , and : @xmath555 where @xmath317 denotes the fractional part of @xmath318 .
mohammad akhtar and alexander m. kasprzyk , _ mutations of fake weighted projective planes _ , to appear in proc .
edinburgh math .
( 2 ) , http://arxiv.org/abs/1302.1152[`arxiv:1302.1152 [ math.ag ] ` ] , 2013 . gennadiy averkov , jan krmpelmann , and benjamin nill , _ largest integral simplices with one interior integral point : solution of hensley s conjecture and related results _ , http://arxiv.org/abs/1309.7967[`arxiv:1309.7967 [ math.co ] ` ] , 2013 .
victor v. batyrev , ionu ciocan - fontanine , bumsig kim , and duco van straten , _ conifold transitions and mirror symmetry for calabi - yau complete intersections in grassmannians _ , nuclear phys .
b * 514 * ( 1998 ) , no . 3 , 640666 . a. r. iano - fletcher , _ working with weighted complete intersections _
, explicit birational geometry of 3-folds , london math .
lecture note ser .
281 , cambridge univ . press , cambridge , 2000 , pp .
101173 .
alexander m. kasprzyk and benjamin nill , _ fano polytopes _ , strings , gauge fields , and the geometry behind the legacy of maximilian kreuzer ( anton rebhan , ludmil katzarkov , johanna knapp , radoslav rashkov , and emanuel scheidegger , eds . ) , world scientific , 2012 , pp . | we characterise mutations between fake weighted projective spaces , and give explicit formulas for how the weights and multiplicity change under mutation .
in particular , we prove that multiplicity - preserving mutations between fake weighted projective spaces are mutations over edges of the corresponding simplices . as an application , we analyse the canonical and terminal fake weighted projective spaces of maximal degree . |
orientational models with disorder and elastic degrees of freedom have been used to investigate phase transitions in a number of systems , as the mixed molecular crystal k(cn)@xmath0br@xmath1 @xcite .
the work on quadrupolar glass models in the presence of random strain fields , which are supposed to mimic the random mismatching of molecular groups , was reviewed by binder and reger @xcite .
a compressible spherical model , with random bonds and in random fields , has been introduced to account for the peculiar behavior of compositionally disordered perovskites , also known as relaxor ferroelectrics @xcite .
disordered ising models , with random competing interactions and elastic degrees of freedom , have also been used to account for the phase diagram and the glassy transition in mixtures of ferro and antiferroelectric hydrogen - bonded crystals of the kdp family @xcite .
more recently , random quadrupolar models in the presence of anisotropic strain fields have been studied in the context of nematic liquid crystal elastomers @xcite .
these investigations on disordered compressible models provided the motivation to introduce a simple compressible sherrington - kirkpatrick ( sk ) ising spin - glass model @xcite , and use standard techniques to obtain the global phase diagram in terms of pressure and temperature .
the investigation of compressible ising models has a long history domb56,salinas75,bergman76 .
mean - field and renormalization - group calculations indicate that the inclusion of elastic degrees of freedom may change the nature of the continuous ising transition .
an ising antiferromagnet on the triangular lattice , with properly chosen elastic degrees of freedom , is known to provide a mechanism to explain the formation of striped phases chen86,gu96 .
although elastic spin models with uniform interactions have been much investigated , disordered compressible spin models are less explored .
nowadays it has been feasible to carry out detailed computer simulations for compressible models @xcite , including disordered compressible models , as in the work of marshall for a compressible edwards - anderson ising spin - glass on a two - dimensional lattice @xcite .
these works have provided further motivation to study the phase diagram of the simple compressible sk spin - glass model .
we treat the mean - field sk system in the pressure ensemble , which is more adequate to analyze the phase diagram in terms of the intensive field variables , temperature @xmath2 and pressure @xmath3 . using the replica method ,
the calculation of the free energy is reduced to the minimization of a functional of three sets of replica variables @xcite .
we obtain a number of results in the replica - symmetric approximation .
in particular , we show that the presence of magneto - elastic couplings introduces a simple shift in the border between paramagnetic and spin - glass phases , and that there appears a tricritical point along the para - ferromagnetic border .
also , we perform an analysis of stability of the replica - symmetric solution and locate the de almeida - thouless line @xcite . in the conclusions ,
we consider a compressible two - sublattice sk model to make contact with the experimental phase diagrams of mixtures of ferro and antiferroelectric hydrogen - bonded crystals .
the compressible sherrington - kirkpatrick spin - glass model is defined by the hamiltonian@xmath4 s_{i}s_{j}-h\sum_{i=1}^{n}s_{i}+\frac{1}{2}kn\left ( v - v_{0}\right ) ^{2},\]]where @xmath5 , with @xmath6 , @xmath7 is the specific volume , @xmath8 is a constant parameter , @xmath9 is the magneto - elastic coupling , @xmath10 is the external field , and @xmath11 is a uniform elastic constant . as usual
, @xmath12 is a set of independent and identically distributed random variables , with suitably scaled mean values , @xmath13 , and variances , @xmath14 .
given a configuration @xmath15 , we write the partition function in the pressure ensemble,@xmath16where @xmath3 is the pressure , @xmath17 , and the trace is a sum over spin configurations . performing the volume integration , and discarding irrelevant terms in the thermodynamic limit , we have @xmath18 \mathbf{tr}\exp \left ( -\beta \mathcal{h}_{eff}\right ) , \]]with an effective four - spin hamiltonian,@xmath19using a gaussian identity , we write the partition function as@xmath20 \int_{-\infty } ^{+\infty } dx\exp \left ( -\frac{n\beta \gamma ^{2}j^{2}}{2k}x^{2}\right ) \nonumber \\ \times \mathbf{tr}\exp \left [ \beta \left ( 1+\frac{\gamma p}{k}+\frac{j\gamma ^{2}x}{k}\right ) \sum_{i
< j}j_{ij}s_{i}s_{j}+\beta h\sum_{i}s_{i}\right ] , \end{aligned}\ ] ] which is in a more convenient form to be treated by replicas . according to the replica method,@xmath21where@xmath22 \int_{-\infty } ^{+\infty }
\prod_{\alpha = 1}^{n}dx^{\alpha } \exp \left [ -\frac{n\beta \gamma ^{2}j^{2}}{2k}\,\sum_{\alpha } \left ( x^{\alpha } \right ) ^{2}\right ] \nonumber \\
\times \left\langle \mathbf{tr}_{n}\exp \left [ \sum_{i < j}j_{ij}\left ( \beta \sum_{\alpha } \xi ^{\alpha } s_{i}^{\alpha } s_{j}^{\alpha } \right ) + \beta h\sum_{i,\alpha } s_{i}^{\alpha } \right ] \right\rangle , \end{aligned}\]]with the definition@xmath23taking the average over the exchange configurations , and discarding irrelevant terms in the thermodynamic limit , we have@xmath22 \int_{-\infty } ^{+\infty } \prod_{\alpha } dx^{\alpha } \exp \left [ -\frac{n\beta \gamma ^{2}}{2k}\,\sum_{\alpha } \left ( x^{\alpha } \right ) ^{2}\right ] \nonumber \\
\fl \times \mathbf{tr}_{n}\exp \left [ \frac{\beta ^{2}j^{2}}{2n}\sum_{\alpha , \beta } \xi ^{\alpha } \xi ^{\beta } \sum_{i < j}s_{i}^{\alpha } s_{i}^{\beta } s_{j}^{\alpha } s_{j}^{\beta } + \frac{\beta j_{0}}{n}\sum_{\alpha } \xi ^{\alpha } \sum_{i < j}s_{i}^{\alpha } s_{j}^{\alpha } + \beta h\sum_{i,\alpha } s_{i}^{\alpha } \right ] , \end{aligned}\]]which can be rewritten as@xmath22 \int_{-\infty } ^{+\infty } \prod_{\alpha } dx^{\alpha } \exp \left\ { \,\sum_{\alpha } \left [ -\frac{n\beta j^{2}\gamma ^{2}}{2k}\left ( x^{\alpha } \right ) ^{2 } \nonumber \right .
+ \frac{n\beta ^{2}j^{2}}{4}\left ( \xi ^{\alpha } \right ) ^{2}\right ] \right\ } \mathbf{tr}_{n}\exp \left\ { \frac{\beta ^{2}j^{2}}{2n}\sum_{\alpha < \beta } \xi ^{\alpha } \xi ^{\beta } \left ( \sum_{i}s_{i}^{\alpha } s_{i}^{\beta } \right ) ^{2 } \nonumber \right .
+ \frac{\beta j_{0}}{2n}\sum_{\alpha } \xi ^{\alpha } \left ( \sum_{i}s_{i}^{\alpha } \right ) ^{2}+\beta h\sum_{i,\alpha } s_{i}^{\alpha } \right\ } .\end{aligned}\ ] ] we now use gaussian identities to introduce the new set of variables @xmath24 , associated with the magnetization , and @xmath25 , associated with the overlap between replicas , so that @xmath26 , \]]where@xmath27 in the thermodynamic limit , the free energy comes from @xmath28where the maximum should be taken with respect to the three sets of replica variables .
in the replica - symmetric solution , with @xmath29 and @xmath30 , for all @xmath31 , and @xmath32 , for all pairs @xmath33 , we have @xmath34 , \end{aligned}\]]with@xmath35the equations of state are given by@xmath36 , \label{q}\ ] ] @xmath37 , \label{m}\ ] ] and@xmath38 we now restrict the analysis to zero field ( @xmath39 ) .
the transition between the spin - glass and the paramagnetic phase comes from the expansions of equations ( [ q ] ) and ( [ x ] ) , @xmath40and @xmath41there is a ( positive ) solution , @xmath42 , for@xmath43introducing the notation@xmath44and using the definition of @xmath45 , given by eq .
( [ def ] ) , we have the second - order boundary between the paramagnetic and the spin - glass phases,@xmath46 in zero field , the calculation of the para - ferromagnetic transition comes from the analysis of the expansion@xmath47which can be written as@xmath48introducing the dimensionless and more compact notation@xmath49eq .
( [ xx ] ) can be rewritten as@xmath50inserting this expansion into the equations of state , it is easy to show that@xmath51from which we have the second - order para - ferromagnetic border,@xmath52 the transition between the spin - glass and the ferromagnetic phases may be calculated from an analysis of the zero - field susceptibility .
it is not difficult to show that this border is given by @xmath53where @xmath54 comes from the equation@xmath55 .\ ] ] in order to investigate the eventual existence of tricritical points , we write the expansions@xmath56and@xmath57for @xmath58 , the coefficients of eq .
( [ q2 ] ) indicate that there is no possibility of a tricritical point along the second - order border between the paramagnetic and the spin - glass phases . at the para - ferromagnetic border
, however , it is not difficult to locate a tricritical point at@xmath59 let us turn to the hessian matrix associated with the replica - symmetric solutions of @xmath60 .
there is a total of thirteen distinct elements in the @xmath61 limit , seven of which coming from derivatives involving the @xmath62 variables . as in the standard sk model ,
the paramagnetic phase is always stable , and the eigenvector subspace is generated by three classes of vectors , which have the general form @xmath63 where @xmath64 , @xmath65 and @xmath66 are column - vectors with @xmath67 , @xmath67 and @xmath68 elements respectively @xcite .
the eigenvalue spectrum is obtained straightforwardly following the same steps of the rigid model . in the ordered region , the change of sign of the transverse eigenvalue leads to an instability .
the de almeida - thouless line is given by@xmath69 , \end{aligned}\]]where @xmath39 , and @xmath70 , @xmath54 , and @xmath71 , are the equations of state .
it is interesting to draw a typical phase diagram in terms of temperature @xmath72 and the ratio @xmath73 ( in zero field ) .
the dotted lines in figure [ figure1 ] are the well - known replica - symmetric second - order boundaries , with the addition of the de almeida - thouless ( at ) instability line , for the rigid sk model ( with no magneto - elastic coupling ) .
the location of these borders and of the at line is shifted in the presence of a magneto - elastic coupling .
the spin - glass and ferromagnetic regions are broadened as a result of a decrease of the effective energy of the system due to the magneto - elastic coupling , as seen in equation ( [ heff ] ) .
the solid and dashed lines were drawn for the parameter values @xmath74 and @xmath75 .
for these particular values of pressure and elastic parameters , the second - order boundary between the paramagnetic and the ferromagnetic phases , given by @xmath76 , ends at a tricritical point , at @xmath77 , beyond which there is a ( dashed ) line of first - order transitions .
these are typical results for the compressible sk model in the replica - symmetric approximation . with the exception of the well - known corrections in the spin - glass region @xcite , in analogy to the treatment of the rigid sk model
, the form of the effective four - spin hamiltonian in the pressure ensemble does not indicate any drastic qualitative changes of the phase diagrams even if we go beyond the simple replica - symmetric approximation .
we now make contact with previous work for mixtures of ferroelectric and antiferroelectric hydrogen - bonded crystals of the kdp family .
a disordered ising model in a random field , with the inclusion of elastic degrees of freedom , has been used to investigate the glassy transition in rb@xmath1(nh@xmath78)@xmath79h@xmath80aso@xmath81 , known as rada @xcite .
also , the phase diagram of rb@xmath1(nh@xmath81)@xmath79h@xmath80po@xmath81 ( radp ) , in terms of temperature @xmath2 and concentration @xmath71 , has been drawn on the basis of a cluster calculation for a model of ising pseudo - spins @xcite .
the rubidium crystals , rbh@xmath80po@xmath81 ( rdp ) and rbh@xmath80aso@xmath81 ( rda ) , display a ferroelectric transition , at about @xmath82 @xmath83 , similar to the well - known ferroelectric phase transition in the isomorphous kh@xmath80po@xmath81 ( kdp ) crystals .
the ammonium crystals , nh@xmath81h@xmath80aso@xmath81 ( ada ) and nh@xmath81h@xmath80po@xmath81 ( adp ) , display a strong first - order antiferroelectric transition . at the mean - field level
, the description of the @xmath84 phase diagrams requires the consideration of two sublattices .
we thus use a two - sublattice compressible sk model , given by the hamiltonian @xmath85 s_{i}s_{j}+\frac{1}{2}kn\left ( v - v_{0}\right ) ^{2},\end{aligned}\ ] ] where the sum refers to all distinct pairs of spins belonging to different sublattices , @xmath86 and @xmath87 .
the random variables @xmath15 are associated with a gaussian distribution , with suitably scaled moments , @xmath13 , and @xmath88 , where we choose @xmath89 and make @xmath90 , in order to mimic ferroelectric ( @xmath91 ) and antiferroelectric ( @xmath92 ) transitions .
two - sublattice sherrington - kirkpatrick models , even in the presence of random fields , have been studied in connection with transitions from antiferromagnetic to spin - glass phases @xcite .
the addition of elastic degrees of freedom , according to the steps of the previous sections , suggests a typical phase diagram as sketched in figure [ figure2 ] . with a convenient choice of the parameters ,
it is possible to predict an elastically induced first - order transition from the paraelectric to the antiferroelectric phase for @xmath93 , turning into second order for smaller values of concentration , and which is in qualitative agreement with the experimental findings .
the glassy transition can still be understood within the framework of the rigid sk models , with the well - known instability of the replica - symmetric solution in the low - temperature region .
it is important to remark that the inclusion of gaussian random fields is not capable of explaining a first - order transition , in contrast to the predictions of the proposed compressible model @xcite .
in fact , first - order transitions are associated to either a double - delta distribution of random fields @xcite or to the consideration of certain discrete quadrupolar - glass and potts - glass models @xcite .
we acknowledge useful remarks of francisco a. da costa .
we also acknowledge the financial support from conselho nacional de desenvolvimento cientfico e tecnolgico ( cnpq ) . | we introduce a sherrington - kirkpatrick spin - glass model with the addition of elastic degrees of freedom .
the problem is formulated in terms of an effective four - spin hamiltonian in the pressure ensemble , which can be treated by the replica method . in the replica - symmetric approximation
, we analyze the pressure - temperature phase diagram , and obtain expressions for the critical boundaries between the disordered and the ordered ( spin - glass and ferromagnetic ) phases . the second - order para - ferromagnetic border ends at a tricritical point ,
beyond which the transition becomes discontinuous .
we use these results to make contact with the temperature - concentration phase diagrams of mixtures of hydrogen - bonded crystals . |
let @xmath0 be a smooth oriented compact surface .
the _ teichmller space _
@xmath1 is defined as the quotient space of the set of smooth integrable complex operators compatible with the orientation ( o.c . ) @xmath2 by @xmath3 , the connected component of the identity in the oriented diffeomorphism group @xmath4 of @xmath0 .
the theory of teichmller spaces is a cornerstone in complex variables and riemann surfaces .
originated by riemann himself and followed by the fundamental works of teichmller , ahlfors and bers , it has moreover implications in many branches of mathematics as algebraic geometry , hyperbolic geometry , complex dynamics , discrete groups , ... perhaps the most basic property of @xmath1 is that it has a natural structure of a complex manifold , making it a global moduli space of complex structures on @xmath0 .
moreover , the mapping class group of @xmath0 acts on @xmath1 and the resulting quotient is a complex orbifold .
this refined quotient coincides with the quotient of @xmath5 by the full group @xmath4 , the so - called riemann moduli space @xmath6 .
+ let now @xmath0 be a smooth oriented compact manifold of even dimension @xmath7 strictly greater than @xmath8 .
the teichmller and riemann moduli spaces can still be defined , but one now has to add the integrability condition in the definition @xmath9\subset t^{1,0}\}\ ] ] for @xmath10 although the literature about these higher dimensional teichmller and riemann moduli spaces is much less developed than that about surfaces , it has grown significantly in the last years and these spaces play an increasing role in complex geometry .
catanese s guide to deformations and moduli @xcite gives some general local properties of @xmath1 and contains many results on the teichmller space of minimal surfaces of general type . and
in the special case of hyperkhler manifolds , the teichmller space is used by verbitsky in a prominent way in his proof of a global torelli theorem @xcite and also to showing some important results on these manifolds @xcite .
however , the main difference with the case of surfaces is that @xmath1 and @xmath6 are _ just topological spaces _ and do not have any good geometric structure . only for special classes as that of hyperkhler manifolds ,
an analytic structure is known on @xmath1 , but even in this case , it is not hausdorff at all points .
perhaps the most dramatic example is given by @xmath0 being @xmath11 .
then @xmath6 , _ as a set _ , is @xmath12 , a point @xmath13 corresponding to the hirzebruch surface @xmath14 ( and each connected component of @xmath1 is @xmath15 , with @xmath16 and @xmath17 encoding the same surface , see examples [ hirzebruch ] and [ hirzebruchbis ] ) .
but , _ as a topological space _ , it is @xmath12 endowed with a non - hausdorff topology .
no two points are separated , as a consequence of the fact that @xmath14 can be deformed onto any @xmath18 with @xmath19 by an arbitrary small deformation .
equivalently , this comes from the fact that the dimension of the automorphism group of hirzebruch surfaces jumps .
moreover , in presence of this jumping phenomenon , @xmath1 and @xmath6 are not even locally hausdorff hence not locally isomorphic to an analytic space ( cf .
example [ hopfbis ] ) .
in fact , it is well - known since the fundamental works of kodaira - spencer and kuranishi that in higher dimension the global point of view must be abandoned for the local point of view ; and the teichmller space replaced with the kuranishi space which must be thought of as the best possible approximation in the analytic category for a local moduli space of complex structures . + nevertheless , to go further in the study of the higher dimensional teichmller and riemann spaces , it is fundamental to understand their global analytic structure .
as we can not expect a structure of analytic space , even a non - hausdorff one , _ we have to view these quotient spaces as stacks_. the aim of this paper is to investigate the structure of these stacks in higher dimension .
since we work with arbitrary complex structures and not only with projective ones , we consider them as analytic stacks , defined over the category of complex manifolds , or that of @xmath20-analytic spaces .
+ for surfaces of genus @xmath21 , the classical results can be reformulated in terms of stacks as : both @xmath1 and @xmath6 are deligne - mumford analytic stacks isomorphic to stacks of families of riemann surfaces of genus @xmath22 . moreover , all isotropy groups of @xmath1 are trivial whereas the isotropy group of @xmath6 at a point coincides with the automorphism group of this point . the case of genus @xmath23 is somewhat more complicated , because of the translations .
both @xmath24 and @xmath25 are artin analytic stacks , isomorphic to stacks of families of complex tori of dimension one .
+ in higher dimension , one has to build these stacks from the local data encoded in the kuranishi space .
one should expect that the local model of the riemann moduli stack , respectively of the teichmller stack , at a complex structure @xmath26 is the kuranishi space of @xmath27 endowed with the action of its automorphisms , respectively of the automorphisms which are @xmath28-isotopic to the identity .
+ this is however not correct .
a third element is missing .
some orbits of @xmath3 may a priori have a complicated geometry and accumulate onto @xmath26 .
this induces additional identifications to be done to pass from the kuranishi space to the riemann or teichmller space , even in the absence of automorphisms .
+ the main problem behind this stack construction is to understand how to glue the bunch of kuranishi spaces , in other words how to keep track of all identifications to be done not only on a single kuranishi space but also between different ones .
+ this is achieved here by describing the space of complex structures @xmath29 as a foliated space transversely modeled at @xmath26 on the translation groupoid encoding the local actions of the automorphisms on the kuranishi space of @xmath27 .
then , we describe the stacky structure of the leaf space . + a natural source of stacks is given by ( leaf spaces of ) foliations .
such stacks admit a presentation by an tale groupoid , the holonomy groupoid @xcite . in general
, the action of @xmath3 onto @xmath29 does not define a foliation , nor a lamination .
but we show that it defines a more complicated foliated structure , that we call a foliation transversely modeled on a translation groupoid , a tg foliation in short , see section [ multifoliate ] .
we then turn to the construction of an associated holonomy groupoid .
it is however much more involved than the classical construction and it constitutes the bulk of the paper .
indeed , the transverse structure of a tg foliation being well defined only up to morita equivalence , the holonomy morphisms are stacks morphisms and do not fit into a nice groupoid .
hopefully , in the case of the teichmller and the riemann stacks , we can do much better and define a holonomy groupoid which moreover enjoys nice properties .
from the one hand , objects and morphisms form a finite - dimensional analytic space , and from the other hand , source and target maps are smooth morphisms .
the teichmller and the riemann stacks admitting these nice groupoids as a presentation , we call them artin analytic stacks . the set of objects is a ( countable ) disjoint union of kuranishi spaces .
we describe completely the set of morphisms encoding the identifications of structures to be done on the collection of kuranishi spaces .
all this can be seen as an answer to question 1.8 of @xcite .
the only condition needed for this result to hold is that the dimension of the automorphism group of all structures of @xmath1 ( or @xmath6 ) is bounded by a fixed integer .
this is nevertheless a mild restriction since we may easily stratify @xmath29 into strata where this dimension is bounded .
we emphasize that @xmath0 can be any compact manifold and that we consider all complex structures and not only projective or khler ones modulo @xmath3 or @xmath4 . ] .
+ the paper is organized as follows .
some notations and classical facts about the kuranishi space are introduced in section [ notations ] .
we then give some general properties of @xmath29 in section [ f - homotopy ] , putting emphasis on connectedness properties , and introducing a graph , called the _ graph of @xmath30-homotopy_. the tg foliation of the space @xmath29 is introduced in section [ tgfoliationsection ] .
the technical core of the paper is constituted by sections [ multifoliate ] and [ teichmueller ] , where we perform the construction of the analogue for the holonomy groupoid .
we call it _ the teichmller groupoid_. to smoothe the difficulties of the construction , a sketch of it is given in section [ holonomy ] and a very simple case is treated in section [ rigidified ] .
all this culminates in the proof of the main theorem [ maintheorem ] , stating that the teichmller groupoid is an analytic smooth presentation of the teichmller stack .
analogous construction and statement for the riemann moduli stack are done in sections [ riemann ] and [ main ] .
we also show that both stacks are isomorphic to stacks of families of complex manifolds .
this is of course not a surprise but it is important to link the theory of teichmller and riemann moduli stacks to the classical deformation theory of kodaira - spencer .
complete examples are given in section [ examples ] .
+ we hope that this paper will serve as a source of motivation for studying global moduli problems in complex analytic geometry and their interplay with artin analytic stacks . from the one hand ,
every abstract result on these stacks might apply to moduli problems and increase our knowledge of complex manifolds . from the other hand ,
examples of teichmller stacks are an unending source of examples of artin analytic stacks , showing all the complexity and richness of their structure , far from finite dimensional group actions and leaf spaces .
let @xmath0 be a smooth ( i.e. @xmath28 ) oriented compact connected manifold of even dimension .
let @xmath31 , respectively @xmath29 , be the space of smooth almost complex , respectively complex operators on @xmath0 which are compatible with the orientation .
the definition of @xmath29 is given in .
we assume that both sets are non - empty .
+ we topologize @xmath31 as a frchet manifold locally modelled onto the smooth sections of a vector bundle over @xmath0 ( cf . @xcite for the encoding of structures , @xcite and @xcite for the frchet topology ) .
we denote by @xmath32 , respectively @xmath33 , a connected component of @xmath31 , respectively @xmath29 .
points of @xmath31 will be denoted generically by @xmath26 .
+ for @xmath34 a topological space , we denote by @xmath35 the set of connected components of @xmath34 .
the previous topology being countable , @xmath36 is a countable set .
+ the diffeomorphism group @xmath4 acts on the right on @xmath31 by pullback of almost complex operators .
it is a frchet lie group @xcite acting analyticallythere is some subtle point here because the complex structure of @xmath4 depends on the choice of a complex structure on @xmath0 .
we will just use the fact that , if we endow locally at identity @xmath4 with chart , then the map @xmath37 is analytic in a neighborhood of @xmath38 . ] onto @xmath31 .
this action preserves @xmath29 .
it is given by @xmath39 we focus on @xmath3 , the connected component of the identity in @xmath4 .
we define the mapping class group @xmath40 and we set @xmath41\ ] ] and @xmath42=\left [ \mathcal t(x)\big /\mathcal m\mathcal c(x)\right ] \ ] ] the brackets mean that we consider these quotient as stacks and not as topological spaces . the aim of the paper is to investigate the structure of these two stacks . here
we just review some definitions about groupoids and stacks .
we define a stack as in ( * ? ? ?
we insist on the fact that we work exclusively in the @xmath20-analytic context . to be more precise we consider only stacks defined over the category of complex manifolds or that of @xmath20-analytic spaces .
we also emphasize that we use the euclidean topology , not the tale or the analytic ones .
hence our families of coverings are just standard topological open coverings .
+ the groupoids we consider are analytic , that is that the set of objects and the set of morphisms are complex analytic spaces ( possibly non - hausdorff for the second one ) and all the structure maps are analytic morphisms .
we emphasize that a smooth analytic groupoid is not a complex lie groupoid , since we allow singularities of both the set of objects and the set of morphisms , but it is the exact singular counterpart of a complex lie groupoid , cf .
+ analytic groupoids are in particular topological so that it makes sense to localize them on an open covering of the set of objects @xcite .
the geometric quotient associated to such a groupoid is the topological space obtained by taking the quotient of the set of objects by the equivalence relation defined by the set of morphisms .
connected components of the groupoid refer to connected components of the geometric quotient .
+ such a groupoid is tale , respectively smooth , if both source and target maps are tale , respectively smooth , morphisms .
we call a stack tale analytic ( respectively artin analytic ) if it admits a presentation by an tale ( respectively smooth ) analytic groupoid ; deligne - mumford analytic if it is tale with finite stabilizers .
we take as definition of morita equivalence that given in @xcite , with the obvious adaptations to the groupoids we use ( e.g. replace @xmath28 map with @xmath20-analytic map , submersion with smooth morphism , ... ) .
+ by a translation groupoid , we mean a groupoid @xmath43 where @xmath44 is a lie group and where the source map is the second projection .
the standard example is that of @xmath44 acting on @xmath45 , including the case of a local action , i.e. @xmath46 is not defined for all @xmath47 .
but the definition is general enough to include also cases with no @xmath44-action , see subsection [ localaction ] .
[ oriented ] in the first version of this paper , we take for @xmath0 an unoriented smooth compact manifold and consider @xmath29 as the set of all integrable complex operators , regardless of orientation .
then @xmath1 is defined as in , and in , we have to replace the oriented diffeomorphism group by the full diffeomorphism group @xmath48 .
this does not change substantially these two sets , and our results apply to this setting .
in fact , the main drawback of forgetting the orientation is that the notion of teichmller space does not coincide to the classical one for surfaces . especially , the unoriented teichmller space of a compact surface has two connected components , corresponding to the two possible orientations . + more generally , if @xmath0 admits a diffeomorphism reversing orientation , then the unoriented teichmller space has twice more connected components as the classical one .
however , the two riemann spaces coincide . finally ,
if @xmath0 does not admit any orientation reversing diffeomorphism , then the unoriented teichmller and riemann spaces are the disjoint union of the classical ones for both orientations . notice that , in this last case , changing the orientation may completely change the teichmller and riemann spaces .
it is even possible that they become empty ( think of @xmath49 and @xmath50 . fix a riemannian metric on @xmath0 and let @xmath51 denote the exponential associated to this metric .
let @xmath52 and set @xmath53 [ notation ] to avoid cumbersome notations , we write @xmath54 for @xmath55 , and @xmath56 for @xmath57 , ... a complex chart for @xmath3 at @xmath58 is given by the map @xmath59 where @xmath60 is the @xmath20-vector space of @xmath61-vector fields of @xmath62 and @xmath63 a neighborhood of @xmath64 .
+ let @xmath65 be the sheaf of germs of holomorphic vector fields on @xmath62 . for @xmath66 , we consider the function @xmath67 we denote by @xmath68 the group of automorphisms of @xmath62 .
the connected component of the identity @xmath69 in @xmath68 is tangent to @xmath70 .
we define @xmath71 [ autattention ] be careful that equal to @xmath69 , cf .
section [ rigidified ] .
let @xmath72 . kuranishi s
theorem @xcite , @xcite , @xcite gives a finite dimensional local model for @xmath29 and the action of @xmath3 , namely * ( kuranishi , 1962)*. [ kuranishitheorem ] for any choice of a closed complex vector space @xmath73 such that @xmath74 there exists a connected open neighborhood @xmath75 of @xmath76 in @xmath29 , a finite - dimensional analytic subspace @xmath77 of @xmath75 and an analytic isomorphism ( onto its image ) @xmath78 such that 1 .
the inverse map is given by @xmath79 2 .
the composition of the maps @xmath80 is the identity
. indeed , kuranishi always uses the @xmath81-orthogonal complement to the space @xmath82 as @xmath73 .
however , it is easy to see that everything works with any other closed complementary subspace , cf .
@xcite .
[ sobolev ] theorem [ kuranishitheorem ] is proved using the inverse function theorem . to do that
, one extends @xmath31 to operators of sobolev class @xmath83 ( with @xmath84 big ) , so that @xmath31 becomes a hilbert manifold .
then one may use the classical inverse function theorem for banach spaces to obtain the isomorphism .
finally , because @xmath77 is tangent to the kernel of a strongly elliptic differential operator , then it only consists of @xmath28 operators and the isomorphism is still valid when restricting to @xmath28 operators , see @xcite , @xcite and @xcite for more details .
following @xcite , we call such a pair @xmath85 a _ kuranishi domain _ based at @xmath76 . moreover , we call @xmath86 the natural retraction map @xmath87 given @xmath88 , we denote by @xmath89 the kuranishi space of @xmath62 .
we use the same convention for @xmath90 as that stated for @xmath0 in remark [ notation ] .
[ kuniqueness ] it is a classical fact that @xmath89 is unique _ as a germ of analytic space_. however , in this paper , we consider @xmath89 as an analytic subspace of @xmath29 , not as a germ . by abuse of terminology , we nevertheless speak of _ the _ kuranishi space .
the complex lie group @xmath91 ( respectively @xmath92 ) is the isotropy group for the action of @xmath3 at @xmath76 ( respectively @xmath4 ) . choosing some closed complementary subset @xmath73 means killing the connected component of the identity @xmath93 in this isotropy group .
it acts on @xmath29 , and so locally on @xmath75 .
this action induces a local action of each @xmath23-parameter subgroup on @xmath77 . in other words ,
if @xmath22 is an element of @xmath92 , it defines an isomorphism of some neighborhood @xmath94 of @xmath76 in @xmath77 as follows @xmath95 in the same way , given @xmath96 in @xmath77 , there exists a connected neighborhood @xmath97 of @xmath98 in @xmath99 such that the map @xmath100 is analytic .
we encode these maps in the translation groupoid @xmath101 [ notanaction ] there is a subtle point which is worth to be remarked here .
although it is the case in many examples , the groupoid does not in general describe a local @xmath44-action .
this comes from the fact that there is no reason for @xmath102 to equal @xmath103 .
in particular , there is no reason for the isotropy groups of the groupoid to be subgroups of @xmath104 .
they are just submanifolds .
we notice that the associated stack is a local moduli space ( also called a universal deformation space ) for germs of deformations of @xmath54 .
this is a direct consequence of theorem [ kuranishitheorem ] and the following lemma .
[ gdecompolemma ] if @xmath63 is a small enough neighborhood of @xmath64 in @xmath73 , then there exist an open and connected neighborhood @xmath34 of the identity in @xmath105 and an open and connected neighborhood @xmath106 of the identity in @xmath107 such that @xmath108 is an isomorphism .
pass to vector fields and diffeomorphisms of sobolev class @xmath109 for some big @xmath84 and extend the map . since @xmath34 is constituted by holomorphic elements , this map is of class @xmath28 and a simple computation shows that its differential at @xmath110 is an isomorphism .
hence we may apply the local inverse theorem and get the result for this sobolev class . to finish with
, it is enough to remark that , since @xmath22 is holomorphic , @xmath111 is of class @xmath28 if and only @xmath112 is .
in other words , is morita equivalent to the translation groupoid @xmath113 , for @xmath75 defined as in .
[ localmodeldifference ] however , the translation groupoid @xmath114 , because some elements in @xmath115 may act on @xmath75 . rephrasing this important remark ,
the teichmller stack is not locally isomorphic to , cf .
remark [ catanese ] .
[ noncommutatif ] notice the order in .
if we consider the map @xmath116 , the above proof does not apply . indeed , this last map is not @xmath117 for vector fields and diffeomorphisms of sobolev class @xmath109 , cf .
* example i.4.4.5 ) .
we now want to investigate more thoroughly the structure of , taking into account the results of @xcite .
the groupoid defines a local foliation .
more precisely , the leaf through a point @xmath96 is the maximal connected subset of @xmath77 all of whose points encode @xmath96 up to isotopy .
it is locally generated by the map in the neighborhood of @xmath96 and has dimension @xmath118 now , consider the decomposition @xmath119 where @xmath120 is the tangent space to the isotropy group of at @xmath96 and @xmath121 is any complementary @xmath20-vector space .
this @xmath120 injects by definition in ( the tangent space of ) the automorphism group of @xmath122 so has dimension less than @xmath123 ; and , following what we just recalled , the @xmath121-orbit at @xmath96 is locally free at @xmath96 and included in the leaf of the foliation of @xcite through @xmath96 , so has dimension less than .
but because of , we must have @xmath124 and we just proved [ autfol ] we have : 1 .
the vector space @xmath120 is tangent to aut@xmath125 at identity .
2 . the local orbit of @xmath121 at @xmath96 defines a plaque of the foliation of @xmath126 defined in @xcite .
observe that kuranishi s theorem [ kuranishitheorem ] implies that @xmath29 is locally @xmath28-pathwise connected in @xmath31 .
therefore , [ countable1 ] we have : 1 .
there are at most a countable number of connected components of @xmath29 in each @xmath32 .
every connected component of @xmath29 is @xmath28-pathwise connected . and [ countableteich ]
the teichmller and riemann moduli stacks have at most a countable number of connected components . moreover , 1 . the natural projection map from @xmath29 onto @xmath1
induces a bijection @xmath127 2 .
the mapping class group @xmath128 acts on both @xmath129 and @xmath130 .
3 . passing to the quotient by the mapping class group @xmath128 , the bijection descends as a bijection @xmath131 just use proposition [ countable1 ] and the fact that @xmath3 preserves the components of @xmath29 . for further use
, we let @xmath132\in\mathcal m\mathcal c(x)\longmapsto [ \mathcal i_0\cdot \phi]\in\pi_0(\mathcal i)\ ] ] denote the map given by the action of the mapping class group onto a fixed component @xmath33 .
[ countableexamples ] for surfaces , the number of connected components of @xmath6 , that is the number of connected components of @xmath29 up to the action of the mapping class group , is finite as soon as it contains a projective manifold @xcite . however , it may be more than one , see @xcite . in dimension
@xmath133 , there are examples of manifolds with @xmath6 having an infinite number of connected components , as @xmath134 for @xmath135 , see @xcite , or the product of a k3 surface with @xmath136 , see @xcite .
+ however , we do not know of any example with an infinite number of connected components of @xmath6 and thus of @xmath29 _ inside a single connected component of @xmath31 _ ( in the above examples , each connected component of @xmath31 contains a single connected component of @xmath29 )
. in particular , it would be interesting to find : [ ccpb ] find a @xmath28 compact manifold @xmath0 with @xmath137 connected and @xmath5 having an infinite number of connected components .
probably , @xmath138 for @xmath135 give such an example .
in particular , it is proven in @xcite that @xmath31 has a single connected component . and
the structures of @xcite should give the countably many connected components of @xmath29 .
since they have pairwise not biholomorphic universal covers , this should give the countably many connected components of @xmath6 .
but proving this is the case seems to be out of reach for the moment .
observe that the first step in showing this result would be to establish that any deformation _ in the large _ of a hopf manifold is a hopf manifold , which is still an open problem as far as we know .
+ the case of surfaces is somewhat different , see remark [ cc2 ] .
recall that kodaira and spencer defined in @xcite the notion of _ @xmath139-homotopy_. taking into account kuranishi s theorem , it turns out that we may equivalently define it by saying that @xmath140 and @xmath141 are @xmath139-homotopic if there exists a smooth path in @xmath29 joining them .
that is if they belong to the same connected component @xmath33 .
similarly , we define let @xmath96 and @xmath142 be two points of the same @xmath33 .
then we say that they are _
@xmath30-homotopic _ if there exists a smooth path in @xmath33 joining them such that the function @xmath143 is constant along it
. recall also that , if @xmath90 denotes the kuranishi space of some @xmath76 , then for any @xmath144 , the sets @xmath145 are analytic subspaces of @xmath90 , cf . @xcite . using kuranishi s theorem
, we immediately obtain that the sets @xmath146 are analytic subspaces as in remark [ sobolev ] to have that @xmath29 and @xmath147 are banach analytic spaces in the sense of @xcite .
] of @xmath29 .
observe that @xmath147 is the union of all @xmath30-homotopy classes whose @xmath143 is greater than or equal to @xmath139 .
+ the analyticity of comes indeed from the fact that the function @xmath143 is upper semi - continuous for the zariski topology , see @xcite .
but this also implies [ countable2 ] there are at most a countable number of @xmath30-homotopy classes in each @xmath33 .
define a weighted and directed graph as follows .
each @xmath30-homotopy class @xmath148 of @xmath29 corresponds to a vertex with weight equal to @xmath149 for @xmath150 .
two vertices @xmath151 and @xmath152 are related by an oriented edge if there exists a smooth path @xmath139 in @xmath29 such that 1 .
the structure @xmath153 belongs to @xmath151 .
2 . for @xmath154
, the structure @xmath155 belongs to the class @xmath152 . observe that the edge is directed from the highest weight to the lowest weight .
[ graph ] the previous graph is called _ the graph of @xmath30-homotopy of @xmath29_. [ graphproperties ] the graph of @xmath30-homotopy has the following properties : 1 .
it has at most a countable number of connected components .
moreover , there is a @xmath156 correspondence between these connected components and the connected components of @xmath29 .
it has at most a countable number of vertices .
each vertex is attached to at most a countable number of edges .
there is no directed loop .
every directed path is finite . items ( i ) , ( ii ) and ( iii ) come from proposition [ countable1 ] , proposition [ countable2 ] and the definitions ; items ( iv ) and ( v ) come from the fact that the weights are strictly decreasing along an edge .
the group @xmath128 acts on the graph of @xmath30-homotopy .
we detail in the following proposition some trivial properties of this action .
[ graphaction ] the action of @xmath128 onto the graph of @xmath30-homotopy 1
. sends a connected component onto a connected component .
2 . sends a vertex to a vertex of same weight .
3 . respects the number and the orientation of the edges attached to a vertex .
hence , the existence of diffeomorphisms acting non trivially on the graph implies strong properties of symmetry of the graph .
if some @xmath30 sends a connected component of @xmath29 onto a different one , then these two connected components must be completely isomorphic .
* hopf surfaces . *
[ hopf ] let @xmath157 . by classical results of kodaira @xcite , @xcite , every complex surface diffeomorphic to @xmath0
is a ( primary ) hopf surface .
there is only one connected component of complex structures up to action of the mapping class group , since any hopf surface is @xmath139-homotopic to any other one , see @xcite .
the mapping class group of @xmath0 is a non trivial group it was pointed out to me by a. hatcher that no mapping class group of a closed @xmath158-manifold seems to be known . ] .
indeed , observe that it contains at least the elements @xmath159 for @xmath160 and @xmath161 a homotopically non trivial loop in @xmath162 , since both have non trivial action in homology . even without knowing the mapping class group
, we can characterize its action on @xmath29 .
following @xcite , we separate hopf surfaces into five classes namely classes iv , iii , iia , iib and iic .
[ hopfcc ] let @xmath30 be a diffeomorphism of @xmath0 .
assume that @xmath30 preserves the connected components of @xmath29 .
then @xmath30 is @xmath28-isotopic to the identity .
let @xmath76 represent a hopf surface of type iic , that is associated to a contracting diagonal matrix @xmath163 with @xmath164 .
+ assume that @xmath165 belongs to the same connected component as @xmath76 .
then there exists a smooth family of hopf surfaces @xmath166 $ ] with endpoints @xmath54 and @xmath167 . by (
* theorem 7.2 ) , there exists an analytic space @xmath90 , obtained by gluing together a finite number of kuranishi spaces of hopf surfaces ( up to taking the product with some vector space ) such that the family @xmath168 maps onto a smooth path into @xmath90 . using the description of the kuranishi spaces of hopf surfaces in ( * ? ? ?
* theorem 2 ) , it is easy to check that 1 .
@xmath90 is a manifold .
the points of @xmath90 encoding the type iia hopf surfaces belongs to a submanifold of codimension @xmath8 .
hence , by transversality , we may replace the initial path with a path with same endpoints and such that all surfaces along this path are linear , that is not of type iia .
a neigborhood of the path in @xmath90 is then obtained by gluing together open subsets of @xmath169 and we may assume that the deformation is isomorphic to the quotient of @xmath170 $ ] by the action generated by @xmath171 for @xmath172 a smooth map from @xmath173 $ ] into @xmath169 which is equal to at @xmath64 .
in particular , this means that @xmath174 is conjugated to by , say , @xmath45 . hence the map @xmath175 induces a biholomorphism between @xmath54 and @xmath167 , which is smoothly isotopic to the identity .
composing @xmath30 with the inverse of this biholomorphism , this gives an automorphism of @xmath54 which corresponds to the same element of the mapping class group as @xmath30 .
+ since every automorphism of every hopf surface is isotopic to the identity ( cf .
@xcite where all the automorphism groups are described ) , we are done . from lemma [ hopfcc ]
, we deduce that @xmath29 decomposes into several identical connected components that are exchanged by action of the mapping class group .
in particular , [ hopfcc2 ] the map is a @xmath156 correspondence between the mapping class group of @xmath0 and the set of connected components of @xmath29 . since all hopf surfaces are @xmath139-homotopic , is surjective . and it is injective by lemma [ hopfcc ] .
let us focus on one of the connected components .
it corresponds to a graph with an infinite number of vertices : one of weight @xmath158 ( class iv ) , one of weight @xmath133 for each value of @xmath176 ( class iii of weight @xmath177 ) and one of weight @xmath8 ( classes iia , iib and iic together ) .
there is an edge joining @xmath158 to @xmath8 and one joining @xmath133 to @xmath8 for each value of @xmath177 .
there is no edge from @xmath158 to any vertex @xmath133 because it is not possible to deform a hopf surface of class iv onto one of class iii without crossing the @xmath30-homotopy class of weight @xmath8 .
in the same way , there is no edge between two different vertices of weight @xmath133 , because every @xmath139-homotopy from a hopf surface of type iii with weight @xmath177 to a hopf surface of type iii with weight @xmath178 must pass through type ii hopf surfaces . in figure [ hopffig ]
, we draw the graph in a synthetic way .
the vertex @xmath179 encodes indeed the uncountable set of vertices of weight @xmath133 labelled by @xmath176 .
the single edge from @xmath133 to @xmath8 remembers all the edges from vertices @xmath133 of label @xmath177 onto the vertex @xmath8 .
-homotopy for hopf surfaces.,scaledwidth=30.0% ] using the five classes of hopf surfaces , one obtains a graph of small deformations which is more precise and complicated than the graph of @xmath30-homotopy , see @xcite , p.31 .
the graph of @xmath30-homotopy must be considered as a very rough decomposition of @xmath29 . * hirzebruch surfaces .
* [ hirzebruch ] consider @xmath180 .
it admits complex structures of even hirzebruch surfaces @xmath14 . by @xcite ,
this exhausts the set of complex surfaces diffeomorphic to @xmath0 .
then there is only one connected component of complex structures up to action of the mapping class group .
the mapping class group is not known ( cf .
footnote [ hatcher ] ) but contains at least four elements generated by @xmath181 where @xmath16 is the antipodal map of @xmath136 .
analogously to lemma [ hopfcc ] and corollary [ hopfcc2 ] , we have [ hirzebruchcc ] let @xmath182 be a diffeomorphism of @xmath0 .
assume that @xmath182 preserves the connected components of @xmath29 .
then @xmath182 is @xmath28-isotopic either to @xmath22 or to the identity .
let @xmath76 represent @xmath183 .
assume that @xmath184 belongs to the same connected component as @xmath76 .
then there exists a smooth family of hirzebruch surfaces @xmath185 $ ] with endpoints @xmath54 and @xmath186 . by (
* theorem 7.2 ) , there exists an analytic space @xmath90 , obtained by gluing together a finite number of kuranishi spaces of hirzebruch surfaces ( up to taking the product with some vector space ) such that the family @xmath168 maps onto a smooth path into @xmath90 . using the description of the kuranishi spaces of hirzebruch surfaces in @xcite
( see also example [ hirzebruchbis ] ) , it is easy to check that 1 . @xmath90 is a manifold .
2 . the points of @xmath90 encoding @xmath14 for @xmath187 belongs to a submanifold of codimension @xmath23 .
hence , by transversality , we may replace the initial path by a path with same endpoints and such that all surfaces along this path are biholomorphic to @xmath183 . by fischer - grauert s theorem ( see @xcite for the version we use ) , such a deformation is locally trivial , hence trivial since the base is an interval , i.e. there exists a smooth isotopy @xmath188)\ ] ] in particular , @xmath189 induces a biholomorphism between @xmath54 and @xmath190 , which is smoothly isotopic to the identity .
composing its inverse with @xmath182 , this gives an automorphism of @xmath54 , that is of @xmath183 , which corresponds to the same element of the mapping class group as @xmath182 .
comparing with the automorphism group of @xmath183 yields the result . and [ hirzebruchcc2 ]
the map is surjective with kernel @xmath191 , [ g]\}$ ]
. since all hirzebruch surfaces are @xmath139-homotopic , is surjective .
lemma [ hirzebruchcc ] gives the kernel .
now , fix a connected component @xmath33 .
we want to describe it more precisely .
observe that @xmath22 corresponds to an automorphism of @xmath183 , but not of the other hirzebruch surfaces since every automorphism of @xmath14 is isotopic to the identity for @xmath187 .
this implies [ gaction ] we have : 1 .
the subset @xmath192 of @xmath33 consisting of structures biholomorphic to @xmath183 is open and connected . 2 .
the closed set @xmath193 has exactly two connected components .
the diffeomorphism @xmath22 acts on @xmath33 by fixing globally @xmath192 ; and by exchanging the two components of @xmath193 .
4 . fix a connected component @xmath194 of @xmath193 .
then the set of points
@xmath195 encoding @xmath196 in @xmath194 is open and connected and its complement is connected .
5 . by induction , for @xmath197 , the set of points @xmath198
encoding @xmath14 in @xmath199 is open and connected and its complement is connected .
observe that @xmath192 is equal to @xmath200 ( we use the notation for @xmath33 ) .
hence it is open .
also we have already observed in the proof of lemma [ hirzebruchcc ] that two @xmath139-homotopic structures both encoding @xmath183 are @xmath139-homotopic through a path all of whose points encode @xmath183 .
this proves ( i ) .
+ to prove ( ii ) and ( iii ) , we need a variation of lemma [ hirzebruchcc ] .
let @xmath76 represent @xmath196 .
call @xmath194 the connected component of @xmath76 in @xmath193 .
assume that @xmath201 belongs to @xmath194 .
then there exists a smooth family of hirzebruch surfaces @xmath185 $ ] with endpoints @xmath54 and @xmath190 and all of whose point are distinct from @xmath183 . using theorem 1 of @xcite and the description of the kuranishi spaces of hirzebruch surfaces in @xcite , p.21 ( see also example [ hirzebruchbis ] ) , it is easy to check that we may assume that all surfaces along this path are biholomorphic to @xmath196 .
arguing as in the proof of lemma [ hirzebruchcc ] , we deduce that @xmath182 must be smoothly isotopic to the identity , since every automorphism of @xmath196 has this property . since we already know that @xmath22 fixes globally @xmath33 , this means that @xmath76 and @xmath202 belongs to two distinct connected components of @xmath193 in @xmath33 .
+ assume now that @xmath96 is another point of @xmath33 encoding @xmath196 .
then there exists @xmath203 such that @xmath96 equals @xmath201 . by corollary [ hirzebruchcc2 ]
, @xmath182 is either isotopic to the identity or to @xmath22 . in the first case
, @xmath96 belongs to also to @xmath194 . in the second case
, it belongs to @xmath204 .
hence , there are exactly two connected components exchanged by @xmath22 , and items ( ii ) and ( iii ) are proved . + finally , similar arguments prove ( iv ) and ( v ) .
-homotopy for hirzebruch surfaces.,scaledwidth=70.0% ] in other words , the associated graph of @xmath30-homotopy has several connected components and each connected component has two branches joined on the vertex corresponding to @xmath183 .
finally , each branch has a countable number of vertices , namely one vertex for each value of @xmath13 .
it has weight @xmath205 , except for @xmath206 which has weight @xmath207 .
given any @xmath208 , there exists an edge from @xmath16 to @xmath209 , because it is possible to deform @xmath14 onto @xmath18 , cf .
@xcite or @xcite .
in particular , every vertex is attached to a countable number of edges .
similar picture is valid for the odd hirzebruch surfaces .
[ ell ] observe that , in the case of elliptic curves , there is a single connected component of complex structures which is fixed by all elements of the mapping class group , although most of them can not be realized as automorphism of some complex structure , cf .
example [ tori ] .
this is completely opposite to the phenomenon described in lemmas [ hopfcc ] and [ hirzebruchcc ] .
[ cc2 ] observe that in examples [ hopf ] and [ hirzebruch ] , the riemann moduli stack @xmath6 is connected because of lemmas [ hopfcc ] and [ hirzebruchcc ] ( cf .
remark [ countableexamples ] ) .
however , we do not know if @xmath1 has a finite number of connected components , because it is not known if the mapping class group of @xmath210 , respectively @xmath11 , is finite or not .
for example , notice that some blow ups of connected sums of @xmath49s have infinite mapping class group , see @xcite .
let @xmath33 be a connected component of @xmath29 .
assume that for all @xmath26 in @xmath33 , we have @xmath149 equal to zero .
then , the action of @xmath3 onto @xmath33 is locally free and one would like to conclude that it defines a foliation of @xmath33 .
this can be made precise as follows .
[ foliationsimple ] assume that the function @xmath143 is identically zero on the connected component @xmath33 .
then , the action of @xmath107 onto @xmath33 induces a holomorphic foliation of @xmath33 whose leaves are frchet submanifolds and whose local transverse section at a point @xmath26 is given by the kuranishi space of @xmath62 .
[ foldef ] be careful that we use the word `` foliation '' in an extended sense .
firstly the leaves are infinite - dimensional and secondly the transverse sections are singular spaces and are not all isomorphic .
we should rather talk of `` lamination '' but we prefer to reserve this terminology for foliated spaces transversely modeled onto a continuous space , e.g. a cantor set .
the condition that the function @xmath143 is zero on the whole @xmath33 implies that , in theorem [ kuranishitheorem ] , we may take @xmath73 to be the full @xmath60 .
this complex vector space is , as a real vector space , the space of vector fields @xmath211 .
its complex structure a priori depends on the base point @xmath26 , but it is easy to check that all @xmath212 are isomorphic as complex vector spaces , ( * ? ? ?
* lemma 7.1 ) . hence the isomorphisms form a foliated atlas of @xmath33 : the plaques representing the local orbits of @xmath3 are preserved by the changes of charts , cf .
the leaves are frchet submanifolds modeled onto @xmath60 and at a point @xmath26 , any germ of transverse section is isomorphic to the kuranishi space of @xmath62 . in this section
, we would like to extend proposition [ foliationsimple ] and show that there still exists a more general type of foliated structure on @xmath33 .
[ tgfoliation ] let @xmath45 be an analytic space ( finite or infinite - dimensional ) .
we say that @xmath45 admits a _ foliation transversely modeled onto a translation groupoid _ , in short a _ tg foliation _ , if there exists 1 .
an open covering @xmath213 of @xmath45 , 2 .
analytic charts @xmath214 for @xmath215 an open ball and @xmath216 an analytic space ( both of finite or infinite dimension ) , 3 .
translation groupoids @xmath217 for connected complex finite dimensional lie groups @xmath218 such that the changes of charts send a plaque @xmath219 onto a plaque @xmath220 and induce isomorphisms of the transverse stacks .
the leaves of a tg foliation are the connected components of sequences of plaques @xmath219 glued through changes of charts . at a point @xmath26 , the transverse section is the groupoid @xmath217 .
the last requirement is that two such groupoids corresponding to two different charts are morita equivalent . in particular
, this means that the transverse structure is well defined only up to morita equivalence .
hence a tg foliation has manifolds as leaves and stacks @xmath221 $ ] as transverse sections .
we emphasize that there are two requirements in the definition : the geometric one allowing to define the leaves as manifolds , and the categorical one , keeping track of the stack nature of the transversals , see example [ locallytrivialaction ] .
standard foliations give examples of tg foliations , but these are of course `` trivial '' examples , with all the @xmath218 being the identity and no needs to refer to groupoids and stacks .
another example is given by a lie group @xmath44 acting holomorphically on a complex manifold or an analytic space @xmath45 .
in that case , a single chart @xmath43 is enough and the leaves are the orbits of @xmath44 .
[ infinite ] this last example is certainly not the typical example we have on mind , since it may be very far from being a classical foliation .
indeed , the notion of tg foliation is of special interest when the leaves have infinite dimension as below in theorem [ tgfoliationtheorem ] .
thanks to the finite dimensionality condition imposed on the groups @xmath218 , a tg foliation is then really `` almost '' a foliation .
the following example shows an important feature of a tg foliation compared with a classical foliation .
[ locallytrivialaction ] let @xmath44 be a finite dimensional complex lie group acting holomorphically onto a complex manifold @xmath45 .
assume that the action is locally trivial .
then we may define two structures of tg foliation on @xmath45 .
firstly that given by the single chart @xmath43 . secondly , that given by the foliation induced by the locally free action of @xmath44 .
observe that they are isomorphic if and only @xmath44 acts freely on @xmath45 , for the classical foliation structure does not keep track of the isotropy groups .
we have : [ tgfoliationtheorem ] the action of @xmath107 onto @xmath33 induces a tg foliation of @xmath33 with the following properties : 1 .
the leaves are infinite - dimensional frchet manifolds modeled onto @xmath60 .
2 . at a point @xmath26
, the transverse structure is modeled onto the groupoid @xmath222 .
this is essentially a reformulation of proposition [ foliationsimple ] .
it makes use of theorem [ kuranishitheorem ] and lemma [ gdecompolemma ] .
we cover @xmath33 by kuranishi domains @xmath223 centered at @xmath224 and we let @xmath218 be @xmath225 and @xmath216 be the kuranishi space of @xmath56 .
theorem [ kuranishitheorem ] gives foliated charts @xmath226 we assume that the image of such a map is @xmath227 for @xmath215 an open and connected neighborhood of zero in @xmath228 .
we also assume that @xmath215 satisfies the property of lemma [ gdecompolemma ] .
+ thanks to this lemma , the plaques @xmath229 are open neighborhoods of @xmath64 in @xmath60 , hence are preserved by the changes of charts . moreover , the universality property of the stack @xmath230 $ ] implies morita equivalence of the transverse structure groupoids as wanted .
let @xmath148 be a foliation of some analytic space
. we may associate to it a holonomy groupoid as follows ( @xcite and @xcite ) .
we choose a set of local transverse sections .
objects of the groupoid are points of the disjoint union of these local sections .
morphisms are generated by holonomy morphisms , obtained by following the leaves from a transverse section to another one , identifying holonomy morphisms having the same germ .
it is an tale groupoid , which encodes the leaf space of the foliation .
+ having proved in theorem [ tgfoliationtheorem ] that the action of @xmath3 induces a tg foliation of each connected component of @xmath29 , we would like to associate to this tg foliation a holonomy groupoid .
as in the classical case , it should be a presentation of the quotient stack , that is here of the teichmller stack .
+ however , this is much more involved than in the classical case .
the problem is that now the transverse sections are groupoids , so that holonomy morphisms are stacks morphisms between stacks of type @xmath231 $ ] .
hence , instead of building a groupoid , we end with a disjoint union of stacks and a set of local stack morphisms .
it is certainly possible to turn this collection into a nice categorical structure .
however , we will not follow this path since we are interested in obtaining a presentation of the teichmller stack .
+ so we forget about the general case of a tg foliation and we stick to the particular case of @xmath29 . to rephrase the problem we face up , the holonomy morphisms inherited from a general tg foliation are defined between topological spaces of type @xmath232 .
the crucial point is to give an atlas of tg foliation with some additional properties .
this will allow us to define holonomy morphisms at the level of @xmath90 and not @xmath232 .
composing them with the @xmath44-morphisms gives us the morphisms of the holonomy groupoid , the objects belonging to a disjoint union of analytic spaces @xmath90 .
+ in a sense , all this process is about lifting to @xmath90 objects and morphisms defined on @xmath232 .
this will be done in four steps , in sections [ multifoliate ] , [ teichmueller ] and [ main ] . + firstly in section [ multifoliate ] , we construct from the tg foliation of @xmath33 partial foliations of @xmath33 .
partial here means that they are not defined on the whole @xmath33 but on an open subset .
we take a countable collection of such foliations whose domains of definition cover @xmath33 .
basically , the transverse structure of these foliations at some point @xmath26 is modeled onto the kuranishi space of the corresponding complex manifold @xmath62 . however , the jumps in the dimension of the automorphism group cause serious problems here , and we start doing the construction in the neighborhood of a @xmath30-homotopy class , where equidimensionality is fulfilled
. then we extend it to the whole @xmath33 , but to achieve that , we are forced to fat the smallest kuranishi spaces to finish with all transversals of the same dimension .
this fatting process was already used in @xcite .
+ secondly , from this set of partial foliations , we obtain a nicer atlas for the tg foliation of @xmath33 .
we call such an atlas a regular one .
we define simple holonomy germs as the classical holonomy germs of each partial foliation .
the main point is that we allow , under certain circumstancies , composition of holonomy germs coming from two different foliations .
the peculiarities of a regular atlas are useful in this process .
we encode all the holonomy data related to a regular atlas in a groupoid .
this is however not the good groupoid to consider , especially because changing of regular atlas does not produce a morita equivalent groupoid .
all this is done in subsections [ regatlas ] , [ simpleholonomy ] and [ firstapp ] .
this preliminary work is essentially notational and technical , but is important to achieve the construction .
+ thirdly , building on the previous sections , we construct in subsection [ teichsub ] the holonomy groupoid of the tg foliation of @xmath33 .
we call it the _ teichmller groupoid_. it is unique up to morita equivalence .
its objects are points of a disjoint union of transverse sections of partial foliations covering @xmath33
. its morphisms are composition of the simple holonomy germs and of the local action of the automorphism group of @xmath62 on its kuranishi space , up to an equivalence relation .
+ fourthly , and last , we prove that the teichmller groupoid is an analytic smooth groupoid and a presentation of the teichmller stack in theorem [ maintheorem ] .
basically there are two points to check . from the one hand
, it must be shown that composition of simple holonomy germs and local automorphisms describes the full action of @xmath3 onto @xmath33 .
this is done in lemma [ isoteich ] . from the other hand
, it must be shown that the source and target maps are smooth morphisms .
the difficulty here comes from the fact that morphisms are _ classes _ of such compositions up to an equivalence relation .
this is done in lemmas [ encodingsimple ] and [ projection ] . +
before developing all this construction , we consider in the next section the rigidified case , in which the tg foliation is just a foliation , and the teichmller groupoid an ordinary holonomy groupoid .
this can be seen as a toy model for the general construction and will serve to fixing some notations and conventions .
recall the ( see @xcite , definition 12 ) .
[ rig ] a compact complex manifold @xmath62 is _ rigidified _ if @xmath233 is equal to the identity . in that case
, the map @xmath234 is injective .
moreover , [ rigidconsequences ] assume that all structures of some connected component @xmath33 are rigidified . then , 1 .
the action of @xmath107 onto @xmath29 is free . 2 .
the tg foliation of @xmath33 is the foliation induced by the free action
. item ( i ) is immediate from and item ( ii ) follows from proposition [ foliationsimple ] and example [ locallytrivialaction ] . in the case of proposition [ rigidconsequences ] ,
the teichmller groupoid is just the standard holonomy groupoid of the foliation .
we give now a complete treatment of this case , which serves as a toy model for section [ teichmueller ] .
we cover @xmath33 by a collection @xmath235 of open subsets .
we assume that each chart @xmath236 is a kuranishi domain associated to the following retraction map ( the composition is the identity , cf . ) @xmath237 we denote by @xmath224 the base point of the kuranishi space @xmath216 .
observe that the index set may be assumed to be countable , due to proposition [ countableteich ] and the countability of the involved topologies .
+ take two points @xmath238 and @xmath239 belonging to the same leaf and choose a path of foliated charts joining @xmath240 to @xmath241 .
a holonomy germ from @xmath240 to @xmath241 is a germ of analytic isomorphism between the pointed spaces @xmath242 and @xmath243 , which is obtained by identifying along the path of foliated charts points belonging to the same leaf , see @xcite or @xcite . +
they can be encoded in a holonomy groupoid @xcite or @xcite as follows .
_ objects _ are points of the disjoint union of transversals @xmath244 we denote by @xmath245 a point of @xmath216 . to encode the morphisms , we first notice that on each non - empty intersection @xmath246 , there exists a unique isomorphism @xmath247 between some open subset @xmath248 of @xmath249 and some open subset @xmath250 of @xmath251 .
it is obtained by following the leaves of the foliation from @xmath249 till meeting @xmath251 ( when this occurs ) .
it satisfies the commutative diagram @xmath252 [ reducedrigidified ] it happens that kuranishi spaces are everywhere non - reduced .
however , since the leaves of the foliation are always reduced , kuranishi s theorem [ kuranishitheorem ] implies that , even in this non - reduced situation , there exists a unique isomorphism making commutative .
we now look at the pseudogroup generated by the @xmath247 .
in other words , we now let @xmath253 be a collection of indices such that each @xmath254 is non - empty and define @xmath255 this composition is defined on some open subset of @xmath256 that we denote by @xmath257 ; and it ranges in some open subset of @xmath258 , that we denote by @xmath259 .
then we represent all holonomy maps as points of @xmath260 here @xmath261 if each @xmath254 is non - empty .
a point @xmath240 in some @xmath257 represents the germ at @xmath240 of the map @xmath262 , the case @xmath263 encoding the identity germs .
we denote such a point by the @xmath264-uple @xmath265 .
+ consider the groupoid whose objects are given in , and morphisms are given in .
observe that both sets are @xmath20-analytic spaces .
the source map sends @xmath266 onto @xmath267 and the target map sends it to @xmath268 .
both are obviously tale analytic maps , since the source map is just the inclusion @xmath269 on the component @xmath257 ; and the target map on the same component is the composition of the isomorphism @xmath262 from @xmath257 onto @xmath270 with the inclusion @xmath271 .
multiplication is given by composition of holonomy germs .
+ however , we are not finished yet .
the previous groupoid is not the holonomy groupoid of the foliation .
we must still identify identical germs .
it may happen for example that such a composition @xmath262 is the identity .
so we take the quotient of by the following equivalence relation @xmath272 that is if they have same source , same target , and are equal as germs .
hence , the set of morphisms is @xmath273 observe that two distinct points of the same component @xmath257 of can not be equivalent .
therefore , the natural projection map from onto is tale .
hence , we define [ teichrigid ] we call _ teichmller groupoid _ of @xmath33 the groupoid whose objects are given by , whose morphisms are given in , and whose source , target maps and multiplication are defined as above .
and we have [ mainrigidified ] assume that the connected component @xmath33 contains only rigidified complex structures .
then , the teichmller groupoid of @xmath33 is an tale analytic presentation of the connected component @xmath274 $ ] of the teichmller stack .
[ moritaequivrigid ] the construction above depends on a choice of a foliated atlas . however , it is easy to show that it is independent of this choice up to morita equivalence .
this can of course be deduced from general arguments , since it represents the stack @xmath274 $ ] , which does not depend on a foliated atlas .
it can also be proved directly as follows .
start with a foliated atlas and construct the associated teichmller groupoid .
take a finer foliated atlas .
then the associated teichmller groupoid is just the localization of the first one over the new atlas , hence both are weakly equivalent @xcite .
start now with two different foliated atlases and their associated teichmller groupoid . since the union of the atlases is a common refinement of both of them
, the two groupoids are morita equivalent .
[ nonhausdorff ] because we took the quotient of by the tale equivalence relation , the set of morphisms may be non - hausdorff .
this is a classical feature of the holonomy groupoid even in finite dimension .
for example , the holonomy groupoid of the reeb foliation of the sphere @xmath275 is non - hausdorff . in classical foliation theory
, this is linked to the existence of so - called vanishing cycles .
recall also that the hausdorffness / non hausdorffness of the set of morphisms is preserved by morita equivalence .
we refer once more to @xcite for more details .
[ aut10bis ] assume that for all structures @xmath26 in @xmath33 , we have @xmath69 equal to the identity .
then proposition [ foliationsimple ] still applies and the action of @xmath3 still defines a foliation of @xmath33 .
so we can still define a holonomy groupoid as above .
morover the geometric quotient of the teichmller stack equals the leaf space , that is the geometric quotient of this holonomy groupoid . nevertheless , they may be different as _ stacks _ , because there may exist a non trivial element in @xmath233 that fixes @xmath33 .
such an element is encoded in the teichmller groupoid we construct in section [ teichmueller ] but not in the holonomy groupoid of definition [ teichrigid ] , cf .
remark [ subtle ] .
+ for many compact complex manifolds @xmath54 , there is no difference between @xmath93 and @xmath91 , cf .
@xcite . indeed , we do not know of any example for which these two groups are distinct , although we do not see any reason for such an example not to exist .
find a compact complex manifold @xmath54 with @xmath93 being reduced to the identity but which is not rigidified .
if @xmath54 is khler , then a result of liebermann implies that @xmath93 has finite index in @xmath91 . in the non - khler case , however , there should exist examples with infinite `` complex mapping class group '' @xmath276 .
in this section , we associate to the tg foliation of a connected component @xmath33 of @xmath29 a collection of standard foliations of open sets of @xmath33 covering it . in subsection
[ simpleholonomy ] , we will associate to these partial foliations their holonomy germs .
this is a crucial step in defining the morphisms of the teichmller groupoid .
the main problem here is that the dimension of the kuranishi spaces may vary inside @xmath33 . to overcome this difficulty
, we proceed in two steps .
it turns out that the dimension we have really to care about in this problem is the dimension of the automorphism group .
hence we first work in the neighborhood of a @xmath30-homotopy class , so that we may assume equidimension of the automorphism groups involved in the choice of foliated atlases .
then , we treat the general case .
we have to fat the kuranishi spaces with small automorphism group , following a process already used in @xcite .
this supposes that the function @xmath143 to be bounded on @xmath33 .
let @xmath148 be a @xmath30-homotopy class in @xmath29 .
let @xmath94 be a connected neighborhood of @xmath148 in @xmath33 .
let @xmath277 be the grassmannian of closed vector subspaces of @xmath278 of codimension @xmath279 .
for each @xmath280 , define @xmath281 [ fadmissible ] we say that @xmath282 is _
@xmath148-admissible _ if @xmath283 is not empty .
assume that @xmath282 is @xmath148-admissible and let @xmath284 .
then , using the isomorphism @xmath285 ( where @xmath60 is the space of @xmath61-vectors for the structure @xmath76 ) , we see that the choice of a @xmath148-admissible @xmath282 is equivalent to the choice of a closed subspace @xmath73 of @xmath60 satisfying and @xmath286 in the sequel , we will denote by the same symbol @xmath282 a closed subspace of @xmath60 and its real part in @xmath278 .
no confusion should arise from this abuse of notation .
observe that all such @xmath282 are complex isomorphic , cf .
* lemma 7.1 ) .
+ so , once chosen such an @xmath282 , we may apply theorem [ kuranishitheorem ] at @xmath76 with @xmath282 .
we define @xmath287 as the maximal open subset of @xmath94 covered by kuranishi domains modelled on @xmath282 and based at points of @xmath283 .
we can interpretate it as follows .
theorem [ kuranishitheorem ] endows each kuranishi domain with a trivial local foliation by copies of @xmath282 and leaf space @xmath77 .
+ now , let us put this interpretation in a global setting .
it tells us that we may cover @xmath287 by kuranishi domains modelled on the same @xmath282 .
hence @xmath282 defines a foliation of @xmath287 by leaves locally isomorphic to a neighborhood of @xmath64 in @xmath282 , see ( * ? ? ?
* theorem 7.2 ) modelling all the kuranishi domains .
since we assume the existence of such a common @xmath282 , the proof applies . ] .
[ lfoliation ] we call this foliation the _ @xmath282-foliation of @xmath94 _ ( even if it is only defined on @xmath287 ) . in the case where @xmath287 is equal to @xmath94 , which is equivalent to saying that @xmath282 is a common complementary subspace to all @xmath70 for @xmath288
, then we obtain a global foliation of @xmath94 . + nevertheless , it is not possible in general to assume this hypothesis .
hence we shall replace this foliated structure by a collection of _ partial foliations _
encoded in a groupoid .
[ coveringfamily ] a set @xmath289 of @xmath148-admissible elements of @xmath290 such that @xmath291 is called _ a covering family _ of @xmath94 .
choose @xmath289 a covering family of @xmath148 .
observe that we may assume @xmath289 to be countable by proposition [ countable1 ] .
to @xmath289 is associated _ a covering set of partial foliations _ of @xmath94 , defined as the set of all @xmath282-foliations of @xmath94 for @xmath292 .
it is useful to encode it in a groupoid as follows .
+ for each @xmath292 , choose an atlas @xmath293 of @xmath287 by @xmath282-foliated charts . define @xmath294 once again , we may assume that @xmath172 is countable , due to the countability of the involved topologies .
then define the groupoid @xmath295 as follows .
_ objects _ are points of the disjoint union @xmath296 hence are encoded by couples @xmath245 . + we insist on seeing each @xmath236 as a @xmath282-foliated frchet space .
we use the notation @xmath297 to denote the vector space @xmath282 associated to @xmath298 . in section [ teichmueller ] , we will enlarge our index set @xmath172 and the interest of this strange notation should be clarified . set now @xmath299 _ morphisms _ are points @xmath300 encoded by triples @xmath301 . + once again , we insist on seeing each @xmath246 as a @xmath282-foliated frchet space .
note that there is no morphism between a point in a @xmath282-foliated chart and the same point in a @xmath302-foliated chart .
we now deal with the definition of a covering set of partial foliations and its encoding in a groupoid for all points of @xmath33 with bounded function @xmath143 .
+ let @xmath13 . recall
let @xmath303 recall it is open .
we assume that it is connected , replacing it with a connected component otherwise .
given a closed subspace @xmath282 of @xmath211 of codimension @xmath16 , define @xmath304 this is an extension of .
we may go on with this generalization .
we say that @xmath282 is _
@xmath16-admissible _ if @xmath283 is not empty .
analogously to what happens in subsection [ multifoliatefclass ] , the choice of an @xmath16-admissible @xmath282 is equivalent to the choice of a closed subspace @xmath73 of @xmath60 satisfying @xmath305 as in subsection [ multifoliatefclass ] , we denote both @xmath282 and @xmath73 by the same symbol @xmath282 .
although this @xmath282 is not complementary to @xmath70 , we may run the proof of kuranishi s theorem after adding some finite - dimensional subspace @xmath306 such that @xmath307 [ cinfini ] we assume that @xmath308 contains only @xmath28 elements , so that we may use the same @xmath306 for all sobolev classes .
this is always possible since @xmath28 diffeomorphisms are dense in @xmath109 diffeomorphisms for @xmath84 big enough .
we thus obtain an isomorphism between a neighborhood @xmath75 of @xmath26 in @xmath29 and a product ( cf .
* theorem 7.2 ) ) @xmath309 setting @xmath310 we obtain a sequence analogous to @xmath311 this is our new definition of kuranishi domains and charts .
let @xmath312 be a covering of @xmath313 by kuranishi domains .
set @xmath314 .
we define @xmath287 as the maximal open subset of @xmath94 covered by kuranishi domains modelled on @xmath282 and based at points of @xmath313 .
we may then define the set and the groupoid @xmath295 of partial foliations of @xmath94 exactly as in subsection [ multifoliatefclass ] .
[ newk ] recall that the local transversal section at some point @xmath76 is not always its kuranishi space @xmath77 .
it is if and only if @xmath315 is equal to @xmath16 .
more generally , it is the product of @xmath77 with an open neighborhood of @xmath64 in @xmath316 .
[ caeti ] observe that , if the function @xmath143 is bounded on a connected component @xmath33 by some integer @xmath16 , then @xmath313 is equal to @xmath33 .
the following proposition shows that the groupoid of partial foliated structures really describes an intrinsic geometric structure .
[ groupoidi ] we have : + i. the groupoid @xmath295 is a foliated frchet tale groupoid , that is 1 . both the set of objects and that of morphisms are foliated frchet manifolds .
the source , target , composition and anchor maps are analytic and respects the foliations .
3 . the source and target maps are local foliated isomorphisms .
the foliated frchet groupoid @xmath295 is independent of @xmath312 up to foliated analytic morita equivalence .
this is completely standard , since this groupoid is very close to the lie groupoid obtained by localization of a smooth manifold over an atlas , see @xcite , 7.1.3 .
starting with i , then ( i ) is obvious from and ; the source map @xmath317 and the target map @xmath318 are given by the following foliation preserving inclusions @xmath319 proving ( iii ) and part of ( ii ) .
composition is given by @xmath320 provided that @xmath321 ( the notation should be clear from ) . assume for simplicity that @xmath298 , @xmath322 and @xmath323 are pairwise distinct .
this is indeed a foliation preserving analytic map from @xmath324 that is @xmath325 onto .
other cases are treated similarly .
this finishes the proof of ( ii ) , hence of i. + as for ii , start from choosing two coverings @xmath312 and @xmath326 of @xmath94 .
the crucial point is contained in i : these groupoids are tale . from that , it is enough to observe that both the localization of @xmath295 over @xmath326 and the localization of @xmath327 over @xmath312 are equal to the groupoid @xmath328 ( see @xcite for the equivalence with the classical definition of morita equivalence ) .
to finish this section , we note that @xmath295 encodes all the possible foliations of open sets of @xmath94 associated to kuranishi domains .
indeed we have [ full ] the full subgroupoid of @xmath295 obtained by restriction to a fixed @xmath292 is the localization over some atlas @xmath287 , hence is morita equivalent to the larger subdomain of @xmath94 foliated by @xmath282 .
in this section , we construct for the tg foliation of theorem [ tgfoliationtheorem ] the analogue for the holonomy groupoid .
we call it _ the teichmller groupoid_. this will be done in several steps . in subsection
[ regatlas ] , we first give for the tg foliation a foliated atlas with good properties .
we call it _ a regular atlas_. we then define in subsection [ simpleholonomy ] the holonomy germs associated to the set of partial foliations . in subsection
[ firstapp ] , we encode these simple holonomy morphisms in a groupoid @xmath329 .
this is however not the right analogue for the holonomy groupoid , since it does not take into account the isotropy groups of the transverse structure of the tg foliation . from the regular atlas ,
we finally build in subsection [ teichsub ] the teichmller groupoid .
theorem [ tgfoliationtheorem ] gives an atlas for the tg foliation of @xmath33 .
however , it is not equidimensional in the sense that the automorphisms groups @xmath218 of the transverse structure may have different dimensions . we need to construct on @xmath94 an equidimensional atlas from the atlas @xmath312 of @xmath329 . besides
, we need this atlas to reflect the partial foliated structure of @xmath33 to be able to define properly the holonomy germs . + as in section [ multifoliate ] , we fix @xmath289 and we define and as well as @xmath295 .
+ we assume that each chart @xmath236 is a kuranishi domain associated to the following retraction map ( the composition is the identity , cf . ) @xmath330 recall remark [ newk ] .
+ the set of holonomy germs of @xmath295 is constructed from the union of all holonomy groupoids when @xmath282 varies .
but in order to mix these holonomies , we first add some charts with common transversal for different foliations .
more precisely , for every couple @xmath331 in @xmath332 with @xmath333 we enlarge the index set @xmath172 to include an atlas @xmath334 of .
we emphasize that _ the same _ analytic set @xmath249 is used as leaf space for both the @xmath282 and the @xmath302-foliations .
this is possible due to the uniqueness properties in kuranishi s theorem [ kuranishitheorem ] .
+ in the same way , for any value of @xmath335 , we enlarge the index set @xmath172 to include atlases @xmath336 for @xmath337 , of @xmath338 once again , we insist on the fact that @xmath249 is a _ common leaf space _ for every @xmath339-foliation restricted to @xmath236 .
we use the notation @xmath340 as a natural extension of .
[ atlases ] it is important to notice that the new covering @xmath312 is constructed from the covering @xmath312 of @xmath295 but has strictly more charts because of and .
moreover , this ( extended ) covering can not be used to construct some @xmath295 , since each chart of @xmath295 has to be explicitely associated to a unique @xmath292 .
however , to avoid cumbersome notations , we use the same symbol for both coverings .
we have now to pay attention to the fact that @xmath216 is no more the kuranishi space of @xmath224 , but its product with some open set in @xmath341 , cf . .
hence the translation groupoid of subsection [ localaction ] is not the good one to consider .
this can be easily fixed by fatting also the group @xmath218 .
recall and remark [ cinfini ] . now define @xmath342 with @xmath306 acting by translations onto itself . with this new definition of @xmath218 ,
it is straightforward to check the [ newgdecompolemma ] let @xmath343 be the map that sends an element @xmath344 of @xmath218 to @xmath345 in @xmath107 . if @xmath63 is a small enough neighborhood of @xmath64 in @xmath282 , then there exist an open and connected neighborhood @xmath34 of the identity in @xmath218 and an open and connected neighborhood @xmath106 of the identity in @xmath107 such that @xmath346 is an isomorphism .
the proof is a copy of that of lemma [ gdecompolemma ] .
[ ig ] in particular , notice that @xmath343 is a local immersion at identity . to simplify notations , we will write in the sequel @xmath22 instead of @xmath347 .
we let @xmath348 act on @xmath216 exactly as in , that is @xmath349 and form the corresponding translation groupoid .
notice that depends on a choice of @xmath282 . + the following proposition is an easy adaptation of theorem [ tgfoliationtheorem ] .
the main point is that @xmath217 is still universal for deformations of @xmath224 , since the extra factor of @xmath218 acts transitively and with no fixed points on the extra factor of @xmath216 .
[ regatlasprop ] the covering @xmath312 is a foliated atlas for the tg foliation of @xmath94 with the following properties 1 .
the foliated charts are given by @xmath350 for every choice of @xmath298 and of @xmath351 , 2 .
the transverse structure is given by @xmath352 letting the second factor of act by translations on the second factor of .
we define [ regatlasdef ] we call _ regular atlas _ of @xmath94 such a foliated atlas @xmath312 . in this subsection
, we associate to the partial foliations of @xmath33 their holonomy germs .
the main point is how to mix the holonomies of the different foliations .
we refer to section [ rigidified ] for comparison .
+ we start with a regular atlas @xmath312 . on each intersection @xmath246 with @xmath353 and for every choice of @xmath339 in , we define the holonomy isomorphism @xmath354 between some open subset @xmath355 of @xmath249 and some open subset @xmath356 of @xmath251 as in section [ rigidified ] .
recall the commutative diagram .
we then look at the pseudogroup generated by the @xmath357 . in other words
, we now let @xmath358 be collections of @xmath359 elements for any value of @xmath359 and define @xmath360 here we assume by convention that both @xmath359 appearing in are the same , allowing repetitions if necessary .
this composition is defined on some open subset of @xmath249 that we still denote by @xmath361 ; and it ranges in some open subset of @xmath362 , that we denote by @xmath363 where @xmath364 note that @xmath365 where this composition is defined , and that @xmath366 we define [ simplegerms ] we call _ simple holonomy morphisms _ of @xmath367 the morphisms .
we may encode the simple holonomy morphisms in a groupoid @xmath329 as follows , compare with the construction of the standard holonomy groupoid in section [ rigidified ] .
it is a first approximation of the teichmller groupoid , but which does not see the automorphism groups .
_ objects _ are points of the disjoint union @xmath368 hence encoded by couples as in .
_ morphisms _ encode germs of holonomy maps . they are defined only between a source object @xmath245 and a target object @xmath369 such that @xmath370 for some collections @xmath322 ( with @xmath371 ) and @xmath282 .
we have first all identity germs , represented by a copy of in the set of morphisms .
then , the _ elementary _ holonomy maps , that is maps for which @xmath322 - and then @xmath282 - has length one . they are encoded as @xmath372 to be precise , a point @xmath240 in some @xmath361 represents the germ at @xmath240 of the map @xmath357 . here
@xmath373 then we represent all holonomy maps as points of @xmath374 for @xmath375 as previously , a point @xmath240 in some @xmath361 represents the germ at @xmath240 of the map @xmath357 , the case @xmath376 encoding the identity germs .
+ however , we are not finished .
we must still identify identical germs .
so we take the quotient of by the following equivalence relation @xmath377 that is if they have same source , same target , and are equal as germs . hence
, the set of morphisms is @xmath378 observe that two distinct points of the same component @xmath361 of can not be equivalent .
therefore , the natural projection map from onto is tale .
+ analogous to proposition [ mainrigidified ] is the following [ groupoidii ] the groupoid @xmath379 is an analytic tale groupoid .
however , and contrary to the case of proposition [ mainrigidified ] and remark [ moritaequivrigid ] , @xmath379 and @xmath380 _ are not always morita equivalent_. this is due to the fact that we mix holonomies of different foliations .
indeed , this is not the good holonomy groupoid to consider , because it does not take into account the groups @xmath218 of the tg structure .
this is quite standard , because @xmath379 is basically just a union of holonomy groupoids ( cf .
@xcite ) . item ( i ) follows from , and . for @xmath298 , @xmath322 and @xmath282 fixed , the source map is the inclusion @xmath381 and the target map is given by @xmath357 , that is @xmath382 this proves ( iii ) and part of ( ii ) .
composition at the level of is given by @xmath383 thanks to .
and it descends on as the composition of germs .
this is analytic as a map from @xmath384 onto @xmath385 in both cases .
this finishes the proof of ( ii ) . to finish this section ,
we want to clarify the relationships between @xmath379 and the holonomy groupoids of the @xmath282-foliations . here
it is important to take special care to remark [ atlases ] . to avoid confusions , we will index the connected components of the objects of @xmath295 by @xmath386 ; and those of @xmath379 by @xmath172 .
we insist on the fact that these two sets are different since we added extra indices to construct @xmath379 . with that difference on mind
, we have immediately [ autregroupoid ] let @xmath292 .
the holonomy groupoid of the @xmath282-foliation is given by the full subgroupoid of @xmath379 over @xmath387 .
in particular , if @xmath289 contains a single element , we have morita equivalence , cf .
section [ rigidified ] .
[ groupoidiibis ] assume that @xmath289 contains a single element @xmath282 , which is equivalent to saying that @xmath282 is a common complementary subspace to all @xmath70 for @xmath288 .
then @xmath379 is the holonomy groupoid of the @xmath282-foliation and it is independent of the covering up to morita equivalence .
[ applicor ] especially , corollary [ groupoidiibis ] applies to the case where @xmath279 is zero , i.e. the automorphism group of all structures of @xmath148 is discrete .
but it also applies to the case of complex tori , since the continuous part of their automorphism group is given by translations and since the associated lie algebra is independent of the complex structure ( as subalgebra of the algebra of smooth vector fields ) .
this shows that this is not the good groupoid to consider , cf .
example [ locallytrivialaction ] . as in the previous subsections , we start from a regular atlas @xmath312 of @xmath94 . for simplicity , we assume that @xmath312 is fine enough so that the following condition is satisfied : the union of @xmath388 and @xmath389 such that @xmath390 is contained in a @xmath282-foliated chart .
so given @xmath391 with @xmath322 of length one , implies that there exists a unique @xmath392 in @xmath282 such that @xmath393 observe that @xmath94 is equal to its saturation @xmath394 hence , given @xmath240 in @xmath94 , its complete @xmath3-orbit is in @xmath94 .
recall also the map .
keep in mind that we need to choose some @xmath351 to define it .
+ let us define the teichmller groupoid @xmath395 as follows .
_ objects _ are points @xmath396 exactly as for @xmath379 .
but we will enlarge the set of morphisms to take into account the action of the automorphism groups .
we proceed as in subsection [ firstapp ] .
+ basically , our morphisms are compositions of holonomy maps @xmath357 and maps .
first , we consider @xmath397 here , the notation @xmath398 means that there exists an isotopy @xmath399\longmapsto g_t\in g_\alpha\ ] ] such that @xmath400 moreover , products @xmath401 are all computed using the same @xmath282 .
observe that such a @xmath402 contains @xmath216 , that is all identity germs and contains @xmath403 .
set @xmath404 whenever it is defined . by convention
, we use @xmath339 to compute the product @xmath405 .
then , we encode the other morphisms as points of @xmath406 for some @xmath407 .
we consider thus the analytic space @xmath408 where @xmath409 is defined in for @xmath410 and @xmath411 is now just the set of @xmath412 with @xmath351 .
+ however , as in subsection [ firstapp ] , we still have to take the quotient of by an appropriate equivalence relation to obtain the set of morphisms . the crucial remark to do that is given by [ diff0encoding ]
we have : 1 . to any @xmath240 in @xmath361 ,
is associated a canonical element in @xmath107 , say @xmath413 , such that @xmath414 2 . to any @xmath415 in @xmath416 , is associated a canonical element in @xmath107 , say @xmath417 , such that @xmath418 notice that , in this lemma , we also consider the case of length zero , that is the case @xmath238 for ( i ) and @xmath419 for ( ii ) .
the meaning of `` canonical '' should be clear from the proof .
we emphasize that and are _
pointwise identities_. changing @xmath240 but keeping @xmath407 or @xmath420 fixed gives a different element in @xmath107 , as suggested by the notations .
hence , from the one hand , and are far from being verified by a unique element of @xmath3 . and from the other hand , a diffeomorphism @xmath413 , respectively @xmath417 , has no reason to send a neighborhood of @xmath240 in @xmath216 onto a neighborhood of @xmath421 , respectively of @xmath422 , in @xmath362 .
it is enough to show the result for @xmath322 of length zero or one .
take @xmath240 in some @xmath216
. then we define @xmath423 as the identity . + for @xmath415 in some @xmath402
, we proceed as follows . by
, there exists a unique @xmath112 in @xmath282 such that @xmath424 in this case , we set @xmath425 let now @xmath426 with @xmath322 of length one .
then using , there exists a unique @xmath112 in @xmath282 such that @xmath427 set @xmath428 finally , if @xmath429 with @xmath322 of length one , we combine and . using , we see that there exist unique @xmath430 and @xmath431 such that @xmath432 set @xmath433 the other cases are obtained by composition .
in other words , any morphism @xmath434 of encodes a canonical diffeomorphism @xmath435 isotopic to the identity .
so this allows us to identify two such morphisms with same source and target if they correspond to the same element of @xmath3 . to be precise
, we define @xmath436 _ morphisms _ are now defined as points @xmath437 [ subtle ] there is a subtle point here we want to emphasize .
equivalence is an equivalence of elements in @xmath3 , whereas equivalence is an equivalence of holonomy maps , the relation between these two type of maps being stated in lemma [ diff0encoding ] . in other words , concerns the geometric orbits of @xmath3 in @xmath33 , whereas concerns the parametrization of the geometric orbits by @xmath3 . in particular ,
if an element of @xmath3 is an automorphism for an open neighborhood of structures in @xmath33 , then it appears as a morphism of but not as a morphism of .
in this short section , we adapt the construction of section [ teichmueller ] to obtain a groupoid that describes the action of the full diffeomorphism group @xmath4 onto @xmath33 .
fix @xmath94 as before .
thanks to , we just have to add the action of the mapping class group on the teichmller groupoid . to do that
, we will first replace @xmath94 by its saturation @xmath438 [ sathyp2 ] from now on , we will always assume that @xmath94 is equal to its saturation . to cover @xmath439 with kuranishi charts , we proceed as follows .
we first choose some covering of with kuranishi charts satisfying .
then we choose some @xmath440 in @xmath4 for every class of @xmath128 .
call @xmath441 the set of indices and set @xmath442 .
we assume that @xmath443 belongs to @xmath444 for all @xmath343 .
but we can not in general assume that @xmath444 is stable under composition .
this would imply that we realize the mapping class group of @xmath0 as a subgroup of @xmath4 , which is not always possible .
+ for any @xmath388 , we define @xmath445 and @xmath446 ( well defined since @xmath216 is included in @xmath388 ) , so that the sequence @xmath447 is a kuranishi chart based at @xmath448 .
+ then we may perform the constructions of section [ teichmueller ] replacing @xmath94 with its saturation .
the riemann moduli groupoid @xmath449 is now defined as the translation groupoid of the action of the mapping class group onto @xmath395 .
more precisely , it is obtained as follows .
we define the set of objects as in . as for the morphisms
, we start with @xmath450 where @xmath451 and we follow the same strategy as in section [ teichmueller ] .
+ setting @xmath452 ( compare with ) . then , as an extension of lemma [ diff0encoding ] , we have [ diffencoding ] let @xmath453 .
then to @xmath415 is canonically associated an element @xmath454 in @xmath455 such that @xmath456 just define for length @xmath64 elements @xmath457 and for length @xmath23 elements @xmath458 as in section [ teichmueller ] , we take the quotient of by the equivalence relation of representing the same diffeomorphism @xmath0 through lemma [ diffencoding ] , cf . .
and we define the set of morphisms as this quotient .
in this section , building on the previous sections , we prove the main results of this paper : the teichmller groupoid is an artin analytic presentation of the teichmller stack , theorem [ maintheorem ] ; the riemann moduli groupoid is an artin analytic presentation of the riemann moduli stack , theorem [ mainr ] ; both of them are isomorphic to stacks of deformations of complex manifolds , theorems [ xvstack ] and [ stack2 ] .
the aim of this subsection is to prove [ maintheorem ] let @xmath33 be a connected component of @xmath29 .
assume that the function @xmath143 is bounded on @xmath33 .
then , 1 .
the teichmller groupoid is an artin analytic presentation of the teichmller stack @xmath274 $ ] .
it is independent of the regular covering of @xmath312 up to morita equivalence .
in the general case , we have [ maincorollary ] let @xmath33 be a connected component of @xmath29 .
then , the teichmller stack @xmath274 $ ] is the direct limit of artin analytic stacks . for every nonnegative integer @xmath16 , we define @xmath313 as in .
we consider the teichmller stack @xmath274 $ ] as the direct limit of stacks @xmath459\hookrightarrow \hdots\hookrightarrow [ \mathcal c_a/\text{diff}^0(x)]\hookrightarrow \hdots\ ] ] applying theorem [ maintheorem ] replacing @xmath33 with @xmath313 for every @xmath16 yields the result .
the manifold @xmath460 gives such an example , cf .
example [ hirzebruchbis ] .
+ going back to theorem [ maintheorem ] , the main difficulty in proving it is to ensure that the source and target maps are _ smooth morphisms_. to achieve that , we first need to establish some auxiliary results on the set of morphisms of the teichmller groupoid .
we begin with showing that it completely describes the action of @xmath3 .
[ isoteich ] we have : 1 .
let @xmath238 an object .
then the set of @xmath240-isomorphisms is @xmath461 .
2 . let @xmath238 and @xmath239 .
then the set of morphisms from @xmath240 to @xmath241 is the set @xmath462
\(ii ) let @xmath463 be an object . through the identification given in lemma [ diff0encoding ]
, @xmath240 is only connected through a morphism to a point @xmath241 which belongs to the same orbit of @xmath3 .
let now @xmath239 such that @xmath464 for some @xmath182 in @xmath3 .
choose also an isotopy @xmath465 from @xmath240 to @xmath241 .
+ _ first assume that @xmath322 is equal to @xmath298 .
_ choose some @xmath466 .
the map @xmath467 is a local isomorphism at @xmath468 by lemma [ newgdecompolemma ] .
so if @xmath182 is small enough , it decomposes uniquely as @xmath469 so that we have @xmath470 proving @xmath182 is in the set of morphisms from @xmath240 to @xmath241 .
this is still true for @xmath182 of arbitrary size , but satisfying @xmath471\ ] ] by decomposing it as a finite product of small elements .
assume now that @xmath472.\ ] ] then , defining @xmath473 we have @xmath474 for some uniquely defined @xmath475 .
notice that @xmath476 , hence @xmath392 is the zero vector field .
we thus may replace the isotopy @xmath477 with the isotopy @xmath478 and @xmath479 with @xmath480 . since @xmath481 satisfies
, the previous proof applies and @xmath182 belongs to the set of morphisms from @xmath240 to @xmath241 .
+ _ assume now that @xmath322 is different from @xmath298 but @xmath389 intersects @xmath388 non trivially_. assume also that @xmath477 satisfies @xmath482 define @xmath483 set also @xmath484 arguing as above shows that @xmath485 , respectively @xmath486 , belongs to the set of morphisms connecting @xmath240 to @xmath487 , respectively @xmath241 to @xmath488 .
but , since @xmath489 we see that @xmath490 belongs to the set of morphisms from @xmath240 to @xmath241 . + in the general case , we can always find @xmath491 such that every @xmath492 falls into one of the previous two cases . by composition , we are done .
+ ( i ) just apply ( ii ) to the case @xmath493 .
[ aut10 ] notice from the proof of lemma [ morphismsxy ] that an element of @xmath494 which is not in @xmath495 decomposes into a non trivial combination of holonomy maps and automorphisms of @xmath495 . to go further , we need to establish another presentation of .
we first have [ encodingsimple ] let @xmath434 be an element of
. then , there exists @xmath496 such that @xmath497 let us explain a little more this technical but important lemma .
it means that any morphism of can be represented by the composition of a single holonomy map with a single automorphism acting , rather than a composition of @xmath359 such maps and automorphisms acting .
let @xmath239 , @xmath238 and @xmath498 such that @xmath499 assume first that @xmath500 and that we can find an isotopy @xmath501 between the identity and @xmath182 such that @xmath502 . then
, by lemma [ isoteich ] , @xmath182 can be decomposed as a finite union of length @xmath64 elements @xmath503 with moreover @xmath504 we want to write as @xmath505 taking into account the definition of @xmath506 ( cf . ) , we see that this amounts to proving a commutation property .
more precisely , this amounts to proving that , given @xmath507 and @xmath430 with @xmath508 , there exists @xmath509 and @xmath431 such that @xmath510 but this is true for @xmath511 and @xmath112 small enough , since is a local diffeomorphism at @xmath468 . if not , just write @xmath511 and/or @xmath512 as a finite composition of small elements . because of , we can repeat the argument on each term of the decomposition and prove the result . + as a consequence , starting once again with but without assuming @xmath500 , we have proven that @xmath182 can be written as a composition of only length one elements .
so it is enough to prove the result for the composition of two length one elements . hence , assuming @xmath513 for @xmath514 we want to rewrite it as @xmath515 for some @xmath516 .
set @xmath517 and observe that @xmath518 satisfies @xmath519 the last assertion coming from the fact that we may choose @xmath520 with @xmath521 but we are thus in position to apply the first part of the proof ; that is , we may find @xmath522 such that @xmath523 combined with , this proves the result . as a consequence of lemma [ encodingsimple ] , we may rewrite as @xmath524 with @xmath525 consider the natural projection @xmath526 we have [ projection ] the projection is tale and @xmath156 on any component @xmath527 .
let @xmath415 and @xmath528 be two elements of the same component @xmath527 .
assume they are equivalent .
then they have same source , hence @xmath241 equals @xmath240 .
moreover , @xmath529 therefore , simplifying both sides by @xmath530 and using the notations of , @xmath531 for some @xmath112 and @xmath532 in @xmath533 .
but this implies , using lemma [ newgdecompolemma ] and assuming that this diffeomorphism is small enough , that @xmath534 if this is not the case , just compose both sides by the same automorphism ( on the left ) to decrease the norm sufficiently , and repeat the argument .
we are now in position to prove theorem [ maintheorem ] .
let us start proving i. because of lemma [ isoteich ] , the teichmller groupoid is a presentation of the teichmller stack .
then , the set of objects and morphisms are analytic spaces thanks to , , and lemma [ projection ] .
the source map is given by @xmath535 so is analytic and a smooth morphism .
+ the target map is given by @xmath536 and is also analytic . to show it is a smooth morphism ,
just observe that the map @xmath537 is a local isomorphism .
hence @xmath318 is locally a projection .
+ the composition map goes as follows .
starting from @xmath538 with @xmath539 one obtains @xmath540 where @xmath541 is just the concatenation of @xmath22 and @xmath511 . + reuse .
this shows that the composition map is analytic and locally defined as a map from an open set of @xmath542 onto @xmath543 .
+ let us prove part ii
. start with @xmath312 and a collection of kuranishi spaces @xmath216 based at @xmath224 ; take then @xmath326 and a collection of kuranishi spaces @xmath544 based at @xmath545 .
form @xmath546 as follows .
define @xmath547 and take as set of objects @xmath548}\ ] ] where @xmath549 $ ] means @xmath550 or @xmath551 and where the union runs over all possible choices .
remark that @xmath552 is based at @xmath224 whereas @xmath553 is based at @xmath545 .
+ define the set of morphisms as previously , that is @xmath554,[\beta , b ] , l}\right ) \bigg /\sim\ ] ] where the union runs over all possible choices .
+ for every pair @xmath550 , choose some @xmath282 belonging to @xmath555 .
then , the holonomy map @xmath556 defines an isomorphism between @xmath552 and @xmath553 . + this allows us to define a weak equivalence @xmath557 between @xmath546 and @xmath395 . on the set of objects ,
send @xmath558 to @xmath238 ( remark that this is @xmath559 ) ; and @xmath560 to @xmath561 .
this is an tale and surjective map .
+ observe that , to @xmath240 , we may associate through lemma [ diff0encoding ] a unique diffeomorphism @xmath562 such that @xmath563 .
+ as for the morphisms , just send a morphism @xmath564 such that @xmath565 onto the morphism @xmath566 connecting @xmath567 to @xmath568 . because of lemma [ isoteich ] , this realizes an isomorphism between the set of morphisms from @xmath240 to @xmath241 and that from @xmath567 to @xmath568 .
more precisely , if @xmath560 and @xmath569 satisfy @xmath570 , the following diagram is cartesian @xmath571 other cases give rise to similar diagrams .
this finishes the proof .
we notice the following [ etale ? ] the teichmller groupoid is an tale analytic presentation of the teichmller stack @xmath274 $ ] if and only if the function @xmath143 is identically zero on @xmath33 .
use theorem [ maintheorem ] and the fact that the isotropy group of a point @xmath26 is @xmath233 by lemma [ isoteich ] .
[ catanese ] it is important to compare the local structure of the teichmller stack at some point @xmath26 with its kuranishi space @xmath89 , or better with the translation groupoid @xmath572 .
the rigidified case is of special interest and amounts to asking if the teichmller stack of @xmath0 is locally isomorphic at @xmath26 to the analytic space @xmath89 , cf @xcite .
+ catanese shows in ( * ? ? ?
* theorem 45 ) , that , for a minimal surface @xmath573 of general type , if @xmath574 is a trivial group , or if @xmath573 is rigidified with ample canonical bundle , then the teichmller space is locally homemorphic to the kuranishi space .
he also shows in ( * ? ? ?
* proposition 15 ) that the same result holds for khler manifolds with trivial canonical bundle .
this is used by verbitsky in @xcite , see example [ hk ] .
+ this question is equivalent to asking if there can be non trivial simple holonomy morphisms . in particular , when _ all _
the structures of a connected component @xmath33 are rigidified , a positive answer means that the holonomy groupoid of the @xmath3-foliation of @xmath33 is trivial , hence that the foliation itself is trivial .
+ this seems however too much to expect in general and suggests the following [ holonomyproblem ] find a compact @xmath28 manifold @xmath0 with a connected component @xmath33 of rigidified structures and with a non - trivial teichmller groupoid . to begin with
, it would be very interesting to have an example of an oriented smooth manifold @xmath0 such that @xmath1 is the leaf space of an irrational foliation of a complex torus .
analogously , we have [ mainr ] let @xmath33 be a connected component of @xmath29 .
assume that the function @xmath143 is bounded on @xmath33 .
then , 1 .
the riemann moduli groupoid is an artin analytic presentation of the riemann moduli stack @xmath274 $ ] .
it is independent of the regular covering of @xmath312 up to morita equivalence . and , in the general case , [ maincorollaryr ] let @xmath33 be a connected component of @xmath29 .
then , the riemann moduli stack @xmath575 $ ] is the direct limit of artin analytic stacks .
the proof of corollary [ maincorollaryr ] is similar to that of corollary [ maincorollary ] .
the proof of theorem [ mainr ] follows that of theorem [ maintheorem ] .
we need adapted versions of lemmas [ morphismsxy ] , [ encodingsimple ] and [ projection ] . as in the previous section
, we first notice that [ morphismsxyr ] pick @xmath240 and @xmath241 in the set of objects .
then , the set of morphisms joining @xmath240 to @xmath241 is @xmath576 let @xmath30 belong to .
then , there exists @xmath577 such that @xmath578 belongs to @xmath3 . by lemma [ morphismsxy ] , we know that @xmath578 belongs to the set of morphisms joining @xmath240 to @xmath579
. hence @xmath580 belongs to the set of morphisms joining @xmath240 to @xmath241 .
then we will rewrite our set of morphisms with help of [ diffsimple ] let @xmath415 be an element of @xmath581 .
then , there exists some @xmath582 such that @xmath583 here @xmath584 denotes the subgroup of @xmath4 generated by @xmath444 .
as already observed , we can not always ensure that @xmath584 is isomorphic to the mapping class group .
this is just a commutation property .
it follows from our choice of kuranishi charts .
indeed , we have @xmath585 taking into account that we have an isomorphism @xmath586 that we extend between @xmath218 and @xmath587 as the identity on the factor @xmath306 .
the lemma follows easily from this property . as in the previous section ,
together with lemma [ encodingsimple ] , this allows us to replace the set of morphisms with @xmath588 where @xmath589 we have [ projection2 ] the projection map @xmath590 is tale and @xmath156 on any component @xmath591 . immediate consequence of and lemma [ projection ] .
let @xmath592 denote the category of @xmath20-analytic spaces .
we give a quick proof that both the teichmller and the riemann moduli stacks are isomorphic to stacks of deformations over @xmath592 .
this is of course not a surprise but it is important to tie the teichmller and the riemann moduli stacks to the classical deformation theory of kodaira - spencer .
+ let @xmath94 be an open set of @xmath29 ( e.g. @xmath94 is some connected component @xmath33 or @xmath94 is some @xmath313 ) . define the following category @xmath593 over @xmath592 .
+ _ objects _ are @xmath594-families @xmath595 that is : 1 .
@xmath596 and @xmath597 .
2 . @xmath168 is a smooth and proper morphism with reduced fibers all diffeomorphic to @xmath0 .
each fiber @xmath598 can be encoded as @xmath599 with @xmath600 .
in other words , a @xmath594-family is nothing else than an analytic deformation of complex structures of @xmath0 such that the structure of each fiber is isomorphic to a point of @xmath601 .
+ _ morphisms _ are cartesian diagrams @xmath602 between @xmath603-families .
observe that the pull - back of a @xmath603-family is a @xmath603-family .
+ it is straightforward to check that this is a stack over @xmath592 .
the fiber over @xmath604 is the groupoid formed by @xmath603-families over @xmath573 and isomorphisms of families .
[ xvstack ] @xmath605 is isomorphic to the restriction @xmath606 $ ] of the riemann moduli stack .
let @xmath312 be a regular atlas of @xmath94 .
we will show that the stackification of @xmath449 over @xmath592 is @xmath605 .
theorem [ mainr ] yields then the result . we assume .
an object over @xmath573 in the stackification of @xmath449 is given by an open covering @xmath607 of @xmath573 , a collection of maps @xmath608 and a collection of gluings @xmath609 satisfying a compatibility condition as well as the usual cocycle condition ( for simplicity , we write @xmath610 for a couple formed by an automorphism and an element of @xmath584 ) . more precisely , the compatibility condition is that , given @xmath240 in @xmath611 , we have @xmath612 and @xmath613 we will show that this is exactly the data we need to construct a @xmath603-family @xmath614 . set @xmath615 where the operator @xmath616 along the fiber @xmath617 is tautologically defined as @xmath26 .
this defines a @xmath594-family over @xmath618 , cf .
+ the main point is that @xmath357 lifts canonically to an isomorphism between the restriction of @xmath619 over @xmath620 and the restriction of @xmath621 over @xmath622 .
this is a consequence of lemmas [ diff0encoding ] and [ diffencoding ] .
define the canonical lifting of as @xmath623 for @xmath624 observe that the cocycle condition just means that the maps @xmath625 verify the cocycle condition in @xmath4 .
now , define @xmath614 as @xmath626 where @xmath627 is the equivalence relation @xmath628 this defines a @xmath603-family thanks to the cocycle condition .
+ hence , every descent datum associated to @xmath449 is a @xmath603-family .
+ let @xmath629 and @xmath630 .
let @xmath631 be a morphism .
let @xmath632 , respectively @xmath633 be a descent datum over @xmath573 , respectively @xmath634 ( we use , and so on ) .
a morphism between them and over @xmath22 is given by a collection of maps @xmath635 from @xmath618 to the set of morphisms of @xmath449 such that 1 .
@xmath636 and @xmath637 .
2 . @xmath638 .
it is straightforward , although awkward , to check that ( i ) shows that @xmath635 induces local cartesian diagrams @xmath639 that is local morphisms between the families associated to the descent data ; and that ( ii ) implies that these local morphisms commute with the gluing , hence define a global morphism of @xmath603-families .
+ all this shows the existence of a functor over @xmath592 from the stackification of @xmath449 to @xmath605 .
but kuranishi s theorem shows that any @xmath603-family is locally isomorphic to a pull - back family @xmath640 .
hence we may choose a covering of the base and a collection of maps @xmath641 as in , with associated gluing maps satisfying and so that it is isomorphic to some family .
hence this functor is essentially surjective .
+ moreover , because of lemma [ morphismsxyr ] , morphisms between two objects of the stackification of @xmath449 coincide with morphisms between them as objects of @xmath593 .
therefore the functor is fully faithful and the two stacks are indeed isomorphic .
this finishes the proof .
hence @xmath605 is the riemann moduli space for complex structures belonging to @xmath94 .
+ we pass now to the construction of @xmath642 , which is more delicate .
as used in the proof of theorem [ xvstack ] , any family @xmath643 can be seen locally over some sufficiently small open set @xmath644 as @xmath645 for some smooth family @xmath616 of complex operators of @xmath0 . as a consequence ,
@xmath646 is diffeomorphically a bundle over @xmath647 with fiber @xmath0 and structural group @xmath4 . in particular
, it makes sense to speak of the structural group of @xmath614 , and to say that the structural group can be reduced to some subgroup @xmath648 of @xmath4 . and it makes also sense to speak of @xmath648-isomorphism of the family @xmath614 , that is isomorphism of @xmath614 such that , in each fiber , the induced diffeomorphism of @xmath0 is in @xmath648 . + we define @xmath649 as the category whose objects are @xmath603-families whose structural group can be reduced to @xmath3 and whose morphisms are cartesian diagrams such that the canonical isomorphism between @xmath614 and @xmath650 is a @xmath3-isomorphism . +
this forms a category over @xmath651 and a subcategory of @xmath593 .
this @xmath649 is the analogue of the teichmller space for complex structures belonging to @xmath94 .
+ we have [ stack2 ] @xmath642 is isomorphic to the restriction @xmath652 $ ] of the teichmller stack .
the proof of theorem [ xvstack ] applies here .
it shows that , given a regular atlas @xmath312 of @xmath94 , the stackification of the teichmller groupoid @xmath395 is a subcategory of @xmath593 .
we are left with proving that it is exactly @xmath649 . but a @xmath603-family with structural group reduced to @xmath3 is equivalently defined as those which are @xmath3-isomorphic to some decomposed family with all gluings in @xmath3 .
taking into account and , this is the case if and only if all @xmath653 can be assumed to be the identity .
+ moreover , the same analysis applies to morphisms . using the notation of the proof of theorem [ xvstack ]
, we have that such a morphism yields a @xmath3-morphism of families if and only if the morphisms @xmath635 map into @xmath494 .
we draw now some consequences of theorems [ xvstack ] and [ stack2 ] on the groupoids @xmath395 , and @xmath449 . recall that their construction depends on the choice of a covering family @xmath289 .
[ lindep ] we have : 1 .
the groupoid @xmath449 is independent of @xmath289 up to analytic morita equivalence .
the groupoid @xmath395 is independent of @xmath289 up to analytic morita equivalence . since theorems [ xvstack ] and [ stack2 ] are completely independent of @xmath289
, we have directly the results . [ rvst ] in the classical case of riemann surfaces ,
the teichmller space is nicer than the riemann moduli space , since the first one is a manifold whereas the second one is an orbifold .
there is no such difference between the teichmller stack and the riemann moduli stack .
both have similar structures of artin analytic stacks . however , the teichmller groupoid has a much more natural geometric interpretation as the holonomy groupoid of the tg foliation of @xmath29 .
the riemann moduli stack is built from this holonomy groupoid and from the action of the mapping class group .
hence , for quite different reasons than for surfaces , the teichmller stack is nicer than the riemann moduli stack .
* tori . * [ tori ] consider firstly the one - dimensional case .
so let @xmath0 be @xmath654 .
then @xmath29 is connected and , as geometric quotients , @xmath1 is the upper half plane @xmath655 , and @xmath6 is the orbifold obtained as the quotient of @xmath655 by the classical action of @xmath656 .
+ however , these are not the teichmller and riemann stacks of @xmath0 , but of @xmath657 where @xmath657 is obtained from @xmath0 by blowing up a point .
equivalently , they are the teichmller and riemann stacks of @xmath0 for structures of elliptic curves , that is with a fixed point .
+ to describe @xmath1 and @xmath6 as stacks , we must incorporate the action of the translations .
this can be done as follows .
consider the quotient @xmath614 of @xmath658 by the group generated by @xmath659 then @xmath660\in\mathcal x\longmapsto \pi [ z,\tau]:=\tau\in\mathbb h\ ] ] is a universal family for all @xmath23-dimensional tori , cf .
@xcite , pp.18 - 19 .
then , we may take as teichmller groupoid , the groupoid @xmath661\ ] ] where the source and target maps are both equal to the projection map @xmath168 of and where composition is just addition .
this must be understood as follows .
the common fibers at a point @xmath318 is the elliptic curve @xmath662 which must be thought of as the translation group of @xmath663 .
observe that even if we are considering tori , the family @xmath614 has a natural section , namely the image of @xmath664 through , allowing a natural identification between @xmath663 and its translation group .
the fact that the source and target maps coincide reflects the stability of the translation groups as explained in remark [ applicor ] .
+ to describe the riemann groupoid , we now just have to add the @xmath656 action . given @xmath665 an element of @xmath656 , recall that @xmath666 just set now @xmath667\ ] ] where the source map is @xmath168 , the target map is given by the @xmath656 action , and composition follows the rule @xmath668 , b\cdot\tau)\circ ( b,[b],\tau)=(ab,[a+b/(rb\cdot \tau+s ) ] , ab\cdot \tau)\ ] ] for @xmath172 defined in .
+ let us treat now the higher dimensional case .
it follows exactly the same pattern .
a universal family is described in @xcite .
one replaces @xmath655 with @xmath669 and one takes the quotient @xmath670 of @xmath671 by the action generated by @xmath672 where @xmath673 is the canonical basis of @xmath674 and @xmath675 the rows of @xmath34
. then the teichmller stack can be presented as @xmath676\ ] ] where the source and target maps are both equal to the projection map and where composition is just addition .
finally , given @xmath677 an element of @xmath678 decomposed into blocks of size @xmath679 , recall that @xmath680 is the action of @xmath678 onto @xmath681 identifying biholomorphic complex tori .
just set now @xmath682\ ] ] where the source map is the projection , the target map is , and composition follows the rule @xmath683 , b\cdot t)\circ ( b,[b],t)=(ab,[a+b(r(b\cdot t)+s)^{-1 } ] , ab\cdot t)\ ] ] the geometric quotients are @xmath681 as teichmller space and the quotient of @xmath681 by the action as riemann space .
notice however that this is far from being an orbifold , cf .
@xcite and @xcite .
* hyperkhler manifolds . * [ hk ] we make the connection between our general results and the beautiful description of the teichmller space for simple hyperkhler manifolds in @xcite , to which we refer for further details .
let @xmath0 be any oriented smooth compact manifold admitting hyperkhler structures .
we restrict @xmath29 to complex structures _ of hyperkhler type_. it has a finite number of connected components .
it follows from proposition 15 of @xcite and the injectivity of the local period map that @xmath1 coincide locally with the kuranishi space .
moreover , we consider only simple hyperkhler structures , that is simply connected ones .
this implies that the first cohomology group with values in the structure sheaf is zero .
so is the group of global @xmath684 holomorphic forms by serre duality .
hence , by pairing , these simple hyperkhler manifolds do not admit any non zero holomorphic vector field .
+ in our setting , this means that 1 .
@xmath1 is tale , see corollary [ etale ? ] , and , taking into account remark [ aut10bis ] , coincides with the holonomy groupoid constructed in subsection [ firstapp ] up to a finite morphism applies and @xmath1 coincides with the holonomy groupoid . ] .
2 . there is no non trivial holonomy germ . in other words , @xmath1 is , up to a finite morphism , the leaf space of a single @xmath282-foliation , and this foliation has no holonomy , so this leaf space is locally hausdorff and coincides locally at each point with the kuranishi space of this point . by bogomolov - tian - todorov theorem ,
the kuranishi space is a manifold , so the leaf space is locally a complex manifold .
+ this is however not enough to imply hausdorffness ; but it forces the inseparable points to lie on a subset of measure zero . and it gives @xmath1 the structure of a non - hausdorff complex manifold .
+ in this particular case , verbitsky shows in theorem 1.15 that the inseparibility condition is an equivalence relation and that the quotient of @xmath1 by this equivalence relation is a hausdorff complex manifold that he calls the birational teichmller space ( taking into account that , following a result by huybrechts , inseparable points correspond to birational hyperkhler manifolds ) .
+ finally , the action of the mapping class group on @xmath1 can be very complicated , see @xcite .
* hopf surfaces . * [ hopfbis ] we go back to the hopf surfaces of example [ hopf ] .
we assume the reader to be acquainted with deformation theory of primary hopf surfaces as detailed in @xcite .
we consider a connected component @xmath33 of @xmath29 .
looking at the @xmath30-homotopy graph of figure [ hopffig ] , we see that it is enough to use kuranishi spaces of type iv and type iii hopf surfaces .
it follows from @xcite and lemma [ hopfcc ] that all type iv can be described as a single kuranishi family constructed as follows , cf .
@xcite and @xcite .
define @xmath687 for @xmath688 and @xmath689 the eigenvalues of @xmath172 .
set @xmath690 then @xmath691 is a versal family for every hopf surface of type iv , which is moreover complete for every surface of type iib and of type iic .
let @xmath176 and define @xmath692 set @xmath693 for @xmath694 .
then @xmath695 is a versal family for every hopf surface of type iii with weight @xmath177 , which is moreover complete for every surface of type iia with weight @xmath177 and of type iic . incorporating the automorphism groups , we define @xmath696 and consider the groupoid @xmath697 where the source map is the projection onto the second factor of ; the target map is the conjugation of the second factor by the first one ; and the composition follows
the rule @xmath698\circ[m , a]=[nm , a]\ ] ] then is a teichmller groupoid for a neighborhood of the @xmath30-homotopy class iv including all type iv , iib and iic hopf surfaces . in the same way ,
let @xmath699 and define @xmath700 with the convention that , given @xmath694 and given @xmath45 with coefficients @xmath701 , then @xmath702 consider the groupoid @xmath703 where the source map is the projection onto the second factor of ; the target map is the conjugation of the second factor by the first one using ; and the composition is given by composition in @xmath704 . then is a teichmller groupoid for a neighborhood of the @xmath30-homotopy class iii of weight @xmath177 including all type iii of weight @xmath177 , iib of weight @xmath177 and iic hopf surfaces .
+ to finish with , we consider the disjoint union of groupoid and of groupoids for all @xmath176 .
we need to add the holonomy morphisms between these groupoids . in this case , it is not even necessary to fat the spaces , since we have natural identifications @xmath705 so we take as presentation of @xmath706 the groupoid whose objects are @xmath707 and whose morphisms are generated by morphisms of and for all @xmath176 from the one hand , and by identifications from the other hand . to be more precise , set @xmath708 and define the supplementary set of morphisms as @xmath709 with source map being the second projection and target map being conjugation of the second factor by the first one composed with identification .
hence the set of morphisms is generated from @xmath710 using the process explained in section [ teichmueller ] .
note that here , the holonomy maps do not make any non trivial loop , hence we finish with a set of morphism which is a hausdorff analytic space . recall that @xmath711 is equal to @xmath6 , hence this gives also a presentation of @xmath6 .
+ finally , we give a model for @xmath711 as a geometric quotient .
consider the map @xmath712 then @xmath713 coincides with the quotient space of @xmath75 by the conjugation action of @xmath169 except for matrices with a single eigenvalue .
+ from this , it is easy to check that the geometric quotient can be constructed as follows .
* start with the domain @xmath714 that is with the image by @xmath182 of the set of invertible matrices with both eigenvalues having modulus strictly less than one .
* double asymmetrically the points of the analytic subspace @xmath715 making @xmath106 non - hausdorff along .
this encodes the fact that above such a point ( for @xmath182 ) , there is two distinct @xmath169-orbits and not a single one .
note that these points correspond to type iv hopf surfaces .
* for each value of @xmath176 , double asymmetrically the points of the analytic subspace @xmath716 making @xmath106 non - hausdorff along .
this encodes the jumping phenomenon of type iii hopf surfaces of weight @xmath177 . by _ doubling asymmetrically the points along some subset _ @xmath717 , we mean that we replace the subset @xmath717 by @xmath718 with the following topology .
the second component is endowed with the topology of @xmath719 . but given any point @xmath161 in the first component of @xmath718 , then every neighborhood of @xmath161 contains also the corresponding point @xmath720 in the other component .
hence , @xmath161 and @xmath720 are not separated , however they do not play the same role and the situation is not symmetric .
[ cstar ] let @xmath721 act by homotheties onto @xmath20 .
then the geometric quotient contains exactly two non - separated points and is obtained from a single point by doubling it asymmetrically .
hence , we can obtain the previous geometric quotient as follows .
consider @xmath722 and take its quotient by @xmath721 acting by homotheties on the @xmath20-factor .
we thus finish with a domain in @xmath723 non - hausdorff along a countable set of analytic curves . at each point corresponding to a type iv or a type iii hopf surface , this space is not locally hausdorff , hence not locally isomorphic to a analytic space .
[ hausdorfftypes ] spaces obtained by doubling asymmetrically the points along some subset @xmath717 are not locally haudorff along @xmath717 since every neighborhood of a point @xmath161 of @xmath717 contains also the double @xmath720 of this point .
in particular , any sequence of points converging onto @xmath161 also converges onto @xmath720 .
this is completely different from the non - hausdorff spaces obtained as leaf spaces of a foliation with no holonomy ( cf .
the teichmller space of simple hyperkhler manifolds , see @xcite and example [ hk ] ) . in this last case ,
given two inseparable points @xmath161 and @xmath720 , we can find neighborhoods of @xmath161 ( respectively @xmath720 ) that do not contain @xmath720 ( respectively @xmath161 ) .
in particular , we can find sequences of points converging to @xmath161 and not converging to @xmath720 ( and vice versa ) .
such spaces are locally hausdorff
. * hirzebruch surfaces . * [ hirzebruchbis ] we go back to the hirzebruch surfaces of example [ hirzebruch ] .
let @xmath187 . to describe @xmath724
, we see from figure [ hirzebruchfig ] that it is enough to use a single kuranishi space , that of @xmath14 . it is equal to @xmath725 and decomposes as a sequence of algebraic cones ( cf .
@xcite , p.21 ) . to be more precise , for any @xmath726 ,
define the algebraic cone @xmath727 of dimension @xmath728 . for any @xmath729 ,
a point @xmath240 of @xmath725 encodes the surface @xmath18
if and only if @xmath730 taking into account that @xmath731 one may check that @xmath732 acts on @xmath725 transitively on each cone ( this follows directly from proposition [ autfol ] ) . + now , we have to take care of the action of the mapping class group , computed in lemma [ hirzebruchcc ] and corollary [ hirzebruchcc2 ] . we can focus on a single connected component of structures , since they are all identified .
hence , we only have to encode the action of the switching map @xmath22 of .
this amounts to consider two copies of @xmath733 and to add the following morphisms : first a holonomy morphism sending a point @xmath734 belonging to the first copy of @xmath725 to the same point in the second copy .
notice that , because of , such a point encodes @xmath183 .
this holonomy morphism is not defined on the points encoding the other hirzebruch surfaces .
this reflects the fact , explained in lemma [ hirzebruchcc ] , that the set of @xmath183 in a connected component of structures is connected whereas that of the other hirzebruch surfaces has two connected components .
then we add the action of @xmath22 , which switches the two copies of @xmath725 .
+ geometrically , we end with a single copy of @xmath725 , but with two ( non - separated ) copies of the cone @xmath735 .
in other words , adapting the vocabulary of example [ hopfbis ] , we double _ symmetrically _ the points of @xmath725 along the cone @xmath735 .
the automorphism group of @xmath14 acts as previously described and the automorphism @xmath22 of @xmath183 fixes @xmath725 but exchanges the two copies of the cone . + more formally , the set of objects of @xmath724 is @xmath736 and the set of morphisms is @xmath737 the third component corresponds to the holonomy morphism and the fourth one to @xmath22 .
source , target and composition can easily be described and we omit the details ( cf . the more complicated example [ hopfbis ] ) . + this describes completely @xmath724 but also @xmath738 . in this last case ,
perform exactly the same construction , but forget about the @xmath22-identification , that is drop the fourth component of .
the geometric quotients of @xmath6 ( respectively @xmath724 ) and @xmath711 ( respectively @xmath738 ) are respectively * @xmath12 ( respectively @xmath739 ) with @xmath740 encoding @xmath18 and with open sets given by @xmath741 , @xmath742 , @xmath743 and so on and * @xmath15 ( respectively @xmath744 ) with @xmath745 encoding @xmath18 and with open sets generated by @xmath741 , @xmath742 , @xmath743 and so on from the one hand , @xmath746 , @xmath747 and so on from the other hand .
meersseman , l. _ feuilletages par varits complexes et problmes duniformisation_. complex manifolds , foliations and uniformization , panoramas & synthses vol .
205 - 257 , editions de la smf , paris , 2011 . | the aim of this paper is to study the structure of the higher - dimensional teichmller and riemann moduli spaces , viewed as stacks over the category of complex manifolds .
we first show that the space of complex operators on a smooth manifold admits a foliation transversely modeled on a translation groupoid , a concept that we define here .
we then show how to construct explicitly a holonomy groupoid for such a structure and show that in this case its objects and morphisms form a finite - dimensional analytic space and its source and target maps are smooth morphisms .
this holonomy data encodes how to glue the local kuranishi spaces to obtain a groupoid presentation of the teichmller and riemann moduli stacks , which can thus be characterized as artin analytic stacks .
this is achieved under the sole condition that the dimension of the automorphism group of each structure is bounded by a fixed integer .
all this can be seen as an answer to question 1.8 of @xcite . |
online analytical processing(olap ) @xcite comprises a set of tools and algorithms that allow efficiently querying multidimensional ( md ) databases containing large amounts of data , usually called data warehouses ( dw ) .
conceptually , in the md model , data can be seen as a _ cube _
, where each cell contains one or more _ measures _ of interest , that quantify _
facts_. measure values can be aggregated along _ dimensions _ , which give context to facts .
at the logical level , olap data are typically organized as a set of _ dimension and fact tables . _
current database technology allows alphanumerical warehouse data to be integrated for example , with geographical or social network data , for decision making . in the era of so - called `` big data '' , the kinds of data that could be handled by data management tools , are likely to increase in the near future .
moreover , olap and business intelligence ( bi ) tools allow to capture , integrate , manage , and query , different kinds of information .
for example , alphanumerical data coming from a local dw , spatial data ( e.g. , temperature ) represented as rasterized images , and/or economical data published on the semantic web . ideally , a bi user would just like to deal with what she knows well , namely the data cube , using only the classical olap operators , like _ roll - up _
, _ drill - down _ , _ slice _ , and _ dice _ ( among other ones ) , regardless the cube s underlying data type .
data types should only be handled at the logical and physical levels , not at the conceptual level .
building on this idea , ciferri et al .
@xcite proposed a _
conceptual _ , _ user - oriented _ model , independent of olap technologies . in this model ,
the user only manipulates a data cube .
associated with the model , there is a query language providing high - level operations over the cube .
this language , called cube algebra , was sketched informally in the mentioned work .
extensive examples on the use of cube algebra presented in @xcite , suggest that this idea can lead to a language much more intuitive and simple than mdx , the _ de facto _ standard for olap .
nevertheless , these works do not give any evidence of the correctness of the languages and operations proposed , other than examples at various degrees of comprehensiveness .
in fact , surprisingly , and in spite of the large corpus of work in the field , a formally - defined reference language for olap is still needed @xcite .
there is not even a well - defined , accepted semantics , for many of the usual olap operations .
we believe that , far for being just a problem of classical olap , this formalization is also needed in current `` big data '' scenarios , where there is a need to efficiently perform real - time olap operations @xcite , that , of course , must be well defined .
[ [ contributions ] ] contributions + + + + + + + + + + + + + in this paper we ( a ) introduce a collection of operators that manipulate a data cube , and clearly define their semantics ; and ( b ) prove , formally , that our operators can be composed , yielding a language powerful enough to express complex queries and cube navigation ( `` _ _ la _ _ olap '' ) paths . we achieve the above representing the data cube as a fixed @xmath0-dimensional matrix , and a set of @xmath1 measures , and expressing each olap operation as a sequence of atomic transformations .
each transformation produces a new measure , and , additionally , when a sequence forms an olap operation , a flag that indicates which are the cells that must be considered as input for the next operation .
this formalism allows us to elegantly define an algebra as a collection of operations , and give a series of properties that show their correctness .
we provide the proofs in the full paper .
we limit ourselves to the most usual operations , namely slice , dice , roll - up and drill - down , which constitute the core of all practical olap tools .
we denote these the _ classical olap operations_. this allows us to focus on our main interest , which is , to prove the feasibility of the approach .
other not - so - usual operations are left for future work .
the main contribution of our work , with respect to other similar efforts in the field is that , for the first time , a formal proof to practical problems is given , so the present work will serve as a basis to build more solid tools for data analysis . existing work either lacks of formalism , or of applicability , and no work of any of these kinds give sound mathematical prove of its claims . in this extended abstract
we present the main properties , and leave the proofs for the full paper .
the remainder of the paper is organized as follows . in section [ sec : data - model ] , we present our md data model , on which we base the rest of our work .
section [ sec : olap - tando ] presents the atomic transformations that we use to build the olap operations . in section [ sec : classicalolap ] we discuss the classical olap operations in terms of the transformations , show how they can be composed to address complex queries .
we conclude in section [ sec : conclusion ] .
in this section we describe the olap data model we use in the sequel .
we next give the definitions of multidimensional matrix schema and instance . in the sequel ,
@xmath0 , with @xmath2 , is a natural number representing the number of dimensions of a data cube .
[ def : matrix - schema ] a _
@xmath0-dimensional matrix schema _ is a sequence @xmath3 of @xmath0 dimension names . as illustrated in the following example
, the convention will be that dimension names start with a capital letter . [
ex : matrix - schema ] the running example we use in this paper , deals with sales information of certain products , at certain locations , at certain moments in time . for this purpose
, we will define a @xmath4-dimensional matrix schema @xmath5 .
[ def : matrix - instance ] a _ @xmath0-dimensional matrix instance _
( _ matrix _
, for short ) over the @xmath0-dimensional matrix schema @xmath6 is the product @xmath7 @xmath8 , where @xmath9 is a non - empty , finite , ordered set , called the _ domain _ , that is associated with the dimension name @xmath10 . for all @xmath8 ,
we denote by @xmath11 , the order that we assume on the elements of @xmath9 . for @xmath12
, we call the tuple @xmath13 , a _ cell _ of the matrix . the cells of a matrix serve as placeholders for the measures that are contained in the data cube ( see definition [ def : data - cube - instance ] below ) .
note that , as it is common practice in olap , we assumed an order @xmath11 on the domain .
the role of the order is further discussed in section [ subsec : order ] . as a notational convention , elements of the domains @xmath9
start with a lower case letter , as it is shown in the following example .
[ ex : matrix - instance ] for the @xmath4-dimensional matrix schema @xmath14 of example [ ex : matrix - schema ] , the non - empty sets @xmath15 , @xmath16 , and @xmath17 produce the matrix instance @xmath18 the cells of the matrix will contain the sales for each combination of values in the domain . in @xmath19 , we have , for instance , the order @xmath20 over the dimension @xmath21 , we have the usual temporal order .
we now define the notions of dimension schema and instance .
[ def : dimension - schema ] let @xmath22 be a name for a dimension .
a _ dimension schema @xmath23 for @xmath22 _ is a lattice , with a unique top - node , called @xmath24 ( which has only incoming edges ) and a unique bottom - node , called @xmath25 ( which has only outgoing edges ) , such that all maximal - length paths in the graph go from @xmath25 to @xmath24 .
any path from @xmath25 to @xmath24 in a dimension schema @xmath23 is called a _ hierarchy _ of @xmath23 .
each node in a hierarchy ( i.e. , in a dimension schema ) is called a _ level _ ( of @xmath23 ) . as a convention ,
level names start with a capital letter .
note that the @xmath25 node is often renamed , depending on the application .
[ ex : dimension - schema ] fig .
[ fig : dimension - schema ] gives examples of dimension schemas @xmath26 and @xmath27 for the dimensions @xmath28 and @xmath21 in example [ ex : matrix - schema ] .
for the dimension @xmath28 , we have @xmath29 , and there is only one hierarchy , denoted @xmath30 the node @xmath31 is an example of a level in this hierarchy . for the dimension @xmath21 , we have @xmath32 , and two hierarchies , namely @xmath33 and @xmath34 . , in ( @xmath35 ) , and @xmath21 , in ( @xmath36 ) . ]
[ def : instancegraph ] let @xmath22 be a dimension with schema @xmath23 , and let @xmath37 be a level of @xmath23 .
a _ level instance of @xmath37 _ is a non - empty , finite set @xmath38 .
if @xmath39 , then @xmath40 is the singleton @xmath41 .
if @xmath42 , then @xmath43 is the the domain of the dimension @xmath22 , that is , @xmath44 ( as in definition [ def : matrix - instance ] ) . a _ dimension graph ( or instance ) _ @xmath45 the dimension schema @xmath23 is a directed acyclic graph with node set @xmath46 where the union is taken over all levels in @xmath23 .
the edge set of this directed acyclic graph is defined as follows .
let @xmath37 and @xmath47 be two levels of @xmath23 , and let @xmath48 and @xmath49 .
then , only if there is a directed edge from @xmath37 to @xmath47 in @xmath23 , there can be a directed edge in @xmath45 from @xmath35 to @xmath50 . if @xmath51 is a hierarchy in @xmath23 , then the _ hierarchy instance _
( relative to the dimension instance @xmath45 ) is the subgraph of @xmath45 with nodes from @xmath38 , for @xmath37 appearing in @xmath51 .
this subgraph is denoted @xmath52 . as notational convention
, the names of objects in a set @xmath38 start with a lower case character .
we remark that a hierarchy instance @xmath52 is always a ( directed ) tree .
also , if @xmath35 and @xmath36 are two nodes in a hierarchy instance @xmath52 , such that @xmath53 is in the transitive closure of the edge relation of @xmath52 , we will say that @xmath35 _ rolls - up _ to @xmath36 and we denote this by @xmath54 ( or @xmath55 if @xmath51 is clear from the context ) .
[ ex : instancegraph ] consider the @xmath28 dimension , whose schema @xmath26 is given in fig .
[ fig : dimension - schema ] ( @xmath35 ) . from example
[ ex : matrix - instance ] , we have @xmath56 @xmath57 @xmath58 @xmath59 , which is @xmath60 , or @xmath61 .
an example of a dimension instance @xmath62 is depicted in fig .
[ fig : dimension - instance ] .
this example expresses , for instance , that the city @xmath63 is located in the region @xmath64 which is part of the country @xmath65 , meaning that @xmath63 rolls - up to @xmath64 and to @xmath65 , that is , @xmath66 and @xmath67 . .
] in a dimension graph , we must guarantee that rolling - up through different paths gives the same results .
this is formalized by the concept of `` sound '' dimension graph .
[ def : sound - graph ] let @xmath45 be a dimension graph ( as in definition [ def : instancegraph ] ) .
we call this dimension graph _ sound _ , if for any level @xmath37 in @xmath23 and any two hierarchies @xmath68 and @xmath69 that reach @xmath37 from the @xmath25 level and any @xmath70 and @xmath71 , we have that @xmath72 and @xmath73 imply that @xmath74 . in this paper , we assume that dimension graphs are always sound . essentially , a data cube is a matrix in which the cells are filled with measures that are taken from some _ value domain _ @xmath75 . for many applications
, @xmath75 will be the set of real or rational numbers , although some other ones may include , e.g. , spatial regions or geometric objects .
[ def : data - cube - schema ] a _ @xmath0-dimensional data cube schema _ consists of ( a ) a @xmath0-dimensional matrix schema @xmath6 ; and ( b ) a hierarchy schema @xmath76 for each dimension @xmath10 , with @xmath77 .
[ def : data - cube - instance ] let @xmath75 be a non - empty set of `` values '' . a _
@xmath0-dimensional , @xmath1-ary data cube instance _ ( or _ data cube _ , for short ) @xmath78 over the @xmath0-dimensional matrix schema @xmath6 and hierarchy schemas @xmath76 for @xmath10 , for @xmath77 , with values from @xmath75 , consists of ( a ) a @xmath0-dimensional matrix instance over the matrix schema @xmath6 , @xmath79 ; ( b ) for each @xmath77 , a _ sound _
dimension graph @xmath80 over @xmath76 ; ( c ) @xmath1 _ measures _ @xmath81 , which are functions from @xmath82 to the value domain @xmath75 ; and ( d ) a _ flag _ @xmath83 , which is a function from @xmath84 to the set @xmath85 . also ,
as a notational convention , we use calligraphic characters , like @xmath78 , to represent data cube instances .
the flag @xmath83 can be considered as a @xmath86-st boolean measure .
the role of @xmath83 is to indicate which of the matrix cells are currently `` active '' .
the active cells have a flag value @xmath87 and the others have a flag value @xmath88 .
when we operate over a data cube , flags are used to indicate the input or output parts of the matrix of the cube .
typically , in the beginning of the operations , all cells have a flag value of @xmath87 .
the role of flags will become more clear in the next sections , when we discuss olap transformations and operations .
when performing olap transformations and operations , we may need to store aggregate information about certain measures up to some level above the @xmath25 one .
we do not want to use extra space for this in the data cube .
instead , we use the available cells of the original data cube to store this information . for this , we make use of the order assumed in definition [ def : matrix - instance ] , for the representation of high - level objects by @xmath25-level objects .
[ def : represents ] let @xmath89 be an arbitrary dimension with domain @xmath90 .
let @xmath37 be a level of @xmath23 .
an element @xmath91 is _ represented _ by the smallest element @xmath92 ( according to @xmath11 ) for which @xmath55 holds .
we denote this as , and say that .
[ ex : order ] continuing with the previous examples , we consider the dimension @xmath28 with @xmath93 ( i.e. , @xmath61 . on this set ,
we _ assume _ the order @xmath94 .
for this dimension , we have the hierarchy and the dimension instance , given in figs . [
fig : dimension - schema ] and [ fig : dimension - instance ] , respectively . at the @xmath29 level ,
cities represent themselves . at higher levels , regions and countries
are represented by their `` first '' city in @xmath95 ( according to @xmath11 ) .
thus , @xmath96 and @xmath65 are represented by @xmath97 , @xmath98 is represented by @xmath99 , and @xmath100 is represented by @xmath101 . at the level @xmath24
, @xmath97 represents @xmath102 note that the @xmath25-level representatives of higher - level objects , will be flagged @xmath87 , and other cells flagged @xmath88 .
also , in our example , if we aggregate information at level @xmath31 , with @xmath103 , then all cities in @xmath95 become flagged .
thus , it would not be clear if the cube contains information at the level @xmath104 or at the level @xmath31 .
the following property shows how the order on the @xmath25 level induces and order on higher levels .
[ prop : order - higher - level ] let @xmath89 be a ( sound ) dimension of a data cube @xmath78 and let @xmath37 be a level in the dimension schema @xmath23 .
the order @xmath11 on @xmath44 induces an order on @xmath38 as follows .
if @xmath71 , then @xmath105 if and only if @xmath106 .
a typical olap user manipulates a data cube by means of well - known operations .
for instance , using our running example , the query `` total sales by region , for regions in belgium or france '' , is actually expressed as a sequence of operations , whose semantics should be clearly defined , and which can be applied in different order . for example , we can first apply a _ roll - up _
( i.e. , an aggregation ) to the _ country _ level , and once at that level apply a _ dice _ operation , which keeps the cube cells corresponding to belgium or france .
finally , a _ drill - down _ can be applied to disaggregate the sales down to the level _ region _ , returning the desired result . in what follows ,
we characterize olap operations as the result of sequences of olap transformations , which are measure - creating updates to a data cube . an _ atomic olap transformation _
acts on a data cube instance , by adding a measure to the existing data cube measures .
olap operations like the ones informally introduced above are defined , in our approach , as a sequence of transformations .
the process of olap transformations starts from a given _ input data cube _ .
we assume that this original data cube has @xmath1 given measures @xmath81 ( as in definition [ def : data - cube - instance ] ) .
these @xmath1 measures have a special status in the sense that they are `` protected '' and can never be altered ( see section [ subsec : olap - operation ] ) . typically , the input - flag @xmath83 of the original data cube @xmath107 is set to @xmath87 in every cell and signals that every cell of @xmath108 is part of the input cube .
atomic olap transformations can be applied to data cubes .
they add ( or create ) new measures to the sequence of existing measures by adding new measure values in each cell of the data cube s matrix . at any moment in this process
, we may assume that the data cube @xmath78 has @xmath109 measures @xmath110 , where the first @xmath1 are the original measures of @xmath107 , and the last @xmath111 ( with @xmath112 ) ones have been created subsequently by @xmath111 olap transformations ( where @xmath113 is the empty sequence of @xmath114 s , for @xmath115 ) .
the next olap transformation adds a new measure @xmath116 to the matrix cells .
we have said that we use olap transformations to compute olap operations .
we indicate that the computation of an olap operation @xmath117 is finished by creating an @xmath118-ary output flag @xmath119 .
this output flag is a boolean measure , that is created via atomic olap transformations .
it indicates which of the cells of @xmath79 should be considered as belonging to the output of @xmath117 .
it is @xmath118-ary in the sense that it keeps the last @xmath118 created measures @xmath120 and `` trashes '' the rest .
it also removes the previous flag , which it replaces .
the initial measures @xmath81 of the input data cube @xmath107 are never removed ( unless they are `` destroyed '' in some cells ) .
they remain in the cube throughout the process of applying one olap operation after another to @xmath107 , and can be used at any stage .
summarizing , after an olap operation of output arity @xmath118 is completed on some cube @xmath78 , the measures in the cells of the output data cube @xmath121 are of the form @xmath122 here , the underlining indicates the protected status of these measures . after each olap operation ,
the unprotected measures with the symbols @xmath123 and the output measures become @xmath124 the next olap operation @xmath125 can then act on @xmath126 and use in its computation all the measures above .
we remark that the dimensions , the hierarchy schemas and instances of @xmath78 remain unaltered during the entire olap process .
we end this description with a remark on _
destructors_. a destructor , optionally , precedes the creation of an output flag .
a destructor @xmath127 takes the value @xmath87 for some cells of the matrix of a data cube , and @xmath88 on other cells .
when @xmath127 is invoked ( and activated by the output flag that follows it ) on a data cube @xmath78 with measures @xmath128 and flag @xmath129 , it empties all cells for which the value of the destructor @xmath127 is @xmath88 by removing all measures from them , even the protected ones , thereby effectively `` destroying '' these cells .
this is the only case where the protected measures are altered ( see operations slice or dice , later ) .
the output of a destructive operation @xmath117 looks like @xmath130 in which the destructor precedes the output flag .
the effect of the presence of a destructor is the following .
a cell such that @xmath131 is emptied , after which it contains no more measures and flag . for cells with @xmath132 , the sequence of measures @xmath133
is transformed to @xmath134 which is renamed as @xmath135 before the next transformation takes place .
this transformation will act , cell per cell , on the matrix of a cube , and it does nothing with emptied cells .
that is , no new measure can ever be added to a destroyed cell .
the following definition specifies how an olap transformation acts on a data cube .
we then address in detail each atomic olap transformation appearing in this definition .
[ def : olap - transformation ] let @xmath78 be a @xmath0-dimensional , @xmath136-ary data cube instance with given ( or protected ) measures @xmath81 , created measures @xmath137 ( with @xmath112 ) and flag @xmath83 over some value domain @xmath75 .
olap transformation _ @xmath138 , applied to @xmath78 , results in the creation of a new measure @xmath116 in @xmath78 .
transformation @xmath138 adds measure @xmath116 ; @xmath116 is produced from : @xmath81 ; @xmath83 ; @xmath139 and the hierarchy schemas and instances of @xmath78 ; and belongs to one of the following classes : ( a ) arithmetic transformations ( definition [ def : trans - arith ] ) ; ( b ) boolean transformations ( definition [ def : trans - boolean ] ) ; ( c ) selectors ( definition [ def : trans - selector ] ) ; ( d ) counting , sum , min - max ( definitions [ def : trans - counting ] , [ def : trans - min - max - revisited ] ) ; ( e ) grouping ( definition [ def : trans - grouping ] ) .
an olap transformation can also result in the creation of a measure that is an output flag @xmath140 of arity @xmath118 .
this should be a measure with a boolean value . to indicate that it is a flag of arity @xmath118
, we use the reserved symbol @xmath140 instead of @xmath116 .
an output flag @xmath140 may ( optionally ) be preceded by a destructor @xmath127 .
this should be a measure with a boolean value ( to indicate which cells are destroyed ) .
we use the reserved symbol @xmath127 instead of @xmath116 .
before we give the definition of an olap operation , we describe the _ input _ to the olap process ( this process may involve multiple olap operations ) .
such input is a @xmath0-dimensional , @xmath1-ary data cube instance @xmath107 , with measures @xmath81 and flag @xmath83 .
these measures are _ protected _ in the sense that they remain the first @xmath1 measures throughout the entire olap process and are never altered or removed unless they are destroyed in some cells .
the cube @xmath107 has also a boolean flag @xmath83 , which typically has value @xmath87 in all cells of @xmath108 .
thus , the measures of the input cube @xmath107 are denoted @xmath141 after applying a sequence of olap operations to @xmath107 , we obtain a data cube @xmath142 . let be a @xmath0-dimensional , @xmath136-ary _ input _ data cube instance with given measures @xmath81 , computed measures @xmath113 and flag @xmath83 .
the data cube @xmath78 acts as the input of an _ olap operation _
@xmath117 ( of arity @xmath118 ) , which consists of a sequence of @xmath143 consecutive olap transformations that create the additional measures @xmath144 , followed by the creation of an @xmath118-ary flag @xmath119 . as the result of the creation of @xmath119 ,
the measures in the cells of the data cube are changed from @xmath145 to @xmath146 which become @xmath147 after renaming .
the output cube @xmath121 has the same dimensions , hierarchy schemas and instances as @xmath142 , and measures @xmath148 @xmath149 in the case where @xmath119 is preceded by a destructor @xmath127 , the same procedure is followed , except for the cells of @xmath79 for which @xmath127 takes the value @xmath88 .
these cells of @xmath79 are emptied , contain no measures , and become inaccessible for future transformations .
we now address the five classes of atomic olap transformations of definition [ def : olap - transformation ] .
we use the following notational convention . for a measure @xmath150
, we write @xmath151 to indicate the value of @xmath150 in the cell @xmath152 we remark that @xmath151 does not exist for empty cells and it is thus not considered in computations .
also , we assume that there are _ protected _ measures @xmath81 , and _ computed _
measures @xmath113 in the non - empty cells , and call @xmath116 the next computed measure .
[ def : trans - arith ] the following creations of a new measure @xmath116 are _ arithmetic transformations _ : 1 .
( * constant * ) @xmath153 , with @xmath154 , a rational number .
( * sum * ) @xmath155 , with @xmath156 .
( * product * ) @xmath157 , with @xmath156 .
( * quotient * ) @xmath158 , with @xmath156 .
[ def : trans - boolean ] the following creations of a new measure @xmath116 are _ boolean transformations _ : 1 .
( * equality test on measures * ) @xmath159 , with @xmath160 . here , the result of @xmath161 is a boolean 1 or 0 ( ) .
( * comparison test on measures * ) @xmath162 , with @xmath163 . here
, the result of the comparison @xmath164 is a boolean 1 or 0 ( ) .
( * equality test on levels * ) 4 .
( * comparison test on levels * ) for a level @xmath37 in the dimension schema @xmath76 of dimension @xmath10 , and a constant @xmath165 , @xmath166 is a `` comparison '' test .
the result of @xmath167 is a boolean 1 or 0 ( ) , such that @xmath168 is @xmath87 if and only if @xmath169 rolls - up to an object @xmath36 at level @xmath37 for which @xmath170 .
the order @xmath171 can be any order that is defined on level @xmath37 .
transformation @xmath172 is defined similarly .
[ ex : trans - boolean-1 ] we illustrate the use of boolean transformations by means of a sequence of transformations that implement a `` dice '' ( see section [ subsec : dice ] for more details ) .
the query @xmath173 asks for the cells in the matrix of @xmath78 which contain sales that are higher than 50 .
this query can be implemented by the following sequence of transformations : * @xmath174 ( rational constant ) ; * @xmath175 ( comparison test on measures ) ; * @xmath176 ( product ) ; * @xmath177 ( destructor ) ; and * @xmath178 ( unary flag ) the measure @xmath179 contains the @xmath180 values larger than or equal to 50 ( and a 0 if the @xmath180 are lower ) .
the destructor @xmath127 destroys the cells that contain a o. finally , the flag @xmath181 selects all cells from the input as output cells ( it will contain a 1 for all such cells that satisfy the condition ) , and concludes the @xmath173 operation .
the output of this operation is @xmath182 which is then renamed to @xmath183 [ def : trans - selector ] the following creations of a new measure @xmath116 are _ selector transformations _ ( or _ selectors _ ) , and their definition is cell per cell of @xmath79 : 1 .
( * constant selector * ) for a level @xmath37 in the dimension schema @xmath76 of a dimension @xmath10 , and @xmath165 , @xmath116 can be a _ constant - selector for @xmath184 _ , denoted @xmath185 , 2 .
( * level selector * ) for a level @xmath37 in the dimension schema @xmath76 of a dimension @xmath10 , @xmath116 can be a _ level - selector for @xmath37 _ , denoted by @xmath186 , which means that we have , for all @xmath187 with @xmath188 , @xmath189 the _ constant _ selector in definition [ def : trans - selector ] , corresponds to the equality test on levels ( see 3 . in definition [ def : trans - boolean ] ) . here
, this transformation appears with a different functionality and we reserve a special notation for it , and we repeated it .
also , note that the _ level _ selector selects all representatives ( at the @xmath25 level ) of objects at level @xmath37 of dimension @xmath10 .
[ ex : trans - selector-2 ] the query @xmath190 asks for the sales in the cities of @xmath97 and @xmath63
. it can be implemented by the following sequence of transformations , where @xmath179 can take values @xmath88 or @xmath87 , since the cities @xmath97 and @xmath63 do not overlap : * @xmath191 ( constant selector ) ; * @xmath192 ( constant selector ) ; * @xmath193 ( sum ) ; * @xmath194 ( product ) ; * * @xmath195 ( unary flag creation ) .
[ def : trans - counting ] the creations of a new measure @xmath116 defined next , are denoted 1 .
( * count - distinct * ) @xmath196 , @xmath197 counts the number of distinct values of measure @xmath150 in the complete matrix @xmath79 of the data cube .
( * @xmath0-dimensional sum * ) @xmath198 with @xmath199 @xmath200 , gives the sum of the measure @xmath150 over all matrix cells .
we abbreviate this operation by writing @xmath201 and call this transformation the _
@xmath0-dimensional sum_. 3 .
( * min - max * ) @xmath202 , with @xmath197 , gives the smallest value of the measure @xmath150 the matrix @xmath79 .
similarly , @xmath203 , gives the largest value of the measure @xmath150 in the matrix @xmath79 .
it is important to remark that the above transformations create the _
same new measure value _ for all cells of the matrix @xmath79 .
[ ex : trans - counting-2 ] now , we look at the query @xmath204 the query can be computed as follows , given @xmath205 : * @xmath191 ( constant selector on @xmath97 ) ; * @xmath206 ( product that selects the sales in @xmath97 , puts a 0 in all other ones ) ; * @xmath207 ( this is the total sales in @xmath97 in every cell ) ; * @xmath208 ( this is the total sales in @xmath97 in the cells of @xmath97 ) ; * @xmath209 ( this flag creation selects the cells of @xmath97 ) .
the output measures are @xmath210 , which are renamed @xmath211 .
thus , the value of the total of sales in @xmath97 is now available in every cell corresponding to @xmath97 . for the cells outside @xmath97
there is a @xmath88 .
the most common olap operations ( e.g. , roll - up , slice ) , require grouping data before aggregating them .
for example , typically we will ask queries like `` total sales by city '' , which requires grouping facts by city , and , for each group , sum all of its sales .
therefore , we need a transformation to express `` grouping '' .
to deal with grouping , we use the concept of `` prime labels '' for sets and products of sets .
we will use these labels to identify elements in dimensions and in dimension levels . before giving the definition of the grouping transformations , we elaborate on . as we show , these prime labels work in the context of measures that take rational values ( as it is often the case , in practice ) .
the following definition specifies our infinite supply of prime labels .
[ def : prime - labels ] let @xmath212 denote the @xmath143-th prime number , for @xmath213 .
we define the sequence of _ prime labels _ as follows : @xmath214 we denote the set of all prime labels by @xmath215 .
[ def : prime - labelling - of - sets ] let @xmath216 , @xmath217 be ( finite ) sets .
a _ prime labeling _ of the set @xmath216 is an injective function @xmath218 . for @xmath219 , we call @xmath220 the _ prime label _ of @xmath35 ( for the prime labeling @xmath221 )
. let @xmath222 be a subset of @xmath223 , which serves as an index set .
a of the cartesian product @xmath224 consists of prime labelings @xmath225 of the sets @xmath226 , for @xmath227 , that satisfy the condition that @xmath228 is empty for @xmath229 and @xmath230 . for @xmath231 , we call @xmath232 the _ prime product @xmath222-label _ of @xmath233 ( given the prime labelings @xmath225 , for @xmath227 )
. when @xmath222 is a strict subset of @xmath223 , we speak about a and when @xmath234 , we speak about a . if we view a cartesian product @xmath224 as a finite matrix , whose cells contain rational - valued measures , we can use prime ( product ) labelings as follows in the aggregation process .
let us assume that the cells of @xmath224 contain rational values of a measure @xmath235 and let us denote the value of this measure in the cell @xmath233 by @xmath236 .
if we have a full prime product labeling on @xmath224 , then we can consider the sum over this cartesian product of the product of the prime product labels with the value of @xmath235 : @xmath237 since each cell of @xmath224 has a unique prime product label , and since these labels are rationally independent ( see property [ prop : prime - sum ] ) , this sum enables us to retrieve the values
@xmath238 if we have a partial prime product labeling on @xmath224 , determined by an index set @xmath222 , then , again , we can consider the sum over this cartesian product of the product of the partial prime product labels with the value of @xmath235 : @xmath239 now , all cells in @xmath224 above a cell in the projection of @xmath224 , receive the same prime label .
this means that these cells are `` grouped '' together and the above sum allows us to retrieve the part of the sum that belongs to each group .
the following definition gives a name to the above sums .
[ def : prime - sums ] the following property can be derived from the well - known fact that the field extension @xmath240 has degree @xmath241 over @xmath242 and corollaries of this property ( see chapter 8 in @xcite ) .
no square root of a prime number is a rational combination of square roots of other primes .
[ prop : prime - sum ] let @xmath213 and let @xmath224 be a cartesian product of finite sets .
we assume that the cells @xmath233 of this set contain rational values @xmath236 of a measure @xmath235 .
let @xmath222 be a subset of @xmath223 and let @xmath225 be prime labelings of the sets @xmath226 , for @xmath227 , that form a prime product @xmath222-labeling .
then , the prime sum @xmath243 uniquely determines the values @xmath244 for all cells of @xmath245 .
we remark that we use these prime ( product ) labels in a purely _ symbolic _ way without actually calculating the square root values in them .
we are now ready to define atomic olap operations that allow us to implement grouping . in what follows
, we apply these prime labels to the case where the sets @xmath226 in @xmath246 are domains of dimensions ( e.g. , at the bottom level ) , or domains of dimensions at some level .
[ def : trans - grouping ] the following creations of a new measure @xmath116 are _ grouping transformations _ :
( * * ) let @xmath10 be a dimension and @xmath37 a level in the dimension schema @xmath76 of a dimension @xmath10 .
let @xmath247 with induced order @xmath248 ( see property [ prop : order - higher - level ] ) .
if the prime labels @xmath249 have been used by previous transformations , then for , we have @xmath250 if @xmath251 .
we denote this transformation by @xmath252 or @xmath253 , for short , and call the result of such a transformation a _ prime labeling_. 2 .
( * projection of a prime sum * ) if the result of some previous transformation @xmath254 is a @xmath255 ( over the complete matrix @xmath79 ) in which prime ( product ) labels @xmath256 ( computed in a previous transformation @xmath257 ) are used , then @xmath116 is a new measure that `` projects '' on the appropriate component from the prime sum , that is , @xmath258 if the prime ( product ) label @xmath259 .
we denote this .
[ ex : trans - grouping-3 ] consider the query @xmath260 @xmath261 this query can be implemented as follows ( explained below , using the data in example [ ex : instancegraph ] ) : * @xmath262 ( this gives each country a prime label ) ; * @xmath263 ( this gives each city a prime label ) ; * @xmath264 ( this gives each city a product of prime labels ) ; * @xmath265 ; * @xmath266 ( gives each product a different prime label ) ; * @xmath267 ( counts the number of products ) ; * @xmath268 ( gives each time moment a different prime label ) ; * @xmath269 ( counts the number of moments in time ) ; * @xmath270 ( is the number of products times the number of time moments ) ; * @xmath271 ( normalization of the sum ) ; * @xmath272 ; ( projection over the prime labels of city ) ; * @xmath273 ( 3-dimensional sum ) ; * @xmath274 ( normalization of the sum ) ; * @xmath275 ( projection over the prime labels of country ) ; * @xmath276 ( this flag creation selects all cells of the matrix ) .
transformation @xmath277 gives each country a next available prime label .
since no labels have been used yet , @xmath65 gets label @xmath87 and @xmath98 gets label @xmath278 .
transformation @xmath279 gives each city a next available prime label .
since @xmath87 and @xmath278 have been used , @xmath97 gets label @xmath280 , @xmath63 gets label @xmath281 , @xmath99 gets label @xmath282 , and @xmath101 gets label @xmath283 .
transformation @xmath284 gives @xmath97 the value @xmath280 ( i.e. , @xmath285 , @xmath63 the value @xmath281(@xmath286 ) , @xmath99 the value @xmath287 ( @xmath288 ) , and @xmath101 the value @xmath289 ( @xmath290 ) .
if there are 10 products and 100 time moments , then @xmath291 puts the value @xmath292 in each cell of the matrix @xmath79 .
transformations @xmath293 and @xmath294 count the number of products and the number of time moments ( using fresh prime labels ) , and the product of these quantities is computed in @xmath295 . in @xmath296 , @xmath179 is divided by this product , putting @xmath297 in every cell .
transformation @xmath298 is a projection on the prime labels of @xmath104 .
since @xmath280 , @xmath281 , @xmath282 , and @xmath283 are the prime labels for the cities , and since @xmath299 , this will put @xmath87 in the cells of @xmath97 and @xmath63 , and @xmath278 in the cells of @xmath99 and @xmath101 .
next , @xmath300 puts @xmath301 in every cell of the cube and @xmath302 puts @xmath303 in every cell of the cube . finally , @xmath304 projects on the prime labels of countries , which are 1 and @xmath278 .
this puts a 2 in every cell of a belgian city and a 2 in every cell in a french city .
this is the result of the query , as the flag indicates , that is returned in every cell .
now every cell of a city in @xmath65 has the count of @xmath305 cities , as has every city in @xmath98 .
we can now extend the transformations of definition [ def : trans - counting ] , in a way that the counting , minimum , and maximum , are taken over cells which share a common prime product label .
[ def : trans - min - max - revisited ] the following creations of a new measure @xmath116 are : 1 .
( * count - distinct * ) 2 .
( * min - max * ) we remark that when there is only one prime label throughout @xmath79 , the above generalization of the counting and min - max transformations correspond to definition [ def : trans - counting ] .
in this section , we prove that the classical olap operations can be expressed using the olap transformations from section [ sec : olap - tando ] .
these classic operations can be combined to express complex analytical queries .
the classical olap operations are dice , slice , slice - and - dice , roll - up and drill - down ( see section [ subsec : rollup ] ) .
we assume in the sequel , that the input data cube has @xmath1 given measures @xmath81 , and that at some point in the olap process this cube is transformed to a cube , having measures @xmath306 where @xmath307 , with @xmath112 , are created measures and @xmath83 is an input / output flag .
before we start , we need to define the notion of a boolean cell - selection condition , and give a lemma about its expressiveness we will use throughout section [ sec : classicalolap ] . [ def : cell - selection ] let @xmath308 be the matrix of @xmath78 .
a _ boolean condition on the cells of @xmath79 _ is a function @xmath309 from @xmath79 to @xmath85 .
we say that a boolean condition @xmath309 is _ transformation - expressible _ if there is a sequence of olap transformations @xmath310 such that @xmath311 for all @xmath312 .
[ lemma : boolean - closure ] if @xmath313 are transformation - expressible boolean conditions on cells , then @xmath314 , @xmath315 , and @xmath316 are transformation - expressible boolean conditions on cells . intuitively , the _ dice _ operation selects the cells in a cube @xmath78 that satisfy a boolean condition @xmath309 on the cells .
the syntax for this operation is @xmath317 where @xmath309 is a boolean condition over level values and measures .
the resulting cube has the same dimensionality as the original cube .
this operation is analogous to a selection in the relational algebra . in a data cube
, it selects the cells that satisfy the condition @xmath309 by flagging them with a @xmath87 in the output cube .
our approach covers all typical cases in real - world olap @xcite .
we next formalize the operator s definition in terms of our transformation language . in the remainder
, we use the term _ olap operation _ to express a sequence of olap transformations .
[ def : dice ] given a data cube @xmath318 , the operation @xmath317 selects all cells of the matrix @xmath79 that satisfy the boolean condition @xmath309 by giving them a @xmath87 flag in the output .
the condition @xmath309 is a boolean combination of conditions of the form : ( a ) a selector on a value @xmath36 at a certain level @xmath37 of some dimension @xmath10 ; ( b ) a comparison condition at some level @xmath37 from a dimension schema @xmath76 of a dimension @xmath10 of the cube of the form @xmath319 or @xmath320 , where @xmath184 is a constant ( at that level @xmath37 ) ; ( c ) an equality or comparison condition on some measure @xmath150 of the form @xmath321 , @xmath322 or @xmath323 , where @xmath184 is a ( rational ) constant . [ prop : dice ] let @xmath78 be a data cube en let @xmath309 be a boolean condition on the cells of @xmath79 ( as in definition [ def : dice ] ) .
the @xmath324 is expressible .
intuitively , the _ slice _
operation takes as input a @xmath0-dimensional , @xmath1-ary data cube @xmath78 and a dimension @xmath10 and returns as output @xmath325 , which is a `` @xmath326-dimensional '' data cube in which the original measures @xmath327 are replaced by their aggregation ( sum ) over different values of elements in @xmath9 . in other words ,
dimension @xmath10 is removed from the data cube , and will not be visible in the next operations .
that means , for instance , that we will not be able to dice on the levels of the removed dimension .
as we will see , the `` removal '' of dimensions is , in our approach , implemented by means of the destroyer measure @xmath127 .
we remark that the aggregation above is due to the fact that , in order to eliminate a dimension @xmath10 , this dimension should have exactly one element @xcite , therefore a roll - up ( which we explain later in section [ subsec : rollup ] ) to the level _ all _ in @xmath10 is performed .
[ def : slice ] given a data cube @xmath318 , and one of its dimensions @xmath10 , the operation @xmath325 `` replaces '' the measures @xmath328 by their aggregation ( sum ) @xmath329 ( for @xmath330 ) as : @xmath331 for all @xmath332 .
we abbreviate the above @xmath87-dimensional sum as @xmath333 [ prop : slice ] let @xmath318 be a data cube and let @xmath10 be one of its dimensions .
the @xmath325 is expressible as an olap operation .
[ ex : slice ] consider dimensions @xmath334 @xmath335 and @xmath336 , and measure @xmath337 in our running example .
the operation @xmath338 returns a cube with @xmath339-cells containing the sums of @xmath340 for each product - time combination , over all location .
all cells not belonging to the representative of @xmath341 in the dimension @xmath28 ( i.e. , @xmath97 ) , are destroyed .
the query is expressed by the following transformations .
* @xmath342 ( prime labels on products ) ; * @xmath343 ( fresh prime labels on time moments ) ; * @xmath344 ( product of the two previous prime labels ) ; * @xmath345 ( product ) ; * @xmath346 ( @xmath4-dimensional sum ) ; * @xmath347 ( projection on prime product labels ) ; * @xmath348 ( selects the representative of @xmath341 in the dimension @xmath28 ) ; * * @xmath349 ( this flag creation selects the relevant cells of the matrix ) .
transformation @xmath350 gives each @xmath339-combination a unique prime product label .
this label is multiplied by the @xmath180 in each cell .
then , @xmath351 is the global sum over @xmath79 ; @xmath347 is the projection over the prime product labels for @xmath339-combinations .
this gives each cell above some fixed @xmath339-combination , the sum of the @xmath180 , over all locations , for that combination .
all cells of @xmath79 that do not belong to @xmath97 ( selected in @xmath352 ) , which represents @xmath341 , are destroyed by @xmath127 .
a particular case of the _ slice _ operation occurs when the dimension to be removed already contains a unique value at the bottom level .
then , we can avoid the roll - up to _ all _ , and define a new operation , called _ slice - and - dice_. although this can be seen as a _ dice _ operation followed by a @xmath353 one , in practice , both operations are usually applied together .
[ def : slice - dice ] [ prop : slice - dice ] let @xmath78 be a data cube , @xmath10 on of its dimensions en let @xmath354 .
the operation @xmath355 is expressible as an olap operation .
[ ex : slice - dice ] in our running example , the operation @xmath356 is implemented by the output flag @xmath357 . intuitively , _
roll - up _ aggregates measure values along a dimension up to a certain level , whereas _ drill - down _ disagregates measure values down to a dimension level . although at first sight it may appear that _ drill - down _ is the inverse of _ roll - up _
@xcite , this is not always the case , e.g. , if a _
roll - up _ is followed by a @xmath353 or a @xmath358 ; here , we can not just undo the _ roll - up _ , but we need to account for the cells that have been eliminated on the way .
more precisely , the _ roll - up _ operation takes as input a data cube @xmath78 , a dimension @xmath10 and a subpath @xmath359 of a hierarchy @xmath51 over @xmath10 , starting in a node @xmath47 and ending in a node @xmath360 , and returns the aggregation of the original cube along @xmath10 up to level @xmath37 for some of the input measures @xmath361 . _
roll - up _ uses one of the classic sql aggregation functions , applied to the indicated protected and computed measures @xmath361 ( selected from @xmath362 ) , namely sum ( @xmath363 ) , average ( @xmath364 ) , minimum /maximum ( @xmath365 and @xmath366 ) , count and count - distinct ( @xmath367 and @xmath368 ) .
usually , measures have an associated _ default _ aggregation function .
the typical aggregation function for the measure @xmath180 , e.g. , is @xmath363 .
we denote the above operation as @xmath369 where @xmath370 is one of the above aggregation functions that is associated to @xmath371 , for @xmath372 .
since we are mainly interested in the expressiveness of this operation as a sequence of atomic transformations , only the destination node @xmath37 in the path @xmath359 is relevant . indeed
, the result of this roll - up remains the same if the subpath @xmath359 is extended to start from the @xmath25 node of dimension @xmath10 .
so , we can simplify the notation , replacing @xmath373 with @xmath374 and assume that the roll - up starts at the @xmath25 level .
the _ drill - down _ operation takes as input a data cube @xmath78 , a dimension @xmath10 and a subpath @xmath359 of a hierarchy @xmath51 over @xmath10 , starting in a node @xmath37 and ending in a node @xmath375 ( at a lower level in the hierarchy ) , and returns the aggregation of the original cube along @xmath10 from the bottom level up to level @xmath47 .
the drill - down uses the same type of aggregation functions as the roll - up . again
, since we are only interested in the expressiveness of this operation , the drill - down operation @xmath376 has the same output as @xmath377 [ def : rollup ] given a data cube @xmath318 , one of its dimensions @xmath10 , and a hierarchy @xmath51 over @xmath10 , ending in a node @xmath360 , the operation @xmath378 computes the aggregation of the measures @xmath371 by their aggregation functions @xmath370 , for @xmath372 , as follows : @xmath379 for all @xmath332 , for which @xmath380 , for some @xmath381 .
this roll - up flags all representative @xmath25-level objects as active .
[ prop : rollup ] let @xmath318 be a data cube , let @xmath10 be one of its dimensions , and let @xmath51 be a hierarchy over @xmath10 ending in a node @xmath37 .
let @xmath382 be a set of selected measures ( taken from the protected measures @xmath383 and the computed measures @xmath384 of @xmath318 ) , with their associated aggregation functions .
the operation @xmath385 is expressible as an olap operation .
[ ex : rollup-1 ] we next express the _ roll - up _ operation , using prime ( product ) labels , sums , projections , and the @xmath4-dimensional sum .
we look at the query `` total sales per country '' .
we use the simplified syntax , only indicating the target level of the roll - up on the _ location _ dimension ( i.e. , _ country _ ) .
the query @xmath386 is the result of the following transformations , given the measure @xmath205 : 1 .
@xmath387 ( prime labels on products ) ; 2 . @xmath388 ( prime labels on time moments ) ; 3 .
@xmath389 ( prime labels on countries ) ; 4 .
@xmath390 ; ( prime product label in one step ) ; 5 .
@xmath391 ( product of labels with @xmath180 ) ; 6 .
@xmath392 ( @xmath4-dimensional sum ) ; 7 . @xmath393 ( projection on prime product labels )
@xmath394 ( output flag on country - representatives ) .
transformation @xmath395 gives every product - date - country combination a unique prime product label .
normally this product takes more steps . above
, we have abbreviated it to one transformation .
the transformation @xmath396 gives the aggregation result , and @xmath397 is the flag that says that only the cities @xmath97 and @xmath99 , which represent the level @xmath398 , are active in the output ( and nothing else of the original cube ) .
the main result of this paper is the proof of the completeness of an olap algebra , composed of the olap operations dice ( section [ subsec : dice ] , slice ( section [ subsec : slice ] ) , slice - and - dice ( section [ subsec : sandd ] ) , roll - up , and drill - down ( section [ subsec : rollup ] ) .
this is summarized by theorem [ theo : main ] .
[ theo : main ] the classical olap operations and their composition are expressible by olap operations ( that is , as sequences of atomic olap transformations ) .
we next illustrate the power and generality of our approach , combining a sequence of olap operations , and expressing them as a sequence of olap transformations .
[ ex : rollup-3 ] an olap user is analyzing sales in different countries and regions .
she wants to compare sales in the north of belgium ( the flanders region ) , and in the south of france ( which we , generically , have denoted _ south _ in our running example ) .
she first filters the cube , keeping just the cells of those two regions .
this is done with the expression : @xmath399 we showed that this can be implemented as a sequence of atomic olap transformations .
now she has a cube with the cells that have not been destroyed .
next , within the same navigation process , she obtains the total sales in france and belgium , only considering the desired regions , by means of : @xmath400 this will only consider the valid cells for rolling up .
after this , our user only wants to keep the sales in france .
thus , she writes : @xmath401 finally , she wants to go back to the details , one level below in the hierarchy , so she writes : @xmath402 implemented as a roll - up from the bottom level to _ region _ , only considering the cells that have not been destroyed .
we have presented a formal , mathematical approach , to solve a practical problem , which is , to provide a formal semantics to a collection of the olap operations most frequently used in real - world practice .
although olap is a very popular field in data analytics , this is the first time a formalization like this is given .
the need for this formalization is clear : in a world being flooded by data of different kinds , users must be provided with tools allowing them to have an abstract `` cube view '' and cube manipulation capabilities , regardless of the underlying data types . without a solid basis and unambiguous definition of cube operations ,
the former could not be achieved .
we claim that our work is the first one of this kind , and will serve as a basis to build more robust practical tools to address the forthcoming challenges in this field .
we have addressed the four core olap operations : slice , dice , roll - up , and drill - down .
this does not harm the value of the work . on the contrary , this approach allows us to focus on our main interest , that is , to study the formal basis of the problem .
our line of work can be extended to address other kinds of olap queries , like queries involving more complex aggregate functions like moving averages , rankings , and the like .
further , cube combination operations , like drill - across , must be included in the picture .
we believe that our contribution provides a solid basis upon which , a complete olap theory can be built .
* acknowledgements : * alejandro vaisman was supported by a travel grant from hasselt university ( korte verblijven inkomende mobiliteit , bof15kv13 ) .
he was also partially supported by pict-2014 project 0787 .
r. agrawal , a. gupta , and s. sarawagi .
modeling multidimensional databases . in _ proceedings of the 15th international conference on data engineering , ( icde ) _ , pages 232243 ,
birmingham , uk , 1997 .
ieee computer society .
f. dehne , q. kong , a. rau - chaplin , h. zaboli , and r. zhou .
scalable real - time olap on cloud architectures .
, 7980:31 41 , 2015 .
special issue on scalable systems for big data management and analytics . o. romero and a. abell . on the need of a reference algebra for olap . in _ proceedings of the 9th international conference on data warehousing and knowledge discovery , dawak07 _ , pages 99110 , regensburg , germany , 2007 . | online analytical processing ( olap ) comprises tools and algorithms that allow querying multidimensional databases .
it is based on the multidimensional model , where data can be seen as a cube , where each cell contains one or more measures can be aggregated along dimensions . despite the extensive corpus of work in the field , a standard language for olap
is still needed , since there is no well - defined , accepted semantics , for many of the usual olap operations . in this paper , we address this problem , and present a set of operations for manipulating a data cube .
we clearly define the semantics of these operations , and prove that they can be composed , yielding a language powerful enough to express complex olap queries .
we express these operations as a sequence of atomic transformations over a fixed multidimensional matrix , whose cells contain a sequence of measures .
each atomic transformation produces a new measure .
when a sequence of transformations defines an olap operation , a flag is produced indicating which cells must be considered as input for the next operation . in this way , an elegant algebra is defined .
our main contribution , with respect to other similar efforts in the field is that , for the first time , a formal proof of the correctness of the operations is given , thus providing a clear semantics for them .
we believe the present work will serve as a basis to build more solid practical tools for data analysis .
* keywords * : olap ; data warehousing ; algebra ; data cube ; dimension hierarchy . |
fluid turbulence is an archetypal phenomenon belonging to out - of - equilibrium and non - linear classical physics . starting probably with the work of reynolds , the complex and multiscale nature of turbulent velocity fluctuations
is usually apprehended in a statistical way . in this spirit ,
@xcite proposed in his seminal article a dimensional based argument explaining the spatial two - point correlation structure of velocity fluctuations , i.e. the 2/3-law ( see for instance classical textbooks * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , as it was observed in early experimental measurements of laboratory flows .
furthermore , he derived rigorously from the navier - stokes equations , using the stationary solution of the von krmn and howarth equation , the behaviour at infinite reynolds number and vanishing scale of the third order moment of velocity increments , known as the 4/5-law @xcite , reminiscent of a non vanishing mean energy transfer across scales .
this gives solid ground to the following phenomenology of three dimensional homogeneous , isotropic and incompressible turbulence : energy is injected at large scale @xmath1 , the so - called integral length scale ( typically the mesh size of a grid generated turbulence in a wind tunnel , or the typical scale of propellers , etc . ) , that is then transferred to smaller scales via a direct cascading process , until it is dissipated by viscosity . in this context , a more complete picture could be given while proposing a _
stochastic _ representation of a velocity field able to reproduce in probability law the formerly described spatial structure of turbulence .
in other words , we ask whether it is possible to build a random vector field @xmath2 , incompressible , statistically homogeneous and isotropic , seen as a statistically stationary solution of the navier - stokes equations , that reproduces in particular the observed 2/3 and 4/5-law
. the very first idea would be to consider a gaussian approximation .
this was first considered by kolmogorov himself , and the respective process belongs to the more general class of fractional brownian motions @xcite .
unfortunately , such a gaussian model fails to reproduce the observed and derived mean energy transfer encoded in the third - order moment of the velocity increments , as previously mentioned .
however , an underlying gaussian velocity field is an appealing starting point , and we will see in the following how to modify it in order to obtain a more realistic picture that includes energy transfer .
hereafter , we consider homogeneous , isotropic and incompressible velocity fields in three dimensional space , i.e. @xmath3 and @xmath4 , with @xmath5 a regularizing scale that plays the role , in a schematic way , of the kolmogorov dissipative length scale . in a gaussian framework ,
let us call @xmath6 such a field . to fully determine this gaussian velocity field
, we have to prescribe its velocity components covariance , that will in particular take into account the self - similar law of kolmogorov ( i.e. the @xmath7-law ) .
homogeneity , isotropy , incompressibility and the self - similar property can be modelled as a stochastic integral @xcite in the following way @xmath8 where @xmath9 is a zero - average vector gaussian white noise whose components are independent with variance equal to the infinitesimal volume @xmath10 , and @xmath11 is a large - scale cut - off of characteristic extension @xmath1 ( i.e. the integral length scale ) ensuring a finite variance of this random velocity field .
it is chosen radially symmetric to ensure isotropy , that is , for any vector @xmath12 , @xmath13 .
the singular kernel @xmath14 is regularized over the scale @xmath15 such that @xmath16 is proportional to @xmath15 when @xmath17 ( see section [ sec : numproc ] and @xcite for further details ) .
this gives differentiability to the vector field @xmath18 for any @xmath19 .
the vector product @xmath20 entering equation ( [ eq : gaussianfield ] ) recalls the structure of the biot - savart law and ensures incompressibility ( i.e. a divergence - free vector field ) .
using standard rescaling techniques ( as done in @xcite , using similar techniques from @xcite ) , it can be shown that the limiting process @xmath21 , which corresponds in a turbulent context to the limit of vanishing viscosity , is a finite variance random vector field when the so - called hurst ( or hlder ) exponent @xmath22 is strictly positive , i.e. @xmath23 .
let us introduce the _ velocity increments _ in order to make a connection with turbulence phenomenology and comment on the free parameter @xmath22 entering the definition of the field from equation ( [ eq : gaussianfield ] ) . as is usually done in the turbulence literature
( see for instance classical textbooks such as * ? ? ?
* ; * ? ? ?
* ) , it is convenient to define the longitudinal @xmath24 and transverse @xmath25 velocity increments . note @xmath26 and @xmath27 , the projections of the vector @xmath28 onto the direction of @xmath29 and , correspondingly , onto any perpendicular direction .
the longitudinal and transverse velocity increments are given by @xmath30 it can be shown , for @xmath31 , that the limiting gaussian random field @xmath32 , as we will recall in this article , is scale - invariant in the sense that moments of velocity increments , the so - called structure functions , behave as power - laws , i.e. for @xmath33 , @xmath34 \build{\sim}_{\ell\rightarrow 0}^ { } c_{2q}^{g,\parallel}\ell^{2qh } \ , \mbox { and } \ , \mathbb e \left[(\delta_\ell^{\perp } u^g)^{2q}\right ] \build{\sim}_{\ell\rightarrow 0}^ { } c^{g,\perp}_{2q}\ell^{2qh},\ ] ] with @xmath35 and @xmath36 two strictly positive constants that are universal in the sense that they depend on the order @xmath37 and only on the value of the cut - off function at the origin @xmath38 ( and not on its entire shape ) .
based on dimensional arguments , kolmogorov phenomenology predicts @xmath39 , that is @xmath40 , justifying the terminology of the 2/3-law .
we gather the proofs in appendix [ ann : gf ] . roughly speaking
, the scale - invariance property comes from the singular power - law shape of the kernel entering the definition of the gaussian velocity field in equation [ eq : gaussianfield ] .
we can see that , as far as kolmogorov s 2/3-law is concerned , we can give a clear meaning ( equation [ eq : gaussianfield ] ) to a stochastic representation of a turbulent homogeneous , isotropic and incompressible velocity field .
unfortunately , such a representation is too naive to reproduce the 4/5-law , that requires a non vanishing ( strictly negative ) third order moment of velocity increments , whereas the gaussian random field ( [ eq : gaussianfield ] ) is such that @xmath41 = \mathbb e \left[(\delta_\ell^{\perp } u^g)^{3}\right ] = 0.\ ] ] as noted , the gaussian velocity field ( [ eq : gaussianfield ] ) fails to give a realistic picture of k41 phenomenology since the gaussian structure leads to vanishing odd - order correlators and thus , to a vanishing mean energy transfer across scales .
furthermore , higher - order even correlators are themselves poorly predicted .
this intrinsic non - gaussianity of the small scales was observed in early experimental measurements @xcite .
this was taken into account by @xcite and @xcite ( hereafter referred to as ko62 ) while refining the k41 theory and setting a very peculiar statistical structure of the dissipation field .
this is the so - called intermittency , i.e. multifractal phenomenon ( see @xcite for a review on the subject ) .
various approaches were developed in the past to provide a stochastic representation of such a dissipation field , starting probably with the _ discrete _ cascades models initiated by the russian school ( @xcite , see also @xcite ) .
we will prefer here to follow a continuous version of these discrete cascade models that ensures homogeneity , known as the limit - lognormal model of @xcite , rigorously studied in the framework of _ gaussian multiplicative chaos _ by @xcite ( see @xcite for recent developments on this matter ) .
the aim is to model the dissipation field as a lognormal process with a long range correlation structure of the fluctuations , as observed in experiments @xcite .
gaussian multiplicative chaos consists in defining such a scalar lognormal process as the exponential of a gaussian field @xmath42 with logarithmic covariance , i.e. @xmath43\sim \ln ( l/|{\boldsymbol{x}}-{\boldsymbol{y}}|)$ ] , with @xmath1 being the integral length scale .
it is then possible to give a clear meaning to the scalar field @xmath44 @xcite , where @xmath45 is a dimensionless free parameter of the theory . in particular , in three dimensional space , it is then possible to show that the local average of this scalar field @xmath46 over a ball of size @xmath29 is a well - posed random field whose moments are scale invariant in the sense that @xmath47 behaves as @xmath48 , with @xmath49 a non - linear ( quadratic ) function of the order @xmath37 .
as we said , such a construction is defined up to a dimensionless free parameter @xmath45 known as the intermittency coefficient that can be precisely estimated on experimental signals ( see for instance @xcite and references therein ) . until now
, the statistical properties of turbulence that we have mentioned concern mainly the fluctuations of the longitudinal velocity profile which is accessible with traditional experimental techniques , hot - wire anemometry in particular , and do not characterize the vector nature of the velocity field .
for example , at this stage , nothing is said about the peculiar correlation structure of the components of the velocity gradients tensor @xmath50 .
we are thus asking if , furthermore , it is possible to build up a differentiable velocity field ( at a finite @xmath15 ) consistent with two important properties of the velocity gradient tensor that are ( i ) the teardrop shape of the joint density of the invariants @xmath51 and @xmath52 @xcite and ( ii ) the preferential alignment of the vorticity vector @xmath53 with the eigenframe of the rate - of - strain matrix .
going beyond the gaussian approximation ( [ eq : gaussianfield ] ) is a difficult matter since the mathematical theory of non - gaussian processes is far more sophisticated . in this direction ,
some recent attempts by @xcite and @xcite are interesting but it is not clear whether these vector fields exhibit energy transfer .
let us also mention the iterative procedure of @xcite that gives a realistic picture but which is not explicit , making analytical results out of reach at the present time .
in a one - dimensional context , several models have been proposed in the literature in order to apply the discrete cascade models to reproduce synthetically the observed fluctuations of longitudinal velocity profiles , including a model for energy transfer @xcite with additional parameters and propositions to extend to spatio - temporal @xcite and lvy - based @xcite stochastic representations . in a different spirit , @xcite propose to reconstruct velocity time series , starting from a time series at a given ( small ) scale and assuming a markov property in scale . as far as we know , @xcite are the first to have proposed a compressible velocity field with non - symmetrical probability laws . to generalize their approach to incompressible velocity fields
, they propose to modify the gaussian field ( [ eq : gaussianfield ] ) in order to include energy transfer and intermittency effects .
this was done while disturbing the vector white noise field @xmath54 by the scalar multifractal measure given by the multiplicative chaos .
unfortunately , for symmetry reasons , @xcite show that this incompressible intermittent velocity field has a vanishing mean energy transfer .
it is tempting to think that the present picture is too heuristic to represent the complex local structure of turbulence .
a further step in this direction was proposed by @xcite in which the euler equations , and more precisely the vorticity stretching mechanism , is used in order to motivate the exponentiation of a homogeneous field of isotropic symmetric trace - free gaussian matrices , that eventually lead to energy transfer .
let us recall how to include some aspects of the vorticity stretching mechanism in the present picture .
the euler equations reads , in the vorticity formulation , @xmath55 where @xmath2 is the velocity field , solution of the euler equation and given by the biot - savart law , i.e. @xmath56 and the deformation field @xmath57 is defined as the symmetric part of the velocity gradient tensor , namely @xmath58 , where @xmath59 stands for matrix transpose . in incompressible flows , the deformation field
is fully determined by the vorticity field and the explicit form reads @xcite @xmath60}{|{\boldsymbol{x}}-{\boldsymbol{z}}|^5 } + \frac { [ ( { \boldsymbol{x}}-{\boldsymbol{z}})\wedge { \boldsymbol{\omega}}({\boldsymbol{z}})]\otimes({\boldsymbol{x}}-{\boldsymbol{z}})}{|{\boldsymbol{x}}-{\boldsymbol{z}}|^5}\right]{\mathrm{d}}z\mbox { , } \ ] ] where the integral is understood as a cauchy principal value ( p.v . ) and @xmath61 is the tensor product , i.e. @xmath62 . the first underlying idea of @xcite is to study the implication of a linearization of the previous formulation of the euler equation on the velocity field generated by the stretching of an initial gaussian vorticity field ( with a k41 structure ) by the initial deformation field .
this first motivates the use of the exponentiation of a gaussian random field of symmetric matrices , although it was not expected from this short time study of the euler equations to reproduce the peculiar intermittent nature of the velocity field .
this structure was introduced by hand using the gaussian multiplicative chaos that is naturally obtained while modifying the integration kernel of the deformation field ( equation [ eq : deformationpv ] ) .
this heuristic procedure , motivated by the short time dynamics of the euler equations , leads to the following proposition of a velocity field representing a realistic local structure of turbulence : @xmath63 where @xmath64 is an isotropic trace - free symmetric random matrix , whose structure recalls the one of the deformation field ( [ eq : deformationpv ] ) , given explicitly by a tensor wiener integral that we will specify later .
the non dimensional constant @xmath0 governs the level of intermittency .
let us finally remark that a crucial step of this construction , as dictated by the short - time dynamics of the euler equations , is the intrinsic dependence of this statistically isotropic matrix @xmath65 on the vector white noise @xmath54 .
we can see , given a hurst exponent that we will take to be @xmath39 to be consistent with k41 phenomenology , that the proposed stochastic model ( [ eq : multifractalfield ] ) does depend on a single free parameter @xmath0 that can be determined empirically .
therefore , if this vector field is to provide a realistic picture of the local structure of turbulence , this unique free parameter should govern at the same time both the intermittency phenomenon and the physics of the energy transfer , which is , as far as we know , a new type of relationship between these phenomena .
we will indeed derive from a perturbative approach ( section [ sec : enertrans ] ) that the third order moment of velocity increments is proportional to the scale , with a multiplicative factor that is itself proportional to this free parameter @xmath0 .
the purpose of this article is to go beyond the results obtained by @xcite in which the field ( [ eq : multifractalfield ] ) has been proposed for the first time and studied mostly numerically for a single value of the intermittency coefficient @xmath0 representing in a satisfactory manner the statistical properties of turbulence . as we will see in the following quick description of the various sections of the article ,
the proposed new material include ( i ) an extensive numerical study of the statistical properties of the velocity field at the smallest @xmath15 resolutions we were able to reach , for several values of the free parameter @xmath0 , ( ii ) an analytical derivation of the spectrum of exponents of the structure functions in the asymptotic limit of vanishing resolutions @xmath66 of a simplified ersatz of the field named @xmath67 and ( iii ) a perturbative approach for small @xmath0 able to capture some aspects of the energy transfer taking place while reconsidering the field @xmath68 ( equation [ eq : multifractalfield ] ) . in section [ sec : notbas ] , we set our notation and define the field of random matrices @xmath69 . in section [ sec : numproc ] , we describe the numerical procedure in order to obtain realizations of the velocity field ( [ eq : multifractalfield ] ) . in short ,
@xmath28 is simulated in a periodic box of size @xmath70 .
we rely then on the discrete fourier transform to perform the convolutions .
the matrix exponential is evaluated at each point of space using a pad approximant with scaling and squaring .
the fast fourier transform ( fft ) algorithm is used in its fully parallel form .
we study then the numerical properties of the velocity field based on realizations up to @xmath71 collocation points . in section [ sec : numa ]
, we use these numerical simulations to compute the joint density of the invariants @xmath72 and @xmath73 at various intermittency coefficients @xmath0 and discuss their comparison with what is obtained in laboratory and numerical flows .
similarly , we show the preferential alignment of vorticity with the intermediate eigendirection of the eigenframe of the deformation , and quantify its dependence on @xmath0 .
section [ sec : intermittency ] is devoted to a joint numerical and analytical study of the intermittency phenomenon observed in the velocity field ( [ eq : multifractalfield ] ) .
we will indeed observe that this field is intermittent ( in a sense that we will make precise in the devoted section ) , and its level of intermittency is given in terms of the coefficient @xmath0 .
a rigorous derivation of the behaviour of the structure functions of the velocity field is mathematically very demanding , and even obtaining the variance of the components is a difficult task .
the reason is related to the strong correlation between the exponentiated gaussian field of matrices @xmath65 and the underlying vector white noise @xmath54 . to obtain analytical results , we study an _ ersatz _ , which has the same structure as the proposed field ( [ eq : multifractalfield ] ) but assuming the independence of the matrix @xmath65 and vector @xmath54 fields .
we will call this case the _ independent _ case and note the respective velocity field @xmath74
. we will show in section [ sec : intermittency ] that indeed , when properly renormalized , the velocity field @xmath74 converges towards a finite - variance process @xmath67 when @xmath75 and we will compute its respective structure functions , obtaining @xmath76 with @xmath77 and @xmath78 two strictly positive constants . note that , asymptotically , higher order longitudinal and transverse structure functions share similar scaling behaviour . note also that we do not obtain perfect power - laws since an additional logarithmic factor appears in the asymptotic behaviour .
this factor is related to the matrix nature of the chaos and was already observed in @xcite .
this former scale dependence of structure functions is based on an exact calculation for @xmath79 , and has been extended to higher orders @xmath80 based on a conjecture proposed in @xcite .
thus , this field @xmath81 allows us to understand the intermittent corrections to the scaling behaviour with respect to the gaussian case @xmath32 .
the independence assumption leads on the other side to vanishing third- and more generally odd - orders structure functions , namely @xmath82 missing all the physics of energy transfer and showing that the intrinsic correlation between the matrix @xmath65 and vector @xmath54 fields in the velocity field @xmath83 ( equation [ eq : multifractalfield ] ) is crucial to reproduce non vanishing third order moment .
nonetheless , we show numerically that , to fourth order , the ersatz @xmath74 and the full velocity field ( [ eq : multifractalfield ] ) share similar intermittent properties .
furthermore , an analytical study that takes into account finite - scale corrections is performed in order to interpret with high precision the numerical results . from the behaviour of the velocity increment flatness that we will define later on ( equation [ eq : flatnesses ] ) , taking into account non trivial finite - scale corrections , we are led to propose the very particular value @xmath84 for turbulent applications in order to be consistent with experimental measurements and numerical simulations . section [ sec : enertrans ] is devoted to the physics of energy transfer . as we said ,
a rigorous study of the statistical properties of the proposed velocity field ( [ eq : multifractalfield ] ) is a difficult task . in order to discuss the important physics of the energy transfer , as required by the @xmath85-law @xcite
, we will rely on a perturbative analysis of this field , at a finite resolution @xmath19 , using the intermittency parameter @xmath0 as the small parameter , which is indeed the case as far as turbulence is concerned .
we show that such a perturbative expansion of the longitudinal third order velocity structure function is given by @xmath86 where @xmath87 stands for a term that depends on @xmath15 but depends on a higher power of @xmath0 than 1 ( typically this term is of order @xmath88 by symmetry ) .
we are then able to show that the dominating term linked to @xmath89 converges when @xmath66 towards a non trivial function @xmath90 which is such that @xmath91 with @xmath92 a constant .
numerical simulations show indeed such a linear behaviour , with @xmath93 , of the third order structure function with both the intermittency coefficient @xmath0 and the scale @xmath29 when @xmath0 is small and when we use the hurst of k41 , namely @xmath39 .
this shows , up to first order in @xmath0 , that the proposed velocity field ( [ eq : multifractalfield ] ) exhibits energy transfer according to kolmogorov phenomenology .
we gather in section [ sec : conclusion ] our conclusion and perspectives .
in what follows , @xmath94 will denote the kronecker delta and @xmath95 the levi - civita symbol .
we adopt einstein s convention of sum over repeated indices , unless explicitly stated , and we note that @xmath96 .
the full vector field ( [ eq : multifractalfield ] ) reads , with index notation , @xmath97 where the kernel @xmath98 encodes the structure of the underlying gaussian velocity field ( [ eq : gaussianfield ] ) and is given by @xmath99 and the following matrix field built from the very same vector white noise @xmath54 that enters the construction of the underlying structure : @xmath100 + [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{w}}({\mathrm{d}}y ) ] \otimes \frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon},\ ] ] which is inspired by the tensor structure of the rate - of - strain matrix @xmath101 ( equation [ eq : deformationpv ] ) that stretches the vorticity vector along its path .
we will motivate the use of the multiplicative factor @xmath102 when we give the variance and covariance of the elements of the matrix @xmath65 . at this stage ,
remark that this matrix is gaussian since it is defined through a linear operation on a gaussian measure @xmath54 ( equation [ eq : xepsilon ] ) .
it is indeed symmetric , and it is easy to check that it is trace free , according to @xmath103=0.\ ] ] the free parameter @xmath0 entering the velocity field @xmath28 ( equation [ eq : multifractalfieldindices ] ) plays the same role as the free parameter @xmath104 used in the field given by equation 12 of @xcite , and their relation is @xmath105 .
the gaussian white vector field @xmath106 , for @xmath107 , follows the following rules of calculation . for
any suitable deterministic function @xmath108 , @xmath109 such that it is integrable along its diagonal , we have @xmath110= 0,\ ] ] and @xmath111= \delta_{ij } \int f({\boldsymbol{x}},{\boldsymbol{x}}){\mathrm{d}}x.\ ] ] let us first show that the matrix field @xmath65 ( [ eq : xepsilon ] ) is indeed homogeneous and isotropic .
homogeneity of the field @xmath112 follows from the convolution with the homogeneous white measure @xmath54 .
consider now a rotation matrix @xmath113 such that @xmath114 , where @xmath115 the @xmath116 identity matrix .
then , it can be shown that for any rotation matrix @xmath117 , we have @xmath118 .
the equality in law @xmath119 stands for equality in probability .
this shows that , in that sense , the matrix field is statistically isotropic . as an important further characterization of the homogeneous field of matrices @xmath65
, we want to obtain its covariance structure , component by component .
let us first remark that all of the elements @xmath120 are of zero mean , which follows from the definition of the field as a convolution with a zero - mean white noise @xmath121=0 $ ] .
we gather all the proofs of the following results in annex [ ann : corrx ] .
we have seen that the field of matrices @xmath65 is statistically isotropic . recall that each element is a gaussian random variable , and @xmath65 is a symmetric matrix .
thus , the covariance structure of its elements is given by the general framework developed in @xcite .
we recall several key properties of this random matrix .
the first property of the elements of @xmath65 is the divergence of their variance with the regularizing parameter @xmath15 . henceforth , we focus only on the element @xmath122 of the matrix .
see annex [ ann : corrx ] for a general discussion on the statistical behaviour of the other elements
. defining the variance of this element as @xmath123 , then it is easy to obtain its asymptotic behaviour when @xmath66 as @xmath124 \build{\sim}_{\epsilon\rightarrow 0}^ { } \ln\frac{l}{\epsilon}\mbox { .}\ ] ] thus , the variance of the elements of @xmath65 diverge logarithmically with @xmath15 .
this situation is classically encountered in the context of multiplicative chaos ( see a review on this topic by @xcite ) .
similarly , the covariance of the element @xmath125 can be computed and we find , taking first the limit @xmath126 and then looking for an equivalent at small distances , @xmath127 \build{\sim}_{|{\boldsymbol{x}}-{\boldsymbol{y}}|\rightarrow 0}^{}\ln\frac{l}{|{\boldsymbol{x}}-{\boldsymbol{y}}| } \mbox { .}\ ] ] in other words , the gaussian random variable @xmath122 converges when @xmath66 towards a random gaussian distribution whose covariance behaves logarithmically at small distances .
we remark that the factor @xmath128 entering the definition of @xmath65 ( [ eq : xepsilon ] ) ensures a unit - factor in front of the logarithmic behaviours seen in equations ( [ eq : asympsigmaeps ] ) and ( [ eq : asympsigmacorr ] ) .
the very purpose of the theory of multiplicative chaos @xcite is to give a meaning to the exponential of such a field .
let us now show that the velocity field ( [ eq : multifractalfieldindices ] ) , defined with the field of matrices @xmath129 ( [ eq : xepsilon ] ) , is indeed homogeneous and isotropic .
again , homogeneity of the vector field @xmath28 follows from the convolution with the homogeneous field @xmath130 .
consider again a rotation matrix @xmath113 .
then , for any rotation matrix @xmath117 , we have @xmath131 .
thus , the velocity field is statistically isotropic . as a consequence ,
the velocity field is of zero - mean , i.e. , for any @xmath19 @xmath132
as we will see in the following , a rigorous derivation of the statistical properties of the velocity field @xmath28 is a difficult matter .
when analytical results are not possible , we will rely on numerical simulations . to do so
, we perform a numerical approximation of @xmath28 in the periodic domain @xmath133 ^ 3 $ ] .
define @xmath134 as the number of collocation points in one direction .
we will typically present results for @xmath135 , using a fully parallelized algorithm of the fast fourier transform @xcite .
the elementary volume is given by @xmath136 .
standard algorithms allow to generate @xmath137 independent realizations of a zero - mean gaussian variable of variance @xmath138 in order to define the vector white noise @xmath139 .
the elements of the matrix exponential entering the construction are calculated with the expokit tool @xcite using the irreducible rational pad approximant .
the remaining convolutions are performed in the fourier space .
we choose as an isotropic cut - off function the following @xmath140 , compactly supported function @xmath141 and we will consider the particular value @xmath142 in order to obtain a couple of integral scales in our simulations .
the precise shape of this function is not important , besides its characteristic length scale @xmath1 , and only the large scale statistical quantities such as the variance depend on it .
we will see that at small scales , explored as an example by velocity increments , only the value at the origin @xmath38 matters . as a regularization mechanism
, we use the following regularized norm latexmath:[\[\label{eq : regulnorm } this regularization procedure makes the continuous field @xmath28 ( [ eq : multifractalfieldindices ] ) differentiable , and divergence free in particular . in the discrete approximation , we can not choose @xmath15 arbitrarily small , since it is bounded from below by the finiteness of the smallest accessible scale @xmath144 . as mentioned
, @xmath15 plays the role , in a schematic way , of the kolmogorov scale , and therefore should depend on the reynolds number .
as far as the gaussian field is concerned ( equation [ eq : gaussianfield ] ) , we can relate @xmath15 to the kinematic viscosity @xmath145 and the hurst exponent @xmath22 in such a way that the average dissipation per unit of mass remains finite and strictly positive when @xmath146 @xcite . in the following
, we will work numerically at a finite viscosity , and we will be interested theoretically in the asymptotic limit @xmath66 , which corresponds to the infinite reynolds number limit .
we thus have to take @xmath15 greater than @xmath144 . when @xmath147 , then the numerical field is smooth and gradients are well approximated . in particular , in standard deviation ,
the divergence of the field @xmath148 is much smaller than the gradient of one of its components . when @xmath149 , the numerical field is rough , and gradients are poorly approximated . in other words ,
the divergence of the field can be of the order of the gradient of one of its components ( in standard deviation ) .
we are also interested in simulations where the _ inertial range _ is wide , i.e. we would like to maximize the ratio @xmath150 . in section [ sec : numa ] , since we will focus on velocity gradients , we will use @xmath151 . in the following sections , we will use @xmath152 . once again
, the precise regularization procedure is not important as long as @xmath16 is of order @xmath15 at the origin , and equal to @xmath153 at a distance @xmath154 from the origin .
we can rigorously show that this is the case for the matrix multiplicative chaos @xcite .
as an important characterization of the velocity gradient tensor @xmath155 , we study its two non vanishing invariants . for instance , the second invariant @xmath72 is given by @xmath156 where @xmath157 is the vorticity vector and @xmath101 the symmetric part of @xmath158 , and can be interpreted as the competition between enstrophy and dissipation ( per unit viscosity ) . then , positive @xmath72 represents rotation - dominated regions and negative @xmath72 dissipation - dominated regions .
analogously , the third invariant @xmath73 is given by @xmath159 representing competition between _ enstrophy production _ and
_ dissipation production_. see @xcite for a discussion on this topic .
we simulate the vector field @xmath28 for four different values of intermittency coefficients @xmath0 , with @xmath142 , @xmath135 and @xmath151 ( see discussion in section [ sec : numproc ] ) and represent the numerical estimation of the joint density of the invariants @xmath72 and @xmath73 in figure [ fig : rqs ] . as is well known , a gaussian velocity field corresponding to @xmath6 ( [ eq : gaussianfield ] ) , or equivalently the velocity field @xmath28 ( [ eq : multifractalfield ] ) with @xmath160 , predicts a joint density of the invariants symmetrical with respect to the @xmath161 line .
this is what we obtain in figure [ fig : rqs](a ) . in figures [ fig :
rqs](b , c , d ) , we study the effect of increasing @xmath0 .
we can see that the bigger the value of @xmath0 , the more elongated is the joint density along the right tail of the zero discriminant ( or vieillefosse ) line , where @xmath162 @xcite .
we will see that increasing @xmath0 corresponds to increasing the level of intermittency . as justified in section [ sec : intermittency ] , we choose the very particular value @xmath84 for turbulence applications whose corresponding joint density of @xmath73 and @xmath72 is displayed in figure [ fig : rqs](c ) .
another striking property of turbulence is the preferential alignment of vorticity with the strain eigendirection associated to the intermediate eigenvalue .
we refer again to @xcite for further discussions .
we represent in figure [ fig : aligns ] the probability density of the cosine of the angle @xmath163 between vorticity and the eigenvectors of the strain .
figure [ fig : aligns](b ) indicates the preferential alignment of vorticity with the correct eigenvector , as observed already in @xcite for a single value of @xmath84 . here , we can see that this alignment is governed by the intermittency coefficient @xmath0 : no preferential alignment is observed when @xmath160 , as expected from a gaussian velocity field , and this preferential alignment increases with increasing @xmath0 .
we observe also in figure [ fig : aligns](a ) that the density of the preferential orthogonality of vorticity with the eigendirection associated to the smallest ( negative ) eigenvalue is barely sensitive to @xmath0 , except in the gaussian case @xmath160 .
as for the angle between vorticity and the eigendirection associated to the biggest ( positive ) eigenvalue ( figure [ fig : aligns](c ) ) , as observed in real flows , the density is almost flat , with a slight dependence on the parameter @xmath0 , showing no preferential orientation .
overall , as far as velocity gradient statistics are concerned , the velocity field @xmath28 predicts non trivial facts of fluid turbulence . in this picture , at least for the range of @xmath0 values studied , the dependence on this parameter is weak .
we will study in the following section the influence of the parameter @xmath0 on the scaling of structure functions , where it will play a key role .
as we have seen in section [ sec : numa ] , at a finite @xmath15 ( or finite reynolds number ) , the velocity field @xmath28 ( [ eq : multifractalfield ] ) predicts realistic velocity gradient statistics . in particular , at any @xmath164 , we reproduce the teardrop shape of the @xmath165-plane ( figure [ fig : rqs ] ) and the preferential alignment of vorticity ( figure [ fig : aligns ] ) .
the precise value of @xmath0 representing realistic turbulent statistics has not been selected yet .
this is the purpose of this section .
we focus now on the intermittency phenomenon , and explore the influence of the parameter @xmath0 on the anomalous scaling of the structure functions @xcite .
to do so , we will discuss this phenomenon based on the flatness of velocity increments ( defined in equation [ eq : defvis ] ) , namely @xmath166 ^ 2},\ ] ] in both the longitudinal ( i.e. @xmath167 ) and transverse cases ( @xmath168 ) .
we perform first simulations of the field for @xmath135 , @xmath142 and several @xmath0 .
we choose @xmath152 in order to maximize the extend of the inertial range , and represent the results of the estimation of the flatness in figure [ fig : flatness ] .
we display in figure [ fig : flatness](a ) ( respectively figure [ fig : flatness]c ) the flatness of the longitudinal ( transverse ) velocity increments as a function of the scale @xmath29 for selected values of the parameter @xmath169 , including the gaussian case @xmath170 .
we indeed notice that in the gaussian case the flatnesses @xmath171 and @xmath172 do not depend on scale and equal 3 .
then , as @xmath0 increases , for any fixed scale , the flatness increases .
we observe a power law with the scale within a limited range , i.e. @xmath173 the regularization over the scale @xmath15 polluting a large part of the accessible scales .
we represent in figure [ fig : flatness_fit]@xmath174 the obtained value for @xmath175 as a function of @xmath169 .
we indeed see that the scaling exponents @xmath175 decrease when @xmath0 increases , showing the augmentation on the level of intermittency @xcite .
we observe also , in this context , that the level of intermittency of transverse velocity increments is higher than the one observed on longitudinal velocity increments . in laboratory and numerical flows
( see for instance * ? ? ?
* ) , we find a universal behaviour ( independent of the flow geometry and the reynolds number ) for the longitudinal case , with @xmath176 .
thus , turbulence statistics seem to be well reproduced for a very particular small value of the parameter @xmath177 , which was already found in @xcite .
the present velocity field ( [ eq : multifractalfield ] ) is an example of random process that exhibits a higher level of intermittency for the transverse case than for the longitudinal case .
this is indeed a surprising effect , also observed in real flows ( see discussions in @xcite ) . in our model
, only an analytical study could give a clear answer to this observed discrepancy , in particular in the asymptotic limit @xmath178 .
unfortunately , the underlying mathematical structure of this field is subtle , the strong correlation between the field @xmath65 and the white measure @xmath54 is difficult to handle . in the following , we will study both numerically and theoretically an _ ersatz _
@xmath74 of the velocity field @xmath83 ( [ eq : multifractalfield ] ) that follows the same rules of construction , except that the field of matrices @xmath65 is built independently of the underlying white measure @xmath54 .
this ersatz is amenable to exact derivations of its statistical properties .
consider now the following velocity field @xmath179 where again @xmath99 and the following _ matrix _ field : @xmath180 + [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{w}}'({\mathrm{d}}y ) ] \otimes \frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon}\ ] ] where now the vector noise @xmath181 is independent of the vector noise @xmath54 of the underlying gaussian structure ( [ eq : multifractalfieldindicesindpt ] ) , namely for any @xmath182 , and any components @xmath183 and @xmath184 , we have @xmath185=0 $ ] .
as we show in appendix [ ann : indeptcase ] , the field ( [ eq : multifractalfieldindicesindpt ] ) needs to be renormalized in order to converge , when @xmath126 , towards a finite - variance process .
this deterministic normalization constant @xmath186 ( which diverges when @xmath66 ) is given by @xmath187.\ ] ] this is a standard way to renormalize the multiplicative chaos and this situation is well understood as far as multiplicative chaos is concerned @xcite .
we can see that the difference between the full vector field ( [ eq : multifractalfield ] ) and the one just mentioned ( [ eq : multifractalfieldindicesindpt ] ) is that the matrix free field @xmath65 and the vector white noise @xmath54 are independent .
this is a strong simplification but we can get simple exact results . we perform similar simulations of the velocity field @xmath74 ( [ eq : multifractalfieldindicesindpt ] ) as we did for the ( dependent ) velocity field @xmath83 ( [ eq : multifractalfieldindices ] ) and compare the estimation of velocity increment flatnesses .
we display our results in figures [ fig : flatness](b , d ) and [ fig : flatness_fit](b ) . overall , at any value of the parameter @xmath0 , both velocity fields @xmath83 and @xmath188 share qualitatively similar levels of intermittency .
this gives confidence in explaining the intermittent properties of @xmath83 ( [ eq : multifractalfieldindices ] ) using results from the intermittent nature of the velocity field ersatz @xmath74 ( [ eq : multifractalfieldindicesindpt ] ) , which is amenable to analytical derivation .
the main difference , as far as intermittency is concerned , comes from the behaviour of the flatness of transverse velocity increments .
we observe indeed that in the independent case , the observed power - laws of the flatnesses @xmath171 and @xmath172 ( equation [ eq : flatnesses ] ) have the same scaling exponent @xmath189 ( [ eq : flatnessesexpo ] ) , a property that we show in the following sections .
also , as we will see , the velocity field ersatz @xmath74 does not exhibit energy transfer .
in other words , @xmath74 is a good ersatz to study the intermittent nature of the velocity field @xmath83 , but fails at describing the physics of energy transfer .
this reveals also the great importance of building up the matrix field @xmath65 from the very same white measure @xmath54 in order to predict energy transfer .
we will come back to this point in section [ sec : enertrans ] . since @xmath54 and @xmath65 are independent , we easily get ( see annex [ ann : indeptcase ] ) @xmath190 = 0.\end{aligned}\ ] ]
thus , the vector field ( [ eq : multifractalfieldindicesindpt ] ) has zero average , as expected from an isotropic vector field . as for the covariance , we find @xmath191 = ( \phi^{\epsilon}_{ik}\star\phi^{\epsilon}_{pk})({\boldsymbol{h}}),\end{aligned}\ ] ] where @xmath192 is the correlation product defined in equation ( [ eq : covgf ] ) .
we can see that the covariance structure of the vector field @xmath74 ( [ eq : multifractalfieldindicesindpt ] ) is the same as the one obtained from the underlying gaussian field @xmath6 ( [ eq : gaussianfield ] ) .
in particular , the field ( [ eq : multifractalfieldindicesindpt ] ) converges in a @xmath193 sense when @xmath66 , and it has same variance and covariance as the underlying gaussian field ( see annex [ ann : gf ] for the properties of the covariance of the underlying gaussian field ) .
we will note the corresponding limiting process as @xmath194 .
similarly , the gaussian field @xmath32 and the field @xmath67 share the same second order structure functions ( both longitudinal and transverse ) .
we define the velocity increment @xmath195 as @xmath196 where we have defined the even function @xmath197 of special interest is the fourth order structure function @xmath198 ( longitudinal and transverse cases ) , and the respective flatnesses , i.e. @xmath199 and @xmath200 ( [ eq : flatnesses ] ) .
we get ( no summation over repeated index @xmath201 implied ) , @xmath202{\mathrm{d}}z_{2}{\mathrm{d}}z_{4},\end{aligned}\ ] ] entering therefore the covariance of the matrix multiplicative chaos @xmath203 , which is analytically derived in @xcite . in annex [ ann :
indeptcase ] , we show that the fourth - order structure function converges when @xmath126 if we choose @xmath204 , and behaves asymptotically in the limit of vanishing scale @xmath205 as @xmath206 where the multiplicative constant @xmath207 is derived in annex [ ann : indeptcase ] .
this shows that longitudinal and transverse fourth order structure functions have the same scaling behaviours .
more precisely , we obtain @xmath208 where the different constants @xmath209 and @xmath210 are also given in annex [ ann : indeptcase ] .
this shows that the velocity field @xmath67 ( [ eq : multifractalfieldindicesindpt ] ) , built assuming independence of @xmath211 and @xmath54 , is intermittent , and the respective flatnesses ( [ eq : flatnesses ] ) behave as power - laws times a logarithmic correction with the scale @xmath29 ( see annex [ ann : indeptcase ] ) : @xmath212 thus , according to this asymptotic prediction ( [ eq : predindflat ] ) , the power - law exponents of the flatness are the same and related to the parameter @xmath0 , namely @xmath213 .
as we can observe in figure [ fig : flatness_fit ] , this asymptotic prediction performs poorly against our numerical results .
we will see in section [ sec : fssflat ] that this quantitative discrepancy can be explained while taking into account finite scale @xmath214 corrections , as it is necessarily done while fitting power - laws of flatnesses estimated in numerical simulations .
it is easy to see , and shown in annex [ ann : indeptcase ] , that all odd order structure functions vanish under the hypothesis of independence : the assumption of independence prevents us from studying the physics of energy transfer .
let us study now the even higher - order structure functions and let us consider the @xmath215-order moment of velocity increments @xmath216 , for @xmath217 , given in equation ( [ eq:2nindptsf ] ) . doing so , we are left considering the @xmath218-correlator of the matrix exponential of @xmath219 : @xmath220.\ ] ] for @xmath221 calculations are tedious , but , based on a conjecture of @xcite and an appropriate range of orders @xmath37 , we can write that the structure functions behave as @xmath222 with a similar spectrum of exponents for both longitudinal and transverse structure functions given by a quadratic function of the order @xmath37 , namely @xmath223 showing that indeed , the parameter @xmath0 fully determines the intermittent properties of the velocity field . as far
as flatnesses are concerned , we have found in our numerical simulations , for which results are displayed in figures [ fig : flatness](b , d ) , a power - law exponent @xmath224 , defined in ( [ eq : flatnessesexpo ] ) , much smaller ( in magnitude ) than our prediction @xmath225 ( [ eq : predindflat ] ) .
this is shown in figure [ fig : flatness_fit](b ) . in this section
we propose to explain this surprising fact while taking into account finite - size corrections based on our predictions before looking at the asymptotic limit @xmath226 .
we have seen while deriving the flatnesses of velocity increments ( section [ sec : derivflatnesses ] and annex [ ann : indeptcase ] ) that the covariance of the matrix chaos @xmath227 enters the expression of the fourth - order structure function , as shown in ( [ eq : fourthordersf ] ) .
the mathematical theory developed in @xcite allows us not only to derive its asymptotic structure in the double limits @xmath66 and then @xmath205 as we have already seen , but also in the limit @xmath126 at a finite scale @xmath29 .
recall that in the double limits , we have found a remaining logarithmic correction to the power - laws of the flatnesses ( see equation [ eq : predindflat ] ) .
the purpose of this section is to explore the behaviour of the matrix chaos covariance when the scale @xmath29 is finite , after taking the limit @xmath66 , that eventually leads to intermittency with logarithmic corrections at vanishing scales @xmath29 .
clearly , the dependence on the scale @xmath29 of the flatnesses is linked to this matrix chaos covariance . under the hypothesis of isotropy
, we can show that the matrix chaos covariance depends on only two scalar functions @xmath228 and @xmath229 defined as @xmath230 \build{\rightarrow}_{\epsilon\to 0}^ { } f(h)\delta_{k_1k_2}\delta_{k_3k_4 } + g(h)\left[\delta_{k_1k_3}\delta_{k_2k_4}+\delta_{k_1k_4}\delta_{k_2k_3 } \right].\ ] ] we give the expressions of @xmath231 and @xmath232 in ( [ eq : ffinitesize ] ) and ( [ eq : gfinitesize ] ) , that are intermediate steps before taking the limit @xmath178 . as we show in annex [ ann : indeptcase ] , the quantities @xmath231 and @xmath232 are responsible for the intermittent correction of the fourth order structure functions , the remaining kernels @xmath233 entering the expressions of the full structure functions ( [ eq : fourthordersf ] ) participate mainly to the scaling @xmath234 .
when renormalizing by the square of the second - order structure function , defining thus the respective flatnesses , we focus only on the intermittent corrections .
thus , we will approximate the flatness exponent @xmath224 by the logarithmic derivative of the contribution associated to @xmath231 and @xmath232 , to write @xmath235}{\partial \ln \ell}.\ ] ] as we have seen also , a key quantity that enters the expression of @xmath231 and @xmath232 ( equations [ eq : ffinitesize ] and [ eq : gfinitesize ] ) is the covariance of the diagonal elements of @xmath211 , i.e. @xmath236 ( equation [ eq : asympsigmacorralpha ] ) , and we will write it as @xmath237 where @xmath238 is a constant independent of the scale @xmath29 ( it is more generally a bounded function of the scale , but we will neglect this functional dependence ) . obviously , the constant @xmath238 is negligible in front of the logarithm ( [ eq : asympsigmacorralphascale ] ) when @xmath205 .
it is not the case when @xmath29 is finite .
we have estimated this constant in our numerical simulation and we find @xmath239 ( data not shown ) .
using equation ( [ eq : approxexpoflat ] ) , we evaluate the logarithmic derivative at the scale @xmath240 , using the form of the covariance given in ( [ eq : asympsigmacorralphascale ] ) with thus @xmath241 .
we display the result of this fit , as a function of @xmath0 , in figure [ fig : flatness_fit](b ) .
the comparison with numerical data is fairly good at low values of @xmath0 and deteriorates at higher values .
several remarks can be made at this stage to justify the level of adequacy of our fit with numerical data .
first , the model used to fit our data has been obtained in the limit of vanishing resolutions @xmath66 , whereas it can remain some finite-@xmath15 corrections when looking at a numerical simulation . secondly , the parameter @xmath238 rigorously should be seen as a bounded function of the scales @xmath29 .
we took it as a constant for the sake of simplicity .
thirdly , relation ( [ eq : approxexpoflat ] ) is only an approximation , and there could be additional finite scale corrections related to the underlying gaussian velocity field .
recall indeed , as shown in appendix [ ann : gf ] , that the gaussian velocity field exhibits exact power laws in the double limits @xmath66 and @xmath205 , whereas we focus here on finite scale corrections . on the theoretical side ,
these corrections are difficult to obtain and will depend on the precise shape of the large - scale cut - off function @xmath242 entering the definition of the velocity field ( [ eq : multifractalfieldindicesindpt ] ) . finally , let us note that the influence of the free parameter @xmath238 entering equation ( [ eq : asympsigmacorralphascale ] ) is only quantitatively substantial for higher values of @xmath0 . we explain in this way
the surprising fact that the estimated values of @xmath224 based on our simulations differ from the asymptotic prediction @xmath243 by taking into account finite - size corrections . under the independence assumption , we are thus able to quantify finite - size corrections to the scalings .
going back to the full vector field ( [ eq : multifractalfield ] ) , we observe in figure [ fig : flatness_fit](a ) that , likewise , longitudinal and transverse structure functions seem to be affected by finite size corrections .
we furthermore observe a slight difference between the longitudinal and transverse cases : according to our numerical simulations , it seems that transverse intermittency corrections are bigger than the longitudinal ones .
the underlying strong correlation between the chaos and the white noise prevents us from deriving analytically the asymptotic regime and thus , we can not conclude at this stage whether this slight difference will remain at vanishing resolution and vanishing scale .
let us now turn back to the full ( i.e. correlated ) vector field ( [ eq : multifractalfield ] ) that we recall here for convenience , @xmath244 where @xmath245 and the following field ( [ eq : xepsilon ] ) is built from the very same vector white noise @xmath54 that enters the definition of the velocity field ( [ eq : multifractalfieldcorr ] ) : @xmath246 + [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{w}}({\mathrm{d}}y ) ] \otimes \frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon}.\notag\ ] ] the same vector noise @xmath54 enters both the velocity field equation ( [ eq : multifractalfieldcorr ] ) and the matrix field implying peculiar correlations , fully given by the correlators @xmath247 , that read , for @xmath248 , @xmath249\notag \\ & = \sqrt{\frac{15}{32{\pi}}}\left[\frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon}\otimes [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{e_p}}]+ [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{e_p } } ] \otimes \frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon}\right ] \mbox { , } \end{aligned}\ ] ] where we have defined the unit vector @xmath250 with @xmath251 .
let us also remark that at a given finite @xmath19 , @xmath252 the components of the @xmath116 matrices @xmath253 are noted @xmath254 .
obtaining the exact statistical properties of the velocity field @xmath28 ( [ eq : multifractalfieldcorr ] ) is a difficult task . the very peculiar correlation between @xmath65 and @xmath54 , fully encoded in the correlator @xmath247 ( [ eq : gammap ] ) , and the non - commutative nature of the field of matrices
@xmath65 make the calculation out of reach at the present time . as an example
, we do not know today how to perform such a calculation even for the variance of the field @xmath28 .
instead , in order to interpret the numerical evidences of energy transfer observed in @xcite , we propose to do a simpler calculation , namely a perturbative expansion in power of @xmath0 that we hope will capture several key ingredients of the physics of energy transfer .
making such an expansion prevents the analysis of intermittent corrections since , in nature , the intermittency phenomenon can not be treated , as far as we know , with such an expansion . as in the independent case @xmath188 ( [ eq : multifractalfieldindicesindpt ] ) , we expect for the full vector field @xmath83 ( [ eq : multifractalfieldcorr ] ) a normalization constant such that @xmath83 is of finite variance .
recall that the elements of @xmath65 are gaussian random variables whose variance diverges logarithmically with @xmath15 ( cf .
equation [ eq : asympsigmaeps ] ) .
indeed , for the independent case , we have shown that the velocity field @xmath188 has to be normalized by a constant @xmath186 that itself diverges with @xmath15 .
it has been shown in @xcite that the multiplicative chaos @xmath255 has to be renormalized in order to define a proper random variable ( see also @xcite for a review on this topic ) . as far as @xmath83 ( [ eq : multifractalfieldcorr ] ) is concerned , since we do not know how to get the variance , we can not make such a normalization constant explicit , but is expected to be of the order of @xmath256 $ ] ( as it has been proved for the chaos in * ? ? ? * ) or @xmath257}$ ] ( as we have shown for the independent case ) . in the following perturbative expansion , we expect then a contribution of order @xmath169 ( and all the following even powers of @xmath0 ) from this constant .
since we do not know it explicitly , we will limit ourselves to a first order expansion in @xmath0 , which will not have any contribution from this possible unknown renormalizing constant .
we will see then that such a first order expansion exhibit energy transfer . at a given finite @xmath15 ,
the covariance of the vector field is given by @xmath258 = \int\phi^\epsilon_{ik}(-{\boldsymbol{z_1}})\phi^\epsilon_{pq}({\boldsymbol{h}}-{\boldsymbol{z_2}})\e \left[\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z_1)}}}\right)_{kl}\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z_2}})}\right)_{qr}w_l({\mathrm{d}}z_1)w_r({\mathrm{d}}z_2)\right].\end{aligned}\ ] ] expanding the matrix exponentials up to first order gives @xmath259=\e
\left[w_k({\mathrm{d}}z_1)w_q({\mathrm{d}}z_2)\right]\\ + & \gamma \left ( \e [ x_{kl}^\epsilon({\boldsymbol{z_1}})w_l({\mathrm{d}}z_1)w_q({\mathrm{d}}z_2)]+\e [ x_{qr}^\epsilon({\boldsymbol{z_2}})w_k({\mathrm{d}}z_1)w_r({\mathrm{d}}z_2)]\right)+o_\epsilon(\gamma),\end{aligned}\ ] ] the @xmath260-order term @xmath261 $ ] gives rise to the underlying gaussian contribution .
it obviously converges when @xmath66 . the first order term proportional to @xmath0 vanishes , since the expectation of the product of an odd number of zero average gaussian random variables always vanishes .
we are thus left with @xmath258 = ( \phi^\epsilon_{ik}\star\phi^\epsilon_{pq})({\boldsymbol{h } } ) + o_\epsilon(\gamma),\end{aligned}\ ] ] which is the covariance of the underlying gaussian velocity field ( see annex [ ann : gf ] ) .
there remains a possible dependence on @xmath15 in the remaining contributions @xmath262 .
we will neglect it and assume that the following limit when @xmath66 makes sense : @xmath263 = ( \phi_{ik}\star\phi_{pq})({\boldsymbol{h } } ) + o(\gamma).\end{aligned}\ ] ] thus , at first order in @xmath0 , the velocity field @xmath28 ( [ eq : multifractalfieldcorr ] ) has the same covariance as the underlying gaussian field @xmath6 .
we recall that the velocity increment @xmath195 is defined as @xmath264 where we have defined the even function @xmath197 assuming no summation over the index @xmath201 , we find @xmath265.\end{aligned}\ ] ] performing similar expansions as for the covariance case , we obtain , up to first order in @xmath0 , @xmath266\\ & = \gamma \e\big [ x^\epsilon_{k_1l_1}({\boldsymbol{z_1}})w_{l_1}({\mathrm{d}}z_1)w_{k_2}({\mathrm{d}}z_2)w_{k_3}({\mathrm{d}}z_3)+x^\epsilon_{k_2l_2}({\boldsymbol{z_2}})w_{k_1}({\mathrm{d}}z_1)w_{l_2}({\mathrm{d}}z_2)w_{k_3}({\mathrm{d}}z_3)\\ & + x^\epsilon_{k_3l_3}({\boldsymbol{z_3}})w_{k_1}({\mathrm{d}}z_1)w_{k_2}({\mathrm{d}}z_2)w_{l_3}({\mathrm{d}}z_3)\big]+ o_\epsilon(\gamma).\end{aligned}\ ] ] the three terms on the right - hand side of the former development give similar contributions once inserted in the expression of @xmath267 .
focusing for example on the first term , using the fact that the product of a even number of zero - averaged gaussian variables factorizes into pair products under an expectation value ( isserlis theorem ) and omitting for convenience the obvious dependence of @xmath268 on @xmath15 , we obtain @xmath269\\ & = \gamma_{k_1l_1}^{k_2}({\boldsymbol{z_1}}-{\boldsymbol{z_2}})\e \big[w_{l_1}({\mathrm{d}}z_1)w_{k_3}({\mathrm{d}}z_3)\big]{\mathrm{d}}z_2+\gamma_{k_1l_1}^{k_3}({\boldsymbol{z_1}}-{\boldsymbol{z_3}})\e \big[w_{l_1}({\mathrm{d}}z_1)w_{k_2}({\mathrm{d}}z_2)\big]{\mathrm{d}}z_3,\end{aligned}\ ] ] where we have used that @xmath270 .
once again , the two terms on the right - hand side of the former equality give similar contributions once inserted in the expression of @xmath267 .
we end up with , using the change of variable @xmath271 and performing the remaining integration over @xmath272 , @xmath273 note that because of the parity of the function @xmath274 , the function @xmath275 is even .
thus , if the correlator @xmath276 was odd , which is the case when @xmath65 is proportional to the identity matrix ( i.e. considering a scalar multiplicative chaos ) , the third order moment of velocity increments would have vanished .
this is consistent with the conclusions of @xcite . in our case ,
the use of a matrix field @xmath65 ensures a non trivial third order moment . without loss of generality
, we work for instance with the first velocity component @xmath277 , and we rename repeated indices as @xmath278 we can write @xmath279 to obtain @xmath280 consider now a longitudinal velocity increment , define the unit - vector @xmath281 along the first direction such that @xmath282 , i.e. @xmath283 .
we have @xmath284,\ ] ] and we obtain @xmath285 with @xmath286 in this expression for @xmath89 , take now the limit @xmath66 , assuming that all the integrals converge . noticing that @xmath287 with @xmath288=-\epsilon_{1qk}\varphi_l(0)x_q \mathcal h_h({\boldsymbol{x}}),\ ] ] we obtain @xmath289 which shows that @xmath290 ^ 2}{|{\boldsymbol{h}}|^{7/2}}{\mathrm{d}}z{\mathrm{d}}h.\end{aligned}\ ] ] this shows that up to first order in @xmath0 , the third order moment of velocity increments ( [ eq : defdepsilon ] ) , if the term @xmath262 remains bounded when @xmath126 , behaves as @xmath291 where the constant @xmath92 given in ( [ eq : predd ] ) is not trivially zero .
we use the same kind of simulations as those displayed in figure [ fig : flatness ] to estimate the numerical value of this constant @xmath92 ( [ eq : predd ] ) .
recall that the explicit value of the constant @xmath92 ( [ eq : predd ] ) involves a six - dimensional integral which is tricky to estimate through numerical integration .
instead , we use the randomness of the velocity field in order to get an estimation of it .
more precisely , in order to get rid of the non - universal factor @xmath38 , we show in figure [ fig : skew ] the results for the skewness @xmath292 , i.e. the non dimensionalized ratio of the third order moment and the power @xmath293 of the second order moment , namely @xmath294^{3/2}}.\ ] ] according to the former @xmath0-expansion ( [ eq : predfinal3rdmom ] ) , we predict , at small @xmath0 , a skewness @xmath295 independent of the scale @xmath29 , and we get @xmath296^{3/2 } } + o(\gamma),\end{aligned}\ ] ] where the constant @xmath297 is given in annex [ ann : gf ] ( [ eq : c2gparr ] ) .
we remark that , at this stage , the value of @xmath298 at first order in @xmath0 is universal , in the sense that it does not depend on the large - scale quantity @xmath38 and on the precise underlying regularization model given in ( [ eq : regulnorm ] ) . in figure
[ fig : skew](a ) we display the numerical estimation of the skewness obtained from simulations of the velocity field @xmath28 ( [ eq : multifractalfield ] ) .
we indeed see that , for small values of @xmath0 , the skewness @xmath299 is independent of the scale @xmath29 .
remark also that the skewness is negative for any @xmath0 , as it is observed in turbulence .
we gather in figure [ fig : skew](b ) values of the skewness at vanishing scale @xmath298 .
we observe a linear dependence of @xmath299 on @xmath0 , which shows ( numerically ) that at first order in @xmath0 , the proposed velocity field @xmath28 ( [ eq : multifractalfield ] ) gives a non trivial , non vanishing and strictly negative skewness .
fitting the linear behaviour of @xmath298 at small @xmath0 , we find @xmath300 , showing that , at first order in @xmath0 , our prediction ( [ eq : predskewness ] ) makes sense . as far
as turbulence is concerned , we have found in section [ sec : intermittencyns ] that the particular value @xmath84 gives a fairly realistic behaviour of the flatness when compared against empirical findings . when inserted into our first order prediction ( [ eq : predskewness ] ) , we obtain @xmath301 , which overestimates typical values obtained in experiments ( see @xcite and @xcite ) . let us comment on this discrepancy .
first of all , real turbulence exhibits intermittent corrections , making the skewness ( equation [ eq : defskew ] ) slightly dependent on the scale @xmath29 and a skewness of derivatives @xmath298 dependent on the reynolds number ( or on @xmath15 in the present picture ) .
these corrections can not be seen with such a perturbative expansion as proposed in equation ( [ eq : predskewness ] ) . furthermore , the typical intermittency parameter @xmath84 , or equivalently @xmath302 is out of the range of observed linear behaviour of @xmath298 with @xmath0 ( figure [ fig : skew]b ) .
these two facts can explain this discrepancy .
if now we take a look at the value of the skewness we are obtaining in the numerical simulation for the parameter @xmath302 , without invoking the perturbative expansion ( see figure [ fig : skew]b ) , we find @xmath303 which is very close to what is obtained in real flows ( see * ? ? ?
this theoretical and numerical study shows that indeed the proposed velocity field @xmath28 ( [ eq : multifractalfield ] ) , and for the first time as far as we know , exhibits non vanishing and realistic energy transfer .
we have studied the statistical properties of an explicit random velocity field ( [ eq : multifractalfield ] ) able to reproduce the main properties of a fully developed turbulent flow , as observed in experiments and numerical simulations .
to do so , we have performed simulations up to @xmath71 grid points , and developed analytical techniques when calculations were possible .
this claimed realistic picture of homogeneous and isotropic turbulence includes the teardrop shape of the joint density of the invariants of the velocity gradient tensor and the preferential alignment of vorticity with the eigenframe of the deformation matrix ( section [ sec : numa ] ) .
furthermore , assuming independence of the building block of the intermittency phenomenon , i.e. the matrix multiplicative chaos ( @xmath255 ) , on the underlying gaussian white measure ( @xmath54 ) , we are able to derive the spectrum of exponents @xmath304 ( see section [ sec : indptcase ] ) , which is thus found to be a non - linear function of the order @xmath37 and gives full meaning to the free parameter @xmath0 .
as we explained , such a strong simplification forbids energy transfer . in order to obtain some insights in the physics of energy transfer
, we perform a perturbative expansion of the full vector field in powers of @xmath0 and show that at first order , the field exhibits a non vanishing third order moment of longitudinal velocity increments , establishing the dissipative nature of the field .
this behaviour is consistent with kolmogorov s @xmath85-law , although the precise value of the prefactor @xmath85 is not predicted since it requires a link ( and thus an additional free parameter ) between the regularization scale @xmath15 and the kinematic viscosity @xmath145 .
nonetheless , we are led to the conclusion that the proposed velocity field is realistic of turbulence when we set @xmath84 , a value that has been obtained when comparing the power - law behaviour of the flatness ( see equation [ eq : flatnessesexpo ] ) with experimental findings , taking into account finite - size corrections to the scalings . as far as we know , the velocity field @xmath28 ( [ eq : multifractalfield ] ) is the first stochastic process proposed in the literature that is able to predict a non vanishing mean energy transfer across scales .
at least numerically , and supported by the perturbative expansion performed in section [ sec : enertrans ] , it seems that the non vanishing nature of the third order moment of longitudinal velocity increments remains in the asymptotic limit @xmath66 .
this justifies the term _ dissipative _ used in the title of the present article and gives hope to define one day rigorously the limiting velocity field @xmath305 as a stochastic representation of weak solutions of the euler equations , as they are depicted in onsager s contribution to turbulence ( see the review article @xcite on this subject ) .
in particular , the modern view of energy transfer of @xcite shows that standard local energy budget is possibly violated by an additional dissipative term , independent of viscosity , related to the non differential nature of the velocity field
. it would be interesting to relate this view to the statistical properties of the proposed field @xmath2 .
to do so , new mathematical techniques are needed in order to handle the tricky correlated nature of the matrix @xmath65 and vector @xmath54 fields . as perspectives ,
let us mention first the usefulness of such a random field in the context of mean square estimation @xcite and related conditional averages as far as turbulent applications are concerned ( see for example * ? ? ?
indeed , such a procedure can be used to provide new types of closures of the subgrid stress tensor while performing a large eddy simulation @xcite .
more recently , it was shown that a gaussian velocity field is able to reproduce non trivial properties of the pressure hessian that enters the dynamics of the velocity gradient tensor ( see * ? ? ?
more precisely , it is shown numerically in @xcite that the following average of the hessian of the pressure @xmath306 conditioned on the local velocity gradient tensor @xmath158 , i.e. @xmath307,\ ] ] is well approximated , when compared against direct numerical simulations of the navier - stokes equations , if one assumes the velocity field to be gaussian , such as @xmath32 ( [ eq : gaussianfield ] ) .
further analytical work in this direction by @xcite shows that indeed the prediction of this conditional average starting from a gaussian velocity field is able , when slightly modified , to regularize the finite - time divergence implied by the self - stretching term .
these results could then be extended while assuming for the velocity field a stochastic structure such as the one we are proposing ( [ eq : multifractalfield ] ) . a first step in this direction
could be reached assuming furthermore independence of the multiplicative chaos on the underlying white measure because this simplification makes calculations tractable ( section [ sec : indptcase ] ) .
finally let us comment on the remaining free parameter @xmath0 that governs both energy transfer and the intermittency phenomenon in our velocity field @xmath28 ( [ eq : multifractalfield ] ) .
it would be much welcome to use further constraints from the equations of motion in order to set its value that would compare in an appropriate way against experiments .
constraints on structure functions of order higher than the third one have been derived by @xcite and @xcite .
they involve the pressure field @xmath306 which is fully determined by the corresponding velocity field through the poisson equation .
we could inquire if these constraints , which include pressure gradient increments , can be used to give a precise range of possible values for the intermittency coefficient @xmath0 .
we keep these perspectives for future investigations .
we thank b. castaing , k. gawedzki and e. lvque for fruitful discussions .
warm acknowledgments to r. robert , r. rhodes and v. vargas for their help at the early stage of this work .
this work was supported by the anr liouville project , grant anr-15-ce40 - 0013 of the french agence nationale de la recherche .
one of the authors ( r.m.p . )
also thanks capes for financial support through the scholarship process number 9497/13 - 7 .
we gratefully acknowledge support from the psmn ( ple scientifique de modlisation numrique ) computing center of ens de lyon .
the homogeneous field of matrices as defined in equation [ eq : xepsilon ] takes on the following structure : * the diagonal entries @xmath308 are independent of the off - diagonal entries @xmath309 for @xmath310 , * the covariance matrix of the diagonal entries @xmath311 $ ] , for @xmath312 , is given by @xmath313 where @xmath314 $ ] , @xmath115 the @xmath116 identity matrix and @xmath315 stands for the @xmath116 matrix filled with the coefficient @xmath316 in all entries . as @xmath15 gets smaller , we get the following asymptotic structure : @xmath317 * the off - diagonal entries @xmath318 are mutually independent with variance @xmath319 .
this covariance structure is peculiar to isotropic matrices , as it is demonstrated in @xcite where it corresponds to the particular case of trace - free matrices .
let us now turn to the characterization of the gaussian field of matrices ( [ eq : xepsilon ] ) .
it is enough , since it is gaussian , to give the covariance structure .
up to an additive independent gaussian matrix with a bounded covariance structure as @xmath320 , we obtain the following asymptotic covariance of the field @xmath65 : * the diagonal entries @xmath321 are independent of the off - diagonal entries @xmath322 , * when @xmath323 , the @xmath116 covariance matrix of the diagonal entries @xmath324 $ ] , for @xmath312 , is given by @xmath325 , where @xmath326 \build{\sim}_{|{\boldsymbol{x}}-{\boldsymbol{y}}|\rightarrow 0}^{}\ln\frac{l}{|{\boldsymbol{x}}-{\boldsymbol{y}}| } \mbox { .}\ ] ] * the off - diagonal entries @xmath327 are mutually independent , each of which with covariance given by , for @xmath328 , @xmath329 \build{\sim}_{\epsilon\rightarrow 0}^ { } \frac{3}{4 } \sigma_{|{\boldsymbol{x}}-{\boldsymbol{y}}|}^2 \mbox { .}\ ] ] the field of isotropic matrices @xmath112 is given by @xmath330 + [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{w}}({\mathrm{d}}y ) ] \otimes \frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon}.\ ] ] next , we are showing that indeed this explicit field leads asymptotically ( @xmath323 ) to the structure previously described in section [ sec : genecovmat ] .
let us consider first the diagonal components @xmath331 of the random matrix - valued field ( [ eq : fieldisomatrix ] ) .
for instance , consider the @xmath332 at the spatial location @xmath333 : @xmath334.\ ] ] we get @xmath335= \frac{15}{8{\pi}}\int \frac{(h_1-y_1)y_1}{|{\boldsymbol{h}}-{\boldsymbol{y}}|^{7/2}_\epsilon|{\boldsymbol{y}}|^{7/2}_\epsilon}[(h_2-y_2)y_2+(h_3-y_3)y_3]{\mathrm{d}}y.\end{aligned}\ ] ] taking @xmath336 in the previous integral , we find , using a spherical integration and the explicit form of the regularized norm ( [ eq : regulnorm ] ) @xmath337&= \frac{15}{4{\pi}}\int \frac{y_1 ^ 2y_2 ^ 2}{|{\boldsymbol{y}}|^{7}_\epsilon}{\mathrm{d}}y \\ & = \frac{15}{4{\pi}}\int_{\theta=0}^{{\pi } } \int_{\varphi=0}^{2{\pi } } \cos^2(\theta)\sin^3(\theta)\sin^2(\varphi)d\theta d\varphi \int_{0}^l \frac{\rho^6d\rho}{\left[\rho^2+\epsilon^2\right]^{7/2}}\\ & = \int_{0}^l \frac{\rho^6d\rho}{\left[\rho^2+\epsilon^2\right]^{7/2}}\\ & \build{\sim}_{\epsilon\to 0}^{}\ln\frac{l}{\epsilon},\end{aligned}\ ] ] which entails equation [ eq : asympgammabeta ] .
the asymptotic logarithmic behaviour can be easily obtained while performing the change of variable @xmath338 .
let us remark that the asymptotic variance of the diagonal elements of the matrix @xmath65 is independent of the precise regularization procedure .
take now @xmath339 .
every integral converges when @xmath126 and we obtain @xmath340=\frac{15}{8{\pi}}\int \frac{(h_1-y_1)y_1}{|{\boldsymbol{h}}-{\boldsymbol{y}}|^{7/2}|{\boldsymbol{y}}|^{7/2}}[(h_2-y_2)y_2+(h_3-y_3)y_3]{\mathrm{d}}y.\end{aligned}\ ] ] we will now compute the equivalent of the former correlation function as @xmath341 .
performing the change of variable @xmath342 , i.e. @xmath343 , we find @xmath344{\mathrm{d}}z \end{aligned}\ ] ] where the integration domain is over @xmath345 .
as @xmath341 , we can always choose a constant @xmath346 such that for @xmath347 , we have @xmath348 .
in the previous integrals , the contribution resulting from the integration over the finite domain @xmath349 gives a bounded function of the norm @xmath350 .
the remaining contributions diverge with @xmath341 .
thus , we get @xmath351{\mathrm{d}}z= \frac{15}{8{\pi}}\int_{c\le |z|\le l / h } \frac{z_1 ^ 2z_2 ^ 2}{|{\boldsymbol{z}}|^{7}}{\mathrm{d}}z \\ & \build{\sim}_{h\to 0}^ { } \ln\frac{l}{h},\end{aligned}\ ] ] which entails equation [ eq : asympgammacorrbeta ] . of great importance
is also the cross - covariance of diagonal components , say @xmath352 $ ] . recall that @xmath334\ ] ] and @xmath353 .\ ] ] in the same manner as followed to compute the correlation of a diagonal component , we get @xmath354= -\frac{15}{8{\pi}}\int \frac{(h_1-y_1)(h_2-y_2)y_1y_2}{|{\boldsymbol{h}}-{\boldsymbol{y}}|^{7/2}_\epsilon|{\boldsymbol{y}}|^{7/2}_\epsilon}{\mathrm{d}}y,\end{aligned}\ ] ] which leads to @xmath355&= -\frac{15}{8{\pi}}\int \frac{y_1 ^ 2y_2 ^ 2}{|{\boldsymbol{y}}|^{7}_\epsilon}{\mathrm{d}}y \build{\sim}_{\epsilon\to 0}^{}-\frac{1}{2}\ln\frac{l}{\epsilon}.\end{aligned}\ ] ] in the same spirit , the cross - covariance of the diagonal element is given by @xmath356= \e [ x_{11}(0)x_{22}({\boldsymbol{h } } ) ] \build{\sim}_{h|\to 0}^ { } -\frac{1}{2}\ln\frac{l}{h}.\ ] ] consider now an off - diagonal element such as @xmath357 .
we have @xmath358 . \end{aligned}\ ] ] in particular @xmath359 , \end{aligned}\ ] ] such that @xmath360 & = \frac{15}{32{\pi}}\int\frac{1}{|{\boldsymbol{y}}|^{7}_\epsilon}\left[y_1 ^ 4+y_2 ^ 4+y_1 ^ 2y_3 ^ 2+y_2 ^ 2y_3 ^ 2 - 2y_1 ^ 2y_2 ^ 2\right]{\mathrm{d}}y\\ & = \frac{15}{16{\pi}}\int\frac{y_1 ^ 4}{|{\boldsymbol{y}}|^{7}_\epsilon}dy= \frac{15}{8}\int_0^{{\pi}}\cos^4(\theta)\sin(\theta)d\theta\int_0^{l}\frac{\rho^6}{[\rho^2+\epsilon^2]^{7/2}}d\rho\\ & = \frac{3}{4}\int_0^{l}\frac{\rho^6}{[\rho^2+\epsilon^2]^{7/2}}d\rho\\ & \build{\sim}_{\epsilon\to 0}^ { } \frac{3}{4}\ln\frac{l}{\epsilon}.\end{aligned}\ ] ] in the same spirit , the cross - covariance of the off - diagonal element is given by @xmath361= \e [ x_{12}(0)x_{12}({\boldsymbol{h } } ) ] \build{\sim}_{h\to 0}^ { } \frac{3}{4}\ln\frac{l}{h}.\ ] ] it can furthermore be shown that the off - diagonal elements are mutually independent , for example @xmath362=0,\ ] ] independent of the diagonal elements , i.e. @xmath363=0,\ ] ] and the respective cross - correlation functions , i.e. @xmath364 $ ] and @xmath365 $ ] are bounded functions of their argument @xmath333 , in particular they do not diverge logarithmically with @xmath350 .
consider the non intermittent case ( i.e. @xmath160 ) .
we are left considering the following gaussian isotropic and homogeneous incompressible velocity field @xmath366 ( [ eq : gaussianfield ] ) : @xmath367 where the kernel @xmath368 is given by @xmath369 it is easily seen that the covariance of this field is ( considering only the covariance between the null vector and @xmath333 by homogeneity ) @xmath370=(\phi^{\epsilon}_{ik}\star\phi^{\epsilon}_{pk})({\boldsymbol{h } } ) = \int \phi^{\epsilon}_{ik}({\boldsymbol{x}})\phi^{\epsilon}_{pk}({\boldsymbol{x}}+{\boldsymbol{h } } ) { \mathrm{d}}x.\ ] ] which defines the correlation product @xmath192 .
the kernel @xmath371 is singular for @xmath372 when @xmath66 .
thus , the limiting integral exists if the strongest singularity obtained while considering the variance ( i.e. @xmath373 ) is integrable in three dimensions .
this singularity @xmath374 is integrable in three dimensions for @xmath23 . in particular , for this range of hurst exponent @xmath23 , the variance of the velocity field converges and we can write @xmath375 since we have assumed that @xmath242 is a radially symmetric function .
remark also that the rapid decay of the cut - off function @xmath242 ensures a finite variance of the process . in the sequel , we will drop the dependence on @xmath15 and consider the limiting process @xmath32 . more generally , as expected from an isotropic velocity field and following the notations of @xcite , the covariance structure of this gaussian velocity field can be written as @xmath376&= \delta_{ip}\int \varphi_l({\boldsymbol{x}})\varphi_l({\boldsymbol{x}}+{\boldsymbol{h}})\frac{{\boldsymbol{x}}\cdot({\boldsymbol{x}}+{\boldsymbol{h}})}{|{\boldsymbol{x}}|^{\frac{5}{2}-h}|{\boldsymbol{x}}+{\boldsymbol{h}}|^{\frac{5}{2}-h}}{\mathrm{d}}x\\ & -\int \varphi_l({\boldsymbol{x}})\varphi_l({\boldsymbol{x}}+{\boldsymbol{h}})\frac{x_p(x_i+h_i)}{|{\boldsymbol{x}}|^{\frac{5}{2}-h}|{\boldsymbol{x}}+{\boldsymbol{h}}|^{\frac{5}{2}-h}}{\mathrm{d}}x\\ & = f(|{\boldsymbol{h}}|)h_ih_p+g(|{\boldsymbol{h}}|)\delta_{ip},\end{aligned}\ ] ] where the functions @xmath377 and @xmath378 are radially symmetric functions . at this stage , as it is usually done in turbulence literature @xcite , it is convenient to decompose the covariance of the field in terms of the _ longitudinal _ and _ transverse _ velocity correlations defined as @xmath379 \mbox { and } r^g_{\perp}({\boldsymbol{h } } ) = \e [ u^{g}_{\perp}(0)u^{g}_{\perp}({\boldsymbol{h}})]\ ] ] where @xmath380 and @xmath381 denote velocity components parallel and normal respectively to the vector separation @xmath333 . in homogeneous and isotropic turbulence ,
the correlations @xmath382 and @xmath383 are functions of the norm @xmath350 only , and @xmath384
. furthermore we have the following decomposition : @xmath385=(\phi_{ik}\star\phi_{pk})({\boldsymbol{h } } ) = \frac{r^g_{\parallel}(h)-r^g_{\perp}(h)}{h^2}h_ih_p+r^g_{\perp}(h)\delta_{ip}.\ ] ] the incompressible condition , i.e. @xmath386 , finally implies that the velocity covariance @xmath387 depends on a single scalar function since @xmath388 the velocity increment @xmath389 reads @xmath390 where we have defined @xmath391.\end{aligned}\ ] ] note that so defined , the kernel @xmath392 is an even function of its argument ( recall that @xmath242 is a radially symmetric function ) , i.e. @xmath393 also , in order to simplify the following expressions , since the field is homogeneous , we will only consider velocity increments at the position @xmath394 and note @xmath395 without loss of generality ( by isotropy ) , consider only the first component @xmath396 of the gaussian vector field @xmath32 .
we get : @xmath397 take @xmath398 as a unit vector , consider then the displacement @xmath399 .
we find , making the change of variables @xmath400 : @xmath401 \frac{y_j+e_j/2}{|{\boldsymbol{y}}+{\boldsymbol{e}}/2|^{\frac{5}{2}-h } } + \varphi_l[\ell ( { \boldsymbol{e}}/2-{\boldsymbol{y}})]\frac{e_j/2-y_j}{|{\boldsymbol{e}}/2-{\boldsymbol{y}}|^{\frac{5}{2}-h } } \right)^2\ , { \mathrm{d}}y.\end{aligned}\ ] ] to conclude regarding scaling behaviour , we need to discuss the remaining @xmath29 dependence in the cutoff functions @xmath242 .
hereafter , we will assume that @xmath402 which ensures that the integrals converge without the need of cutoff functions .
thus , for this range of parameters @xmath4030,1[$ ] , longitudinal and transverse velocity increments have the same scaling behaviour given by the spectrum of exponents @xmath404 .
considering for instance the unit vectors for the longitudinal case @xmath405 and @xmath406 for the transverse one , we finally obtain @xmath407 with @xmath408 and @xmath409 notice that the constants @xmath410 and @xmath411 are non null and positive . using the notation @xmath412 we can write the constant @xmath297 in the following convenient way @xmath413
since the field is gaussian , it is straightforward to get the higher order structure functions as : @xmath414 with @xmath415 we observe also that all odd moments vanish for a gaussian process , i.e. for @xmath217 @xmath416
we recall the definition of the velocity field @xmath188 ( [ eq : multifractalfieldindicesindpt ] ) @xmath417 where again @xmath99 and the following _ matrix _ field : @xmath418 + [ ( { \boldsymbol{x}}-{\boldsymbol{y}})\wedge { \boldsymbol{w}}'({\mathrm{d}}y ) ] \otimes \frac{{\boldsymbol{x}}-{\boldsymbol{y}}}{|{\boldsymbol{x}}-{\boldsymbol{y}}|^{7/2}_\epsilon}\ ] ] where now the vector noise @xmath181 is independent of the vector noise @xmath54 , namely for any @xmath182 and any components @xmath183 and
@xmath184 we have @xmath419 .
as we will see in the following , the field ( [ eq : annmultifractalfieldindicesindpt ] ) needs to be renormalized in order to converge , when @xmath126 , towards a finite - variance process . to ensure such a finite variance
, we introduce a deterministic normalization constant ( which depends on @xmath420 @xmath421.\ ] ] this is a standard way to renormalize the multiplicative chaos @xcite .
since @xmath54 and @xmath65 are independent , we easily obtain @xmath422\\ & = \frac{1}{c_\epsilon}\int\phi_{ik}^\epsilon({\boldsymbol{x}}-{\boldsymbol{z}})\e\left[\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z}})}\right)_{kl}\right]\e\left[w_l({\mathrm{d}}z)\right]\\ & = 0.\end{aligned}\ ] ] thus , the vector field ( [ eq : annmultifractalfieldindicesindpt ] ) is of zero average , as expected for an isotropic vector field . for the covariance
we get @xmath191 & = \frac{1}{c_\epsilon^2}\int\phi^\epsilon_{ik}(-{\boldsymbol{z_1}})\phi^\epsilon_{pq}({\boldsymbol{h}}-{\boldsymbol{z_2}})\e \left[\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z_1}})}\right)_{kl}\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z_2}})}\right)_{qr}w_l({\mathrm{d}}z_1)w_r({\mathrm{d}}z_2)\right]\\ & = \frac{1}{c_\epsilon^2}\int\phi^\epsilon_{ik}(-{\boldsymbol{z_1}})\phi^\epsilon_{pq}({\boldsymbol{h}}-{\boldsymbol{z_2}})\e \left[\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z_1}})}\right)_{kl}\left({\mathrm{e}}^{\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z_2}})}\right)_{qr}\right]\e \left[w_l({\mathrm{d}}z_1)w_r({\mathrm{d}}z_2)\right]\\ & = \frac{1}{c_\epsilon^2}\int\phi^\epsilon_{ik}(-{\boldsymbol{z}})\phi^\epsilon_{pq}({\boldsymbol{h}}-{\boldsymbol{z}})\e \left[\left({\mathrm{e}}^{2\gamma { \mathsfbi{x}}^\epsilon({\boldsymbol{z}})}\right)_{kq}\right]{\mathrm{d}}z.\end{aligned}\ ] ] the field of matrices @xmath65 is isotropic , thus , according to @xcite , we get @xmath423 = c_\epsilon^2 \delta_{kq},\ ] ] where the renormalizing constant is given in ( [ eq : annrenormindpt ] ) .
we can see that the covariance structure of the vector field ( [ eq : annmultifractalfieldindicesindpt ] ) is the same as the one obtained from the underlying gaussian field ( [ eq : gaussianfield ] ) , namely @xmath191 = \e [ u^{g,\epsilon}_i(0)u^{g,\epsilon}_p({\boldsymbol{h}})]=(\phi^{\epsilon}_{ik}\star\phi^{\epsilon}_{pk})({\boldsymbol{h}}).\end{aligned}\ ] ] in particular , the field ( [ eq : annmultifractalfieldindicesindpt ] ) converges in a @xmath193 sense when @xmath66 , and it has same variance and covariance as the underlying gaussian field . in the sequel , we note @xmath424 .
as we have seen , the covariance of the vector field ( [ eq : annmultifractalfieldindicesindpt ] ) is the same as the one obtained from the underlying gaussian field .
thus it has also the same second order structure functions , both longitudinal and transverse ( see annex [ ann : gf ] ) .
take @xmath426 . using the independence between @xmath65 and @xmath54
, we can write ( hereafter , no summation over the index @xmath201 ) , @xmath427\e \left[\prod_{q=1}^nw_{l_q}({\mathrm{d}}z_q)\right].\ ] ] we can see that all odd - order structure functions vanish since the expectation of an odd product of the vector white noise components is always zero , i.e. @xmath428=0.\ ] ] thus @xmath429 this field ( [ eq : annmultifractalfieldindicesindpt ] ) is not dissipative , and does not exhibit energy transfer .
we consider in the sequel only even - order structure functions to get : @xmath430\e \left[\prod_{q=1}^{2n}w_{l_q}({\mathrm{d}}z_q)\right].\ ] ] factorizing the @xmath215-product of the white noise vector ( isserlis theorem ) as @xmath431 & = \frac{1}{2^nn!}\sum_{\mathcal s\in s_{2n } } \prod_{i=1}^n \e \left[w_{l_{\mathcal s(2i-1)}}({\mathrm{d}}z_{\mathcal s(2i-1)})w_{l_{\mathcal s(2i)}}({\mathrm{d}}z_{\mathcal s(2i)})\right]\\ & = \frac{1}{2^nn!}\sum_{\mathcal s\in s_{2n } } \prod_{i=1}^n \delta_{l_{\mathcal s(2i-1)},l_{\mathcal s(2i)}}\delta_{z_{\mathcal s(2i-1)},z_{\mathcal s(2i)}}dz_{\mathcal s(2i)}\end{aligned}\ ] ] where @xmath432 is the permutation ensemble of @xmath433 , which has cardinality @xmath434 .
thus @xmath435\prod_{q=1}^n { \mathrm{d}}z_{\mathcal s(2q)}.\end{aligned}\ ] ] it is clear that the sum over the permutations gives @xmath434 equal contributions , and we are left with @xmath436\prod_{q=1}^n { \mathrm{d}}z_{2q}.\end{aligned}\ ] ] note that , in the gaussian case , i.e. @xmath160 , which is equivalent to @xmath437 and @xmath438 , we recover the statistics obtained for the underlying gaussian field ( see annex [ ann : gf ] ) , namely @xmath439^n\\ & = \frac{(2n)!}{2^nn!}\left[\left(\phi^{\epsilon,{\boldsymbol{\ell}}}_{ik}\star\phi^{\epsilon,{\boldsymbol{\ell}}}_{ik}\right)(0)\right]^n= \frac{(2n)!}{2^nn!}\left[\e ( \delta_\ell u_i^{ind,\epsilon})^{2}\right]^n= \frac{(2n)!}{2^nn!}\left[\e ( \delta_\ell u_i^{\epsilon , g})^{2}\right]^n.\end{aligned}\ ] ] of special interest are the fourth order structure functions @xmath198 ( longitudinal and transverse cases ) , and the respective flatnesses , i.e. @xmath199 and @xmath200 ( [ eq : flatnesses ] ) .
we get ( no summation over repeated index @xmath201 implied ) , @xmath440{\mathrm{d}}z_{2}{\mathrm{d}}z_{4}.\end{aligned}\ ] ] let us massage slightly the covariance of the matrix exponentials that enters the former expression .
to do so , we will use the theory developed by @xcite .
the field of isotropic symmetric matrices @xmath65 is homogeneous , and furthermore , the joint density of two matrices at the locations @xmath441 and @xmath442 depends only on the distance @xmath443 .
thus , it can be shown that @xmath230 = f_\epsilon ( h)\delta_{k_1k_2}\delta_{k_3k_4 } + g_\epsilon(h)\left[\delta_{k_1k_3}\delta_{k_2k_4}+\delta_{k_1k_4}\delta_{k_2k_3 } \right],\ ] ] where @xmath444\ ] ] and @xmath445.\ ] ] we remark that if @xmath65 is proportional to the identity , the function @xmath446 vanishes . using this isotropic form of a fourth - order tensor
, we obtain @xmath447 using the results of @xcite , we get @xmath448\build{\sim}_{\epsilon\to 0}^{}4(2\gamma\sigma_\epsilon)^4\left(1+\frac{1}{2}\right)^2{\mathrm{e}}^{4\gamma^2\sigma_\epsilon^2}{\mathrm{e}}^{-4\frac{1}{2}\gamma^2\sigma_{h}^2}\int_{o_3(\mathbb r)}|o_{11}|^2{\mathrm{e}}^{4(1+\frac{1}{2})\gamma^2\sigma^2_{h}|o_{11}|^2}{\mathrm{d}}o,\end{aligned}\ ] ] and similarly , @xmath449\build{\sim}_{\epsilon\to 0}^{}4(2\gamma\sigma_\epsilon)^4\left(1+\frac{1}{2}\right)^2{\mathrm{e}}^{4\gamma^2\sigma_\epsilon^2}{\mathrm{e}}^{-4\frac{1}{2}\gamma^2\sigma_{h}^2}\int_{o_3(\mathbb r)}{\mathrm{e}}^{4(1+\frac{1}{2})\gamma^2\sigma^2_{h}|o_{11}|^2}{\mathrm{d}}o,\end{aligned}\ ] ] where the remaining integration is performed over the orthogonal group @xmath450 , and recall that asymptotically @xmath451 ( when @xmath66 ) and @xmath452 ( when @xmath66 and after @xmath341 ) . in the same fashion , we can find @xmath453\build{\sim}_{\epsilon\to 0}^ { } \frac{8}{3}\gamma^2\sigma_\epsilon^2\left(1+\frac{1}{2}\right){\mathrm{e}}^{2\gamma^2\sigma_\epsilon^2}.\ ] ] thus , @xmath454=3 ^ 2{\mathrm{e}}^{-4\frac{1}{2}\gamma^2\sigma_{h}^2}\int_{o_3(\mathbb r)}|o_{11}|^2{\mathrm{e}}^{4(1+\frac{1}{2})\gamma^2\sigma^2_{h}|o_{11}|^2}{\mathrm{d}}o,\end{aligned}\ ] ] and similarly , @xmath455 = 3 ^ 2{\mathrm{e}}^{-4\frac{1}{2}\gamma^2\sigma_{h}^2}\int_{o_3(\mathbb r)}e^{4(1+\frac{1}{2})\gamma^2\sigma^2_{h}|o_{11}|^2}{\mathrm{d}}o.\end{aligned}\ ] ] this shows that the quantities @xmath456 and @xmath446 converge when @xmath66 .
we will note @xmath231 and @xmath232 their respective limits .
thus , @xmath457,\end{aligned}\ ] ] and @xmath458.\end{aligned}\ ] ] introducing the dawson integral @xmath459 @xmath460 we can obtain an explicit expression of former angular integrals at a finite scale @xmath29 as @xmath461,\end{aligned}\ ] ] and @xmath462.\end{aligned}\ ] ] using the asymptotic behaviour of the dawson integral , namely @xmath463 when @xmath464 , we get the asymptotic behaviour of the functions @xmath231 and @xmath232 , that is @xmath465 recall that asymptotically , the covariance of the diagonal elements of @xmath211 is logarithmic ( see equation [ eq : asympsigmacorr ] ) .
we can thus write it as @xmath466 where @xmath467 is a bounded function of its argument , showing that @xmath468 if the integrals exist , we can obtain the limit @xmath66 of the fourth order moment of velocity increments as @xmath469 with @xmath470 we must still make sure that the integrals exist . clearly
, the strongest singularities are obtained when the integration variable @xmath350 is such that its modulus is zero . while performing the integration over @xmath333 in the asymptotic form of @xmath471 , we thus encounter the singularity @xmath472 which is integrable in three dimensions only if @xmath473 let us get the scaling behaviour of the fourth order structure function .
define first the unit - vector @xmath398 and write @xmath399 .
recall too that ( see the scaling arguments developed in annex [ ann : gf ] ) @xmath474 with @xmath475.\ ] ] in the same spirit , the intermittent correction to the fourth order structure function comes from the asymptotic behaviour of @xmath231 and @xmath232 at small arguments , namely @xmath476 we thus obtain the following scaling behaviour of the fourth order structure function @xmath477({\boldsymbol{h}})\left ( \frac{l}{h}\right)^{4\gamma^2}{\mathrm{d}}h.\end{aligned}\ ] ] this shows that longitudinal and transverse fourth order structure functions have the same scaling behaviours .
more precisely , without loss of generality , consider the first velocity component @xmath478 and the two unit vectors @xmath405 and @xmath406 for the transverse one , we finally get @xmath208 with @xmath479({\boldsymbol{h}})\left ( \frac{l}{h}\right)^{4\gamma^2}{\mathrm{d}}h\ ] ] and @xmath480({\boldsymbol{h}})\left ( \frac{l}{h}\right)^{4\gamma^2}{\mathrm{d}}h \mbox { .}\ ] ] this shows that the velocity field @xmath67 ( [ eq : multifractalfieldindicesindpt ] ) , built assuming independence of @xmath211 and @xmath54 , is intermittent , and the respective flatnesses ( [ eq : flatnesses ] ) behave as power - laws times a logarithmic correction with ( see annex [ ann : gf ] for the expression of the second - order structure functions that are the same as the ones obtained from the underlying gaussian velocity field ) @xmath481 1972 possible refinement of the lognormal hypothesis concerning the distribution of energy dissipation in intermittent turbulence . in _
statistical models and turbulence _
m. rosenblatt & c. van atta ) , _ lecture notes in physics _
springer berlin heidelberg . | we investigate the statistical properties , based on numerical simulations and analytical calculations , of a recently proposed stochastic model for the velocity field of an incompressible , homogeneous , isotropic and fully developed turbulent flow .
a key step in the construction of this model is the introduction of some aspects of the vorticity stretching mechanism that governs the dynamics of fluid particles along their trajectories .
an additional further phenomenological step aimed at including the long range correlated nature of turbulence makes this model dependent on a single free parameter @xmath0 that can be estimated from experimental measurements .
we confirm the realism of the model regarding the geometry of the velocity gradient tensor , the power - law behaviour of the moments of velocity increments ( i.e. the structure functions ) including the intermittent corrections and the existence of energy transfer across scales .
we quantify the dependence of these basic properties of turbulent flows on the free parameter @xmath0 and derive analytically the spectrum of exponents of the structure functions in a simplified non dissipative case . a perturbative expansion in power of @xmath0 shows that energy transfer , at leading order , indeed take place , justifying the dissipative nature of this random field . |
there is a sustained interest in beams of light carrying orbital angular momentum , particularly in those with screw topological wave front dislocations or vortices .
these vortex light beams have open new perspectives in information encoding , quantum entanglement , imaging , optical trapping , and in diverse forms of transference of optical angular momentum to matter , e. g. , to micro- or nano - particles , to bose - einstein condensates or atoms @xcite .
although most of the research initially focused on laguerre - gauss beams @xcite , many other vortex beams have been explored , many of them endowed of the advantageous property of being diffraction - free , as high - order bessel beams ( bbs ) in linear media @xcite , vortex solitons in transparent media with suitable nonlinearities @xcite , or dissipative vortex solitons in nonlinear media with gain and losses @xcite . in applications involving pumping of orbital angular momentum to matter ,
the net transference of energy and angular momentum implies a power loss of the pumping vortex beam , whose effect in its propagation is not usually taken into account . in ( dissipative )
vortex solitons , for example , perturbation of the precise balance between nonlinearities and diffraction and between gain and losses may result in the quenching of the solitary regime . although there is no known light beam that can overcome the attenuation effects of linear absorption , there are light beams , known as nonlinear unbalanced bessel beams ( nl - ubbs ) that can propagate without any diffraction and attenuation while their energy is transferred to matter nonlinearly via multiphoton absorption @xcite .
nl - ubbs belong to the family of conical waves and as such they transport infinite power .
this power reservoir flows permanently from the linear outer rings towards the central peak of intensity , where most of nonlinear power losses take place , refilling it .
nl - ubbs has been shown to play an important role in the filamentation of ultrahsort pulses , @xcite to be able to create long - lived fluorescence channels excited by multiphoton absorption of the nl - ubb power @xcite , and to be easily generated experimentaly from a zero - order linear bb @xcite . in this paper
, we point out that the above nl - ubb is the fundamental member of an infinite family of nl - ubbs with helical wave fronts and carrying orbital angular momentum . unlike linear vortex beams and vortex solitons
, these vortex nl - ubbs incorporate a mechanism of transference of their angular momentum to matter : the nonlinear absorption just supporting their stationarity . light power and angular momentum spiral inward permanently ,
and they are transferred to matter within a thin ring surrounding the phase singularity .
simultaneous absorption of @xmath0 photons , each photon carrying ( arbitrarily high ) orbital angular momentum @xmath1 ( @xmath2 ) could be used as an efficient method of optical pumping of angular momentum @xcite .
we also stress that self - focusing or self - defocusing nonlinearities @xcite , saturable or not @xcite , local or non - local @xcite , are accessory nonlinearities for vortex nl - ubbs that do not contribute substantially to the their propagation invariance property , and that unlike dissipative vortex solitons , a balancing gain is not required .
experiments in ref .
@xcite demonstrated that a zero - order bb launched in a nonlinear medium reshapes spontaneously into a fundamental nl - ubb of the same cone angle as the input bb and of certain peak intensity .
here we show that a high - order bb transforms similarly into a vortex nl - ubb that preserves the cone angle and the topological charge .
we further find a third preserved quantity of the propagation that allows us to provide with a solution to the so - called selection problem , " or determination the specific ( vortex or vortex - less ) nl - ubb that is selected as the final stage of the nonlinear propagation of the input bb , a problem that remained open in ref .
accordingly , we consider a light beam @xmath3 $ ] of angular frequency @xmath4 and linear propagation constant @xmath5 , where @xmath6 is the linear refractive index and @xmath7 is the speed of light in vacuum . in the paraxial approximation
the change of the complex envelope @xmath8 along the propagation direction @xmath9 can be described by a nonlinear schrdinger equation ( nlse ) of the type @xmath10 where @xmath11 is the transversal laplace operator , @xmath12 are polar coordinates in the transversal plane , @xmath13 is the @xmath0-photon absorption coefficient , and dispersive nonlinearities as self - focusing or self - defocusing are included in the term with @xmath14 . in absence of all nonlinear terms , eq .
( [ nlse ] ) is satisfied by the high - order ( vortex ) bb @xmath15 with any @xmath16 , of cone angle @xmath17 and topological charge @xmath18 . for a more comprehensive analysis of the propagation of bbs in nonlinear media ,
we introduce dimensionless radius @xmath19 , propagation distance @xmath20 and envelope @xmath21 , with which eq .
( [ nlse ] ) rewrites as @xmath22 where now @xmath23 , and the vortex bb launched at the entrance plane @xmath24 of the medium rewrites as @xmath25 , where the parameter @xmath26 controls its intensity . for illustration purposes ,
we consider cubic and quintic dispersive nonlinearities , in which case @xmath27 in eq .
( [ nlse ] ) , where @xmath28 ( @xmath29 ) are the nonlinear refractive indexes , and @xmath30 in eq .
( [ nlsen ] ) , where @xmath31 .
( a ) nonlinear dynamics of an vortex bb @xmath32 with topological charge @xmath33 , amplitude @xmath34 in a medium with four - photon absorption @xmath35 and @xmath36 and @xmath37 .
the gray curves represent the intensity of the input vortex bb , and the dashed curves the intensity of the propagation invariant beam reached at large @xmath38 ( the vortex nl - ubb with @xmath39 , @xmath33 and @xmath35 ) .
( b ) nonlinear power losses per unit propagation length as a function of propagation distance.,title="fig:",width=234 ] ( a ) nonlinear dynamics of an vortex bb @xmath32 with topological charge @xmath33 , amplitude @xmath34 in a medium with four - photon absorption @xmath35 and @xmath36 and @xmath37 .
the gray curves represent the intensity of the input vortex bb , and the dashed curves the intensity of the propagation invariant beam reached at large @xmath38 ( the vortex nl - ubb with @xmath39 , @xmath33 and @xmath35 ) .
( b ) nonlinear power losses per unit propagation length as a function of propagation distance.,title="fig:",width=234 ] nonlinear dynamics of the vortex bb @xmath40 with topological charge @xmath33 , amplitude @xmath34 in media with four photon absorption @xmath35 and ( a ) @xmath36 and @xmath37 , ( b ) @xmath41 and @xmath42 , and ( c ) @xmath43 and @xmath44.,title="fig:",width=234 ] nonlinear dynamics of the vortex bb @xmath40 with topological charge @xmath33 , amplitude @xmath34 in media with four photon absorption @xmath35 and ( a ) @xmath36 and @xmath37 , ( b ) @xmath41 and @xmath42 , and ( c ) @xmath43 and @xmath44.,title="fig:",width=234 ] nonlinear dynamics of the vortex bb @xmath40 with topological charge @xmath33 , amplitude @xmath34 in media with four photon absorption @xmath35 and ( a ) @xmath36 and @xmath37 , ( b ) @xmath41 and @xmath42 , and ( c ) @xmath43 and @xmath44.,title="fig:",width=234 ] numerical solution of the nlse in eq .
( [ nlsen ] ) shows that high - order bbs are not completely depleted by nonlinear absorption , as would happen with a plane wave or with a gaussian beam , but they transform into new non - diffracting and non - attenuating beams .
figure [ fig1](a ) illustrates the transformation dynamics in the simplest situation of negligible dispersive nonlinearities ( @xmath45 ) . after an initial stage of sudden absorption ( provided
@xmath26 is large ) , the beam profile stabilizes into the new propagation - invariant beam , characterized by the same cone angle and topological charge as the input vortex bb , but with a reduced contrast of the annuli surrounding the vortex .
remarkably , the beam do not attenuate in spite that the nonlinear power losses per unit propagation length @xmath46 in the final propagation - invariant regime beam are constant and non - negligible , as seen in fig .
[ fig1](b ) . the radial profile , particularly the contrast of the annuli of the final propagation invariant vortex beam , and also its nonlinear power losses , strongly depend on the amplitude @xmath26 of the input bb .
the gradual transformation into the new vortex beam always starts at the beam center , and fills the cone @xmath47 at a propagation distance @xmath38 ( the cone @xmath48 in real - world variables ) , as can be appreciated in fig .
[ fig2](a ) .
similar dynamics is also observed when dispersive nonlinearities are included , e. g. , in fig . [ fig2](b ) for a saturable self - focusing nonlinearity , or in fig . [ fig2](c ) for a saturable self - defocusing nonlinearity .
the only significant difference is that the inner rings in the final stationary regime may be thinner or thicker , and that the outer rings are shifted radially to match the inner ones .
an analogous dynamics has been experimentally observed @xcite for the transformation of a zero - order bessel beam into a vortex - less nl - ubb .
the above simulations generalize those observations to vortex beams , and evidence the existence of non - diffracting beams that can transfer their energy and orbital angular momentum without any attenuation . according to these simulations
, these non - diffracting beams do not require specific dispersive nonlinearities balancing ( more or less stably ) diffraction spreading , as for vortex solitons in media with competing self - focusing and self - defocusing nonlinearities @xcite , or with non - local self - focusing nonlinearity @xcite . in sec .
[ stationary ] we investigate more in - depth on the properties of these vortex nl - ubbs .
this will allow us to complete in sec .
[ selection ] the description of the ( vortex or vortex - less ) bb nonlinear dynamics by fully specifying the particular propagation invariant nl - ubb that is formed for each particular amplitude @xmath26 of the input bb .
we above analysis suggests to search for solutions to eq .
( [ nlse ] ) of the form @xmath49 , or in dimensionless variables , solutions to eq .
( [ nlsen ] ) of the form @xmath50 , where @xmath51 is the sign of @xmath52 . according to eq .
( [ nlsen ] ) the amplitude @xmath53 and phase @xmath54 profiles must satisfy @xmath55 the vortex of charge @xmath56 at @xmath57 requires the amplitude to behave as @xmath58 as @xmath59 , and the localization condition requires @xmath60 as @xmath61 . in absence of nonlinear absorption , and depending on the specific dispersive nonlinearities , the above boundary problem with the positive sign ( @xmath62 ) has a discrete spectrum of solutions [ @xmath63 as @xmath59 with discrete values of @xmath64 , or vortex solitons . with the negative sign ( @xmath65 ) , the above problem has a continuous spectrum of solutions [ @xmath66 for any @xmath64 , or vortex nonlinear bbs . since eq .
( [ phas ] ) without the absorption term is satisfied by @xmath67 , only eq .
( [ amp ] ) is usually written explicitly @xcite .
the introduction of nonlinear absorption washes out the discrete spectrum in the above problem with the positive sign , i. e. , vortex solitons cease to exist in presence of nonlinear absorption .
the continuous spectrum for the negative sign , however , continues to exist , but limited to a maximum value of the constant @xmath7 , which constitute the continuous set of vortex nl - ubbs .
values of @xmath68 of nl - ubbs with different topological charge @xmath56 in media with different @xmath0 and with ( a ) @xmath69 , ( b ) @xmath70 and @xmath71.,title="fig:",width=158 ] values of @xmath68 of nl - ubbs with different topological charge @xmath56 in media with different @xmath0 and with ( a ) @xmath69 , ( b ) @xmath70 and @xmath71.,title="fig:",width=158 ] for @xmath69 , intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( a ) @xmath73 , ( b ) @xmath74 and ( c ) @xmath75 . for @xmath70 ,
@xmath71 intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( d ) @xmath76 , ( e ) @xmath77 and ( f ) @xmath78 .
the dotted curves represent linear vortex bbs with @xmath72 and the same @xmath26 as the vortex nl - ubbs.,title="fig:",width=154 ] for @xmath69 , intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( a ) @xmath73 , ( b ) @xmath74 and ( c ) @xmath75 . for @xmath70 , @xmath71 intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( d ) @xmath76 , ( e ) @xmath77 and ( f ) @xmath78 .
the dotted curves represent linear vortex bbs with @xmath72 and the same @xmath26 as the vortex nl - ubbs.,title="fig:",width=154 ] for @xmath69 , intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( a ) @xmath73 , ( b ) @xmath74 and ( c ) @xmath75 . for @xmath70 , @xmath71 intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( d ) @xmath76 , ( e ) @xmath77 and ( f ) @xmath78 .
the dotted curves represent linear vortex bbs with @xmath72 and the same @xmath26 as the vortex nl - ubbs.,title="fig:",width=154 ] for @xmath69 , intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( a ) @xmath73 , ( b ) @xmath74 and ( c ) @xmath75 . for @xmath70 , @xmath71 intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( d ) @xmath76 , ( e ) @xmath77 and ( f ) @xmath78 .
the dotted curves represent linear vortex bbs with @xmath72 and the same @xmath26 as the vortex nl - ubbs.,title="fig:",width=154 ] for @xmath69 , intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( a ) @xmath73 , ( b ) @xmath74 and ( c ) @xmath75 . for @xmath70 , @xmath71 intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( d ) @xmath76 , ( e ) @xmath77 and ( f ) @xmath78 .
the dotted curves represent linear vortex bbs with @xmath72 and the same @xmath26 as the vortex nl - ubbs.,title="fig:",width=154 ] for @xmath69 , intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( a ) @xmath73 , ( b ) @xmath74 and ( c ) @xmath75 .
for @xmath70 , @xmath71 intensity profiles of vortex nl - ubbs with @xmath72 , @xmath35 , and ( d ) @xmath76 , ( e ) @xmath77 and ( f ) @xmath78 .
the dotted curves represent linear vortex bbs with @xmath72 and the same @xmath26 as the vortex nl - ubbs.,title="fig:",width=154 ] for a physically meaningful classification of the vortex nll - ubbs , we first note that all nonlinear terms in eqs .
( [ amp ] ) and ( [ phas ] ) are negligible in the vortex core @xmath59 .
therefore vortex nl - ubbs behave as a linear vortex bb @xmath79 as @xmath59 . since @xmath80 as @xmath59 , the solution of the nonlinear problem in eqs .
( [ amp ] ) and ( [ phas ] ) satisfying the boundary condition @xmath81 represents the vortex nl - ubb whose vortex core is that of the vortex bb @xmath82 .
as said , nl - ubbs in nonlinearly absorbing are seen to exist ( satisfy the localization condition ) only for @xmath83 $ ] , where @xmath68 depends on @xmath56 , @xmath0 and the specific dispersive nonlinearities .
numerically evaluated values of @xmath68 are displayed in fig .
[ fig3 ] ( a ) and ( b ) as functions of @xmath0 for different values of @xmath56 , in absence of dispersive nonlinearities and in a particular self - focusing medium . a few examples of their intensity radial profiles @xmath84 are shown in fig .
[ fig4 ] ( solid curves ) and compared to the vortex bb with the same vortex core ( same @xmath26 ) . at low amplitude
( small @xmath26 ) , the nl - ubb behaves as the bb not only at the vortex core but at any radial distance .
when @xmath26 increases up to @xmath68 , the contrast of the rings is gradually lost , and the inner rings become compressed ( enlarged ) in self - focusing ( self - defocusing ) media . in all cases , the outer rings at @xmath85 decay in amplitude as @xmath86 and oscillate at the same radial frequency as those of the linear bb . in self - focusing or self - defocusing media , these outer rings
are therefore radially shifted with respect to those of the linear vortex bb so that they match the compressed or enlarged inner rings .
integration in @xmath87 in eq .
( [ phas ] ) leads to @xmath88 or @xmath89 for short .
this relation expresses that the nonlinear power losses in any circle of radius @xmath87 , @xmath90 , are refuelled by an inward power flux @xmath91 crossing its circumference and coming from a power reservoir at large radial distance .
this is substantially the mechanism of stationarity of nl - ubb , which is only possible in beams carrying infinite power as conical beams , and applies equally to vortex nl - ubbs .
a closer look reveals some differences . rewriting the nlse in eq .
( [ nlsen ] ) in terms of the amplitude and phase ( @xmath92 ) of a light beam ( not necessarily stationary ) leads to the continuity equation @xmath93 for the intensity @xmath94 , where the transversal current of the intensity is given by @xmath95 and where the nonlinear losses density @xmath96 act as a distributed sink .
the general condition of stationarity for the intensity is then @xmath97 expressing , by virtue of the divergence theorem , that the power losses in any finite region of the transversal plane is refuelled by an inward power flux through its contour . the above eq .
( [ phas ] ) is the expression of this fact for the particular case of stationarity for beams with radially symmetric intensity @xmath84 and phase @xmath98 , in which case @xmath99 and eq .
( [ refillinge ] ) is obtained by integrating eq .
( [ refillingdif ] ) over a circle of radius @xmath87 . for comparison ,
[ fig5 ] shows the intensity , phase and the intensity current for previously known stationary beams and for a vortex nl - ubb . for vortex
bbs ( and also for vortex solitons ) in transparent media , the intensity current @xmath100 is azimuthal and solenoidal ( @xmath101 ) . for the fundamental nl - ubbs ,
the transversal intensity current @xmath102 is radial inwards with a divergence that equals the nonlinear losses density @xmath96 . for vortex nl - ubbs
the intensity current has a solenoidal azimuthal component and a non - solenoidal radial inwards component , so that the power spirals inwards permanently from the power reservoir at infinity towards the inner rings , where most of nonlinear power losses take place .
unlike the fundamental ubb , the vortex nl - ubb carries an axial orbital angular momentum with a density @xcite @xmath103 that is proportional to the intensity and to the topological charge .
thus , as the intensity , the stationary angular momentum density is permanently flowing spirally with a current @xmath104 proportional to that of the intensity , and this current refills continuously the angular momentum losses @xmath105 associated to the nonlinear absorption process according to the continuity equation @xmath106 for the angular momentum density .
in particular , integration over a circle of radius @xmath87 leads to @xmath107 which expresses , by analogy with eq .
( [ refillinge ] ) , that the radial flux of angular momentum through any circle of radius @xmath87 refills the angular momentum losses within that circle , an angular momentum that is transferred to the material medium . a complete understanding of the reshaping of input high - order bbs into vortex nl - ubbs requires , as we will see in sec .
[ selection ] , a detailed analysis of the structure of vortex nl - ubbs at large radial distances .
vortex nl - ubbs are linear not only at the vortex center , but also at large radial distances . at the vortex core
the nl - ubb behaves as the linear bessel beam @xmath108 , or what is the same , @xmath109e^{il\varphi}e^{-i\zeta}\ , , \ ] ] where the high - order bb of amplitude @xmath26 is decomposed in a balanced " superposition of two high - order hnkel beams of equal amplitudes @xmath26 , the hnkel beam @xmath110 carrying power spirally outwards , and the hnkel beam @xmath111 carrying the same amount of power spirally outwards @xcite . as the fundamental nl - ubb @xcite
, a vortex nl - ubb behaves at large radius as the unbalanced high - order bessel beam ( ubb ) @xmath112e^{il\varphi}e^{-i\zeta}\ , , \ ] ] [ fig .
[ fig6](a),solid and dashed curves ] where the two interfering high - order hnkel beams have different amplitudes @xmath113 and @xmath114 [ fig .
[ fig6](a),gray dashed curves ] . using that @xmath115}$ ] at large @xmath87 @xcite , the condition that the inward radial flux @xmath116 $ ] matches at large @xmath87 the total nonlinear losses @xmath117 is readily seen to impose the constraint latexmath:[\[\label{refilling2 } @xmath114 and @xmath113 , from which it follows that @xmath119 for any vortex nl - ubb . also , the radial intensity at large @xmath87 is given by @xmath120\right\}\ ] ] with @xmath121 .
thus , @xmath122 represents harmonic oscillations of contrast @xmath123 about the mean value @xmath124 .
the values of @xmath125 and @xmath126 for a given vortex nl - ubb with @xmath26 can be easily extracted from the above properties and its numerically evaluated profile , and are represented as functions of @xmath26 in fig . [
fig6](b ) and ( c ) for nl - ubbs with different charges in different media .
it turns out that @xmath114 and @xmath113 are real in absence of dispersive nonlinearities , and @xmath127 with dispersive nonlinearities as self - focusing or self - defocusing .
( a ) amplitude of a nl - ubb ( @xmath35 , @xmath69 , @xmath128 and @xmath129 ) ( solid curve ) , its asymptotic linear ubb ( dashed curve ) , and its hnkel inward and outward components ( gray dashed curves ) .
( b , c ) values of @xmath126 and @xmath125 for nl - ubbs in the indicated values of @xmath130 , @xmath131 , @xmath56 and @xmath35 , as functions of their amplitude @xmath26.,title="fig:",width=234 ] ( a ) amplitude of a nl - ubb ( @xmath35 , @xmath69 , @xmath128 and @xmath129 ) ( solid curve ) , its asymptotic linear ubb ( dashed curve ) , and its hnkel inward and outward components ( gray dashed curves ) .
( b , c ) values of @xmath126 and @xmath125 for nl - ubbs in the indicated values of @xmath130 , @xmath131 , @xmath56 and @xmath35 , as functions of their amplitude @xmath26.,title="fig:",width=230 ] ( a ) amplitude of a nl - ubb ( @xmath35 , @xmath69 , @xmath128 and @xmath129 ) ( solid curve ) , its asymptotic linear ubb ( dashed curve ) , and its hnkel inward and outward components ( gray dashed curves ) .
( b , c ) values of @xmath126 and @xmath125 for nl - ubbs in the indicated values of @xmath130 , @xmath131 , @xmath56 and @xmath35 , as functions of their amplitude @xmath26.,title="fig:",width=230 ]
in sec . [ dynamics1 ] we showed that a high - order linear bb survives as a stationary beam in a nonlinear medium with nonlinear absorption by reaching spontaneously a new propagation invariant state in the form of a vortex nl - ubb of same cone angle and topological charge as the input bb , but in general of much lower intensity . with the notation introduced in sec .
[ stationary ] , the input bb specified by a certain value of @xmath26 , say @xmath132 , transforms at large @xmath38 into a nl - ubb specified by a different value of @xmath26 , say @xmath133 [ e. g. , the bb with @xmath134 transforms into the nl - ubb with @xmath135 in the simulation of fig .
[ fig1 ] ] . for a complete description of the dynamics ,
the question that remains unsolved is to determine which particular nl - ubb is selected as the final stationary state given the input bb , i. e. , the value of @xmath133 given the value of @xmath132 , and to understand why .
this problem arose in ref .
@xcite and remained unsolved in relation of vortex - less bbs and nl - ubbs , where the peak intensities of all input bbs and final nl - ubbs are normalized to unity , but their ratio is actually @xmath136 .
values of @xmath133 of the final vortex nl - ubb as a function of @xmath132 of the input vortex bb for @xmath137 and @xmath33 in a medium with @xmath35 , @xmath70 and @xmath138 , obtained by solving numerically the nlse in eq .
( [ nlsen ] ) .
the dashed curves represent the amplitudes @xmath139 and @xmath140 of the inward and outward hnkel components of the final vortex nl - ubb with the amplitude @xmath133 . as a function of @xmath132 , @xmath139
is found numerically to be the identity function , i. e. , @xmath141.,title="fig:",width=158 ] values of @xmath133 of the final vortex nl - ubb as a function of @xmath132 of the input vortex bb for @xmath137 and @xmath33 in a medium with @xmath35 , @xmath70 and @xmath138 , obtained by solving numerically the nlse in eq .
( [ nlsen ] ) .
the dashed curves represent the amplitudes @xmath139 and @xmath140 of the inward and outward hnkel components of the final vortex nl - ubb with the amplitude @xmath133 . as a function of @xmath132 , @xmath139
is found numerically to be the identity function , i. e. , @xmath141.,title="fig:",width=158 ] in a purely numerical approach , the values of @xmath133 for each @xmath132 can be found by solving numerically the nlse in eq .
( [ nlsen ] ) , as done in sec .
[ dynamics1 ] .
the pairs [ @xmath142 $ ] extracted from these simulations are displayed in fig .
[ fig7 ] for @xmath143 in a particular medium .
a first conclusion is that the final nl - ubb approaches that of maximum amplitude supported by the medium at very high intensities of the input bb , i. e. , @xmath144 when @xmath145 . in the opposite side ,
the selected nl - ubb does not differ substantially from the launched bb when the intensity is low enough so that nll are negligible , i. e. , @xmath146 when @xmath147 . at intermediate amplitudes , a certain nl - ubb with @xmath133 between @xmath148 and @xmath68
is selected .
the law of the nl - ubb selection is suggested by the dashed lines in fig .
they represent the amplitudes @xmath139 and @xmath140 of the asymptotic inward and outward hnkel components of the final nl - ubb .
it turns out from these numerical simulations that @xmath141 .
this result is seen to hold irrespective of the medium ( the values of @xmath0 and the dispersive nonlinearities ) and the topological charge @xmath56 . since for the input bb @xmath149 too , we infer the law that _ the amplitude of the asymptotic inward hnkel component is conserved in the nonlinear propagation , _ and only the amplitude of the outward hnkel component diminishes .
the inward hnkel beam is indeed a linear beam that continuously brings power from the energy and angular momentum reservoir at infinity , and therefore is not affected by the nonlinear effects that take place in the inner rings .
thus , given the value of @xmath132 of the input bb , the nl - ubb selected as the final stage of the dynamics is that with @xmath133 whose inward component equals @xmath132 , that is , @xmath150 figures such as figs.[fig6](b ) and ( c ) for the stationary nl - ubbs provide therefore a solution to the selection problem . given a medium ( @xmath0 and the dispersive nonlinearities ) and charge @xmath56 , @xmath133 is the value of the abscissa that intersects the curve @xmath125 at the ordinate value @xmath132 .
peak intensity of a zero - order bb launched in fused silica as a function of propagation distance , obtained by numerical solution of eq .
( [ nlse ] ) .
details of the input bb and of the optical properties of fused silica are given in the text.,width=264 ] as an example , and to express these results with physical magnitudes , we consider the input zero - order bb @xmath151 at @xmath152 nm carrier wave length ( @xmath153 fs@xmath154 ) and of cone angle @xmath155 in fused silica , considered as a pure cubic kerr medium with four - photon absorption ( @xmath156 , @xmath157 cm@xmath154 , @xmath159 @xmath160/w , @xmath35 , and @xmath161 @xmath162/w@xmath163 @xcite , so that @xmath164 , @xmath165 ) . for the input intensity
@xmath166 tw/@xmath160 , the dimensionless amplitude is @xmath167 .
the final stationary state is then the nl - ubb with dimensionless amplitude @xmath133 such that its inward hnkel component is @xmath168 .
unfortunately , there is no simple ( analytical ) relation between @xmath125 and @xmath26 for nl - ubbs , and a numerical analysis of the stationary states for the involved values of @xmath0 , @xmath56 , and @xmath169 is necessary .
this analysis is performed in fig .
[ fig6](c ) , showing that the nl - ubb with @xmath168 is that with @xmath170 , corresponding to a peak intensity @xmath171 w/@xmath160 . as a verification , fig .
[ fig8 ] depicts the peak intensity of the input bb as it propagates in fused silica as obtained from numerical solution of the nlse in eq .
( [ nlse ] ) . after a sudden diminution
, the peak intensity is seen to stabilize into the value predicted from the above theory .
we have reported on the existence of non - diffracting and non - attenuating vortex beams supported by nonlinear absorption .
these vortex nl - ubb form narrow and ( ideally ) infinitely long tubes of light where energy and orbital angular momentum can be transferred to matter , but the beam energy and angular momentum are continuously restored by spiral currents coming from a reservoir at large radial distances . in real settings , a vortex nl - ubb can be formed only up to a finite radial distance , say @xmath172 , by which the reservoir is of limited capacity .
it will then be depleted , and the refilling mechanism will cease , after a distance of the order of the diffraction - free distance @xmath173 , as described for other conical light beams @xcite . vortex nl - ubbs arise naturally in the propagation of linear bbs at intensities at which multiphoton absorption and possibly other nonlinearities are significant . in this sense , we have provided an accurate description of the nonlinear propagation of ( vortex or vortex - less ) bessel beams in nonlinear media in terms of their transformation into the nl - ubb that preserve the cone angle , the topological charge and the inward current coming from the reservoir .
m. a. p. acknowledges financial support from project mtm2012 - 39101-c02 - 01 of ministerio de economa y competitividad of spain .
99 l. allen , m. padgett , and m. babiker , the orbital angular momentum of light , " prog . opt . * 39 , * 291372 ( 1999 ) . j. w. r. tabosa and d. v. petrov , optical pumping of orbital angular momentum of light in cold cesium atoms , " phys . rev .
e * 83 , * 4967 - 4970 ( 1999 ) .
a. v. masalov , transfer of a high angular momentum from a photon to an atom , " laser physics * 7 , * 751754 ( 1997 ) .
l. allen , m. w. beijersbergen , r.j.c .
spreeuw , and j. p. woerdman
orbital angular momentum of light and the transformation of laguerre gaussian laser modes , " phys .
rev . a * 45 , * 81858189 ( 1992 ) .
j. durnin , j. j. miceli , and j. h. eberly j h , diffraction free beams , " phys .
* 58 * 1499501 ( 1987 ) .
k. volke - sepulveda , v. garcs - chvez , s. chvez - cerda , j. arlt , and k. dholakia , orbital angular momentum of a high - order bessel light beam , " j. opt .
b : quantum semiclass . opt .
* 4 , * s82-s89 ( 2002 ) .
a. desyatnikov , y. kivshar , and l. torner , optical vortices and vortex solitons , " prog
* 47 , * 291391 ( 2005 ) .
g. a. swartzlander , jr . , c. t. law ,
optical vortex solitons observed in kerr nonlinear media , " phys .
lett * 69 , * 2503 - 2506 ( 1992 ) .
b. a. malomed , l. c. crasovan , d. mihalache , stability of vortex solitons in the cubic - quintic model , " physica d * 161 , * 187201 ( 2002 ) .
i. yakimenko , y. a. zaliznyak , and y. kivshar , stable vortex solitons in nonlocal self - focusing nonlinear media , " phys .
e * 71 , * 065603(r ) ( 2005 ) .
l. c. crasovan , b. a. malomed , and d. mihalache , stable vortex solitons in the two - dimensional ginzburg - landau equation , " phys .
a * 63 , * 016605 ( 2000 ) .
o. v. borovkova , v. e. lobanov , y. v. kartashov , and l. torner , rotating vortex solitons supported by localized gain , " opt .
* 36 , * 1936 - 1938 ( 2011 ) .
j. m. soto - crespo , n. akhmediev , c. meja - corts , n. devine , dissipative ring solitons with vorticity , " opt . express * 17 , * 4236 - 4250 ( 2009 ) .
m. a. porras , a. parola , d. faccio , a. dubietis , and p. di trapani , nonlinear unbalanced bessel beams : stationary conical waves supported by nonlinear losses , " phys .
lett * 93 , * 153902 ( 2004 ) .
e. gaizauskas , a. dubietis , v. kudriasov , v. sirutkaitis , a. couairon , d. faccio , and p. di trapani , on the role of conical waves in self - focusing and filamentation of femtosecond pulses with nonlinear losses , " , in _ self - focusing : past and present .
fundamentals and prospects _ , part ii , topics in applied physics * 114 , * 457 - 479 ( 2009 ) .
p. polesana , d. faccio , p. di trapani , a. dubietis , a. piskarskas , a. couairon , and m. a. porras,high localization , focal depth and contrast by means of nonlinear bessel beams , " opt .
express * 13 , * 6160 - 6167 ( 2005 ) .
p. polesana , a. dubietis , m. a. porras , e. kucinskas , d. faccio , a. couairon , and p. di trapani , near - field dynamics of ultrashort pulsed bessel beams in media with kerr nonlinearity , " phys . rev . a * 73 , * 056612 ( 2006 )
. j. salo , j. fagerholm , a. t. friberg , and m. m. salomaa , unified description of nondiffracting x and y waves , " phys .
e * 62 , * 4261 - 4275 ( 2000 ) . i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series , and products _
( academic press , new york , 1965 ) . | we show that a high - order bessel beam propagating in a medium with nonlinear absorption is not completely absorbed , but survives in the form of a new propagation invariant vortex beam in which the beam energy and orbital angular momentum are permanently transferred to matter and at the same time refueled by spiral inward currents of energy and angular momentum . unlike vortex solitons and dissipative vortex solitons ,
these vortex beams are not supported by specific dispersive nonlinearities ( self - focusing or self - defocusing ) and do not require gain .
propagation invariance in presence of multiple absorption of photons carrying ( possibly high ) orbital angular momentum makes these beams attractive for optical pumping of angular momentum over long distances . |
it is widely believed that the distribution of galaxies is different from the overall distribution of dark matter ( dm ) .
this difference , _ the bias _ , is crucial for comparisons between observations and predictions of cosmological models .
observations provide information about the distribution of _ objects _ such as galaxies and galaxy clusters .
the models , however , most readily predict the evolution of the dark matter distribution , which can not be observed directly .
the models , therefore , should be able to predict the distribution of _ objects _ or , conversely , predict how this distribution is different from that of the dark matter ( i.e. , the bias ) . the notion of bias
was introduced by @xcite to explain the large difference in clustering strength between galaxies and abell clusters .
later @xcite and @xcite applied this argument to galaxies themselves .
davis et al .
( 1985 ) showed that the cdm model disagreed with observations , if galaxies were distributed like dark matter . however ,
if a biased galaxy formation scenario was assumed , the `` galaxy '' correlations substantially exceeded the correlations of mass at all scales and agreed with observations for certain values of bias .
this work was followed by other studies which tried to account for the phenomenon of galaxy bias ( @xcite ; @xcite ; @xcite ; @xcite ) . the bias can be defined and understood differently . in this paper
we will use the conventional statistical definition of the bias as the ratio of the correlation functions of objects and dark matter : @xmath17 here , @xmath18 is the square of the _ bias function _ , @xmath19 and @xmath20 are the 2-point spatial correlation functions of objects and dark matter , respectively .
dependencies in the above equation indicate that in general the bias may depend on the epoch @xmath16 , scale @xmath21 , and properties of the objects such as their mass @xmath22 .
it is also expected that the functional form @xmath23 depends on the cosmological model .
the bias in the above definition is most closely related to the observations because the _ observed _ @xmath19 can be compared to the @xmath20 _ predicted _ within a framework of a given cosmological model .
this gives an estimate of _ the amount of bias needed _ for a cosmological model to agree with observations .
while it is clear that this approach can not be used to test a model , it provides some insights into the nature of bias and its evolution .
for example , steidel et al . (
1998 ) used the observed clustering strength of high - redshift ( @xmath24 ) lyman - break galaxies to derive the implied value of bias in different cosmological models .
this value was found to be quite large in all models ( @xmath25 ) . on the other hand ,
a number of theoretical studies ( e.g. , klypin , primack & holtzman 1996 ; cole et al .
1997 ; jenkins et al .
1998 ; and references therein ) estimated in different cosmological models at @xmath12 and made comparisons with accurate measurements of from local galaxy surveys .
these studies indicate that significant _ anti - bias _ ( @xmath26 ) is required for the open and flat low- models at scales @xmath27 , while @xmath28 cluster - normalized models indicate positive bias ( @xmath29 ) .
comparison of the low- and high-@xmath16 results implies that , regardless of cosmological model , the bias has decreased significantly from the early epochs to the present .
the distribution of the dark matter and @xmath20 can not be observed directly .
therefore , a test of a cosmological model is possible only if the model can predict the @xmath30 of observed objects .
unfortunately , this is not an easy and straightforward task .
first of all , we should fully understand what are the observed objects and where / how they form .
the standard lore is that observed galaxies form dissipatively inside dark matter halos .
the properties of galaxies will then depend on the mass of the parent halo , its spin , details of dissipative processes and mass accretion history , and other factors ( e.g. , mo , mao & white 1998 ) .
therefore , in order to predict the type of galaxy and its properties , the relevant processes must be included in the model .
however , it seems likely that in _ every _ sufficiently massive ( @xmath31 ) _ gravitationally bound _
halo baryons will cool , form stars , and produce an object ressembling a galaxy ( e.g. , kauffman , nusser & steinmetz 1997 ; roukema et al .
1997 ; yepes et al . 1997 ; @xcite ) .
the galaxy population _ as a whole _ can be then viewed as a population of _ galaxy - size _ dark matter halos .
the clustering of the latter can be studied without inclusion of complicated physics ; it has been modelled using both direct numerical simulations and analytical methods .
typical ingredients of the analytical models ( e.g. , matarrese et al .
1997 ; moscardini et al . 1998 ; @xcite ) are the extended press - schechter formalism ( @xcite ; @xcite ) used to follow mass evolution of halos , analytical approximations to the non - linear clustering evolution of the dark matter ( e.g. , hamilton et al .
1991 ; peacock & dodds 1994,1996 ; jain , mo & white 1995 ; smith et al .
1998 ) , and a model for bias evolution ( e.g. , mo & white 1996 ; matarrese et al . 1997 ) .
the evolution of bias in such models can be calculated quickly which makes extensive parameter - space studies possible .
the main disadvantages are large uncertainties ( especially at small , @xmath32 , scales ) introduced by bias prescriptions and limited applicability of the non - linear clustering approximations .
direct numerical simulations should , ideally , predict halo - halo clustering without any additional assumptions and uncertainties .
however , until very recently the predictions of numerical simulations were also quite uncertain .
the main reason for the uncertainty was that dissipationless @xmath0-body simulations had been consistently failing to produce galaxy - size dark matter halos in dense environments typical for galaxy groups and clusters .
recently , it was shown that this effect , known as `` the overmerging problem '' ( @xcite ; summers , davis , & evrard 1995 ) , is due mainly to the insufficient force and mass resolution of such simulations ( moore , katz & lake 1996 ; @xcite , hereafter kgkk ; @xcite ) .
the lack of sufficient resolution leads to artificial disruption of halos in clusters .
this , in turn , leads to a strong artificial anti - bias ( especially at small scales @xmath33 , but larger scales are also affected ) .
there are several ways to deal with this problem .
one possible way is to break up massive structureless halos into subhalos using some kind of observationally motivated prescription ( @xcite ; klypin , nolthenius , & primack 1997 ) .
these subhalos can then be included into halo catalogs used to compute the correlation function and other halo statistics .
another common approach is to overcome overmerging by weighting the massive halos according to their mass and compute thus a _ weighted _ correlation function ( @xcite ) .
both of these approaches are useful . nevertheless , it is not clear whether the unavoidable heuristic assumptions take the processes driving the small - scale bias evolution correctly into account .
the weighting technique , for example , ignores the real physical effects in groups and clusters such as tidal stripping and dynamical friction .
we will argue below that these effects are likely to be driving the small - scale bias evolution at low ( @xmath34 ) redshifts .
hydrodynamic simulations that include gas cooling are affected by overmerging to a significantly lesser degree ( e.g. , summers et al .
1995 ; @xcite ) .
the cooling creates compact dense objects inside halos which can survive in clusters .
these simulations , therefore , can be used to study the halo clustering directly .
unfortunately , the computational cost required to simulate a large volume with sufficiently high mass resolution is prohibitively high .
moreover , such simulations usually oversimplify the gas dynamic by including only cooling mechanism .
this leads to `` overcooling '' : without a heating process to regulate it , the cooling produces very compact and dense baryonic blobs in the halo centers .
these blobs do survive successfully in clusters , but they also suffer much less from the tidal stripping of material as compared to a realistic galaxy with a more extended distribution of baryons
. the mass of the objects may thus be higher than it should be which may lead to excessive dynamical friction and thus incorrect dynamics of halos .
the very high resolution @xmath0-body simulations are thus a viable alternative , if the required resolution can be reached at an affordable computational cost . analytical arguments and numerical experiments ( moore et al .
1996 ; kgkk ) indicate that the required force and mass resolution are @xmath35 and @xmath36 , respectively .
this force resolution is sufficiently high for most halos to survive in high - density regions .
the mass resolution is determined by the requirement that a galaxy - size halo should have @xmath37 particles to avoid relaxation effects and to assure a robust identification by a halo finding algorithm .
the numerical simulations that reach such resolution ( ghigna et al . 1998 ; kgkk ) show that most halos do survive even in the richest clusters . the dynamic range required to reach
the needed resolution in a statistically large volume , @xmath38 , is quite high : @xmath39 .
nevertheless , the advances in computer hardware and in the numerical algorithms make it now possible to carry out such simulations at an affordable computational cost . in this paper
we use such high dynamic range simulations to calculate the 2-point correlation function of halos and corresponding bias using _ all _ halos in the simulated volume : i.e. , both isolated dark matter halos and satellites of massive halos , and halos inside group- and cluster - size systems .
the simulations of different cosmological models and box sizes were made using the adaptive refinement tree ( art ; kravtsov , klypin & khokhlov 1997 ) and the ap@xmath40 m ( couchman 1991 ) @xmath0-body codes .
the absence of the overmerging ) of clusters are affected by the overmerging , even with resolution this high , is still a matter of debate .
ghigna et al .
( 1998 ) , for example , argue that these regions are still completely overmerged .
they find , however , that halos survive at smaller radii ( @xmath41 kpc ) in their @xmath42 kpc resolution run1 , as compared to the @xmath43 kpc run2 .
the resolution of the simulations presented in this paper is comparable to the run1 .
we do find halos within central @xmath44 ( see 4.3 and figs .
although some halos do survive at @xmath45 from cluster center , our tests show that these regions may be affected by the overmerging .
the larger scales @xmath46 , are not affected ] in these simulations means that the additional steps such as breaking - up of clusters , or mass - weighting ( see above ) are not necessary .
therefore , the correlation function of halos is measured _ directly _ down to unprecedentedly small scales ( @xmath47 ) without the usual uncertainties associated with these steps . for the first time this opens the possibility to study the effects of dynamical processes such as tidal destruction and dynamical friction on the evolution of small - scale bias .
the results on the evolution of bias and on the halo correlation function can be used as a basis for comparisons and interpretations of the existing and upcoming observations , as well as a check and/or input for the analytic models of clustering evolution .
the paper is organized as follows . in
2 we briefly review the definitions of bias and current analytical models of its evolution .
the cosmological models studied in this paper are described in 3 .
the details of the numerical simulations and discussion of the construction of halo catalogs and halo survival in the high - density regions is given in
4 . in 5
we present our results on the evolution of halo clustering and bias in different cosmological models .
we also present a comparison of our @xmath12 results with the galaxy correlation function measured using the apm galaxy survey .
a discussion of the main results is presented in 6 .
we summarize our main results and conclusions in 7 .
the notion of bias is more complicated than eq . ( 1 ) might suggest .
formally , the bias is defined as a function relating fluctuations in the dark matter density , @xmath48 , to the fluctuations in the number density of objects , @xmath49 . in general , this relationship can be complicated and may depend on a large number of factors : scale , mass and/or type of objects , time , etc .
the large number of the dependencies or discreteness of the object distribution can result in the _ stochasticity _ of the bias ( dekel & lahav 1998 ) : a scatter in the relationship between @xmath49 and @xmath48 .
finally , the bias can be _ non - local _ , if probability to form an object at a given point is not fully determined by local factors .
for example , in addition to the local density , temperature , etc . , the probability to form a galaxy may depend on environment .
analytical results of catelan et al .
( 1998b ) indicate that the bias is expected to be non - local even in the linear regime .
the lack of general understanding of these dependencies of the bias usually results in the use of the simplest assumptions .
the most common approach is to assume that bias is _ local _ , depends only on the local matter density , and is _ linear _ : @xmath50 .
an obvious consequence of the latter assumption is that bias is scale - independent . in this case , the correlation function and power spectrum of halos and dm are simply related by a constant scaling factor . although the linearity of bias would greatly simplify the theoretical interpretation of the clustering data , it is likely that bias depends on a variety of processes that may lead to nonlinearity .
these processes include merging , tidal disruption ( e.g. , dubinski 1998 ) , suppression of galaxy formation in small halos due to supernovae feedback ( dekel & silk 1986 ; yepes et al .
1997 ) , etc . the observations , in fact ,
indicate that at small scales galactic bias _ is _ nonlinear .
the correlation functions of different types of galaxies differ in amplitude and shape suggesting that at least some of the galaxies are nonlinearly biased . during the last years
, a significant progress has been made in the analytical modelling of the bias and its evolution ( e.g. , coles 1993 ; fry 1996 ; mo & white 1996 , hereafter mw ; matarrese et al .
1997 ; mann et al . 1997 ; catelan et al .
1998a , b ) .
although the current analytical models may have some drawbacks and limitations , they provide an insight and interpretation for numerical simulations . at this point
it is also important to check how well the prediction of the analytical models agree with the results of simulations ( see , e.g. , mw ; mo , jing & white 1996 ; jing 1998 ) .
we will therefore review briefly some results of the analytical models concerning evolution of the bias . in a seminal paper , kaiser ( 1984 )
showed that if the observed systems in the universe ( such as galaxies and galaxy clusters ) form in the _ peaks _ of the density field , their distribution is biased .
this is an example of _ statistical _ bias : the bias determined simply by the fact that objects do not sample the distribution of matter but that of the peaks , the latter having a statistically different distribution from the former .
thus , the bias is introduced from the start by the way the objects form . however , the bias changes subsequently as the distribution evolves driven by the gravitation . in hierarchical models
this evolution results in growth of objects via multiple mergers .
however , if the merger rate is low , the evolution of bias can be modeled by the _ object - conserving model_. in this model the objects form with an initial statistical bias and after that are dragged without merging by a gravitational pull from the surrounding density fluctuations ( dekel & rees 1987 ; nusser & davis 1994 ; fry 1996 ) .
mw used the extended press - schechter formalism to derive an expression for the bias in lagrangian coordinates ( comoving radius @xmath51 of the region from which halos form or halo mass @xmath22 ) . at linear scales , this expression or @xmath52 , as long as the scale @xmath21 is larger than the lagrangian radius @xmath51 .
they find that at these scales the correlation function of halos and the bias in their numerical simulations are well described by equation ( 2 ) ( see , however , catelan et al .
1998a , and jing 1998 ) . ] for the bias at redshift @xmath16 for dm halos of mass @xmath22 formed at redshift @xmath53 is ( matarrese et al . 1997 ) : @xmath54 here @xmath55 , @xmath56 is the growing mode of linear perturbations normalized to unity at @xmath12 , @xmath57 , @xmath58 is the critical overdensity for spherical collapse at @xmath12 , and @xmath59 is the rms linear mass fluctuation on the scale @xmath22 of halos linearly extrapolated to redshift @xmath16 .
this expression is similar to that obtained in earlier studies from a peak - background split argument ( e.g. , cole & kaiser 1989 ) .
the standard interpretation of the press - schechter description of the hierarchical evolution is that at any epoch @xmath16 all halos merge immediately to form more massive halos .
thus , if observed objects are identified with host halos at any epoch , then @xmath60 in eq .
matarrese et al .
( 1997 ) call this _ the merging model_. the object - conserving and merging models are two extremes pictures of clustering evolution , although they may be applicable to the evolution of galaxy clustering at certain epochs . at some epochs the halos may neither survive nor merge instantly . moreover , other processes such as halo dynamics in galaxy clusters and groups may become important when clustering reaches highly nonlinear stages . the wealth of potentially important processes may make a one - to - one identification between galaxies and dm halos very difficult ( see , for example , discussion in moscardini et al .
1998 ) .
( 2 ) gives an estimate of the bias for objects of a single mass .
the bias of a sample of objects ( `` effective '' bias ) with a range of masses @xmath61 should be calculated as a weighted _ average _ over the mass distribution of objects @xmath62 : @xmath63 where @xmath62 is given by the press - schechter ( 1974 ) distribution and @xmath64 is the mean number density of objects with masses @xmath61 at redshift @xmath16 .
moscardini et al .
( 1998 ) give a useful fitting formula for @xmath65 : @xmath66/d_+(z)^{\beta},\ ] ] and provide best - fit parameters @xmath67 and @xmath68 for a variety of different cosmological models and values of @xmath69 .
the generic feature of the evolution of the effective bias described by eq .
( 4 ) is its rapid decrease with decreasing redshift and increase with increasing @xmath69 . for galaxy - size halos , @xmath70 and @xmath71 for @xmath72 .
this gives @xmath73 at @xmath74 and @xmath75 at the present epoch .
we will use eq .
( 4 ) in 6 to interpret and compare our results with predictions of the analytical models discussed above .
lcccccl scdm & 1.0 & 0.0 & 0.50 & 13.1 & 1.1 & efstathiou et al .
ocdm & 0.3 & 0.0 & 0.65 & 12.2 & 0.9 & bbks+sugiyama3cdm@xmath76 & 0.3 & 0.7 & 0.70 & 13.4 & 1.17 & bbks+sugiyama3cdm@xmath77 & 0.3 & 0.7 & 0.70 & 13.4 & 1.0 & klypin & holtzman & 1.0 & 0.0 & 0.5 & 13.1 & 1.0 & efstathiou et al .
we have chosen to study the evolution of bias in four representative variants of the cdm models ( see details in table 1 ) : 1 ) the standard _
cobe_-normalized cdm model ( scdm ) ; 2 ) a variant of the @xmath78 cdm model ( ) with a different shape of the power spectrum ( the shape parameter @xmath79 ) ; 3 ) a flat low - density model with @xmath80 ( ) ; 4 ) an open model with @xmath81 ( ocdm ) .
the observations of the galaxy clustering indicate that the power spectrum of the galaxy distribution has a shape different from that of the standard cdm model : @xmath82 instead of @xmath83 of the scdm ( e.g. , maddox , efstathiou , & sutherland 1996 ; peacock & dodds 1994 ) .
this has motivated jenkins et al .
( 1998 ) to study the model , in which @xmath84 indicates the fact that lower values of @xmath85 may be obtained with a late decay of the massive @xmath84-neutrino .
the physics ( additional to the physics of the scdm model ) responsible for the change of the shape is , in fact , irrelevant for study of structure formation .
although the model in that respect is somewhat heuristic , it is interesting as an example of a model with @xmath86 and an approximately correct shape of the power spectrum .
the observations of the galaxy cluster evolution ( eke et al .
1998 ) and of the baryon fraction in clusters ( @xcite ) strongly indicate value of matter density @xmath87 , while various observational measurements of the hubble constant ( e.g. , kim et al . 1997 ; falco et al . 1997 ; salaris & cassisi 1998 ) tend to converge on the values of @xmath88 .
therefore , we have considered two models ( open and flat ) with @xmath89 and @xmath90 .
the age of the universe in all of our models is given in table 1 and is in good agreement with the ages of the oldest globular clusters ( chaboyer 1998 ) .
we have used different approximations for the power spectrum of density fluctuations for different considered models . for the ocdm and @xmath76 models ( see table 1 ) we used the bardeen et al .
( 1986 ; bbks ) fit for the power spectrum @xmath91 with corrections of @xcite : @xmath92^{-1/4}\end{aligned}\ ] ] and @xmath93^{-1}\ ] ] where @xmath94 ( @xcite we used a slightly higher value of @xmath95 . ] ) .
the power spectra for both the scdm and the models were approximated by the fitting formula of @xcite @xmath96^{1.13 } \right)^{2/1.13}}\ ] ] where @xmath97 , @xmath98 , @xmath99 , and @xmath100 is the normalization constant .
the shape parameter @xmath85 is @xmath101 and @xmath102 for the scdm and models , respectively .
these two analytic fits provide fairly good approximations to the power spectra of these models in the limit @xmath103 .
for the @xmath77 models we have used an approximation to the power spectrum different from that of the @xmath76 model . the approximation @xmath104 where @xmath105 , @xmath106 , @xmath107 , @xmath108 , and @xmath109 , is given by klypin & holtzman ( 1997 ) and was obtained by a direct fit to the power spectrum estimated using a boltzmann code .
the accuracy of this approximation is @xmath110 ( see klypin & holtzman 1997 for details ) .
a small shift in the normalization makes the approximations for the @xmath111 and @xmath112 models very similar in the range of wavenumbers probed in our simulations ( @xmath113 ) .
this shift can be expressed using the value of 8 ( amplitude of fluctuations on @xmath114 scale ) : @xmath115 for the bbks approximation versus @xmath116 for the holtzman approximation . with these values of 8 the differences in power
are negligible at large scales ( @xmath117 ) and are within 10% at the smaller scales .
our scdm model was normalized to the two - year _
cobe_-dmr data , @xmath118 ( a somewhat higher value , @xmath119 , is obtained from the four - year cobe - dmr data , see e.g. , bunn & white 1997 ) .
this normalization is inconsistent with normalization , @xmath120 , deduced from the observed cluster abundances ( e.g. , @xcite ) , which reflects the well - known failure of this model to account for both _
cobe _ and cluster data . our normalization of the ocdm is higher than that implied by the four - year _ cobe_-dmr data ( @xmath121 ) , but is consistent with the normalization , @xmath122 , implied by the cluster abundances .
the normalization of our @xmath123cdm models , on the other hand , is in good agreement with both cluster and the 4-year _ cobe _ data .
as explained above , the small difference in normalization between the @xmath111 and @xmath112 was introduced to minimize differences between the two power spectrum approximations used in these models .
normalization of the model is inconsistent with both the _ cobe _ normalization ( @xmath124 ) and with cluster abundance normalization ( @xmath125 ) .
this is motivated by our intent to use as a toy - model rather than a reasonable approximation to the real universe .
the normalization of the is similar to normalizations of the cdm and models .
therefore , comparison between results of these models allows us to study effects of changing the shape of the power spectrum and value of @xmath126 .
lllccrcccc 3 m & scdm & scdm & 49 & @xmath127 & 8000 & 3.0 & 30 & @xmath128 3 m & ocdm & ocdm & 109 & @xmath129 & 7000 & 4.7 & 30 & @xmath1283 m & 3cdm@xmath76 & & 64 & @xmath129 & 4000 & 3.0 & 30 & @xmath128art & 3cdm@xmath77 & & 30 & @xmath129 & 41300 & 1.8 & 60 & @xmath130art & 3cdm@xmath77&30art & 45 & @xmath131 & 13800 & 0.9 & 30 & @xmath130 3 m & & & 50 & @xmath132 & 8000 & 3.0 & 30 & @xmath128 we have used two different @xmath0-body codes to carry out our simulations : the adaptive refinement tree code ( art ; kravtsov et al .
1997 ) and the ap@xmath40 m code models . ]
( @xcite ) .
comparison of the results obtained with different numerical codes allows us to insure the results are robust .
ap@xmath40 m code is an extension of the well - known p@xmath40 m algorithm ( hockney & eastwood 1981 ) .
the code performs hierarchical rectangular refinements in the high - density regions to reduce expensive particle - particle calculations .
the gravitational force is obtained by matching the gravitational forces calculated using fft solver on the base and refinement grids and the small - scale force calculated using direct particle - particle summation ( see couchman 1991 for details ) .
a total of four refinement levels were allowed during the course of the 3 m simulations presented here .
the art code also reaches high force resolution by refining all high - density regions with an automated refinement algorithm .
the refinements are recursive : the refined regions can also be refined , each subsequent refinement having half of the previous level s cell size .
this creates an hierarchy of refinement meshes of different resolution covering regions of interest .
the refinement is done cell - by - cell ( individual cells can be refined or de - refined ) and meshes are not constrained to have a rectangular ( or any other ) shape .
this allows us to refine the required regions in an efficient manner .
the criterion for refinement is _ local overdensity _ of particles : in the simulations presented in this paper the code refined an individual cell only if the density of particles ( smoothed with the cloud - in - cell scheme ; hockney & eastwood 1981 ) was higher than @xmath133 particles .
therefore , _ all _ regions with overdensity higher than @xmath134 , where @xmath135 is the average number density of particles in the cube , were refined to the refinement level @xmath136 .
for the two art simulations presented here , and 30art , @xmath135 is @xmath137 and @xmath138 , respectively .
the poisson equation on the hierarchy of meshes is solved first on the base grid and then on the subsequent refinement levels .
on each refinement level the code obtaines potential by solving a dirichlet boundary problem with boundary conditions provided by the already existing solution at the previous level .
there is no particle - particle summation in the art code .
the detailed description of the code is given in kravtsov et al .
note , however , that the present version of the code uses multiple time steps on different refinement levels , as opposed to the constant time stepping in the original version of the code .
the multiple time stepping scheme is described in some detail in kravtsov et al .
( 1998 ; also see below ) .
the information about the numerical parameters of the simulations is given in table 2 .
the ap@xmath40 m code was used to produce four simulations of different cosmological models . ] with the box size @xmath139 .
the size of the box side is a compromise between requirements of the high spatial resolution ( @xmath140 ) and good statistics of halos .
nevertheless , for our most realistic model ( @xmath123cdm ) we also use the art code to simulate a @xmath141 box ( run ) .
we estimate effects of the finite box size ( see 5.1 ) on our results by comparing results of @xmath142 box simulation of @xcite with results of our @xmath141 and @xmath143 boxes .
all 3 m runs were done with @xmath128 particles .
the two art runs used @xmath130 particles .
the initial conditions were set using the zeldovich approximation on uniform 128@xmath40 , @xmath130 , and @xmath144 meshes for 3 m runs , 30art run , and run , respectively .
the seed used to generate the gaussian random density field was the same in all of our 3 m runs , but different for each of the two art runs .
all of the simulations are started at the moment of time when the rms density fluctuations at the nyquist wavelength @xmath145 are still linear : @xmath146 .
all 3 m runs evolved during a period in which the linear growth factor increased by a factor of 50 .
this explains different values of @xmath147 for different models in table 2 .
as was explained in 1 , the purpose of our study was to compute the correlation function and the bias accounting for _ all _ dm halos , including those inside groups and clusters . to assure that halos do survive in clusters the force resolution
should be @xmath148 ( moore et al .
1996 ; kgkk ) .
furthermore , if we aim to study galaxy - size halos , the mass resolution should be @xmath36 to resolve galaxy - size halos ( @xmath149 ) with at least @xmath150 particles .
the compromise between these considerations and the computational expense determined the force and mass resolution of our simulations ( see table 2 ) .
the art code integrates the equations of motion in _ comoving _ coordinates .
however , the refinement strategy of the art code is designed to effectively preserve the initial _ physical _ resolution of the simulation ( see below ) .
the peak resolution is reached by creating a refinement hierarchy with six levels of refinement . in the 3 m
runs the force resolution @xmath151 ( spline softening length ) was kept constant in comoving coordinates while fluctuations are still in the linear regime and is then set to be constant in _ physical _ units .
the switch occurs at the moment when first galaxy - size halos start to collapse ( @xmath152 ) for our simulations .
we chose to maintain fixed comoving resolution until it reaches @xmath153 ( physical ) ( at @xmath154 ) . at later moments
the resolution is fixed to this value in physical coordinates ( the exception is the ocdm model in which the resolution was set to @xmath155 by mistake ) .
the dynamical range @xmath156 of the simulations implied by the force resolution is @xmath157 ( @xmath158 formal ) for the art runs and @xmath159 for the 3 m runs . the dynamic range of the 3 m runs is just enough to keep the initial physical resolution ( @xmath160 ) .
the art code integrates the evolution in comoving coordinates .
therefore , in order to prevent degradation of force resolution in _ physical _ coordinates , the dynamic range between the start and the end ( @xmath12 ) of the simulation should increase by @xmath161 : i.e. , for our simulations reach @xmath162 .
this is accomplished with the prompt successive refinements during the simulations .
the time stepping of the 3 m and art codes is rather different .
first of all , the codes integrate the equation of motion using different time variables : the time in the 3 m code and the expansion factor in the art code .
in the 3 m runs the time step is _ constant _ and is _ the same _ for all particles . in the art runs , as was noted above , the particles residing on different refinement levels move with different time steps .
the particles on the same level , however , move with the same step .
the refinement of the time integration mimicks spatial refinement and step for each subsequent refinement level is two times smaller than the step on the previous level .
the global time step hierarchy is thus set by the step @xmath163 at the zeroth level ( uniform base grid ) .
the step on level @xmath136 is then @xmath164 .
the choice of an appropriate time step for a simulation is dictated by the adopted force resolution .
the number of time steps in our simulations is such that the _ rms _ displacement of particles during a single time - step is always less than @xmath165 ( less than 1/4 of a cell in the art code ) , especially at late times and in the run . in the art code particles
do not move further than @xmath166 cells in a single time step , where the cell size and time step for particles located on the refinement level @xmath136 are @xmath167 and @xmath168 , respectively . ] .
the size of the time - step , @xmath169 , for the 3 m runs was chosen to be sufficiently small to satisfy the stability criteria of the numerical integration ( @xcite ) throughout the entire run . in the case of art runs
, the value of @xmath170 was determined in a convergence study using a set of smaller @xmath171 particle simulations described in kravtsov et al .
( 1998 ) . in both 3 m and art
runs the energy was conserved with an accuracy @xmath172 .
llcccccc scdm & 180 & @xmath173 & 13 ocdm & 180 - 400 & @xmath174 & 13 & 180 - 340 & @xmath174 & 13 & 180 - 340 & @xmath175 & 20 30art & 180 - 340 & @xmath176 & 10 & 180 & @xmath173 & 13 identification of dm halos in the very high - density environments ( e.g. , inside groups and clusters ) is a challenging problem .
traditional halo finding algorithms , such as friends - of - friends ( e.g. , davis et al .
1985 ) or `` overdensity-200 '' ( e.g. , lacey & cole 1994 ) , can not be used .
these algorithms are not designed to search for substructure ; they identify an isolated halo above virial overdensity as a single object and can not account for the internal substructure .
our goal , however , is to identify both isolated halos and halos orbiting within larger systems ( `` sub - halos '' ) .
the problems associated with halo identification within high - density regions are discussed in kgkk . in this study
we use a halo finding algorithm called bound density maxima ( bdm ; see kgkk ) . a detailed description of the working version of the bdm algorithm used here can be found in klypin & holtzman ( 1997 ) .
other recently developed algorithms capable of identifying satellite halos are described in ghigna et al .
( 1998 ) and kgkk .
the main idea of the bdm algorithm is to find positions of local maxima in the density field smoothed at a certain scale and to apply physically motivated criteria to test whether the identified site corresponds to a gravitationally bound halo . in the following we describe specific parameters of the bdm used to construct halo catalogs used in our study ( the main parameters are listed in table 3 ) .
the radius of a halo assigned to it by the algorithm is either its virial radius is set in accord with the prediction of the top - hat collapse model .
note that @xmath177 depends on the cosmological model ( e.g. , kitayama & suto 1996 ) . ] or @xmath178 , whichever is smaller .
the latter is approximately the maximum virial radius we would expect for a galaxy - size halo .
the mass and radius are very poorly defined for the satellite halos due to the tidal stripping which alters a halo s mass and physical extent .
therefore , in this study we will use maximum circular velocity @xmath179 as a proxy for halo mass .
this allows us to avoid complications related to the mass and radius determination for satellite halos . for isolated halos @xmath179 and the halo s virial mass are directly related .
for example , for a halo with a density distribution described by the navarro , frenk & white ( hereafter nfw ; 1996 ) profile @xmath180 ( @xmath181 ; @xmath182 is scale - radius ) : @xmath183 where @xmath184 and @xmath185 are virial mass and radius , @xmath186 , @xmath187 .
while for the sub - halos @xmath179 may not be related to mass in any obvious way , it is still the most physically and observationally motivated halo quantity .
the limiting radius of @xmath178 is sufficient to determine the @xmath179 for galaxy - size halos .
the cluster - size halos are not explicitly excluded from the halo catalogs .
we assume therefore that the center of each cluster can be associated with a central cluster galaxy . the latter ( due to the lack of hydrodynamics and other relevant processes )
can not be identified in our simulations in any other way .
the density maxima are identified using a top - hat filter with radius @xmath188 ( `` search radius '' ) .
the search is performed starting from a large number of randomly placed positions ( `` seeds '' ) and proceeds by moving the center of mass within a sphere of radius @xmath188 iteratively until convergence . in order to make sure that we use a sufficiently large number of seeds
, we used the position of every tenth particle as a seed .
therefore , the number of seeds by far exceeds the number of expected halos .
the search radius @xmath188 also defines the minimum allowed distance between two halos . if the distance between centers of any of the two halos is @xmath189 , only one halo ( the more massive of the two ) is left in the catalog .
a typical value for the search radius is @xmath190 .
we set a lower limit for the mass inside the search radius @xmath191 : halos with @xmath192 are not included in the catalog .
this is done to exclude pure poisson fluctuations from the list of halo candidates .
some halos may have significant substructure in their cores due , for example , to an incomplete merger .
such cases appear in the catalogs as multiple ( 2 - 3 ) halos with very similar properties ( mass , velocity , radius ) at small separations .
our strategy is to count these as a single halo .
specific criteria used to identify such cases are : ( 1 ) distance between halo centers is @xmath193 , ( 2 ) their relative velocity in units of the rms velocity of particles in the halos @xmath194 is less than 0.15 , and ( 3 ) the difference in mass is less than factor 1.5 .
only the most massive halo is kept in the catalog .
it is obvious that for a statistical study it is important to be confident that our bdm algorithm does not miss a large fraction of halos . while a substantial effort is made to reject fake halos , it is important to make sure that all real halos are included in the catalog .
the algorithm is very efficient at finding isolated halos and the major difficulties are in identification of halos in crowded regions .
halo identification in such regions is complicated by the large number of halos and by the high - density background of fast moving particles .
therefore , performance of the algorithm in such regions is a good indicator of its overall performance .
in fact , the parameters of the halo finder used to construct our halo catalogs were tuned by visual inspections of the most difficult and complicated regions .
an example of such regions is shown in figure 1 .
the figure shows the distribution of the dm particles in a virgo - type cluster in the model ( run ) . to enhance the substructure , particles are color - coded on a grey - scale with a rendering algorithm based on the local particle density ( see caption ) .
a large number of distinct and compact dm halos is clearly present within the virial radius of the cluster .
figure 2 shows positions of dm halos identified by the bdm code in the same volume .
there are 121 halos in the plot , each halo having more than 20 bound particles .
more stringent limits of more than 30 particles and a maximum circular velocity , @xmath179 , larger than @xmath195 produce 98 halos .
all distinct halos visible in figure 1 , are identified .
( 0,8.0)(15.0,22.6 )
( -2.0,22.8)=11.5 cm ( 8.5,22.8)=11.5 cm ( 0.0,11.3 ) = 3.5 mm fig . 1. distribution of dark matter in a virgo - like cluster in the @xmath196 simulations .
the cluster virial mass is @xmath197 and the corresponding 3d velocity dispersion is @xmath198 .
the particles inside a sphere of the radius of 1.5 ( solid circle ) are shown .
the size of the small box , shown to provide the comparison scale , is 100 . to enhance the contrast ,
we have color - coded dm particles on a grey scale according to their local density : intensity of each particle is scaled as the logarithm of the density difference @xmath199 , where the densities were obtained using top - hat filter with radii 15 and 75 .
( 9.6,11.3 ) = 3.5 mm fig .
2. the region of the space shown in figure 1 with circles representing the dm halos found by the bdm halo finder .
area of each circle representing a halo is proportional to the halo s maximum circular velocity .
there are 121 halos in the plot , each of which has mass @xmath200 ( more than 20 bound particles ) . in 1
we have stressed that the main new feature of the analysis presented in this paper is the identification and use of halos located inside the virial radius of other halos ( satellites or sub - halos ) .
particularly , one of the main goals is to include halos inside cluster - size halos .
this would allow an estimation of the halo - halo correlation function down to unprecedentedly small scales ( @xmath201 ) .
however , we can consider this estimate robust only if we are confident that no halos ( or a very small fraction of them ) are artificially destroyed in clusters due to the insufficient resolution .
kgkk have discussed analytical estimates and numerical experiments that could be used to address this issue . following the approach of kgkk
, we have run a series of small @xmath0-body simulationscdm model . ]
using the direct - summation aarseth s code ( binney & tremain 1987 ) .
the basic setup of the simulations is as follows .
a dm halo of virial mass @xmath202 ( containing a few thousand particles within the virial radius ) is constructed . the initial equilibrium density profile of the halo
is described by the nfw formula .
the halo is then placed on an orbit in a constant potential corresponding to a galaxy cluster of mass @xmath203with the nfw density distribution . the particular numbers quoted here are intended to mimic the orbital evolution in the cluster presented in figs
the orbital evolution of the halo was studied for different orbits and different ( mass and force ) resolutions .
both the mass and the force resolution were varied by more than a factor of ten .
the orbital evolution of mass bound to the halo converges at the force resolution of the 3 m , the corresponding resolution is @xmath204 for the art . ]
@xmath205 : i.e. , the evolution of bound mass in runs with higher resolution is _
identical_. in fact , during the first @xmath206 gyrs the resolution of @xmath207 is adequate .
the mass loss in this case is somewhat higher which leads to total destruction after 5 gyrs , whereas resolution of @xmath208 allows halo to survive during the hubble time .
these experiments have also shown that mass resolution of @xmath209 is sufficient , provided that force resolution is high .
two runs with force resolution of @xmath208 and with mass resolutions of @xmath210 and @xmath211 resulted in identical mass evolution .
it is worth noting that in these experiments the halo was followed until it was totally destroyed by the tidal field ( @xmath212 gyrs in most cases ) . in real simulations , however , clusters form only at @xmath13 , and most of the accreted halos spend @xmath213 gyrs in clusters . during this
time halos lose 80 - 90% ( depending on the orbit ) of their initial mass .
thus , if the mass resolution is @xmath209 , the halos with initial mass @xmath214 can be identified even after spending a substantial time in a cluster . details of the evolution of halos in clusters depend sensitively on the parameters of the halo orbit .
specifically , the mass loss rate depends on the pericenter and on the eccentricity of the orbit . a halo on a very eccentric orbit survives for a considerably longer time than a halo on a circular orbit , even if the radius of the latter is larger than the pericenter .
for example , a halo on a circular orbit with a radius of @xmath215 was _ totally destroyed _ in less than 5 gyrs , while a halo on a very eccentric orbit with the pericenter of only @xmath216 _ survived _ for more than 10 gyrs .
the explanation for this is simple : a halo on a circular orbit spends all of its time in a high - density cluster core suffering a steady mass loss , while a halo on the eccentric orbit spends only a small fraction of its orbital period in the core . with the resolution quoted above
, only a relatively small fraction of halos can be tidally destroyed in clusters . only the halos with the _ apocenter _
@xmath217 are subject to the destruction .
it can be expected that halos which are accreted when the cluster was young and its radius was small ( @xmath218 ) have small apocenters .
the fraction of such halos in a @xmath12 cluster is small .
our estimates show that for halos with large apocenters the dynamical friction can not bring the apocenter considerably closer to the cluster center .
the tidal stripping reduces the halo mass very efficiently , thus increasing the friction time .
the tidal stripping that halos suffer in clusters has an important effect on their density profiles .
it appears that the halo profile is affected at all radii and not only at radii close to the tidal radius .
indeed , the trajectories of particles in a halo are mostly eccentric and after a crossing time the absence of the stripped particles will be `` felt '' by the whole halo .
the change in the density profile means , of course , a change in the maximum circular velocity of the halo .
the effect is not dramatic because the fraction of particles on highly eccentric orbits is rather small and the central regions of halos are thus affected the least .
typically , the circular velocity of the halos is reduced during their evolution in the cluster by 20%30% .
this correction was taken into account in figure 3 by reducing the velocity cut from 120 km / sec at z=1.0 to 100 km / sec at z=0.0 .
the tidal destruction is important only for cluster - size halos ( where tidal fields are the strongest ) .
there are three massive clusters in our simulation with mass @xmath219 .
there is one `` coma cluster '' with velocity dispersion @xmath220 and mass @xmath221 inside @xmath222 radius .
the cluster contains 201 halos of mass @xmath223 inside a 2 box and 9 halos inside 0.3 .
unfortunately , the cluster has suffered a recent major merger and it was difficult to study the radial distribution of halos .
one of the other two clusters also shows an indication of an ongoing merger .
it has two sub - clusters in the central region separated by @xmath224 .
therefore , we will focus on the third cluster which has a relatively regular appearance ( see fig.1 ) .
the cluster contains 121 bound halos with @xmath225 particles inside 1.5 .
there are 231,200 dark matter particles inside the virial radius of 1.28 .
the virial mass of the cluster has increased from @xmath226 at @xmath227 to @xmath228 at @xmath12 .
figure 3 shows the density profile of matter and the radial number density profile of dm halos inside the cluster . the dm density profile at @xmath12 is well approximated by the nfw profile .
for comparison , we also present the density profile of this cluster at redshift one . note that at @xmath227 the radius is given in proper units and the mean matter density ( @xmath229 ) is estimated at @xmath12 .
similarly , the bottom panel of figure 3 shows the number density profile of halos in the cluster at @xmath12 and at @xmath227 .
halos with more than 30 _ bound _ particles and with limits on the maximum circular velocity @xmath230 and @xmath231 ( the change is for the reasons explained above ) for @xmath12 and @xmath227 , respectively , were used .
the mean number density of halos , @xmath232 , was estimated using all halos in the simulation within the above velocity limits ( @xmath233 at @xmath12 ) .
the profile at @xmath227 is also rescaled into proper units .
figure 3 clearly indicates that the number of halos in the central @xmath234 ( proper radius ) has declined substantially from @xmath227 to the present epoch . in the central @xmath234
there are three times as many halos ( 24 ) at @xmath235 than at @xmath12 .
if we interpret the difference as due to tidal destruction , then 16 halos were destroyed in the central part .
the situation is different at larger radii : there are 57 and 50 halos at @xmath236 for @xmath227 and @xmath12 , respectively .
note , that we compare the number of halos in the _ same proper volume _ at these two moments : the volume corresponding to the virial radius of the cluster at @xmath12 ( the virial radius at @xmath227 is smaller ) .
the number density profile is thus virtually unaffected at @xmath237 .
in this section we will present results on the evolution of the 2-point correlation function and bias ( as defined by eq . [ 1 ] ) in the simulations described in the previous section .
the presentation of the results is split in two subsections . in the first of them we present the results of our largest simulation : @xmath123cdm@xmath196 . in the second section we present the results of the rest of our simulations of different cosmological models .
thus , we will first focus on the results of the most realistic of the studied models , ( see 3 ) , and then will discuss the differences between cosmological models .
however , before we proceed with the presentation of the results , we will first compare our estimate of the dark matter correlation function with estimates which have been performed by different authors ( 0.5,-1.5)(13.0,17.0 ) ( -1.0,16.7 ) ( 0.4,4.5 ) = 3.5 mm fig .
3. the density profile of the cluster shown in figures 1 and 2 .
_ top panel : _ dark matter density in units of the mean matter density at @xmath12 ( _ solid line _ ) and at @xmath227 ( _ dot - dashed line _ ) ; the _ dashed line _ shows the navarro , frenk & white ( 1996 ) fit to the @xmath12 density profile .
the @xmath227 density profile is given in _
proper _ units : the radius is given in proper scale and the mean matter density is estimated at @xmath12 . _
bottom panel : _ number density profiles of halos in the cluster at @xmath12 ( _ solid circles _ ) and at @xmath227 ( _ open circles _ ) as compared to the @xmath12 density profile of dark matter ( _ solid curve _ ) .
the error - bars show @xmath238 poisson errors .
halos with more than 30 bound particles and with maximum circular velocity larger than @xmath239 at @xmath12 and larger than @xmath240 at @xmath227 were used to estimate the average density @xmath232 ( see 4.3 for details ) .
the profile at @xmath227 is rescaled into proper units similarly to that of the dark matter in the top panel . and with different numerical codes . in figure 4
we compare the 2-point correlation function of the _ dark matter _ in the model ( the art run @xmath123cdm@xmath196 ) with similar estimates presented by klypin et al .
( 1996 ) and by jenkins et al .
the two latter estimates were done with different codes ( pm in klypin et al . ; and ap@xmath40 m in jenkins et al . ) and different resolutions ( cell size of @xmath241 for the pm , and @xmath242 ( plummer ) for the ap@xmath40 m ) .
all simulations followed @xmath130 particles , although physical mass resolution was different due to the different box sizes ( @xmath243 , @xmath141 , and @xmath244 for the pm , art , and 3 m simulations ) .
( 0.5,-1.5)(13.0,13.0 ) ( -0.5,13.3)=10.5 cm ( 0.4,3.5 ) = 3.5 mm fig .
4. the comparison of the correlation functions of the dark matter in the model estimated by different authors with different numerical resolutions and codes .
the _ solid curve _ shows in our @xmath196 run , simulated using the art code .
the _ dashed curve _ shows estimated by jenkins et al .
( 1998 ) using the 3 m code .
the _ crosses _ show estimated by klypin et al .
( 1996 ) using the pm code .
the 3 m and pm correlation functions were rescaled as described in 5.1 to account for the difference in normalization between different simulations .
all simulations followed evolution of @xmath130 particles .
the vertical lines indicate formal force resolution for each code ( the line for the art code at @xmath245 is off the plot ) . although the cosmological model was exactly the same in all three estimates , the normalization of the power spectrum was slightly different .
the rms mass fluctuations on scale of @xmath114 @xmath246 was 0.9 , 1.0 , and 1.1 for the ap@xmath40 m , art , and pm simulations , respectively .
therefore , we have multiplied ( divided ) the dm correlation function of ap@xmath40 m ( pm ) simulation by @xmath247 in order to account for the differences in normalization .
figure 4 indicates that there is a very good agreement between all estimates at scales @xmath248 .
the art and 3 m estimates agree to better than 10% at scales
@xmath249 ! on larger scales the art correlation function has a smaller amplitude than that of the 3 m due to a factor of 2.35 smaller box size of the former .
this result is in agreement with previous studies ( e.g. , coln et al .
1997 and references therein ) which concluded that the correlation function is underestimated at scales @xmath250 of the simulation box size .
nevertheless , the agreement is striking at smaller scales , given all the differences ( _ including cosmic variance _ ) between the estimates .
we conclude , therefore , that the correlation functions presented in the next subsection are robust at scales @xmath7 .
this scale is , of course , lower ( @xmath251 ) for the @xmath143 simulations presented in 5.3 .
ccccccccc 0.0 & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] 0.5 & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] 1.0 & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] 1.5 & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] 3.0 & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] & [email protected] & [email protected] & & [email protected] & [email protected] & & [email protected] & [email protected] ( 0.5,-1.5)(13.0,15.0 ) ( -2.3,14.7 ) ( 0.0,1.5 ) = 3.5 mm fig .
5. the evolution of the 2-point correlation function of the dark matter ( bottom panel ) and halos ( top panel ) in the @xmath196 simulation .
only halos with maximum circular velocity @xmath253 were used to estimate the halo correlation functions .
poisson errors for the halo correlation functions are negligible at scales @xmath254 and are not shown for clarity ; at scales @xmath255 the errorbars are @xmath256 ( see 5.2 for details ) .
the best fit parameters of the power - law fit to the halo correlation functions are given in table 4 .
in figure 5 we plot the evolution of the correlation function of both the dark matter @xmath257 and the dm halos @xmath258 in our simulation .
all results in this and the following sections are presented in _
comoving _ coordinates .
the halo - halo correlation function was constructed using halos with @xmath231 .
note that although the earliest moment at which we show the correlation function is @xmath259 , the first halos in this simulation collapsed at redshift @xmath260 and hundreds of halos are identified at @xmath261 .
the number of halos in the @xmath231 catalog is approximately @xmath262 , @xmath263 , and @xmath264 for redshifts of 5 , 2 , and 0 , respectively .
the good statistics result in a very accurate estimate of the correlation function .
the number of pairs per scale bin used to estimate @xmath258 is @xmath265 in all cases .
typically , @xmath266 for @xmath267 , @xmath268 for @xmath269 , and @xmath270 for the larger scales .
the pure poisson errors associated with each of the points are thus neglegibly small , except for the first 2 - 3 bins ( where poisson errors are still small : @xmath271 ) .
there are subtler errors associated with radial binning , but these are also less than a few per cent .
figure 5 shows that the shapes of the matter and halo - halo correlation functions are quite different .
the matter correlation function changes its shape from almost a power - law to a complicated shape .
the slope of @xmath257 at scales @xmath272 stays approximately constant throughout the evolution , while at the larger scales @xmath257 significantly _ steepens_. the amplitude of the dm correlation function increases from @xmath259 to @xmath12 by factors of @xmath273 and @xmath274 at small and large scales , respectively .
the halo - halo correlation function behaves very differently .
its shape can be well - described by a power - law at all epochs .
the amplitude of @xmath258 evolves non - monotonically : it _ decreases _ somewhat from @xmath259 to @xmath74 , and then gradually increases .
the evolution , however , is much more modest than the dramatic evolution of the @xmath257 : the maximum difference in amplitude among any two epochs is only a factor of two .
the details of the @xmath258 evolution are illustrated in figure 6 .
this figure shows the evolution of the @xmath258 amplitude at a variety of different _ comoving _ scales ( indicated on the right ) for the three runs from our set of simulations ( see table 2 ) .
different initial conditions of these runs allow us to evaluate the cosmic variance , while different particle masses and spatial resolutions ( by a factor of 8 and 2 , respectively ) of @xmath275 and @xmath276 allow us to check for the resolution effects .
comparison of @xmath277 in the @xmath143 runs simulated with different codes and with significantly different resolutions shows that the two runs agree very well .
the @xmath276 simulation had the best mass and force resolution , yet , we do not find any visible systematic differences at scales @xmath278 with the other two simulations .
there is an indication that at @xmath279 the amplitude in the @xmath143 simulations is systematically lower than that in the @xmath275 , which can be explained by the effects of finite box size .
this is in agreement with expectation that the finite size effects become important at scales @xmath280 ( @xmath281 of the box size ) . at smaller scales , where we expect the @xmath143 simulations to produce correct results ,
the agreement is very good .
besides the illustration of a very mild evolution of @xmath258 , figure 6 also shows that there is a common feature of the evolution .
although the exact evolution depends to some extent on the scale , the amplitudes at all scales are quite high ( as high or higher as they are at @xmath12 ) at very high redshift ( @xmath282 ) . the amplitude then _ decreases _ until @xmath283 , and grows steadily at lower redshifts .
it is important to note that this evolution is more complicated than simple evolution models often used in the observational and theoretical analyses : @xmath284 .
figure 6 shows that parameter @xmath285 estimated by such analyses would depend not only on the redshift range used , but also on the scale at which the amplitude is measured ( as well as on other parameters such as the object s mass ) .
this calls into question the usefulness of such a simplistic approach ( see also arguments in moscardini et al .
note that there is also some observational evidence ( giavalisco et al .
1998 ) indicating that the above parameterization is a poor description of the observed galaxy clustering evolution . for a limited range of redshifts @xmath286
, we find only very weak evolution of halo clustering in _ comoving _ coordinates indicating a value of @xmath287 .
this value seems to be favored by observations of galaxy clustering at these redshifts ( e.g. , postman et al .
1998 ) .
( 0.5,-1.5)(13.0,17.0 ) ( -2.7,17.0)=14.5 cm ( 0.0,2.5 ) = 3.5 mm fig .
6.the evolution of the correlation function of halos in @xmath196 simulation at various _ comoving _ scales .
each curve shows the amplitude of the correlation function at a fixed comoving radius indicated on the right side of the panels .
the correlation functions were estimated for halos with circular velocity larger than @xmath288 ( _ top _ ) and @xmath289 ( _ bottom _ ) .
the _ solid curves _ and _ open circles _ indicate amplitudes in the art @xmath141 and @xmath143 simulations , respectively ; the _ triangles _ show the amplitude in the @xmath143 3 m simulation ( see 5.2 for details and discussion ) .
the power - law fits to @xmath258 of the form @xmath290 for various epochs and for halo catalogs with different cuts in the maximum circular velocity are presented in table 4 .
direct ( rather than linear fits to @xmath291 ) weighted power - law fits were done to @xmath258 with the levenberg - marquardt method described in press et al .
each bin of the correlation function was weighted with its poisson error @xmath292 , where @xmath293 is the number of halo pairs in the bin .
visual inspection shows that power - law fits are very successful at scales @xmath294 , while at smaller scales there are @xmath295 deviations in some cases .
the goodness of the fits is represented in rather small _ formal _ errors of the best fit parameters @xmath296 and @xmath297 .
we have estimated how these parameters change if the correlation function is re - binned differently and found that the ( 0.5,-1.5)(13.0,16.0 ) ( -1.0,16.0 ) ( 0.4,3.5 ) = 3.5 mm fig .
7. _ bottom panel : _ comparison of the halo correlation function in the @xmath196 simulation with the correlation function of the apm galaxies ( baugh 1996 ) .
results for halos with maximum circular velocity larger than @xmath240 , @xmath288 , and @xmath298 are presented by the _ solid _ ,
_ dot - dashed _ , and _ dashed _ curves , respectively .
the _ dotted curve _ shows the dark matter correlation function .
note that at scales @xmath294 the halo correlation function does not depend on the limit in the maximum circular velocity ( see 5.2 for details ) .
_ top panel : _ dependence of bias on scale and maximum circular velocity .
the curve labeling is the same as in the bottom panel , except that the _ dotted line _ now represents the bias of halos with @xmath299 .
change is always @xmath300 .
the examination of the table 4 shows that typical ranges of @xmath296 and @xmath297 are @xmath301 and @xmath302 , respectively . for all halo catalogs ,
parameters evolve slowly with redshift .
the correlation length @xmath296 decreases somewhat between redshifts of 5 and 3 , and then increases steadily until @xmath12 .
the evolution of @xmath297 is even slower with the tendency for @xmath297 to increase by @xmath303 from @xmath259 to the present epoch .
an important and interesting point is that the correlation amplitude , and hence the value of @xmath296 , are quite different for halo catalogs with different @xmath179 cuts .
the correlation lengths for @xmath231 and @xmath304 catalogs differ by 25% at @xmath305 , while the difference is only @xmath306 at @xmath12 .
this means there is a mass segregation of halo clustering properties at high redshift , which , however , is erased during the subsequent evolution . during the last few years
, there has been tremendous progress in the observational studies of high - redshift galaxy clustering .
we will discuss how our results on the _ halo _ clustering evolution compare with the results of observations in 6 . here ,
however , we present a comparison of the @xmath258 with the most accurate measurement of the galaxy correlation function @xmath307 ( at @xmath308 ) made using the apm galaxy survey ( baugh 1996 ) .
figure 7 shows the @xmath12 correlation functions of halos and dark matter in the model and the real space apm galaxy correlation function .
the halo - halo correlation function was estimated for halo catalogs with cuts in the maximum circular velocity of @xmath240 , @xmath288 , and @xmath298 .
the figure shows striking agreement between the halo and galaxy correlation functions : at scales @xmath294 the correlation functions of all halo catalogs match _ both the shape and the amplitude _ of the @xmath307 .
the correlation function for @xmath309 catalog agrees with apm @xmath307 within errors at all probed scales . as we noted above , the differences that exist between
the catalogs at high redshifts virtually vanish during the course of the evolution .
this is manifested in the similarity of @xmath258 for different catalogs on scales @xmath294 .
note , however , that this does not mean that @xmath258 has no mass dependency .
rather , the result means that by @xmath12 any mass dependence of the correlation function vanishes when averaged over a range of galactic masses .
as was explained above , the poisson errors of the halo correlation functions shown in figure 7 are very small and were not shown for clarity .
the robustness of the result can be estimated , however , by comparing @xmath258 of @xmath310 and @xmath309 catalogs .
the number of halos in these catalogs is significantly different : 4708 and 2480 , respectively .
this makes the halo samples largely independent .
the correlation functions agree , however , within the poisson errors .
while the correlation function of halos matches that of galaxies very accurately , the correlation function of matter @xmath257 matches @xmath307 neither in shape nor in amplitude .
the amplitude is matched only at scales @xmath311 . at smaller scales
it is much higher than the amplitude of the apm @xmath307 , implying that dm halos are _ anti - biased _ at these scales with respect to the dark matter .
moreover , the difference in shape between @xmath258 and @xmath257 implies that the _ bias is scale - dependent_. the scale dependence of the bias ( @xmath312 ) for the halo correlation functions is shown in the top panel of figure 7 .
the bias varies significantly at scales @xmath11 in the range @xmath313 .
moreover , as was shown in figure 5 , the shape of the correlation function of dark matter differs from that of @xmath258 _ at all epochs _ and evolves much more strongly than the correlation function of halos .
the former fact implies that _ the bias is scale - dependent at all epochs _ , while the latter means that _ the bias evolves rapidly with cosmic time_. the evolution of bias in the @xmath275 run is illustrated in figure 8 .
the evolution is shown for different halo catalogs and at different scales .
the bias evolves very rapidly from value of @xmath314 at @xmath315 to @xmath313 at @xmath12 .
the evolution depends on the velocity ( or mass ) cut of the catalog at high @xmath16 : the halos in the catalogs with higher velocity cuts exhibit stronger clustering .
this difference vanishes , however , at @xmath316 .
the evolution of the scale dependence ( 0.5,-1.5)(13.0,14.5 ) ( -2.2,14.8 ) ( 0.4,0.7 ) = 3.5 mm fig .
8. _ top panel : _ the evolution of bias at _ comoving _ scale of @xmath317 for halos with different lower limit on the maximum circular velocity in the @xmath196 simulation . the _ bottom panel _ shows dependence of the bias on ( _ comoving _ ) scale for halos with maximum circular velocity @xmath289 . of the bias
is also interesting . at high redshifts the bias was larger at small scales . at small redshifts the halos are almost unbiased ( @xmath318 ) on a few megaparsec scales and are anti - biased ( @xmath319 ) on small scales .
( 0,8.0)(15.0,22.5 ) ( -0.2,22.3 ) ( 9.5,22.3 ) ( -0.2,11.0 ) = 3.5 mm fig
. 9.the evolution of the correlation function of the dark matter ( _ solid lines _ ) and halos ( _ dot - dashed line _ ) for the scdm model in the @xmath143 3 m simulation .
the panels show the correlation functions at different redshifts .
only halos with maximum circular velocity larger than @xmath298 were used to compute the halo correlation function .
the number of halos used to estimate the correlation function is indicated in each panel .
the poisson errors ( _ dotted _ ) and bootstrap errors ( _ solid _ ) are shown by vertical bars ( see 5.3 for details ) .
( 9.6,11.0 ) = 3.5 mm fig .
10. the same as figure 9 , but for the model .
as in figure 9 , only halos with @xmath320 were used .
( 0,8.0)(15.0,20.0 ) ( -0.2,19.8 ) ( 9.5,19.8 ) ( -0.2,8.5 ) = 3.5 mm fig .
11.the same as figure 9 , but for the model .
only halos with @xmath321 were used to compute the correlation function ( see 5.3 ) .
( 9.6,8.5 ) = 3.5 mm fig .
12. the same as figure 9 , but for the ocdm model .
as in figure 11 , only halos with @xmath321 were used ( see 5.3 ) .
( 0,8.0)(15.0,24.0 ) ( -1.5,24.0)=12.5 cm ( 8.5,24.0)=12.5 cm ( -0.2,11.2 ) = 3.5 mm fig . 13.the evolution of the halo correlation function at various scales for all models .
the correlation functions shown in the upper panel were computed using a fixed number density of halos ( implying different limits on the maximum circular velocity cut ) .
the correlation functions shown in the lower panel were computed using all halos with @xmath179 larger than @xmath298 .
results for only two 30 simulations are presented in this panel .
the other two simulations had too small numbers of halos .
the solid lines in both panels represent results for the model with the art code in the @xmath141 box .
( 9.6,11.2 ) = 3.5 mm fig .
14. the evolution of bias for different models at two scales , @xmath322 ( upper panel ) and @xmath323 ( lower panel ) .
the markers show results with the 3 m code , the solid curves represent results for the model with the art code in the @xmath141 box .
is evolution of bias observed in the simulations specific to this model or is this evolution similar for all of the models ?
we address this question by comparing results presented in the previous section with results of the @xmath143 simulations of other cosmological models ( see table 2 ) .
figures @xmath324 show the correlation functions of halos and the dark matter for the four 3 m @xmath325 runs at four epochs .
halos with @xmath326 were used to compute @xmath327 for the scdm and the models .
a lower @xmath310 limit was used for the ocdm and models . the difference in the @xmath179 limits
is explained by the difference in the matter density that results in different mass resolution of the simulations . in the ocdm and
the runs halos with @xmath328 are scarce , while the poorer mass resolution does not allow us to reduce the limit to 120 in the scdm or the runs .
the total number of halos found in the simulations is indicated in each panel .
the number depends on the epoch and model and varies from the maximum of 1942 ( in the scdm run at @xmath329 ) to the minimum of 611 ( in the run at @xmath330 ) .
the statistics of halos are poorer than in the @xmath196 simulation .
therefore , we plot the errorbars associated with each point of .
we estimate both poisson and bootstrap errorbars and plot the largest of the two .
the bootstrap error bars have been estimated as follows .
for each run and each epoch we have drawn 5 randomly selected samples of halos from the corresponding halo population .
the number of halos in each sample is one half of the total .
we then compute the rms fluctuation between the samples and divide it by @xmath331 to get the 1@xmath332 errorbar .
figures @xmath324 show that in all models the evolution of the correlation function is qualitatively similar to that observed in the model .
for example , the shape of the is similar ( power - law ) in all models and is always different from the corresponding .
this means that _ the scale - dependent bias is universal in cosmological models_. details of the evolution are , however , model dependent .
the most drastic differences are seen at the highest redshifts .
the figures , for example , clearly show that the bias at @xmath333 has very different values in different models . while bias in the two low- models is very similar , the distribution of halos at this redshift is only weakly biased in the scdm model , as opposed to the strongly biased distribution in the model . the evolution of the halo correlation function at three different comoving scales ( @xmath334 , 1 , and @xmath335 ) for all models is plotted in figure 13 .
the solid lines in both panels represent results for the model with the art code in the @xmath141 box . in the upper panel ,
the evolution is shown for the halo catalogs with a fixed number density of halos , which was achieved by varying the @xmath179 limit in different models .
note that in this case we compare correlations of halo samples with different mass functions : the and ocdm halo samples contain many low - mass halos , while samples in the @xmath336 models contain only massive halos .
such comparison is interesting for comparisons with observations when we know the number density of objects in the sample rather than their mass ( or type ) . at scales
@xmath337 , differences between the models are not significant . at 3 and at high @xmath16
the amplitude of in the scdm model is significantly lower than in other models .
the amplitude in the rest of the models is surprisingly similar .
therefore , if the biased galaxy formation scenario is correct and galaxies can be associated with host halos , this result may have interesting implications for the interpretation of clustering observations . to be able to differentiate between the models , we _ must _ know what type of the objects was used to estimate the clustering signal .
the knowledge of the number density of objects in the sample is not sufficient .
the point is to some extent illustrated by the lower panel of figure 13 , where we compare the evolution of the amplitude in @xmath275 , scdm , and models for the halo catalogs with the same selection criterion ( @xmath338 ) .
it is obvious that in this case the differences between the models are significant . although the differences are smaller at low redshifts , at @xmath339 the difference in the amplitude between and scdm models is almost an order of magnitude .
this difference can probably be explained by the delayed formation of galaxy - size halos in the model as compared with the scdm model .
the halos in the form at lower redshifts with high _ statistical bias _ , while halos in the scdm form systematically earlier and thus have had time to go through merging evolution .
the effect of the latter is to decrease the bias ( e.g. , moscardini et al .
note also that merging rates are higher in @xmath340 models ( e.g. , @xcite ) .
these results show that predictions of cosmological models are very different for samples of objects selected with the same set of criteria for all models .
the evolution of bias at scales @xmath334 and @xmath341 is shown in figure 14 for all models . here
again we compute the halo correlation function for the fixed number density of halos .
evolution of bias in all models is qualitatively similar to that of the model discussed above : the bias is a very strong function of redshift .
however , unlike the amplitude , the value of bias at these scales is very different among the models .
this is not very surprising because when the number density of halos is fixed , different models have very similar amplitudes of ( see fig.13 ) , but very different amplitudes of .
the latter is explained by the differences in the cosmological parameters , normalization , and the shape of the power spectrum .
a more interesting implication of the figure 14 is that differences in bias get smaller at low redshifts , virtually disappearing at @xmath12 .
the same effect can be observed in the evolution of the amplitude in figure 13 . as we will argue in the next section ,
the evolution of the halo correlations and bias at these scales is likely to be driven by the halo dynamics within nonlinear structures , in which case the differences between different cosmologies are largely erased .
the evolution shown in figures 13 and 14 provides , therefore , indirect support for this point : the differences in clustering amplitude and bias between the models disappear at @xmath13 , where most of the clustering signal comes from the halos located in nonlinear structures .
it is interesting to compare the evolution of the halo correlation function and the bias observed in our simulations with predictions of the analytical models and results of previous numerical simulations .
the fact that clustering strength of halos at high redshifts is comparable to that at the present epoch , has been noted in results of many simulations ( e.g. , @xcite ; brainerd & villumsen 1994 ; coln et al .
1997 ; and references therein ) .
bagla ( 1998 ) summarizes the generic behaviour of the correlation amplitude of halos above certain mass : the amplitude is high at very high redshifts , when halos are being formed , decreases thereafter and reaches a minimum , and then increases slowly and steadily until the present epoch .
the results presented in the previous section ( see figs . 6 and 13 )
are in agreement with this picture .
thus , there seems to be a good qualitative ( although , in some cases , not quantitative ) agreement among results of different numerical simulations concerning the evolution of the halo correlation function .
analytical models have reached a sufficient degree of sophistication to be able to predict the evolution of halo clustering in mildly nonlinear regimes ( see 2 ) .
the halos are found to form at the peaks of the density field ( e.g. , @xcite ) and their bias exhibits a simple scaling relation with the height of these peaks ( kaiser 1984 ; @xcite ) . at any given epoch
the halo population represents a mix of halos formed at different redshifts : newly - born or already evolved through merging .
the evolution of the correlation of halos in such a hierarchical framework is described using the extended press - schechter formalism ( mw ) . to compare predictions of the analytical models with our results
, we will use the approximation to the evolution of the _ effective bias _ given by eq .
( 4 ) ( moscardini et al . 1998 ) .
this approximation describes the evolution of bias of a sample of all halos above a certain mass .
this is roughly equivalent to our definition of halo samples with a limiting maximum circular velocity .
the prediction of this approximation is shown in figure 8 with the dotted line , where we used @xmath342 and @xmath343 ( see eq.[4 ] ) appropriate for our model and for the mass limit of @xmath344 ( moscardini et al .
the analytical model is expected to provide a good approximation at scales where @xmath345 ( mo et al .
1996 ) , i.e. at @xmath346 ( see table 4 ) .
therefore , the analytical prediction should be compared with the curve showing evolution of bias at @xmath347 .
the comparison shows that both the numerical result and the analytical model predict a rapid decrease of bias with cosmic time .
moreover , at high redshifts ( @xmath305 ) they agree well _ quantitatively_. at lower redshifts , however , the two predictions deviate from each other and are different by a factor of 2 at @xmath12 : at @xmath347 almost no bias is observed in the simulation , while strong anti - bias of @xmath348 is predicted by the analytical model .
the differences are not surprising , given the differences in our definition of a halo from that of the press - schechter halo .
the definition of the latter does not include `` satellite '' halos ; a halo ceases to exist , once it becomes bound to another halo ( i.e. , _ `` merges '' _ ) and orbits inside that halo s virial radius . our definition , on the other hand ,
does take satellite halos into consideration , because we include in our halo list every gravitationally bound clump of particles , regardless of whether it is also bound to a larger system or not .
the similarity between our numerical result and the prediction of the analytical model at @xmath305 is then an indication that the two definitions are equivalent at these high redshifts . indeed , large systems such as clusters and groups have not yet formed at these redshifts , and the fraction of satellite halos in our catalogs ( i.e. , all halos above a certain mass limit ) is relatively small . at smaller redshifts
the ever larger fraction of halos become satellites to more massive halos and the two halo definitions result in rather different halo samples .
this explains the large difference predicted for the value of bias at @xmath12 .
we have found that in our simulation the @xmath12 amplitude of the 2-point correlation function of the press - schecter halos ( i.e. , _ isolated _ in terms of their virial radius ) at @xmath347 is approximately _ two times as small _ as the amplitude of the correlation function of the bdm halos shown in fig .
7 , resulting thus in the _ twice as small bias _ of @xmath349 . in this respect
, we believe that there is no contradiction between our results and predictions of the analytical model , once the difference in the halo definition is taken into account .
this also explains the difference in the bias value with the result of jing ( 1998 ) , who found bias of @xmath350 for the galaxy - size halos defined with the fof algorithm .
we find also that , contrary to the assumption of the analytical models , the _ small - scale _ bias of the halo distribution is _ scale - dependent _ regardless of the halo definition .
the large - scale bias ( @xmath351 ) , on the other hand , is not probed in our simulations and may be independent of scale , in accord with assumptions of the analytical models .
this indeed was suggested by the simulations of jing ( 1998 ) .
we think that the most encouraging result is the agreement between numerical and analytical modeling on the general form of the bias evolution demonstrated in fig .
this indicates that we now have a solid general understanding of the nature of bias and of the processes driving its evolution at redshifts @xmath352 . at smaller redshifts ,
the merging rate is considerably smaller and small - scale correlations are sensitive to the dynamics of halos inside the nonlinear structures .
particularly , the dynamics and the clustering evolution of satellite halos in high - density regions are essentially independent of the background cosmology and are driven by such processes as dynamical friction and tidal stripping ( e.g. , kgkk ) .
these processes tend to suppress the growth of the correlation amplitude , thus counteracting the clustering growth due to the gravitational pull .
this leads to the anti - bias observed in our simulations at nonlinear scales and at small redshifts .
indeed , the correlation amplitude at small scales ( @xmath353 ) is approximately constant at redshifts @xmath354 ( see figs . 6 and 13 ) .
moreover , anti - bias is observed in all cosmological models studied in this paper ( see figs .
the fact that differences in the correlation amplitude and bias , existing between the cosmological models at @xmath24 , virtually vanish by the present epoch ( see figs
. 13 and 14 ) argues for the importance of the nonlinear halo dynamics .
this result also implies that low - redshift clustering depends only weakly on the background cosmology .
therefore , the information about the underlying cosmological model can probably be extracted only from the high-@xmath16 ( @xmath305 ) clustering data . as was discussed in 4.3 ,
the numerical resolution required to assure the survival of halos in high - density regions is high , and has not been reached in previous simulations .
our results therefore indicate that high resolution is important for correct modelling of the bias evolution at small redshifts .
although our definition of a halo is different from that used in the conventional press - schechter framework , we believe that it is closer to what can be identified in a simulation as a galaxy location .
it seems likely that in _ every _ sufficiently massive ( @xmath31 ) _ gravitationally bound _
halo baryons will cool , form stars , and produce an object ressembling a galaxy ( e.g. , kauffman , nusser & steinmetz 1997 ; roukema et al .
1997 ; yepes et al . 1997 ; @xcite ) .
we believe , therefore , that each of the halos in our catalog can be associated with a `` galaxy '' .
observationally , many distinct galaxies are located well inside the virial radii of massive galaxies , groups , and clusters .
these galaxies , non - existent by definition in the `` virial overdensity '' halo catalogs , are included in galaxy surveys and are used to compute the correlation function . our definition , therefore , is natural , if the goal is to compare the observations with predictions of the numerical simulations .
although , as we discussed above ( 4.3 ) , the resolution ( and , correspondingly , the computational costs ) required for galaxy - size halos to survive in the tidal fields of high - density regions is quite high , the subsequent comparison with the observations is straightforward and does not require ambiguous corrections for the `` overmerging '' . in 5.2
we have compared the correlation function of halos @xmath5 in our @xmath141 simulation with the @xmath12 correlation function of galaxies @xmath30 ( baugh 1996 ) in the apm catalog ( loveday et al .
1995 ) and found a very good agreement between the two .
the apm galaxy correlation function is measured very accurately , which makes the agreement within the @xmath238-errorbars very striking ( see fig.7 ) .
recently , the advent of new faint galaxy surveys allowed the measurement of the clustering surveys at redshifts of @xmath354 ( e.g. , le fvre et al .
1996 ; shepherd et al .
1997 ; carlberg et al . 1997 ; connolly et al .
1998 ; postman et al . 1998 ; carlberg et al .
unfortunately , there is some disagreement between these studies concerning the amplitude of the clustering signal , which possibly indicates that there is morphology and/or luminosity segregation in the clustering of the intermediate redshift galaxies .
it is indeed true that the correlation amplitude depends on the luminosity of the galaxies ( e.g. , carlberg et al .
1998 ; postman et al . 1998 ) .
the comoving correlation length @xmath296 measured in these surveys varies , generally lying in the range @xmath355 with an average value of about @xmath356 .
the correlation length of bright galaxies is , however , somewhat larger and consistent with the correlation length of the local galaxies @xmath357 ( carlberg et al .
1998 ; postman et al .
comparison of the above values of @xmath296 with correlations in our simulation ( table 4 ) , shows that we predict a correlation length in good agreement with that of bright galaxies and somewhat larger than the @xmath296 value of the faint galaxies .
the latter , however , is measured with rather large uncertainties and our values of @xmath296 are actually consistent with the observed ones within @xmath238 .
remarkable progress of the high - redshift galaxy detection techniques , based on the search for the signatures of the lyman break in the colors of faint galaxies ( steidel & hamilton 1992 , 1993 ) , resulted in a rapid growth of amount and quality of the clustering observations at @xmath24 ( steidel et al .
1998 ; giavalisco et al .
1998 ; adelberger et al . 1998 ) .
these `` lyman break galaxies '' ( lbgs ) were found to be clustered at @xmath24 as strongly as the present day galaxies .
the real - space correlation function of these galaxies was well - described by the conventional power - law form with the value of the slope @xmath297 and the correlation length @xmath296 consistent with the @xmath308 values ( giavalisco et al .
1998 ) : @xmath358 and @xmath356 .
these values are in reasonably good agreement with @xmath74 values for our @xmath359 catalog ( table 4 ) .
the value of bias that is measured for the lbgs at scales of @xmath360 is @xmath361 , for the @xmath340 , @xmath362 ( open ) , and @xmath363 ( flat ) models , respectively ( giavalisco et al .
1998 ; adelberger et al . 1998 ) .
these values can be compared with @xmath364 for halos in our simulations , shown in figs . 8 and 14 .
all models agree with the observations , within the uncertainties of the galaxy - to - halo mapping .
this result is in general agreement with the results of the numerous recent numerical studies that modeled the clustering of lbgs ( e.g. , wechsler et al . 1998 ; jing & suto 1998 ; bagla 1998 ; governato et al .
however , there appear to be some puzzling details in comparisons with the data .
the observed value of @xmath365 can be reproduced in the for massive halos with @xmath366 ( see top panel of fig.8 ) .
this is in agreement with almost all other theoretical studies .
however , the correlation function of the lbgs measured by giavalisco et al .
( 1998 ) , does not agree with the correlation function measured for the @xmath366 of our halos : @xmath356 for lbgs versus @xmath367 for the halos .
this disagreement was actually noticed in the study of adelberger et al .
( 1998 ) who used the count - in - cells analysis to derive the value of bias .
the values of the parameters of the correlation function that were derived from the observed rms fluctuations of galaxies in cells of @xmath368 ( @xmath369 ) , are considerably higher than those measured directly by giavalisco et al .
the corresponding value of @xmath370 in our simulation is @xmath371 for halos with @xmath372 , respectively .
this is consistent with the interpretation of the lbgs as objects residing inside massive ( @xmath366 or @xmath373 ) dm halos .
this result supports the interpretation of adelberger et al .
( 1998 ) and suggests that the correlation amplitude of the lbgs may be higher than that obtained from the observed angular correlation function ( giavalisco et al . 1998 ) .
overall , we believe that the comparisons discussed above indicate that there is good agreement between our results and the clustering data at both low and high redshifts .
this implies that hierarchical models in which observed galaxies form in the host dm halos naturally explain the observed galaxy clustering at different epochs , including excellent agreement with the accurately measured @xmath12 correlation function . on the other hand ,
the generic form of the bias evolution observed in the numerical simulations at high redshifts agrees well with the prediction of the analytical models based on the extended press - schechter formalism .
this implies that we understand the nature of the bias and the processes that drive its evolution at high @xmath16 . at low redshifts ,
the bias evolution of gravitationally bound halos is driven by the dynamical processes inside the nonlinear structures which are largely independent of cosmology .
the study of these processes is important for a successful modelling of galaxy clustering at @xmath13 .
we have studied the evolution of the correlation function and bias of galaxy - size halos in different cosmological models ( , ocdm , , and scdm ) .
the high - resolution of our numerical simulations allowed us to avoid the overmerging in the high - density regions and estimate the correlation amplitude and bias _ directly _ at small ( down to @xmath374 ) scales .
the main results and conclusions presented in this paper are as follows .
at all epochs , the 2-point correlation function of galaxy - size halos @xmath5 is well approximated by a power - law @xmath375 with the slope @xmath376 .
the correlation length @xmath296 at @xmath12 is @xmath377 , regardless of the minimal mass limit of the halo samples . at high redshifts
, the correlation function evolves non - monotonically : @xmath296 decreases somewhat between redshifts of 5 and 3 , and then increases steadily until @xmath12 . for the most massive halos ,
the correlation length at @xmath315 is comparable to that at @xmath12 .
the difference between the shape of the @xmath5 and the shape of the correlation function of matter results in a _ scale - dependent bias _ at scales @xmath7 .
we find this to be a generic prediction of the hierarchical models independent of the epoch and of the model details .
another generic prediction is that the _ comoving _ amplitude of the correlation function for halos above a certain mass evolves non - monotonically : it decreases from an initially high value at @xmath378 , and very slowly increases at @xmath13 .
this behaviour at large scales was demonstrated by a number of authors ( see 6 ) . here
, we have shown that this behaviour also applies to the correlation amplitude at small scales ( @xmath337 ) .
the non - monotonic evolution of the correlation function calls into question the usefulness of the simplistic `` @xmath285-models '' as a description of the clustering evolution .
we note , however , that at @xmath13 the evolution of the halo correlation function is approximately monotonic ( albeit dependent on scale ) .
the very slow evolution of the halo correlation amplitude in _ comoving _ coordinates at these redshifts implies a value of @xmath287 , which is in agreement with the values preferred by the observations ( e.g. , postman et al .
1998 ) .
the evolution of the halo correlation function is _ very mild _ compared to evolution of the dark matter correlation function .
the latter evolves by a factor of @xmath379 ( depending on scale ) between redshifts of @xmath380 and @xmath381 , while the difference in amplitude of the former between any two epochs is less than a factor of 2 .
the large difference in the evolution rates of the matter and halo correlation functions means that the bias evolves rapidly with cosmic time : it changes from high @xmath382 values of @xmath8 at @xmath9 to _ anti - bias _
@xmath382 values of @xmath313 on small @xmath11 scales at @xmath12 .
we find that our results agree well with existing clustering data at different redshifts , indicating a general success of the hierarchical models of structure formation in which galaxies form inside the host dm halos .
particularly , we find excellent agreement in both slope and amplitude between @xmath14 in our simulation and the galaxy correlation function measured using the apm galaxy survey . at high redshifts ,
all models reproduce well the observed clustering of the lyman - break galaxies .
our results imply that for high - redshift clustering to be used as a cosmological test , it is crucial that we know what type of objects are used to estimate the clustering signal .
the knowledge of the number density of objects _ is not sufficient _ ( see 5 ) .
we find good agreement at @xmath15 between our results and predictions of the analytical models of bias evolution ( mw ; matarrese et al .
this indicates that we now have a solid understanding of the nature of the bias and of the processes that drive its evolution at these @xmath16 .
we argue , however , that at lower redshifts the evolution of bias is driven by dynamical processes , i.e. , dynamical friction and tidal stripping , inside the nonlinear high - density regions such as galaxy clusters and groups .
these processes do not depend on cosmology and tend to erase the differences in clustering properties of halos that exist between cosmological models at high @xmath16 .
the latter result implies that low - redshift clustering is probably not a very strong discriminator between cosmological models .
we believe that the success of the current theoretical models in interpreting the clustering data forms a solid foundation for further sophistication of the models by including the processes important for galaxy formation ( such as dynamics of baryons , cooling , star formation , and stellar feedback )
. these models would allow one to predict the observed properties of galaxies and thus mimick the observational selection criteria , allowing for a robust comparison between the model and the data .
we believe that the differences between high-@xmath16 clustering properties of objects in different cosmological models demonstrated in this study ( see 5 ) , the improved theoretical models , and the ever increasing amount of new clustering data can be successfully combined in the near future to put useful constraints on the cosmological parameters describing our universe .
this work was funded by nsf grant ast-9319970 , nasa grant nag-5 - 3842 , and nato grant crg 972148 to the nmsu .
was partially supported by dgapa / unam through project in-109896 .
we are grateful to stefan gottlber for the help in computing the correlation function of isolated halos , to carlton baugh for sending us the apm correlation function in electronic form , and to adrian jenkins for providing the virgo dark matter correlation function in electronic form .
we thank elizabeth rizza for the help in improving the language of the manuscript .
our art simulations were done at the national center for supercomputing applications ( urbana - champaign , illinois ) and on the origin2000 computer at the naval research laboratory .
the 3 m simulations have been carried out on the origin-2000 at the direccin general de servicios de cmputo , unam , mexico and at the dec alpha stations at cita ( toronto , canada ) .
is very grateful to r. carlberg for an access to the cita computing resources , where part of the 3 m simulations were run .
dum adelberger , k.l . ,
steidel , c.c .
, giavalisco , m. , dickinson , m. , pettini , m. , kellogg , m. 1998 , apj , 505 , 18 bagla , j.s .
1998 , mnras , submitted ( astro - ph/9711081 ) bardeen , j.m .
1986 , in inner space / outer space , ed .
kolb et al .
( chicago : university of chicago press ) , 212 bardeen , j.m . , bond , j.r . ,
kaiser , n. , & szalay , a.s .
1986 , apj , 304 , 15 ( bbks ) baugh , c.m .
1996 , mnras , 280 , 267 binney , j. , & tremain , s. 1987 , _ galactic dynamics _ ( princeton : princeton university press ) bond , j.r . , cole , s. , efstathiou , g. , kaiser , n. 1991 , apj , 379 , 440 bower , r.j .
1991 , mnras , 248 , 332 brainerd , t.g . , & villumsen , j.v .
1994 , apj , 431 , 477 bunn , e.f .
, & white , m. 1997 , apj , 480 , 6 carlberg , r.g .
1990 , apj , 359 , l1 carlberg , r.g . ,
cowie , l.l . , songaila , a. , & hu , e.m .
1997 , apj , 484 , 538 carlberg , r.g . ,
1998 , in `` large scale structure in the universe '' , proceedings of the royal society discussion meeting ( astro - ph/9805131 ) catelan , p. , matarrese , s. , porciani , c. 1998a , apj , 502 , l1 catelan , p. , lucchin , f. , matarrese , s. , porciani , c. 1998b , mnras submitted ( astro - ph/9708067 ) chaboyer , b. 1998 , phys .
reports , in press ( astro - ph/9808200 ) cole , s. , & kaiser , n. 1989 , mnras , 237 , 1127 cole , s. , weinberg , d.h . , frenk , c.s . , & ratra , b. 1997 , mnras , 289 , 37 coles , p. 1993 , mnras , 262 , 1065 coln , p. , carlberg , r.g . , &
couchman , h.m.p .
1997 , apj , 490 , 1 connolly , a.j . , szalay , a.s . , brunner , r.j .
1998 , apj , in press ( astro - ph/9803047 ) couchman , h.m.p .
1991 , apj , 368 , l23 davis , m. , efstathiou , g. , frenk , c.s .
, & white , s.d.m .
1985 , apj , 292 , 371 dekel , a. , & rees , m.j .
1987 , nature , 326 , 455 dekel , a. , & silk , j. 1986 , apj , 303 , 39 dekel , a. , & lahav , o. 1998 , apj submitted ( astro - ph/9806193 ) dubinski , j. 1998 , apj , 502 , 141 efstathiou , g. , davis , m. , frenk , c. s. , & white , s. d. m. 1985 , apjs , 57 , 241 efstathiou , g. , bond , j.r . , white , s.d.m .
1992 , mnras , 258 , 1p eke , v.r . ,
cole , s. , & frenk , c.s .
1996 , mnras , 282 , 263 eke , v.r . ,
cole , s. , frenk , c.s .
, & henry , p.j .
1998 , mnras , 298 , 1145 evrard , a.e .
1997 , mnras , 292 , 289 falco e.e . ,
shapiro , i.i . ,
moustakas , l.a . ,
davis , m. 1997 , apj , 484 , 70 frenk , c.s . ,
white , s.d.m .
, davis , m. , & efstathiou , g. 1988 , apj , 327 , 507 fry , j.n .
1996 , apj , 461 , l65 ghigna , s. , moore , b. , governato , f. , lake , g. , quinn , t. , stadel , j. 1998 , mnras , submitted ( astro - ph/9801192 ) giavalisco , m. , steidel , c.c . , adelberger , k.l . ,
dickinson , m.e .
, pettini , m. , kellogg , m. 1998 , apj , 503 , 543 governato , f. , baugh , c.m . , frenk , c.s . ,
cole , s. , lacey , c.g . ,
quinn , t. , stadel , j. 1998 , nature , in press ( preprint astro - ph/9803030 ) hamilton , a.j.s . , kumar , p. , lu , e. , mathews , a. 1991 , apj , 374 , l1 hockney , r.w . , & eastwood , j.w .
1981 , _ computer simulations using particles _
( new york : mcgraw - hill ) jain , b. , mo , h.j . ,
white , s.d.m .
1995 , mnras , 276 , l25 jenkins et al .
( the virgo consortium ) 1998 , apj , 499 , 20 jing , y.p . , & suto , y. 1998 , apj , 494 , l5 jing , y.p .
1998 , apj , 503 , l9 kaiser , n. 1984 , apj , 284 , l9 kaiser , j.m .
1986 , in inner space / outer space , ed .
kolb et al .
( chicago : university of chicago press ) , 258 katz , n. , hernquist , l. , weinberg , d.h .
1998 , preprint ( astro - ph/9806257 ) kauffmann , g. , nusser , a. , & steinmetz , m. 1997 , mnras , 286 , 795 kim , a.g . , et al .
1997 , apj , 476 , l63 kitayama , t. , & suto , y. 1996 , apj , 469 , 480 klypin , a. , primack , j. , & holtzman , j. 1996 , apj , 466 , 13 klypin , a. , nolthenius , r. , primack , j. 1997 , apj , 474 , 533 klypin , a.a . ,
gottlber , s. , kravtsov , a.v . , khokhlov , a.m
1998 , apj submitted ( astro - ph/9708191 ) ( kgkk ) klypin , a. , & holtzman , j. 1997 , preprint ( astro - ph/9712217 ) kravtsov , a.v .
, klypin , a. , khokhlov , a.m. 1997 , apjs , 111 , 73 kravtsov , a.v . ,
klypin , a.a . ,
bullock , j.s .
, primack j.r . 1998
apj , 502 , 48 lacey , c. , & cole , s. 1994 , , 271 , 676 le fvre , o. , hudon , d. , lilly , s.j . , crampton , d. , hammer , f. , & tresse , l. 1996 , apj , 461 , 534 loveday , j. , maddox , s.j . ,
efstathiou , g. , & peterson , b.a .
1995 , apj , 442 , 457 maddox , s.j . , efstathiou , g. , sutherland , w.j .
1996 , mnras , 283 , 1227 mann , r.g . , peacock , j.a .
, & heavens , a.f .
1998 , mnras , 293 , 209 matarrese , s. , coles , p. , lucchin , f. , & moscardini , l. 1997 , mnras , 286 , 115 mo , h.j .
, & white , s.d.m . 1996 , mnras , 282 , 347 ( mw ) mo , h.j . , jing , y.p . , & white , s.d.m .
1996 , mnras , 282 , 1096 mo , h.j . ,
mao , s. , & white , s.d.m .
1998 , mnras , 295 , 319 moore , b. , katz , n. , & lake , g. 1996 , apj , 457 , 455 moscardini , l. , coles , p. , lucchin , f. , & matarrese , s. 1998 , mnras , submitted ( astro - ph/9712184 ) navarro , j. , frenk , c. , & white , s.d.m .
1996 , , 462 , 563 ( nfw ) nolthenius , r. , klypin , a. , & primack , j.r .
1997 , nusser , a. , & davis , m. 1994 , apj , 421 , l1 peacock , j.a . , & dodds , s.j .
1994 , mnras , 267 , 1020 peacock , j.a . ,
& dodds , s.j .
1996 , mnras , 280 , l19 postman , m. , lauer , t.r . ,
szapudi , i. , & oegerle , w. 1998 , apj , in press ( astro - ph/9804141 ) press w.h . , & schechter , p. 1974 , apj , 187 , 425 press , w.h .
, teukolsky , s.a . , vetterling , w.t . , & flannery , b.p .
1992 , numerical recipes , 2nd ed .
( cambridge : cambridge univ . press ) rees , m.j .
1985 , mnras , 213 , 75p roukema , b.f . ,
peterson , b.a . ,
quinn , p.j . , & rocca - volmerange , b. 1997 , mnras 292 , 835 salaris , m. , & cassisi , s. 1998 , mnras , 298 , 166 salucci , p. , & persic , m. 1997 , in dark and visible matter in galaxies , eds
. m. persic , & p. salucci ( san francisco : asp conference series ) , 1 schaeffer , r. , & silk , j. 1985 , apj , 292 , 319 shepherd , c.w . ,
carlberg , r.g . ,
yee , h.k.c .
, & ellingson , e. 1997 , apj , 479 , 82 silk , j. 1985 , apj , 297 , 1 smith , c.c .
, klypin , a.a . , gross , m.a.k . , primack , j.r . , & holtzman , j. 1998 , mnras , 297 , 910 steidel , c.c . , & hamilton , d. 1992 , aj , 104 , 941 steidel , c.c . , & hamilton , d. 1993 , aj , 105 , 2017 steidel , c.c . ,
adelberger , k.l . , dickinson , m. , giavalisco , m. , pettini , m. , & kellogg , m. 1998 , apj , 492 , 428 sugiyama , n. 1995 , apjs , 100 , 281 summers , f.j . , davis , m. , & evrard , a.e .
1995 , apj , 454 , 1 walker , t.p . ,
steigman , g. , schramm , d.n . , olive , k.a . , kang , h .- s .
1991 , apj , 376 , 51 wechsler , r.h . , gross , m.a.k . , primack , j.r . ,
blumenthal , g.r .
, & dekel , a. 1998 , apj , submitted ( preprint astro - ph/9712141 ) yepes , g. , kates , r. , khokhlov , a. , & klypin , a. 1997 , mnras , 284 , 235 | we study the evolution of the halo - halo correlation function and bias in four cosmological models ( , ocdm , , and scdm ) using very high - resolution @xmath0-body simulations with dynamical range of @xmath1 ( force resolution of @xmath2 and particle mass of @xmath3 ) .
the high force and mass resolution allows dark matter ( dm ) halos to survive in the tidal fields of high - density regions and thus prevents the ambiguities related with the `` overmerging problem . ''
this allows us to estimate for the first time the evolution of the correlation function and bias at small ( down to @xmath4 ) scales .
we find that at all epochs the 2-point correlation function of galaxy - size halos @xmath5 is well approximated by a power - law with slope @xmath6 .
the difference between the shape of @xmath5 and the shape of the correlation function of matter results in the _ scale - dependent bias _ at scales @xmath7 , which we find to be a generic prediction of the hierarchical models , independent of the epoch and of the model details .
the bias evolves rapidly from a high value of @xmath8 at @xmath9 to the _ anti - bias _ of @xmath10 at small @xmath11 scales at @xmath12 .
another generic prediction is that the _ comoving _ amplitude of the correlation function for halos above a certain mass evolves non - monotonically : it decreases from an initially high value at @xmath9 , and very slowly increases at @xmath13 .
we find that our results agree well with existing clustering data at different redshifts , indicating the general success of the hierarchical models of structure formation in which galaxies form inside the host dm halos .
particularly , we find an excellent agreement in both slope and the amplitude between @xmath14 in our simulation and the galaxy correlation function measured using the apm galaxy survey . at high redshifts ,
the observed clustering of the lyman - break galaxies is also well reproduced by the models .
we find good agreement at @xmath15 between our results and predictions of the analytical models of bias evolution .
this indicates that we have a solid understanding of the nature of the bias and of the processes that drive its evolution at these epochs .
we argue , however , that at lower redshifts the evolution of the bias is driven by dynamical processes inside the nonlinear high - density regions such as galaxy clusters and groups .
these processes do not depend on cosmology and tend to erase the differences in clustering properties of halos that exist between cosmological models at high @xmath16 .
3 m 30art 8 # 1#2.5pt ' # 1#1i"13i |
as a possible approach to the functional integral with complex measure , such as the one encountered in the finite density qcd @xcite , the complex langevin method @xcite has attracted much attention in recent years .
this recent interest was triggered mainly by the discovery of sufficient conditions for the convergence of the method to a correct answer @xcite .
reference @xcite is a review on recent developments . roughly speaking , if the probability distribution of configurations generated by the langevin dynamics damps sufficiently fast at infinity of configuration space , the statistical average over the configurations is shown to be identical to the integration over the original complex measure .
it has been observed that , in systems for which the complex langevin ( cl ) method converges to a wrong answer ( such as the three - dimensional xy model @xcite ) , this requirement of a sufficiently localized distribution is broken , typically in `` imaginary directions '' in configuration space .
after the above understanding , a prescription in lattice gauge theory that makes the probability distribution well localized was proposed in ref .
@xcite ; the prescription is termed `` gauge cooling '' and it proceeds as follows : the link variables in lattice gauge theory are originally elements of the compact gauge group @xmath1 . when the ( effective ) action is complex , however , the corresponding langevin evolution drives link variables into imaginary directions and link variables become elements of @xmath2 , a _ noncompact _ gauge group .
this evolution tends to make the distribution wide in noncompact directions ; in terms of the @xmath1 lie algebra , those noncompact directions are parametrized by imaginary coordinates . at this point ,
one notes that the definition of a physical observable that is invariant under the original @xmath1 gauge transformations can always be tailored so that it is invariant also under the noncompact @xmath2 gauge transformations .
the idea of the gauge cooling is that by applying the @xmath2 gauge transformations appropriately along the complex langevin evolution , one squeezes the distribution well localized so that the prerequisite of the convergence theorem @xcite is fulfilled without changing physical observables . in a one - dimensional gauge model and in the four - dimensional qcd with heavy quarks
, it has been confirmed that the gauge cooling makes the distribution well localized and the complex langevin method gives rise to correct answers @xcite .
more recently , this method was applied to the full qcd at finite density @xcite .
see also refs .
one should note , however , that the langevin dynamics itself is defined on gauge noninvariant variables ( i.e. , link variables ) and also that the gauge cooling step can not be regarded as a langevin evolution that is induced by a _
action ; the latter is assumed in the convergence theorem @xcite .
strictly speaking , therefore , the convergence theorem does not apply when the gauge cooling is employed .
the method should still be carefully examined in various possible ways . in the present paper
, we apply the complex langevin method in conjunction with the gauge cooling to the two - dimensional lattice yang - mills theory which can be analytically solved @xcite . by doing this ,
we examine the validity of the method .
the partition function of the two - dimensional yang - mills theory on the lattice is given by @xmath3 \mathrm{e}^{-s } , \label{eq:(1.1)}\ ] ] where @xmath4 are link variables defined on a two - dimensional rectangular lattice , @xmath5 is the lattice action , . ] @xmath6 , \label{eq:(1.2)}\ ] ] and the plaquette variable is defined by @xmath7 for simplicity , we assume that the gauge group is @xmath0 , that is , @xmath8 in the original integral . on the other hand ,
when the gauge coupling @xmath9 is complex , the corresponding langevin equation [ eq . below ] evolves link variables as elements of @xmath10 .
thus , the distinction between @xmath11 and @xmath12 becomes very important in the complex langevin dynamics .
for the convergence theorem in refs .
@xcite to apply , the action @xmath5 that generates the drift force in the langevin equation and physical observables must be a holomorphic function of dynamical variables ; our above definitions that entirely use @xmath12 not @xmath11 are chosen by this criterion .
note also that the plaquette action is invariant under the @xmath10 lattice gauge transformations [ such as the one in eq . ] .
we consider the expectation value of the plaquette variable : @xmath13\right\rangle = \frac{1}{\mathcal{z } } \int\left[\prod_{x,\mu}\mathrm{d}u_{x,\mu}\right ] \mathrm{e}^{-s } \operatorname{tr}\left[u_{01}(x)\right ] .
\label{eq:(1.4)}\ ] ] even if the gauge coupling @xmath9 is complex , this can be exactly computed by the character expansion @xcite . under periodic boundary conditions ,
one yields @xmath13\right\rangle = -\frac{n}{v}\frac{\partial}{\partial\beta}\ln\mathcal{z},\qquad \mathcal{z}=\sum_{n=1}^\infty\left[\frac{2}{\beta}i_n(\beta)\right]^v , \label{eq:(1.5)}\ ] ] where @xmath14 denotes the modified bessel function of the first kind and @xmath15 is the number of lattice points .
the following procedures are basically identical to the ones adopted for the four - dimensional lattice qcd in ref .
@xcite for example , although our two - dimensional pure - gauge system is much simpler . for the link variable ,
the langevin equation with a discretized langevin time @xmath16 with the time step @xmath17 is defined by @xmath18u_{x,\mu}(t ) , \label{eq:(2.1)}\ ] ] where @xmath19 ( @xmath20 , @xmath21 , @xmath22 ) are pauli matrices , @xmath23 are gaussian real random numbers of the variant @xmath24 and @xmath25 is the drift force generated by the action @xmath5 in eq . ; the derivative with respect to the link variable is given by @xmath26 when the gauge coupling @xmath9 is complex , the drift force becomes complex and the langevin evolution evolves link variables as elements of @xmath10 . the above complex langevin dynamics tends to make the probability distribution function of link variables wide in noncompact directions of @xmath10 . to squeeze the distribution well localized without changing gauge invariant quantities
, we apply the following @xmath10 gauge transformation ( this step is the gauge cooling ) @xmath27 where @xmath28 , \label{eq:(2.5)}\ ] ] and @xmath29 is a real parameter .
the distance defined by @xcite @xmath30 measures how a @xmath10 gauge field is far away from the subspace of @xmath0 gauge fields .
it is then straightforward to see that for a sufficiently small @xmath17 , the gauge cooling decreases or does not change the distance @xmath31 .
note that @xmath32 in eq .
is not a holomorphic function of link variables and thus the step can not be regarded as a part of the complex langevin dynamics in which the drift force is generated by a holomorphic action ; this fact prevents us from applying the convergence theorem @xcite to the above procedures .
we numerically solved the langevin equation with the discretized langevin time , eq . , on a @xmath33 lattice .
periodic boundary conditions are imposed .
the maximal size of the time step @xmath17 we adopted was @xmath34 and , when the drift force becomes large , we further reduced @xmath17 `` adaptively '' according to the prescription in ref .
@xcite . as the parameter @xmath35 in the gauge cooling , we tried both @xmath36 and the adaptive choice ( see ref .
@xcite ) @xmath37 our numerical results did not show any notable difference in these two choices and we present the results with the latter choice in what follows . to determine an appropriate rate of the gauge cooling along the langevin evolution
, we observed the time evolution of the distance @xmath31 by changing the number of the gauge cooling steps per one langevin update . in fig .
[ fig:1 ] , for @xmath38 , we plotted the evolution of the distance @xmath31 as the function of the langevin time @xmath16 by changing the number of the gauge cooling steps per one langevin update as , @xmath39 , @xmath40 , and @xmath41 .
such that it does not elapse during the gauge cooling steps .
] since this plot shows the evolution including the langevin stochastic dynamics , the distance @xmath31 does not necessarily decrease . for @xmath38 with various numbers of the gauge cooling steps per one langevin update , @xmath39 , @xmath40 , and @xmath41 .
] since we do not see much difference for those three choices , we adopted ten gauge cooling steps per one langevin update . with this choice , the evolution of the distance @xmath31 for various complex gauge couplings , @xmath42 , @xmath43 , @xmath44 , and
@xmath45 , looks as depicted in fig .
[ fig:2 ] . for various complex gauge couplings , @xmath42 , @xmath43 , @xmath44 , and @xmath45 .
the number of the gauge cooling steps per one langevin update is @xmath39 . ]
it appears that the gauge cooling is working perfectly for those complex gauge couplings , suppressing the evolution to noncompact imaginary directions .
now , we turn to the computation of the expectation value of the plaquette , eq . , by the complex langevin method . starting from a configuration of random @xmath0 matrices , we discarded configurations until the langevin time @xmath46 for thermalization .
then @xmath47 configurations separated by @xmath48 from @xmath46 to @xmath49 are used to compute the expectation value . for typical values of the complex gauge coupling
, we confirmed that the plaquette values between configurations separated by @xmath48 practically have no autocorrelation . in figs .
[ fig:3][fig:4 ] , we plotted the real and imaginary parts of the expectation value obtained by the cl method .
the error bars are statistical ones .
the horizontal axis is the the complex phase @xmath50 of the gauge coupling , the partition function is invariant under @xmath51 . because of this invariance and the complex conjugation , it is sufficient to consider the range of the phase , @xmath52 . ] with the modulus @xmath53 : @xmath54 of the gauge coupling in eq . . ] of the gauge coupling in eq . .
] the solid line curves are exact values given by eq . .
we see that the complex langevin method reproduces the real part fairly well , while it clearly fails to converge to the correct value of the imaginary part when the complex phase of the gauge coupling is large . the gradation plot in fig .
[ fig:5 ] shows the relative error @xmath55\right\rangle_{\text{cl } } -\left\langle\operatorname{tr}\left[u_{01}(x)\right]\right\rangle_{\text{exact } } \right| } { \left| \left\langle\operatorname{tr}\left[u_{01}(x)\right ] \right\rangle_{\text{exact } } \right| } \label{eq:(3.3)}\ ] ] on the first quadrant of the complex @xmath9 plane .
is omitted from the plot because @xmath56\rangle\sim0 $ ] for @xmath57 . ]
( the gauge coupling ) plane .
the block around the origin @xmath58 is omitted from the plot .
the quadrant is eq .
along which figs .
[ fig:3][fig:4 ] are plotted .
four black crosses indicate complex gauge couplings we used in fig .
[ fig:2 ] . ]
the ( quadrant ) circle in the figure is eq .
along which figs .
[ fig:3][fig:4 ] are plotted .
clearly , the relative error of the complex langevin method becomes large when the complex phase of the gauge coupling becomes large .
four black crosses in the figure indicate complex gauge couplings we used in fig .
[ fig:2 ] ; the behavior in fig .
[ fig:2 ] thus suggests that the gauge cooling is correctly working for the region of the complex gauge coupling shown in fig .
[ fig:5 ] .
nevertheless , the complex langevin method shows large deviation from the correct value as in fig .
[ fig:4 ] .
this is the main result of the present paper .
a similar failure of the complex langevin method for large complex @xmath9 has been observed ; see fig .
8 of ref .
this result of ref .
@xcite is , however , for a one - dimensional integral ( not a gauge theory ) and the validity of the gauge cooling , which is our main issue in this paper , is not relevant to this result of ref .
@xcite .
it is of interest how configurations generated by the langevin dynamics distribute in configuration space .
to give some idea on this point , in figs .
[ fig:6][fig:7 ] , we present scatter plots of the plaquette variable ( averaged over the lattice volume ) for each configuration .
( @xmath59 ) . ] ( @xmath60 ) . ] both cases , with and without the gauge cooling , are shown .
figure [ fig:6 ] is for @xmath61 ( i.e. , @xmath59 ) and corresponds to points in figs .
[ fig:3][fig:4 ] with a relatively small complex phase .
figure [ fig:7 ] is , on the other hand , for @xmath62 ( i.e. , @xmath60 ) and corresponds to points in figs .
[ fig:3][fig:4 ] with a large complex phase and with large deviation .
although there is a tendency when the complex phase of the gauge coupling is large for the distribution to become somewhat wider even after the gauge cooling , it is not clear from the scatter plot in fig .
[ fig:4 ] alone whether the distribution is so poorly localized as to break the prerequisite of the convergence theorem @xcite .
more detailed study is needed on this point .
in the present paper , we applied the complex langevin method in conjunction with the gauge cooling to the two - dimensional lattice @xmath0 yang - mills theory .
our intention was to examine the validity of the method by using this analytically solvable model .
somewhat unexpectedly , as shown in figs .
[ fig:4][fig:5 ] , we obtained strong numerical evidence that the method fails to converge to the correct value when the complex phase of the gauge coupling is large . as we emphasized in the introduction , the convergence proof of refs .
@xcite does not necessarily apply when the gauge cooling is employed ; thus there is no contradiction even if the method leads to a wrong answer .
nevertheless , it is not yet clear what causes the failure for the gauge coupling with a large complex phase . to find the resolution of the problem we found in the present study , first we have to pin down what the real source of the failure is . for this , consideration
on the basis of another approach to the functional integral with complex measure , the lefschetz thimble @xcite , might provide useful insight .
see also ref .
@xcite for suggestive observations .
we thank erhard seiler , kim splittorff , and ion - olimpiu stamatescu for the introduction to the complex langevin method .
we are grateful to masanobu yahiro for encouragement . the work of h. s. is supported in part by grant - in - aid for scientific research grant no .
23540330 . p.
de forcrand , pos lat * 2009 * , 010 ( 2009 ) [ arxiv:1005.0539 [ hep - lat ] ] .
c. gattringer , pos lattice * 2013 * , 002 ( 2014 ) [ arxiv:1401.7788 [ hep - lat ] ] .
g. parisi , phys .
b * 131 * , 393 ( 1983 ) .
j. r. klauder , j. phys .
a * 16 * , l317 ( 1983 ) .
j. r. klauder , phys .
a * 29 * , 2036 ( 1984 ) .
g. aarts , e. seiler and i. o. stamatescu , phys .
d * 81 * , 054508 ( 2010 ) [ arxiv:0912.3360 [ hep - lat ] ] .
g. aarts , f. a. james , e. seiler and i. o. stamatescu , eur .
j. c * 71 * , 1756 ( 2011 ) [ arxiv:1101.3270 [ hep - lat ] ]
. g. aarts , l. bongiovanni , e. seiler , d. sexty and i. o. stamatescu , eur .
phys . j. a * 49 * , 89 ( 2013 ) [ arxiv:1303.6425 [ hep - lat ] ] .
g. aarts and f. a. james , jhep * 1008 * , 020 ( 2010 ) [ arxiv:1005.3468 [ hep - lat ] ] .
e. seiler , d. sexty and i. o. stamatescu , phys .
b * 723 * , 213 ( 2013 ) [ arxiv:1211.3709 [ hep - lat ] ] .
d. sexty , phys .
b * 729 * , 108 ( 2014 ) [ arxiv:1307.7748 [ hep - lat ] ] .
a. mollgaard and k. splittorff , phys .
d * 88 * , no .
11 , 116007 ( 2013 ) [ arxiv:1309.4335 [ hep - lat ] ] .
k. splittorff , phys .
d * 91 * , no . 3 , 034507 ( 2015 ) [ arxiv:1412.0502 [ hep - lat ] ] .
a. mollgaard and k. splittorff , phys .
d * 91 * , no . 3 , 036007 ( 2015 ) [ arxiv:1412.2729 [ hep - lat ] ] .
r. balian , j. m. drouffe and c. itzykson , phys .
d * 10 * , 3376 ( 1974 ) .
a. a. migdal , sov .
jetp * 42 * , 413 ( 1975 ) [ zh . eksp .
fiz . * 69 * , 810 ( 1975 ) ] .
h. j. rothe , world sci .
notes phys . *
43 * , 1 ( 1992 ) [ world sci .
notes phys .
* 59 * , 1 ( 1997 ) ] [ world sci .
notes phys . *
74 * , 1 ( 2005 ) ] [ world sci .
notes phys .
* 82 * , 1 ( 2012 ) ] .
g. aarts and i. o. stamatescu , jhep * 0809 * , 018 ( 2008 ) [ arxiv:0807.1597 [ hep - lat ] ] . g. aarts , f. a. james , e. seiler and i. o. stamatescu , phys
. lett .
b * 687 * , 154 ( 2010 ) [ arxiv:0912.0617 [ hep - lat ] ] .
g. aarts , f. a. james , j. m. pawlowski , e. seiler , d. sexty and i. o. stamatescu , jhep * 1303 * , 073 ( 2013 ) [ arxiv:1212.5231 [ hep - lat ] ] .
m. cristoforetti _ et al . _
[ aurorascience collaboration ] , phys .
d * 86 * , 074506 ( 2012 ) [ arxiv:1205.3996 [ hep - lat ] ] .
m. cristoforetti , f. di renzo , a. mukherjee and l. scorzato , phys .
d * 88 * , no .
5 , 051501 ( 2013 ) [ arxiv:1303.7204 [ hep - lat ] ] .
h. fujii , d. honda , m. kato , y. kikukawa , s. komatsu and t. sano , jhep * 1310 * , 147 ( 2013 ) [ arxiv:1309.4371 [ hep - lat ] ] .
g. aarts , l. bongiovanni , e. seiler and d. sexty , jhep * 1410 * , 159 ( 2014 ) [ arxiv:1407.2090 [ hep - lat ] ] .
t. kanazawa and y. tanizaki , jhep * 1503 * , 044 ( 2015 ) [ arxiv:1412.2802 [ hep - th ] ] .
a. cherman , p. koroteev and m. nsal , j. math .
* 56 * , no . 5 , 053505 ( 2015 ) [ arxiv:1410.0388 [ hep - th ] ] | the complex langevin method in conjunction with the gauge cooling is applied to the two - dimensional lattice @xmath0 yang - mills theory that is analytically solvable .
we obtain strong numerical evidence that at large langevin time the expectation value of the plaquette variable converges , but to a wrong value when the complex phase of the gauge coupling is large . |
for any graph @xmath3 , let @xmath8 ( resp .
@xmath9 ) denote the number of its vertices ( resp .
if it is clear from the context , we simply write @xmath2 and @xmath10 instead of @xmath8 and @xmath9 .
the crossing number @xmath11 of a graph @xmath3 is the minimum number of edge crossings over all drawings of @xmath3 in the plane . in the optimal drawing of @xmath3
, crossings are not necessarily distributed uniformly among the edges .
some edges could be more `` responsible '' for the crossing number than some other edges . for any fixed @xmath12 ,
it is not hard to construct a graph @xmath3 whose crossing number is @xmath12 , but @xmath3 has an edge @xmath13 such that @xmath14 is planar . richter and
thomassen @xcite started to investigate the following general problem .
we have a graph @xmath3 , and we want to remove a given number of edges . by _
at least _ how much does the crossing number decrease ?
they conjectured that there is a constant @xmath15 such that every graph @xmath3 with @xmath16 has an edge @xmath13 with @xmath17 .
they only proved that @xmath3 has an edge with @xmath18 .
pach , radoii , tardos , and tth @xcite proved that for every graph @xmath3 with @xmath19 , we have @xmath20 .
it is not hard to see @xcite that for _ any _ edge @xmath13 , we have @xmath21 .
these two results imply an improvement of the richter
thomassen bound if @xmath22 , and also imply the richter
thomassen conjecture for graphs of @xmath23 edges .
j. fox and cs .
tth @xcite investigated the case where we want to delete a _ positive fraction _ of the edges .
for every @xmath0 , there is an @xmath24 such that every graph @xmath3 with @xmath25 vertices and @xmath26 edges has a subgraph @xmath5 with @xmath27 and @xmath28 in this note we generalize theorem a. for every @xmath29 , there is an @xmath30 such that every graph @xmath3 with @xmath31 vertices and @xmath26 edges has a subgraph @xmath5 with @xmath32 and @xmath33
our proof is based on the argument of fox and tth @xcite , the only new ingredient is lemma [ lemma_rcycles ] .
let @xmath34 be integers .
a _ @xmath35-earring of size @xmath36 _ is a graph which is a union of an edge @xmath37 and @xmath36 edge - disjoint paths between @xmath38 and @xmath39 , each of length at most @xmath40 .
edge @xmath37 is called the _
main edge _ of the @xmath35-earring .
[ lemma_rcycles ] let @xmath34 be integers
. there exists @xmath41 such that every graph @xmath3 with @xmath42 vertices and @xmath43 edges contains at least @xmath44 edge - disjoint @xmath35-earrings , each of size @xmath36 . by the result of alon , hoory , and linial @xcite , for some @xmath41 , every graph with @xmath45 vertices and at least @xmath46 edges contains a cycle of length at most @xmath35 .
suppose that @xmath3 has @xmath42 vertices and @xmath43 edges .
maximal _ edge - disjoint set @xmath47 of @xmath35-earrings , each of size @xmath36 .
let @xmath48 , the set of all edges of the earrings and let @xmath49 .
now let @xmath50 be a @xmath35-earring of @xmath5 of maximum size .
note that this size is less than @xmath36 .
let @xmath51 .
similarly , let @xmath52 be a @xmath35-earring of @xmath53 of maximum size and let @xmath54 .
continue analogously , as long as there is a @xmath35-earring in the remaining graph .
we obtain the @xmath35-earrings @xmath55 , and the remaining graph @xmath56 does not contain any @xmath35-earring .
let @xmath57 .
we claim that @xmath58 .
suppose on the contrary that @xmath59 .
take the main edges of @xmath55 .
we have at least @xmath46 edges so by the result of alon , hoory , and linial @xcite some of them form a cycle @xmath60 of length at most @xmath35 .
let @xmath61 be the smallest index with the property that @xmath60 contains the main edge of @xmath62 .
then @xmath60 , together with @xmath62 would be a @xmath35-earring of @xmath63 of greater size than @xmath62 , contradicting the maximality of @xmath62 .
each of the earrings @xmath55 has at most @xmath64 edges so we have @xmath65 .
the remaining graph , @xmath66 does not contain any @xmath35-earring , in particular , it does not contain any cycle of length at most @xmath35 , since it is a @xmath35-earring of size one .
therefore , by @xcite , for the number of its edges we have @xmath67 .
it follows that the set @xmath68 contains at least @xmath69 @xmath70 edges .
each of @xmath71 has at most @xmath72 edges , therefore , @xmath73 .
[ lemma_ft ] let @xmath3 be a graph with @xmath2 vertices , @xmath10 edges , and degree sequence @xmath74 .
let @xmath75 be the integer such that @xmath76 but @xmath77 . if @xmath2 is large enough and @xmath78 then @xmath79 _ proof of the theorem .
_ let @xmath80 be fixed .
choose integers @xmath81 such that @xmath82 , and @xmath83 .
it follows that we have @xmath84 , and @xmath85 .
then there is an @xmath30 with the following properties : ( a ) @xmath86 from lemma [ lemma_rcycles ] , ( b ) @xmath87 .
let @xmath3 be a graph with @xmath88 vertices and @xmath89 edges .
let @xmath90 be the vertices of @xmath3 , of degrees @xmath74 and define @xmath75 as in lemma [ lemma_ft ] , that is , @xmath76 but @xmath77 .
let @xmath91 be the subgraph of @xmath3 induced by @xmath92 .
observe that @xmath91 has @xmath93 edges .
therefore , by lemma [ lemma_rcycles ] @xmath91 contains at least @xmath94 edge - disjoint @xmath35-earrings , each of size @xmath36 .
let @xmath95 be the set of the main edges of these @xmath35-earrings .
we have @xmath96 .
let @xmath97 and @xmath98 .
take an optimal drawing @xmath99 of the subgraph @xmath100 .
we have to draw the missing edges to obtain a drawing of @xmath3 .
our method is a randomized variation of the embedding method , which has been applied by leighton @xcite , richter and thomassen @xcite , shahrokhi et al .
@xcite , szkely @xcite , and most recently by fox and tth @xcite . for every missing edge @xmath101
, @xmath102 is the deleted main edge of a @xmath35-earring @xmath103 .
so there are @xmath36 edge - disjoint paths in @xmath91 from @xmath104 to @xmath105 . for each of these paths , draw a curve from @xmath104 to @xmath105 infinitesimally close to that path , on either side .
call these @xmath36 curves _
potential @xmath106-edges _ and call the resulting drawing @xmath107 . note that a potential @xmath106-edge crosses itself if the corresponding path does . in such cases , we redraw the potential @xmath106-edge in the neighborhood of each self - crossing to get a noncrossing curve . to get a drawing of @xmath3 , for each @xmath108 , choose one of the @xmath36 potential @xmath106-edges at random , independently and uniformly , with probability @xmath109 , and draw the edge @xmath106 as that curve .
there are two types of new crossings in the obtained drawing of @xmath3 .
first category crossings are infinitesimally close to a crossing in @xmath99 , second category crossings are infinitesimally close to a vertex of @xmath91 in @xmath99 .
the expected number of first category crossings is at most @xmath110 indeed , for each edge of @xmath5 , there can be at most one new edge drawn next to it , and that is drawn with probability at most @xmath109 .
therefore , in the close neighborhood of a crossing in @xmath99 , the expected number of crossings is at most @xmath111 .
see figure [ fig : embedding](a ) .
@xmath112 in order to estimate the expected number of second category crossings , consider the drawing @xmath107 near a vertex @xmath105 of @xmath91 . in the neighborhood of vertex @xmath105 we have at most @xmath113 original edges .
since we draw at most one potential edge along each original edge , there can be at most @xmath113 potential edges in the neighborhood .
each potential edge can cross each original edge at most once , and any two potential edges can cross at most twice . see figure [ fig : embedding](b ) .
therefore , the total number of first category crossings in @xmath107 in the neighborhood of @xmath105 is at most @xmath114 .
( this bound can be substantially improved with a more careful argument , see e. g. @xcite , but we do not need anything better here . ) to obtain the drawing of @xmath3 , we keep each of the potential edges with probability @xmath109 , so the expected number of crossings in the neighborhood of @xmath105 is at most @xmath115 , using the fact that the self - crossings of the potential @xmath37-edges have been eliminated
. therefore , the total expected number of crossings in the random drawing of @xmath3 is at most @xmath116 .
there exists an embedding with at most this many crossings , therefore , by lemma [ lemma_ft ] we have @xmath117 it follows that @xmath118 so @xmath119 consequently , @xmath120
in the statement of our theorem we can not require that _ every _ subgraph @xmath5 with @xmath121 edges has crossing number @xmath122 , instead of just _ one _ such subgraph @xmath5 .
in fact , the following statement holds . roughly speaking
, @xmath124 will be the disjoint union of a large graph @xmath130 with low crossing number and a small graph @xmath131 with large crossing number .
more precisely , let @xmath132 be a disjoint union of graphs @xmath133 and @xmath134 , where @xmath5 is a disjoint union of @xmath135 complete graphs , each with @xmath136 vertices and @xmath137 is a complete graph with @xmath138 vertices .
we have @xmath139 and @xmath140 , since @xmath141 . by the crossing lemma ( see e. g. @xcite ) , @xmath142 , but @xmath143 , because @xmath144 . in the preliminary version of this paper
@xcite we conjectured that we can require that a positive fraction of all subgraphs @xmath5 of @xmath3 with @xmath121 edges has crossing number @xmath122 .
the following construction shows that the conjecture does not hold in general for graphs with less than @xmath145 edges .
[ prop_4/3 ] for every @xmath146 and @xmath147 there exist graphs @xmath124 with @xmath125 vertices and @xmath126 edges with the following property .
let @xmath130 be a random subgraph of @xmath124 such that we choose each edge of @xmath124 independently with probability @xmath148 .
then as in proposition [ prop_1 ] , the idea is to build the graph @xmath132 from two disjoint graphs @xmath150 and @xmath137 , where @xmath150 is a large graph with low crossing number and @xmath137 is a small graph with high crossing number . in addition , deleting a random constant fraction of edges from @xmath137 will break all the crossings in @xmath137 with high probability .
now we describe the constructions more precisely .
let @xmath151 be a constant such that @xmath152 and @xmath153 .
let @xmath150 be a disjoint union of @xmath135 complete graphs , each with @xmath154 vertices ( we omit the explicit rounding to keep the notation simple ) .
we have @xmath155 and @xmath156 . the graph @xmath137 consists of five _ main _ vertices @xmath157 and @xmath158 internally vertex disjoint paths of length @xmath159 connecting each pair @xmath160
. the graph @xmath137 has @xmath161 vertices and @xmath162 edges .
we claim that @xmath163 .
the upper bound follows from the fact that the crossing number of @xmath164 is @xmath165 .
we take a drawing of @xmath164 with one crossing and replace each edge @xmath13 by @xmath158 paths drawn close to @xmath13 . for
the lower bound , take a drawing of @xmath137 minimizing the number of crossings .
let @xmath166 be a path with the minimum number of crossings among the paths connecting @xmath105 and @xmath167 .
by redrawing all the other paths connecting @xmath105 and @xmath167 along @xmath166 the crossing number of the drawing does not change .
the paths @xmath166 together form a subdivision of @xmath164 , therefore at least one pair @xmath168 of the paths crosses . due to the redrawing , every path connecting @xmath105 and @xmath167 crosses every path connecting @xmath169 and @xmath170 , which makes @xmath171 crossings . by the choice of @xmath172
, we have @xmath173 , therefore @xmath174 and @xmath175 .
let @xmath5 be a random subgraph of @xmath3 where each edge of @xmath3 is taken independently with probability @xmath148 .
let @xmath176 .
we show that with high probability , @xmath177 is a forest , in particular @xmath178 .
this happens if at least one edge is missing from every path connecting two main vertices of @xmath137 .
the probability of such an event is at least @xmath179 it follows that with this probability , @xmath180 .
note that in the previous construction the number @xmath181 does not have to be constant : it is enough to delete a random @xmath182 fraction of the edges to get the same conclusion with probability almost @xmath165 .
the question whether deleting a small random constant fraction of the edges of a graph @xmath3 decreases the crossing number only by a small constant fraction remains open for graphs with more than @xmath183 edges .
we do not know the answer even to the following weaker version of the question .
let @xmath184 and @xmath185 be constants .
does there exist @xmath186 and @xmath41 such that for every graph @xmath3 with @xmath187 and @xmath188 , a random subgraph @xmath5 of @xmath3 with each edge taken with probability @xmath36 has crossing number at least @xmath189 , with probability at least @xmath190 ?
the graphs in proposition [ prop_4/3 ] have small number of edges responsible for almost all the crossings .
is this the only way how to force a random subgraph of @xmath3 to have crossing number @xmath191 ?
let @xmath192 .
does there exist @xmath41 and @xmath181 such that every graph @xmath3 with @xmath193 and @xmath26 has a subset @xmath194 of @xmath195 edges such that every subgraph @xmath5 of @xmath3 with @xmath196 and @xmath197 has @xmath198 ?
j. ern , j. kynl and gza tth , improvement on the decay of crossing numbers , _ proceedings of the 15th international symposium on graph drawing ( graph drawing 2007 ) , lecture notes in computer science _ 4875 , 2530 , springer , berlin , 2008 .
j. pach , r. radoii , g. tardos , g. tth : improving the crossing lemma by finding more crossings in sparse graphs , _ proc .
19th acm symposium on computational geometry 2004 _ , 68 - 75 .
also in : _ discrete and computational geometry _
* 36 * ( 2006 ) , 527552 . | we prove that the crossing number of a graph decays in a `` continuous fashion '' in the following sense . for any @xmath0
there is a @xmath1 such that for a sufficiently large @xmath2 , every graph @xmath3 with @xmath2 vertices and @xmath4 edges , has a subgraph @xmath5 of at most @xmath6 edges and crossing number at least @xmath7 .
this generalizes the result of j. fox and cs .
tth .
department of applied mathematics , + charles university , faculty of mathematics and physics , + malostransk nm .
25 , 118 00 praha 1 , czech republic ; + ` [email protected] ` + + department of applied mathematics and institute for theoretical computer science , + charles university , faculty of mathematics and physics , + malostransk nm . 25 , 118 00 praha 1 , czech republic ; + ` [email protected] ` + + rnyi institute , hungarian academy of sciences , budapest ; + ` [email protected] ` |
both bars and central mass concentrations ( hereafter cmcs ) are features that are present in the majority of disc galaxies .
thus , bars harbouring a cmc should be the rule rather than the exception and it is important to study the effects of the latter on the evolution of the former ( and vice versa ) .
observations in near infrared wavelengths , where the effect of dust is largely suppressed , reveal that between 70% and 90% of all disc galaxies have a bar component @xcite .
cmcs such as massive black holes , central discs , dense central stellar clusters are very frequent , too .
super - massive black holes have a mass of the order of @xmath2 , which correlates with the luminosity , the mass , and , in particular , the velocity dispersion of the hosting bulge @xcite .
such correlations lead to a ratio of super - massive black hole mass to bulge mass of the order of @xmath3 ( see also * ? ? ?
the mass of the cmc is in fact larger than that of the black hole , since the black hole pulls inwards the bulge and disc material in its immediate neighbourhood and thus forms a cusp , which contributes to the cmc mass @xcite . compact massive central discs , stellar and/or gaseous , are also examples of cmcs .
gaseous such discs are expected to form particularly frequently in barred galaxies , due to the gas inflow driven by the bar @xcite .
indeed , observations have shown that large concentrations of molecular gas can be found in many disc galaxies , and particularly in barred ones @xcite .
the mass of such components is of the order of @xmath4 , while their radial extent varies from several tens of parsecs to a couple of kpc .
discy bulges @xcite are examples of mainly stellar such discs ( * ? ? ?
* and references therein ) .
cmcs can also be central dense stellar clusters , with masses of @xmath5 .
orbital - structure studies , aimed at understanding the effect of the cmc on the bar - supporting orbits , were made first in 2d @xcite and then extended to 3d @xcite .
they include the calculation of the main families of periodic orbits and of their stability .
the latter is particularly important , since stable periodic orbits trap around them regular orbits , while unstable periodic orbits generate chaos .
these studies showed that with increasing cmc mass or concentration the unstable fraction of the bar - supporting orbits increases , reducing the possibility for barred equilibria .
orbital - structure studies have thus made the first crucial step in our understanding of the effect of a cmc on a bar by revealing the mechanism by which a cmc could destroy a bar .
these studies can inform us whether a given bar model ( i.e. density profile and pattern speed ) may co - exist with a given cmc .
however , the parameter space of bar models is huge and an extensive exploration is neither possible nor meaningful , because nothing will be learned about the _ evolution _ of a given bar in presence of a growing cmc .
such information can only be achieved by appropriate @xmath0-body simulations , which were pioneered by @xcite ( ) .
these authors used particles on a grid to represent the ( barred ) disc , which was exposed to the gravitational field of the cmc and of a spheroidal ( bulge - like ) component . for both 2d and 3d simulations ,
nsh found that a cmc of mass 5% of that of the galaxy is sufficient for bar dissolution . in the @xmath0-body simulations of @xcite ( (
* hereafter ss ) ) the disc is represented by particles on a 3d cylindrical polar grid and , similarly to nsh , exposed to the gravity of the cmc and of a rigid spherical halo component .
these authors found that a cmc mass of a few percent is necessary to destroy the bar , by and large , consistent with the results of nsh .
more precisely , in the models displayed in their fig . 5 , a cmc mass of approximately 4% that of the disc was necessary for a hard ( compact ) cmc and more than 10% for a soft ( extended ) cmc .
they compared two models with different bar strengths and concluded that the strength of the bar hardly influences these numbers .
recently , @xcite ( ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* hereafter hh ) ) re - examined the problem , solving the poisson equation by expanding the density and the potential in a set of bi - orthonormal basis functions .
their approach is 2d , which implies the necessity for a rigid halo potential , similar to nsh & ss .
these authors find that a cmc of mass 0.5% of the disc mass is sufficient to destroy the bar .
thus , their limiting cmc mass is an order of magnitude lower than that given by nsh .
the difference between their results and those of nsh&ss is surely significant . indeed ,
if the mass necessary to destroy a bar is as small as claimed by hh , then the masses of cmcs observed in disc galaxies are sufficient to dissolve observed bars in relatively short time scales . the opposite is true if the nsh & ss picture is correct .
this disagreement between previous studies , as well as their lack of realism in modeling the galactic halo , prompted us to revisit the problem .
hh argued that the differences between the two sets of simulations are due to the different initial density distributions in the disc , since they use an exponential disc , while nsh&ss employed a kuzmin - toomre disc @xcite .
hh argue that , since the latter is more centrally concentrated than the former , the cmc could influence a larger fraction of the disc mass and thus achieve bar dissolution easier .
however , preliminary simulations @xcite showed that bars are rather robust , and so we opted for an exponential disc model which is known to represent well the observed light distribution of disc galaxies @xcite .
this may give a better chance for the cmc to destroy the bar and allows us to check whether indeed the initial density distribution in the disc is of crucial importance for the result .
we employ a poisson solver different from both types previously used in this problem , namely a tree code described in section [ subsec : code ] .
unlike the previous studies , we use a live halo in our simulations , which makes them _ fully _ self - consistent .
this has been shown to be essential for bar formation in presence of a strong halo component .
indeed , @xcite ( ( * ? ? ?
* hereafter a02 ) ) compared two simulations with initially identical disc components and haloes of identical mass distribution . in one case
the halo was rigid ( represented as an imposed spherical potential ) and in the other it was composed of about @xmath6 particles .
the difference between the evolution in the two cases is stunning ( fig . 1 of a02 ) .
in the first case , the disc formed only a minor oval confined to the centre - most parts of the disc , while in the second case it grew a strong bar with very realistic observed properties ( * ? ? ?
* hereafter am02 ) .
this important difference is due to the fact that the angular momentum absorption by the halo is artificially suppressed in one of the two cases , while in the other it is allowed to freely take place ( for a comprehensive discussion of the effect of the halo on the angular momentum exchange see * ? ? ?
* hereafter a03 ) .
the paper is organised as follows . in section [ sec : numerical ] we present our numerical methods and our simulations .
sections [ sec : before_cmc ] and [ sec : after_cmc ] describe the evolution before and after the introduction of the cmc , respectively .
we discuss our results and conclude in section [ sec : discussion ] .
we first follow the formation and evolution of a bar in two bar - unstable discs .
we choose a time after the initial fast bar formation period and during the phase of slow secular evolution . at this time
we start growing a cmc in the centre of the disc .
the various steps in this work are described in the following subsections .
the models are composed of a disc and a halo component .
the disc has an initial volume density profile @xmath7 where @xmath8 is the cylindrical radius , @xmath9 is the disc mass and @xmath10 and @xmath11 are the disc radial and vertical scale - lengths , respectively and @xmath12 axes are on the disc equatorial plane and the @xmath13 axis is perpendicular to that plane . in all figures the bar will be along the @xmath12 axis . ] .
the initial halo density profile is @xmath14 where @xmath15 is the radius , @xmath16 is the halo mass and @xmath17 and @xmath18 are the halo scale - lengths .
the former can be considered as the core radius of the halo .
the constant @xmath19 is defined by @xmath20^{-1},\ ] ] where @xmath21 @xcite .
information on how the initial conditions were set up can be found in @xcite and am02 .
both simulations have @xmath22 , @xmath23 , @xmath24 , @xmath25 and @xmath26 .
the first one has a halo with a small core ( @xmath27 ) and is thus , following the definition of am02 , mh - type .
the second one has a halo with a large core ( @xmath28 ) and is thus of md - type ( am02 ) .
the discs of the two simulations have @xmath29 and 1 , respectively , where @xmath30 is the stability parameter introduced by @xcite .
our model units of mass and length are simply the disc mass and disc scale length . in order to convert them to @xmath31 and kpc ,
we need to set the units of length and mass to values representative of the object in consideration , a barred galaxy .
am02 used a length unit of 3.5 kpc and a mass unit of @xmath32 . with @xmath33 ,
this implies that the model unit of velocity is 248kms@xmath34 and the model unit of time is @xmath35yr .
thus , time 400 corresponds to @xmath36yr .
this choice , however , is in no way unique , and we can choose quite different values .
for example , we can use a smaller length unit since the scale - length of the disc increases with time @xcite and thus the initial length unit could be chosen to be 2 kpc .
because of this rather wide range of possibilities , we will restrict ourselves to model units and leave it to the readers to convert units according to their needs .
for the simulations described in this paper , we used the publicly available @xmath0-body code , which employs the tree - based force solver @xcite .
unlike the standard @xcite tree code , which has complexity @xmath37 , uses cell - cell interactions , has complexity @xmath38 , and is about ten times faster at comparable accuracy .
some of the simulations were run with a proprietary version of the code , which uses the sse instruction set supported on x86 chips .
these instructions allow a substantial speed - up for the computation of body - body forces via direct summation .
this method is used within and between small cells and results in a gain of a factor of @xmath39 for this part of the code . the total gain for the force computation for @xmath40 bodies ( including the tree build )
is about 25% . in all @xmath0-body simulations
it is necessary to use a softening kernel in order to avoid excessive noise and diverging forces .
a @xcite softening , which has density kernel @xmath41 , has been used in most simulations made so far .
this , however , has the disadvantage that the softened force converges relatively slowly to the newtonian force @xcite , although the difference becomes negligible , and beyond any numerical importance , at large radii .
here we adopt a softening with a density kernel @xmath42 @xcite , which converges much faster to the accurate newtonian force .
this is particularly interesting for the problem at hand , as will be discussed in the last section .
we adopted a softening length @xmath43 of 0.03 , which roughly corresponds to a plummer softening length of 0.02 ( in the sense that the maximum two - body force is the same ) .
a small softening length is useful in this problem , since the cmc will create an increased central concentration of the disc and halo components , and more centrally concentrated structures necessitate a smaller softening length @xcite . the tree code used in our simulations allows for an approximate interaction ( as opposed to the direct summation force ) between two tree nodes ( cells or single bodies ) only if their critical spheres do not overlap @xcite .
the critical spheres are centred on the centre of mass of the nodes and have radii @xmath44 , where @xmath45 is the radius containing all particles in the node and @xmath46 the tolerance parameter ( opening angle ) . in our simulations we use @xmath47 . in order to have sufficiently small time steps in the area around the cmc
, we used adaptive time stepping with the block - step scheme ( time steps differ by factors of two and are hierarchically nested ) . in most simulations
we used 6 levels , in such a way that the longest time step is @xmath48 and the shortest @xmath49 .
for the numerically more demanding simulations ( with a cmc of very high mass or very small radius ) we used 7 levels and a smallest time step of @xmath50 .
the time step levels are adapted in an ( almost ) time symmetric fashion to be on average @xmath51 where @xmath52 and @xmath53 are constants , @xmath54 is the modulus of the acceleration and @xmath55 is the potential .
we adopted @xmath52 = 0.01 and @xmath53 = 0.015 , thus ensuring that the smallest time step bin contains of the order of a few percent of the particles .
we model the disc with 200000 and the halo with roughly 1000000 particles . with these settings ,
the energy conservation after the cmc growth ( during cmc growth energy is not conserved ) was always better than one part in 1000 , and , in most cases , of the order of a couple of parts in 10000 .
following previous studies ( nsh&ss ) , we model the cmc by a plummer sphere of potential @xmath56 where @xmath57 is the mass of the cmc at time @xmath58 and @xmath59 its radius . in order to avoid transients ,
the cmc is introduced gradually with @xmath60 where @xmath61 and @xmath62 is the time at which the cmc is introduced @xcite .
thus the mass , as well as its first and second derivatives , are continuous functions of @xmath58 .
the radius @xmath59 of the cmc is kept constant with time . in the following ,
since we are interested in the time evolution after the cmc is introduced , we measure time from the time when the cmc has been introduced , i.e. from @xmath62 . some of the software used in the analysis has already been described in am02 and we refer the reader to that paper for more information . our runs are listed in table [ tab : initcond_cmc ] .
the first column gives the name of the run and the second one the type of the model .
the mass of the cmc , @xmath63 , its radius , @xmath59 , and its growth time , @xmath64 , are given in the third , fourth and fifth columns , respectively .
[ tab : initcond_cmc ] .parameters for the cmc [ cols="<,<,<,<,<",options="header " , ]
the formation and evolution of the bar in @xmath0-body simulations has already been extensively described in the literature . here
we will only summarise very briefly some results from am02 , a02 , and a03 , which are necessary for discussing the simulations at hand , and will refer the interested reader to these papers for more information ( see also * ? ? ?
barred galaxies evolve by re - distributing their angular momentum .
this is emitted mainly by near - resonant material in the inner disc and absorbed by near - resonant material in the outer disc and in the halo .
the amount of angular momentum that can be emitted and absorbed depends on the amount of near - resonant material and on how dynamically hot / cold this is .
since there is relatively little material in the outer disc , the halo can play a major role in this exchange , even though the material that constitutes it is relatively hot .
thus , bars immersed in massive haloes can grow stronger than bars immersed in weak haloes , and very much stronger than bars immersed in rigid haloes
. models where the halo plays a substantial role in the dynamics within the inner 4 or 5 disc scale lengths , while being also sufficiently cool to absorb considerable amounts of angular momentum were termed mh ( for massive halo ) in am02 .
such models form strong bars . on the contrary ,
models in which the disc dominates the dynamics within that region form less strong bars and were termed md ( for massive disc ) in am02 . due to the considerably different dynamics in the two cases
, we decided to study the effect of a cmc in two models , one of mh - type and one of md - type .
the observable properties of the bars which grow in these two types of models are quite different ( see am02 for a more complete discussion ) .
mh - type bars are stronger ( longer , thinner and more massive ) than md - type bars .
viewed face - on , they have a near - rectangular shape , while md - type bars are more elliptical . viewed side - on , they show stronger peanuts and sometimes ( particularly towards the end of the simulation ) even ` x ' shapes . on the other hand ,
bars in md - type models are predominantly boxy when viewed side - on .
[ fig : faceon_prof ] shows the projected density radial profiles of the two types of bars seen face - on .
the results are in good agreement with what was found by am02 ( see their fig . 5 and section 5 ) .
the radial profile along the major axis of the md bar decreases with radius without showing any clear change of slope at the end of the bar . on the other hand , for the mh type the profile along the bar major axis is quite different .
it has a relatively flat part in the bar region which is followed by a sharp drop at the end of the bar .
these two different types of profiles have been also found by observations of barred galaxies .
@xcite , using blue and near - infrared surface photometry of a sample of 15 barred galaxies , classified the photometric profiles in the bar region in two classes : the flat ones and the exponentially decreasing ones , i.e. the two types found in our two fiducial simulations .
the importance of these two types of profiles for bar destruction will be discussed in the last section .
mh - typemd - type + + the effect of a cmc on a mh - type bar is shown in the left panels of fig .
[ fig:3v_mhmd ] .
the upper left panel shows the three views of the disc component before the cmc is introduced .
the face - on view shows a strong bar with ansae at its ends , similar to what is seen in many early types strongly barred galaxies ( good examples can be seen on pages 42 and 43 of the hubble atlas , * ? ? ?
seen side - on , it shows a clear peanut , or mild ` x ' , shape , whose extent is somewhat shorter than the bar , as expected ( for a discussion see * ? ? ?
seen end - on , the bar can easily be mistaken for a bulge and can only be distinguished from it by kinematical measurements @xcite .
the middle and lower panels on the left of fig .
[ fig:3v_mhmd ] show the disc component a @xmath65 after the cmc has been introduced . with the units proposed in section [ subsec : initcond ] ,
this is equivalent to 5.6gyrs .
the mass of the cmc in these two cases is 0.05 and 0.1 , respectively and @xmath66 .
we note that the smaller of the two masses did not destroy the bar , but made it considerably less strong .
in particular , the final bar is considerably shorter .
since its width has not changed considerably , its axial ratio ( minor to major axis ) has become considerably larger .
the heaviest of the two cmcs has also not destroyed the bar , although its mass is as much as 10% of that of the disc .
the resultant bar , however , is rather short and oval .
a cmc with a mass of 0.2 is sufficiently massive to destroy the bar and make the disc near - axisymmetric ( not shown here ) .
@xmath67 @xmath68 @xmath69 the cmc changes also the vertical structure of the bar . before the introduction of the cmc and seen side - on
, the bar has a peanut - shape , or mild ` x'-form , with a maximum vertical extent at a distance from the centre roughly equal to three quarters of the bar length and a minimum in between the two maxima .
the middle left panel , corresponding to @xmath68 , show that the bar seen side - on has a boxy shape , with a roughly constant vertical extent .
the radial extent of this boxy feature is roughly the same as that of the peanut before the cmc was introduced .
thus , although the cmc brought a considerable decrease of the bar length seen face - on , it brought little , if any , decrease of the radial extent of the vertically extended feature .
the more massive cmc , @xmath69 , brings a stronger change to the vertical structure .
seen side - on , the shape is more elliptical than boxy .
so the introduction of a cmc changes the side - on outline from peanut to boxy , or , for very massive cmcs , to elliptical .
the radial extent of these features , however , does not change much . in the end - on view
also , the cmc makes changes in the outline .
the vertical extent increases in the central region and also the bulge - like feature becomes considerably more extended radially .
the right panels of fig . [ fig:3v_mhmd ] are similar , but for an initially md - type bar .
before the cmc is introduced the bar is both fatter and shorter , i.e. less strong than that of the previous example , in agreement with the results of am02 .
it also has no ansae .
seen side - on , it has a boxy structure , but no peanut .
the middle and lower right panels show the disc component a @xmath65 after the cmc has been introduced .
the mass of the cmc in these two cases is again 0.05 and 0.1 , respectively and @xmath66 .
we note that the introduction of the cmc makes major changes .
seen face - on , the bar has nearly disappeared even for @xmath68 , where only close scrutiny allows one to see that the isophotes are not circular , but slightly elongated in what used to be the bar region . for @xmath69 , there is no trace of the bar anymore , the isophotes being circular within the measuring errors .
seen side - on , the boxiness has more or less disappeared for @xmath68 , and there is definitely no trace left by @xmath69 .
thus , judging from the morphology of the three orthogonal views , the models with @xmath69 or 0.05 , qualify as an sa galaxy , i.e. a galaxy with no bar .
thus , even moderately light cmc can destroy the bar in md - type models .
this is contrary to what was found above for mh - type models and constitutes an important difference between the two . to check whether a less massive cmc can still destroy the bar we ran four more simulations with @xmath63 equal to 0.04 , 0.03 , 0.02 and 0.01 , respectively . in all four cases
the bar is clearly visible by the end of the simulation , although considerably weakened .
we also ran simulations with @xmath63 = 0.1 and @xmath70 to check whether the bar is destroyed even for cmcs with such large radii .
we found that @xmath59 has to be at least as large as 0.1 for the bar not to be destroyed .
figure [ fig:2damplitude_mhmd ] gives , for six simulations , the amplitude of the @xmath71 ( bisymmetric ) component of the mass / density distribution as a function of both time and radius .
this is a simple measure of the bar strength .
the results for higher @xmath1 values will be discussed in the next subsection .
the left panels correspond to two reference simulations in which no cmc has been introduced .
they show clearly a secular evolution , during which the bar gets slowly and steadily stronger .
we note that the maximum amplitude increases , but also that both the innermost and the outermost parts become gradually more bisymmetric .
the picture is different for simulations with a cmc . at early times ,
up to 20 or 30 time units after the cmc has been introduced , the evolution continues more or less as if no cmc had been introduced .
this is presumably because the cmc has to reach a limiting mass before its effect can be felt by the bar and/or because the orbits in the bar region take some time to adjust .
this time interval is longer in the case of less massive cmcs .
after that starts a second time interval , which lasts roughly till the time the cmc reaches its maximum mass . during this second phase the amplitude drops abruptly with time .
a third time interval starts after that and lasts till the end of the simulation .
unless the bar has totally disappeared after the end of the second interval , its amplitude continues to decrease in this third phase , but at a considerably slower rate than in the second one .
the distinction between the second and the third time interval is sharper for md - type than for mh - type models and for more massive cmcs .
figure [ fig:2damplitude_mhmd ] shows clearly that the maximum amplitude decreases steadily with time .
but also that both the innermost and the outermost parts of the disc become gradually more axisymmetric .
this is in agreement with what is seen in figure [ fig:3v_mhmd ] .
indeed , it was clear there that at @xmath72 after the cmc was introduced the bar has become shorter and also that the innermost isophotes have become rounder .
the corresponding time evolution can be followed in figure [ fig:2damplitude_mhmd ] .
comparison of the upper and lower panels of fig .
[ fig:2damplitude_mhmd ] shows that a given cmc can destroy md - type bars more efficiently than mh - type ones .
also comparison of the central and right rows of panels shows that more massive cmcs are more efficient for bar destruction , as expected .
[ fig : mcomponmh ] shows the amplitude of the relative @xmath1 = 2 , 4 , 6 , and 8 fourier components of the mass / density for mh - type simulations .
they have been calculated as described in am02 . before the introduction of the cmc
( left panel ) all components have high amplitudes .
in particular , their maxima are 0.7 , 0.48 , 0.33 and 0.25 , for @xmath1 = 2 , 4 , 6 and 8 , respectively .
the introduction of the cmc brings a considerable decrease of all these values .
thus , for a @xmath63 = 0.05 and @xmath73 = 400 the maxima of the amplitudes for the various @xmath1s fall to 0.54 ( @xmath1 = 2 ) , 0.43 ( @xmath1 = 4 ) , 0.32 ( @xmath1 = 6 ) and 0.25 ( @xmath1 = 8) of the initial maximum values , respectively . for @xmath63
= 0.1 and the same time lapse the @xmath1 = 2 amplitude falls to 0.3 of its initial value , while the other components drop to the noise level .
[ fig : mcomponmh ] quantifies the effect of the cmc and , more importantly , shows that it has a stronger effect on the higher @xmath1 components .
the latter can be understood as follows : as can be seen from the left panel of fig .
[ fig : mcomponmh ] , before the cmc is introduced the radius at which the relative amplitude has its maximum increases with @xmath1 and comes closer to the outer regions of the bar .
these regions , as we saw in section [ subsec : existence ] and [ subsubsec : bstrengthtr ] , are affected first and tend faster to axisymmetry . in this way
, the higher @xmath1 components will be affected before the lower ones .
this argument is further re - enforced when we look at the positions of the maxima . for @xmath71
the maximum is at roughly 2.1 and stays there after the cmc has been introduced , as can be seen by comparing the left panel to the middle and right ones . on the other hand
the maximum of the @xmath74 , and 6 components is roughly at 2.4 at the time the cmc is introduced , and moves to 2.1 , i.e. the same radius as that of the @xmath71 maximum , 400 time units after a cmc with @xmath68 is introduced . since the higher @xmath1 components are affected less than the lower @xmath1 ones , the iso - density contours become fatter and/or less rectangular ( and thus more elliptical ) . fig .
[ fig:3v_mhmd ] confirms this .
[ fig : mcomponmd ] gives similar information , but for md - type simulations .
as already discussed in am02 , md - type models have lower values of all the amplitudes and also lower values of the relative amplitude of the higher @xmath1 components than mh - type models .
thus , for md - types , the cmc destroys the higher @xmath1 components even for relatively small values of the cmc mass .
this is illustrated in fig .
[ fig : mcomponmd ] , where we show results for two models with @xmath63 equal to 0.01 and 0.02 , respectively . fig .
[ fig : sigma_mh ] shows the three components of the velocity dispersion as a function of distance from the centre .
they have been calculated as described in am02 .
namely , we isolate a thin strip of particles centred on the bar major axis and having a width of 0.07 .
the distance from the centre is calculated along this strip . in this way
, we avoid making an integration along the line of sight , since this would impose the choice of a particular viewing angle .
we also bring out clearer the information on the motions of particles near the bar major axis , which is of particular use when comparing with periodic orbit structure .
nevertheless , any comparison with observations can only be rough , since our presentation gives information only on a subset of the orbits . at the time the cmc is introduced
, the velocity dispersions of model mh show features characteristic of mh - type models ( see am02 for a more complete description ) .
the @xmath75 component ( perpendicular to the bar ) has a clear maximum at the centre and two secondary maxima on either side of it .
these can be due to orbits trapped around x@xmath76 periodic orbits with loops at their apo - centres , and/or on the superposition of orbits trapped around elliptical and around rectangular shaped periodic orbits .
the cmc turns these maxima into plateaus and moves them nearer to the centre .
both effects could be expected .
indeed these maxima are only found for strong bars ( am02 ) , while the cmc has reduced the bar strength .
also these structures are linked to the end of the bar and the cmc makes this shorter ( sections [ subsec : existence ] and [ subsubsec : bstrengthtr ] ) .
of course the clearest effect of the cmc is to raise very substantially the central velocity dispersion , and more strongly so for the most massive cmcs . at the time the cmc is introduced
, the @xmath77 component has a small , but characteristic , minimum at the centre , surrounded closely by two maxima on either side . since the @xmath77 is the component parallel to the bar major axis , this feature should be mainly seen in edge - on barred galaxies which are viewed nearer to end - on than to side - on .
the cmc obliterates this feature and creates , instead , a clear central maximum , as for @xmath75 .
the @xmath75 and @xmath77 components have quite different values at the centre in the case with no cmc . as expected , however , the velocity distribution in the central regions becomes more isotropic in the cases with cmc , and more strongly so for the most massive cmcs .
the md - type models show no clear characteristic features on their velocity dispersion profiles ( see fig .
13 of am02 ) .
thus the cmc only introduces a sharp maximum at the centre , clear in all three components of the velocity dispersion , as was seen already in the mh - type models .
we ran two simulations with smaller values of @xmath59 ( see table [ tab : initcond_cmc ] ) and found that more centrally concentrated cmcs are more efficient for bar destruction , thus confirming a result already found by @xcite , with orbital calculations , and by ss , with simulations .
we also ran two simulations with a larger value of @xmath64 ( see table [ tab : initcond_cmc ] ) and confirmed a result already found by ss and by hh , namely that the value of @xmath64 does not influence much the final bar weakening .
it only influences the duration of the initial strong decrease .
the transition between the initial strong decrease and the later milder one is sharper in simulations with larger @xmath63 and/or smaller @xmath59 ( see bottom panels fig . [
fig : omegap_sb_md ] ) .
nsh mentioned that the introduction of the cmc results in an increase of the bar pattern speed . in order to assess further the effect of the cmc on the bar pattern speed
, we measured this quantity in our simulations .
this was not always easy , since in many cases the bar amplitude is so weak that the measurement was either not possible , or gave a very uncertain value .
in most cases , some smoothing was necessary .
nevertheless , from the cases where the measurement is sufficiently precise , we can clearly see that the pattern speed _ increases _ with time , as the bar gets weaker .
this is shown , for two sets of md - type simulations , in fig .
[ fig : omegap_sb_md ] and for mh - type simulations in fig .
[ fig : omegap_sb_mh ] .
for comparison , we also plot a measure of the bar strength as a function of time .
as such we have chosen the maximum value of the amplitude of the @xmath1 = 2 component of the mass .
results found with other measures of the bar strength are similar .
the left panels of fig .
[ fig : omegap_sb_md ] give results from a series of simulations with different values of the mass of the cmc , namely simulations md@xmath78
( @xmath63 = 0.01 ) , md@xmath79 ( 0.02 ) , md@xmath80 ( 0.03 ) , md@xmath81 ( 0.04 ) and md@xmath76 ( 0.05 ) . in the very first 20 time units or so , the bar strength increases , as in simulations with no cmc . during this time
, at least in simulations with small cmcs , the pattern speed decreases , again as in simulations with no cmc .
this is presumably due to the fact that the cmc has not grown enough to influence the evolution sufficiently to reverse the trend .
this reversal indeed occurs at somewhat later times .
we note that the bar speed increases noticeably up to time 100 , while the cmc mass increases and the bar strength decreases most strongly . at later times there may still be an increase in the pattern speed , but it is very small and the data are too noisy to assess it with any certainty .
the right panels fig .
[ fig : omegap_sb_md ] give similar results , now from a series of simulations with different values of the radius of the cmc , namely simulations md@xmath82
( @xmath59 = 0.01 ) , md@xmath83 ( 0.02 ) , md@xmath84 ( 0.05 ) , md@xmath85 ( 0,08 ) and md@xmath86 ( 0.1 ) . for the value of @xmath63 used in this sequence
( @xmath63 = 0.01 ) and the smallest of our @xmath59 , the bar is practically destroyed by the time the mass of the cmc has reached its final value .
thus cmcs with a yet smaller radius give the same result .
the very short - lived initial increase of the bar strength is less clear here than in the previous set of simulations .
the basic result , however , stays the same .
[ fig : omegap_sb_mh ] shows similar plots , but now for mh - type simulations .
the results are similar , i.e. there is an increase in the pattern speed during the time the strength of the bar decreases sharply .
this increase is quite noticeable in cases where the decrease of the bar strength is important , and considerably less so in cases with a milder decrease of the bar strength ( as md@xmath78 or mh@xmath76 ) .
in this paper we studied the effect of a central mass concentration ( cmc ) on the bar that harbours it .
we find that the cmc leads to a decrease of the bar strength , which in some cases can be sufficiently important to lead to bar destruction .
more massive and/or more concentrated cmcs are more efficient . the effect of the cmc depends also on the type of the bar model .
strong bars , which we call mh - type , because they form in simulations with massive haloes and which are morphologically similar to @xcite s ( @xcite ) flat - profile bars , are less prone to destruction than weaker bars , which we call md - type , because they form in simulations with massive discs and which are morphologically similar to @xcite s exponential - profile bars . the effect of the cmc is strongest in the innermost and the outermost parts of the bar , where it makes the density distribution more axisymmetric .
this leads to shorter and fatter bars , whose iso - density contours are more elliptical than rectangular .
the cmc also affects the pattern speed of the bar , which grows noticeably while the mass of the cmc grows and the strength of the bar decreases strongly .
this is in good agreement with the anti - correlation between bar strength and pattern speed found in a03 ( see figures 16 and 17 in that paper ) .
the cmc also influences the velocity dispersions .
in particular , after the introduction of the cmc , the velocity dispersion profiles along the bar major axis show a strong central maximum , while the secondary maxima on either side of the centre , seen in the side - on projection of strong bars , become less strong and approach the centre . in order to understand why a given cmc can destroy a bar in some cases , while only weakening it in others , it is necessary to study the time evolution of the orbital structure in the two cases . for this , one needs to study the orbital structure at a sequence of times during the evolution by calculating the potential and pattern speed at these times and using the corresponding positions and velocities of the simulation particles ( as done e.g. in a02 and a03 ) .
the evolution of the orbital structure along this time sequence will give information on the evolution of the orbital structure during the simulation and should help explain the difference between the effects of a cmc on mh - type and on md - type bars .
this analysis is beyond the scope of this paper and will be the subject of a future paper .
it is possible , however , to make some remarks at this stage .
a mh - type bar which survived with a decrease of its amplitude was initially thinner , i.e. had a smaller ratio of minor to major axis , than a md - type bar which was destroyed .
this means that the orbits in the former are more eccentric than those in the latter @xcite , i.e. for a given orbital major axis , the particles in the former will come nearer to the cmc than in the latter .
thus , one would expect that the bar structure in the mh - type bar would be more affected than the bar structure in the md - type , and this is indeed what is found by @xcite in their orbital calculations .
furthermore , as can be seen in fig .
[ fig : faceon_prof ] , the density distribution in mh - type bars is more centrally concentrated than in md - types , so that the particles come closer to the cmc .
both effects suggest that mh - type bars could be destroyed more easily , and yet fully self - consistent simulations show the opposite .
it is not entirely clear what is wrong with this naive reasoning .
possibly , the orbital structure of planar bars , as studied by @xcite and hh , is a bad representation of the orbits actually populated , which extend to non - negligible heights .
indeed , @xcite have shown that the backbone of the bar is formed by the 3d families of the x@xmath76 tree , and not the 2d x@xmath76 family .
thus , the fact that the planar orbits are unstable does not necessarily imply that the bar has to be destroyed .
finally , the 3d orbital - structure study of @xcite is restricted to a single bar model and thus can not shed any light on the differences between mh - type and md - type bars .
additionally , the halo may play a role .
indeed , in galaxies with no cmc , mh - type haloes can absorb much more angular momentum from the bar than md - type haloes , thus causing the bar to grow stronger in mh - type galaxies ( a03 ) .
a similar effect could also occur in disc galaxies with a cmc .
namely , in galaxies with a cmc , the extra angular momentum that can be absorbed by the mh - type haloes could well cause the bar to decrease less in mh - type than in md - type galaxies , which is indeed what our simulations show .
this alternative could also explain why ss find only mild differences between their strong and their weak bar .
indeed , if the difference is due to the halo , it would not show in these simulations which have a rigid halo .
there is good qualitative agreement between the results of all @xmath0-body studies so far , namely nsh ( @xcite ) , ss ( @xcite ) , hh ( @xcite ) , and our study .
first and foremost , all find that the growing cmc results in a decrease of the bar strength .
the disagreement is more quantitative than qualitative and concerns whether the decrease in bar amplitude is sufficiently strong to destroy the bar .
there are further agreements , in that the pattern speed increases as the bar strength decreases ( nsh & this study ) and that more massive and/or more centrally concentrated cmcs weaken the bar more than less concentrated ones ( all studies ) .
all the above are also in agreement with orbital - structure studies @xcite . finally , there is agreement that @xmath64 does not influence much the final bar weakening , but only influences the rate at which the bar strength decreases .
quantitative comparisons are much more difficult to make , mainly because of the differences between the various studies , affecting ( 1 ) the bar and ( 2 ) halo models ( life or rigid ) , ( 3 ) the poisson solvers and time steps , ( 4 ) the dimensionality of the motion ( 2d or 3d ) , and ( 5 ) the initial conditions . despite all these differences , nsh , ss , and our study all
agree in that a cmc mass of at least a few per cent of the disc mass is required to destroy a bar .
ss find that a cmc with @xmath87 and @xmath88 are necessary to destroy the bar , while for @xmath69 and @xmath89 the bar is weakened but not destroyed .
our fiducial @xmath59 is between these values and , for one of our bar models , the bar is destroyed and for the other it is considerably weakened for @xmath68 or 0.1 .
the most notable difference between these previous studies and ours is the treatment of the halo ( rigid vs. fully self - consistent ) , since ss used a rigid halo for computational economy . in support of the adequacy of this choice they run one simulation with a live halo and
compare it to a rigid halo case .
unfortunately the two runs are not very similar initially , so that an assessment of the importance of the halo response does not follow straightforwardly from a comparison .
indeed , fig . 8 in ss shows a difference in the initial bar amplitude of the same order as the difference between their strong and weak bar cases ( their fig . 5 ) .
moreover , their comparison is done for a halo whose density is low in the disc region , which will indeed restrict the influence of the halo on the bar evolution .
simulations with a dense halo would necessarily show a much stronger influence .
in particular , a rigid halo will not allow halo material to form a cusp around the cmc .
furthermore , the importance of a live halo for the correct description of bar evolution has been clearly demonstrated in some cases by comparing rigid and live halo simulations ( a02 ) . for these reasons , we have adopted a live halo in our simulations and believe this to be crucial for a correct astrophysical understanding .
ss studied two bars of different strength and found only mild differences . on the other hand
, we find significant differences between our md - type and mh - type bars : the former is destroyed for an @xmath69 and an @xmath66 , while the latter is only weakened .
this is not a real disagreement , since the difference between the weak and strong bar of ss is considerably less than the difference between our mh - type and md - type bars .
however , if the robustness of our mh - type bars originates ( at least partly ) from the interaction with its halo , simulations like those of ss may not be suitable for interpreting the evolution of galaxies with massive haloes .
there are notable differences between the two sets of studies with respect to their softening .
we use a softening of 0.03 ( i.e. 0.02 plummer equivalent ) , a fiducial @xmath59 of 0.01 and a range of @xmath59 values between 0.005 and 0.1 .
thus our fiducial @xmath59 is half the softening , while in most of our adopted range it is equal to or bigger than the softening .
we also use a softening kernel which decreases faster than the standard plummer one , thus limiting its effect at larger radii . to test further the effect of the softening
, we ran a series of simulations with larger softening , namely 0.06 and 0.09 ( i.e. equivalent plummer softening of 0.04 and 0.06 ) and found only very small differences , e.g. in the bar strength .
this argues strongly that our softening is adequate for the problem at hand .
ss have been bolder than us in the use of the softening .
they use a softening of 0.02 , for their strong bar case , or 0.05 for their weak bar case , while studying the effect of cmcs with @xmath59 ranging between 0.0003 and 0.1 , with a fiducial value of 0.001 .
this makes their softening 20 to 50 times larger than their fiducial @xmath59 , the ratio reaching , for their most concentrated cmc , roughly 67 for their strong bar case and 167 for their weak bar . in their figure 7
, they compare three simulations with softening 3 , 7 and 17 times the cmc radius respectively and find differences in the bar strength only of the order of , or smaller than , 10% .
the recent study by hh differs notably from both previous studies ( nsh & ss ) and from ours .
in particular , hh find that a cmc with 1% or even 0.5% of the disc mass is sufficient to destroy a bar ( for @xmath66 disc scale lengths ) .
it is not clear at this stage what actually causes this discrepancy .
hh blamed the different initial conditions , in particular the fact that they have an exponential disc , while nsh & ss use a kuzmin - toomre disc . our discs , however , are also initially exponential and yet our results are in agreement with those of ss , but not with those of hh .
we discuss below some other possible explanations of the discrepancy .
one alternative is that there is a subtle numerical difference between the two sets of simulations , in particular the resolution of the scf code and the time stepping . a poor resolution , corresponding to large effective softening ,
should , however , decrease the mass of any cusp forming around the cmc and hence reduce the destructive effect .
furthermore , hh have tested the number of terms they use and found it to be sufficient .
an insufficiently small time step may entail an artificial bar destruction .
both ss and this study use a variable time step , which is very small in the vicinity of the cmc , contrary to hh , who use a constant time step . in the same units as we use here ,
hh have a @xmath90 , or 0.005 , depending on the model .
this contrasts with the time step which we have adopted , which , in the vicinity of the cmc , is roughly 0.0005 or 0.0002 .
the time steps may be compared with the dynamical time at @xmath91 due to the cmc alone @xmath92 for a plummer model ( used by nsh , ss , and us ) and @xmath93 shorter for the cmc model of hh .
for the @xmath94 model of hh , we find @xmath95 .
an obvious difference between hh on the one hand and ss and our work on the other is that the former simulations are 2d , while the latter are 3d .
@xcite , however , compared 2d and 3d simulations and found no important differences , capable of accounting for the discrepancy we wish to explain .
given that @xcite consider only a single bar model , it would be desirable to have more in order to better understand the relation between 2d and 3d results .
gas flowing in a barred galaxy potential will form shocks along the leading edges of the bar @xcite , leading to an important inflow which can create a cmc .
a number of @xmath0-body+sph calculations have followed this evolution and found that the cmc that is thus formed can destroy the bar @xcite .
@xcite found that , when the gas is sufficiently concentrated so as to form a cmc of 2% of the disc mass , the bar is destroyed in about 1 gyr .
@xcite found a very similar result when the cmc formed by the inflow had a mass equal to 1.6% that of the galaxy within 10 kpc . and a characteristic radius of the order of a few 100 pcs .
the bar is destroyed also in the simulations of @xcite .
the two first studies use sph @xcite to model the gas , while the third one uses a sticky particles algorithm @xcite .
the cmc mass and radius estimates obtained from such simulations can not be compared with those obtained by purely @xmath0-body simulations .
of course , in both cases the growth of the cmc makes a large fraction of the x@xmath76 orbits unstable , and this family is known to be the backbone of the bar @xcite . in cases including gas , however , there is an extra effect , absent from the purely stellar cases .
it is now known that the growth of the bar is governed by the angular momentum exchanged between the inner parts of the disc ( bar ) and the halo plus outer disc ( a03 ) . in cases with gas
, there is an extra component taking part in this exchange process . @xcite
have shown that the gas , by giving angular momentum to the inner disc , will damp the bar growth .
thus the presence of a gaseous component should favour bar destruction so that the cmc mass necessary to achieve this should be smaller in cases with a sizeable gaseous component than in purely stellar ones . as noted in the introduction ,
the mass of observed super - massive black holes is of the order of @xmath3 times the mass of the bulge that harbours them .
unfortunately , neither our simulations , nor those of nsh&ss and hh , include any case with a live bulge .
bulges , however , are in general less massive than the corresponding discs , so the value of @xmath96 can be considered as an upper limit of the super - massive black hole mass .
this is two orders of magnitude lower than our fiducial value and one order of magnitude smaller than the smallest value considered here .
making comparisons with the cmc radius is less straightforward . following previous studies
, we assume that our results are relevant to super - massive black holes since the cmc radius is much smaller than its influence radius ; more than an order of magnitude for our fiducial cases .
therefore , in as much as our results can be applied to super - massive black holes , they argue that these can not destroy the bar that harbours them .
observations show that important mass concentrations , sometimes as large as @xmath97 , can often be found in the central parts of strongly barred galaxies @xcite .
these are either in the form of molecular gas discs , or of discy bulges ( for a description see * ? ? ?
simulations that do not include gas give only partial information on the effect of such a cmc on the bar , since they can not evaluate the role of the gas in the angular momentum exchange .
in other words , they can give information on the effect of the cmc as such , but not on the effect of its formation .
the following discussion will therefore , by necessity , neglect the latter effect . assuming a characteristic disc mass of 5 @xmath98
, we get a ratio @xmath99 , i.e. five times smaller than our fiducial @xmath63 .
the sizes of these cmcs are between 0.1 and 2 kpc , i.e. between 0.03 and 0.6 times the exponential disc scale length of the milky way . assuming that the characteristic scale length is half the extent , we find that the corresponding values for the @xmath59 , in the units used here , are between 0.015 and 0.3
. since in this paper we have used values of @xmath59 between 0.005 and 0.02 , the observed gaseous cmc will , for a same mass , be less destructive than the cmcs used here .
we thus come to the conclusion that such cmcs , seen their masses and not withstanding the effect of their formation , are not capable of destroying bars .
@xcite and @xcite show that the degree of gas concentration in the central kpc is higher in barred than in unbarred galaxies
. this would not be possible if the molecular gas cmcs could destroy a bar .
it is , however , in good agreement with our results .
the torques from a bar can push gas to the centre and , for a given gas reservoir , the stronger the bar the more important this central gaseous concentration will be @xcite .
it is thus reasonable , if the cmc does not destroy the bar , to expect that stronger gas concentrations will be found in barred rather than in unbarred galaxies , as is indeed borne out by the observations of @xcite and @xcite .
we thank a. bosma and s. hozumi for stimulating discussions and an anonymous referee for comments that improved the presentation .
e.a . and j.c.l .
thank the insu / cnrs , the region paca and the university of aix - marseille i for funds to develop the computing facilities used for the calculations in this paper .
athanassoula e. , 1992a , mnras , 259 , 328 athanassoula e. , 1992b , mnras , 259 , 345 athanassoula e. , 2002 , apj , 569 , l83 ( a02 ) athanassoula e. , 2003 , mnras , 341 , 1179 ( a03 ) athanassoula e. , 2005a , cel .
mech . & dyn .
astr . , 45 , 9 athanassoula e. , 2005b , mnras , 358 , 1477 athanassoula e. , misiriotis a. , 2002 , mnras , 330 , 35 ( am02 ) athanassoula e. , bienayme o. , martinet l. , pfenniger d. , 1983 , a&a , 127 , 349 athanassoula e. , fady e. , lambert j. c. , bosma a. , 2000 , mnras , 314 , 475 athanassoula e. , dehnen w. , lambert j. c. , 2003 , in engvold o. , ed . , highlights of astronomy vol
bahcall j. n. , wolf r. a. , 1976 , apj , 209 , 214 barnes j. , hut , p. , 1986
, nature , 324 , 446 berentzen i. , heller c. h. , shlosman i. , fricke k. j. , 1998 , mnras , 300 , 49 berentzen i. , athanassoula e. , heller c. h. , fricke k. j. , 2004 , mnras , 347 , 220 bournaud f. , combes f. , 2002 , a&a , 392 , 83 bureau m. , athanassoula e. , 2005 , apj , submitted contopoulos g. , papayannopoulos t. , 1980 , a&a , 92 , 33 dehnen w. , 2000a , apj , 536 , l39 dehnen w. , 2000b , aj , 119 , 800 dehnen w. , 2001 , mnras , 324 , 273 dehnen w. , 2002 , j. comp .
phys . , 179 , 27 dehnen w. , odenkirchen m. , grebel e. k. , rix h .- w . , 2004 ,
aj , 127 , 2753 elmegreen b. g. , elmegreen d. m. , 1985 , apj , 288 , 438 eskridge p. b. et al . , 2000 ,
aj , 119 , 536 ferrarese l. , merritt d. , 2000 , apj , 539 , l9 freeman k. c. , 1970 , apj , 160 , 811 friedli d. , 1994 , in shlosman i. , ed , mass - transfer induced activity in galaxies .
cambridge univ . press , p. 268
friedli d. , benz w. , 1993 , a&a , 268 , 65 gebhardt k. et al . , 2000 , apj , 539 , l13 goodman j. , binney j. , 1984 , mnras , 207 , 511 grosbl p. , patsis p. a. , pompei e. , 2004 , a&a , 423 , 849 hasan h. , norman c. , 1990 , apj , 361 , 69 hasan h. , pfenniger d. , norman c. , 1993 , apj , 409 , 91 helfer t. t. , thornley m. d. , regan m. w. , wong t. , sheth k. , vogel s. n. , blitz l. , bock d. c .- j . , 2003 , apjs , 145 , 259 heller c. h. , shlosman i. , 1994 , apj , 424 , 84 hernquist l. , 1993 , apjs , 86 , 389 hozumi s. , hernquist l. , 1998 , astro - ph 9806002 ( hh ) hozumi s. , hernquist l. , 1999 , in merritt d. , sellwood j. a. , valluri m. , eds , galaxy dynamics .
asp conf .
182 , p. 259
( hh ) hozumi s. , hernquist l. , 2005 , pasj , submitted ( hh ) kormendy j. , gebhardt k. , 2001 , in wheeler j. c. , martel h. , eds , 20th texas symposium on relativistic astrophysics , aip conf .
586 , p. 363 kormendy j. , kennicutt r. c. , 2004 , ara&a , 42 , 603 kormendy j. , richstone , d. , 1995 , ara&a , 33 , 581 kuzmin g.g . , 1956 ,
azh , 33 , 27 leeuwin f. , athanassoula e. , 2000 , mnras , 317 , 79 magorrian j. et al . , 1998 ,
aj , 115 , 2285 marconi a. , hunt l. k. , 2003 , apj , 589 , l21 mclure r. j. , dunlop j. s. , 2002 , mnras , 331 , 795 monaghan j. j. , 1992 , ara&a , 30 , 543 norman c. a. , sellwood j. a. , hasan h. , 1996 , apj , 462 , 114 ( nsh ) oneill j. k. , dubinski j. , 2003 , mnras , 346 , 251 peebles p. j. e. , 1972 , ge .
gravitation , 3 , 61 plummer , h. c. , 1911 , mnras , 71 , 460 quinlan g. d. , hernquist l. , sigurdsson s. , 1995 , apj , 440 , 554 regan m. w. , teuben p. j. , 2004 , apj , 600 , 595 regan m. w. , thornley m. d. , helfer t. t. , sheth k. , wong t. , vogel s. n. , blitz l. , bock d. c .- j .
, 2001 , apj , 561 , 218 sakamoto k. , okumura s. k. , ishizuki s. , scoville n. z. , 1999 , apj , 525 , 691 sandage a. , 1961 , the hubble atlas of galaxies . washington : carnegie institution schwarz m. p. , 1981
, apj , 247 , 77 seigar , m. s. , james , p. a. , 1988 , mnras , 299 , 672 shapiro s. l. , lightman a. p. , 1976
, nature , 262 , 743 sheth k. , stuart n. v. , regan m. w. , thornley m. d. , teuben p. j. , 2005 , astro - ph/0505393 shen j. , sellwood j. a. , 2004 , apj , 604 , 614 ( ss ) skokos h. , patsis p. a. & athanassoula e. , 2002 , mnras , 333 , 861 toomre a. , 1963 , apj , 138 , 385 toomre a. , 1964 , apj , 139 , 1217 tremaine s. et al .
, 2002 , apj , 574 , 740 valenzuela o. , klypin a. , 2003 , mnras , 345 , 406 wada k. , habe a. , 1992 , mnras , 258 , 82 wada k. , habe a. , 1995 , mnras , 277 , 433 young p. , 1980 ,
apj , 242 , 1232 | we study the effect of a central mass concentration ( cmc ) on the secular evolution of a barred disc galaxy . unlike previous studies
, we use fully self - consistent 3d @xmath0-body simulations with live haloes , which are known to be important for bar evolution .
the cmc is introduced gradually , to avoid transients . in all cases where the mass of the cmc is of the order of , or more than , a few per cent of the mass of the disc , the strength of the bar decreases noticeably .
the amount of this decrease depends strongly on the bar type . for the same cmc ,
bars with exponential surface - density profile , which formed in a disk - dominated galaxy ( md - type bars ) , can be totally destroyed , while strong bars with a flat surface - density profile , whose evolution is largely due to the halo ( mh - type bars ) , witness only a decrease of their strength .
this decrease occurs simultaneously from both the innermost and outermost parts of the bar .
the cmc has a stronger effect on the fourier components of higher azimuthal wave number @xmath1 , leading to fatter and/or less rectangular bars .
furthermore , the cmc changes the side - on outline from peanut - shaped to boxy or , for massive cmcs , to elliptical .
similarly , side - on initially boxy outlines can be destroyed .
the cmc also influences the velocity dispersion profiles .
most of the decrease of the bar strength occurs while the mass of the cmc increases and it is accompanied by an increase of the pattern speed . in all our simulations , the mass of the cmc necessary in order to destroy the bar is at least several per cent of the mass of the disc .
this argues that observed super - massive black holes are not likely to destroy pre - existing bars .
[ firstpage ] galaxies : evolution galaxies : bulges galaxies : structure galaxies : kinematics and dynamics methods : @xmath0-body simulations . |
intense tropical cyclones ( tcs ) are among the most devastating of natural phenomena , and considerable effort is spent in estimating the risk of tc landfall .
landfall risk assessments are used by the insurance industry for setting rates and by governments for establishing building regulations and planning emergency procedures .
recently , there has been much interest in documenting and understanding trends in tc frequency and intensity , and whether such trends are related to anthropogenic climate change and increasing sea - surface temperatures ( sst ) .
tc frequency has increased in the north atlantic in recent years , as has the number of tcs reaching the most intense categories ( webster et al . , 2005 ) , though the global tc frequency has been approximately steady .
theory points to a large role for sst in the maximum potential tc intensity ( emanuel , 1987 ) , and emanuel ( 2005 ) found a rapid increase since 1970 in global tc power dissipation that is highly correlated with sst . rising sst in the main development regions ( mdrs ) of tcs over recent decades is largely due to anthropogenic greenhouse warming ( santer et al .
, 2006 ; elsner , 2006 ) . less attention has been focused on variations in landfall risk and its geographic distribution with sst .
changes in tc frequency are likely to affect landfall rates , but so might changes in geographic distribution of tc genesis and paths of propagation
. lyons ( 2004 ) found that most of the difference between high and low u.s .
landfall years is due to changes in the fraction of tcs making landfall , rather than changes in basin - wide tc number .
in contrast to the atlantic - wide increase in tc power dissipation shown by emanuel ( 2005 ) , tc power dissipation at u.s .
landfall has not shown any trend ( landsea , 2005 ) , although there are many times fewer data on which to base the analysis . the most straightforward approach to analyzing
the sst dependence of landfall risk is to use historical landfall events .
this approach is sound if there are sufficient events , such as is found over large sections of coast and over many years .
however , if one aims to study changes in geographic landfall distribution and , in addition , use subsets of data years based on climate state , then sampling error becomes a major issue .
basin - wide statistical track models are an attractive alternative .
the primary advantage of a basin - wide model is that it utilizes historical track information over the full basin , roughly 100 times more data that just at landfall .
a second advantage is that by using a track model landfall changes can be decomposed into changes in various tc properties , such as number , genesis site , and propagation .
here we use the track model developed by hall and jewson ( 2007a ) to explore regional variations of landfall rates with sst . in section 2
we describe the historical data , in section 3 we briefly review the statistical tc track model , and in section 4 we discuss the conditioning of the model on sst .
the impact of sst on each model component is presented in section 5 . in section 6
we discuss the sst impact on north american landfall and relate the impact to tc number , genesis site and propagation .
we conclude in section 7 .
our tc analysis is based on hurdat data over the north atlantic ( jarvinen et al . , 1983 ) .
we use hurdat tcs of all intensity from 19502005 , which encompasses 595 tcs .
observations prior to 1950 are less reliable , as they precede the era of routine aircraft reconnaisance .
we also use sst data obtained from the uk met office hadley centre ( rayner et al . , 2003 ) .
1 shows the evolution from 1950 to 2005 of july - august - september sst averaged over the region 290@xmath1w to 345@xmath1e and 10@xmath1n to 20@xmath1n .
there is a well - documented upward trend , particularly in the past two decades , which is largely due to anthropogenic greenhouse warming ( santer et al .
, 2006 ; elsner , 2006 ) .
from the 56 years we extract the hottest 19 years , with sst @xmath2 27.38@xmath1c , and the coldest 19 years , with sst @xmath3 27.07@xmath1c , roughly 1/3 each of all years .
other geographic zones for sst averaging over the tropical and mid - latitude north atlantic result in very similar partitioning of data years . figs .
2a and 2b shows the historical tc tracks in the 19 cold years and the 19 hot years , of which there are 165 and 239 , respectively . in order to compute landfall rate we divide the north american continental coastline into 39 segments of varying length from maine to the yucatan peninsula , as shown in fig .
mileposts a through j are shown for reference in subsequent figures .
`` landfall '' of a tc is recorded when a track segment , heading landward , intersects a coastline segment .
a tc can make multiple landfalls .
4a shows the total landfall counts for hot and cold years for each coastline segment , running from maine to yucatan . in fig
4b these raw counts are converted to rates per year per 100 km of segmented coastline and plotted versus distance along the segmented coast .
the landfall counts and rates of fig .
4 show large variations along the coast and between hot and cold .
summed over the full coast there are 69 landfalls in 165 cold - year tcs and 110 landfalls in 239 hot - year tcs .
some components of these hot - cold differences reflect a geophysical relationship with sst , while other components are simply due to the sampling error inherent in the low total landfall counts .
many coastline segments have experienced only a few or no landfalls .
our goal is to extract the components of the hot - cold differences that are geophysical by minimizing the sampling error . in the statistical track analysis that follows
we exploit the much larger tc data set over the full atlantic basin , not just the data at landfall .
in essence , historical data from the full basin is projected onto the coastline to reduce the landfall sampling error .
hall and jewson ( 2007a ) described and evaluated a statistical model of tc tracks in the north atlantic from genesis to lysis based on hurdat data .
the track model consists of three components : ( 1 ) genesis , ( 2 ) propagation , and ( 3 ) lysis ( death ) .
the number of tcs in a simulation year is determined by random resampling of the historical annual tc number .
genesis sites are simulated by sampling a kernel pdf built around historical sites .
the kernel bandwidth is optimized by jackknife out - of - sample log - likelihood maximization . for propagation , we compute mean latitude and longitude 6-hourly displacements and their variances , by averaging of `` nearby '' historical displacements . nearby is defined optimally by jackknife out - of - sample log - likelihood maximization .
standardized displacement anomalies are modeled as a lag - one autoregressive model , with latitude and longitude treated independently .
the autocorrelation coefficients are determined from `` nearby '' historical anomalies , with `` near '' optimized as above .
finally , tcs suffer lysis with a probability determined by optimal averaging of nearby historical lysis rates .
tracks are stochastic , and two tracks originating from the same point can follow very different trajectories . at present no tc intensity is modeled , and no intensity information is used in the track model . clearly , intensity is a critical component of tc risk assessment . in this study , however , we restrict attention to landfall by tcs of any intensity .
the advantage of a track model over direct analysis of landfall is that information is brought to bear on landfall from the full trajectories of all tcs , not just the single landfall points of landfalling tcs .
this increases the data by a factor of 100 .
not all the data is equal , and the most relevant information comes from portions of tc trajectories near the coast .
this exploitation of the most relevant data for estimating landfall is implicit in the track model .
the track - model advantage is particularly important for analysis of small coast sections ; analysis of regions with low activity , where there are few or no historical landfalls ; and analysis in which only subsets of data years are used , for example , conditioned on a climate state . hall and jewson ( 2007b ) compared the landfall rates of the track model to those of a bayesian model based solely on local landfall observations , a type of statistical analysis often used in tc landfall studies .
they found that on regional and smaller scales the track model performs better in a jackknife out - of - sample evaluation with historical landfalls , despite the fact that over the entire north american coast the track model underestimates landfall by 12% , an amount that is significant . on a local basis , the decreased sampling error of the track model more than compensates for the increased bias .
or goal is to examine the effect of sst on tc landfall rates . to do
so we construct the track model separately on each historical sst set ; that is , the averaging to obtain means , variances , and autocorrelation coefficients of 6-hourly track displacements , and the track lysis probabilities are restricted to the particular cold or hot subset .
the track model components are thus `` conditioned '' on data being in a particular subset based on sst .
additionally , the kernel pdf of genesis sites is built on data only in the subset , and the annual tc number is drawn randomly only from the subset .
the hot- and cold - conditioned models are then run for a large number of years ( 1000 ) , and the synthetic tracks and their landfalls analyzed .
figs . 2c and
2d show the synthetic tracks based on the hot and cold years . in addition , in order to isolate individual components of the sst influence on tracks we generate synthetic tracks with single model components ( tc number , genesis site , propagation ) conditioned on sst , while the other components are unconditioned . in order to test the significance of hot - cold differences compared to other sources of variability
we have also constructed the track model from random subsets of data , as follows : we choose randomly a 19 year subset of the 56-year data , construct the track model , and generate 1000 synthetic tracks .
we then choose another random but non - overlapping 19-year subset , again construct the track model , and generate 1000 new synthetic tracks .
various properties ( e.g. , landfall rates ) of the two sets of synthetic tracks are differenced .
the selection of model components conditioned on the random 19-year periods always matches the selection of model components for sst conditioning ; for example , when we condition tc propagation alone on sst we compare the hot - cold difference to random differences in which propagation alone is conditioned on the random 19-year sets , to be consistent .
the comparison of hot - cold differences to random differences calculated in this way addresses the significance of hot - cold differences compared to other sources of variability . if the hot - cold difference of some rate stands out sufficiently from the random differences the null hypothesis , that sst has no effect on that rate , is disproved .
as we shall see , in many , but not all , regions hot - cold differences are significantly larger than random differences .
once hot - cold significance is established , however , a complementary analysis is needed to estimate the uncertainty on the estimate of a hot - year or cold - year rate . to this end
we successively remove one of the 19 years from the hot ( or cold ) set , for each removal conditioning the model on the 18 remaining years , generating synthetic tcs , and computing rates .
the range of resulting 18-year rates provides a measure of uncertainty of the 19-year rate .
here we examine the impact of sst conditioning on each model component individually . the resulting impact on landfall rates , both individually and in combination , is presented in section 6 .
our model for annual tc number is simple , consisting of random resampling from the tc annual numbers in the historical years of the data - year subset in question ( e.g. , hot , cold , or random ) .
the mean annual tc number over 1000 simulated hot years is 12.8 yr@xmath0 , while for cold years it is 8.5 yr@xmath0 .
the hot - cold difference , 4.3 yr@xmath0 ( 50% ) , is highly significant . by comparison ,
the mean absolute - magnitude of differences across the 20 random 19-year pairs is 0.8 yr@xmath0 , with an rms deviation of 0.6 yr@xmath0 .
in fact , none of the 40 random 19-year subsets has as few tcs as the cold years and none has as many tcs as the hot years . as we shall see below this difference in tc number has the single greatest influence on landfall rates . by construction
the effect of number on tc tracks and landfall has no geographic structure . given an annual number of tcs , where should they originate ? figs . 5a and
5b show the kernel pdfs of genesis sites for the cold years and the hot years .
not surprisingly , the hot years have greater probability for genesis in the mdr .
ocean heat is a key ingredient to cyclogenesis and tc intensification ( emanuel , 1987 ) , and the mdr is precisely the region whose sst we have used to condition the model genesis .
this highlights the fact that our separation of tc number and genesis site is to some degree artificial : much of the hot - year number increase is concentrated in the defining region of hot years . however , there is additional hot - cold genesis - site structure outside the mdr .
hot years also have greater probability for genesis off the mexican gulf coast and just east of the yucatan peninsula . in contrast , individual events in cold years have greater probability for genesis in the northern gulf of mexico and in the western atlantic off the southeastern u.s coast . for comparison ,
5c shows the genesis pdf for all 56 data years , which is intermediate .
( we emphasize that these pdfs are normalized spatially .
they provide the probability of a genesis site , given a genesis occurrence .
if a pdf increases in one region , it must decrease elsewhere . )
5d shows shows a measure of the significance of the hot - cold difference in genesis - site pdfs .
the pdfs of the 20 random 19-year pairs are differenced and the rms deviation of the differences computed at each plotted point in the basin ( @xmath4 ) .
the hot - cold difference is plotted where ( 1 ) its magnitude is greater than one standard deviation across the random differences ( i.e. , it is significant ) , and ( 2 ) its magnitude is within 50% of the maximum hot - cold magnitude ( i.e. , it is large ) .
the differences noted qualitatively comparing fig .
5a to 5b are confirmed to be significant .
the hot - cold differences in genesis site impact the distribution of modeled tc tracks throughout the basin and also affect the modeled tc landfall rate .
landfall rates are analyzed in section 6 . here , we determine the impact of genesis site changes on tc tracks by analyzing the number of tcs crossing lines of constant longitude and latitude . for this purpose ,
only genesis site pdfs are conditioned on sst , while tc number and propagation are unconditional .
6 shows the zonal `` flux '' of hot- and cold - year tcs ; that is , the number of tcs crossing lines of constant latitude per year counted in 5@xmath1 latitude bins .
eastward and westward fluxes are shown separately , as are the eastward and westward hot - cold flux differences .
plotted with the flux differences are the random - difference standard deviations as a measure of statistical significance .
7 shows a comparable plot for the meridional tc flux . the greater probability of genesis in the mdr in hot years results in more tcs propagating westward in the subtropics .
the hot - cold difference is significant .
many tcs then curve northward , followed by eastward . a significantly greater hot - year northward
flux is seen across 20@xmath1n . the eastward midlatitude tc flux is greater in hot years , but the hot - cold difference is in most places not significant compared to the random differences . by contrast , the influence of the cold - year genesis - site peak off the southeastern u.s .
coast can be seen in the significantly greater cold - year northward flux off the u.s .
coast through 30@xmath1n and 40@xmath1n .
we now analyze the impact of sst on tc propagation by conditioning on sst the mean 6-hourly displacements , the variance about the mean , the autocorrelation , and the lysis .
the genesis - site pdfs and the tc numbers are unconditional . first , to illustrate the sst influence we generate 1000 hot and cold synthetic tcs from the same genesis site and compute the density of track points .
fig . 8a and
8b shows the resulting regions of high track - point density from eastern and western genesis sites in the mdr .
the hot - cold tc difference is most apparent for the western genesis site .
hot - year tcs are more likely to bend northward and eastward earlier , while cold - year tcs tend to propagate further west , including into the caribbean , before heading north .
the zonal and meridional fluxes are shown in figs . 9 and 10 .
these figures are analogous to figs . 6 and 7 , but now for the case of sst - conditioned propagation and lysis , but unconditional tc number and genesis site .
consistent with fig .
8 westward cold - year flux is greater than hot - year flux through 290@xmath1e , although the difference is only marginally significant .
by contrast the northward hot - year flux is greater through 20@xmath1n , 30@xmath1n , and 40@xmath1n in the western atlantic .
the most significant propagation differences are seen in the eastward flux in midlatitudes , which is significantly greater in hot years than cold .
the physical mechanisms relating mdr sst to tc propagation are unclear .
it may be that the signal seen here is a projection of the north - atlantic oscillation ( nao ) on our sst data - year partitioning .
in the nao s high phase sst in the tropical north atlantic is lower , and the subtropical surface - pressure high extends further west ( wang , 2002 ) .
this would tend to cause tcs to propagate further west before curving north . during the nao s low phase the sst is higher , and
is therefore associated with more tcs ( elsner et al . , 2006 ) , but the subtropical high does not extend as far west , and tcs curve north sooner .
we now analyze the landfall - rate characteristics of the synthetic tc data .
figs 1114 illustrate landfall rates ( counts per year per 100 km ) and rate differences as a function of distance along the segmented coastline running southward from maine , as in fig .
4 . the general features of the model s landfall rate and its evaluation against historical rates are discussed in hall and jewson ( 2007a ; 2007b ) . accumulated over the full coast
the model suffers from a low landfall bias of about 12% compared to the 19502005 historical rate , a difference which is marginally significant compared to the rms deviation across many 56-year simulations .
landfall rates show strong geographic variations , with several local maxima .
the rates are sensitive to the orientation of coast segments , with segments perpendicular to the local mean tc path experiencing more landfalls .
hence , there is a local landfall - rate minimum off the southeast u.s .
atlantic coast ( between mileposts c and d ) , where the coast segments are close to parallel with the mean tc tracks , while there is a maximum near cape hatteras ( mileposts b
c ) and long island ( mileposts a
b ) , where south - facing coast segments jut out into oncoming tcs .
one implication is that different segmented models of the coastline will result in different distributions , making it difficult to compare directly the quantitative landfall rates from one study to another .
11 shows the landfall rates and the hot - cold difference for the case with only tc number conditioned on sst .
tc - number conditioning on sst has no geographic structure , and the hot - cold landfall - rate ratio is roughly uniform , with the hot - year rate everywhere significantly greater than the cold - year rate .
( the ratio is uniform in the limit of large number of simulation years . )
accumulated over the entire segmented coast there are 3.2 cold landfalls per year and 4.8 hot landfalls per year , a ratio that simply reflects the historical hot - cold tc - number ratio 239/165 .
the hot - cold landfall - rate difference , 1.5 yr@xmath0 , is highly significant . across the 20 random 19-year pairs
the rms deviation of landfall - rate differences is only 0.4 yr@xmath0 . fig .
12 is comparable to fig .
11 , but now for the case of genesis - site pdf conditioned on sst , while the other components are unconditional . accumulated over the entire coastline the hot - year landfall rate is 3.9 yr@xmath0 compared to the cold - year rate 3.8 yr@xmath0 , a difference that is well within the range of random differences , which has an rms deviation of 0.5 yr@xmath0 . a few regions , however , have significant hot - cold differences . on the mexican gulf coast ( milepost g ) and the eastern yucatan ( milepost j )
, hot - year genesis sites result in significantly greater landfall , consistent with the hot - year genesis - site peaks just off the coasts of these regions ( fig . 5 ) .
by contrast , on the northern u.s .
gulf coast ( between mileposts e and f ) the cold - year landfall rate is significantly higher , consistent with the cold - year genesis - site peak in the central and northern gulf of mexico .
in addition , the peak in cold - year genesis sites off the southeastern u.s .
coast causes cold - year landfall in the mid - atlantic u.s .
coast to be greater ( between mileposts b and c ) , a difference that is marginally significant .
note that these landfall rate differences do not take tc intensity into account .
cold - year genesis sites cause more cold - year landfalls on the u.s .
mid - atlantic coast , but these landfalls are not far from their genesis sites , and have not had much time to intensify .
by contrast , the more frequent hot - year genesis in the mdr results in lower probability per tc of making landfall there is more opportunity to veer away from coast but also allows more time for tcs to intensify .
13 shows the landfall rates for sst conditioning of propagation and lysis , with the genesis - site and tc number unconditional .
accumulated over the entire coastline the hot - year landfall rate is 3.7 yr@xmath0 and the cold - year landfall rate is 3.9 yr@xmath0 , a difference that is well within the random variability of random differences , which now has an rms deviation of 0.6 yr@xmath0 . once again , however , on some regions the hot - cold differences are significant , albeit marginally so . on the u.s .
northeast and mid - atlantic coast segments facing the oncoming mean tc tracks ( long island and cape hatteras ) cold - year landfall rates are greater , reflecting the fact that in cold years tcs propagate further west before veering north ( figs 810 ) .
note in fig .
10 that across 40@xmath1n just off the u.s .
coast the northward flux of tcs is greater in cold years than hot ( marginally significant ) , even though only a few degrees further east the opposite is true .
by contrast on the eastern yucatan ( milepost j ) the sst effect on propagation causes hot - years to have greater landfall rate .
this effect must be well localized , because there is little hot - cold difference in the westward tc flux across 280@xmath1e , about 10@xmath1 east of yucatan .
we have not separated the sst influence on propagation and lysis , in order to keep the number of figures manageable .
separate analysis ( not shown ) indicates no significant hot - cold landfall - rate differences due to lysis alone .
14 shows the hot - year and cold - year landfall rates when all model components are conditioned on sst simultaneously .
accumulated over the entire coastline the sst effects on genesis site and propagation are negligible , and the fully - conditioned hot - cold landfall rates are nearly identical to the tc - number only case : 3.1 yr@xmath0 in cold years and 4.7 yr@xmath0 in hot years .
the difference , 1.5 yr@xmath0 , is significant , about 2 times the random difference rms deviation of 0.7 yr@xmath0 .
note that compared to the tc - number conditioning , the random variation is now larger , reflecting variability in more model components .
hot - cold differences in genesis site and propagation affect regional landfall rates , however , in some places amplifying the tc - number effect and in other places ameliorating it . on the eastern yucatan coast
the landfall effects of sst on tc number , genesis site and propagation all have the same sign : greater hot - year than cold - year landfall . with all components conditioned on sst
the yucatan landfall rate is about 3 times higher in hot years than cold , a difference which is highly significant .
by contrast , the greater landfall rates in hot years on the northeast and mid - atlantic u.s .
coast due to tc number alone is countered by the greater cold - year rates due to genesis site and propagation .
the net effect is no significant hot - cold landfall - rate difference in this region .
most of the sst effect via tc - number on florida and the u.s .
gulf coast survives under full conditioning , although its significance is now marginal , as the random variability is larger . note that significance increases when landfall is accumulated over larger regions .
table 1 lists the hot - year and cold - year rates on the 6 regions indicated in fig .
3 . also listed are the @xmath5 scores of the differences ; that is , the hot - year minus cold - year rate difference divided by the rms deviation of the random differences . on the u.s .
northeast and mid - atlantic coastlines the hot - cold differences are not significant . on florida , the u.s gulf coast and the mexican gulf coast the hot - cold differences are marginally significant ( @xmath5 scores of 0.9 , 1.1 , and 0.9 , respectively ) . on the yucatan peninsula
the hot - cold difference is highly significant , with @xmath6 .
the fact that genesis - site and propagation conditioning on sst has little effect on landfall rates accumulated over large sections of the coast is reflected in the small hot - cold difference in landfall fraction ( fraction of tcs making landfall ) . over the the u.s .
portion of the coastline ( segments 123 ) the modeled landfall fractions are 0.28 in cold years and 0.29 in hot years .
the actual historical record reveals a larger difference : 0.30 in cold years and 0.35 in hot years ( 50 u.s .
landfalls in 165 historical cold - year tcs and 84 u.s .
landfalls in 239 historical hot - year tcs ) . here
we examine this apparent discrepancy .
the historical hot and cold years subsets analyzed here each have 19 years .
we divide the 1000-year hot and cold simulations into 52 19-year periods and compute the landfall fractions in each period .
these fractions exhibit considerable variability .
15a shows the pdf of these landfall fraction for hot and cold years , along with the total simulated hot and cold fractions and the historical hot and cold fractions .
the historical fractions each fall within the variability of the simulated fractions .
we then take many random hot - cold differences of landfall fractions and compute the pdf of the differences ( fig .
the historical difference ( hot @xmath2 cold by 0.05 ) is well within the range of random differences in the simulations .
this suggests that the apparent discrepancy between the historical and simulated hot - cold landfall fraction differences is not significant .
it also suggests that no significant impact of sst on u.s .
landfall fraction can be gleaned from observations over the period 19502005 .
we have demonstrated that for many features and on many regions the landfall rates in hot years are significantly greater than those in cold .
we now convert the hot - year rates to landfall probabilities with confidence limits and compare to the same probabilities in all years .
landfall can be modeled as a poisson process ( bove et al . , 1998 ) whose rate is uncertain , due to the finite historical record .
( we have verified that the distributions of annual landfall counts in the model are poisson . ) incorporating this uncertainty into the landfall probability results in a negative binomial distribution ( elsner and bossak , 2001 ; hall and jewson , 2007b ) .
given @xmath7 landfalls in @xmath8 years the probability of @xmath9 landfalls in a subsequent year is @xmath10 expression ( [ e : prob ] ) reduces to a poisson distribution for large @xmath8 and @xmath7 .
in the case of the synthetic tcs @xmath11 , and @xmath7 on 100 km coastline sections varies from 1 to 144 ( using all years ) , depending on region . fig .
16a shows the probability of at having at least one landfall in a year in 100 km sections along the segmented coastline for the hot years and for all years , based on these distributional assumptions .
fig 16b shows the comparable distribution for the probability of at least two landfalls .
plotted for the hot - years are the 90% confidence limits of the probability determined from a distribution of probabilities .
the distribution is determined , in turn , by successively dropping out one of the 19 hot years , conditioning the model on the remaining 18 hot years , generating synthetic tcs , and computing landfall rates and probabilities .
for example , on segment 19 ( 295 km in length facing southeast of the louisiana coast between mileposts e and f in fig .
3 ) there is a probability of 0.0850.117 of at last one landfall in hot years in 100 km sections , versus a probability of 0.065 in all years , an increase of 31% to 80% . for two or more landfalls
there is a probability of 0.00380.0072 for hot years versus 0.0022 for all years , an increase of 73% to 327% . by contrast , on segment 8 ( 91 km in length facing south from cape hatteras between mileposts b and c ) the probability range for at least one landfall in 100 km is 0.1100.152 , which encompasses the probability for all years of 0.134 .
the probability here for at least two landfalls in hot years is 0.00640.0122 , which again encompasses the probability , 0.0094 , for at least two landfalls in all years .
listed in table 2 are the hot - year and all - year landfall probabilities on 6 larger regions of coast : the u.s .
northeast , the u.s .
mid - atlantic , the florida peninsula , the u.s .
gulf coast ( excepting florida ) , the mexican gulf coast ( excepting yucatan ) , and the yucatan peninsula .
mileposts delineating these regions are indicated in fig .
3 . compared to fig .
16 , here the probabilities are accumulated over the entire region , rather than in 100 km sections . on the u.s .
northeast coast there is apparently a slight reduction in landfall probability in hot years compared to all years .
however , as indicated in table 1 , the hot - cold difference is not significant , and we conclude that there is no change in landfall probability in hot years . on the u.s .
mid - atlantic coast the hot - year and all - year probabilities are not different , and the hot - cold difference is within the range of random differences . on florida , the u.s .
gulf coast , and the mexican gulf coast the hot - cold differences are marginally significant ( @xmath12 ) , while on the yucatan coast the hot - cold difference is highly significant ( @xmath6 ) . on florida , the u.s .
gulf , and the yucatan peninsula , the hot - year landfall probabilities are higher than the probabilities for all years , with at least 95% confidence ( i.e. , the lower 90% confidence limit of hot years is greater than the probability for all years ) .
d the increase in landfall probability in hot years is greatest on the yucatan peninsula ; e.g. , there is a 90% to 250% increase in probability of two or more tcs making landfall in hot years compared to all years . on the u.s .
gulf coast the comparable increase is 14% to 36% .
the hot - year increase on the u.s .
gulf coast for making three or more landfalls is 20% to 70% .
we have performed additional limited analysis with alternative spatial and temporal averaging of sst to test the robustness of our conclusions .
increasing the latitude range from 10@xmath120@xmath1n to as much as 0@xmath140@xmath1n , the longitude range from 290@xmath1345@xmath1e to as much as 270@xmath1345@xmath1e , and altering the time period from jul sep to aug oct in various combinations never changes the 19 hottest and 19 coldest years among 19502005 by more than 3 years .
thus , our tc model conditioning will not be strongly affected . for the case 10@xmath130@xmath1n , 270@xmath1345@xmath1e , and aug oct ( change in 3 of the 19 hottest and coldest years ) we have fully conditioned the model and simulated 1000 hot- and cold - year tc tracks .
the genesis - site pdfs are altered quantitatively , though not qualitatively , and the hot - cold landfall - rate difference over the entire coastline is reduced ( 4.5 yr@xmath0 in hot years , 3.5 yr@xmath0 in cold years )
. however , the basic features described above are unchanged .
there is no significant hot - cold difference for the u.s .
ne and mid - atlantic coasts , while the remaining regions exhibit significant hot - cold differences .
we have used a statistical tropical cyclone ( tc ) track model to elucidate the relationship between tropical atlantic sst and tc landfall rates on north america .
the advantage of a track model over direct analysis of historical landfall rates is the reduction in sampling error from using the much larger quantity of data over the entire ocean basin .
this reduced sampling error allows meaningful examination of detailed geographic structure in landfall rate and its dependence on sst .
we have constructed the model components ( annual tc number , genesis site , and propagation ) individually and together from the 19 lowest and 19 highest sst years in 19502005 , generated 1000s of synthetic tc tracks , and analyzed their landfall characteristics .
we have constructed the model on randomly selected pairs of 19-year periods and compared the range of landfall differences among these pairs to the hot - cold differences as a test of hot - cold significance .
in addition , we have constructed the model on sub - samples of the 19 hot years in order to estimate confidence limits on the hot - year rates .
tc number has a large and well known dependency on sst : there are 45% more tcs in the hot years than cold , a difference which is highly significant compared to random sampling .
alone , this sst dependency of tc number translates to an identical fractional landfall - rate increase , uniform along the coast .
possibly modifying this tc - number effect is variation with sst of the geographic distribution of genesis and the propagation of tcs .
it turns out , however , that accumulated over the entire coastline there is little additional effect on landfall of genesis - site and propagation variations . when all components of the model are conditioned on being in either the cold or hot sst years we find a landfall rate of 4.7 yr@xmath0 in hot years and 3.1 yr@xmath0 in cold years , rates whose ratio is not significantly different than the historical hot - cold tc - number ratio 239/165 . over the period 19502005 tcs in the 19 hot years have a higher u.s .
landfall fraction ( 0.35 ) than tcs in the 19 cold years ( 0.30 ) , but we have argued that this difference is not significant .
there are , however , significant differences in the geographic distribution of landfall in hot and cold years .
tc number alone causes a uniform increase in landfall rate in hot years .
changes in genesis site and propagation amplify this increase in some regions , most notably the eastern yucatan peninsula , whose landfall rate is 3 times higher in hot years than cold .
in contrast , genesis - site and propagation dependence on sst ameliorate the increase in landfall on the u.s . northeast and mid - atlantic coast , with a net result of no significant hot - cold difference in landfall rate .
landfall rates on the florida and northern u.s .
gulf coasts are higher in hot years than cold , although the difference is less significant than would be concluded from tc number alone .
we have also computed the hot - year increase in landfall probabilities compared to all years with confidence limits determined from subsampling the 19 hot data years .
landfall probabilities on florida , the northern u.s .
gulf coast , and the yucatan peninsula are higher in hot years with at least 95% confidence .
we have verified that these qualitative features of regional hot - cold differences are not sensitive to the sst averaging .
it is important to note that no intensity model is included here , nor is any intensity used in the analysis .
our estimated landfall rates refer to any named tc .
overall , landfall rates are , of course , lower if attention is restricted to the most intense tcs .
in addition , regional hot - cold differences might be quite different .
for example , cold - year genesis sites cause greater cold - year landfall rates on u.s .
mid - atlantic coast . but
these landfalls are not far from their genesis sites , and so have not had much time to intensify . by contrast , more frequent genesis in the tropical eastern atlantic in hot years may result in a lower chance per tc of making landfall
there is more opportunity to veer away from coast but the tcs that do make landfall have had more time to intensify .
increases in sst in regions of tropical cyclogenesis over recent decades is due primarily to anthropogenic greenhouse forcing ( santer et al .
, 2006 ; elsner , 2006 ) .
these increases are likely to persist and to intensify because ( 1 ) the ocean has large thermal intertia , so that ( 2 ) the ocean is still responding to greenhouse forcing of the last few decades ( hansen et al . , 2005 ) , and ( 3 ) greenhouse gas forcing will almost certainly continue to grow over the next few decades .
it must be acknowledged that the mechanisms for cyclogenesis and tc intensification are poorly understood , and factors in addition to sst play important roles .
tc frequency in other ocean basins has not increased as it has in the atlantic . moreover ,
the association we see between sst and tc propagation is not understood , and may not be directly causal ; i.e. , there may be an unidentified mechanism responsible for variability in both sst and propagation
. however , our results show a significant increase of north american tc landfall with tropical sst in rough proportion to the increase in tc number .
the implication is clear for higher tc landfall risk with future sst increases .
finally , we note that the vast majority of the upward trend in financial loss due to tc landfall is explained by increased exposure ; that is , increased coastal population and development ( pielke and landsea , 1998 ) .
it is sometimes argued that this fact trivializes any climate - change impact on tc frequency and intensity .
the argument is wrong , in our view . regardless of past damages
, coastal regions will continue to be populated and developed . to plan development , to plan emergency procedures , and to set insurance rates
developers , governments and insurers need to know the risk of tc damage and its potential for future change .
this is the question that is addressed by this study and other studies seeking to understand the connection between tcs and climate variation , including secular anthropogenic change .
nasa is acknowledged for support of this research .
we also thank noaa s atlantic oceanographic and meteorological laboratory for maintenance of the hurdat data base .
bove , m. c. , j. b. elsner , c. w. landsea , and j. j. obrien , 1998 , effect of el nino on u.s .
landfalling hurricanes , revisited , _ bull .
_ , * 79 * , 24772482 .
jarvinen , b. r. , c. j. neumann , and m. a. s. davis , 1984 : a tropical cyclone data tape for the north atlantic basin , 18861983 , contents , limitations , and uses , _ noaa tech
_ , nws nhc 22 , miami , florida .
rayner , n. a. , et al .
, 2003 , global analysis of sea - surface temperature , sea ice , and night marine air temperature since the late nineteenth century , _ j. geophys
_ , * 108 * , doi:10.1029/2002jd002670 .
e to 290@xmath1e and 10@xmath1n to 20@xmath1n from the hadley centre sst data ( rayner et al . , 2003 ) .
the horizontal dashed lines indicate the thresholds for hot and cold years ( 27.38@xmath1c and 27.07@xmath1c ) used for model conditioning .
the averaging region is plotted in the map ( right).,title="fig : " ] + cold by an amount at least 50% of the maximum difference and the difference is greater than one rms deviation across the random differences .
blue indicates regions with the same magnitude and significance thresholds , but for cold @xmath2 hot.,title="fig : " ] + w through 340@xmath1w every 10@xmath1 from left to right ) as a function of latitude ( vertical axes ) . in this figure sst conditioning
is performed for the genesis - site pdfs only ; i.e. , there is no conditioning for the tc number , propagation or lysis .
units are numbers per year of tcs crossing the longitude lines accumulated in 5@xmath1 latitude bins .
curves are shifted 10@xmath1 for each longitude line . shown
are ( a ) the eastward tc fluxes for cold - years ( blue ) and hot - years ( red ) ; ( b ) the difference ( solid lines ) of the hot - cold eastward fluxes ; ( c ) the westward fluxes ; and ( d ) the difference of the hot - cold westward fluxes . to test significance of the hot - cold differences , also shown ( orange filled ) in ( b ) and ( d ) are the mean @xmath13 rms deviation of the eastward and westward random flux differences.,title="fig : " ] + n to 50@xmath1n every 10@xmath1 latitude , bottom to top ) as a function of longitude ( horizontal axes ) . in this figure sst conditioning
is performed for the genesis - site pdfs only ; i.e. , there is no conditioning for the tc annual number , propagation or lysis .
units are numbers per year of tcs crossing the latitude lines accumulated in 5@xmath1 longitude bins .
curves are shifted 10@xmath1 for each latitude line . shown
are ( a ) the northward tc fluxes for cold - years ( blue ) and hot - years ( red ) ; ( b ) the southward fluxes ; ( c ) the difference ( solid lines ) of the hot - cold northward fluxes ; ( d ) the difference of the hot - cold southward fluxes . to test significance of the hot - cold differences , also shown ( orange filled ) in ( c ) and ( d )
are the mean @xmath13 rms deviation of the northward and southward random flux differences.,title="fig : " ] + rms deviation difference ( orange fill ) . here
, only the annual tc number is conditioned on sst .
the other track model components ( genesis site , propagation , and lysis ) are unconditional .
mileposts ( a j ) and regions are labeled for ease of reference to fig .
vertical shaded regions correspond to the regions defined in fig .
3.,title="fig : " ] + .landfall rates ( counts per year ) on 6 regions and the full coast for hot years ( @xmath14 ) and cold years ( @xmath15 ) and the @xmath5 score of their difference , given by @xmath16 divided by the rms deviation of random differences .
[ cols="^,^,^,^,^,^,^,^",options="header " , ] | we employ a statistical model of north atlantic tropical cyclone ( tc ) tracks to investigate the relationship between sea - surface temperature ( sst ) and north american tc landfall rates .
the track model is conditioned on summer sst in the tropical north atlantic being in either the 19 hottest or the 19 coldest years in the period 19502005 . for each conditioning
many synthetic tcs are generated and landfall rates computed . compared to direct analysis of historical landfall , the track model reduces the sampling error by projecting information from the entire basin onto the coast .
there are 46% more tcs in hot years than cold in the model , which is highly significant compared to random sampling and corroborates well documented trends in north atlantic tc number in recent decades . in the absence of other effects
, this difference results in a significant increase in model landfall rates in hot years , uniform along the coast .
hot - cold differences in the geographic distribution of genesis and in tc propagation do not significantly alter the overall landfall - rate difference in the model , and the net landfall rate is 4.7 yr@xmath0 in hot years and 3.1 yr@xmath0 in cold years .
sst influence on genesis site and propagation does modify the geographic distribution of landfall , however .
the yucatan suffers 3 times greater landfall rate in hot years than cold , while the u.s .
mid - atlantic coast exhibits no significant change .
landfall probabilities increase in hot years compared to all years in florida , the u.s gulf coast , the mexican gulf coast , and yucatan with at least 95% confidence . |
we first introduce a three - strategy version of the spatial inspection game , where in addition to criminals @xmath1 and punishers @xmath2 , also ordinary people @xmath3 compete for space on a @xmath4 square lattice with periodic boundary conditions .
we use the latter as the simplest network to account for the fact that the interaction range among individuals in human societies is limited . the payoff matrix + [ cols=">,^,^,^",options="header " , ] contains the same three main parameters as the three - strategy payoff matrix , with the key difference being that punishers @xmath5 and @xmath6 are willing to bear only @xmath7 and @xmath8 of the full punishment cost @xmath9 , respectively .
naturally , they also receive a proportionally smaller reward @xmath10 .
punishers @xmath11 correspond to punishers @xmath2 in the three - strategy model in terms of their commitment to sanctioning criminals , but we introduce a different notation for convenience . both the uniform three - strategy and the heterogeneous five - strategy spatial inspection game are studied by means of monte carlo simulations , as described in the methods section .
we begin by presenting the complete @xmath12 phase diagram at a representative value of the punishment cost @xmath9 in fig .
[ phase ] .
it can be observed that criminals dominate if the reward for their punishment @xmath10 is small .
if the reward exceeds a certain value at a fixed temptation / loss @xmath13 , then the punishers become viable . at moderate @xmath13 values , however , their presence is also accompanied by the emergence of ordinary players .
the stability of the @xmath14 phase is due to cyclic dominance between the three competing strategies @xcite . in particular , within the @xmath14 region ordinary people outperform the punishers , the punishers defeat the criminals , while the criminals beat ordinary people , thus closing the @xmath15 loop of dominance .
conversely , for larger values of @xmath13 , in particular if @xmath16 , the pure @xmath1 phase becomes the two - strategy @xmath17 phase via a second - order continuous phase transition as @xmath10 increases . moreover , at sufficiently large values of the reward @xmath10 , the three - strategy @xmath14 phase and the two - strategy @xmath17 phase are separated by a second - order continuous phase transition . for a more quantitative view , we present in fig .
[ cross ] characteristic cross - sections of the phase diagram shown in fig .
[ phase ] .
these cross - sections confirm that criminals can dominate in the high temptation / loss region or in the low reward region .
moreover , it can be observed that larger rewards are beneficial for the punishers , but only up to a certain point . if @xmath10 increases beyond a critical point
ordinary people emerge , and as second - order free - riders they flourish on the expense of those that punish criminal behavior .
we emphasize that , interestingly , the payoffs of ordinary people are independent of @xmath10 , yet still their fraction increases as @xmath10 increases .
this counterintuitive result is due to cyclic dominance , where feeding the prey , in this case the punishers who do get larger payoffs for larger @xmath10 values , directly benefits the predator , which in this case are the ordinary people @xcite .
we can thus conclude that the real obstacle in the fight against criminal behavior is the possibility of ordinary people to free - ride on the efforts of punishers .
a similar conclusion has been reached before for the evolution of cooperation in the public goods game with punishers , where the free - riding problem of defectors is simply deferred to the second - order free - riding problem of cooperators @xcite . as a natural response of punishers to the harmful exploitation of ordinary people
, we next consider the five - strategy spatial inspection game with heterogeneous punishment .
in particular , strategies @xmath5 and @xmath6 try to eschew the exploitation by reducing the amount they contribute for sanctioning to @xmath7 and @xmath8 of the full cost , respectively .
however , their reward is proportionally smaller as well ( see the extended payoff matrix in section 2 for details ) . due to the large number of competing strategies and the resulting multitude of possible subsystem solutions we focus on the most important parameter region where ordinary players survive in the uniform , three - strategy , model .
accordingly , we explore a representative cross section when the reward is high enough for punishing strategies to survive , and we explore how the system responds to the diversity of punishment .
results presented in the left panel of fig .
[ opext ] confirm the effectiveness of resorting to heterogeneous punishment in that second - order free - riders are able to survive only in a significantly narrower interval of the temptation / loss @xmath13 if compared to the uniform punishment model .
furthermore , results presented in the right panel of fig . [ opext ] also give credence to the expectation that the reduced viability of ordinary people will promote the evolution of punishers . more precisely , we find that the uniform punishment strategy is significantly less effective than heterogeneous punishment for almost the entire range of the temptation / loos @xmath13 , except for a narrow interval in the @xmath16 region . as we will show in fig .
[ cext ] , this fact has important consequences for the mitigation of criminal behavior in the population .
another peculiarity that can be observed in the right panel of fig .
[ opext ] is the zig - zag outlay of the aggregate fraction of all punishers in the five - strategy model . yet
this can be understood thoroughly simply by looking at the fraction of punishers @xmath5 , @xmath6 and @xmath11 individually .
the mentioned panel reveals clearly that low values of @xmath13 are able to sustain only those punishers who are willing to invest the lowest cost towards sanctioning criminals .
the rank of the most viable punishers subsequently increases from @xmath5 over @xmath6 to @xmath11 as we increase @xmath13 , and the solution of the five - strategy model thus eventually becomes identical to the the solution of the three - strategy model .
remarkably , we can observe six consecutive phase transitions [ @xmath18 as we increase a single parameter , @xmath13 .
it is worth pointing out that the reported increment of the punisher rank with increasing the temptation / loss @xmath13 resonates with the outcome of a recent human experiment @xcite , where , in the realm of a social dilemma , it was shown that if cooperation is likely one should punish mildly .
we continue with the results presented in fig .
[ cext ] , where we compare the effectiveness of uniform and heterogeneous punishment to deter criminal behavior . to a degree
unexpected , it can be observed that the possibility to resort to different levels of punishment does not necessarily work better than uniform punishment in reducing crime . on the contrary ,
the fraction of @xmath1 players is generally higher over a large interval of @xmath13 values when the heterogeneous punishment model is used .
more precisely , the fraction of criminals is lower only in the low temptation / loss region where @xmath5 punishers can adjust to this favorable condition .
this observation is related to the failure of heterogeneous punishment to eliminate second - order free - riding more effectively than uniform punishment , and it indicates that sophisticatedly adjusted punishers may win a battle against ordinary people , but loose the main war against the actual enemy , the criminals .
while punishers can lower the amount they invest towards sanctioning criminals , such a reduced effort also yields smaller rewards .
interestingly , the positive side of lower costs can be utilized only if the heterogeneity of punishers is maintained .
the said effect becomes visible if we mark the borders of different phases on the curve of criminals , as shown in the right panel of fig .
as it is illustrated , the fraction of criminals can be a decaying function even if we increase the temptation / loss @xmath13 , but only as long as different types of punishers exist and compete against the criminals .
as soon as evolution favors a single punisher type , an effective response to an increase of the value of @xmath13 becomes absent .
lastly , we note that the conclusions attained with the results presented in figs . [ opext ] and [ cext ] remain generally valid also for all high temptation values . to obtain a better understanding of the origin of the zig - zag outlay of criminals depicted in fig .
[ cext ] , we monitor the time evolution of the distribution of strategies in the population for three different combinations of payoff parameters , as shown in fig .
[ snapshots ] .
we emphasize that the main mechanism responsible for the formation of different stationary states is due to the different motion of interfaces that separate the possible solutions of the system .
accordingly , we follow the evolution of interfaces starting from a prepared initial state , but for clarity only two types of punishers are present because this minimal model is sufficient to capture the essence of the emerging effect .
the extrapolation to the full five - strategy model , however , is straightforward . for comparison
, we use an identical prepared initial state , as shown in the leftmost panel , for three representative values of @xmath13 . as in previous figures ,
red color depicts @xmath1 players while light and dark blue depict the @xmath5 and @xmath6 punishers , respectively . before discussing each specific case , we note that , individually , @xmath5 always beats @xmath6 due to the lower cost of inspection
. when the temptation / loss is low , as shown in panels ( a)-(d ) , @xmath6 can beat @xmath1 very efficiently , while @xmath5 is unable to do the same but simply coexists with the criminals .
the superiority of @xmath5 over @xmath6 , however , will result in a shrinking area of the @xmath6 domain , as shown in panel ( b ) .
ultimately , this fact leads to the extinction of strategy @xmath6 , despite the fact that it is more successful in deterring criminals than strategy @xmath5 .
as soon as @xmath6 die out , as shown in panel ( c ) , criminals can exploit the milder punishment from strategy @xmath5 and spread towards the stationary state , as shown in panel ( d ) .
a seemingly surprising and counterintuitive result is that criminals , who can coexist with @xmath5 players but are defeated by @xmath6 players , are able to survive while their `` predators '' ( @xmath6 ) go extinct .
but in fact , the evolution depicted in the panels ( a)-(d ) simply illustrates the actual consequence of second - order free - riding .
namely , @xmath5 players exploit the more altruistic @xmath6 players by contributing less to sanctioning criminals . in the absence of @xmath5 players , however , the common enemy ( @xmath1 ) can spread relatively free and reach a significantly high level ( @xmath19 ) .
interestingly , when @xmath6 players are less successful in deterring @xmath1 players , the outcome is completely the opposite , as shown in panels ( e)-(h ) of fig .
[ snapshots ] . since the temptation / loss @xmath20
, @xmath1 are able to coexist with @xmath6 .
the coexistence of @xmath1 and @xmath5 strategies is also still possible , and at the same time @xmath5 continue to invade the pure @xmath6 phase [ the invasion ends in panel ( f ) ] .
however , @xmath5 become ineffective against the @xmath21 alliance .
indeed , this two - strategy alliance is so powerful that it beats the other @xmath22 alliance completely .
the competition between the two alliances starts in panel ( g ) , and it terminates with the total victory of the @xmath21 alliance in panel ( h ) .
the conclusion is similar as in the preceding case .
namely , when the evolution selects only one type of punishers , then criminals have a reasonable chance to survive .
note that the fraction of criminals in the stationary state is again relatively high , @xmath23 , despite of substantial punishment .
the most favorable outcome can be obtained at an intermediate temptation / loss value , as shown in panels ( i)-(l ) of fig .
[ snapshots ] .
the @xmath24 value is still high enough to maintain the coexistence of the @xmath21 alliance , but it lessens its evolutionary advantage in that the @xmath22 alliance is able to survive .
the stationary state thus contains three strategies , whereby a relatively small portion of the population , @xmath25 , is occupied by criminals .
we thus conclude that , in the long - run , if different punisher strategies survive in the stationary state , heterogeneous punishment may be utilized successfully to mitigate crime better than uniform punishment .
note that @xmath26 is a decreasing function of @xmath13 in the three - strategy phase in fig .
[ cext ] , while it always increasing when homogeneous punishment is applied ( in @xmath22 , @xmath21 , or in the @xmath27 phases ) .
this is because heterogeneous punishment enables the validation of the most effective approach against crime : sometimes moderate efforts , yielding milder fines , serve the interest of whole population better than severe punishment . even more importantly , the simultaneous presence of different types of punishers enables a synergy among them in that one strategy ( in our case @xmath6 ) can lower the payoff of criminals significantly while the other strategy ( @xmath5 ) can still enjoy a more competitive payoff due to a smaller cost .
this multi - point effect is conceptually similar to when the duty of punishment is shared stochastically among cooperative players @xcite . of course
, as we have already emphasized , these conclusions remain valid and can be extrapolated to a larger number of different punisher strategies .
we have studied the effectiveness of punishment in abating criminal behavior in the spatial inspection game with three and five competing strategies , entailing criminals , ordinary people and punishers . in the five - strategy game , we have introduced three different types of punishers , depending on the amount they are willing to contribute towards sanctioning criminals .
we have shown that cyclic dominance plays an important role in that it maintains the survivability of seemingly subordinate strategies through indirect support . for example , increasing the reward for punishing criminals might promote second - order free - riding of ordinary people , despite of the fact that it should in fact support the punishers .
this is due to cyclic dominance , where directly promoting the prey , in this case the punishers , benefits the predator , which in this case are the ordinary people .
moreover , we have shown that the actual obstacle in the fight against criminal behavior is the possibility of ordinary people to free - ride on the efforts of punishers , which is also the main culprit behind the establishment of cyclic dominance .
in general , sanctioning criminal behavior is thus a double - edged sword .
the obvious benefit is that the evolution of crime is contained and is unable to dominate in the population .
the pitfall is that , in conjunction with ordinary people , punishment creates conditions that support cyclic dominance , which prevents the complete abolishment of crime even if the sanctions are severe and effective .
in addition to these observations , we have shown that the possibility of heterogeneous punishment yields a highly ambiguous measure against criminal behavior . at specific parameter values
it can happen that milder punishers play the role of second - order free riders , which ultimately prevents to eliminate crime completely [ see panels ( a)-(d ) in fig .
[ snapshots ] ] .
evidently , the reverse process is also possible in structured populations where the more altruistic punishers can separate from second - order free riders and win the indirect territorial battle @xcite .
but in the realm of the studied inspection game , we have also observed that the diversity of punishers can yield a more favorable social outcome even as the temptation to do crime is growing . in the latter case , the simultaneous presence of different punishers provides an advantageous coexistence : some punishers ensure a higher fine to criminal players while other punishers can benefit from a lower cost due to a less intensive engagement .
importantly , neither of these two options is effective on its own right , but together they improve the effectiveness of combating crime .
notably , the emergence of cyclic dominance due to strategic complexity has been reported before , for example in public goods games with volunteering @xcite , peer punishment @xcite , pool punishment @xcite and reward @xcite , but also in pairwise social dilemmas with coevolution @xcite .
other counterintuitive phenomena that are due to cyclic dominance @xcite include the survival of the weakest @xcite , the emergence of labyrinthine clustering @xcite , and the segregation along interfaces that have internal structure @xcite , to name but a few examples .
cyclical interactions are thus in many ways the culmination of evolutionary complexity @xcite , and we here show that they likely play a prominent role in deterring crime as well . however , while the beneficial role of cyclic dominance for maintaining biodiversity is undeniable , one has to concur that it is a rather unsatisfactory outcome in terms of fighting criminal behavior . that is the sort of diversity in behavior that human societies could happily do without , yet it seems that this is precisely the trap the current system has fallen into .
indeed , data from the federal bureau of investigation ( see fig . 2 in ref .
@xcite ) indicate that crime , regardless of type and severity , is remarkably recurrent .
although positive and negative trends may be inferred , crime events between 1960 and 2010 fluctuate across time and space , and there is no evidence to support that crime rates are permanently decreasing .
the search for more effective crime mitigation strategies is thus in order , in particularly for such where the permanent elimination of crime is not an a priori impossibility .
for both the 3-strategy and the 5-strategy spatial inspection game the monte carlo simulation procedure is the same . initially all competing strategies are distributed uniformly at random on the square lattice .
we note , however , that the reported final stationary states are largely independent of the initial fractions of strategies .
subsequently , in agreement with the random sequential update protocol , a randomly selected player @xmath28 acquires its payoff @xmath29 by playing the game pairwise with all its four neighbors .
next , player @xmath28 randomly chooses one neighbor @xmath30 , who then also acquires its payoff @xmath31 in the same way as previously player @xmath28 .
once both players acquire their payoffs , player @xmath28 adopts the strategy @xmath32 from player @xmath30 with a probability determined by the fermi function @xmath33},\ ] ] where @xmath34 quantifies the uncertainty related to the strategy adoption process @xcite . in agreement with previous works
, the selected value ensures that strategies of better - performing players are readily adopted by their neighbors , although adopting the strategy of a player that performs worse is also possible @xcite .
this accounts for imperfect information and errors in the evaluation of the opponent .
each full monte carlo step ( mcs ) consists of @xmath35 elementary steps as described above , which are repeated consecutively , thus giving a chance to every player to change its strategy once on average .
we typically use lattices with @xmath36 players , although close to the phase transition points up to @xmath37 players had to be used in this case to avoid accidental extinctions , and thus to arrive at results that are valid in the large - size limit .
the fractions of competing strategies @xmath38 are determined in the stationary state after a sufficiently long relaxation time lasting up to @xmath39 mcs .
in general , the stationary state is reached when the average of the strategy fractions becomes time - independent .
moreover , to account for the differences in initial conditions and to further improve accuracy , the final results are averaged over up to @xmath40 independent runs for each set of parameter values .
this research was supported by the slovenian research agency ( grant p5 - 0027 ) , the hungarian national research fund ( grant k-101490 ) , and by the deanship of scientific research , king abdulaziz university ( grant 76 - 130 - 35-hici ) . | as a simple model for criminal behavior , the traditional two - strategy inspection game yields counterintuitive results that fail to describe empirical data .
the latter shows that crime is often recurrent , and that crime rates do not respond linearly to mitigation attempts .
a more apt model entails ordinary people who neither commit nor sanction crime as the third strategy besides the criminals and punishers . since ordinary people free - ride on the sanctioning efforts of punishers , they may introduce cyclic dominance that enables the coexistence of all three competing strategies . in this setup
ordinary individuals become the biggest impediment to crime abatement .
we therefore also consider heterogeneous punisher strategies , which seek to reduce their investment into fighting crime in order to attain a more competitive payoff .
we show that this diversity of punishment leads to an explosion of complexity in the system , where the benefits and pitfalls of criminal behavior are revealed in the most unexpected ways . due to the raise and fall of different alliances
no less than six consecutive phase transitions occur in dependence on solely the temptation to succumb to criminal behavior , leading the population from ordinary people - dominated across punisher - dominated to crime - dominated phases , yet always failing to abolish crime completely . in @xmath0 wilson and kelling
@xcite introduced the `` broken windows theory '' , explaining how seemingly unimportant and harmless signals of urban disorder may over time elicit antisocial behavior and serious crime .
the central premise of the theory is simple yet powerful , and it is reminiscent of preferential attachment or the matthew effect @xcite with a negative connotation .
just like the more connected nodes attract more new links during network growth @xcite , so does an unattended broken window invite bypassers to behave mischievously or even disorderly .
similarly , a graffiti might point to an unkept environment , signaling that more egregious damage will likely be tolerated as well .
one broken window is thus likely to become many broken windows , and the inception of urban decay and criminal behavior is in place .
the simplicity of this widely adopted criminological theory invites mathematicians and physicists to adopt a complex systems approach @xcite to study criminal behavior @xcite , in particular since the collective behavior of the system in this case can hardly be inferred from the relatively simple individual actions .
emergent phenomena such as pattern formation including percolation @xcite and phase transitions are commonly associated with complex social and biological systems @xcite , and in this realm the mitigation of crime is certainly no exception .
recent research highlights that crime is far from being uniformly distributed across space and time @xcite , and this is confirmed also by the dynamic nucleation and dissipation of crime hotspots @xcite and the emergence of complex criminal networks @xcite .
the emergence of crime can also be treated as a social dilemma @xcite , in as far that social order is the common good that is threatened by criminal activity , with competition arising between criminals and those trying to prevent crime .
an adversarial evolutionary game with four competing strategies has recently been proposed @xcite , where paladins are model citizens that do not commit crimes and collaborate with authorities , while villains , at the other extreme of the spectrum , commit crimes and do not report them .
intermediate figures are informants who report on other offenders while still committing crimes , and apathetics who neither commit crimes nor report to authorities .
apathetics are similar to second - order free - riders in the context of the public goods game with punishment @xcite , in that they cooperate at first order by not committing crimes , but defect at second order by not punishing offenders .
simulations have revealed that in the realm of the adversarial game informants are key to the emergence of a crime - free society , and this has subsequently been confirmed also with human experiments @xcite . in general , the mitigation of crime can be framed as an evolutionary game with punishment , although recent research has raised doubts on the use of sanctions as a means to promote prosocial behavior @xcite .
rewards for not doing and reporting crime are a viable alternative , and in this case the `` stick versus carrot '' dilemma becomes an important consideration @xcite . in the context of rehabilitating criminals , the question is also how much punishment for the crime and how much reward for eschewing wrongdoing in the future is in order for optimal results , as well as whether these efforts should be the responsibility of individuals or institutions @xcite under the assumption of a limited budget @xcite .
it is at this intersection of statistical physics of complex system and evolutionary games that we aim to contribute in the present paper by considering a three - strategy spatial inspection game with uniform punishment as well as a five - strategy spatial inspection game with heterogeneous punishment .
the inspection game is a recognized model in the sociological literature for the dynamics of crime @xcite .
the game addresses the question of why anybody would be willing to invest into costly punishment of criminals , given that individuals are tempted to benefit from the punishing activities of others without actively contributing to them .
as soon as ordinary people are introduced who neither commit crimes nor contribute to their mitigation , one is thus faced with the second - order free - rider problem @xcite .
as we will show in what follows , this may introduce cyclic dominance that enables the coexistence of all three competing strategies in the uniform punishment model .
more importantly , the consideration of heterogeneous punisher strategies drastically elevates the complexity of possible solutions , revealing on the one hand a more effective solution to the second - order free - rider problem , yet still failing to abolish crime completely . as a consequence ,
the diversity of punishment allows the formation of different alliances between competing strategies , which gives rise to a sophisticated range of solutions in dependence on the payoffs . in the next section we first present the details of the considered 3-strategy and 5-strategy spatial inspection game , and then demonstrate how systematic monte carlo simulations reveal the benefits and pitfalls of punishing criminal behavior .
simulation details are described in the methods section .
we conclude by discussing the presented results and their wider implications . |
inclusive lepton scattering has for many decades been the most important tool with which to probe the internal quark and gluon ( or parton ) structure of nucleons and nuclei .
structure functions extracted from inclusive deep - inelastic scattering ( dis ) experiments display the central features of quantum chromodynamics ( qcd ) asymptotic freedom at short distances ( via structure function scaling and its violation ) and confinement at large distance scales ( via parton momentum distributions ) .
since the late 1960s , dis experiments have yielded an impressive data set that maps nucleon structure functions over several orders of magnitude in the bjorken scaling variable , @xmath0 , and the squared four - momentum transfer , @xmath1 .
these data , supplemented by cross sections from hadronic collisions and other high - energy processes , have enabled a detailed picture of the parton distribution functions ( pdfs ) of the nucleon which are extracted through global qcd analyses ( see ref .
@xcite and references therein ) . at lower energies , where nonperturbative quark gluon interactions are important and the inclusive lepton
nucleon cross section is dominated by nucleon resonances , the structure functions reveal another intriguing feature of qcd , namely , _
quark - hadron duality_. here , the low energy cross section , when averaged over appropriate energy intervals , is found to resemble the high energy result , whose @xmath1 dependence is described by perturbative qcd . in this context ,
quark hadron duality provides a unique perspective on the relationship between confinement and asymptotic freedom , and establishes a critical link between the perturbative and nonperturbative regimes of qcd . in the framework of qcd , quark hadron duality
can be formally interpreted in terms of structure function moments @xcite . from the operator product expansion ( ope )
, the moments can be expressed as a series in @xmath2 , with coefficients given by matrix elements of local quark
gluon operators of a given twist .
the leading ( twist 2 ) term corresponds to scattering from a single parton , while higher twist terms correspond to involving multi
quark and quark gluon interactions .
since at low @xmath1 the resonance region makes a significant contribution to the structure function moments , one might expect a strong @xmath1 dependence of the low-@xmath1 moments arising from the higher twist terms of the ope . in practice
, however , the similarity of the structure function moments at low @xmath1 with the moments extracted at high energies suggests the dominance of the leading twist contribution , with the higher twist , multi - parton contributions playing a relatively minor role or that the large higher twists cancel each other .
this non - trivial relationship between the low - energy cross section and its deep - inelastic counterpart was first observed by bloom and gilman @xcite in the early dis measurements that were instrumental in establishing structure function scaling .
more recently , the availability of extensive , precise structure function data from jefferson lab , over a wide range of kinematics , has opened up the possibility for in - depth studies of quark - hadron duality .
duality has now been observed in the proton @xmath3 and @xmath4 structure functions @xcite , the nuclear structure function @xmath3 @xcite , the spin - dependent @xmath5 structure functions of the proton and @xmath6he @xcite , the individual helicity-1/2 and 3/2 virtual photoproduction cross sections for the proton @xcite . to establish the dynamical origin of quark - hadron duality one must also study the _
neutron_. four - quark higher twists contributions suggest that duality in the proton could arise from accidental cancellations between quark charges , which would not occur for the neutron @xcite .
unfortunately , the absence of high - density free neutron targets means that essentially all information on the structure functions of the neutron has had to be derived from measurements on deuterium .
typically , the deuterium data are corrected for fermi smearing and other nuclear effects @xcite , which introduces an element of model dependence into the procedure .
this is particularly problematic in the nucleon resonance region , where fermi motion leads to significant smearing of the resonant structures .
the existence of duality in the neutron @xmath3 structure function was suggested recently @xcite in an analysis which used an iterative deconvolution method @xcite to extract neutron resonance spectra from inclusive proton and deuteron @xmath3 data @xcite .
a _ direct _ confirmation of duality in the neutron , however , was not possible .
recently , a new experimental technique , based on spectator nucleon tagging @xcite , has been used to extract the free neutron @xmath3 structure function @xcite . by detecting low - momentum protons at backward angles in electron deuteron scattering , the bonus experiment at jefferson lab measured @xmath7 in both the resonance and dis regions , with minimal uncertainty from nuclear smearing and rescattering corrections @xcite . in the present paper ,
we use the bonus data to quantify for the first time the degree to which duality holds for the @xmath3 structure function of the free neutron . because the results reported here use data from an experimentally isolated neutron target
, one can expect significantly reduced systematic uncertainties compared with those in the model - dependent analysis of inclusive deuterium data @xcite . for the theoretical analysis of duality we use the method of _ truncated _ structure function moments developed by psaker _ et al . _
@xcite . here
, the @xmath8-th truncated moment of the @xmath3 structure function is defined as @xmath9 [ eq : moments2 ] where the integration over @xmath0 is restricted to an interval between @xmath10 and @xmath11 .
this method avoids extrapolation of the integrand into poorly mapped kinematic regions , and is particularly suited for the study of duality where an @xmath0 region can be defined by a resonance width around an invariant mass @xmath12 , where @xmath13 is the nucleon mass . as the position of the resonance peak varies with @xmath0 for different @xmath1 values , the values for @xmath10 and @xmath11 evolve to correspond to the appropriate invariant mass squared region .
for the bonus data , four ranges in @xmath14 were considered , corresponding to the three prominent resonance regions ( @xmath15 gev@xmath16 for the first or @xmath17 resonance region , @xmath18 gev@xmath16 for the second resonance region , and @xmath19 gev@xmath16 for the third resonance region ) , as well as the combined resonance region ( @xmath20 gev@xmath16 ) .
results on the lowest three non - trivial moments are reported and compared with recent global pdf parametrizations , as well as with previous model - dependent data analyses . in sec .
[ sec : experiment ] , we review the bonus experiment and the results for the neutron @xmath3 structure function .
the analysis of the truncated moments is discussed in sec .
[ sec : truncated ] , together with the implications for duality and its violation . finally , in sec .
[ sec : conclusion ] we summarize our results and discuss their wider implications .
the results reported here rely on a novel experimental technique aimed at eliminating or substantially reducing the theoretical uncertainties involved in extracting neutron data from nuclear targets .
the bonus ( barely off shell nucleon structure ) experiment @xcite used a radial time projection chamber ( rtpc ) to detect backward moving , low momentum spectator protons produced in electron
deuterium scattering in conjunction with electrons detected using clas @xcite in hall b at jefferson lab . by tagging low momentum backward moving spectator protons
one minimizes final state interactions @xcite and ensures that the neutron is just barely off
shell @xcite .
additionally , fermi smearing effects are essentially eliminated .
the bonus experiment ran in 2005 and acquired electron
deuteron scattering data at two electron beam energies : 4.223 , and 5.262 gev .
the rtpc consists of three layers of gas electron multipliers surrounding a thin , pressurized gas deuterium target which can detect spectator protons with momenta as low as 70 mev / c .
the scattered electron was detected by clas .
the experiment and data analysis are described in detail in @xcite . ratios of neutron to proton @xmath3 structure functions and the neutron @xmath3 structure function by itself were extracted over a wide kinematic range and for proton spectator momenta between 70 and 100 mev / c .
the total systematic uncertainty in the neutron structure function extracted is 8.7% @xcite .
additionally , there is an overall 10% scale uncertainty due to cross normalization of the bonus data to existing @xmath21 parameterizations .
the kinematic coverage shown in figure [ fig : bonus_kinematics ] ( 4.223 and 5.262 gev results combined ) extends from the quasielastic peak to the deep inelastic region corresponding to final state invariant masses of @xmath22 gev@xmath16 .
the curves shown represent the @xmath14 thresholds for the resonance mass regions mentioned above .
typical @xmath7 results for @xmath23 and @xmath24 gev@xmath16 are shown in figure [ fig : spectra ] .
the open / closed symbols correspond to data from 4.2 and 5.2 gev electron beam energies , respectively .
predictions , with and without higher twist effects , of the abkm qcd fit @xcite are also shown .
qualitatively , one can see evidence here for quark hadron duality in that the curves generally go through the resonance data .
the data used in this analysis had spectator angles with respect to the momentum transfer greater than 100@xmath25 and momenta between 70 and 100 mev / c .
thresholds for the four resonance mass regions mentioned in the text . ]
[ fig : bonus_kinematics ] gev@xmath16 ( top panel ) and @xmath26 gev@xmath16 ( bottom panel ) .
the open circles represent data obtained with a beam energy of @xmath27 gev , while the closed circles were obtained with @xmath28 gev beam energy .
the curves shown are the abkm dis parameterizations @xcite with ( solid line ) and without ( dashed line ) higher twist effects and target mass corrections . ]
[ fig : spectra ]
since @xmath1 , @xmath0 , and @xmath14 are not independent of each other , a range in @xmath14 at fixed @xmath1 implies a corresponding range in @xmath0 .
this allows for a straightforward integration of the experimental @xmath7 structure function data to obtain truncated moments . in order to minimize model dependence ,
the integrals were evaluated based solely on the experimentally measured points without using any inter- or extrapolating function . the second ( @xmath29 ) truncated moments , @xmath30 , obtained from these data
are shown in fig .
[ fig : moments ] as a function of @xmath1 , and are listed in table [ tab : moments ] .
the uncertainties quoted take into account the experimental statistical and systematic uncertainties but do not the include the 10% scale uncertainty due to cross normalization of the bonus data .
the corresponding higher order truncated moments ( @xmath31 and @xmath32 ) are provided in tables [ tab : moments4 ] and [ tab : moments6 ] , respectively .
, as a function of @xmath1 .
the closed circles represent the moments obtained from the bonus data , while the blue rectangles are the moments obtained from the abkm parameterization @xcite which includes target mass and higher twist corrections . ]
[ fig : moments ] .second order truncated moments of the neutron @xmath3 structure function based on the bonus data for the three resonance regions studied , as well as for the whole region @xmath33 gev@xmath16 . for @xmath34
gev@xmath16 the moment for the whole region was not computed due to the lack of data in the highest @xmath14 region .
[ cols= " < , < , < , < , < " , ] in order to study local duality , we formed the ratio of the truncated moments of @xmath7 in the resonance region obtained from the bonus data to the neutron moments calculated over the same @xmath0 range based on the abkm qcd fit @xcite .
these truncated moments are shown as a function of @xmath1 in fig .
[ fig : data_theory_m2 ] ( closed circles ) for @xmath30 . for the @xmath30 the earlier , model dependent results from @xcite
are also shown ( open circles ) .
the four panels correspond to the four invariant mass regions previously defined .
the computed abkm moments include higher twist effects and target mass corrections , which allows for a direct comparison of the present results with the studies from ref .
the figure also presents the ratio of truncated @xmath3 neutron moments computed using the cteq
jlab ( cj12 ) parameterization @xcite with respect to the same abkm moments .
the explicit inclusion of higher twist terms in these parameterization allows for extending the region of validity into the large@xmath0 region .
the interplay between higher twist and target mass corrections and their influence on the pdfs at large @xmath0 is studied in depth in ref .
in addition , the cj12 parameterization includes nuclear effects in part to improve the neutron , and hence @xmath35-quark pdf . to study the possible influence of target mass corrections and higher twists the cj12 model was used to obtain both the leading twist ( lt , thin dashed line ) and the target mass and higher twist corrected ( tmc+ht , thick solid line ) predictions
. the differences between the two models including higher twist and target mass corrections presented are within the uncertainty of the data . in the resonance region ,
target mass corrections and higher twist effects are expected to be sizable and so it may not be surprising that agreement of the data moments with leading twist parameterizations is improved by the inclusion of ht and tmc effects .
however , the abkm and cj12 parameterizations extracted tmc and ht corrections from deep inelastic scattering data at higher @xmath1 . therefore confirmation of duality indicates that these contributions are similar in the resonance region and in the dis .
an alternative possibility is that the pdf fits have a very large uncertainty at high @xmath0 and perhaps need to increase the @xmath36 values to better describe the data . in fig .
[ fig : data_theory_m2 ] the first @xmath37 range corresponding to the @xmath17 resonance displays the same @xmath1 behavior , but is 2040% larger than the model expectation .
this underscores the likelihood that the pdf based fits underestimate the large @xmath0 regime , to which this @xmath37 region corresponds .
although the bonus results reported here are in good agreement with the earlier data extracted from deuteron using modeling ( open circles ) @xcite , and both are consistent with unity in the second and third resonance region , the bonus data are consistently higher in the first resonance region .
although all local duality studies on the neutron are sensitive to binding effects , fermi motion , and final state interactions , obtaining neutron structure functions by subtracting ( smeared ) proton from deuteron data is especially sensitive to the nuclear models employed at the largest @xmath0 ( i.e. smallest @xmath37 ) values .
therefore , this discrepancy could be due to either a difference in the free neutron / proton @xmath17 resonance excitation or a manifestation of nuclear theory uncertainties .
interestingly , the bonus / cj12 ratios assuming no higher twist , would fall consistently below unity , as evidenced by the dashed line .
moments in the resonance region display the same @xmath1 dependence as the pdf predictions , i.e. the ratio of the two shown in fig .
[ fig : data_theory_m2 ] is independent of @xmath1 .
this is a remarkable confirmation of one aspect of duality , _
i.e. _ that the large @xmath0 and lower @xmath1 , region can be described by the same evolution equations displaying the same dynamics as the higher @xmath1 , larger @xmath37 , scattering regime of single quark deep inelastic scattering .
this is true even for the very largest @xmath0 , lowest @xmath37 , area . here ,
the uncertainties on the pdfs are large and could easily account for the @xmath38% strength difference , but not the @xmath1dependence .
truncated moments for the neutron structure function @xmath3 obtained from the bonus data and the same quantity obtained from the abkm parameterization @xcite ( closed circles ) .
the abkm calculation includes both higher twists and target mass corrections .
the open circles represent the model
dependent results obtained by @xcite .
the lines represent comparisons of the abkm with the cj12 parameterizations@xcite : thick solid line for the case when tmc and ht effects are included in both models , thin dashed line when the cj model is evaluated only for leading twist . ]
[ fig : data_theory_m2 ] local quark
hadron duality in the proton @xmath3 structure function has been studied extensively @xcite , showing possible violations mainly in the first resonance region . to compare the proton with the neutron results ,
truncated @xmath3 moments were constructed for the proton using the abkm model to calculate the deep inelastic moment and a global fit @xcite for the resonance region .
[ fig : data_f2nf2p ] shows the ratio of the neutron to proton truncated moments ( closed circles ) as a function of @xmath1 for the four resonance regions .
the line shown is the ratio of neutron and proton truncated moments calculated using the abkm ( including ht and tmc ) model .
the model agrees well with the data in the second and third resonance regions but substantially underestimates the data in the first resonance region .
this can be indicative of a fundamental difference between the proton and the neutron at the very largest @xmath0 values corresponding to the @xmath17 resonance region .
alternatively , this can be the result of the fact that qcd fits for either proton or neutron are not well constrained at large @xmath0 , which is the kinematic range of the first resonance region .
truncated moments , @xmath39 for the four regions studied .
the closed circles represent the bonus results , while the solid line corresponds to the abkm qcd fit prediction including ht and tmcs . ]
[ fig : data_f2nf2p ] in conventional parameterizations of parton distribution functions the @xmath40 ratio is often assumed to go either to zero or infinity as @xmath41 , depending on the parameterization of the @xmath35 pdf @xcite . at @xmath0 above 0.8
the uncertainties on the pdfs , especially @xmath35 , are considerable , due to deuteron nuclear model corrections and to the lack of large @xmath0 data .
the ratio @xmath40 , and correspondingly @xmath42 , at large @xmath0 is however of fundamental interest @xcite and there are numerous qcd based theoretical predictions for this quantity .
all require @xmath43 and hence @xmath44 for a given @xmath0 and @xmath1 .
this is in contrast to naive predictions for the @xmath17 resonance . for the latter ,
given that the proton and neutron transitions to the @xmath17 are isovector , the resonant contributions should be identical , which would lead to an @xmath42 ratio close to unity .
it is perhaps not surprising that the resonance behavior dominates in this @xmath37 region because the @xmath17 region contains only one resonance with less continuum background than the other regions .
similar tension between the resonance data and dis/ pqcd - derived moments could exist in the second and third resonance regions as well . assuming a dominance of magnetic couplings , as some quark models do @xcite
, the proton resonance data should overestimate the dis prediction in these regions due to the odd
parity resonances , such as the prominent spin@xmath45 octet , resulting in a measured neutron to proton truncated moment ratio that should fall below the corresponding dis curve . as shown in fig .
[ fig : data_f2nf2p ] the results presented here do not support this argument , and duality is apparently the predominant effect for regions above the @xmath17 .
in conclusion , this paper investigates local quark
hadron duality in the neutron structure function based on data obtained by the bonus experiment at jefferson lab , which used a novel experimental technique to tag the spectator proton in a deuterium target and thereby create an effective neutron target .
this technique provides smaller systematic uncertainties than earlier studies that relied on the subtraction of smeared hydrogen data and nuclear modeling from deuterium .
truncated @xmath3 structure function moments were compared to pdf fits based on pqcd and largely deep inelastic scattering data , as well as to similar truncated moments obtained for the proton .
the results indicate that quark
hadron duality holds dynamically everywhere , as well as in sum for the second and third resonance regions .
this data provides further confirmation that duality , still not well understood , appears to be a fundamental aspect of nucleon structure . | recent jefferson lab results on quark - hadron duality are presented . |
a remarkable achievement of theoretical cosmology has been the construction , by belinski , khalatnikov and lifshitz ( bkl ) , of a general solution to the @xmath0-dimensional vacuum einstein equations in the vicinity of a spacelike ( `` cosmological '' ) singularity @xcite , @xcite @xcite , @xcite .
they found that this solution exhibits a never - ending oscillatory behavior , with strong chaotic properties .
they could describe in detail the statistical properties of this never - ending oscillatory behavior by approximating the einstein field equations ( near the singularity ) by a system of ode s for three variables @xmath1 , @xmath2 , @xmath3 ( ` anisotropic scale factors ' ) , namely [ abc ] @xmath4 where @xmath5 , and by approximately reducing the continuous dynamics of @xmath1 , @xmath2 , @xmath3 to a sequence of discrete maps .
the crucial discrete map introduced by bkl relates the ` kasner exponents ' @xmath6 , @xmath7 , @xmath8 describing the ( approximately linear ) @xmath9-evolutions of the three scale factors @xmath1 , @xmath2 , @xmath3 during two successive ` epochs ' ( i.e. two successive segments of the dynamics ( [ abc ] ) during which the influence of the right - hand side is negligible ) .
later studies have refined the description of the statistical properties of the chaotic bkl oscillations , notably by introducing and studying more complete discrete iteration maps ( involving several real variables ) , and notably by a two - dimensional discrete map @xcite , @xcite , @xcite .
+ separately from the work of bkl , and , with a different aim and motivation , misner realized that generic bianchi ix homogeneous cosmological models have a `` very complex singularity '' @xcite : a reformulation of this dynamics @xcite , @xcite led to the simple picture of a point moving on a lobachevsky plane and reflecting upon _ fixed _ billiard - type cushions .
this led chitre to remark that the dynamics of the system point is ergodic and mixing , with unique invariant liouville measure .
the problem of relating the statistical properties of the discrete bkl map to the invariance of the liouville measure , in the continuous billiard dynamics , la misner - chitre has been considered in some detail by kirillov and montani @xcite . + the description of cosmological singularities in term of _ billiards _ in ( higher dimensional ) lobachevsky ( or lorentzian ) spaces has recently received a new impetus from the discovery that the billiard chambers corresponding to many interesting physical theories can be identified with the `` weyl chambers '' of certain ( infinite - dimensional ) lorentzian kac - moody algebras @xcite , @xcite , @xcite
. this has raised the conjecture that , hidden below the bkl `` chaos '' , there lies a remarkable `` kac - moody symmetry '' , akin to the duality symmetries of supergravity and string theories @xcite , @xcite , @xcite .
more precisely , in lorentzian space , the lagrangian @xmath10 for these models reads @xmath11 where @xmath12 , and where the potential term @xmath13 determines the billiard walls @xmath14 . in the case of pure gravity , the only walls that can appear are either the gravitational walls @xmath15 or the symmetry walls @xmath16 . in @xmath17 dimensions ,
the usual bkl billiard table is obtained by considering the three gravitational walls @xmath18 , @xmath19 , @xmath20 .
this big billiard corresponds to considering a _ diagonal metric _
, described by the three scale factors @xmath21 , @xmath22 , @xmath23 .
the more general case of a _ non - diagonal _ metric corresponds to a small billiard enclosed by the gravitational wall @xmath24 , together with the two leading symmetry walls @xmath25 and @xmath26 .
projecting the @xmath27-dimensional lorentzian @xmath28 space on a lobachevsky plane allows one to obtain some useful representations of the billiard dynamics .
+ we focus here on a few results from ref . .
the dynamics on a billiard table consists of free - flight evolution and of bounces on the boundaries of the billiard .
the trajectories describing the free - flight evolution are called ( kasner ) epochs , while a set of trajectories taking place in the same corner of the billiard table is called a ( kasner ) era .
six kinds of eras can be defined , according to the corner where an era takes place , and the orientation of the first epoch of the era .
all the information necessary to describe the dynamics can be encoded in maps that take into account collisions on given walls .
+ ] , height=153 ] there are two complementary descriptions of billiards : the unit disk and the poincar half plane , sketched in fig .
[ fb ] and fig .
[ fc ] , respectively .
+ the unit disk is the @xmath29-dimensional version of the general @xmath30-dimensional ` ball ' conformal representation , geometrically realized [ within the @xmath31-dimensional lorentzian @xmath32-space ] by stereographically projecting ( from the ` south pole ' @xmath33 , i.e. a center of projection located on the _ past _ unit hyperboloid ) the _ future _ unit hyperboloid onto a ( @xmath30-dimensional ) hyperplane passing through the origin in @xmath32-space .
this representation is powerful in outlining the symmetry group of order six that characterizes the @xmath34-dimensional version of this model .
+ the poincar model can be obtained by a suitable geometric inversion of the ball model .
the poincar half plane offers a less symmetric representation , but it offers a convenient parametrization both of the dynamics and of the oriented endpoints of the geodesics . each collision can be described by a linear fractional transformation of the variables ( of the plane ) : @xmath35 for each given wall , there exists an application of the form ( [ 3 ] ) which transforms each geodesic incident on the considered wall in the reflected geodesic . when parametrizing a geodesic by the two real numbers ( @xmath36 ) parametrizing its end points on the absolute ( see fig .
[ fc ] ) , the transformation ( [ 3 ] ) also applies , separately , to @xmath37 and @xmath38 .
note , in particular , that the single real parameter @xmath38 parametrizes the kasner exponents of the corresponding kasner epoch .
this @xmath38 parametrization of the kasner exponents actually coincides ( as emphasized by kirillov and montani ) with the usual bkl parametrization , namely @xmath39 several integral invariants can be constructed for billiards .
their role is to provide conserved quantities ( that can be used in the statistical analysis ) charactering the dynamics .
+ the usually considered energy - shell liouville measure is a three - form obtained by reducing to the hypersurface @xmath40 the four - form @xmath41 , where @xmath42 is the ( invariant ) poincar - cartan @xmath29-form .
however , one can also construct another invariant measure by directly restricting the poincar - cartan two - form on any transverse section of the hamiltonian flow . in the case of the poincar plane ( asymptotic )
hamiltonian for billiards , @xmath43 the transversally reduced poincar - cartan two - form reads @xmath44_{v = const}=du\wedge d p_u=2\frac{du^+\wedge du^-}{(u^+-u^-)^2}.\ ] ] the two - form ( [ 7 ] ) is then found ( both by general arguments , and by explicit calculation ) to be invariant under the discrete billiard map , which connects one collision to the next one .
each billiard map is given by eq .
( [ 3 ] ) and is therefore of the form @xmath45 .
+ the invariant two - form ( [ 7 ] ) is linked to the liouville @xmath27-form by @xmath46 , where @xmath47 is a phase - space parameter measuring the length along each geodesic segment .
we describe the dynamics of the _ big billiard _ as a hopscotch game at different levels , such that different properties can be stressed upon . + the epoch - hopscotch description of the billiard dynamics
is obtained by dividing the complete billiard table into six regions @xmath48 , each denoting the portion of the reduced phase space where all the epochs going from the wall @xmath49 to the wall @xmath50 take place .
[ fe ] illustrates the jumping around of successive epochs in the ( @xmath51 ) plane : from the epoch number @xmath52 ( from @xmath2 to @xmath1 ) to the epoch number @xmath53 ( from @xmath2 to @xmath3 ) .
each transformation between successive epochs is given by linear fractional transformations of ( @xmath51 ) .
+ we can also use the symmetry group @xmath54 ( of order @xmath55 ) to _ quotient _ the dynamics of the big billiard .
one way would be to consider a _ kaleidoscopic _ version of the big billiard dynamics in which the single `` moving ball '' of the billiard is augmented by its @xmath53 ( generically distinct ) images under @xmath54 .
this leads to a billiard game where @xmath55 ( symmetry related ) balls simultaneously move within the same billiard table , and ( simultaneously ) bounce on its bounding walls .
a second way to look at the quotiented dynamics is to replace the latter kaleidoscopic phase - space point + @xmath56 by its unique representative , say @xmath57 , within a fundamental domain of @xmath54 , together with its corresponding momenta @xmath58 .
when passing from the continuous billiard dynamics to the discrete billiard map from an epoch to the next epoch , the quotienting of the big billiard leads to a quotiented version of the epoch hopscotch game of fig .
[ fe ] . in the alternative , fundamental - domain , version of the quotiented dynamics
we could replace each @xmath54 orbit in the ( @xmath59 ) plane by its unique representative located within , say , the box @xmath60 .
this is achieved by means of a suitable kasner transformation .
[ indeed , the six boxes @xmath48 of fig
. [ fe ] are exchanged under @xmath54 . ] in that view , the discrete quotiented big billiard would become a map from the box @xmath60 onto itself .
if we ignore the @xmath61 coordinate , we see that the law giving the successive values of the @xmath62 coordinate coincides with the law found long ago by bkl .
this shows that the bkl discrete dynamics of the variable @xmath63 is obtained by : ( i ) quotienting our more complete hopscotch dynamics by the permutation group @xmath54 , and ( ii ) ignoring the @xmath61 coordinate and identifying the bkl variable @xmath63 with @xmath62 .
+ the discrete epoch hopscotch map ( both in its unquotiented and in its quotiented versions ) leaves invariant the measure @xmath64 , eq .
( [ 7 ] ) .
in addition , by marginalizing @xmath65 over @xmath61 , this leads to the following invariant one - dimensional measure for the usual bkl map on the @xmath63 parameter ( in the @xmath66 convention ) : @xmath67 as far as we know , this result has not been explicitly discussed before in the literature .
+ however , one can not avail oneself of usual results of ergodic theory because the integral of @xmath65 on its total domain ( e.g @xmath60 for the quotiented map @xmath68 ) is _ infinite _ ( similarly the integral of @xmath69 is logarithmically divergent ) . in order to be able to use the usual results of ergodic theory
, one needs a ( discrete ) invariant map preserving a _ finite _ measure
. this can be achieved by considering the _ era _ hopscotch map ( either in the unquotiented or quotiented forms).an _ era _ hopscotch dynamics is obtained simply by ignoring the intermediate epochs and focusing on the map transforming the ( possibly quotiented ) first epoch of an era @xmath70 into the first epoch @xmath71 of the next era .
the unquotiented era - hopscotch court is sketched in fig .
[ ff ] , and the explicit expression of the unquotiented era map @xmath72 is listed in table [ t1 ] . in the _ quotiented case _ ,
the resulting discrete ( two - dimensional ) map , say @xmath73 , coincides with the map studied by chernoff - barrow and by lifshitz - khalatnikov - sinai - khanin - shchur .
when using a representative of the @xmath54 orbit within the @xmath60 box , the quotiented era map @xmath73 is a map of @xmath74 onto itself , i.e. the rectangular domain @xmath75 , @xmath76 .
the explicit expression of the map @xmath73 is @xmath77}-1.\ ] ] this maps leaves invariant the _ restriction _ of the two - form @xmath65 to the domain @xmath74 ; by contrast with the original measure @xmath65 on the full hopscotch court , this restricted measure has now a _ finite integral _ , namely @xmath78 .
quotienting the dynamics of the big billiard implies a loss of information with respect to the full unquotiented dynamics .
we analyse here what kind of information is lost , and show how to recover this information . + within the notation we have developed , the continued - fraction expansions$ ] for @xmath79 denotes the usual integer part of @xmath49 when @xmath80 ( e.g. @xmath81=3 $ ] ) , and @xmath82\le0 $ ] when @xmath83 ( so that @xmath84=-3 $ ] ) ; we define the fractional part of a negative number accordingly , e.g. @xmath85 .
furthermore , we have denoted the first integer of the decomposition of @xmath86 as @xmath87 $ ] , so that @xmath88 + 1 $ ] denotes the length of the era starting with @xmath86 .
] of the first ( @xmath59 ) values of a ( quotiented ) era starting in the box @xmath74 can be written as furthermore , if we start an unquotiented era hopscotch dynamics in some specific region @xmath94 , we can first map it to the reference region @xmath74 by some specific ` kasner transformation ' @xmath95 ( listed in table [ table6 ] ) to get its @xmath96-representative , @xmath97}_{f_{xy}}$ ] , namely @xmath98}_{f_{xy}}\equiv k_{xy}u^+_{f_{xy}}\equiv n_{xy}-1+[n_2 , n_3 , ... ] .\ ] ] we see that the quotiented era dynamics encodes the lengths of successive eras , but ignores the fuller information contained in the original , unquotiented era hopscotch dynamics , namely the precise _ corner _ , @xmath99 ( with @xmath100 ) and _ orientation _ , ( @xmath101 ) , ( i.e. @xmath102 ) , of the first epoch of the @xmath90-th hopscotch era , such that one could express the explicit result of iterating @xmath103 times the unquotiented @xmath72 map in terms of the simpler action of the quotiented @xmath73 map , namely @xmath104 the missing information is recovered by considering the quantities where @xmath106 , @xmath107 denotes the number of epochs contained in the @xmath108-th era , and @xmath109 $ ] is the determinant of the first kasner transformation @xmath95 involved in ( [ un ] ) .
the procedure determining @xmath110 is as follows : ( i ) starting from @xmath95 and the continued - fraction decomposition of @xmath111 $ ] , one determines the @xmath112 , and the @xmath113 and @xmath114 ; ( ii ) then @xmath113 determines the rotation ( in the unit - disc representation ) between @xmath115 and the final corner @xmath101 , and @xmath114 determines whether this corner is `` parallel '' ( @xmath116 ) or `` antiparallel '' ( @xmath117 ) to the initial @xmath115 ; ( iii ) finally , knowing the ordered corner @xmath118 , table [ table6 ] determines the transformation @xmath119 that maps it onto the @xmath96 corner .
periodic orbits offer another way to contrast the unquotiented dynamics with its quotiented version .
first , any @xmath120-periodic orbit of the one - dimensional bkl map @xmath121 , parametrized by the special values of @xmath62 that admit a ( regular ) periodic continued - fraction expansion of the type @xmath122 \ , , \ ] ] gives rise to a unique corresponding @xmath120-periodic orbit of the two - dimensional @xmath73 map .
second , any @xmath120-periodic orbit of @xmath121 and @xmath73 can be lifted to some periodic orbit of period @xmath123 of the full , unquotiented billiard , according to the condition @xmath124 where @xmath125 is a suitable kasner transformation ( which depends on @xmath120 and on the considered periodic orbit of @xmath73 ) , and @xmath126 is the order of @xmath125 .
the _ order _ of a particular group element , such as @xmath125 , is the smallest integer @xmath126 such that @xmath127 ; as a transposition is of order @xmath29 , and a cyclic permutation , ( @xmath128 ) or ( @xmath129 ) , of order @xmath27 , we see that the order @xmath126 of @xmath125 must be equal to @xmath130 or @xmath27 . in other words ,
the period in the unquotiented billiard can be up to three times larger than the one in the quotiented one .
for example , the simplest @xmath52-periodic orbit of @xmath73 given by the golden ratio @xmath131=(1+\sqrt{5})/2 $ ] corresponds to a @xmath27-periodic orbit of @xmath72 , as shown in fig .
[ fa ] . 0 v. a. belinskii and i. m. khalatnikov , _ sov .
* 29 * ( 1969 ) 911 [ _ zh .
* 56 * ( 1969 ) 1710 ] .
i. m. khalatnikov and e. m. lifshitz , _ phys .
* 24 * ( 1970 ) 76 .
v. a. belinskii and i. m. khalatnikov , _ sov .
. jetp _ * 30 * ( 1970 ) 1174 [ _ zh .
_ * 57 * ( 1969 ) 2163 ] .
v. a. belinskii , e. m. lifshitz and i. m. khalatnikov , _ sov .
_ * 13 * ( 1971 ) 745 [ _ usp .
* 102 * ( 1970 ) 463 ] .
d. f. chernoff and j. d. barrow , _ phys .
* 50 * ( 1983 ) 134 .
e. m. lifshitz , i. m. khalatnikov , ya .
g. sinai , k. m. khanin , l. n. shchur , _ jetp letters _ * 38 * ( 1983 ) 91 [ _ p .
* 38 * ( 1983 ) 79 ]
. i. m. khalatnikov , e. m. lifshitz , k. m. khanin , l. n. shchur and ya . g. sinai , _ j. stat
_ * 38 * ( 1985 ) 97 . c. w. misner , _ phys .
* 22 * ( 1969 ) 1071 . c. w. misner , k. s. thorne and j. a. wheeler , _ gravitation _ , ( w. h. freeman and company , san francisco , 1973 ) .
d. m. chitre , phd thesis ( university of maryland , 1972 ) .
a. a. kirillov and g. montani , _ phys .
* d 56 * ( 1997 ) 6225 .
t. damour and m. henneaux , _ phys .
_ * d 86 * ( 2001 ) 4749 .
t. damour , m. henneaux , b. julia and h. nicolai _ phys .
* b 509 * ( 2001 ) 323 .
t. damour , s. de buyl , m. henneaux and c. schomblond , _ jhetp _ * 0208 * ( 2002 ) 030 .
t. damour , m. henneaux and h. nicolai , _ phys .
_ * 89 * ( 2002 ) 221601 .
t. damour , m. henneaux and h. nicolai , _ class .
grav . _ * 20 * ( 2003 ) r145 .
m. henneaux , d. persson and p. spindel , _ living rev .
* 11 * ( 2008 ) 1 .
t. damour and o. m. lecian , statistical properties of cosmological billiards , [ arxiv:1011.5797 ] , to appear in _ phys . | we summarize some recent progress in the understanding of the statistical properties of cosmological billiards . |
he field of visual media has been witnessing explosive growth in recent years , driven by significant advances in technology that have been made by camera and mobile device manufacturers , and by the synergistic development of very large photo - centric social networking websites , which allow consumers to efficiently capture , store , and share high - resolution images with their friends or the community at large .
the vast majority of these digital pictures are taken by casual , inexpert users , where the capture process is affected by delicate variables such as lighting , exposure , aperture , noise sensitivity , and lens limitations , each of which could perturb an image s perceived visual quality .
though cameras typically allow users to control the parameters of image acquisition to a certain extent , the unsure eyes and hands of most amateur photographers frequently lead to occurrences of annoying image artifacts during capture .
this leads to large numbers of images of unsatisfactory perceptual quality being captured and stored along with more desirable ones .
being able to automatically identify and cull low quality images , or to prevent their occurrence by suitable quality correction processes during capture are thus highly desirable goals that could be enabled by automatic quality prediction tools @xcite .
thus , the development of objective image quality models from which accurate predictions of the quality of digital pictures as perceived by human observers can be derived has greatly accelerated .
advances in practical methods that can efficiently predict the perceptual quality of images have the potential to significantly impact protocols for monitoring and controlling multimedia services on wired and wireless networks and devices .
these methods have the potential to also improve the quality of visual signals by acquiring or transporting them via `` quality - aware '' processes .
such `` quality - aware '' processes could perceptually optimize the capture process and modify transmission rates to ensure good quality across wired or wireless networks .
such strategies could help ensure that end users have a satisfactory quality of experience ( qoe ) .
the goal of an objective no - reference image quality assessment ( nr iqa ) model is as follows : given an image ( possibly distorted ) and no other additional information , automatically and accurately predict its perceptual quality . given that the ultimate receivers of these images are humans ,
the only reliable way to understand and predict the effect of distortions on a typical person s viewing experience is to capture opinions from a large sample of human subjects , which is termed _ subjective image quality assessment_. while these subjective scores are vital for understanding human perception of image quality , they are also crucial for designing and evaluating reliable iqa models that are consistent with subjective human evaluations , regardless of the type and severity of the distortions
. the most efficient nr iqa algorithms to date are founded on the statistical properties of natural images .
natural scene statistics ( nss ) models @xcite are based on the well - established observation that good quality real - world photographic images obey certain _ perceptually _ relevant statistical laws that are violated by the presence of common image distortions .
some state - of - the - art nr iqa models @xcite that are based on nss models attempt to quantify the degree of ` naturalness ' or ` unnaturalness ' of images by exploiting these statistical perturbations .
this is also true of competitive reduced - reference iqa models @xcite .
such statistical ` naturalness ' metrics serve as image features which are typically deployed in a supervised learning paradigm , where a kernel function is learned to map the features to ground truth subjective quality scores .
a good summary of such models and their quality prediction performance can be found in @xcite .
* authentic distortions : * current blind iqa models @xcite use legacy benchmark databases such as the live image quality database @xcite and the tid2008 database @xcite to train low - level statistical image quality cues against recorded subjective quality judgements .
these databases , however , have been designed to contain images corrupted by only one of a few synthetically introduced distortions , e.g. , images containing only jpeg compression artifacts , images corrupted by simulated camera sensor noise , or by simulated blur .
though the existing legacy image quality databases have played an important role in advancing the field of image quality prediction , we contend that determining image quality databases such that the distorted images are _ derived _ from a set of high - quality source images and by simulating image impairments on them is much too limiting .
in particular , traditional databases fail to account for difficult mixtures of distortions that are inherently introduced during image acquisition and subsequent processing and transmission .
for instance , consider the images shown in fig .
[ sampleimgs](a ) - fig .
[ sampleimgs](d ) .
figure [ sampleimgs](d ) was captured using a mobile device and can be observed to be distorted by both low - light noise and compression errors .
figure 1(b ) and ( c ) are from the legacy live iqa database @xcite where jpeg compression and gaussian blur distortions were synthetically introduced on a pristine image ( fig .
[ sampleimgs](a ) ) .
although such singly distorted images ( and datasets ) facilitate the study of the effects of distortion - specific parameters on human perception , they omit important and frequently occurring mixtures of distortions that occur in images captured using mobile devices .
this limitation is especially problematic for _ blind _ iqa models which have great potential to be employed in large - scale user - centric visual media applications .
designing , training , and evaluating iqa models based only on the statistical perturbations observed on these restrictive and non - representative datasets might result in quality prediction models that inadvertently assume that every image has a `` single '' distortion that most objective viewers could agree upon . although top - performing algorithms perform exceedingly well on these legacy databases ( e.g. , the median spearmann correlation of @xmath0 on the legacy live iqa database @xcite reported by brisque @xcite and @xmath1 reported by tang _ et .
al _ in @xcite ) , their performance is questionable when tested on naturally distorted images that are normally captured using mobile devices under highly variable illumination conditions .
indeed , we will show in sec .
[ sec : expt ] that the performance of several top - performing algorithms staggers when tested on images corrupted by diverse authentic and mixed , multipartite distortions such as those contained in the new live in the wild image quality challenge database .
@xmath2{womanhat - eps - converted - to.pdf } & \includegraphics[width=1.3in , height=2.0in]{womanhat_jpeg - eps - converted - to.pdf } & \includegraphics[width=1.3in , height=2.0in]{womanhat_wn - eps - converted - to.pdf } & \includegraphics[width=1.5in , height=2.0in]{challenge - eps - converted - to.pdf } \end{array}$ ] with this in mind , we formulated the goal of designing a unique and challenging database containing a large number of _ authentically _ distorted images .
of course , we had to consider the question : what makes an image quality database representative ?
beyond capturing a wide variety of image content including pictures of people , objects , and both indoor and outdoor scenes , such a database should also account for the various and diverse frequently used commercial image capture devices .
the corpus of pictures should also have been obtained under varied illumination conditions and , given the propensity of users to acquire their pictures in imperfect ways , they should exhibit a broad spectrum of authentic quality `` types , '' mixtures , and distortion severities .
the result of our effort is a difficult new resource called the * live in the wild image quality challenge database*. of course , since we did not synthetically distort images , no pristine references are available ; hence the new database is only suitable for no - reference iqa research .
* large - scale subjective study : * since existing records of human quality judgements are associated with legacy `` synthetic '' databases , another important contribution that we made is to acquire human opinion scores on authentically distorted images .
most existing no - reference image quality assessment models follow a supervised learning paradigm , hence their accuracy is strongly linked to the quality and quantity of the annotated training data available . having access to more ratings per image means that variations of perceived quality can be more easily captured . the human opinion scores in most of the legacy datasets
were collected by conducting subjective studies in fixed laboratory setups , where images were displayed on a single device with a fixed display resolution and which the subjects viewed from a fixed distance .
however , significant advances in technology made by camera and mobile device manufacturers now allow users to efficiently access visual media over wired and wireless networks .
thus , the subjective image quality opinions gathered under artificially controlled settings do not necessarily mirror the picture quality perceived on widely used portable display devices having varied resolutions . gathering representative subjective opinions by simulating different viewing conditions would be exceedingly time - consuming , cumbersome , and would require substantial manual effort .
on the other hand , exploring novel ways to collect subjective scores online requires dealing with noisy data , establishing the reliability of the obtained human opinions , etc . - a challenging research topic .
as we describe in the following , we have conducted such a study to obtain a very large number of human opinion scores on the live challenge image data using a sophisticated crowdsourcing system design .
we explain the human study , the method of validating the data obtained , and demonstrate how it can be used to fruitfully advance the field of no - reference / blind image quality prediction .
* contributions : * we describe our attack on the difficult problem of blind image quality assessment on authentically distorted images from the ground up and summarize our contributions below : first , we introduce the content and characteristics of the new live in the wild image quality challenge database , which contains @xmath3 authentically distorted images captured from many diverse mobile devices .
each image was collected without artificially introducing any distortions beyond those occurring during capture , processing , and storage by a user s device .
next , we aimed to gather very rich human data , so we designed and implemented an extensive online subjective study by leveraging amazon s crowdsourcing system , the mechanical turk
. we will describe the design and infrastructure of our online crowdsourcing system and how we used it to conduct a very large - scale , multi - month image quality assessment subjective study , wherein a wide range of diverse observers recorded their judgments of image quality .
we also discuss the critical factors that are involved in successfully crowdsourcing human iqa judgments , such as the overall system design of the online study , methods for subject validation and rejection , task remuneration , influence of the subjective study conditions on end users assessment of perceptual quality , and so on . as a demonstration of the usefulness of the study outcomes
, we also conducted extensive empirical studies on the performance of several top - performing nr iqa models ( sec .
[ sec : expt ] ) , both on a legacy benchmark dataset @xcite as well as on the new live in the wild image quality challenge database .
our results demonstrate that : ( 1 ) state - of - the - art nr iqa algorithms perform poorly on the live in the wild image quality challenge database , which has a high percentage of images distorted by multiple processes , all of which are authentic ( sec .
[ sec : difftech ] ) .
( 2 ) our human - powered , crowdsourcing framework proved to be an effective way to gather a large number of opinions from a diverse , distributed populace over the web .
so far , we have collected over @xmath4 human opinion scores on 1,162 naturally distorted images from over @xmath5 distinct subjects , making it the world s largest , most comprehensive online study of perceptual image quality ever conducted .
( 3 ) a correlation of @xmath6 was obtained between the mos values gathered from the proposed crowdsourcing platform and those from a different study conducted by the creators of the live multiply distorted image quality database @xcite .
this high correlation advocates the veracity of our designed online system in gathering reliable human opinion scores ( sec .
[ goldstandard ] ) .
* benchmark iqa databases - content and test methodologies : * [ sec : databases ] most of the top - performing iqa models ( full , reduced , and no - reference ) have been extensively evaluated on two benchmark databases : the live iqa database which was designed in @xmath7 and the tid2008 database , designed and released in @xmath8 .
the live iqa database , one of the first comprehensive iqa databases , consists of @xmath9 images , much larger than the small databases that existed at the time of its introduction @xcite -@xcite .
this legacy database contains @xmath10 pristine reference images and models five distortion types - jp2k , jpeg , gaussian blur , white noise , and fast fading noise @xcite .
the tid2008 database is larger , consisting of @xmath11 reference and @xmath12 distorted images over @xmath13 distortion categories .
tid2013 @xcite is a very recently introduced image quality database with an end goal to include the peculiarities of color distortions in addition to the 17 simulated spatial distortions included in tid2008 .
it consists of 3000 images and includes seven new types of distortions , thus modeling a total of 24 distortions .
we refer the reader to @xcite for more details on the categories and severities of image distortions contained in this database .
these databases @xcite contain quality ratings obtained by conducting subjective studies in controlled laboratory settings .
the tid2008 opinion scores were obtained from @xmath14 observers by conducting batches of large scale subjective studies , whereby a total of @xmath15 comparisons of the visual quality of distorted images were performed .
although this is a large database , some of the test methodologies that were adopted do not abide by the itu recommendations .
for instance , the authors followed a _ swiss competition principle _ and presented three images , wherein two of them are the distorted versions of the third one .
a subject was asked to choose one image of superior quality amongst the two distorted images .
we believe that this kind of presentation does not accurately reflect the experience of viewing and assessing distorted images in the most common ( e.g. mobile ) viewing scenarios .
furthermore , in each experiment , a subject would view and compare @xmath16 instances of the same reference image containing multiple types and degrees of distortions , introducing the significant possibility of serious _ hysteresis effects _ that are not accounted for when processing the individual opinion scores . in pairwise comparison studies , the method for calculating preferential ranking of the data
can often dictate the reliability of the results .
certain probabilistic choice model - based ranking approaches @xcite offer sophisticated ways to accurately generate quality rankings of images .
however , the opinion scores in the tid2008 database were obtained by first accumulating the points `` won '' by each image .
these points are driven by the preferential choices of different observers during the comparative study .
the mean values of the winning points on each image were computed in the range @xmath17 $ ] and are referred to as mean opinion scores .
this method of gathering opinion scores , which diverges from accepted practice , is in our view questionable .
conversely , the live iqa database was created following an itu recommended single - stimulus methodology .
both the reference images as well as their distorted versions were evaluated by each subject during each session .
thus , _ quality difference scores _ which address user biases were derived for all the distorted images and for all the subjects .
although the live test methodology and subject rejection method adheres to the itu recommendations , the test sessions were designed to present a subject with a set of images , all afflicted by the same type of distortion ( for instance , all the images in a given session consisted of different degrees of jpeg 2000 distortion ) that were artificially added to different reference images .
we suspect that this could have led to over - learning of each distortion type by the subjects as the study session progressed .
since cameras on mobile devices make it extremely easy to snap images spontaneously under varied conditions , the complex mixtures of image distortions that occur are not well - represented by the distorted image content in either of these legacy image databases .
this greatly motivated us to acquire real images suffering the natural gamut of authentic distortion mixtures as the basis for a large new database and human study .
such a resource could prove quite valuable for the design of next - generation robust iqa prediction models that will be used to ensure future end users high quality of viewing experience .
* online subjective studies * [ sec : onlinestudies ] most subjective image quality studies have been conducted in laboratory settings with stringent controls on the experimental environment and involving small , non - representative subject samples ( typically graduate and undergraduate university students ) .
for instance , the creators of the live iqa database used two 21-inch crt monitors with display resolutions of @xmath18 pixels in a normally lit room , which the subjects viewed from a viewing distance of @xmath19 screen heights . however , the highly variable ambient conditions and the wide array of display devices on which a user might potentially view images will have a considerable influence on her perception of picture quality .
this greatly motivates our interest in conducting iqa studies on the internet , which can enable us to access a much larger and more diverse subject pool while allowing for more flexible study conditions .
however , the lack of control on the subjective study environment introduces several challenges ( more in sec .
[ crowdsourcechallenges ] ) , some of which can be handled by employing counter measures ( such as gathering details of the subject s display monitor , room illumination , and so on ) @xcite .
a few studies have recently been reported that used web - based image , video , or audio rating platforms @xcite . some of these studies employed pairwise comparisons followed by ranking techniques @xcite to derive quality scores , while others adopted the single stimulus technique and an absolute category rating ( acr ) scale .
since performing a complete set of paired comparisons ( and ranking ) is time - consuming and monetarily expensive when applied on a large scale , xu _ et al . _
@xcite introduced the hodgerank on random graphs ( hrrg ) test , where random sampling methods based on erds - rnyi random graphs were used to sample pairs and the hodgerank @xcite was used to recover the underlying quality scores from the incomplete and imbalanced set of paired comparisons .
more recently , an active sampling method @xcite was proposed that actively constructs a set of queries consisting of single and pair - wise tests based on the expected information gain provided by each test with a goal to reduce the number of tests required to achieve a target accuracy .
however , all of these studies were conducted on small sets of images taken from publicly available databases of synthetically distorted images @xcite , mostly to study the reliability and quality of the opinion scores obtained via the internet testing methodology . in most cases ,
the subjective data from these online studies is publicly unavailable . to the best of our knowledge , we are aware of only one other project @xcite reporting efforts made in the same spirit as our work , that is , _ crowdsourcing _ the image subjective study on mechanical turk by following a single - stimulus methodology . however , the authors of @xcite tested their crowdsourcing system on only @xmath20 jpeg compressed images from the legacy live image quality database of synthetically distorted images @xcite and gathered opinion scores from only forty subjects .
by contrast , the new live in the wild image quality challenge database has @xmath3 challenging images and engaged more than @xmath21 unique subjects . also , we wanted our web - based online study to be similar to the subjective studies conducted under laboratory settings with instructions , training , and test phases ( more details in sec .
[ sec : inst ] ) .
we also wanted unique participants to view and rate the images on a continuous rating scale ( as opposed to using the acr scale ) .
thus we chose to design our own crowdsourcing framework incorporating all of the above design choices , as none of the existing successful crowdsourcing frameworks @xcite seemed to offer us the flexibility and control that we desired .
in practice , every image captured by a typical mobile digital camera device passes several processing stages , each of which can introduce visual artifacts . authentically distorted images captured using modern cameras are likely to be impaired by sundry and mixed artifacts such as low - light noise and blur , motion - induced blur , over and underexposure , compression errors , and so on .
the lack of content diversity and mixtures of bonafide distortions in existing , widely - used image quality databases @xcite is a continuing barrier to the development of better iqa models and prediction algorithms of the perception of real - world image distortions . to overcome these limitations and towards creating a holistic resource for designing the next generation of robust , perceptually - aware image assessment models , we designed and created the live in the wild image quality challenge database , containing images afflicted by diverse authentic distortion mixtures on a variety of commercial devices .
figure [ fig : challengeimgs ] presents a few images from this database .
the images in the database were captured using a wide variety of mobile device cameras as shown in fig .
[ fig : devicedist ] .
the images include pictures of faces , people , animals , close - up shots , wide - angle shots , nature scenes , man - made objects , images with distinct foreground / background configurations , and images without any specific object of interest .
some images contain high luminance and/or color activity , while some are mostly smooth .
since these images are naturally distorted as opposed to being artificially distorted post - acquisition pristine reference images , they often contain mixtures of multiple distortions creating an even broader spectrum of perceivable impairments .
since the images in our database contain mixtures of unknown distortions , in addition to gathering perceptual quality opinion scores on them ( as discussed in detail in sec .
[ sec : onlinecrowdsource ] ) , we also wanted to understand to what extent the subjects could supply a sense of distortion type against a few categories of common impairments .
thus we also conducted a crowdsourcing study wherein the subjects were asked to select the single option from among a list of distortion categories that they think represented the _ most dominant distortion in each presented image_. the categories available to choose from were - `` blurry , '' `` grainy , '' `` overexposed , '' `` underexposed , '' `` no apparent distortion .
its a great image , '' and `` i do nt understand the question . ''
we adopted a majority voting policy to aggregate the distortion category labels obtained on every image from several subjects .
a few images along with the category labels gathered on them are shown in fig .
[ fig : categoryclashes ] . images presented in the left column of fig .
[ fig : categoryclashes ] were sampled from an image pool where a majority of the subjects were in full agreement with regard to their opinion of the specific distortion present in those images .
the images presented in the second column are from a pool of images that received an approximately equal number of votes for two different classes of distortions .
that is , about 50% of the subjects who viewed these images perceived one kind of dominant distortion while the remaining subjects perceived a completely different distortion to be the most dominating one .
the confusion of choosing a dominant distortion was more difficult for some images , a few of which are presented in the last column . here ,
nearly a third of the total subjects who were presented with these images labeled them as belonging to a distortion category different from the two other dominant labels obtained from the other subjects .
figure [ fig : categoryclashes ] highlights the risk of forcing a consensus on image distortion categories through majority voting on our dataset .
multiple objective viewers appeared to have different sensitivities to different types of distortions which , in combination with several other factors such as display device , viewer distance from the screen , and image content , invariably affect his / her interpretation of the underlying image distortion .
this non - negligible disagreement among human annotators sheds light on the extent of distortion variability and the difficulty of the data contained in the current database .
we hope to build on these insights to develop a holistic identifier of mixtures of authentic distortions in the near future . for
now , we take this as more direct evidence of the overall complexity of the problem . + + * no well - defined distortion categories in real - world pictures : * the above study highlights an important characteristic of real - world , authentically distorted images captured by nave users of consumer camera devices - that these pictures can not be accurately described as generally suffering from _ single _ distortions
. normally , inexpert camera users will acquire pictures under highly varied illumination conditions , with unsteady hands , and with unpredictable behavior on the part of the photographic subjects .
further , the overall distortion of an image also depends on other factors such as device and lens configurations .
furthermore , authentic mixtures of distortions are even more difficult to model when they interact , creating new agglomerated distortions not resembling any of the constituent distortions .
indeed real - world images sufer from a many - dimensional continuum of distortion perturbations . for this reason ,
it is not meaningful to attempt to segregate the images in the live in the wild image quality challenge database into discrete distortion categories .
crowdsourcing systems like amazon mechanical turk ( amt ) , crowd flower @xcite , and so on , have emerged as effective , human - powered platforms that make it feasible to gather a large number of opinions from a diverse , distributed populace over the web . on these platforms , `` requesters '' broadcast their task to a selected pool of registered `` workers '' in the form of an open call for data collection .
workers who select the task are motivated primarily by the monetary compensation offered by the requesters and also by the enjoyment they experience through participation . despite the advantages offered by crowdsourcing frameworks ,
there are a number of well - studied limitations of the same .
for example , requesters have limited control over the study setup and on factors such as the illumination of the room and the display devices being used by the workers .
since these factors could be relevant to the subjective evaluation of perceived image quality , we gathered information on these factors in a compulsory survey session presented towards the end of the study ( more details in sec .
[ sec : inst ] ) . the basic study structure and
procedures of subjective testing in a crowdsourcing framework differ from those of traditional subjective studies conducted in a laboratory .
subjective tests conducted in a lab environment typically last for many minutes with a goal of gathering ratings on every image in the dataset and are usually conducted in multiple sessions to avoid subject fatigue .
for instance , the study reported in @xcite was conducted in two sessions where each session lasted for 30 minutes .
however , crowdsourced tasks should be small enough that they can be completed by workers quickly and with ease .
it has been observed that it is difficult to find workers to participate in large and more time consuming tasks , since many workers prefer high rewards per hour @xcite .
thus , an online test needs to be partitioned into smaller chunks .
further , although requesters can control the maximum number of tasks each worker can participate in , they can not control the exact number of times a worker selects a task .
thus , it is very likely that all the images in the dataset will not be viewed and rated by every participating worker . despite these limitations imposed on any crowdsourcing framework , our online subjective study , which we describe in great detail below
has enabled us to gather a large number of highly reliable opinion scores on all the images in our dataset .
image aesthetics are closely tied to perceived quality and crowdsourcing platforms have been used in the past to study the aesthetic appeal of images @xcite .
here we have focused on gathering subjective quality scores using highly diverse aesthetic content .
we also informed users how to focus on quality and not aesthetics . in future studies , it will be of value to gather associated side information from each subject regarding the content and aesthetics of each presented image ( or video ) .
the data collection tasks on amt are packaged as hits ( human intelligence tasks ) by requesters and are presented to workers , who first visit an instructions page which explains the details of the task . if the worker understands and likes the task , she needs to click the `` accept hit '' button which then directs her to the actual task page at the end of which , she clicks a `` submit results '' button for the requester to capture the data .
crowdsourcing has been extensively and successfully used on several object identification tasks @xcite to gather segmented objects and their labels .
however , the task of labeling objects is often more clearly defined and fairly straightforward to perform , by contrast with the more subtle , challenging , and highly subjective task of gathering opinion scores on the perceived quality of images .
the generally naive level of experience of the workers with respect to understanding the concept of image quality and their geographical diversity made it important that detailed instructions be provided to assist them in understanding how to undertake the task without biasing their perceptual scores .
thus , every unique participating subject on amt that selects our hit was first provided with detailed instructions to help them assimilate the task .
a screenshot of this web page is shown in fig .
[ fig : instructions - b ] . specifically , after defining the objective of the study , a few sample images were presented which are broadly representative of the kinds of distortions contained in the database , to help draw the attention of the workers to the study and help them understand the task at hand .
a screenshot of the rating interface was also given on the instructions page , to better inform the workers of the task and to help them decide if they would like to proceed with it .
[ fig : instructions - b ] * ensuring unique participants : * after reading the instructions , if a worker accepted the task , and did so for the first time , a rating interface was displayed that contains a slider by which opinion scores could be interactively provided .
a screenshot of this interface is also shown in fig .
[ fig : instructions - a ] . in the event that this worker had already picked our task earlier , we informed the worker that we are in need of unique participants and this worker was not allowed to proceed beyond the instructions page .
only workers with a confidence value greater than 0.75 were allowed to participate .
even with such stringent subject criteria , we gathered more than 350,000 ratings overall .
* study framework : * we adopted a single stimulus continuous procedure @xcite to obtain quality ratings on images where subjects reported their quality judgments by dragging the slider located below the image on the rating interface . this continuous rating bar is divided into five equal portions , which are labeled `` bad , '' `` poor , '' `` fair , '' `` good , '' and `` excellent . '' after the subject moved the slider to rate an image and pressed the _ next image _
button , the position of the slider was converted to an integer quality score in the range @xmath22 , then the next image was presented .
before the actual study began , each participant is first presented with @xmath23 images that were selected by us as being reasonably representative of the approximate range of image qualities and distortion types that might be encountered .
we call this the * training phase*. next , in the * testing phase * , the subject is presented with @xmath24 images in a random order where the randomization is different for each subject .
this is followed by a quick survey session which involves the subject answering a few questions .
thus , each hit involves rating a total of 50 images and the subject receives a remuneration of @xmath25 cents for the task .
figure [ fig : hit ] illustrates the detailed design of our hit on iqa and fig .
[ fig : turkflow ] illustrates how we package the task of rating images as a hit and effectively disperse it online via amt to gather thousands of human opinion scores .
crowdsourcing has empowered us to efficiently collect large amounts of ratings .
however , it raises interesting issues such as dealing with noisy ratings and addressing the reliability of the amt workers . to gather high quality ratings ,
only those workers on amt with a confidence value greater than 75% were allowed to select our task .
also , in order to not bias the ratings due to a single worker picking our hit multiple times , we imposed a restriction that each worker could select our task no more than once .
5 of each group of 43 test images were randomly presented twice to each subject in the testing phase . if the difference between the two ratings that a subject provided to the same image each time it was presented exceeded a threshold on at least 3 of the 5 images , then that subject was rejected .
this served to eliminate workers that were providing unreliable , `` random '' scores .
prior to the full - fledged study , we conducted an initial subjective study and obtained ratings from 300 unique workers .
we then computed the average standard deviation of these ratings on all the images . rounding this value to
the closest integer yielded 20 which we then used as our threshold for subject rejection .
5 of the remaining 38 test images were drawn from the live multiply distorted image quality database @xcite to supply a control .
these images along with their corresponding mos from that database were treated as a _ gold standard_. the mean of the spearman s rank ordered correlation values computed between the mos obtained from the workers on the gold standard images and the corresponding ground truth mos values from the database was found to be * 0.9851*. the mean of the absolute difference between the mos values obtained from our crowdsourced study and the ground truth mos values of the gold standard images was found to be * 4.65*. furthermore , we conducted a paired - sampled t - test and observed that this difference between gold standard and crowdsourced mos values is not statistically significant .
this high degree of agreement between the scores gathered in a traditional laboratory setting and those gathered via an uncontrolled online platform with several noise parameters is critical to us .
although the uncontrolled test settings of an online subjective study could be perceived as a challenge to the authenticity of the obtained opinion scores , this high correlation value indicates a high degree of reliability of the scores that are being collected by us using amt , reaffirming the efficacy of our approach of gathering opinion scores and the high quality of the obtained subject data .
in addition to measuring correlations against the gold standard image data as discussed above , we further analyzed the subjective scores in the following two ways : to evaluate subject consistency , we split the ratings obtained on an image into two disjoint equal sets , and computed two mos values on every image , one from each set . when repeated over 25 random splits , an average spearman s rank ordered correlation between the mean opinion scores between the two sets was found to be * 0.9896*. evaluating intra - subject reliability is a way to understand the degree of consistency of the ratings provided by individual subjects @xcite .
we thus measured the spearman s rank ordered correlation ( srocc ) between the individual opinion scores and the mos values of the gold standard images .
a median srocc of * 0.8721 * was obtained over all of the subjects .
all of these additional experiments further highlight the high degree of reliability and consistency of the gathered subjective scores and of our test framework .
the database currently comprises of more than @xmath4 ratings obtained from more than @xmath5 unique subjects ( after rejecting unreliable subjects ) .
enforcing the aforementioned rejection strategies led us to reject 134 participants who had accepted our hit .
each image was viewed and rated by an average of @xmath26 unique subjects , while the minimum and maximum number of ratings obtained per image were @xmath27 and @xmath28 , respectively . while computing these statistics , we excluded the 7 images used in the training phase and the 5 gold standard images as they were viewed and rated by all of the participating subjects
. workers took a median duration of @xmath29 minutes to view and rate all 50 images presented to them .
the mean opinion scores ( mos ) after subject rejection was computed for each image by averaging the individual opinion scores from multiple workers .
mos is representative of the _ perceived viewing experience _ of each image .
the mos values range between @xmath30 $ ] .
figure [ fig : mos ] is a scatter plot of the mos computed from the individual scores we have collected . in order to compare the mos values with single opinion scores ( sos ) , we computed the standard deviation of the subjective scores obtained on every image and obtained an average standard deviation of @xmath31 .
the uncontrolled online test environment poses certain unique challenges : a test subject of any gender or age may be viewing the image content on any kind of a display , under any sort of lighting , from an unknown distance , and an unknown level of concentration , each of which can affect her choice of quality score
. figures [ fig : demographics ] ( a ) and [ fig : demographics ] ( b ) illustrate the demographic details of the unique subjects who have participated in our study .
most of them reported in the final survey that they are inexperienced with image quality assessment but do get annoyed by image impairments they come across on the internet .
since we did not test the subjects for vision problems , they were instructed to wear corrective lenses during the study if they do so in their day - to - day life .
later in the survey , the subjects were asked if they usually wore corrective lenses and whether they wore the lenses while participating in the study .
the ratings given by those subjects who were not wearing their corrective lenses they were otherwise supposed to wear were rejected .
figures [ fig : demographics ] ( c ) and [ fig : demographics ] ( d ) illustrate the distribution of the distances from which workers have viewed the images and the broad classes of different display devices used by them .
these four plots illustrate the highly varied testing conditions that exist during the online study and also highlight the diversity of the subjects .
figure [ fig : influences1 ] ( a ) illustrates the distribution of the types of consumer image capture devices that are preferred by the users .
it is evident from this plot that most of the workers reported that they prefer using mobile devices to capture photographs in their daily use .
one of the questions we posed to our subjects in the survey was whether the poor quality of pictures that they encounter on the internet bothers them .
subjects chose between the following four options - `` yes , '' `` no , '' `` i do nt really care , '' and `` i do nt know . ''
the distribution of the responses to this question is plotted in fig.[fig : influences1 ] ( b ) which clearly indicates that a large population of the workers are bothered by poor quality internet pictures .
we next present our analysis of the influence of several factors such as age , gender , and display devices on user s perceptual quality . in all cases , we study the effect of each factor independently while fixing the values of the rest of the factors .
we believe this strategy helped us to closely study the influence of each factor independently and to help avoid combined effects caused by the interplay of several factors on a user s perceptual quality .
note that the results presented in the following sections are consistent irrespective of the specific values that were fixed for the factors .
@xmath32{preferreddevices - eps - converted - to.pdf } & \includegraphics[width=1.6in , height=1.9in]{quality - eps - converted - to.pdf } \end{array}$ ] @xmath33{gender - eps - converted - to.pdf } & \includegraphics[width=1.8in , height=2.0in]{age - eps - converted - to.pdf } & \includegraphics[width=1.8in , height=2.0in]{distranges - eps - converted - to.pdf } & \includegraphics[width=1.8in , height=1.99in]{devices - eps - converted - to.pdf } \end{array}$ ] @xmath34{img1-eps - converted - to.pdf } & \includegraphics[width=1.3in]{img2-eps - converted - to.pdf } & \includegraphics[width=1.3in]{img3-eps - converted - to.pdf } & \includegraphics[width=1.3in]{img4-eps - converted - to.pdf } & \includegraphics[width=1.3in]{img5-eps - converted - to.pdf } \end{array}$ ] to understand to what extent gender had an affect on our quality scores , we separately analyzed the ratings obtained from male and female workers on five randomly chosen images ( figures [ fig : infimages](a)-(e ) ) while maintaining all the other factors constant .
specifically , we separately captured the opinion scores of male and female subjects who are between @xmath35 years old , and reported in our survey to be using a desktop and sitting about @xmath36 inches from the screen . under this setting and on the chosen set of images , both male and female workers appeared to have rated the images in a similar manner .
this is illustrated in figure [ fig : influences](a ) .
next , we considered both male and female workers who reported using a laptop during the study and were sitting about @xmath36 inches away from their display screen .
we grouped their individual ratings on these 5 images ( fig .
[ fig : infimages ] ) according to their age and computed the mos of each group and plotted them in fig [ fig : influences](b ) . for the images under consideration ,
again , subjects belonging to different _ age categories _ appeared to have rated them in a similar manner .
@xmath37{genderinf_c - eps - converted - to.pdf } & \includegraphics[width=2.1in , height=2.3in]{ageinf_c - eps - converted - to.pdf } \\
\includegraphics[width=2.2in , height=2.3in]{distinf_c - eps - converted - to.pdf } & \includegraphics[width=2.2in , height=2.3in]{displayinf_c - eps - converted - to.pdf } \end{array}$ ] although gender and age did not seem to significantly affect the ratings gathered on the randomly chosen images discussed above , we believe that other factors such as the content in the image can play a significant role in being appealing to one group more than to another .
a systematic study focussed exclusively on understanding the interplay of image content , gender , and age using this database might help us better understand the impact of each of these factors on perceptual quality .
we next explored the influence of the distance between a subject and her monitor , on the perception of quality .
one of the questions in the survey asked the subjects to report which of the three _ distance categories _ best described a subject s location relative to the viewing screen - `` less than 15 inches , '' `` between 15 to 30 inches , '' and `` greater than 30 inches . ''
we gathered the ratings of subjects who reported to be between @xmath38 years old and were participating in the study using their desktop computer .
we grouped their ratings on the five test images ( fig .
[ fig : infimages ] ) according to these distance categories and report the results in fig .
[ fig : influences](c )
. it may be noticed that the difference between the mean of the ratings obtained on the same image when viewed from a closer distance as compared to when the same image was viewed by subjects from a greater distance is not statistically significant .
however , we do not rule out the possible influences that viewing distance may have on distortion perception from an analysis of 5 random images .
the observed indifference to viewing distance could be due to an interplay of the resolution of the display devices , image content , and viewing distances which is a broad topic worthy of future study . to better understand the influence of display devices on qoe , we focussed on workers between @xmath35 years old and who reported to be @xmath36 inches away from the screen while participating in the study .
we grouped the ratings of these subjects on the five images in fig .
[ fig : infimages ] according to the display device that the subjects reported to have used while participating in the study . as illustrated in fig .
[ fig : influences](d ) , the influence of the specific display device that was used for the study appears to have had little effect on the recorded subjective ratings .
of course , we are not suggesting that the perceptual quality of images is unaffected by the display devices on which they are viewed .
it is possible that more fine - grained detail regarding the type of display device used by the study participants ( e.g. , screen resolution , display technology involved , shape of the screen etc . )
could deepen our understanding of the dependency between display device and perceptual image quality .
however , we chose to focus as much of each participants effort on the visual tasks as reasonable , and so did not poll them on these details , leaving it for future studies . i as mentioned earlier , one of the questions posed to the subjects in the survey was whether the quality of pictures they encounter on the internet bothers them ( distribution of the responses in fig .
[ fig : influences1 ] ( b ) ) .
when we grouped our ratings according to these three answers , we noticed that the subjects from each of these three response categories were almost equally sensitive to the visual distortions present in the images from our dataset .
this is illustrated in fig .
[ fig : qualitybothersinf ] ( b ) .
* observation * + & age : 20 - 30 years , display device : desktop , distance from the screen : 15 - 30 inches & both male and female workers appeared to rate images in a similar manner .
+ & gender : both male and female workers , display device : desktop , distance from the screen : 15 - 30 inches & very little difference was noticed in the ratings of the subjects of different age groups .
+ & gender : both male and female workers , age : 30 - 40 years , display device : desktop & little effect on the subjective ratings .
+ & gender : both male and female workers , age : 20 - 30 years , distance from the screen : 15 - 30 inches . & little effect on subjective ratings .
+ & none & people who claimed to differ in their level of annoyance in response to image distortions appeared to rate images in a similar manner .
+ figure [ fig : numberofworkersinf ] illustrates how mos values flatten out with increases in the number of subjects rating the images .
it is interesting to note that there is much more consistency on images with very high and very low mos values than on intermediate - quality images .
of course , the opinion scores of subjects are affected by several external factors such as the order in which images are presented , a subject s viewing conditions , and so on , and the mos thus exhibit variability with respect to the number of workers who have rated them .
we summarize all the factors whose influence we studied and presented in this section ( by controlling the other factors ) in table [ table : infsummary ] .
the crowdsourcing image quality study allowed our diverse subjects to participate at their convenience , and in diverse , uncontrolled viewing circumstances , enhancing our ability to investigate the effects of each of these factors on perceived picture quality .
the results of our studies of those factors affecting data reliability , and our observations of the high correlations of the objective quality scores against the mos values of the gold standard images that were obtained under controlled laboratory conditions , both strongly support the efficacy of our online crowdsourcing system for gathering large scale , reliable data .
crowdsourcing is a relatively new tool with considerable potential to help in the production of highly valuable and generalized subjective databases representative of human judgments of perceptual quality .
however , the approach involves many complexities and potential pitfalls which could affect the veracity of the subject results .
a good summary and analysis of these concerns may be found in @xcite .
for example , while we have a high degree of faith in our subject results , it is based on a deep analysis of them rather than simply because the participants were screened to have high amt confidence values . as mentioned earlier , the confidence values of the workers computed by amt is an aggregate that is measured over all the tasks in which a worker has participated .
this metric thus is not necessarily an indicator of reliability with regards to any specific task and should be accompanied by rigorous , task - specific subject reliability methods .
future studies would benefit by a more detailed data collection and analysis of the details of workers display devices @xcite and viewing conditions .
while our current philosophy , even in laboratory studies , is to not screen the subjects for visual problems , given the newness of the crowdsourcing modality , it might be argued that visual tests could be used to improve subject reliability checks .
many other environmental details could be useful , such as reports of the time spent by a worker in viewing and rating images , to further measure worker reliability .
we also explored the usefulness of the new database by using it to evaluate the quality prediction performance of a number of leading blind iqa algorithms .
these algorithms are almost invariably machine learning - based training procedures applied on perceptual and/or statistical image features .
therefore , in all of the experiments described below , we randomly divided the data , @xmath39 images that were captured during the night were excluded from the dataset .
thus , the total number of images for this experiment consisted of only @xmath40 images . in sec .
[ sec : nightimages ] , the experiments were repeated including these images . ] into content - separated disjoint 80% training and 20% testing sets , learned a model on the training data , and validated its performance on the test data . to mitigate any bias due to the division of data
, we repeated the process of randomly splitting the dataset over 50 iterations and computed spearman s rank ordered correlation coefficient ( srocc ) and pearson s linear correlation coefficient ( plcc ) between the predicted and the ground truth quality scores at the end of every iteration .
we report the median of these correlations across 50 iterations . a higher value of each of these metrics indicates better performance both in terms of correlation with human opinion as well as the performance of the model .
we recently proposed a novel blind image quality assessment model @xcite , dubbed * * f**eature maps based * * r**eferenceless * * i**mage * * qu**ality * * e**valuation * * e**ngine ( friquee ) , that seeks to overcome the limitations of existing blind iqa techniques with regards to mixtures of authentic picture distortions , such as those contained in the live in the wild image quality challenge database .
friquee is a natural scene statistics ( nss ) based model that is founded on the hypothesis that different existing statistical image models capture distinctive aspects of the loss of the perceived quality of a given image .
friquee embodies a total of @xmath41 statistical features that have been observed to contribute significant information regarding image distortion visibility and perceived image quality .
we designed a model that combines a deep belief net and an svm and achieves superior quality prediction performance when compared to the state - of - the - art .
our proposed deep belief net ( dbn ) @xcite has four hidden layers formed by stacking multiple restricted boltzmann machines ( rbm ) and by learning weight matrices at every level by treating the hidden layer s activities of one rbm as the visible input data for training a higher level rbm in a greedy layer - by - layer manner @xcite .
our dbn builds more complex representations of the simple statistical friquee features provided as input and remarkably generalizes over different distortion types , mixtures , and severities .
the _ deep feature representations _ learned from our dbn together with subjective opinion scores are later used to train a regressor such that , given a unique test image , its quality is accurately predicted .
we extracted the quality - predictive features proposed by several prominent blind iqa algorithms ( whose code was publicly available ) on the images of the live in the wild image quality challenge database and used the same learning model that was originally presented in their work under the 80 - 20 train - test setting . in this experiment , we excluded @xmath39 images that were captured during the night .
we studied the influence of including night images in the data in sec .
[ sec : nightimages ] . in the case of friquee ,
a combination of dbn and svm was used and the dbn s input layer has @xmath41 units , which is equal to the number of features extracted from an image .
we report the results in table [ dbn - perf ] from which we conclude that the performance of our proposed model on unseen test data is significantly better than current top - performing state - of - the - art methods when tested on the live in the wild image quality challenge database .
brisque @xcite & 0.61 & 0.60 diivine @xcite & 0.56 & 0.51 bliinds - ii @xcite & 0.45 & 0.40 niqe @xcite & 0.48 & 0.42 s3 index @xcite & 0.32 & 0.31c - diivine @xcite & 0.66 & 0.63 [ dbn - perf ] to further highlight the challenges that the authentic distortions present in our database pose to top - performing algorithms , we also computed the median correlation values when the algorithms were tested on the standard legacy benchmark database @xcite .
friquee was implemented using the dbn model , by extracting @xmath41 features on all the images of the dataset and repeating the same evaluation procedure over @xmath42 iterations . for the other blind iqa models , we use the same learner that was presented originally in their work .
we present the median correlation values obtained in table [ legacy - results ] .
it may be observed that all of the top - performing models , when trained and tested on the legacy live iqa database which is comprised of singly distorted images , perform remarkably well as compared to their performance on our difficult database of images suffering from unknown mixtures of distortions ( table [ dbn - perf ] ) .
& 0.86 & 0.86ssim @xcite & 0.91 & 0.91 ms - ssim @xcite & 0.95 & 0.95 _ cbiq _ @xcite & _ 0.89 _ & _ 0.90 _ _ lbiq _ @xcite & _ 0.91 _ & _ 0.91diivine _ @xcite & _ 0.92 _ & _ 0.93 _ _ bliinds - ii _
@xcite & _ 0.91 _ & _ 0.92 _ _ brisque _ @xcite & 0.94 & _ 0.94 _ _ niqe _ @xcite & 0.91 & 0.92 _ c - diivine _
@xcite & 0.94 & 0.95 _ tang et.als model _
@xcite & _ * 0.96 * _ & _ -na- _ _ friquee - all _ & _ 0.93 _ & _ 0.95 _ [ legacy - results ] friquee & * 0.71 * & * 0.70 * brisque & 0.60 & 0.60 diivine & 0.60 & 0.59 bliinds - ii & 0.44 & 0.46 niqe & 0.47 & 0.45 c - diivine @xcite & 0.65 & 0.64 [ dbn - night ] @xmath43{night1-eps - converted - to.pdf } & \includegraphics[width=0.7in]{night2-eps - converted - to.pdf } & \includegraphics[width=0.7in]{night3-eps - converted - to.pdf } & \includegraphics[width=0.7in]{night4-eps - converted - to.pdf } \\ \end{array}$ ] of the total of @xmath44 images , @xmath39 pictures were captured at night and suffer from severe low - light distortions ( fig .
[ fig : nightimgs ] ) .
it should be noted that none of the legacy benchmark databases have images captured under such low illumination conditions and it follows that the nss - based features used in other models were created by training natural images under normal lighting conditions . here , we probe the predictive capabilities of top - performing blind iqa models when such low - light images are included in the training and testing data .
we therefore included the night - time pictures into the data pool and trained friquee and the other blind iqa models .
the results are given in table [ dbn - night ] . despite such challenging image content
, friquee still performed well in comparison with the other state - of - the - art models .
this further supports the idea that a generalizable blind iqa model should be trained over mixtures of complex distortions , and under different lighting conditions .
@xcite+ live challenge & 0.82 & 0.78 live challenge + live multiply @xcite & 0.80 & 0.79 live challenge + live multiply @xcite + live iqa @xcite & 0.85 & 0.83 [ tbl : combine - perf ] since nr iqa algorithms are generally trained and tested on various splits of a single dataset ( as described above ) , it is natural to wonder if the trained set of parameters are database specific . in order to demonstrate that the training process is simply a calibration , and
that once such training is performed , an ideal blind iqa model should be able to assess the quality of any distorted image ( from the set of distortions it is trained for ) , we evaluated the performance of the multi - model friquee algorithm on combinations of different image databases - the live iqa database @xcite and the live multiply distorted iqa database @xcite , as well as the new live in the wild image quality challenge database .
the same 80 - 20 training setup was followed , i.e. , after combining images from the different databases , 80% of the randomly chosen images were used to train our dbn model and the trained model was then tested on the remaining 20% of the image data .
we present the results in table [ tbl : combine - perf ] .
it is clear from table [ tbl : combine - perf ] that the performance of friquee is _ not _ database dependent and that once trained , it is capable of accurately assessing the quality of images across the distortions ( both single and multiple , of different severities ) that it is trained for .
the results clearly show friquee s potential to tackle the imminent deluge of visual data and the unavoidable distortions they are bound to contain .
with more than @xmath4 subjective judgments overall , we believe that the study described here is the largest , most comprehensive study of perceptual image quality ever conducted . of course , digital videos ( moving pictures ) are also being captured with increasing frequency by both professional and casual users . in the increasingly mobile environment
, these spatial - temporal signals will be subject to an even larger variety of distortions @xcite arising from a multiplicity of natural and artifical processes @xcite .
predicting , monitoring , and controlling the perceptual effects of these distortions will require the development of powerful blind video quality assessment models , such as @xcite , and new vqa databases representative of human opinions of modern , realistic videos captured by current mobile video camera devices and exhibiting contemporary distortions .
current legacy vqa databases , such as @xcite are useful tools but are limited in regard to content diversity , numbers of subjects , and distortion realism and variability . therefore , we plan to conduct large - scale crowdsourced _
video _ quality studies in the future , mirroring the effort described here , and building on our expertise in conducting the current study .
we acknowledge prof .
sanghoon lee , dr . anish mittal , dr .
rajiv soundararajan , numerous unnamed photographers from ut austin and yonsei university , among others , for helping to collect the original images in the live in the wild image quality challenge database .
this work was supported in part by the national science foundation under grant iis-1116656 .
m. saad , a. c. bovik , and c. charrier , `` blind image quality assessment : a natural scene statistics approach in the dct domain , '' _ ieee trans . image process .
21 , no . 8 , pp . 3339 - 3352 , aug .
2012 .
y. zhang , a. k. moorthy , d. m. chandler , and a. c. bovik , `` c - diivine : no - reference image quality assessment based on local magnitude and phase statistics of natural scenes , '' _ sig .
image commun _ , vol .
29 , issue 7 , pp .
725 - 747 , august . 2014 .
sheikh , m.f .
sabir , and a.c .
bovik , `` a statistical evaluation of recent full reference image quality assessment algorithms , '' _ ieee trans . image process .
3440 - 3451 , nov 2006 .
n. ponomarenko , v. lukin , a. zelensky , k. egiazarian , m. carli , and f. battisti , `` tid2008-a database for evaluation of full - reference visual quality assessment metrics , '' _ adv of modern radio electron._,vol .
4 , pp . 30 - 45 , 2009 .
n. ponomarenko , o. ieremeiev , v. lukin , k. egiazarian , l. jin , j. astola , b. vozel , k. chehdi , m. carli , f. battisti , and c .- c .
jay kuo , `` color image database tid2013 : peculiarities and preliminary results , '' _ proc . of 4th european workshop on visual info .
_ , pp . 106 - 111 , june 2013 .
t. hofield , p. korshunov , f. mazza , i. povoa , and c. keimel , `` crowdsourcing - based multimedia subjective evaluations : a case study on image recognizability and aesthetic appeal , '' _ int .
workshop on crowdsourcing for multimedia _ , pp .
29 - 34 , oct .
2013 t. hofield , c. keimel , m. hirth , b. gardlo , j. habigt , k. diepold , and p. tran - gia , `` best practices for qoe crowdtesting : qoe assessment with crowdsourcing , '' _ ieee trans .
multimedia _ , 16(2 ) , 541 - 558 , 2014 .
j. m. foley and g. m. boynton,``a new model of human luminance pattern vision mechanisms : analysis of the effects of pattern orientation , spatial phase , and temporal frequency , '' _ in spie proceedings _ , vol . 2054 , 1993 .
z. wang , a. c. bovik , h. r. sheikh , and e. p. simoncelli , `` image quality assessment : from error measurement to structural similarity , '' _ ieee trans . on image proc .
_ , vol . 13 , no .
600 - 612 , april 2004 .
k. seshadrinathan , r. soundararajan , a.c .
bovik , and l.k .
cormack , `` study of subjective and objective quality assessment of video , '' _ ieee trans .
image process .
, _ vol . 19 , no . 6 , pp . 1427 - 1441 , june 2010
moorthy , k. seshadrinathan , r. soundararajan , and a.c .
bovik , `` wireless video quality assessment : a study of subjective scores and objective algorithms , '' _ ieee trans .
video technol .
587 - 599 , april 2010 .
deepti ghadiyaram received the b.tech .
degree in computer science from international institute of information technology ( iiit ) , hyderabad in 2009 , and the m.s .
degree from the university of texas at austin in 2013 .
she joined the laboratory for image and video engineering ( live ) in january 2013 , and is currently pursuing her ph.d .
she is the recipient of the microelectronics and computer development ( mcd ) fellowship , 2013 - 2014 .
her research interests include image and video processing , computer vision , and machine learning , and their applications to the aspects of information retrieval such as search and storage .
alan c. bovik holds the cockrell family endowed regents chair in engineering at the university of texas at austin , where he is director of the laboratory for image and video engineering ( live ) in the department of electrical and computer engineering and the institute for neuroscience .
his research interests include image and video processing , digital television and digital cinema , computational vision , and visual perception .
he has published over 750 technical articles in these areas and holds several u.s .
his publications have been cited more than 45,000 times in the literature , his current h - index is about 75 , and he is listed as a highly - cited researcher by thompson reuters .
his several books include the companion volumes _ the essential guides to image and video processing _
( academic press , 2009 ) .
bovik received a primetime emmy award for outstanding achievement in engineering development from the the television academy in october 2015 , for his work on the development of video quality prediction models which have become standard tools in broadcast and post - production houses throughout the television industry .
he has also received a number of major awards from the ieee signal processing society , including : the society award ( 2013 ) ; the technical achievement award ( 2005 ) ; the best paper award ( 2009 ) ; the signal processing magazine best paper award ( 2013 ) ; the education award ( 2007 ) ; the meritorious service award ( 1998 ) and ( co - author ) the young author best paper award ( 2013 ) .
he also was named recipient of the honorary member award of the society for imaging science and technology for 2013 , received the spie technology achievement award for 2012 , and was the is&t / spie imaging scientist of the year for 2011 .
he is also a recipient of the joe j. king professional engineering achievement award ( 2015 ) and the hocott award for distinguished engineering research ( 2008 ) , both from the cockrell school of engineering at the university of texas at austin , the distinguished alumni award from the university of illinois at champaign - urbana ( 2008 ) .
he is a fellow of the ieee , the optical society of america ( osa ) , and the society of photo - optical and instrumentation engineers ( spie ) , and a member of the television academy , the national academy of television arts and sciences ( natas ) and the royal society of photography .
professor bovik also co - founded and was the longest - serving editor - in - chief of the _ ieee transactions on image processing _ ( 1996 - 2002 ) , and created and served as the first general chair of the _ ieee international conference on image processing _ , held in austin , texas , in november , 1994 .
his many other professional society activities include : board of governors , ieee signal processing society , 1996 - 1998 ; editorial board , _ the proceedings of the ieee _ , 1998 - 2004 ; and series editor for image , video , and multimedia processing , morgan and claypool publishing company , 2003-present .
his was also the general chair of the 2014 texas wireless symposium , held in austin in november of 2014 . | most publicly available image quality databases have been created under highly controlled conditions by introducing graded simulated distortions onto high - quality photographs . however , images captured using typical real - world mobile camera devices are usually afflicted by complex mixtures of multiple distortions , which are not necessarily well - modeled by the synthetic distortions found in existing databases . the originators of existing legacy databases usually conducted human psychometric studies to obtain statistically meaningful sets of human opinion scores on images in a stringently controlled visual environment , resulting in small data collections relative to other kinds of image analysis databases .
towards overcoming these limitations , we designed and created a new database that we call the _ live in the wild image quality challenge database _ , which contains widely diverse authentic image distortions on a large number of images captured using a representative variety of modern mobile devices .
we also designed and implemented a new online crowdsourcing system , which we have used to conduct a very large - scale , multi - month image quality assessment subjective study .
our database consists of over 350,000 opinion scores on 1,162 images evaluated by over 8100 unique human observers . despite the lack of control over the experimental environments of the numerous study participants , we demonstrate excellent internal consistency of the subjective dataset .
we also evaluate several top - performing blind iqa algorithms on it and present insights on how mixtures of distortions challenge both end users as well as automatic perceptual quality prediction models .
the new database is available for public use at https://www.cs.utexas.edu/~deepti/challengedb.zip .
ghadiyaram : characterizing the perception quality of real distorted images using natural scene statistics and deep belief nets .
perceptual image quality , subjective image quality assessment , crowdsourcing , authentic distortions . |
in the 1960s-1980 s high quality secondary beams became available at proton synchrotrons of increasing energies ( ps , ags , ihep , fnal , sps ) .
the secondary charged hadron beams contained the six stable or quasi stable charged particles @xmath0 , k@xmath1 , @xmath2 , @xmath3 .
the main experimental lines of research were : i ) total hadron
hadron ( @xmath4 ) cross section measurements at high energies ( @xmath5 gev / c ) ; ii ) elastic @xmath4 scattering measurements ; iii ) measurements of absorption cross sections of hadrons in nuclei ( as byproducts of @xmath4 total cross section measurements ) ; iv ) hadron production at forward angles in @xmath6 collisions , and v ) inelastic @xmath4 collisions [ 1 - 5 ]
. in some high intensity beams many particle searches were made , which also lead to the study of @xmath7 and @xmath8 production [ 6 ] . in order to reach higher energies it became necessary to build hadron colliders .
the first was the isr @xmath9 and @xmath10 collider at cern , which allowed to reach c.m .
energies between 22 and 63 gev ; then the @xmath11 collider at cern allowed @xmath10 collisions from 600 to 900 gev ; finally the fermilab tevatron collider allowed @xmath12 collisions up to 1.8 tev .
the first experiments performed at the isr were relatively simple dedicated experiments , like those for total cross section and elastic scattering measurements , and single arm spectrometers for the study of inelastic collisions [ 7 - 8 ] . then followed general purpose detectors : the sfm at the isr [ 9 ] , the general purpose detectors ua1 and ua2 and experiment ua4 at the @xmath11 collider , and finally the general purpose detectors cdf , d0 and the specialized experiments e710 , f8 at fermilab [ 10 ] .
the highest energies were and are still obtainable only with cosmic rays .
we shall discuss total cross sections at high energies , the high energy low@xmath13 parameters and some features of inelastic collisions [ 11 , 12 ] .
at fixed target accelerators ( ags , ihep , fnal ) the total cross sections were measured with the transmission method , with relative precisions smaller than @xmath14 and systematic scale errors of about @xmath15 .
the measurements of the total cross sections at the @xmath9 and @xmath16 colliders required the development of new experimental techniques : the scattering of particles was measured at very small angles , with detectors positioned in reentrant containers ( roman pots " ) located very close to the circulating beams .
the combination of statistical and systematic uncertainties are @xmath17 .
fig 1 summarizes the present status of high energy total cross section measurements : for @xmath18 gev all total @xmath4 cross sections decrease , reach a minimum and then increase with increasing energy ( the @xmath19 total cross section was already increasing at serpukhov energies [ 1 ] ) .
moreover the differences between the cross sections @xmath20 and @xmath21 decrease with increasing energy [ 5 ] .
there is not a unique interpretation for this rise , though in many qcd inspired models it seems to be connected with the increase of the number of minijets and thus to semi - hard gluon interactions . most of the high energy elastic and total cross section data have been usually interpreted in terms of regge poles , and thus in terms of pomeron exchange .
even if the pomeron was introduced long time ago we do not have a consensus on its exact definition and on its detailed substructure .
some authors view it as a `` gluon ladder '' .
future experiments on hadron - hadron total cross sections remain centered at the fermilab collider ( for @xmath12 ) .
the near future will rely on the bnl - rich collider , and later , on the lhc proton - proton and heavy ion collider at cern .
large area cosmic ray experiments may be able to improve the data in the ultra high energy region [ 12 ] .
the differential cross section for the elastic scattering of unpolarized particles from unpolarized targets has a simple structure in the high energy region .
it depends on two variables : the energy and an angular variable , which is usually chosen to be the square of the four momentum transfer @xmath22 .
the energy dependence is of the @xmath23 type .
the angular distribution may be divided in four regions : i ) the coulomb region for @xmath24 ( gev / c)@xmath25 ; ii ) the coulomb - nuclear interference region for @xmath26 ( gev / c)@xmath25 ; measurements in this region yield information on the ratio @xmath27 of the real to the imaginary part of the forward scattering amplitude .
iii ) the nuclear diffraction region proper for @xmath28 ( gev / c)@xmath25 ; here the most important parameter is the slope b , or the slopes @xmath29 , of the diffraction pattern .
iv ) the large angle region for @xmath30 ( gev / c)@xmath25 : for @xmath31 gev it is characterized by a dip - bump structure which resembles that from diffraction from an opaque disc .
most of these features are observed in the @xmath9 and @xmath16 elastic scattering angular distributions at @xmath32 , 546 and 1400 gev , see fig .
one usually defines the following high energy scattering parameters : i ) the total cross section , @xmath33 ; ii ) the slope @xmath34 of the differential elastic nuclear cross section @xmath35 , at @xmath36 ( gev / c)@xmath25 ; iii ) the ratio @xmath37 ; iv ) the opacity @xmath38 . at very high energies
@xmath39 , @xmath34 and @xmath40 rise with energy , while @xmath27 is slightly positive . from the behaviour of these parameters emerges the picture of a proton which becomes bigger ( larger @xmath34 ) and more opaque ( larger @xmath40 ) as the energy increases . at the highest energy @xmath41 , which is still smaller than the black disk value of 1 . +
3 shows pictorially various inelastic processes at low @xmath42 : single and double diffraction dissociation and inelastic processes ; these are the dominant processes with the largest cross sections , concentrated in particular in the very forward angular region : this is the region which is most important for cosmic ray experiments . at high energies
the total @xmath16 cross section may be written as @xmath43 where @xmath44 is the elastic cross section , @xmath45 is the single diffractive cross section when the incoming proton fragments into a number of particles , @xmath46 is the single diffractive cross section for the fragmentation of the antiproton ( at high energies @xmath47 ) .
@xmath48 is the double diffractive cross section , @xmath49 is the non - diffractive part of the inelastic cross section ( fig .
the elastic , single diffractive and double diffractive processes give rise to low multiplicity events with particles emitted in the very forward region in the c.m .
the non - diffractive cross section is the main part of the inelastic cross section ; it gives rise to high multipliciy events and to particles emitted at all angles .
most of the non - diffractive cross section concerns particles emitted with low transverse momentum ( _ low @xmath50 physics _ ) with properties which change slowly with c.m .
energy ( _ @xmath51 physics)_. a small part of the non - diffractive cross section is due to central collisions among the colliding particles and gives rise to high @xmath50 jets of particles emitted at relatively large angles ( _ large @xmath50 physics)_. the contribution of jet physics increases with c.m . energy . in fig .
4a are shown , vs @xmath52 , the average charged multiplicities in @xmath9 and @xmath16 collisions . the data may be fitted to a power law of @xmath23 : @xmath53 at @xmath54 tev are produced on average 40 charged and 20 neutral particles . fig .
4b shows the average number of @xmath0 , @xmath55 , @xmath2 and @xmath3 produced in @xmath9 collisions up to @xmath52 @xmath56 gev .
pions are the dominantly produced hadrons .
hadrons are mainly produced at low transverse momenta .
the average @xmath50 increases slowly with @xmath52 : it is @xmath57 gev / c in @xmath58 gev ; then it increases slowly up to @xmath59 gev / c at @xmath54 tev .
the simplest interpretation of these features is in terms of thermodynamic models : hadrons are emitted from a region with a temperature @xmath60 mev .
qcd , at present can not be used to calculate the data at low @xmath50 , because the strong coupling constant is large and perturbative methods can not be used .
therefore one has to use models .
hard collisions can instead be calculated by perturbative qcd .
usually , as byproducts of @xmath4 total cross section measurements , the absorption cross sections of @xmath0 , @xmath55 , @xmath2 and @xmath3 on various nuclei ( li , c , al , cu , sn , and pb ) were measured at incident lab momenta up to 280 gev / c [ 2 , 4 , 11 ] .
most absorption cross sections decrease slowly for p@xmath61 up to @xmath62 50 gev / c ; for higher p@xmath61 they are almost constant .
the data at each energy were fitted to the simple expression @xmath63 where @xmath64 is the atomic weight of the target nucleus .
examples are shown in fig .
5a at three lab momenta for three different incoming hadrons .
for all incident particles except antiprotons , the value of @xmath65 increases by up to 10% as the incident p@xmath61 increases from 60 to 280 gev / c , with the largest increase for @xmath66 . for @xmath3
, @xmath65 decreases with increasing p@xmath61 .
5b shows the parameters @xmath65 and @xmath67 vs the corresponding @xmath68 total cross section : @xmath65 rises monotonically with @xmath69 ; the values of @xmath67 are consistent with @xmath67 approaching 0.67 for large values of @xmath69 as would be expected for an opaque nucleus .
the fragmentation cross sections of various nuclei were measured at many energies using nuclear track detectors , which yield high resolution measurements of the restricted energy loss .
this arises from the relatively low energy required to break the polymeric bonds of the detectors and because fluctuations due to energetic @xmath70-electrons do not contribute to the latent track formation .
the radiation damage along the path of an incoming nucleus may be developed to microscope - visible cones by chemical etching .
6a shows the charge distribution obtained with 200 gev / nucleon s@xmath71 ions and their fragments produced in a cu target [ 13 ] : notice the good charge resolution and the absence of nuclei with fractional charges .
the cross sections in fig .
6b are relative to fragments of 158 gev / nucleon pb@xmath72 ions in a target with @xmath73 with a variation of atomic number @xmath74 with respect to the @xmath75 of the incoming ions .
a wealth of experimental information has been obtained since the early 1960 s on high - energy hadron - hadron and hadron - nucleus cross sections , starting with simple equipments and proceeding to more complex apparata and better beams .
the higher energy data coming from collider experiments have large systematic uncertainties , mainly connected to the poor knowledge of the collider luminosity , the difficulty of measuring with some precision the single and double diffraction cross sections , etc .
most of the information on hadron - hadron collisions was interpreted in terms of phenomenological models .
much more work is needed to interpret and systemize the available data , since only the small part of the cross section corresponding to large transverse momenta may be interpreted in terms of qcd .
few and not too precise data are available at very low @xmath42 and very small angles , which is the region where more interest lies for cosmic rays .
the information is here codified in monte carlos , which are becoming progressively more complicated .
the monte carlos include hadron - nucleus collisions [ for which the data are much less abundant than for @xmath4 collisions ] ; hadron - nucleus effects are mainly interpreted in the context of the gluber model [ 14 ] .
nucleus - nucleus data are even less abundant , and the monte carlos have extra complications coming from the used models [ 14 ] . in the near future
the rich collider at bnl should provide data on nucleus - nucleus collisions . in the future the different experiments at the lhc collider should provide data on @xmath4 , hadron - nuclei and nuclei - nuclei collisions at much higher energies .
we thank many colleagues for their collaboration , for many discussions and for explanations of cosmic ray monte carlos .
we thank ms l. de angelis for typing the manuscript .
j. v. allaby et al . , phys . lett .
30b ( 1969 ) 500 ; s. p.
denisov et al . phys .
lett . 36b ( 1971 ) 415 ; yad fis 14 ( 1971 ) 998 . 2 .
f. binon et al . , phys .
lett . 31b ( 1970 ) 230 . 3 .
w. f. baker et al .
b51 ( 1974 ) 303 ; a. s. carroll et al .
33 ( 1974 ) 928 ; phys .
( 1974 ) 932 ; phys .
lett . 61b ( 1976 ) 303 ; phys .
lett . 80b ( 1979 ) 423 .
a. s. carroll et al . , phys .
lett . 80b ( 1979 ) 319 .
g. giacomelli , progress in nuclear physics 12 ( 1970 ) 77 . 6 .
w. bozzoli et al . , nucl . phys .
b140 ( 1978 ) 271 ; nucl .
b144 ( 1978 ) 317 ; nucl .
b159 ( 1979 ) 363 ; a. bussiere et al .
b174 ( 1980 ) 1 . 7 .
u. amaldi et al . , phys .
lett . 44b ( 1973 ) 112 ; s. r. amendolia et al .
lett . 44b ( 1973 ) 119 ; g. carboni et al .
b254 ( 1985)69 ; g. giacomelli and m. jacob , phys .
( 1979 ) 1 . 8 .
a. bertin et al . , phys .
( 1972 ) 260 ; phys .
lett . 42b ( 1972 ) 493 ; m. antinucci et al . , nuovo cimento lett .
6 ( 1973 ) 121 ; p. capiluppi et al .
b70 ( 1974 ) 1 ; nucl .
b79 ( 1974 ) 189 ; e. albini et al .
, nuovo cimento 32 ( 1976 ) 101 .
a. breakstone et al . , phys .
132b ( 1983 ) 458 ; nucl .
b248 ( 1984 ) 253 ; phys .
54 ( 1985 ) 2180 . 10 .
n. a. amos et al . , phys .
63 ( 1989 ) 2784 ; phys .
b247 ( 1990 127 ; phys .
b243 ( 1990)158 . 11 .
w. f. baker et al .
, ( production of @xmath76 and @xmath3 by 400 gev / c protons ) fermilab-78/79-exp ( 1978 ) . 12 .
g. giacomelli , int . j. mod
a5 ( 1990 ) 223 ; hadron - hadron and hadron - nuclei interactions at intermediate and high energies 1994 marshak memorial , dfub 9/94 ( 1994 ) ; total cross sections , kycia memorial symposium , 19/5/2000 , bnl , hep - ex/0006038 ; progress in nuclear physic s 12 ( 1970 ) 77 . 13 .
s. cecchini et al .
phys . 1 ( 1993 ) 369 ; g. giacomelli et al .
instr . meth .
a411 ( 1998 ) 41 ; h. dekhissi et al . ,
a662 ( 2000 ) 207 . 14 .
a. capella et al . , phys .
236 ( 1994 ) ; j. j. glauber et al .
, nucl . phys .
b21 ( 1970 ) 135 ; j. ranft , phys . rev .
d51 ( 1995 ) 64 . | a brief review is made of the present situation of hadron - hadron and hadron - nuclei total , elastic and inelastic cross sections at high energies .
dfub 00 - 20 \2 .
cm * hadron - hadron and hadron - nuclei collisions at high energies * 0.4 cm * invited paper at chacaltaya meeting on cosmic ray physics , + lapaz , 23 - 27 july , 2000 + * 0.4 cm g. giacomelli and r. giacomelli + 0.2 cm dipartimento di fisica delluniversit di bologna and infn , i-40127 bologna , italy |
the invariant variety of periodic points , or the ivpp for short@xcite@xcite@xcite , of a @xmath0 dimensional map @xmath1 , @xmath2 with @xmath3 invariants @xmath4 of period @xmath5 is a `` variety '' of periodic points @xmath6 which is given by only the invariants @xmath7 the ivpp has an important property which is called the ivpp theorem@xcite@xcite@xcite : _ let f be a @xmath0 dimensional `` rational '' map with @xmath3 invariants . if @xmath8 , an ivpp and discrete periodic points on a level set of any period do not exist in one map , simultaneously . _
the ivpp theorem gives an essential information of the integrability of the map . because a julia set@xcite which characterizes a non integrable system is given by the closure of the sum set of discrete `` repulsive '' periodic points on a level set .
therefore it seems that the existence of ivpp / julia set is incompatible with the existence of julia set / ivpp .
this conjecture leads to the sufficient conditions of the integrability of a map as follows : _ if a map @xmath9 has an ivpp / julia set then the map @xmath9 is integrable / non integrable . _ in particular , the reason why the ivpp is interesting , is a phenomenon that periodic points become a set of continuum points .
there have been derived many ivpps of various maps in the form of algebraic varieties , however , there have never been investigated the structure of an ivpp in detail .
one of our purposes of this paper is to explore the component structure of an ivpp . on the other hand it was known@xcite
that , when a map is restricted on the ivpp of period @xmath10 it provides a recurrence equation@xcite@xcite of period @xmath10 , an equation whose solutions are @xmath10 periodic for all initial points .
we show that a @xmath0 dimensional recurrence equation has a discrete symmetry , hence it gives a decomposition of the @xmath0 dimensional space to `` fundamental domains '' .
in other words , components of ivpp are given by fundamental domains of the recurrence equation which is a restriction map on ivpp . hence comparing these two objects
we will find a relation of components of ivpp and fundamental domains of recurrence equation as follows : _ components of ivpp @xmath11 fundamental domains of recurrence equation _
in addition , we give an algorithm to derive components of ivpp for all rational maps at the final section .
this structure tells us an important information , how points move on ivpps . in this paper , we show two concrete examples of such component structure of ivpps which are restricted by the cyclotomic polynomial@xcite of the recurrence maps .
first , we introduce a 2 dimensional map and its ivpps@xcite . @xmath12 with an invariant @xmath13 @xmath14 we know general formula of the ivpps([ivpp ] ) of the map([map ] ) as follows@xcite : @xmath15 and explicit forms , which will be discussed in this paper , as follows : @xmath16 ) ] in this subsection , we give components of the ivpp of the map([map ] ) .
first , we give the flow of points on the ivpp of period 3 by symbolic calculus which is parametrized by @xmath17 as follows : @xmath18 now , we can give boundaries of components of the ivpp of period 3 in the @xmath17 direction by the substitution @xmath19 for @xmath17 of ( [ seq ] ) .
therefore we get the components of the ivpp of period 3 in the @xmath17 direction @xmath20 as follows , @xmath21 @xmath22 we can draw a picture of `` tiling '' of the ivpp of period 3 . :
red , @xmath23 : blue , @xmath24 : green . ]
all other periods are to be discussed in almost the same way of the case of period 3 : flow of points on the ivpp of period 4 : @xmath25 components of the ivpp of period 4 in the @xmath17 direction @xmath26 as follows : @xmath27 @xmath28 : red , @xmath29 : blue , @xmath30 : green , @xmath31 : pink . ]
the flow of points on the ivpp of period 5 are as follows , @xmath32 where @xmath33 .
we notice that the ivpp of period 5 is multivalued dependent on @xmath34 , hence we must discuss them separately .
we get components of ivpp of period 5 @xmath35 as follows , @xmath36 @xmath37 of components of ivpp of period 5 , @xmath38 : red , @xmath39 : blue , @xmath40 : green , @xmath41 : pink , @xmath42 : gold . ] of components of ivpp of period 5 , @xmath43 : red , @xmath44 : blue , @xmath45 : green , @xmath46 : pink , @xmath47 : gold . ] the flow of points on the ivpp of period 6 is as follows : @xmath48 therefore we get components of ivpp of period 6 as follows : @xmath49 @xmath50 : red , @xmath51 : blue , @xmath52 : green , @xmath53 : pink , @xmath54 : gold , @xmath55 : yellow . ]
in this section , we discuss the structure of the components of ivpps of the map ( [ map ] ) .
the map([map ] ) is reduced by the invariant as follows : @xmath56 or in the linear map form @xmath57 where @xmath58 .
it is a mbius map .
furthermore , we can transform @xmath59 to a diagonal matrix by coordinate transformation as @xmath60 therefore , we get a diagonal form of ( [ mob ] ) as follows : @xmath61 where the @xmath62 are eigenvalues of @xmath59 .
in other words , if we take a new invariant @xmath63 then the map ( [ mob2 ] ) is just a scale transformation @xmath64 this fact means that the ivpp of ( [ mob2 ] ) of period @xmath10 is given by a cyclotomic polynomial@xcite : @xmath65 hence the components of the ivpp of period @xmath10 of the map([mob2 ] ) is equivalent to an orbifold @xmath66 : cc ] ] we give a formula of @xmath17 direction boundaries @xmath67 of the components of ivpps of period @xmath10 as follows : @xmath68 where @xmath69 satisfies the @xmath10-order cyclotomic polynomial .
furthermore , a formula in @xmath70 space is also given by @xmath71 these formulas are given by the pull buck the @xmath72 iterated map ( [ mob2 ] ) to the @xmath72 iterated map ( [ mob ] ) as follows , @xmath73 where @xmath74 is the diagonalization matrix of @xmath59 .
the @xmath17 direction boundaries are given by the substitution @xmath19 for @xmath17 .
therefore we can get the formula([formula ] ) . finally , we give one more point of view about the boundaries of components of the ivpp .
we substitute @xmath19 for @xmath17 of the flow on the ivpp , hence the boundaries are given by the `` intersections of the ivpp and zero points set of denominators of iterations of the map '' .
we draw a graph of zero points set of the denominators of iterations of the map ( [ map ] ) as follows : ) , @xmath9 : red , @xmath75 : green , @xmath76 : yellow , @xmath77 : blue , @xmath78 : purple , @xmath79 : light blue . ]
next , we try to discuss 3 dimensional case . in particular , we discuss about the 3 dimensional lotka - volterra map@xcite , and only the ivpp of period 2 , because higher dimensional cases and higher period cases are difficult to write symbolic calculus and decompositions .
@xmath80 with invariants @xmath81 , @xmath82 we know some conditions of the ivpps of the map ( [ lvmap ] ) as follows@xcite : @xmath83 we draw the ivpp of period 2 in fig . [ lvivpp ] .
similar to the case of 2 dimensional map , we give a symbolic calculus on ivpp of period 2 .
@xmath84 where @xmath85 this case also is multivalued .
therefore we get components of ivpp of period 2 as follows : @xmath86_x= ( -\infty , 0 ] , \quad [ ( c_{2_\pm 2})_1]_x= ( 0 , 1 ] , \quad [ ( c_{2_\pm 1})_2]_x= ( 1 , \infty ] , \quad [ ( c_{2_\pm 2})_2]_x= ( 1 , \infty].\ ] ] of components of ivpp of period 2 , @xmath87 : red , @xmath88 : blue , @xmath87 : green , @xmath89 : pink . ] of components of ivpp of period 2 , @xmath90 : red , @xmath91 : blue , @xmath90 : green , @xmath92 : pink . ] here , we notice that each graph has two kinds of tiling , _
i.e. , _ red and blue pair and green pink pair in both the fig .
11 and fig . 12 .
it is a new phenomenon .
in addition , we should give a remark about the reason of appearance of the striped pattern of tilings of the ivpp .
the ivpp of period 2 is given by the condition as follows , @xmath93 then the ivpp of period 2 has a free parameter @xmath94 which is not restricted by this condition
. therefore we draw graphs by showing stripes corresponding to the parameter of invariant @xmath94 which is stepped by @xmath95 for visibility . in other words ,
each stripe is one of a level set corresponding to a part of ivpp . in general , any ivpps of the map([lvmap ] ) are given by one condition , hence we can solve the condition about @xmath96 .
in other words , any ivpps are described by one parameter @xmath94 .
we give a small remark about the relation of decompositions of ivpp and fundamental domains of the recurrence equation .
the recurrence equation of period 2 is a restriction on ivpp of period 2 of the map([lvmap ] ) as follows : @xmath97 we notice that the map([lvre ] ) is independent on the parameter @xmath94 . in a similar way as the case of 2 dimensional map([map ] ) , we can get a diagonal form @xmath98 and the fundamental domain is given by @xmath99 .
in order to conclude this paper we would like to give our algorithm to determine the components of ivpp for all rational maps : _ algorithm _ * give an initial point @xmath100 on ivpp .
if ivpp is multivalued , then we must use our algorithm for all parts .
* symbolic calculation of the mapping on the ivpp of period @xmath10 : @xmath101 * substitution for a minimum value of domain of parameters @xmath102 of the ivpp of period @xmath10 for any points @xmath103 . * changing the parameters of step 2 to positive direction ,
until go to other components .
* if there exists a non discussed domain when finished the above operations , then go to step 3 for non discussed domain .
* therefore we get component family of the ivpp of period @xmath10 in the @xmath104 direction about parameters @xmath105 . in conclusion , above discussions give a relation of decomposition of ivpp and the tiling as follows : @xmath106
i would like to thank dr .
s. saito and mr .
y. wakimoto for useful discussion
. 00 s. saito and n. saitoh , _ j. phys
_ , * 76 * no.2 p.024006 , 2007 . s. saito and n. saitoh
_ j. math .
phys _ , * 51 * 063501 , 2010 . s. saito and n. saitoh , _
`` invariant varieties of periodic points '' _ in mathematical physics research developments , 2008 nova science publishers , inc . ,
capt.3 pp 85 - 139 , 2008 .
wolfram mathworld , mathworld.wolfram.com/cyclotomicpolynomial.html .
r. l. devaney , _ `` an introduction to chaotic dynamical systems '' _ , westview press , 2003 .
r. l. graham , d. e. knuth and o. patashnik , _ concrete mathematics _
( addison - wesley ) , 1994 .
r. hirota and h. yahagi , _ j. phys
_ , * 71 * , 2867 , 2002 .
s. saito and n. saitoh , _ j. phys . a : math .
_ , * 40 * , 12775 - 12787 , 2007 .
s. saito , n. saitoh , h. harada , t. yumibayashi and y. wakimoto , aip advances , aip i d : 003306adv , 2013 .
r. hirota , s. tsujimoto and t. imai , _ future directions of nonlinear dynamics in physical and biological systems _ , ed . by p.l.christiansen at al . , p.7 ( plenum press , new york , 1993 ) .
r. hirota , and s. tsujimoto , _ j.phys.soc.jpn .
_ * 64 * 3125 - 3127 ( 1995 ) .
we do not discuss a reason of missing the ivpp of the map([mob ] ) of period 2 .
this reason is given from the transformation of invariant([newinv ] ) @xmath107 by the cyclotomic polynomial , the ivpp of the map([mob2 ] ) of period 2 is given @xmath108 . therefore if @xmath108 then @xmath94 is satisfied @xmath109 by ( [ newinv ] ) .
we check the values of @xmath70 when @xmath17 is ( real ) boundary of components of ivpps .
these are given by the formula ( [ formula2 ] ) : | in this paper , we discuss duality about components of invariant variety of periodic points(ivpp)@xcite@xcite@xcite and fundamental domain of recurrence equation , and present an algorithm for the derivation of all components of ivpps of any rational maps .
it is based on the study of two examples of a 2 dimensional map and a 3 dimensional map . in particular , all components of ivpps of the 2 dimensional example are completely determined by means of the cyclotomic polynomials@xcite . |
the production of heavy quarkonium at hadron colliders provides useful laboratory for testing the high - energy limit of quantum chromodynamics ( qcd ) as well as the interplay of perturbative and nonperturbative phenomena in qcd .
the total collision energies , @xmath6 tev and 1.96 tev in tevatron runs i and ii , respectively , and @xmath7 tev or 14 tev at the lhc , sufficiently exceed the characteristic scale @xmath8 of the relevant hard processes , which is of order of quarkonium transverse mass @xmath9 , _ i.e. _ we have @xmath10 . in this high - energy regime , so called `` regge limit '' , the contribution of partonic subprocesses involving @xmath11-channel parton ( gluon or quark ) exchanges to the production cross section can become dominant
. then the transverse momenta of the incoming partons and their off - shell properties can no longer be neglected , and we deal with `` reggeized '' @xmath11-channel partons .
these @xmath1channel exchanges obey multi - regge kinematics ( mrk ) , when the particles produced in the collision are strongly separated in rapidity .
if the same situation is realized with groups of particles , then quasi - multi - regge kinematics ( qmrk ) is at work . in the case of @xmath0meson inclusive production
, this means the following : @xmath0meson ( mrk ) or @xmath4-meson plus gluon jet ( qmrk ) is produced in the central region of rapidity , while other particles are produced with large modula of rapidities . the parton reggeization approach ( pra ) @xcite is particularly appropriate for high - energy phenomenology .
we see , the assumption of a dominant role of mrk or qmrk production mechanisms at high energy works well .
pra is based on an effective quantum field theory implemented with the non - abelian gauge - invariant action including fields of reggeized gluons @xcite and reggeized quarks @xcite .
reggeized partons interact with quarks and yang - mills gluons in a specific way .
recently , in ref.@xcite , the feynman rules for the effective theory of reggeized gluons were derived for the induced and some important effective vertices .
this approach was successfully applied to interpret the production of isolated jets @xcite , prompt photons @xcite , diphotons @xcite , charmed mesons @xcite , bottom - flavored jets @xcite measured at the fermilab tevatron , at the desy hera and at the cern lhc , in the small-@xmath12 regime , where @xmath13 .
we suggest the mrk ( qmrk ) production mechanism to be the dominant one at small @xmath12 values . using the feynman rules for the effective theory
, we can construct heavy quarkonium production amplitudes in framework of non - relativistic qcd ( nrqcd)@xcite .
the factorization formalism of the nrqcd is a rigorous theoretical framework for the description of heavy - quarkonium production and decay .
the factorization hypothesis of nrqcd assumes the separation of the effects of long and short distances in heavy - quarkonium production .
nrqcd is organized as a perturbative expansion in two small parameters , the strong - coupling constant @xmath2 and the relative velocity @xmath3 of the heavy quarks inside a heavy quarkonium .
our previous analysis of charmonium @xcite and bottomonium @xcite production at the fermilab tevatron using the high - energy factorization scheme and nrqcd approach has shown the efficiency of such type of high - energy phenomenology . in this paper
we repeat our calculations for the direct and prompt @xmath0meson transverse momentum spectra at the fermilab tevatron @xcite to obtain by fitting procedure octet nonperturbative matrix elements ( nmes ) , than we calculate prompt @xmath0meson spectra , which were measured recently at the cern lhc collider at the energy of @xmath5 tev .
we find a good agreement of our calculations and experimental data from atlas @xcite , cms @xcite and lhcb @xcite collaborations .
working at the leading order ( lo ) in @xmath2 and @xmath3 we consider the following partonic subprocesses , which describe charmonium production at high energy : @xmath14(p ) , \label{eq : rrtoh}\\ r(q_1 ) + r(q_2 ) & \to & { \cal h } [ { ^3s}_1^{(1)}](p ) + g(p ' ) , \label{eq : rrtohg}\end{aligned}\ ] ] where @xmath15 is the reggeeized gluon and @xmath16 is the yang - mills gluon , respectively , with four - momenta indicated in parentheses , @xmath17 $ ] is the physical charmonium state , @xmath18 is the included @xmath19 fock state with the spin @xmath20 , total angular momentum @xmath21 , orbital angular momentum @xmath22 and in the singlet @xmath23 or in the octet @xmath24 color state . in the general case , the partonic cross section of charmonium production receives from the @xmath19 fock state @xmath25=[{}^{2s+1}l_j^{(1,8)}]$ ] the contribution @xcite @xmath26 \to { \cal h})=d\hat \sigma ( r + r \to c\bar c[^{2s+1}l_j^{(1,8)}])\frac{\langle { \cal o}^{\cal h}[^{2s+1}l_j^{(1,8)}]\rangle}{n_\mathrm{col}n_\mathrm{pol}},\ ] ] where @xmath27 for the color - singlet state , @xmath28 for the color - octet state , and @xmath29 , @xmath30\rangle$ ] are the nmes .
they satisfy the multiplicity relations @xmath31\rangle&=&(2j+1)\langle{\cal o}^{\psi(ns)}[^3p_0^{(8)}]\rangle,\nonumber\\ \langle{\cal o}^{\chi_{cj}}[^3p_j^{(1)}]\rangle&=&(2j+1)\langle{\cal o}^{\chi_{c0}}[^3p_0^{(1)}]\rangle,\nonumber\\ \langle{\cal o}^{\chi_{cj}}[^3s_1^{(8)}]\rangle&=&(2j+1)\langle{\cal o}^{\chi_{c0}}[^3s_1^{(8)}]\rangle,\end{aligned}\ ] ] which follow from heavy - quark spin symmetry to lo in @xmath3 .
the partonic cross section of @xmath19 production is defined as @xmath32)=\frac{1}{i}\overline{|{\cal a}(r + r \to c\bar c[^{2s+1}l_j^{(1,8)}])|^2}d\phi,\label{eq : dsigma}\ ] ] where @xmath33 is the flux factor of the incoming particles , which is taken as in the collinear parton model @xcite , @xmath34)$ ] is the production amplitude , the bar indicates average ( summation ) over initial - state ( final - state ) spins and colors , and @xmath35 is the invariant phase space volume of the outgoing particles .
this convention implies that the cross section in the high - energy factorization scheme is normalized approximately to the cross section for on - shell gluons in the collinear parton model when @xmath36 .
the lo results for the squared amplitudes of subprocesses ( [ eq : rrtoh ] ) and ( [ eq : rrtohg ] ) that we found by using the feynman rules of ref .
@xcite coincide with those we obtained in ref .
the formulas for the squared amplitudes @xmath37)|^2}$ ] for the @xmath38 subprocesses ( [ eq : rrtoh ] ) are listed in eq .
( 27 ) of ref
. @xcite .
the analytical result in case of the @xmath39 subprocess ( [ eq : rrtohg ] ) is presented in ref.@xcite , where the results for the @xmath38 subprocesses are also listed , but in another equivalent form . the relation between these forms
is discussed in ref.@xcite . exploiting the hypothesis of high - energy factorization ,
we may write the hadronic cross section @xmath40 as convolution of partonic cross section @xmath41 with unintegrated parton distribution functions ( pdfs ) @xmath42 of reggeized gluon in the proton , as @xmath43 @xmath44 , @xmath45 , @xmath46 and @xmath47 are the fractions of the proton momenta passed on to the reggeized gluons , and the factorization scale @xmath8 is chosen to be of order @xmath48 .
the collinear and unintegrated gluon distribution functions are formally related as @xmath49 so that , for @xmath50 , we recover the conventional factorization formula of the collinear parton model , @xmath51 we now describe how to evaluate the differential hadronic cross section from eq .
( [ eq : kt ] ) combined with the squared amplitudes of the @xmath52 and @xmath53 subprocesses ( [ eq : rrtoh ] ) and ( [ eq : rrtohg ] ) , respectively . the rapidity and pseudorapidity of a charmonium state with four - momentum @xmath54
are given by @xmath55 respectively , and @xmath56 .
the invariant phase volume @xmath35 in the eq .
( [ eq : dsigma ] ) for @xmath38 subprocess ( [ eq : rrtoh ] ) can be presented as follows : @xmath57 from the eqs .
( [ eq : dsigma ] ) , ( [ eq : kt ] ) and ( [ eq:21 ] ) we obtain the master formula for the @xmath52 subprocess ( [ eq : rrtoh ] ) : @xmath58 where @xmath59 and the relation @xmath60 has been taken into account .
the invariant phase volume @xmath35 in the eq .
( [ eq : dsigma ] ) for @xmath39 subprocess ( [ eq : rrtohg ] ) can be presented as follows : @xmath61 such a way , accordingly the eqs .
( [ eq : dsigma ] ) , ( [ eq : kt ] ) and ( [ eq:22 ] ) , we have the master formula for the @xmath53 subprocess ( [ eq : rrtohg ] ) : @xmath62 where @xmath63 are the angles enclosed between @xmath64 and the transverse momentum @xmath65 of @xmath66 , @xmath67.\ ] ] in our numerical analysis , we adopt as our default the prescription proposed by kimber , martin , and ryskin ( kmr ) @xcite to obtain unintegrated gluon pdf of the proton from the conventional integrated one , as implemented in watt s code @xcite .
as is well known @xcite , other popular prescriptions , such as those by blmlein @xcite or by jung and salam @xcite , produce unintegrated pdfs with distinctly different @xmath11 dependences . in order assess the resulting theoretical uncertainty
, we also evaluate the unintegrated gluon pdf using the blmlein approach , which resums small-@xmath68 effects according to the balitsky - fadin - kuraev - lipatov ( bfkl ) equation @xcite . as input for these procedures ,
we use the lo set of the martin - roberts - stirling - thorne ( mrst ) @xcite proton pdf as our default . throughout our analysis
the renormalization and factorization scales are identified and chosen to be @xmath69 , where @xmath70 is varied between 1/2 and 2 about its default value 1 to estimate the theoretical uncertainty due to the freedom in the choice of scales .
the resulting errors are indicated as shaded bands in the figures .
at first , we examine direct and prompt ( sum of direct production , production via radiative decays of @xmath71 mesons and production via decays of @xmath72 mesons contributions ) @xmath0meson production in proton - antiproton collisions at the tevatron at the energies @xmath6 tev ( run i)@xcite and @xmath73 tev ( run ii)@xcite in the central region of pseudorapidity @xmath74 .
the data of run ii includes @xmath12 distributions of prompt @xmath4 mesons , so far without separation into direct , @xmath75decay , and @xmath76decay contributions . in the ref.@xcite , we have performed a joint fit to the run i and run ii cdf data @xcite to obtain the color - octet nme s for @xmath4 , @xmath72 , and @xmath71 mesons .
our fits included five experimental data sets , which come as @xmath12 distributions of @xmath4 mesons from direct production , prompt production , @xmath71 decays , and @xmath72 decays .
since in the previous calculations we have used the old set mrst98 @xcite as a collinear input for unintegrated pdf , in the present study we repeat our fit with the next - generation mrst set @xcite . we find small differences between the old and new fit parameters , however , it is important for precise description of the data .
the results of our fit are presented in the table i along with results of the fit in the next to leading order ( nlo ) of collinear parton model ( pm ) and nrqcd approach @xcite .
oppositely the ref.@xcite , we perform a fit procedure by assumption for nmes to be only positive .
than , using the cdf data for a prompt @xmath4 production @xcite , presented separately for direct @xmath4 mesons , @xmath4 from @xmath77 decays , and @xmath4 from @xmath71 decays , we obtain color - octet nmes @xmath78\rangle$ ] , @xmath79\rangle$ ] , and @xmath80\rangle$ ] independently from each other .
looking at the table i , we find a good agreement with the nlo fit in collinear parton model performed in the ref .
@xcite , which strongly improves if we take into account that a sum of contributions of nmes @xmath81\rangle$ ] and @xmath82\rangle$ ] from the ref .
@xcite , leading to almost parallel @xmath4 transverse momenta spectra , corresponds to our contribution of the nme @xmath81\rangle$ ] .
such an agreement demonstrates a validity of factorization in the charmonium production in hadronic collisions , i.e. an independence of the @xmath19 production mechanism from the nonperturbative charmonium formation at the last step .
it is necessary to note that a same consent between lo results obtained in the uncollinear factorization scheme and nlo results obtained in the collinear parton model is also observed when describing other relevant processes , see refs .
@xcite . in figs . 14
, we compare the cdf data on @xmath4 mesons from direct production , @xmath72 decays and @xmath71 decays in run i @xcite and from prompt production in run ii @xcite , with the respective theoretical results evaluated with the nmes listed in table i. as default , we present in all figures the theoretical results which are obtained using kmr unintegrated gluon density @xcite .
the comparison between kmr @xcite and blmlein @xcite pdfs is made in fig . 4 for a prompt @xmath4 production only . in fig . 1
one can find a dominance of color - octet contributions at all values of direct @xmath4 meson transverse momentum : @xmath83 $ ] contribution dominates at the large values @xmath84 gev , and @xmath85 $ ] at the small @xmath86 gev .
the situation is very similar for @xmath4 production from @xmath72 decay , considered in fig .
it is also important , that the @xmath87 $ ] contribution vanishes in direct @xmath4 and @xmath72 production .
the obtained results are in agreement with previous calculations of ref .
@xcite with a slight difference . in ref .
@xcite it was also shown that in the direct @xmath4 production at the tevatron color - octet contribution dominates .
oppositely our conclusions , in ref .
@xcite , the main contribution comes from the @xmath85
$ ] nme at all transverse momenta .
the reason of this discrepancy arises from the fact that in ref .
@xcite the color - octet nmes for @xmath4 meson have been obtained by a fit of direct @xmath4 production data for @xmath88 gev only . in our fit
we take into account both direct @xcite and prompt @xcite @xmath4 production data , the last ones contain points in a small @xmath12 region .
we observe , the inclusion of prompt @xmath4 production data in the fit to change the relative weight of different color - octet nmes .
this fact should be important in study of polarized @xmath4 mesons . in case of @xmath4 production via decays of @xmath71 mesons , considered in fig . 3
, we confirm the conclusion of refs.@xcite , which reads , that in the high - energy factorization scheme , the color - singlet contribution is sufficient to describe the data for production of @xmath89wave charmonium states . in fig .
4 , the @xmath12 distribution of prompt @xmath4 production in tevatron run ii is presented as a sum of the contributions from direct production , @xmath72 decays , and @xmath71 decays .
we observe at the @xmath90 gev the @xmath4 mesons to be produced preferably directly .
the contribution from @xmath71 decays dominates at small @xmath91 gev .
the contribution from @xmath72 is smaller than other ones and it exceeds the contribution from @xmath71 decays only at the @xmath92 gev . the curve number ( 5 ) in fig .
4 is obtained using blmlein unpdf @xcite .
the visible difference between this curve and the curve ( 4 ) , which is obtained using kmr unpdf @xcite , takes place only in the region of small @xmath4 transverse momentum . in the range of @xmath93 gev
, there is no difference between them as in the prompt production as in the direct production or in the production via decays of high charmonium states , @xmath72 and @xmath71 . as it is obvious from figs .
14 , the theoretical uncertainties associated with the variation of the factorization scale @xmath8 are large at the small @xmath12 region , taking a value of about factor 5 between upper and lower boundaries , and they sufficiently decrease down to a factor 2 at the @xmath94 gev .
the uncertainties from errors in the color - octet nmes are small , they are about 7 - 10% . moving on from tevatron to the lhc , which is currently running at the total energy being about 3.5 times larger than at the tevatron , we expect the range of validity of our approach to be extended by the same factor , to @xmath95 gev , as we describe well the tevatron data at the range of @xmath96 gev .
this expectation is nicely confirmed in figs .
56 , where the recent measurements of the prompt @xmath4 production by the atlas collaboration at the cern lhc @xcite , which cover the kinematic region 1 gev @xmath97 gev and @xmath98 , are compared with our predictions based on the particle reggeization approach and nrqcd formalism .
the measurements of the cms collaboration @xcite were performed in the similar kinematic range 6.5 gev @xmath99 gev and @xmath98 , see fig . 7 .
we observe a dominant role of direct production mechanism in the prompt @xmath4 hadroproduction at the all values of @xmath4 meson transverse momentum .
concerning the relative contributions of @xmath72 decays and @xmath71 decays into a prompt @xmath4 production , we found the contribution from @xmath72 decays to dominate at the large @xmath100 gev , and the contribution from @xmath71 decays to dominate at the small @xmath12 , respectively .
additionally , we compare our predictions with the data from lhcb collaboration @xcite , which were extracted in the range @xmath101 gev and @xmath102 . we find a good agreement between our predictions and
prompt @xmath4 production data at the moderate rapidity interval @xmath103 , see figs .
at the same time our theoretical result overestimates the data of at most factor 2 in the range of large rapidity @xmath104 .
this distinction is expected in the parton reggeization approach , because the multi - regge kinematics conditions to be broken if @xmath4 mesons are produced with large rapidity .
we observe , in fig .
10 , that relative contributions of the color - singlet ( curve 1 ) and color - octet ( curve 2 ) production mechanisms to the prompt @xmath4 spectrum strongly depend on the @xmath4 transverse momentum . similarly to the nlo calculations in the collinear parton model , the color - octet contribution dominates at the large @xmath12 region , basically via the contributions of the color - octet nmes @xmath105\rangle$ ] and @xmath106\rangle$ ] .
it is significant , the experimental data @xcite depend on the assumption of polarization of produced @xmath4 mesons slightly .
we perform calculations and make a comparison to the data in a case of non - polarized @xmath4 meson production . comparing our results with the recent studies of @xmath4 meson hadroproduction in the conventional collinear pm , which were performed in full nlo approximation of nrqcd formalism @xcite or in the non - complete nnlo@xmath107 approximation of color - singlet model @xcite
, we should emphasize the following . at first , oppositely to nlo and nnlo@xmath107 calculations , which provide a good description of data only at non - small @xmath88 gev , we can reproduce data well at all transverse momenta @xmath12 . at second , the present study along with the previous investigations in the parton reggeization approach @xcite demonstrate the important role of ( quasi)multi - regge kinematics in particle production at high energies , this feature is out of account in the collinear pm .
such a way , we find the approach based on the effective theory of reggeized partons @xcite and high - energy factorization scheme with unintegrated pdfs , which in the large logarithmic terms ( @xmath108 ) are resummed in all orders of strong coupling constant @xmath2 , to be more adequate for the description of experimental data than fixed order in @xmath2 calculations in the frameworks of collinear pm .
the fermilab tevatron and , even more so , the cern lhc are currently probing particle physics at terascale c.m .
energies @xmath109 , so that the hierarchy @xmath10 , which defines the mrk and qmrk regimes , is satisfied for processes of heavy quark and heavy quarkonium production in the central region of rapidity , where @xmath8 is of order of their transverse mass . in this paper , we studied qcd processes of particular interest , namely prompt @xmath4 hadroproduction , at los in the parton reggeization approach and nrqcd approach , in which they are mediated by @xmath52 and @xmath39 partonic subprocesses initiated by reggeized gluon collisions .
we found by the fit of tevatron data that numerical values of the color - octet nmes are very similar to ones obtained in the full nlo calculations based on nrqcd approach @xcite . using these nmes
, we nicely described recent lhc data for prompt @xmath4 meson production measured by atlas @xcite , cms @xcite and lhcb @xcite collaborations at the whole presented range of @xmath4 transverse momenta .
we found only one exclusion , the region of large modulo of rapidity @xmath110 , where lhcb data are by a factor 2 smaller than our predictions .
however , this kinematical region is out of the applicability limits of the mrk or qmrk pictures . here and in refs .
@xcite , parton reggeization approach was demonstrated to be a powerful tool for the theoretical description of qcd processes in the high - energy limit .
we are grateful to b. a. kniehl , m. butenschn , m. bscher and n. n. nikolaev for useful discussions .
the work of m. a. n. and a. v. s. was supported by the federal ministry for science and education of the russian federation under contract no .
14.740.11.0894 and in part by the grant dfg 436 rus 113/940 ; the work of v. a. s. was supported in part by the russian foundation for basic research under grant 11 - 02 - 00769-a and by sfb fellowship of hamburg university ( sfb-676 ) .
l. n. lipatov , sov .
j. nucl .
* 23 * , 338 ( 1976 ) [ yad .
* 23 * , 642 ( 1976 ) ] ; e. a. kuraev , l. n. lipatov , and v. s. fadin , sov .
jetp * 44 * , 443 ( 1976 ) [ zh .
teor . fiz .
* 71 * , 840 ( 1976 ) ] ; sov .
jetp * 45 * , 199 ( 1977 ) [ zh . eksp .
. fiz . * 72 * , 377 ( 1977 ) ] ; i. i. balitsky and l. n. lipatov , sov .
j. nucl .
* 28 * , 822 ( 1978 ) [ yad .
* 28 * , 1597 ( 1978 ) ] ; sov .
jetp * 63 * , 904 ( 1986 ) [ zh .
fiz . * 90 * , 1536 ( 1986 ) ] .
b. a. kniehl , v. a. saleev , and a. v. shipilova , phys . rev .
d * 81 * , 094010 ( 2010 ) [ arxiv:1003.0346 [ hep - ph ] ] ; pos(dis2010 ) , 177 ( 2010 ) ; in proceedings of physics at the lhc 2010 ( plhc2010 ) , hamburg , germany , 2010 , edited by m. diehl , j. haller , t. schrner - sadenius , and g. steinbrueck , doi : ` http://dx.doi.org/10.3204/desy-proc-2010-01/shipilova ` .
b. a. kniehl , d. v. vasin , and v. a. saleev , phys .
d * 73 * , 074022 ( 2006 ) [ arxiv : hep - ph/0602179 ] ; in proceedings of the 15th international workshop on deep - inelastic scattering and related subjects ( dis 2007 ) , munich , germany , 2007 , edited by g. grindhammer and k. sachs , doi : ` http://dx.doi.org/10.3360/dis.2007.169 ` .
m. a. kimber , a. d. martin , and m. g. ryskin , eur .
j. c * 12 * , 655 ( 2000 ) [ arxiv : hep - ph/9911379 ] ; phys .
d * 63 * , 114027 ( 2001 ) [ arxiv : hep - ph/0101348 ] ; g. watt , a. d. martin , and m. g. ryskin , eur .
j. c * 31 * , 73 ( 2003 ) [ arxiv : hep - ph/0306169 ] ; phys .
d * 70 * , 014012 ( 2004 ) ; * 70 * , 079902(e ) ( 2004 ) [ arxiv : hep - ph/0309096 ] .
small-@xmath68 collaboration , b. andersson _ et al .
_ , eur .
j. c * 25 * , 77 ( 2002 ) [ arxiv : hep - ph/0204115 ] ; f. hautmann and h. jung , nucl .
b ( proc . suppl . ) * 184 * , 64 ( 2008 ) [ arxiv:0712.0568 [ hep - ph ] ] .
m. butenschoen and b. a. kniehl , phys .
lett . * 106 * , 022003 ( 2011 ) [ arxiv:1009.5662 [ hep - ph ] ] ; phys .
d * 84 * , 051501 ( 2011 ) [ arxiv:1105.0820 [ hep - ph ] ] .
artoisenet et al .
rev . lett .
* 101 * , 152001 ( 2008 ) ; j. p. lansberg , phys .
lett . b * 695 * , 149 ( 2011 ) ; e - print [ arxiv:1107.0292 ] .
s. brandt .
statistical and computational methods in data analysis .
heidelberg university , north holland publishing company , 1970 .
.[tab : nme ] nmes for @xmath4 , @xmath77 , and @xmath71 mesons from fits of the cdf data @xcite in the nlo collinear parton model @xcite and in the parton reggeization approach using the blmlein @xcite , and kmr @xcite unintegrated gluon distribution functions .
the errors on our fit results are determined by the varying in turn each nme up and down about its central value until the value of @xmath111 is increased by unity keeping all other nmes fixed at their central values .
when we obtained the value of @xmath112 , we have used normalizing multiplier approach @xcite to reduce this value to unity . [
cols="^,^,^,^",options="header " , ] direct @xmath4 transverse momentum spectrum from cdf collaboration @xcite , @xmath6 tev , @xmath74 , ( 1 ) is the contribution of @xmath113 $ ] state , ( 2 ) @xmath114 $ ] , ( 3 ) @xmath115 $ ] , ( 4 ) sum of their all.,scaledwidth=95.0% ] @xmath4 transverse momentum spectrum from @xmath72 decays from cdf collaboration @xcite , @xmath6 tev , @xmath74 , ( 1 ) is the contribution of @xmath113 $ ] state , ( 2 ) @xmath115 $ ] , ( 3 )
@xmath114 $ ] , ( 4 ) sum of their all.,scaledwidth=95.0% ] @xmath4 transverse momentum spectrum from @xmath71 decays from cdf collaboration @xcite , @xmath6 tev , @xmath74 , ( 1 ) is the contribution of @xmath116 $ ] state , ( 2 ) @xmath117 $ ] , ( 3 ) @xmath118 $ ] , ( 4 )
@xmath113 $ ] , ( 5 ) sum of their all.,scaledwidth=95.0% ] prompt @xmath4 transverse momentum spectrum from cdf collaboration @xcite , @xmath73 tev , @xmath119 , ( 1 ) is the direct production , ( 2 ) from @xmath71 decays , ( 3 ) from @xmath72 decays , ( 4 ) sum of their all ( kmr unpdf ) , ( 5 ) sum of all contributions ( blmlein unpdf).,scaledwidth=95.0% ] prompt @xmath4 transverse momentum spectrum from atlas collaboration @xcite , @xmath7 tev , ( 1 ) is the direct production , ( 2 ) from @xmath71 decays , ( 3 ) from @xmath72 decays , ( 4 ) sum of their all . for the different range in the rapidity : ( a) @xmath120 , ( b ) @xmath121.,scaledwidth=80.0% ] prompt @xmath4 transverse momentum spectrum from atlas collaboration @xcite , @xmath7 tev ,
( 1 ) is the direct production , ( 2 ) from @xmath71 decays , ( 3 ) from @xmath72 decays , ( 4 ) sum of their all . for the different range in the rapidity : a )
@xmath122 , ( b )
@xmath123 .,scaledwidth=80.0% ] prompt @xmath4 transverse momentum spectrum from cms collaboration @xcite , @xmath7 tev , ( 1 ) is the direct production , ( 2 ) from @xmath71 decays , ( 3 ) from @xmath72 decays , ( 4 ) sum of their all . for the different range in the rapidity : ( a) @xmath124 , ( b )
@xmath125 , ( c )
@xmath126,scaledwidth=90.0% ] prompt @xmath4 transverse momentum spectrum from lhcb collaboration @xcite , @xmath7 tev , ( 1 ) is the direct production , ( 2 ) from @xmath71 decays , ( 3 ) from @xmath72 decays , ( 4 ) sum of their all .
for the different range in the rapidity : ( a)- @xmath127 , ( b ) - @xmath128 , ( c ) - @xmath129,scaledwidth=90.0% ] prompt @xmath4 transverse momentum spectrum from lhcb collaboration @xcite , @xmath7 tev , ( 1 ) is the direct production , ( 2 ) from @xmath71 decays , ( 3 ) from @xmath72 decays , ( 4 ) sum of their all . for the different range in the rapidity : ( a) @xmath130 , ( b )
@xmath131,scaledwidth=90.0% ] the relative contributions of the color - singlet ( curve 1 ) and color - octet ( curve 2 ) production mechanisms to the prompt @xmath4 transverse momentum spectrum at the @xmath7 tev , @xmath122.,scaledwidth=95.0% ] | we study prompt @xmath0meson hadroproduction invoking the hypothesis of gluon reggeization in @xmath1channel exchanges at high energy and the factorization formalism of nonrelativistic quantum chromodynamics at the leading order in the strong - coupling constant @xmath2 and the relative velocity of quarks @xmath3 .
the transverse - momentum distribution of direct and prompt @xmath0meson production measured at the fermilab tevatron fitted to obtain the nonperturbative long - distance matrix elements , which are used to predict prompt @xmath4 production spectra at the cern lhc . at the numerical calculation
, we adopt the kimber - martin - ryskin and blmlein prescriptions to derive unintegrated gluon distribution function of the proton from their collinear counterpart , for which we use the martin - roberts - stirling - thorne set . without adjusting any free parameters
, we find good agreement with measurements by the atlas , cms and lhcb collaborations at the lhc at the hadronic c.m.energy @xmath5 tev . -2 cm |
given probability densities @xmath14 with respect to lebesgue measure @xmath15 on @xmath16 , and a cost function @xmath17 $ ] , monge s transportation problem is to find a map @xmath18 pushing @xmath19 forward to @xmath20 which minimizes the expected transportation cost @xcite @xmath21 where @xmath22 means @xmath23 = \mu^+[g^{-1}(y)]$ ] for each borel @xmath24 . in this context
it is interesting to know when a map attaining this infimum exists ; sufficient conditions for this were found by gangbo @xcite and by levin @xcite , extending work of a number of authors described in @xcite @xcite .
one may also ask when @xmath25 will be smooth , in which case it must satisfy the prescribed jacobian equation @xmath26 , which turns out to reduce to a degenerate elliptic partial differential equation of monge - ampre type for a scalar potential @xmath27 satisfying @xmath28 .
sufficient conditions for this were discovered by ma , trudinger and wang @xcite and trudinger and wang @xcite @xcite , after results for the special case @xmath29 had been worked out by brenier @xcite , delane @xcite , caffarelli @xcite @xcite @xcite @xcite @xcite , and urbas @xcite , and for the cost @xmath30 and measures supported on the unit sphere by wang @xcite . if the ratio @xmath31 although bounded away from zero and infinity
is not continuous , the map @xmath25 will not generally be differentiable , though one may still hope for it to be continuous .
this question is not merely of academic interest , since discontinuities in @xmath11 arise unavoidably in applications such as partial transport problems @xcite
@xcite @xcite @xcite .
such results were established for the classical cost @xmath32 by caffarelli @xcite
@xcite @xcite , for its restriction to the product of the boundaries of two strongly convex sets by gangbo and mccann @xcite , and for more general costs satisfying the strong regularity hypothesis * ( a3 ) * of ma , trudinger and wang @xcite which excludes the cost @xmath32 by loeper @xcite ; see also @xcite @xcite @xcite . under the weaker and
degenerate hypothesis * ( a3w ) * of trudinger and wang @xcite , which includes the cost @xmath32 ( and whose necessity for regularity was shown by loeper @xcite ) , such a result remains absent from the literature ; we aim to provide it below under a slight strengthening of their condition ( still including the quadratic cost ) which appeared in kim and mccann @xcite@xcite , called _ non - negative cross - curvature_. ( related but different families of strengthenings were investigated by loeper and villani @xcite and figalli and rifford @xcite . )
our main result is stated in theorem [ t : hoelder ] .
a number of interesting cost functions do satisfy non - negative cross - curvature hypothesis , and have applications in economics @xcite and statistics @xcite .
examples include the euclidean distance between two convex graphs over two sufficiently convex sets in @xmath16 @xcite , the riemannian distance squared on multiple products of round spheres ( and their riemannian submersion quotients , including products of complex projective spaces @xmath33 ) @xcite , and the simple harmonic oscillator action @xcite . in a sequel
, we apply the techniques developed here to deduce regularity of optimal maps in the latter setting @xcite .
moreover , theorem [ t : hoelder ] allows one to apply the higher interior regularity results established by liu , trudinger and wang @xcite , ensuring in particular that the transport map is @xmath34-smooth if @xmath35 and @xmath36 are
. + most of the regularity results quoted above derive from one of two approaches .
the continuity method , used by delano , urbas , ma , trudinger and wang , is a time - honored technique for solving nonlinear equations .
here one perturbs a manifestly soluble problem ( such as @xmath37 with @xmath38 , so that @xmath39 ) to the problem of interest ( @xmath40 , @xmath41 ) along a family @xmath42 designed to ensure the set of @xmath43 $ ] for which it is soluble is both open and closed .
openness follows from linearization and non - degenerate ellipticity using an implicit function theorem . for the non - degenerate ellipticity and closedness
, it is required to establish estimates on the size of derivatives of the solutions ( assuming such solutions exist ) which depend only on information known a priori about the data @xmath44 . in this way one obtains smoothness of the solution @xmath45 from the same argument which shows @xmath46 to exist .
the alternative approach relies on first knowing existence and uniqueness of a borel map which solves the problem in great generality , and then deducing continuity or smoothness by close examination of this map after imposing additional conditions on the data @xmath47 .
although precursors can be traced back to alexandrov @xcite , in the present context this method was largely developed and refined by caffarelli @xcite
@xcite @xcite , who used convexity of @xmath27 crucially to localize the map @xmath48 and renormalize its behaviour near a point @xmath49 of interest in the borderline case @xmath50 . for non - borderline
* ( a3 ) * costs , simpler estimates suffice to deduce continuity of @xmath25 , as in @xcite @xcite @xcite @xcite ; in this case loeper was actually able to deduce an explicit bound @xmath51 on the hlder exponent of @xmath25 when @xmath52 , which was recently improved to its sharp value @xmath53 by liu @xcite using a technique related to the one we develop below and discovered independently from us ; both loeper and liu also obtained explicit exponents @xmath54 for @xmath55 with @xmath56 @xcite or @xmath57 @xcite and @xmath58 .
explicit bounds on the exponent are much worse in the classical case @xmath50 @xcite , when such exponents do not even exist unless @xmath59 @xcite @xcite .
+ below we extend the approach of caffarelli to non - negatively cross - curved costs , a class which includes the classical quadratic cost .
our idea is to add a null lagrangian term to the cost and exploit diffeomorphism ( i.e. gauge ) invariance to choose coordinates which depend on the point of interest that restore convexity of @xmath60 ; our strengthened hypothesis then permits us to exploit caffarelli s approach more systematically than liu was able to do @xcite .
however , we still need to overcome serious difficulties , such as getting an alexandrov estimate for @xmath7-subdifferentials ( see section [ s : alexandrov ] ) and dealing with the fact that the domain of the cost function ( where it is smooth and satisfies appropriate cross - curvature conditions ) may not be the whole of @xmath16 .
( this situation arises , for example , when optimal transportation occurs between domains in riemannian manifolds for the distance squared cost or similar type . )
the latter is accomplished using theorem [ thm : bdry - inter ] , where it is first established that optimal transport does not send interior points to boundary points , and vice versa , under the strong @xmath7-convexity hypothesis * ( b2u ) * described in the next section .
( for this result to hold , the cost needs not to satisfy the condition * ( a3w)*. ) without our strengthening of trudinger and wang s hypothesis @xcite ( i.e. with only * ( a3w ) * ) , we obtain the convexity of all level sets of @xmath60 in our chosen coordinates as liu also did ; this yields some hope of applying caffarelli s method and the full body of techniques systematized in gutierrez @xcite , but we have not been successful at overcoming the remaining difficulties in such generality . in two dimensions
however , there is an alternate approach to establishing continuity of optimal maps which applies to this more general case ; it was carried out by figalli and loeper @xcite , but relies on theorem [ thm : bdry - inter ] , first proved below .
let us begin by formulating the relevant hypothesis on the cost function @xmath2 in a slightly different format than ma , trudinger and wang @xcite . for each @xmath61 assume
: + * ( b0 ) * @xmath62 and @xmath63 are open and bounded and @xmath64 ; + * ( b1 ) * ( bi - twist ) @xmath65 are diffeomorphisms onto their ranges ; + * ( b2 ) * ( bi - convex ) @xmath66 are convex subsets of @xmath16 ; + * ( b3 ) * ( non - negative cross - curvature ) @xmath67 : = -\frac{\partial^4}{\partial s^2 \partial t^2}\bigg|_{(s , t)=(0,0 ) } c(x(s),y(t ) ) \ge 0\ ] ] for every curve @xmath68 \longmapsto \big(d_y c(x(t),y(0)),d_x c(x(0),y(t))\big ) \in { \mathbf{r}}^{2n}$ ] which is an affinely parameterized line segment .
+ if the convex domains @xmath69 and @xmath70 in * ( b2 ) * are all strongly convex , we say * ( b2u ) * holds . here a convex set @xmath62 is said to be _ strongly _ convex if there exists a radius @xmath71 ( depending only on @xmath72 , ) such that each boundary point @xmath73 can be touched from outside by a sphere of radius @xmath74 enclosing @xmath75 ; i.e. @xmath76 where @xmath77 is an outer unit normal to a hyperplane supporting @xmath75 at @xmath78 .
when @xmath75 is smooth , this means all principal curvatures of its boundary are bounded below by @xmath79 .
hereafter @xmath80 denotes the closure of @xmath75 , @xmath81 denotes its interior , @xmath82 its diameter , and for any measure @xmath83 on @xmath80 , we use the term _ support _ and the notation @xmath84 to refer to the smallest closed set carrying the full mass of @xmath85 . condition * ( b3 ) * is the above - mentioned strengthening of trudinger and wang s criterion * ( a3w ) * guaranteeing smoothness of optimal maps in the monge transportation problem ( [ monge ] ) ; unlike us , they require ( [ mtw ] ) only if , in addition @xcite , @xmath86 necessity of trudinger and wang s condition for continuity was shown by loeper @xcite , who noted its covariance ( as did @xcite @xcite ) and some relations to curvature .
their condition relaxes the hypothesis * ( a3 ) * proposed earlier with ma @xcite , which required strict positivity of ( [ mtw ] ) when ( [ nullity ] ) holds .
the strengthening considered here was first studied in a different but equivalent form by kim and mccann @xcite , where both the original and the modified conditions were shown to correspond to pseudo - riemannian sectional curvature conditions induced by the cost @xmath7 on @xmath87 , highlighting their invariance under reparametrization of either @xmath75 or @xmath88 by diffeomorphism ; see ( * ? ? ?
* lemma 4.5 ) .
the convexity of @xmath69 required in * ( b2 ) * is called _
@xmath7-convexity of @xmath72 with respect to @xmath89 _ by ma , trudinger and wang ( or strong @xmath7-convexity if * ( b2u ) * holds ) ; they call curves @xmath90 , for which @xmath91 \longmapsto -d_y c(x(s),{{\tilde y}})$ ] is a line segment ,
_ @xmath7-segments with respect to @xmath89_. similarly , @xmath88 is said to be strongly @xmath92-convex with respect to @xmath78 or with respect to @xmath80 when it holds for all @xmath93 and the curve @xmath94 from is said to be a @xmath92-segment with respect to @xmath78 .
such curves correspond to geodesics @xmath95 and @xmath96 in the geometry of kim and mccann . here and throughout , _
line segments _ are always presumed to be affinely parameterized .
we are now in a position to summarize our main result : [ t : hoelder ] let @xmath97 satisfy * ( b0)**(b3 ) * and * ( b2u)*. fix probability densities @xmath98 and @xmath99 with @xmath100 and set @xmath101 .
if the ratio @xmath102 for some open set @xmath103 , then the minimum ( [ monge ] ) is attained by a map @xmath104 whose restriction to @xmath8 is continuous and one - to - one . as recalled below in section [
s : background ] ( or see e.g. @xcite ) it is well - known by kantorovich duality that the optimal joint measure @xmath105 from vanishes outside the @xmath7-subdifferential of a potential @xmath106 satisfying the @xmath7-convexity hypothesis , and that the map @xmath104 which we seek is uniquely recovered from this potential using the diffeomorphism * ( b1 ) * to solve .
thus the continuity claimed in theorem [ t : hoelder ] is equivalent to @xmath107 .
since @xmath108 do not charge the boundaries of @xmath75 ( or of @xmath88 ) , lemma [ l : cma properties](e ) shows the @xmath7-monge - ampre measure defined in has density satisfying @xmath109 on @xmath80 and @xmath110 on @xmath8 . thus @xmath111 according to theorem [ t : continuity ] .
injectivity of @xmath25 follows from theorem [ t : strict convex ] , and the fact that the graph of @xmath25 is contained in the set @xmath112 of . note that in case @xmath113 is continuous and compactly supported , choosing @xmath114 for all @xmath115 , yields continuity and injectivity of the optimal map @xmath116 throughout @xmath117 .
theorem [ t : hoelder ] provides a necessary prerequisite for the higher interior regularity results established by liu , trudinger and wang in @xcite a prerequisite which one would prefer to have under the weaker hypotheses * ( b0)**(b2 ) * and * ( a3w)*. note that these interior regularity results can be applied to manifolds , after getting suitable stay - away - from - the - cut - locus results : this is accomplished for multiple products of round spheres in @xcite , to yield the first regularity result that we know for optimal maps on riemannian manifolds which are not flat , yet have some vanishing sectional curvatures .
kantorovich discerned @xcite @xcite that monge s problem ( [ monge ] ) could be attacked by studying the linear programming problem @xmath118 here @xmath119 consists of the joint probability measures on @xmath120 having @xmath108 for marginals . according to the duality theorem from linear programming ,
the optimizing measures @xmath121 vanish outside the zero set of @xmath122 for some pair of functions @xmath123 satisfying @xmath124 these arise as optimizers of the dual program .
this zero set is called the @xmath7-subdifferential of @xmath27 , and denoted by @xmath125 we also write @xmath126 , and @xmath127 , and @xmath128 for @xmath129 . formula ( [ c - transform ] ) defines a generalized legendre - fenchel transform called the _
@xmath7-transform _ ; any function satisfying @xmath130 is said to be _
@xmath7-convex _ , which reduces to ordinary convexity in the case of the cost @xmath50 . in that case
@xmath131 reduces to the ordinary subdifferential @xmath132 of the convex function @xmath27 , but more generally we define @xmath133 @xmath134 , and @xmath135 . assuming @xmath136 ( which is the case if * ( b0 ) * holds ) , any @xmath7-convex function @xmath137 will be semi - convex , meaning its hessian admits a bound from below @xmath138 in the distributional sense ; equivalently , @xmath139 is convex on each ball in @xmath75 @xcite
. in particular , @xmath27 will be twice - differentiable @xmath15-a.e .
on @xmath75 in the sense of alexandrov . as in @xcite @xcite @xcite , hypothesis * ( b1 ) * shows the map @xmath140 is uniquely defined on the set @xmath141 of differentiability for @xmath27 by @xmath142 the graph of @xmath25 , so - defined , lies in @xmath131 .
the task at hand is to show continuity and injectivity of @xmath25 the former being equivalent to @xmath143 by studying the relation @xmath144 . to this end
, we define a borel measure @xmath145 on @xmath16 associated to @xmath27 by latexmath:[\[\label{c - monge - ampere measure } @xmath129 ; it will be called the _ @xmath7-monge - ampre measure _ of @xmath27 .
( similarly , we define @xmath147 . )
we use the notation @xmath148 on @xmath8 as a shorthand to indicate @xmath149 for each @xmath150 ; similarly , @xmath151 indicates @xmath152 . as the next lemma shows , uniform bounds above and below on the marginal densities of a probability measure @xmath121 vanishing outside @xmath131 imply similar bounds on @xmath153 .
[ l : cma properties ] let @xmath7 satisfy * ( b0)*-*(b1 ) * , while @xmath9 and @xmath154 denote @xmath7-convex functions for each @xmath155 . fix @xmath156 and constants @xmath157 . + ( a )
then @xmath158 and @xmath159 is a borel measure of total mass @xmath160 on @xmath161 .
( b ) if @xmath162 uniformly , then @xmath163 is @xmath7-convex and @xmath164 weakly-@xmath165 in the duality against continuous functions on @xmath166 .
+ ( c ) if @xmath167 for all @xmath168 , then the functions @xmath154 converge uniformly if and only if the measures @xmath169 converge weakly-@xmath165 .
+ ( d ) if @xmath170 on @xmath80 , then @xmath171 on @xmath172 . +
( e ) if a probability measure @xmath173 vanishes outside @xmath174 , and has marginal densities @xmath11 , then @xmath175 on @xmath176 and @xmath177 on @xmath172 imply @xmath178 on @xmath8 , whereas @xmath179 on @xmath8 and @xmath180 on @xmath172 imply @xmath181 on @xmath8 .
\(a ) the fact @xmath182 is an immediate consequence of definition ( [ c - subdifferential ] ) . since @xmath183
, the @xmath7-transform @xmath184 defined by ( [ c - transform ] ) can be extended to a lipschitz function on a neighbourhood of @xmath172 , hence rademacher s theorem asserts @xmath185 is a set of full lebesgue measure in @xmath172 .
use * ( b1 ) * to define the unique solution @xmath186 to @xmath187 as in @xcite @xcite , the vanishing of @xmath122 implies @xmath188 , at least for all points @xmath189 where @xmath172 has lebesgue density greater than one half . for borel @xmath129
, this shows @xmath190 differs from the borel set @xmath191 by a @xmath15 negligible subset of @xmath172 , whence @xmath192 so claim ( a ) of the lemma is established .
\(b ) let @xmath193 .
it is not hard to deduce @xmath7-convexity of @xmath163 , as in e.g. @xcite .
define @xmath194 and @xmath195 on @xmath196 as above , so that @xmath197 .
moreover , @xmath198 in @xmath199 , where @xmath200 is the @xmath201-dual to @xmath202 .
the uniform semiconvexity of @xmath203 ( i.e. convexity of @xmath204 ) ensures pointwise convergence of @xmath205 @xmath15-a.e . on @xmath172 . from @xmath206
we deduce @xmath207 @xmath15-a.e . on @xmath172 .
this is enough to conclude @xmath208 , by testing the convergence against continuous functions and applying lebesgue s dominated convergence theorem .
\(c ) to prove the converse , suppose @xmath154 is a sequence of @xmath7-convex functions which vanish at @xmath78 and @xmath209 weakly-@xmath165 .
since the @xmath154 have lipschitz constants dominated by @xmath210 and @xmath161 is compact , any subsequence of the @xmath154 admits a convergent further subsequence by the ascoli - arzel theorem . a priori ,
the limit @xmath163 might depend on the subsequences , but ( b ) guarantees @xmath211 , after which ( * ? ? ?
* proposition 4.1 ) identifies @xmath163 uniquely in terms of @xmath212 and @xmath213 , up to an additive constant ; this arbitrary additive constant is fixed by the condition @xmath214 .
thus the whole sequence @xmath154 converges uniformly .
\(e ) now assume a finite measure @xmath173 vanishes outside @xmath215 and has marginal densities @xmath11 .
then the second marginal @xmath216 of @xmath121 is absolutely continuous with respect to lebesgue and @xmath121 vanishes outside the graph of @xmath217 , whence @xmath218 by e.g. ( * ? ? ?
* lemma 2.1 ) .
( here @xmath219 denotes the identity map , restricted to the domain @xmath220 of definition of @xmath221 . )
recalling that @xmath192 ( see the proof of ( a ) above ) , for any borel @xmath222 we have @xmath223whenever @xmath224 and @xmath179 .
we can also reverse the last four inequalities and interchange @xmath225 with @xmath226 to establish claim ( e ) of the lemma .
\(d ) the last point remaining follows from ( e ) by taking @xmath227 .
indeed an upper bound @xmath225 on @xmath228 throughout @xmath161 and lower bound @xmath229 on @xmath15 translate into a lower bound @xmath230 on @xmath231 , since the reflection @xmath232 defined by @xmath233 for each @xmath234 vanishes outside @xmath235 and has second marginal absolutely continuous with respect to lebesgue by the hypothesis @xmath236 . differentiating ( [ implicit g ] )
formally with respect to @xmath78 and recalling @xmath237 yields the monge - ampre type equation @xmath238 } { |\det d^2_{xy } c({{\tilde x}},g({{\tilde x}}))| } = \frac{f^+({{\tilde x}})}{f^-(g({{\tilde x}}))}\ ] ] on @xmath75 , where @xmath239 is given as a function of @xmath78 and @xmath240 by ( [ implicit g ] ) .
degenerate ellipticity follows from the fact that @xmath116 produces equality in @xmath241 .
a condition under which @xmath7-convex weak-@xmath165 solutions are known to exist is given by @xmath242 the boundary condition @xmath243 which then guarantees @xmath244 to be uniquely determined @xmath35-a.e .
is built into our definition of @xmath7-convexity .
in fact , ( * ? ? ? * proposition 4.1 ) shows @xmath9 to be uniquely determined up to additive constant if either @xmath245 or @xmath246 @xmath15-a.e . on its connected domain , @xmath75 or @xmath88 .
a key result we shall exploit several times is a maximum principle first deduced from trudinger and wang s work @xcite by loeper ; see ( * ? ? ?
* theorem 3.2 ) . a simple and direct proof , and also an extension can be found in ( * ? ? ?
* theorem 4.10 ) , where the principle was also called ` double - mountain above sliding - mountain ' ( * dasm * ) . other proofs and extensions appear in @xcite @xcite @xcite @xcite @xcite : [ t : dasm ] assume * ( b0)**(b2 ) * and * ( a3w ) * and fix @xmath247 . if @xmath248 \longmapsto -d_x c({{\tilde x}},y(t))$ ] is a line segment then @xmath249 for all @xmath250 $ ] .
it is through this theorem and the next that hypothesis * ( a3w ) * and the non - negative cross - curvature hypothesis * ( b3)*enter crucially . among the many corollaries loeper deduced from this result
, we shall need two .
proved in ( * ? ? ?
* theorem 3.1 and proposition 4.4 ) ( alternately ( * ? ? ?
* theorem 3.1 ) and ( * ? ? ? * a.10 ) ) , they include the @xmath7-convexity of the so - called _ contact set _ ( meaning the @xmath251-subdifferential at a point ) , and a local to global principle .
[ c : local - global ] assume * ( b0)**(b2 ) * and * ( a3w ) * and fix @xmath252 .
if @xmath9 is @xmath7-convex then @xmath253 is @xmath251-convex with respect to @xmath254 , i.e. @xmath255 forms a convex subset of @xmath256 .
furthermore , any local minimum of the map @xmath257 is a global minimum . as shown in (
* corollary 2.11 ) , the strengthening * ( b3 ) * of hypothesis * ( a3w ) * improves the conclusion of loeper s maximum principle .
this improvement asserts that the altitude @xmath258 at each point of the evolving landscape then accelerates as a function of @xmath248 $ ] : [ t : time - convex dasm ] assume * ( b0)**(b3 ) * and fix @xmath247 . if @xmath248 \longmapsto -d_xc({{\tilde x}},y(t))$ ] is a line segment then @xmath250 \longmapsto f(t ) : = -c(x , y(t ) ) + c({{\tilde x}},y(t))$ ] is convex .
[ rmk : symmetry]since all assumptions * ( b0)**(b3 ) * and * ( a3w ) * on the cost are symmetric in @xmath0 and @xmath1 , all the results above still hold when exchanging @xmath0 with @xmath1 .
in this section , we set up the notation for the rest of the paper . recall that @xmath259 is a non - negatively cross - curved cost function satisfying * ( b1)**(b3 ) * on a pair of bounded domains @xmath75 and @xmath88 which are strongly @xmath7-convex with respect to each other * ( b2u)*. fix @xmath260 and an open domain @xmath261 , and let @xmath27 be a @xmath7-convex solution of the @xmath7-monge - ampre equation @xmath262 we sometimes abbreviate by writing @xmath263 $ ] . in the following sections , we will prove interior differentiability of @xmath27 on @xmath264 , that is @xmath265 ; see theorem [ t : continuity ]
. throughout @xmath266 will denote the derivative with respect to the variable @xmath1 , and iterated subscripts as in @xmath267 denote iterated derivatives .
we also use @xmath268 to denote the bi - lipschitz constants @xmath269 of the coordinate changes ( [ q of x ] ) and the jacobian bounds @xmath270 for the same transformation .
notice @xmath271 for any cost satisfying * ( b1 ) * , and equality holds whenever the cost function @xmath2 is quadratic .
so the parameter @xmath272 crudely quantifies the departure from the quadratic case .
the inequality @xmath273 is much more rigid , equality implying @xmath274 is the identity matrix , and not merely constant . in the current subsection
, we introduce an important transformation ( mixing dependent and independent variables ) for the cost @xmath2 and potential @xmath60 , which plays a crucial role in the subsequent analysis .
this change of variables and its most relevant properties are encapsulated in the following definition and theorem .
_ in the sequel , whenever we use the expression @xmath275 or @xmath276 we refer to the modified cost function and convex potential defined here , unless otherwise stated .
_ since properties * ( b0)**(b3 ) * , * ( a3w )
* and * ( b2u ) * were shown to be tensorial in nature ( i.e. coordinate independent ) in @xcite @xcite , the modified cost @xmath277 inherits these properties from the original cost @xmath7 with one exception : ( [ x of q ] ) defines a @xmath278 diffeomorphism @xmath279 , so the cost @xmath280 may not be @xmath281 smooth .
however , its definition reveals that we may still differentiate @xmath277 four times as long as no more than three of the four derivatives fall on the variable @xmath282 , and it leads to the same geometrical structure ( pseudo - riemannian curvatures , including ( [ mtw ] ) ) as the original cost @xmath7 since the metric tensor and symplectic form defined in @xcite involve only mixed derivatives @xmath283 , and therefore remain @xmath284 functions of the coordinates @xmath285 .
[ d : cost exponential ] given @xmath64 strongly twisted * ( b0)**(b1 ) * , we refer to the coordinates @xmath286 defined by @xmath287 as the _ cost exponential coordinates _ from @xmath288 and @xmath93 respectively .
we denote the inverse diffeomorphisms by @xmath289 and @xmath290 ; they satisfy @xmath291 the cost @xmath292 is called the _ modified cost at @xmath89_. a subset of @xmath80 or function thereon is said to _ appear from @xmath89 _ to have property @xmath293 , if it has property @xmath293 when expressed in the coordinates @xmath294 . identifying the cotangent vector @xmath295 with the tangent vector @xmath296 to @xmath87 using the pseudo - metric of kim and
mccann @xcite shows @xmath297 to be the projection to @xmath75 of the pseudo - riemannian exponential map @xmath298 ; similarly @xmath299 is the projection to @xmath88 of @xmath300 . also , @xmath301-@xmath302 and @xmath303-@xmath304 in the notation of loeper @xcite .
our first contribution is the following theorem . for a non - negatively cross - curved cost * ( b3 )
* , it shows that any @xmath277-convex potential appears convex from @xmath288 .
even if the cost function is weakly regular * ( a3w ) * , the level sets of the @xmath277-convex potential appear convex from @xmath89 , as was discovered independently from us by liu @xcite , and exploited by liu with trudinger and wang @xcite .
note that although the difference between the cost @xmath2 and the modified cost @xmath305 depends on @xmath89 , they differ by a null lagrangian @xmath306 which being independent of @xmath307 does not affect the question of which maps @xmath25 attain the infimum ( [ monge ] ) . having a function with convex level sets is a useful starting point , since it enables us to apply caffarelli s affine renormalization of convex sets approach and a full range of techniques from gutierrez @xcite to address the regularity of @xmath7-convex potentials .
[ thm : apparently convex ] let @xmath97 satisfying * ( b0)**(b2 ) * be weakly regular * ( a3w)*. if @xmath308 is @xmath7-convex on @xmath80 , then @xmath309 has convex level sets , as a function of the cost exponential coordinates @xmath310 from @xmath288 . if , in addition , @xmath7 is non - negatively cross - curved * ( b3 ) * then @xmath311 is convex on @xmath312 . in either case
@xmath311 is minimized at @xmath313 if @xmath314 .
furthermore , @xmath311 is @xmath277-convex with respect to the modified cost @xmath315 on @xmath316 , and @xmath317 for all @xmath318 .
the final sentences of the theorem are elementary : @xmath7-convexity @xmath137 asserts @xmath319from ( [ c - transform ] ) , hence @xmath320 and @xmath317 since all three suprema above are attained at the same @xmath321 .
taking @xmath322 reduces the inequality @xmath323 to @xmath324 , with equality precisely if @xmath325 .
it remains to address the convexity claims .
since the supremum @xmath326 of a family of convex functions is again convex , it suffices to establish the convexity of @xmath327 for each @xmath321 under hypothesis * ( b3)*. for a similar reason , it suffices to establish the level - set convexity of the same family of functions under hypothesis * ( a3w)*. first assume * ( a3w)*. since @xmath328 we see that @xmath277-segments in @xmath312 with respect to @xmath89 coincide with ordinary line segments .
let @xmath329 be any line segment in the convex set @xmath312 .
define @xmath330 .
loeper s maximum principal ( theorem [ t : dasm ] above , see also remark [ rmk : symmetry ] ) asserts @xmath331 , which implies convexity of each set @xmath332 . under hypothesis
* ( b3 ) * , theorem [ t : time - convex dasm ] goes on to assert convexity of @xmath91 \longmapsto f(s , y)$ ] as desired . the effect of this change of gauge on jacobian inequalities is summarized in a corollary : [ c : jacobian transform ] using the hypotheses and notation of theorem [ thm : apparently convex ] , if @xmath333 \subset [ 0,\infty]$ ] on @xmath176 , then @xmath334 $ ] on @xmath335 , where @xmath336 and @xmath337 are defined in . furthermore , @xmath338 and @xmath339 . from the jacobian bounds
@xmath340 $ ] on @xmath8 , we find @xmath341 for each @xmath150 . on the other hand , theorem [ thm : apparently convex ] asserts @xmath342 , so the claim @xmath334 $ ] follows from the hypothesis @xmath333 $ ] , by definition ( [ c - monge - ampere measure ] ) and the fact that @xmath343 from is a diffeomorphism ; see * ( b1)*. the bounds @xmath344 and @xmath345 follow from @xmath346 and @xmath347 .
the renormalization of a function @xmath311 by an affine transformation @xmath348 will be useful in section [ s : alexandrov ] to prove our alexandrov type estimates .
let us therefore record the following observations .
define @xmath349 here @xmath350 denotes the jacobian determinant of @xmath351 , i.e. the determinant of the linear part of @xmath351 . assuming * ( b0)**(b1 )
* , given a @xmath352-convex function @xmath353 and affine bijection @xmath354 , define the renormalized potential @xmath355 by and renormalized cost @xmath356 using the adjoint @xmath357 to the linear part of @xmath351 .
then , for all borel @xmath358 , @xmath359 from we see @xmath360 if and only if @xmath361 , thus follows from @xmath362 .
similarly , since yields @xmath363 , we see @xmath364 is equivalent to @xmath365 from ( and theorem [ thm : apparently convex ] ) , whence @xmath366 to establish . as a corollary to this lemma
, we recover the affine invariance not only of the monge - ampre equation satisfied by @xmath326 but also of the @xmath352-monge - ampre equation it satisfies under coordinate changes on @xmath88 ( which induce linear transformations @xmath351 on @xmath367 and @xmath357 on @xmath368 ) : for @xmath369 , @xmath370
the subsequent sections of this paper are largely devoted to ruling out exposed points in @xmath69 of sets on which ordinary convexity of the @xmath352-convex potential from theorem [ thm : apparently convex ] fails to be strict .
this current section rules out exposed points on the boundary of @xmath69 .
we do this by proving an important topological property of the ( multi - valued ) mapping @xmath174 .
namely , we show that the subdifferential @xmath215 maps interior points of @xmath371 only to interior points of @xmath88 , under hypothesis , and conversely that @xmath215 maps boundary points of @xmath75 only to boundary points of @xmath88 . this theorem may be of independent interest , and was required by figalli and loeper to conclude their continuity result concerning maps of the plane which optimize * ( a3w ) * costs @xcite
. this section does not use the cross - curvature condition * ( b3)*(nor * a3w * ) on the cost function @xmath259 , but relies crucially on the _ strong _ @xmath7-convexity * ( b2u ) * of its domains @xmath72 and @xmath372 ( but importantly , not on @xmath373 ) .
no analog for theorem [ thm : bdry - inter ] was needed by caffarelli to establish @xmath374 regularity of convex potentials @xmath60 whose gradients optimize the classical cost @xmath375 @xcite , since in that case he was able to take advantage of the fact that the cost function is smooth on the whole of @xmath16 to chase potentially singular behaviour to infinity .
( one general approach to showing regularity of solutions for degenerate elliptic partial differential equations is to exploit the threshold - hyperbolic nature of the solution to try to follow either its singularities or its degeneracies to the boundary , where they can hopefully be shown to be in contradiction with boundary conditions ; the _ degenerate _ nature of the ellipticity precludes the possibility of _ purely local _ regularizing effects . ) [ thm : bdry - inter ] let @xmath7 satisfy * ( b0)**(b1 ) * and @xmath137 be a @xmath7-convex function ( which implies @xmath376 ) , and @xmath377 . 1 .
if @xmath378 on @xmath379 and @xmath88 is strongly @xmath251-convex with respect to @xmath380 , then interior points of @xmath380 can not be mapped by @xmath381 to boundary points of @xmath88 : i.e. @xmath382 .
2 . if @xmath383 on @xmath80 , and @xmath75 is strongly @xmath7-convex with respect to @xmath88 , then boundary points of @xmath75 can not be mapped by @xmath215 into interior points of @xmath88 : i.e. @xmath384 is disjoint from @xmath215 .
note that when @xmath380 is open the conclusion of ( a ) implies @xmath385 is disjoint from @xmath386 .
we therefore remark that it suffices to prove ( a ) , since ( b ) follows from ( a ) exchanging the role @xmath0 and @xmath1 and observing that @xmath383 implies @xmath387 as in lemma [ l : cma properties](d ) .
let us prove ( a ) . fix any point @xmath388 in the interior of @xmath380 , and @xmath389 .
assume by contradiction that @xmath390 .
at @xmath391 we use * ( b0)**(b1 ) * to define cost - exponential coordinates @xmath392 by @xmath393 and define a modified cost and potential by subtracting null lagrangian terms : @xmath394 similarly to corollary [ c : jacobian transform ] , @xmath395 , where @xmath270 denote the jacobian bounds for the coordinate change .
note @xmath396 corresponds to @xmath397 . since @xmath7-segments with respect to @xmath398
correspond to line segments in @xmath399 we see @xmath400 depends linearly on @xmath282 , whence @xmath401 ; similarly @xmath251-segments with respect to @xmath388 become line segments in the @xmath402 variables , @xmath403 depends linearly on @xmath402 , @xmath404 , and the extra factor @xmath405 in our definition of @xmath297 makes @xmath406 the identity matrix ( whence @xmath407 and @xmath408 for all @xmath282 in @xmath409 and @xmath402 in @xmath410 .
although the change of variables @xmath411 is only a @xmath278 diffeomorphism , we can still take four derivatives of the modified cost provided at least one of the four derivatives is with respect to @xmath282 and another is with respect to @xmath402 .
we denote @xmath412 and choose orthogonal coordinates on @xmath72 which make @xmath413 the outer unit normal to @xmath414 at @xmath415 .
note that @xmath416 is strongly convex by hypothesis ( a ) . in these variables , consider a small cone of height @xmath417 and angle @xmath418 around the @xmath413 axis : @xmath419 observe that , if @xmath420 are small enough , then @xmath421 , and its measure is of order @xmath422 .
consider now a slight enlargement @xmath423 of the polar dual cone , where @xmath417 will be chosen sufficiently small depending on the large parameter @xmath424 forced on us later .
the strong convexity ensures @xmath70 is contained in a ball @xmath425 of some radius @xmath426 contained in the half - space @xmath427 with boundary sphere passing through the origin .
as long as @xmath428 we claim @xmath429 intersects this ball a fortiori @xmath70 in a set whose volume tends to zero like @xmath430 as @xmath431 . indeed , from the inequality @xmath432satisfied by any @xmath433 we deduce @xmath434 , i.e. @xmath435
if @xmath418 is small enough . combined with the further inequalities @xmath436 ( the first inequality follows by the strong convexity of @xmath70 ) , this yields @xmath437 and @xmath438 as @xmath431 .
thus @xmath439 for a dimension dependent constant @xmath440 , provided @xmath441 .
the contradiction now will come from the fact that , thanks to the @xmath277-cyclical monotonicity of @xmath442 , if we first choose @xmath424 big and then we take @xmath417 sufficiently small , the image of all @xmath443 by @xmath444 has to be contained in @xmath429 for @xmath418 small enough . since @xmath445 this will imply latexmath:[\[{\varepsilon}^n\theta^{n-1 } \sim \tilde \lambda { { \mathscr l}^{n } } ( e_{\theta,{\varepsilon } } ) \leq which gives a contradiction as @xmath431 , for @xmath447 small but fixed .
thus all we need to prove is that , if @xmath424 is big enough , then @xmath448 for any @xmath417 sufficiently small .
let @xmath443 and @xmath449 .
combining @xmath450 = \tilde c(q , p ) + \tilde c({{\mathbf 0}},{{\mathbf 0}})- \tilde c(q,{{\mathbf 0}})-c({{\mathbf 0}},p ) \le 0\end{aligned}\ ] ] ( where the last inequality is a consequence of @xmath451-monotonicity of @xmath452 ; see for instance ( * ? ? ?
* definitions 5.1 and 5.7 ) ) with @xmath453 \\ = & d^2_{qp } \tilde c({{\mathbf 0}},{{\mathbf 0 } } ) + \int_0^t dt ' d^3_{qpp } \tilde c({{\mathbf 0}},t'p)[p ] \\ &
+ \int_0^s ds ' d^3_{qqp } \tilde c(s'q,{{\mathbf 0}})[q ] + \int_0^s ds ' \int_0^t dt ' d^4_{qqpp } \tilde c(s'q , t'p)[q , p]\end{aligned}\ ] ] yields @xmath454 \\ & \le c_0 |q|^2 |p|^2\end{aligned}\ ] ] since @xmath455 and @xmath456 vanish in our chosen coordinates and @xmath457 is the identity matrix . due to the tensorial nature of the cross - curvature ,
@xmath424 depends on @xmath458 and the bi - lipschitz constants @xmath269 from . from the above inequality and the definition of @xmath459 we deduce
@xmath460 so @xmath461 as desired .
in this section we shall prove that up to constants the ordinary monge - ampre measure @xmath462 dominates the @xmath352-monge - ampre measure @xmath463 , when defined in the coordinates introduced in theorem [ thm : apparently convex ] .
let us begin with a lemma which motivates our proposition heuristically .
the conclusions of the lemma extend easily from smooth to non - smooth functions by an approximation argument combining lemma [ l : cma properties](b)(c ) with results of trudinger and wang @xcite .
however this approach would require the domains @xmath75 and @xmath88 to be smooth , so in proposition [ p : c - ma less ma ] we prefer to construct an explicit approximation which proves the statement we need , requires no additional smoothness hypotheses , and is logically independent of both lemma [ lem : c - ma less ma ] and @xcite .
[ lem : c - ma less ma ] assume * ( b0)**(b3 ) * and let @xmath464 be a convex @xmath352-convex function as in theorem [ thm : apparently convex ] . if @xmath465 for some open set @xmath466 , then @xmath467 on @xmath468 , where @xmath469 are defined as in ( [ jacobian bound ] ) .
in addition to the convexity of @xmath276 , for any @xmath321 theorem [ thm : apparently convex ] asserts the convexity of the @xmath277-convex function @xmath470 .
thus @xmath471by the concavity of @xmath472 on symmetric non - negative definite matrices . on the other hand , for @xmath277-convex @xmath473 ,
the measure @xmath474 is absolutely continuous , with lebesgue density given by the left hand side of ( [ monge - ampere type equation ] ) .
thus at any @xmath475 , @xmath476 as desired .
we now prove the proposition that we actually need subsequently .
[ p : c - ma less ma ] assume * ( b0)**(b3 ) * , and let @xmath464 be a convex @xmath352-convex function from theorem [ thm : apparently convex ] .
then @xmath467 on @xmath477 , where @xmath469 are defined as in ( [ jacobian bound ] ) .
it suffices to prove @xmath478 for each ball whose closure is contained in @xmath468 .
given such a ball , let @xmath479 , and let @xmath480 be a smooth mollifier vanishing outside @xmath481 and carrying unit mass . for @xmath447 sufficiently small , we can define the smooth convex function @xmath482 on @xmath483 . since @xmath311 and @xmath484 are locally lipschitz , letting @xmath74 denote a bound for @xmath485 inside @xmath486 yields @xmath487 for all @xmath488 .
_ claim : _ fix @xmath489 and @xmath490 . for @xmath447
sufficiently small , we claim that to each @xmath491 corresponds some @xmath492 such that @xmath493 .
indeed , for @xmath494 with @xmath495 , observe @xmath496 with equality at @xmath313 .
moreover @xmath497 .
therefore @xmath498 on the other hand , if @xmath499 , then @xmath500 , and so @xmath501 thus , if for fixed @xmath502 small we choose @xmath417 small enough so that @xmath503 we deduce that if we lower the graph of the function @xmath504 to the lowest level at which it intersects the graph of @xmath505 , then the point of intersection must lie over @xmath506 .
this proves the claim .
having established the claim , let @xmath507 denote the ( borel ) set of all @xmath508 which arise from @xmath509 in this way .
since @xmath505 is smooth , the condition @xmath510 implies @xmath511 as well as @xmath512 by and * ( b1 ) * we can define a smooth map @xmath513 throughout @xmath514 using the relation @xmath515 and find that @xmath516 is a singleton . in this way we obtain @xmath517 where in the last equality we used to deduce that @xmath518 is non - negative definite . hence the inequality @xmath519 holds ( similarly to above ) , and
so @xmath520 letting first @xmath521 and then @xmath522 , we finally deduce @xmath523 here , to see the last inequality one may , for instance , use lemma [ l : cma properties](b ) with @xmath524 .
arbitrariness of @xmath525 yields the desired result .
in this section we prove the key estimates for @xmath7-convex potential functions which will eventually lead to the continuity and injectivity of optimal maps .
namely , we extend alexandrov type estimates commonly used in the analysis of monge - ampre equations ( thus for the cost @xmath526 ) , to general non - negatively curved cost functions .
this is established in lemma [ lemma - infgeq ] ( plus proposition [ p : c - ma less ma ] ) and lemma [ l : prop - infleq ] .
these estimates are used to compare the infimum of @xmath7-convex function on a section with the size of the section , which are the key ingredients in the proof of our main results ; see propositions [ prop : alex1 ] and [ prop : estimate ] .
lemma [ l : prop - infleq ] represents the most nontrivial and technical result we obtain in this section .
we recall a basic lemma for convex sets due to fritz john @xcite , which will play an essential role in the rest of the paper .
[ l : john ] for a compact convex set @xmath527 , there exists an affine transformation @xmath528 such that @xmath529 .
we now estimate the infimum of @xmath311 in terms of the monge - ampre measure in a section .
the following lemma is a standard fact for convex functions .
with lemma [ l : john ] in mind , we state it for normalized functions @xmath355 and sections @xmath530 .
however , the estimate ( [ inf geq ] ) is invariant under the affine renormalization ; according to , it holds with or without stars .
[ lemma - infgeq ] let @xmath531 be a convex function whose section @xmath532 satisfies @xmath533 .
assume that @xmath534 on @xmath535 .
then , for all @xmath536 , @xmath537 where @xmath538 denotes the dilation of @xmath530 by a factor @xmath539 with respect to the origin .
although the proof of this result is classical ( see for instance @xcite ) , for sake of completeness we prefer to give all the details .
we can assume that @xmath540 , otherwise the estimate is trivial .
it is not difficult to prove that @xmath541 indeed , if @xmath542 and @xmath543 , then @xmath544 and taking the supremum in the left hand side among all @xmath545 follows .
thus , since @xmath546 , we conclude @xmath547 combining the above lemmas , we obtain : [ prop : alex1 ] assume * ( b0)**(b3 ) * and define @xmath548 as in ( [ jacobian bound ] ) . any convex @xmath277-convex function @xmath549 from theorem [ thm : apparently convex ] which satisfies @xmath550 $ ] in a section of the form @xmath551 and @xmath552 on @xmath553 , also satisfies @xmath554 first use the affine map @xmath351 as given in lemma [ l : john ] to renormalize @xmath311 into @xmath355 using .
this does not change the bound @xmath555 $ ] , but allows us to apply lemma [ lemma - infgeq ] with @xmath556 .
its conclusion ( [ inf geq ] ) has been expressed in a form which holds with or without the stars , in view of .
proposition [ p : c - ma less ma ] now yields the desired inequality .
we now progress toward the alexandrov type estimate in lemma [ l : prop - infleq ] . in this subsection
we construct and study the @xmath352-cone associated to the section of a @xmath352-convex function .
this @xmath352-cone plays an essential role in our proof of lemma [ l : prop - infleq ] .
[ def : ccone ] assume * ( b0)**(b2 ) * and * ( a3w ) * , and let @xmath549 be the @xmath277-convex function with convex level sets from theorem [ thm : apparently convex ] .
let @xmath557 denote the section @xmath558 , fix @xmath559 , and assume @xmath552 on @xmath560 .
@xmath352-cone @xmath561 generated by @xmath562 and @xmath557 with height @xmath563 _ is given by @xmath564 notice the @xmath352-cone @xmath565 depends only on the convex set @xmath358 , @xmath566 , and the value @xmath567 , but is otherwise independent of @xmath568 .
recalling that @xmath569 on @xmath570 , we record several key properties of the @xmath352-cone : [ lemma : propccone ] adopting the notation and hypotheses of definition [ def : ccone ] , let @xmath571 be the @xmath352-cone generated by @xmath572 and @xmath557 with height @xmath573 . then 1 .
@xmath565 has convex level sets ; it is a convex function if * ( b3 ) * holds ; 2 .
@xmath574 for all @xmath575 ; 3 .
@xmath576 ; 4 .
@xmath577 .
property ( a ) is a consequence of the level - set convexity of @xmath578 proved in theorem [ thm : apparently convex ] , or its convexity assuming * ( b3)*. moreover , since @xmath579 for all @xmath580 , ( b ) follows . for each pair @xmath581 and @xmath582 , consider the supporting mountain @xmath583 , i.e. @xmath584 and @xmath585 . consider the @xmath352-segment @xmath586 connecting @xmath587 and @xmath588 in @xmath88 with respect to @xmath589 .
since @xmath590 , by continuity there exists some @xmath5910 , 1]$ ] for which @xmath592 satisfies @xmath593 . from loeper s maximum principle ( theorem [ t : dasm ] above ) , we have @xmath594= \max[m_z , 0],\ ] ] and therefore , from @xmath585 , @xmath595 by the construction , @xmath596 is of the form @xmath597 and vanishes at @xmath589 .
this proves ( c ) .
finally ( d ) follows from ( c ) and the fact that @xmath598 .
indeed , it suffices to move down the supporting mountain of @xmath565 at @xmath572 until the last moment at which it touches the graph of @xmath311 on @xmath557 from below .
the conclusion then follows from loeper s local to global principle , corollary [ c : local - global ] above .
the following estimate shows that the monge - ampre measure , and the relative location of the vertex within the section which generates it , control the height of any well - localized @xmath352-cone .
afficionados of the monge - ampre theory may be less surprised by this estimate once it is recognized that the localization in coordinates ensures the cost is approximately affine , at least in one of its two variables .
still , it is vital that the approximation be controlled !
together with lemma [ lemma : propccone](d ) , this proposition plays a key role in the proof of our alexandrov type estimate ( lemma [ l : prop - infleq ] ) .
[ lower bound on the monge - ampre measure of a small @xmath352-cone ] [ p : partial c - cone ] assume * ( b0)**(b3 ) * and define @xmath599 as in definition [ d : cost exponential ] .
let @xmath358 be a closed convex set and @xmath565 the @xmath352-cone generated by @xmath566 of height @xmath600 over @xmath557 .
let @xmath601 be two parallel hyperplanes contained in @xmath602 and touching @xmath560 from two opposite sides .
then there exists @xmath603 small , depending only on the cost ( and given by lemma [ l : c estimate ] ) , and a constant @xmath604 depending only on dimension , such that if @xmath605 then @xmath606 where @xmath607 denotes the maximal length among all the segments obtained by intersecting @xmath557 with a line orthogonal to @xmath608 . to prove this , we first observe a basic estimate on the cost function @xmath7 .
[ l : c estimate ] assume * ( b0)**(b2)*. for @xmath599 from definition [ d : cost exponential ] and each @xmath321 and @xmath609 , @xmath610 where @xmath611 is given by @xmath612 in the notation . for fixed @xmath613 introduce the @xmath451-exponential coordinates @xmath614 .
the bi - lipschitz constants of this coordinate change are estimated by @xmath615 as in corollary [ c : jacobian transform ] .
thus @xmath616 where @xmath617 from definition [ d : cost exponential ] has been used .
similarly , noting the convexity * ( b2)*of @xmath618 , @xmath619 the result follows since @xmath620 .
( the last inequality follows from @xmath621 . )
we fix @xmath622 .
let @xmath623 , @xmath624 , ( with @xmath625 equal either @xmath626 or @xmath627 ) be hyperplanes contained in @xmath628 , touching @xmath560 , and such that @xmath629 are all mutually orthogonal ( so that also @xmath630 are mutually orthogonal ) .
moreover we choose @xmath631 in such a way that , if @xmath632 denotes the projection of @xmath557 on @xmath625 and @xmath633 denotes its @xmath634-dimensional hausdorff measure , then @xmath635 for some universal constant @xmath636 .
indeed , as @xmath632 is convex , by lemma [ l : john ] we can find an ellipsoid @xmath514 such that @xmath637 , and for instance we can choose @xmath631 among the hyperplanes orthogonal to the axes of the ellipsoid ( for each axis we have two possible hyperplanes , and we can always choose the furthest one so that holds ) .
each hyperplane @xmath638 touches @xmath557 from outside , say at @xmath639 .
let @xmath640 be the outward ( from @xmath557 ) unit vector at @xmath641 orthogonal to @xmath623 .
then @xmath642 for some @xmath643 , and by corollary [ c : local - global ] there exists @xmath644 such that @xmath645 define @xmath646 as @xmath647 i.e. @xmath646 is the @xmath352-segment from @xmath89 to @xmath648 with respect to @xmath641 . as in the proof of lemma [ lemma : propccone ] ( c ) ,
the intermediate value theorem yields @xmath649 such that @xmath650 with equality at @xmath641 .
thus , by the definition of @xmath565 , @xmath651 , @xmath652 therefore by the convexity of @xmath565 shown in lemma [ lemma : propccone](a ) , the affine function @xmath653 with slope @xmath654 and with @xmath655 satisfies @xmath656 .
this shows @xmath657 therefore if @xmath658 with @xmath659 small , each vector @xmath660 is close to @xmath654 , say @xmath661 since the vectors @xmath662 are mutually orthogonal , the above estimate implies that for @xmath663 small enough the convex hull of @xmath664 has measure of order @xmath665 .
thus , by the lower bound and the convexity of @xmath666 , we obtain @xmath667 since @xmath625 was either @xmath608 or @xmath627 , we have proved that @xmath668 to conclude the proof , we apply lemma [ lemma : orthogonal sections ] below with @xmath669 given by the segment obtained intersecting @xmath557 with a line orthogonal to @xmath608 . combining that lemma with , we obtain @xmath670 and last two inequalities prove the proposition ( taking @xmath671 larger if necessary ) .
[ lemma : orthogonal sections ] let @xmath557 be a convex set in @xmath672 .
let @xmath673 denote the projections to the components @xmath674 , @xmath675 , respectively .
let @xmath669 be a slice orthogonal to the second component , that is @xmath676 then there exists a constant @xmath636 , depending only on @xmath677 , such that @xmath678 where @xmath679 denotes the @xmath680-dimensional hausdorff measure .
let @xmath681 be an affine map with determinant @xmath229 given by lemma [ l : john ] such that @xmath682 for some @xmath683 .
then , if we extend @xmath351 to the whole @xmath16 as @xmath684 , we have @xmath685 , @xmath686 , and @xmath687 hence , we can assume from the beginning that @xmath688 .
let us now consider the point @xmath689 , and we fix an orthonormal basis @xmath690 in @xmath691 such that @xmath692 for some @xmath693 .
since @xmath694 , there exist points @xmath695 such that @xmath696 .
let @xmath697 denote the convex hull of @xmath669 with @xmath698 , and let @xmath699 denote the @xmath700-dimensional strip obtained taking the convex hull of @xmath701 with @xmath698 .
observe that @xmath702 , and so @xmath703 we now remark that , since @xmath696 and @xmath704 for @xmath705 , we have @xmath706 for all @xmath705 . moreover , if @xmath707 denotes the closest point to @xmath708 , then the segments joining @xmath708 to @xmath648 parallels @xmath709 , hence these segments are all mutually orthogonal , and they are all orthogonal to @xmath699 too . from this fact it is easy to see that , if we define the convex hull @xmath710 then , since @xmath711 for @xmath705 , by and the inclusion @xmath712 we get @xmath713 this concludes the proof , as @xmath714 .
the next alexandrov type lemma holds for localized sections @xmath557 of @xmath352-convex functions .
[ l : prop - infleq ] assume * ( b0)**(b3 ) * and let @xmath549 be a convex @xmath277-convex function from theorem [ thm : apparently convex ] .
let @xmath557 denote the section @xmath558 , assume @xmath715 on @xmath553 , and fix @xmath716 .
let @xmath601 be two parallel hyperplanes contained in @xmath628 and touching @xmath560 from two opposite sides .
then there exists @xmath717 small , depending only on dimension and the cost function ( with @xmath718 given by proposition [ p : partial c - cone ] ) such that if @xmath719 then @xmath720 where @xmath607 denotes the maximal length among all the segments obtained by intersecting @xmath557 with a line orthogonal to @xmath608 , and @xmath721 is defined as in ( [ jacobian bound ] ) .
fix @xmath622 .
observe that @xmath552 on @xmath560 and consider the @xmath352-cone @xmath565 generated by @xmath572 and @xmath557 of height @xmath722 as in .
from lemma [ lemma : propccone](d ) we have @xmath723 and from loeper s local to global principle , corollary [ c : local - global ] above , @xmath724 therefore @xmath725 the lower bound on @xmath726 comes from .
this finishes the proof . combining the results of proposition [ prop : alex1 ] and lemma [ l : prop - infleq ] yields : [ prop : estimate ]
assume * ( b0)**(b3 ) * and let @xmath549 be a convex @xmath277-convex function from theorem [ thm : apparently convex ] .
there exists @xmath717 small , depending only on dimension and the cost function ( and given by lemma [ l : prop - infleq ] ) , and constants @xmath636 , @xmath728 , @xmath729 , depending only on dimension , such that the following holds : letting @xmath557 denote the section @xmath558 , assume @xmath730 $ ] in @xmath557 and @xmath552 on @xmath560 .
let @xmath731 be parallel hyperplanes contained in @xmath732 and supporting @xmath557 from two opposite sides . if @xmath733 then @xmath734 and @xmath735 where @xmath607 denotes the maximal length among all the segments obtained by intersecting @xmath557 with a line orthogonal to @xmath608 , and @xmath721 is defined as in ( [ jacobian bound ] ) .
equation follows from lemma [ l : prop - infleq ] and the assumption @xmath736 .
now , by lemma [ l : john ] , we deduce that there exists an ellipsoid @xmath514 such that @xmath737 , where @xmath738 denotes the dilation of @xmath514 by a factor @xmath739 with respect to its barycenter @xmath740 . taking @xmath608 and
@xmath627 orthogonal to one of the longest axes of @xmath514 and @xmath741 in yields @xmath742 on the other hand , convexity of @xmath311 along the segment which crosses @xmath557 and passes through both @xmath740 and the point @xmath562 where @xmath743 is attained implies @xmath744 since the barycenter of @xmath514 divides the segment into a ratio at most @xmath745 .
combining these two estimates with we obtain , to complete the proof .
noting @xmath746 from corollary [ c : jacobian transform ] , we observe that the estimate is stable under affine renormalization : let @xmath351 be an affine transformation , and recall the renormalization @xmath747 then @xmath748 is a section for @xmath355 and @xmath749 on the other hand , estimate is not stable under affine renormalization ( a line orthogonal to @xmath608 is not an affinely invariant concept ) . for this reason ,
both in the proof of the @xmath7-strict convexity ( section [ s : contact ] ) and in the proof of differentiability @xmath750 ( section [ s : continuousinjection ] ) we apply our alexandrov estimates directly to the original sections , without renormalizing them . using this strategy ,
our estimates turn out to be strong enough to adapt to our situation the strict convexity and interior continuity theory of caffarelli @xcite @xcite .
we perform this in the remainder of the manuscript .
in this section and the final one , we complete the crucial step of proving the strict @xmath7-convexity of the @xmath7-convex potentials @xmath751 arising in optimal transport , meaning @xmath752 should be disjoint from @xmath253 whenever @xmath753 are distinct .
this is accomplished in theorem [ t : strict convex ] . in this section ,
we show that , if the contact set does not consist of a single point , then it extends to the boundary of @xmath75 .
our method relies on the non - negative cross - curvature * ( b3 ) * of the cost @xmath7 . from now on
we adopt the following notation : @xmath754 means that there exist two positive constants @xmath755 and @xmath756 , depending on @xmath739 and @xmath757 only , such that @xmath758 .
analogously we will say that @xmath759 ( resp .
@xmath760 ) if there exists a positive constant @xmath440 , depending on @xmath739 and @xmath757 only , such that @xmath761 ( resp .
@xmath762 ) . recall that a point @xmath0 of a convex set @xmath527 is _ exposed _ if there is a hyperplane supporting @xmath763 exclusively at @xmath0 .
although the _ contact set _
@xmath764 may not be convex , it appears convex from @xmath89 by corollary [ c : local - global ] , meaning its image @xmath765 in the coordinates is convex .
the following theorem shows this convex set is either a singleton , or contains a segment which stretches across the domain .
we prove it by showing the solution geometry near certain exposed points of @xmath766 inside @xmath69 would be inconsistent with the bounds established in the previous section .
[ thm : noexposed ] assume * ( b0)**(b3 ) * and let @xmath27 be a @xmath7-convex solution of with @xmath767 open .
fix @xmath768 and @xmath769 , and define the contact set @xmath770 .
assume that @xmath771 , i.e. it is not a singleton .
then @xmath763 intersects @xmath772 . as in definition
[ d : cost exponential ] , we transform @xmath773 with respect to @xmath89 , i.e. we consider the transformation @xmath774 , defined on @xmath775 by the relation @xmath776 and the modified cost function @xmath777 on @xmath778 , for which the @xmath352-convex potential function @xmath779 is convex . we observe that @xmath569 for all @xmath282 , and moreover the set @xmath780 appears convex from @xmath89 , meaning @xmath781 is convex , by the corollary [ c : local - global ] to loeper s maximum principle .
our proof is reminiscent of caffarelli s for the cost @xmath782 ( * ? ? ?
* lemma 3 ) .
observe @xmath783 lies in the interior of the set @xmath784 where @xmath785 $ ] , according to corollary [ c : jacobian transform ] .
choose the point @xmath786 furthest from @xmath562 ; it is an exposed point of @xmath787 .
we claim either @xmath788 or @xmath789 . to derive a contradiction ,
suppose the preceding claim fails , meaning @xmath790
. for a suitable choice of cartesian coordinates on @xmath88 we may , without loss of generality , take @xmath791 parallel to the positive @xmath792 axis .
denote by @xmath709 the associated orthogonal basis for @xmath793 , and set @xmath794 and @xmath795 , so the halfspace @xmath796 of @xmath797 intersects @xmath787 only at @xmath798 .
use the fact that @xmath799 is an exposed point of @xmath787 to cut a corner @xmath800 off the contact set @xmath763 by choosing @xmath801 small enough that @xmath802 satisfies : 1 .
@xmath803 is a compact convex set in the interior of @xmath69 ; 2
. @xmath804 , where @xmath805 is from lemma [ l : prop - infleq ] .
defining @xmath806 , @xmath807 the corresponding @xmath7-segment with respect to @xmath89 , and @xmath808 , note that @xmath809 contains @xmath810 , and @xmath800 contains @xmath811 and @xmath812 . since the corner @xmath800 needs not intersect the support of @xmath813 ( especially , when @xmath799 is not an interior point of @xmath814 ) , we shall need to cut a larger corner @xmath815 as well , defined by @xmath816 , which intersects @xmath817 at @xmath78 . by tilting the supporting function slightly
, we shall now define sections @xmath818 of @xmath9 whose interiors include the extreme point @xmath812 and whose boundaries pass through @xmath819 and @xmath78 respectively , but which converge to @xmath800 and @xmath815 respectively as @xmath521 . indeed , set @xmath820 and observe @xmath821 taking @xmath822 in this formula and @xmath447 shows the sections defined by @xmath823 both include a neighbourhood of @xmath824 but converge to @xmath800 and @xmath815 respectively as @xmath521 .
we remark that there exist a priori no coordinates in which all set @xmath825 are convex .
however for each fixed @xmath447 , we can change coordinates so that both @xmath825 and @xmath826 become convex : use @xmath827 to make the transformations @xmath828 so that the functions @xmath829 are convex on @xmath830 .
observe that , in these coordinates , @xmath825 and @xmath826 become convex : @xmath831 and either @xmath832 or @xmath833 since @xmath834 .
for @xmath447 small , the inclusion must be the first of the two since the limits satisfy @xmath835 and @xmath836 .
in the new coordinates , our original point @xmath768 , the exposed point @xmath812 , and the @xmath7-convex combination @xmath819 with respect to @xmath89 , correspond to @xmath837 thanks to ( ii ) , we have @xmath838 for @xmath417 sufficiently small , so that the estimate of lemma [ l : prop - infleq ] applies . in these coordinates ( chosen for each @xmath417 )
we consider the parallel hyperplanes @xmath839 which support @xmath840 from opposite sides and which are orthogonal to the segment joining @xmath841 with @xmath842 .
since @xmath843 paralles the @xmath844 axis the limiting hyperplanes @xmath845 must coincide with @xmath846 and @xmath847 . thus @xmath848 and @xmath849 observing that @xmath850 is shorter than segment obtained intersecting @xmath851 with the line orthogonal to @xmath852 and passing through @xmath853 , lemma [ l : prop - infleq ] combines with @xmath854 and @xmath855 from and corollary [ c : jacobian transform ] to yield @xmath856 on the other hand
, @xmath857 implies @xmath858 , and @xmath859 implies @xmath860 similarly . thus yields
@xmath861 our contradiction with will be established by bounding the ratio @xmath862 away from zero .
recall that @xmath863 and @xmath864 with equality at @xmath812 . from the convergence of @xmath826 to @xmath865 and the asymptotic behaviour of
@xmath866 we get @xmath867}\\ & \ge \frac{- c(x^0 , y_{\varepsilon } ) + c({{\tilde x } } , y_{\varepsilon } ) + c(x^0 , { { \tilde y } } ) - c({{\tilde x } } , { { \tilde y } } ) } { \sup_{x \in k^1_{\varepsilon}}[-c(x , y_{\varepsilon } ) + c({{\tilde x } } , y_{\varepsilon } ) + c(x , { { \tilde y}})- c({{\tilde x } } , { { \tilde y } } ) ] } \\ & \geq \frac{{\varepsilon}(\langle - d_y c(x^0,{{\tilde y } } ) , e_1\rangle - { { \tilde b}})+o({\varepsilon } ) } { { \varepsilon}(\max \{q_1 \mid q \in -d_y c(k^1_{\varepsilon},{{\tilde y}})\ } - { { \tilde b}})+o({\varepsilon } ) } \\ & \geq \frac{1}{2 } \end{split}\ ] ] for @xmath417 sufficiently small .
this shows @xmath868 is close to the minimum value of @xmath869 .
we would like to appeal to lemma [ lemma - infgeq ] to conclude the proof , but are unable to do so since we only have bounds @xmath870 $ ] on the potentially small intersection of @xmath871 with @xmath826 .
however , this intersection occupies a stable fraction of @xmath826 as @xmath521 , which we shall prove as in ( * ? ? ?
* lemma 3 ) .
since @xmath826 converges to @xmath865 for sufficiently small @xmath417 , observe that @xmath826 is uniformly bounded .
therefore the affine transformation @xmath872 that sends @xmath873 to @xmath874 as in lemma [ l : john ] is an expansion , i.e. @xmath875 , with a constant @xmath876 independent of @xmath417 . since @xmath78 is an interior point of @xmath817 , @xmath877 for sufficiently small @xmath878 ,
hence @xmath879 with @xmath880 from
. defining @xmath881 , we have @xmath882 reducing @xmath502 if necessary to ensure @xmath883 , define ( to apply lemma [ lemma - infgeq ] later ) @xmath884 ( as in lemma [ lemma - infgeq ] , @xmath885 denotes the dilation of @xmath886 of a factor @xmath887 with respect to the origin . )
since @xmath886 is convex , it contains the convex hull of @xmath888 , and so @xmath889 for some constant @xmath890 depending on dimension only .
letting @xmath891 this implies @xmath892 recalling that @xmath893 and @xmath894 denote inequalities which hold up to multiplicative constants depending on @xmath895 and @xmath896 , proposition [ p : c - ma less ma ] combines with this estimate to yield @xmath897where ( [ eq : cma ] ) and corollary [ c : jacobian transform ] have been used .
finally , since the conclusion of lemma [ lemma - infgeq ] holds with or without stars in light of , taking @xmath898 in yields @xmath899 since @xmath900 is independent of @xmath417 this contradicts to complete the proof . as can be easily seen from the proof
, one can actually show that if @xmath901 and @xmath763 is not a singleton , then @xmath787 has no exposed points in the interior of @xmath902 . indeed ,
if by contradiction there exists @xmath798 an exposed point of @xmath787 belonging to the interior of @xmath902 , we can choose a point @xmath903 in the interior of @xmath904 such that the segment @xmath791 is orthogonal to a hyperplane supporting @xmath787 at @xmath798 .
then it can immediately checked that the above proof ( which could even be simplified in this particular case ) shows that such a point @xmath798 can not exist .
the first theorem below combines results of sections [ s : mapping ] and [ s : contact ] to deduce strict @xmath7-convexity of the @xmath7-potential for an optimal map , if its target is strongly @xmath7-convex .
this strict @xmath7-convexity which is equivalent to injectivity of the map will then be combined with an adaptation of caffarelli s argument ( * ? ? ?
* corollary 1 ) to obtain interior continuity of the map or equivalently @xmath905-regularity of its @xmath7-potential function for non - negatively cross - curved costs , yielding the concluding theorem of the paper .
[ t : strict convex ] let @xmath7 satisfy * ( b0)**(b3 ) * and * ( b2u)*. if @xmath27 is a @xmath7-convex solution of on @xmath767 open , then @xmath27 is strictly @xmath7-convex on @xmath817 , meaning @xmath906 and @xmath907 are disjoint whenever @xmath753 are distinct .
suppose by contradiction that @xmath908 for two distinct points @xmath753 , and set @xmath909 . according to theorem [ thm : noexposed ] , the set @xmath763 intersects the boundary of @xmath75 at a point @xmath910 .
since asserts latexmath:[${\lambda}\le assume by contradiction that is not . as @xmath916 is compact , one can find a point @xmath918 in the set @xmath919 such that @xmath920 is an exposed point of the compact convex set @xmath921 .
similarly to definition [ d : cost exponential ] , we transform @xmath922 with respect to @xmath918 , i.e. we consider the transformation @xmath923 , defined on @xmath924 by the relation @xmath925 and the modified cost function @xmath926 on @xmath927 , for which the @xmath352-convex potential function @xmath928 is convex .
we observe that @xmath929 for all @xmath282 , the point @xmath78 is sent to @xmath930 , @xmath931 , and @xmath311 is strictly convex thanks to theorem [ t : strict convex ] .
moreover , since @xmath920 was an exposed point of @xmath921 , @xmath932 is an exposed point of @xmath933 .
hence , we can find a vector @xmath934 such that the hyperplane orthogonal to @xmath935 is a supporting hyperplane for @xmath933 at @xmath936 .
let us now consider the section @xmath938 .
since @xmath939 , @xmath940 and @xmath311 is strictly convex , @xmath941 as @xmath521 .
thus by it is easily seen that for @xmath417 sufficiently small the following hold : @xmath942 where @xmath943 is a positive constant depending on @xmath417 and such that @xmath944 as @xmath521 .
since @xmath930 is the minimum point of @xmath311 , this immediately implies that one between our alexandrov estimates or must be violated by @xmath311 inside @xmath825 for @xmath417 sufficiently small , which is the desired contradiction .
a. figalli , y .- h
. kim and r.j .
when is a multidimensional screening a convex program ?
uniqueness and stability of optimal strategies in the principal - agent problem .
preprint at www.math.toronto.edu/mccann .
f. john .
extremum problems with inequalities as subsidiary conditions . in
_ studies and essays presented to r. courant on his 60th birthday , january 8 , 1948 _ , pages 187204 .
interscience , new york , 1948 . | consider transportation of one distribution of mass onto another , chosen to optimize the total expected cost , where cost per unit mass transported from @xmath0 to @xmath1 is given by a smooth function @xmath2 .
if the source density @xmath3 is bounded away from zero and infinity in an open region @xmath4 , and the target density @xmath5 is bounded away from zero and infinity on its support @xmath6 , which is strongly @xmath7-convex with respect to @xmath8 , and the transportation cost @xmath7 is non - negatively cross - curved , we deduce continuity and injectivity of the optimal map inside @xmath8 ( so that the associated potential @xmath9 belongs to @xmath10 ) .
this result provides a crucial step in the low / interior regularity setting : in a subsequent paper @xcite , we use it to establish regularity of optimal maps with respect to the riemannian distance squared on arbitrary products of spheres .
the present paper also provides an argument required by figalli and loeper to conclude in two dimensions continuity of optimal maps under the weaker ( in fact , necessary ) hypothesis * ( a3w)*@xcite . in higher dimensions , if the densities @xmath11 are hlder continuous , our result permits continuous differentiability of the map inside @xmath12 ( in fact , @xmath13 regularity of the associated potential ) to be deduced from the work of liu , trudinger and wang @xcite . |
recent single - dish and _ herschel _ spectroscopic observations successfully detected bright high-@xmath0 co emission ( @xmath0 = 4 3 up to 30 29 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) from nearby starburst galaxies and ( ultra-)luminous infrared galaxies ( u / lirgs ) , as well as lower-@xmath0 co lines ( e.g. , * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the observational evidence of extreme co excitation in u / lirgs is explained by a combination of several heating models of interstellar medium ( ism ) which consists of cosmic - ray dominated regions ( crdrs ) , photon dominated regions ( pdrs ) , x - ray dominated regions ( xdrs ) , and/or mechanically dominated regions ( mdrs ) , because coarse single - dish beams ( @xmath11 10 kpc ) are not possible to distinguish many molecular conditions @xcite .
high - resolution imaging of various co lines is , thus , an important way to investigate the multiple phases of the molecular ism and physical processes involved in obscured central activities of galaxies .
interferometric studies of co @xmath0=65 transition ( hereafter co ( 65 ) ) in nearby lirgs showed that the distribution of the co ( 65 ) emission is compact compared to the extended lower-@xmath0 co emission @xcite , suggesting that co ( 65 ) probes warmer and denser gas directly related to the nuclear activities .
however , a direct comparison among distributions of multiple co transitions is not straightforward as we are limited in angular resolution for the low-@xmath0 lines , sensitivities and @xmath12-coverages vary , and the high-@xmath0 lines require good weather making observations scarce from the ground . in this paper
, we present high - resolution , high - sensitivity , and @xmath12-matched alma observations of the nearby ir - bright galaxy ngc 1614 in order to understand the distribution and kinematics of multiple co lines and their excitation conditions in the nuclear region .
we investigate lower-@xmath0 co lines up to @xmath13 = 6 , which are thought to be mainly excited by star - forming activities .
ngc 1614 is a nearby lirg ( @xmath14 = 10@xmath15 @xmath16 ; * ? ? ?
* ) at a distance of 67.8 mpc ( 1= 330 pc ) . the total star formation rate ( sfr ) is @xmath1 75 @xmath17 yr@xmath18 based on the measurement of extinction - corrected paschen @xmath19 emission @xcite . merging signatures such as a tidal tail
are clearly seen in the hi and optical images ( figure [ figure_1]b ; * ? ? ?
numerical simulations performed by @xcite suggested that the system is a minor merger with the mass ratio of 1:3 1:5 ( i.e. , nearly major merger ) .
a strong hi absorption is detected toward the central region of ngc 1614 surrounded by an arc - like structure in hi emission .
visually , ngc 1614 has been classified as a merger remnant because of no nearby companion , and is one of the remnants with a rotating molecular gas disk which possibly evolves into a late - type galaxy @xcite .
hcn ( 43 ) and hco@xmath20 ( 43 ) observations show that ngc 1614 is consistent with a starburst - dominated galaxy without strong active galactic nuclei ( agn ) signatures .
this is consistent with the radio , mid - ir , and x - ray observations @xcite .
the global co spectral line energy distribution ( sled ) is characterized by a steep decline towards higher-@xmath0 transitions @xcite .
high - resolution ( @xmath1 025 @xmath1 83 pc ) alma band 9 observations revealed that the co ( 65 ) distribution coincides with a starburst ring ( r @xmath21 350 pc ) detected in pa @xmath19 ( figure [ figure_1]c ; * ? ? ?
* ) and 8.4 ghz radio continuum emission @xcite , while the co ( 21 ) and co ( 32 ) emission are more extended @xcite .
recent high - sensitivity co ( 21 ) observations revealed the presence of a high - velocity wing , which might be associated with a nuclear molecular outflow @xcite , whereas deep co ( 10 ) imaging found that extended molecular gas components associated with the southern tidal tail and a dust lane @xcite .
continuum emission from 4.81 ghz to 691 ghz is confined within the central 17 radius (= 560 pc ) and the radio - to - fir spectral energy distribution ( sed ) can be explained by star - forming activities ( @xcite , hereafter ) . this paper is organized as follows : the observations and data reduction are summarized in section [ obs ] .
all co maps , position - velocity diagrams , and line intensity ratios are presented in section [ results ] . in section [ model ] , we describe modeling procedures of the spatially - resolved co sled using three different models .
we provide a discussion of our modeling results in section [ discuss ] , including heating source , co - to - h@xmath22 conversion factor , a possible two - phase ism model and its heating source , cold molecular gas kinematics ( inflow and outflow ) , and overall picture of molecular gas in ngc 1614 .
finally , we summarize our main findings in section [ conclusion ] .
we have adopted h@xmath23 = 70 km s@xmath18 mpc@xmath18 , @xmath24 = 0.3 , @xmath25 = 0.7 throughout this paper .
the co ( 10 ) and co ( 21 ) observations toward ngc 1614 were carried out for our alma cycle 2 program ( i d : 2013.1.01172.s ) using thirty - five 12 m antennas .
the band 3 and band 6 receivers were tuned to the co ( 10 ) and co ( 21 ) emission lines in the upper sideband , respectively .
the band 3 data ( single pointing ) , with a projected baseline length ( @xmath26 ) of 28 - 1060 m ( c34 - 6 configuration ) , were obtained on 2014 august 30 ( on - source time of @xmath27 = 16.9 min . ) , and the band 6 data ( 3-point mosaic ) with @xmath26 = 15 - 349 m ( c34 - 2/1 configuration ) were obtained on 2014 december 8 ( @xmath27 = 7.3 min per pointing ) .
the correlator was configured to have four spectral windows ( spw ) , two of which were set to each sideband , each of the spws with 1.875 ghz bandwidth and 1.129 mhz resolution ( @xmath1 3.0 km s@xmath18 and 1.4 km s@xmath18 for band 3 and band 6 , respectively ) .
both bands used a strong quasar j0423 - 0120 for bandpass and phase calibration . j0423
- 0120 (= j0423 - 013 ) was also the flux calibrator in band 3 , while uranus was that in band 6 .
ngc 1614 was observed during cycle 0 using alma in bands 7 and 9 ( i d : 2011.0.00182.s and 2011.0.00768.s ) , and the data were originally published in @xcite and @xcite .
we obtained the calibrated archival visibility data of the co ( 32 ) and the co ( 65 ) from the alma archive .
detailed information of the cycle 0 and 2 data are shown in table [ table_obs ] ( see also ) .
the co ( 10 ) observations toward ngc 1614 were also carried out during cycle 2 ( i d : 2013.1.00991.s ; * ? ? ?
* ) using both ten 7 m antennas of aca @xcite and thirty - one 12 m antennas of the 12 m array .
these ancillary data are used to increase the sensitivity and recover extended emission that we ignore due to the @xmath12-clipping ( section [ 12 ] ) .
thus , we use those data to study the global gas kinematics of ngc 1614 .
we retrieved those data from the alma archive , and used one spw which was tuned to the co ( 10 ) line .
detailed description of this project is found in @xcite , and we list some parameters in table [ table_data ]
. we performed calibration and imaging using casa ( version 3.4 , 4.1.0 , and 4.2.2 for band 7 , 9 , and 3/6 , respectively ; * ? ? ? * ) .
detailed procedure for data processing is summarized in . before imaging
, we flagged the @xmath12 range below 45 k@xmath28 , so that the maximum recoverable scale ( mrs ; * ? ? ?
* ) of each alma observation is consistent .
the minimum @xmath12 distance is determined by the configuration of the co ( 65 ) observation .
since the truncated @xmath12 range at 45 k@xmath28 corresponds to the mrs of @xmath1 46 ( @xmath29 1.5 kpc ) , we ignore the missing flux effect on structures smaller than 46 throughout this paper .
all of the images are convolved to the same resolution ( 10 @xmath30 10 ) .
we made all maps in this paper without the primary beam correction , although flux measurements were done after that .
the systematic errors on the absolute flux calibration using a solar system object are estimated to be 5% , 10% , 10% , and 15% for both sidebands in band 3 , band 6 , band 7 , and band 9 , respectively @xcite .
the band 3 observation has a quasar as the flux calibrator , so we estimate the absolute systematic error on the flux calibration for our band 3 data by using the alma calibrator source catalogue . the flux uncertainty of j0423 - 013 on 2014 august 30 is 1.9 - 3.7% at 103.5 ghz
therefore , we adopt the flux uncertainty of 5% to be conservative . since the recovered flux of each co transition compared to single - dish observations @xcite is very low ( 27 - 59 % ) , ngc
1614 has large amounts of ambient molecular gas structures larger than 1.5 kpc even in the co ( 65 ) .
the detailed data properties and line information are listed in table [ table_data ] .
we used casa ( version 4.2.1 and 4.2.2 ) to process the data . in order to combine the co ( 10 ) data shown in section [ aca ] with the data described in section [ 12 ]
, we used statwt task to recalculate visibility weights for all the data .
the resultant @xmath26 is 7.3 - 1060 m. this corresponds to the mrs of 176 ( @xmath1 5.8 kpc ) .
we constructed a data cube with the velocity resolution of 20 km s@xmath18 .
the flux calibrator for the aca observation is a quasar j0501 - 0159 . using the alma calibrator source catalogue , the flux uncertainty of j0501 - 0159 on 2015 june 30
is estimated to be @xmath1 12% at 97.4 ghz .
therefore , we adopt the flux uncertainty of 12% when we use the aca - combined co ( 10 ) data .
the achieved rms noise level is 1.4 mjy beam@xmath18 with the synthesized beam size of 102 @xmath30 057 ( position angle = @xmath3188 ) . comparing with the co
( 10 ) flux obtained by nrao 12 m @xcite , the recovered flux is 113 @xmath32 26 % indicating that we recovered most ( or probably all ) of the co ( 10 ) emission .
the total h@xmath22 mass is ( 4.1 @xmath32 0.5 ) @xmath30 10@xmath33 @xmath17 assuming 1/5 of the milky way co - to - h@xmath22 conversion factor .
when comparing with aca - only co ( 10 ) flux measured by @xcite ( @xmath1 241 jy km s@xmath18 ) , the recovered flux is @xmath1 123 % .
this difference might be due to the differences of mask shape to measure the co flux in r.a .
, decl . , and velocity axes ( i.e. , low surface brightness , high velocity wings ) .
detailed information of the aca and the 12 m array data is shown in table [ table_obs ] , and the line information are listed in table [ table_data ] .
the @xmath12-matched co integrated intensity , velocity field , and velocity dispersion maps are shown in figure [ figure_2 ] , except for the aca - combined co ( 10 ) image ( figure [ figure_2b]a ) .
the channel maps are shown in appendix [ a1 ] .
the @xmath12-matched co distributions of ngc 1614 can be divided into two regions : the starburst ring ( central 590 pc in radius ) and an outer disk ( @xmath1 3.3 kpc in radius ) . the starburst ring is detected in x - ray , polycyclic aromatic hydrocarbon @xcite , pa @xmath19 @xcite , radio - to - fir continuum ( @xcite ; ) , hcn ( 43 ) , and hco@xmath20 ( 43 ) emission .
most of the co emission in the starburst ring arise from the western side .
the asymmetry in the starburst ring coincides with the radio - to - fir continuum and the hco@xmath20 ( 43 ) emission . on the other hand ,
the outer disk is only detected in extended pa @xmath19 and soft x - ray emission @xcite , indicating a site of moderate star formation relative to the ring ( but we still have missing flux for extended , low surface brightness co ( 65 ) emission ) .
all co ( 10 ) peaks outside the starburst ring coincide with the peaks detected in pa @xmath19 ( figure [ figure_1]c ) .
the overall velocity structure of all the co transitions seems to be dominated by rotation .
this is consistent with the velocity modeling with tilted concentric rings provided by @xcite .
the velocity field and dispersion of the starburst ring smoothly connect to those of the outer disk , which indicates that the kinematics ( e.g. , position angle , inclination ) are similar .
we identify non - circular motions in a region with a length of 9(@xmath1 3 kpc ) that is extended from north to south of the disk ( dashed lines in figure [ figure_2]b ) in all co images , except for the co ( 65 ) .
the apparent s "- shape feature in the velocity field suggests the presence of non - circular motions ( e.g. , * ? ? ?
* ) in the region where the dust lane connects with the starburst ring ( figure [ figure_1]b ) or a warped disk ( e.g. , centaurus a ; * ? ? ?
the apparent characteristics are similar to the nearby barred spiral galaxy ngc 1097 which is a minor merger system @xcite .
we investigate the position - velocity diagrams ( pvds ) per the different co transitions along the north - south direction with a width of 134 , length of 150 through the nucleus ( figure [ figure_5]a , [ figure_5]b , [ figure_5]c , and [ figure_5]d ) . the width and length of these pvds covers all of the co emission ( figure [ figure_2 ] ) . the pvds show rigid rotation ( i.e. , starburst ring ) in all co transitions and the s "- shape feature ( dashed ellipses in figure [ figure_5]a ) except for the co ( 65 ) .
the direction of the pvd cut is similar to the kinematical major axis determined from the co ( 21 ) velocity field ( 352@xmath32 1 ; * ? ? ?
while the large scale kinematics can be explained by both circular and non - circular motion , the central region appears to show outflows .
@xcite found high velocity co ( 10 ) wings ( 210 km s@xmath18
@xmath21 @xmath34 @xmath21 450 km s@xmath18 , where @xmath35 = 4763 km s@xmath18 ) with the p.a . of 82@xmath36 in the central region .
they suggested these components are a putative bipolar outflow with an outflow rate of @xmath1 40 @xmath17 yr@xmath18 .
the direction and the outflow rate is similar to those of the ionized gas outflow @xcite .
we did not detect the line wings in all co transitions , although they are detected in the aca - combined co ( 10 ) image ( see section [ aca_result ] ) . in figure [ figure_3]a
, we show the normalized radial distribution of the co sled of ngc 1614 .
the co ( 65 ) emission is centrally concentrated , while the lower-@xmath0 co transitions show an extended component ( i.e. , the outer disk ) .
the peak radius shifts toward outside as the co transition shifts to lower-@xmath0 , showing a radial excitation gradient in the starburst ring ( see section [ radex ] ) . in figure [ figure_3]b , as a comparison , we plot the normalized radial distribution of the radio - to - fir sed of ngc 1614 with the same angular resolution and mrs .
all of the continuum emission ( synchrotron , free - free , and cold dust ) are more compact than the lower-@xmath0 co transitions , but the sizes are similar to the extent of the co ( 65 ) emission .
this suggests that the co ( 65 ) is a better tracer of star - forming activity than lower-@xmath0 co transitions ( e.g. , * ? ? ?
although differences are seen in the spatial distribution between the transitions , there are no significant differences in the velocity profile within the central 23 radius as shown in figure [ figure_3]c , supporting that the rigid rotating component dominates the nuclear region .
we show the co ( 10 ) image that is generated by combining the aca data with the 12 array data in figure [ figure_2b]a .
the global velocity field does not change significantly , although additional extended structures toward the north - east ( redshifted ) and south - west ( blueshifted ) appear .
these arm - like structures are located along the optical dust lanes , as already found by aca - only co ( 10 ) image @xcite , and pa @xmath19 ( figure [ figure_1]c ) .
the low s / n high - velocity wings ( figure [ figure_2b]b and [ figure_2b]c ) are consistent with the presence of a putative outflow suggested by @xcite .
we investigate the pvd of the aca - combined image along the north - south ( along the kinematical major axis with a width of 134 ; figure [ figure_5]e ) and east - west direction ( along the kinematical minor axis with a width of 46 ; figure [ figure_12b]a ) through the radio nucleus @xcite . along the major axis , we found more extended components than the @xmath12-clipped co ( 10 ) image .
the aca data shows features that are consistent with bar - induced inflow motions predicted by @xcite . in contrast , as shown in the pvd along the minor axis which emphasizes low surface brightness components within the 46 width , there are high - velocity components around the nuclear region . since the high - velocity components show the same spatial and spectral distributions as the molecular gas outflow suggested by @xcite , we regard the high - velocity components as the molecular gas outflow .
we constructed the integrated intensity image of the blueshifted ( 4340 - 4460 km s@xmath18 ) and redshifted ( 4940 - 5000 km s@xmath18 ) outflows in figure [ figure_12b]b .
the extracted velocity ranges contain high - velocity wings stronger than 3@xmath37 .
the co ( 10 ) outflow shows an extended blueshifted ( redshifted ) emission toward the south - east ( north - west ) direction .
the molecular gas outflow seems to encompass the ionized gas outflow @xcite .
this picture is consistent with the molecular gas outflow entrained by the ionized gas outflow in the nearby starburst galaxy ngc 253 and m51 @xcite .
we show co ( 21)/co ( 10 ) , co ( 32)/co ( 10 ) , and co ( 65)/co ( 10 ) line ratio images in figure [ figure_4]a , b , and c , respectively .
we divide the integrated intensity maps ( figure [ figure_2]d , g , and j ) by the co(1 - 0 ) integrated intensity map ( figure [ figure_2]a ) to make the line ratio maps . after that
, we masked all the line ratio maps by using the co ( 10 ) outline which is shown in figure [ figure_4 ] .
the co ( 21)/co ( 10 ) and co ( 32)/co ( 10 ) line ratios show higher values than the optically - thick thermalized value , while the co ( 65)/co ( 10 ) line ratio is moderate . those high co ( 21)/co ( 10 ) and co ( 32)/co ( 10 ) line ratios indicate low optical depths for the co ( 10 ) line and/or the presence of multiple phase ism ( e.g. , co ( 32)/co ( 10 ) ratio in arp 220 ; * ? ? ?
the peak position of each line ratio is located at 1east of the nucleus .
since it coincides with the eastern part of a possible ionized and molecular gas outflow @xcite , the high line ratio component might be due to the interaction between the outflow and the starburst ring .
strong outflow can heat surrounding ism kinematically .
this results in highly excited ( i.e. , low-@xmath0 depopulation ) or highly turbulent ( i.e. , low optical depth ) molecular gas around the shock interface , and thus high line ratios .
high co ( 32)/co ( 10 ) ratio is also observed in the circumnuclear disk ( cnd ) of seyfert 2 galaxy ngc 1068 ( see section 7.2 of * ? ? ? * ) .
the high ratio in the cnd are located parallel to the bipolar jet , which may indicate that the cnd is affected by the nuclear jet .
a similar situation is found in the nuclear region of ngc 1614 ( figure [ figure_4 ] ) , which may indicate outflow - ring interaction
. however , further molecular gas and ionized gas observations are necessary to confirm this scenario quantitatively .
the co sleds in u / lirgs are often explained by a combination of crdr , pdr , xdr , and mdr .
however , especially for ngc 1614 , previous studies suggested that contributions from crdr and xdr are small or those models can not reproduce observed quantities ( see * ? ? ?
* ; * ? ? ?
* in details ) , so we only focus on the effect of both pdr and mdr in this paper .
we used the non - lte radiative transfer code radex @xcite and varied the parameters until the residuals between the observed line fluxes and the modeled line fluxes are minimized in a @xmath38 sense . assuming a single - phase ism ( i.e.
, the gas physics can be represented by a single set of the excitation parameters ) , an expanding sphere geometry ( @xmath39 = 200 km s@xmath18 ) , the cosmic microwave background temperature ( @xmath40 = @xmath41 = 2.73 k ) , and a [ co]/[h@xmath22 ] abundance ( 3 @xmath30 10@xmath43 ) which is similar to the standard value observed in galactic molecular clouds @xcite , we derived the physical conditions of molecular gas .
the upper state energies and the einstein coefficients were taken from the _ leiden atomic and molecular database _ ( lamda ; * ? ? ?
we varied the gas kinetic temperature within a range of @xmath5 = 10 - 300 k ( d@xmath5 = 1 k ) , the gas density of @xmath2 = 10@xmath44 - 10@xmath45 @xmath4 ( d@xmath2 = 10@xmath46 @xmath4 ) , and the gas column density of @xmath47(h@xmath22 ) = 10@xmath48 - 10@xmath49 @xmath50 ( d@xmath47(h@xmath22 ) = 10@xmath46 @xmath50 ) .
for the input parameters , we used the radial flux distribution of each co line shown in figure [ figure_3 ] , which were radially binned at 02 intervals .
we note that the @xmath51 = @xmath41 is a reasonable assumption although one might expect that , given the intense dusty star formation , a higher @xmath51 is more suitable ( i.e. , @xmath51 @xmath1 @xmath52 ) .
we estimated the dust opacity at the starburst ring in @xcite and found a low dust opacity ( @xmath21 0.21 ) even at 690 ghz .
therefore , at this low frequency regime ( 115 - 690 ghz ) , the background temperature can not be as high as @xmath52 .
we show results of the radex modeling at the outer region contained within a radius of 28 - 30 in figure [ figure_14 ] .
line ratios related to co ( 65 ) show different distributions to lower-@xmath0 co line ratios in the parameter planes , demonstrating the importance of the co ( 65 ) data .
the results as a function of radius are shown in figure [ figure_6]a , [ figure_6]b , and [ figure_6]c .
the minimum reduced-@xmath38 of the central 18 is relatively poor ( @xmath53 3.6 ) , whereas that of the outer region ( r @xmath10 18 ) is @xmath11 3.6 .
gas density is high ( @xmath1 10@xmath54 @xmath4 ) independently of the distance from the nucleus . on the other hand , both the @xmath5 (= 20 - 70
k ) and @xmath47(h@xmath22 ) (= 10@xmath55 @xmath4 ) decrease as the radius increases .
their peaks coincide with the radius of the starburst ring ( 100 - 350 pc ; * ? ? ?
* ; * ? ? ?
the derived @xmath5 are about two times higher than the results of a similar modeling for ngc 1614 by @xcite .
this can be explained by the different spatial sampling between our analysis and theirs . in order to match the mrs between data sets , @xcite used single - dish data to compensate for the missing short baselines of their interferometric co ( 21 ) and co ( 32 ) data , while we clipped the inner @xmath12 data at 45 k@xmath28 ( see section [ obs ] ) .
we calculated the optical depth for each transition ( figure [ figure_7 ] ) , yielding that @xmath56 shows moderate optical depth ( @xmath11 1 ) , @xmath57 and @xmath58 are optically thick ( 1 - 10 ) , and @xmath59 becomes thin as the radius increases ( @xmath59 = 8 to 0.1 ) . the derived @xmath56 for ngc 1614 is consistent with that is estimated for a mid - stage merger vv 114 ( @xmath1 1 ; * ? ? ? * ) , the central region of a close galaxy pair ngc 6240 ( 0.2 - 2 ; * ? ? ?
* ) , the overlap region of the mid - stage merger arp 299 ( 0.5 - 1.5 ; * ? ? ? * ) , and the local mid - stage merger the antennae ( 0.2 - 1.5 ; * ? ? ? * ) .
this suggests that luminous mergers have co ( 10 ) line of moderate optical depths and thick high-@xmath0 co condition .
we reconstructed the best - fit co sled from models in order to compare with the observational data ( figure [ figure_8]a and [ figure_8]b ) .
the co sleds in the outer disk ( @xmath10 18 ) show good agreement with the observations , while those in the starburst ring ( @xmath11 18 ) systematically underestimated the co ( 32)/co ( 10 ) and co ( 65)/co ( 10 ) ratios .
one possibility to explain the systematic difference is that the starburst ring of ngc 1614 has an additional warmer ism , as suggested by @xcite .
the additional warm gas can move the modeled co ( 32)/co ( 10 ) and co ( 65)/co ( 10 ) ratios toward a higher value by fixing the lower-@xmath0 co fluxes .
we note that the flux differences between the observations and the model are unlikely to be caused by dust extinction . although some u / lirgs show high dust opacities even in the submillimeter wavelengths ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , the cold dust in the starburst ring of ngc 1614 is still optically thin at 691 ghz .
we compare the observed co sled ( figure [ figure_8]a ) with the model of , which can parameterize the co sled of galaxies by a power - law function of @xmath60 .
using the @xmath60 derived by , we reconstructed the co sled ( figure [ figure_8]c ) .
the @xmath60 were derived by fitting the radial radio - to - fir seds binned at 02 intervals up to 18 radius assuming the absence of a strong or a hearvily obscured agn .
each continuum images have the same angular resolution of 10 and mrs of 46 as the co images , so we can compare the observed co sleds with the reconstructed co sleds dicrectly .
as clearly seen in figure [ figure_8]a and [ figure_8]c , this model is inconsistent with the observed co sleds at the starburst ring ( @xmath21 18 ) .
the co ( 32)/co ( 10 ) and co ( 65)/co ( 10 ) ratios are also underestimated as already described in section [ radex ] .
a multi - phase ism model is one of the possible explanations .
because our analysis includes only four co transitions and can only constrain a small number of free parameters , we will focus on a two - phase ism model rather than a multiple - phase ism ( section [ ism ] ) .
the @xmath60 used here is extinction - free because used the low frequency continuum from 4.81 to 691 ghz .
thus , we can ignore the extinction effect in this comparison .
we used the pdr toolbox @xcite , which can model the beam - averaged incident far - ultraviolet ( fuv ; 6 ev @xmath21 @xmath61 @xmath21 13.6 ev ) radiation field intensity ( @xmath6 in the unit of 1.6 @xmath30 10@xmath62 erg @xmath50 s@xmath18 ) and pdr hydrogen nucleus density ( @xmath63 @xmath4 = @xmath64 + 2@xmath2 ) from the flux density of each co transition , in order to better understand the starburst activities of ngc 1614 .
we varied @xmath63 within the 90% confidence range of 2@xmath2 derived by the single - phase radex ( d@xmath63 = 10@xmath65 @xmath4 ) and @xmath6 within a range of 10@xmath66 - 10@xmath67 ( d@xmath6 = 10@xmath65 ) .
here we assumed that the @xmath64 is negligible compared to the 2@xmath2 term because the central kpc of mergers are often dominated by the molecular phase ( e.g. , * ? ? ?
* ; * ? ? ?
this assumption is consistent with the compact co distributions ( figure [ figure_2 ] ) and the extended arc - like hi distribution ( figure [ figure_1]a ) around the kpc region of ngc 1614 . the results are shown in figure [ figure_6]d .
the minimum reduced-@xmath38 is @xmath1 1 for both grids .
the starburst ring has higher @xmath6 ( @xmath10 10@xmath68 , average = 10@xmath7 ) than the outer disk ( @xmath11 10@xmath68 , average = 10@xmath9 ) .
the average value of the whole galaxy is @xmath1 10@xmath69 . in order to check the reliability of the derived @xmath6
, we roughly estimate the averaged fuv radiation field on dust - obscured star - forming region using the prescription provided by @xcite , @xmath70,\end{aligned}\ ] ] where @xmath71 is the mean distance of fuv photons propagation before being absorbed , @xmath72 is the size of the starburst ring ( @xmath1 350 pc ; * ? ? ?
for the clumpy pdr model , @xmath71 is @xmath11 1 pc .
the derived @xmath73 is @xmath11 10@xmath74 , which is consistent with the average @xmath6 derived from the co sled .
the s "- shape morphology in the velocity field suggests the presence of a north - south bar or a warped disk morphology .
it coincides with the leading edge of the main near - ir bar suggested by @xcite , so we suggest that the north - south bar results in the
s "- shape non - circular motions ( i.e. , inflow along a bar ) detected in the lower-@xmath0 co transitions .
the presence of the cold gas inflow along the bar is also suggested by @xcite .
in addition , analysis of n - body / sph simulation performed by @xcite revealed that tidal interaction between gas - rich disks can produce gas flows directly toward the central regions due to the production of both transient arms and long - lived @xmath75 = 2 bars ( i.e. , bar instability ) .
they also found that the inflowing gas is characterized by distinct diffuse gas clumps in the pvds , while dense gas only present in the central rigid rotating component .
this simulation result is consistent with our observations including non - detection of the s "- shape feature in the co ( 65 ) image .
we compare @xmath5 and @xmath76 with @xmath60 , which is derived by modeling the radio - to - fir sed , in figure [ figure_9 ] .
there is a positive correlation with a correlation coefficient of @xmath77 = 0.91 between @xmath78 and @xmath79 .
this is a sensible result because both parameters are related to star formation .
the relation between @xmath78 and @xmath80 also shows a positive correlation ( @xmath77 = 0.93 ) . using a pdr model described in equation 3 of @xcite and assuming @xmath81 which is found by pdr models
, the derived @xmath5 of 20 - 70 k corresponds to @xmath79 of @xmath11 10@xmath82 .
this is consistent with the result of the co sled model described in section [ pdr ] .
these models and a simple comparison indicates that the dusty starburst in ngc 1614 is enough to power the co excitation up to @xmath13 = 6 even if we just assume a single - phase molecular gas ism for the starburst ring .
molecular hydrogen mass is one of the most fundamental parameters to characterize galaxies in relation to their starbursts and agns .
although the galactic - scale co ( 10 ) luminosity to h@xmath22 mass conversion factor ( @xmath83/@xmath84 @xmath85 @xmath86 ) in nearby u / lirgs has been well studied ( * ? ? ?
* for a review ) , the spatial properties are not fully understood .
we measure the @xmath86 as a function of radius using , conversion factor (= @xmath86 = @xmath83/@xmath87 ) as a function of the radius derived by the signle - phase radex model ( i.e. , @xmath47(h@xmath22 ) ) .
the dashed lines show the approximate inner and outer radii of the nuclear starburst ring @xcite .
the shaded areas show the 90% confidence .
the blue and red regions show approximate @xmath86 for the milky way and ulirgs @xcite . , width=226 ] conversion factor for co ( 1 - 0 ) , co ( 32 ) , and co ( 65 ) lines as a function of the radius .
the dashed lines show the approximate inner and outer radii of the nuclear starburst ring @xcite .
the shaded areas show the 90% confidence .
, width=226 ] @xmath88 where @xmath89 is the mass of the hydrogen molecule , @xmath90 is the column density derived by radex ( section [ radex ] ) , @xmath91 is the area of the concentric rings , and @xmath84 is the co ( 10 ) luminosity .
we multiply 1.36 to account for the helium abundance relative to hydrogen .
the derived @xmath86 at the starburst ring is 0.2 - 0.5 ( @xmath92/[h_2]}}$ ] ) except for the central hole ( central 02 radius ) which yields @xmath1 0.05 ( @xmath92/[h_2]}}$ ] ) ( figure [ figure_10 ] ) .
this is lower than 0.9 - 1.5 ( @xmath92/[h_2]}}$ ] ) at the starburst ring and 0.1 ( @xmath92/[h_2]}}$ ] ) at the central hole derived by @xcite .
the difference between both models can be explained by the presence of virialized giant molecular clouds ( gmcs ) and the different treatment of the missing flux . assuming that gmcs in the central region of ngc 1614 are virialized , @xmath86 depends on gas density and kinetic temperature . as a consequence of the different missing flux treatment , the results of our radex modelings for the starburst ring are warmer ( 27 - 120 k and 10@xmath93 @xmath4 ) than the results in @xcite ( 19 - 64 k and 10@xmath94 @xmath4 ) ( see also section [ radex ] ) .
assuming the same gas density ( @xmath95 ) and beam filling factor for both models , @xmath86 derived from our model shall indeed be @xmath1 2 times lower than the @xcite value , indicating that both models are consistent .
we note that the treatment of the missing flux ( i.e. , how to achieve the same mrs for all transitions ) should be different depending on the situation .
when we have short - spacing data for all " transitions , we should combine the data with data obtained with the compact configurations , which allows us to recover extended structures
. in the case of this paper , single - dish or aca data for several transitions are lacking , and thus we chose to clip the inner @xmath12 , making @xmath86 different from @xcite . comparing to local lirgs , it is known that nearby spirals show larger @xmath86 ( e.g. , * ? ? ?
@xcite constructed @xmath86 as a function of galactocentric radius ( @xmath96 = @xmath77/@xmath97 , where @xmath97 is the b - band isophotal radii at 25 mag arcsec@xmath98 ) for 26 spiral galaxies , and found lower @xmath86 at the galaxy centers ( @xmath96 @xmath21 0.2 ) by a factor of 1 @xmath1 5 relative to their disks ( @xmath96 = 0.4 - 1.0 ) .
using @xmath97 of 6.5 kpc ( @xmath1 20 ) derived by @xmath97 = 1.9 @xmath30 @xmath99 @xcite and @xmath100 = 3.44 kpc @xcite for ngc 1614 , the outer radius of the central hole , the starburst ring , and the co disk of ngc 1614 are @xmath96 @xmath1 0.01 , 0.05 , and 0.14 , respectively ( figure [ figure_10 ] ) .
this indicates that ngc 1614 shows a similar radial @xmath86 trend to nearby spirals , but it has a more compact distribution and lower @xmath86 value than that of nearby spirals for a given galactocentric radius .
we calculate the conversion factor for co ( 32 ) ( @xmath83/@xmath101 @xmath85 @xmath102 ) and co ( 65 ) ( @xmath83/@xmath103 @xmath85 @xmath104 ) in figure [ figure_10b ] assuming the single - phase ism .
the radial distribution of @xmath102 is almost the same as that of @xmath105 , while the @xmath104 shows a steeper shape , except for the nuclear region ( @xmath21 10 ) , which shows a similar value of @xmath86 because of the sub - thermalized conditions in the outer region . _
herschel _ and other single - dish observations revealed that the co sled of nearby u / lirgs can be represented by multi - phase molecular gas ism model ( e.g. , * ? ? ?
* ; * ? ? ?
although the spatial properties of the co sled for u / lirgs are unclear because of the limited number of high - resolution co data , our high - resolution co ( 10 ) and co ( 21 ) observations , as well as the archival co ( 32 ) and co ( 65 ) data , allow us to study this for a lirg for the first time . before discussing the two - phase molecular gas
ism in ngc 1614 , we note that the dust ism in the starburst ring can be described by warm and cold components .
@xcite reported the presence of a warm dust component in the starburst ring ( @xmath1 110 k ) revealed by modeling of the mid - ir emission , although there is also a cold dust component ( @xmath1 35 k ; * ? ? ? * ) . in an ism where the gas and dust are intermingled
, the molecular gas component separates into a warm component of @xmath106 @xmath1 110 k and a cold component of @xmath107 @xmath1 35 k. this can be addressed by radex with simple assumptions . in order to model the warm and cold gas components ,
we minimize the number of free parameters by fixing the cold gas component using the observed line ratios at the outer region contained within a radius of 28 - 30 .
this region shows a good agreement with a single - phase model with a temperature of @xmath107 @xmath1 19 k ( figure [ figure_6]a ) .
here we assume that cold molecular clouds , which dominate the outer region , dominate the cold gas component in the starburst ring .
we calculate the contribution of the warm gas component to the total co luminosity by using , @xmath108 where @xmath109 is the co ( 21 ) luminosity of the warm gas component and @xmath110 is the observed co ( 21 ) luminosity within the central 18 radius . in the fitting routine , we varied @xmath77 within a range of 0 - 0.009 ( d@xmath77 = 0.001 ) , 0.01 - 0.09 ( d@xmath77 = 0.01 ) , and 0.1 - 0.9 ( d@xmath77 = 0.1 ) to estimate the co luminosities of the warm gas and cold gas components .
then , we varied the warm gas kinetic temperature within a range of @xmath106 = 10 - 800 k ( d@xmath106 = 10 k ) and the warm gas column density of @xmath47(h@xmath22)@xmath111 = 10@xmath112 - 10@xmath113 @xmath50 ( d@xmath47(h@xmath22)@xmath111 = 10@xmath46 @xmath50 ) .
we fixed [ co]/[h@xmath22 ] (= 3 @xmath30 10@xmath43 ) and the gas density of the warm gas component as @xmath114 = 10@xmath115 ( grid a ) , 10@xmath116 ( grid b ) , and 10@xmath117 @xmath4 ( grid c ) .
the results are shown in figure [ figure_11 ] .
the minimum @xmath38 are 1.30 ( grid a ) , 1.05 ( grid b ) , and 0.85 ( grid c ) , which are smaller than the @xmath38 values of the single - phase modeling for the starburst ring ( r @xmath21 18 ) .
the 90% confidence ranges of @xmath106 are 300 - 370 , @xmath10 250 , and 70 - 450 k , whereas those of @xmath47(h@xmath22)@xmath111 are @xmath10 10@xmath118 , @xmath10 10@xmath119 , and 10@xmath120 for the grid a , b , and c , respectively .
the 90% confidence ranges of @xmath77 are @xmath21 0.1 ( best - fit = 0 ) , @xmath21 0.8 ( best - fit = 0 ) , and @xmath21 0.6 ( best - fit = 0.003 ) for the grids a , b , and c , respectively , indicating less contribution of the warm gas component to the lower-@xmath0 co fluxes .
this two - phase ism model in ngc 1614 is similar to that found in the nearby ulirg arp 220 , showing that a warm gas component has almost no contribution to co lines lower than @xmath0 = 32 @xcite . using the pdr model ( section [ pdr ] ) assuming @xmath63 = 10@xmath121 @xmath4
, we find that the fuv field ( @xmath76 ) of the inferred warm gas component is higher than 10@xmath122 .
considering the relatively low @xmath60 and the absence of a heavily obscured agn , such extreme radiation field is unreasonable for ngc 1614 . in summary
, the starburst ring of ngc 1614 can be described by a two - phase molecular gas ism with a cold component of @xmath1 19 k ( dominated by normal gmcs ) and a warm component of @xmath10 70 k ( warm ism directly related to powerful activities ) .
uv - radiation produced by star forming activities ( i.e. , clumpy pdrs ) are not possible heating sources of the warm gas .
we note that these models are consistent with a two - phase dust ism ( @xmath1 35 k and @xmath1 110
k ) of the starburst ring @xcite and multi - phase models for other u / lirgs ( e.g. , * ? ? ?
* ; * ? ? ?
the observed high-@xmath0 co fluxes are systematically higher than those in the narayanan & krumholz model ( section [ nk ] ) .
also , fuv radiation heating due to the observed star - forming activities can not reproduce the possible warm gas component ( section [ two ] ) assuming the two - phase model . here
we employ the mechanical heating ( i.e. , shock excitation ) as an alternative mechanism to power the warm gas .
we found the largest values of the co ( 21)/co ( 10 ) and co ( 32)/co ( 10 ) ratios ( @xmath1 10 and @xmath1 27 , respectively ) along with the putative outflow as described in section [ ratio ] , which are similar to what has been observed in the nearby seyfert galaxy ngc 1068 ( @xmath1 10 and @xmath1 35 respectively at the e - knot ; * ? ? ? * ) .
ngc 1068 has a cnd which is thought to be heated by the nuclear jet @xcite . the similarity between these galaxies may suggest that in the central 18 of ngc 1614 the co emission is significantly affected by the kinematical interaction between the starburst ring and the putative outflow .
the outflow can heat the starburst ring kinematically , producing highly excited molecular gas components around the shocked region .
this scenario is consistent with the detection of h@xmath22 10 s(1 ) line within the central 18 radius @xcite , suggesting that nearly 1/3 of the total near - ir h@xmath22 is produced by shocks .
stellar feedback such as supernova explosions and stellar winds are also possible sources of mechanical heating .
therefore , we compare the kinetic luminosity of the outflow and the degree of stellar feedback with the total co luminosity of the warm gas .
we derive the kinetic luminosity from the supernovae ( @xmath123 ) using @xcite , @xmath124 where @xmath125 is the supernova rate and @xmath126 is the energy of a supernova ( @xmath1 10@xmath127 erg ; * ? ? ?
using the @xmath125 of @xmath1 0.9 @xcite , we obtain @xmath123 of @xmath1 2.7 @xmath30 10@xmath128 erg s@xmath18 . assuming the energy from stellar winds is similar to supernovae , the total kinetic luminosity from the stellar feedback is @xmath1 5.4 @xmath30 10@xmath128 erg s@xmath18 . on the other hand ,
the kinetic luminosity from the putative outflow is @xmath1 2.5 @xmath30 10@xmath129 erg s@xmath18 @xcite . comparing with the inferred co luminosity of the warm gas component ( @xmath1 1.2 @xmath30 10@xmath130 erg s@xmath18 ;
sum up to @xmath13 = 23 ) , both the outflow and the stellar feedback are possible heating sources of the warm gas component even when we adopt a low energy injection efficiency of @xmath11 5% .
although the molecular outflow is a possible source in terms of the energy , the high resolution co ( 6 - 5 ) image @xcite shows that the distribution is closely related to the starburst ring , not the outflow ( figure [ figure_12b]a ) .
this indicates that stellar feedback is more likely to be the heating source for higher-@xmath0 co rather than the putative molecular outflow when we consider both energy and spatial co ( 65 ) distribution .
we note that the mechanical energy is also a possible heating source in arp 220 @xcite , ngc 6240 @xcite , and other u / lirgs @xcite . in order to evaluate the effect of the molecular gas inflow
, we estimate molecular gas inflow rate ( ifr ) using , @xmath131 where @xmath132 is the inflowing h@xmath22 mass , @xmath133 is the inflowing velocity , and @xmath134 is the distance from the starburst ring ( e.g. , * ? ? ? * ) . assuming the inflowing component is along the gas disk , equation [ eqn5 ]
can be calculated as follows . @xmath135 where @xmath136 is the line - of - sight velocity relative to the rotation velocity at the projected distance ( @xmath137 ) , @xmath138 is the inclination of the disk , and
the @xmath139 is the co - to - h@xmath22 conversion factor for the galaxy (= 4.3 ; * ? ? ?
here we adopt @xmath138 of 33and @xmath140 of 0.2@xmath139 which are the same values to derive the molecular gas outflow rate ( ofr ; * ? ? ?
we estimate the flux density of the inflowing gas components using figure [ figure_5]e .
the northern ( southern ) inflowing component has 25.8 ( 19.3 ) jy km s@xmath18 , which corresponds to @xmath141 of 10@xmath142 ( 10@xmath143 ) @xmath17 .
the @xmath136 and @xmath137 of the northern ( southern ) component are 80 ( 40 ) km s@xmath18 and 500 ( 500 ) pc , respectively .
the derived ifr is 54 @xmath17 yr@xmath18 with the uncertainty by a factor of three , which mainly due to the uncertainties of @xmath136 and @xmath137 . comparing the sfr in the starburst ring ( 32.8 @xmath17 yr@xmath18 ; * ? ? ?
* ) and ofr ( 40 @xmath32 12 @xmath17 yr@xmath18 ; * ? ? ?
* ) , and assuming that the ifr , ofr , and sfr continue constantly , we can investigate the balance between molecular gas input and output in the starburst ring of ngc 1614 .
@xmath144 independent of the adopted @xmath145 , the molecular gas input and output are roughly balanced , indicating that ngc 1614 can sustain both star - forming activities and an outflow in the starburst ring until exhausting the molecular gas reservoir in the starburst ring and the outer region via bar - driven inflow .
the total molecular gas depletion time due to star formation and outflow (= @xmath83/(sfr + ofr ) ) ranges from 17.7 @xmath32 4.6 myr ( @xmath145 = @xmath146 ) to 56.6 @xmath32 9.3 myr ( @xmath145 = 0.2@xmath146 ) .
this is consistent with the typical duration timescale of u / lirg phase predicted by numerical merger simulations ( dozens of myr ; * ? ? ?
* ; * ? ? ?
the luminous merger remnant ngc 1614 may quench itself until the next dozens of myr due to depletion of the cold molecular gas reservoir ( if the molecular outflow and star formation continue constantly ) .
we summarize all results provided by this paper and previous studies into a simple schematic picture ( figure [ figure_13 ] ) .
we used a position angle of @xmath1 -352and an inclination of @xmath14736as estimated by @xcite , for the cold gas disk and the starburst ring .
this is consistent with the results provided by @xcite .
we found s "- shape non - circular motions along the leading edge of the north - south near - ir bar in the lower-@xmath0 co velocity fields and pvds .
the near - ir bar connects to the near - ir spiral arms @xcite .
moreover , the northern part of the molecular bar was detected in sma co ( 21 ) observations ( dashed line in figure 3 of * ? ? ?
@xcite also suggested that the northern molecular gas components are diffuse gas clouds which are now funneled into the nuclear region , although the flowing direction is different from the picture shown in figure [ figure_13 ] . the central starburst ring ( sfr @xmath1 32.8 @xmath148 yr@xmath18 ; * ? ? ?
* ) was detected in star formation tracers @xcite and dense / warm gas tracers @xcite .
the nucleus has no strong agn or heavily obscured agn @xcite .
the upper limit of the agn luminosity suggests that it might be possible for driving the putative molecular gas outflow with the outflow rate of 40 @xmath148 yr@xmath18 as well as the starburst activities @xcite .
the detection of the molecular outflow coincides with the geometry of the ionized gas outflow with no redshifted component due to the extinction by the large column of dust in the inclined disk @xcite .
radiative transfer modeling with radex suggest that the starburst ring will have a single - phase ( @xmath1 42 k ) or two - phase ( @xmath1 19 k and @xmath10 70 k ) molecular gas ism , while the outer disk has a single cold gas ism ( @xmath1 22 k ) .
the single - phase model is consistent with the fuv radiation field estimated by pdr models .
however , star - forming activities are not enough to power the warm gas component in the starburst ring if the two - phase ism is valid .
alternatively , we suggest the mechanical heating from the supernovae and stellar winds as a heating source of the warm gas component .
in this paper , we present a detailed study of the molecular gas ism in the luminous merger remnant ngc 1614 through high - resolution , high - sensitivity , and @xmath12-matched alma observations of the co ( 10 ) , co ( 21 ) , co ( 32 ) , and co ( 65 ) lines .
the results are summarized as follows : 1 .
the co ( 65 ) line shows a compact distribution which coincides with the starburst ring detected in pa @xmath19 and radio - to - fir continuum emission , while the other lower-@xmath0 co lines are extended .
this indicates that the co ( 65 ) can be used as a better tracer of star formation as already suggested by @xcite .
we find a s "-
shape non - circular motion along the near - ir north - south bar in the co ( 10 ) , co ( 21 ) , and co ( 32 ) velocity fields . comparing with numerical simulations , this may be a bar - driven cold gas inflow connecting to the starburst ring .
radiative transfer modeling with radex reveal that the molecular gas ism in ngc 1614 can be described in a single - phase or two - phase model .
the single - phase model shows a radial gradient of the gas kinetic temperature from 70 to 20 k and the h@xmath22 column density from 10@xmath119 to 10@xmath149 @xmath50 with peaks at the starburst ring .
this single - phase model is applicable to explain the observed co sleds , while there is a systematic underestimation of the co ( 32 ) and co ( 65 ) flux . the narayanan & krumholz co sled model using
the observed @xmath60 also shows the same trend . to account for the underestimation in both models , we verify the two - phase model to the line ratios in the central 18 radius .
4 . the two - phase radex model for the starburst ring shows a good agreement with a cold gas component of @xmath1 19 k and a warm gas component of @xmath10 70 k , while there are number of assumptions at this stage .
this is consistent with the two - phase dust ism ( @xmath1 35 k and @xmath1 110 k ) revealed by high - resolution ( @xmath21 05 ) mid - ir and far - ir observations and two - phase modelings for other nearby u / lirgs .
co observation ( @xmath13 @xmath150 7 ) is critical to characterize the molecular gas ism in the starburst ring of ngc 1614 .
considering that the observed star - forming activities are not enough to power the warm gas component inferred from the two - phase radex , we suggest mechanical heating from supernovae and stellar winds as an alternative power source .
the summation between the cold molecular gas outflow rate ( ofr ) and the star formation rate ( sfr ) in the starburst ring is comparable to the cold gas inflow rate ( ifr ) , showing evidence of the balance between the molecular gas input and output at the starburst ring . assuming ifr , ofr , and sfr continue constantly , it takes dozens of myr to exhaust all h@xmath22 mass in ngc 1614
this is consistent with the typical u / lirg duration timescale suggested by numerical merger simulations .
we will present our high - resolution multi - transition observations , including co , @xmath151co , cn , cs , etc .
, toward ngc 1614 using band 3 and band 6 of alma in a forthcoming paper ( ando et al . in preparation ) .
the authors thank an anonymous referee for coments that improved the contents of this paper .
ts thanks n. lu , e. w. pellegrini , and e. schinnerer for useful discussion .
ts and the other authors thank alma staff for their kind support and h. nagai for the instruction of alma data reduction .
ts and ml are financially supported by a research fellowship from the japan society for the promotion of science for young scientists .
ts was supported by the alma japan research grant of naoj chile observatory , naoj - alma-0114 .
di was supported by the alma japan research grant of naoj chile observatory , naoj - alma-0011 , and jsps kakenhi grant number 15h02074 .
this paper makes use of the following alma data : ads / jao.alma#2011.0.00182.s , ads / jao.alma#2011.0.00768.s , ads / jao.alma#2013.1.00991.s , and ads / jao.alma#2013.1.01172.s .
alma is a partnership of eso ( representing its member states ) , nsf ( usa ) and nins ( japan ) , together with nrc ( canada ) , nsc and asiaa ( taiwan ) , and kasi ( republic of korea ) , in cooperation with the republic of chile .
the joint alma observatory is operated by eso , aui / nrao and naoj .
this research has made extensive use of 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 .
99 aalto , s. 2013 , iau symposium , 292 , 199 albrecht , m. , krgel , e. , & chini , r. 2007 , , 462 , 575 alonso - herrero , a. , engelbracht , c. w. , rieke , m. j. , rieke , g. h. , & quillen , a. c. 2001 , , 546 , 952 armus , l. , mazzarella , j. m. , evans , a. s. , et al .
2009 , , 121 , 559 bayet , e. , williams , d. a. , hartquist , t. w. , & viti , s. 2011 , , 414 , 1583 bellocchi , e. , arribas , s. , & colina , l. 2012 , , 542 , a54 bellocchi , e. , arribas , s. , colina , l. , & miralles - caballero , d. 2013 , , 557 , a59 blake , g. a. , sutton , e. c. , masson , c. r. , & phillips , t. g. 1987 , , 315 , 621 bolatto , a. d. , warren , s. r. , leroy , a. k. , et al .
2013 , , 499 , 450 bolatto , a. d. , wolfire , m. , & leroy , a. k. 2013 , , 51 , 207 combes , f. , garca - burillo , s. , casasola , v. , et al .
2013 , , 558 , a124 costagliola , f. , aalto , s. , rodriguez , m. i. , et al .
2011 , , 528 , a30 downes , d. , & solomon , p. m. 1998 , , 507 , 615 espada , d. , matsushita , s. , peck , a. b. , et al .
2012 , , 756 , l10 falstad , n. , gonzlez - alfonso , e. , aalto , s. , et al .
2015 , , 580 , a52 garca - burillo , s. , combes , f. , usero , a. , et al .
2014 , , 567 , a125 garca - burillo , s. , combes , f. , usero , a. , et al .
2015 , , 580 , a35 greve , t. r. , leonidaki , i. , xilouris , e. m. , et al .
2014 , , 794 , 142 hailey - dunsheath , s. , sturm , e. , fischer , j. , et al .
2012 , , 755 , 57 herrero - illana , r. , prez - torres , m. .
, alonso - herrero , a. , et al .
2014 , , 786 , 156 hibbard , j. e. , & yun , m. s. 1996 , cold gas at high redshift , 206 , 47 hollenbach , d. j. , & tielens , a. g. g. m. 1999 , reviews of modern physics , 71 , 173 iguchi , s. , morita , k .-
i . , sugimoto , m. , et al .
2009 , , 61 , 1 imanishi , m. , & nakanishi , k. 2013 , , 146 , 47 iono , d. , yun , m. s. , & mihos , j. c. 2004 , , 616 , 199 iono , d. , yun , m. s. , & ho , p. t. p. 2005
, , 158 , 1 iono , d. , wilson , c. d. , takakuwa , s. , et al .
2007 , , 659 , 283 iono , d. , wilson , c. d. , yun , m. s. , et al .
2009 , , 695 , 1537 kalnajs , a. j. 1978 , structure and properties of nearby galaxies , 77 , 113 kamenetzky , j. , glenn , j. , rangwala , n. , et al .
2012 , , 753 , 70 kamenetzky , j. , rangwala , n. , glenn , j. , maloney , p. r. , & conley , a. 2014 , , 795 , 174 kamenetzky , j. , rangwala , n. , glenn , j. , maloney , p. r. , & conley , a. 2016 , , 829 , 93 kaneko , h. , kuno , n. , iono , d. , et al .
2014 , arxiv:1411.2660 knig , s. , aalto , s. , muller , s. , beswick , r. j. , & gallagher , j. s. 2013 , , 553 , a72 knig , s. , aalto , s. , muller , s. , et al .
2016 , , 594 , a70 kotilainen , j. k. , reunanen , j. , laine , s. , & ryder , s. d. 2001 , , 366 , 439 kaufman , m. j. , wolfire , m. g. , & hollenbach , d. j. 2006 , , 644 , 283 leech , j. , isaak , k. g. , papadopoulos , p. p. , gao , y. , & davis , g. r. 2010 , , 406 , 1364 lin , l .- h . ,
wang , h .- h . ,
hsieh , p .- y .
, et al .
2013 , , 771 , 8 liu , d. , gao , y. , isaak , k. , et al .
2015 , , 810 , l14 lu , n. , zhao , y. , xu , c. k. , et al .
2014 , , 787 , l23 a. lundgren , 2013 , alma cycle 2 technical handbook version 1.1 , alma maloney , p. r. 1999 , , 266 , 207 mao , r .- q . ,
schulz , a. , henkel , c. , et al .
2010 , , 724 , 1336 mashian , n. , sturm , e. , sternberg , a. , et al .
2015 , , 802 , 81 matsushita , s. , iono , d. , petitpas , g. r. , et al .
2009 , , 693 , 56 mccray , r. , & kafatos , m. 1987 , , 317 , 190 mcmullin , j. p. , waters , b. , schiebel , d. , young , w. , & golap , k. 2007 , astronomical data analysis software and systems xvi , 376 , 127 meijerink , r. , kristensen , l. e. , wei , a. , et al .
2013 , , 762 , ll16 meixner , m. , & tielens , a. g. g. m. 1993 , , 405 , 216 michiyama , t. , iono , d. , nakanishi , k. , et al .
2016 , , narayanan , d. , groppi , c. e. , kulesa , c. a. , & walker , c. k. 2005 , , 630 , 269 narayanan , d. , & krumholz , m. r. 2014 , , 442 , 1411 olsson , e. , aalto , s. , thomasson , m. , & beswick , r. 2010 , , 513 , a11 panuzzo , p. , rangwala , n. , rykala , a. , et al .
2010 , , 518 , l37 papadopoulos , p. p. 2010
, , 720 , 226 papadopoulos , p. p. , van der werf , p. p. , xilouris , e. m. , et al . 2012 , , 426 , 2601 papadopoulos , p. p. , zhang , z .- y . , xilouris , e. m. , et al .
2014 , , 788 , 153 pearson , c. , rigopoulou , d. , hurley , p. , et al .
2016 , arxiv:1610.06206 pellegrini , e. w. , smith , j. d. , wolfire , m. g. , et al .
2013 , , 779 , l19 pereira - santaella , m. , spinoglio , l. , busquet , g. , et al .
2013 , , 768 , 55 pereira - santaella , m. , spinoglio , l. , van der werf , p. p. , & piqueras lpez , j. 2014 , , 566 , a49 pereira - santaella , m. , colina , l. , alonso - herrero , a. , et al .
2015 , , 454 , 3679 querejeta , m. , schinnerer , e. , garca - burillo , s. , et al .
2016 , , 593 , a118 rangwala , n. , maloney , p. r. , glenn , j. , et al .
2011 , , 743 , 94 rangwala , n. , maloney , p. r. , wilson , c. d. , et al .
2015 , , 806 , 17 rigopoulou , d. , hurley , p. d. , swinyard , b. m. , et al .
2013 , , 434 , 2051 rosenberg , m. j. f. , van der werf , p. p. , & israel , f. p. 2012
, , 540 , a116 rosenberg , m. j. f. , van der werf , p. p. , aalto , s. , et al .
2015 , , 801 , 72 rothberg , b. , & joseph , r. d. 2004 , , 128 , 2098 saito , t. , iono , d. , yun , m. s. , et al .
2015 , , 803 , 60 saito , t. , iono , d. , xu , c. k. , et al .
2016 , , 68 , 20 saitoh , t. r. , daisaka , h. , kokubo , e. , et al .
2009 , , 61 , 481 sakamoto , k. , aalto , s. , combes , f. , evans , a. , & peck , a. 2014 , , 797 , 90 sanders , d. b. , scoville , n. z. , & soifer , b. t. 1991 , , 370 , 158 sandstrom , k. m. , leroy , a. k. , walter , f. , et al .
2013 , , 777 , 5 schirm , m. r. p. , wilson , c. d. , parkin , t. j. , et al .
2014 , , 781 , 101 schier , f. l. , van der tak , f. f. s. , van dishoeck , e. f. , & black , j. h. 2005 , , 432 , 369 sliwa , k. , wilson , c. d. , petitpas , g. r. , et al .
2012 , , 753 , 46 sliwa , k. , wilson , c. d. , krips , m. , et al .
2013 , apj , 777 , 126 sliwa , k. , wilson , c. d. , iono , d. , peck , a. , & matsushita , s. 2014 , , 796 , l15 solomon , p. m. , & vanden bout , p. a. 2005 , , 43 , 677 spinoglio , l. , pereira - santaella , m. , busquet , g. , et al .
2012 , , 758 , 108 tateuchi , k. , konishi , m. , motohara , k. , et al .
2015 , , 217 , 1 teyssier , r. , chapon , d. , & bournaud , f. 2010 , , 720 , l149 topal , s. , bayet , e. , bureau , m. , davis , t. a. , & walsh , w. 2014 , , 437 , 1434 ueda , j. , iono , d. , yun , m. s. , et al .
2014 , , 214 , 1 visnen , p. , rajpaul , v. , zijlstra , a. a. , reunanen , j. , & kotilainen , j. 2012 , , 420 , 2209 van der tak , f. f. s. , black , j. h. , schier , f. l. , jansen , d. j. , & van dishoeck , e. f. 2007 , , 468 , 627 van der werf , p. p. , isaak , k. g. , meijerink , r. , et al .
2010 , , 518 , ll42 viti , s. , garca - burillo , s. , fuente , a. , et al .
2014 , , 570 , a28 wilson , c. d. , petitpas , g. r. , iono , d. , et al .
2008 , , 178 , 189 wilson , c. d. , warren , b. e. , israel , f. p. , et al .
2012 , , 424 , 3050 wilson , c. d. , rangwala , n. , glenn , j. , et al .
2014 , , 789 , l36 wu , r. , madden , s. c. , galliano , f. , et al .
2015 , , 575 , a88 xu , c. k. , cao , c. , lu , n. , et al .
2014 , , 787 , 48 xu , c. k. , cao , c. , lu , n. , et al .
2015 , , 799 , 11 yao , l. , seaquist , e. r. , kuno , n. , & dunne , l. 2003 , , 588 , 771 zhao , y. , lu , n. , xu , c. k. , et al .
2016 , , 820 , 118 zhu , m. , seaquist , e. r. , & kuno , n. 2003 , , 588 , 243 lcccccccccrl b3 & 2014 jun .
30 & aca 7 m & 95.6 & 10 & 7.3 - 48 & 42 - 170 & j0501 - 0159 & j0423 - 013 & j0427 - 0700 & 49.1 + b3 & 2014 dec .
04 & alma 12 m & 55.7 & 31 & 15 - 290 & 31 - 120 & uranus & j0423 - 0120 & j0423 - 0120 & 13.1 + b3&2014 aug .
30 & alma 12 m & 55.7 & 35 & 28 - 1060 & 50 - 130 & j0423 - 013 & j0423 - 0120 & j0423 - 0120 & 16.9 + b6&2014 aug .
14 & alma 12 m & 27.8 & 35 & 15 - 349 & 70 - 160 & uranus & j0423 - 0120 & j0423 - 0120 & 7.3 @xmath30 3 + b7 & 2012 jul .
14 & alma 12 m & 18.6 & 27 & 18 - 431 & 100 - 210 & callisto & j0522 - 364 & j0423 - 013 & 88.8 + b9 & 2012 aug .
13/28 & alma 12 m & 9.3 & 23 & 20 - 394 & 500 - 1100 & ceres & j0423 - 013 & j0423 - 013 & 49.4 cccccccccccc @xmath0 = 10 ( aca+12 m ) & 115.271 & 5.53 & b3 & 17.6 & briggs & 1.0 @xmath30 0.6 & 1.4 & 295.3 @xmath32 1.8 & 113 @xmath32 26 & 1 , a + @xmath0 = 10 ( 12 m ) & 115.271 & 5.53 & b3 & 4.6 & briggs & 1.0 @xmath30 1.0 & 1.6 & 69.3 @xmath32 1.1 & 27 @xmath32 5 & this work , a + @xmath0 = 21 ( 12 m ) & 230.542 & 16.60 & b6 & 4.6 & uniform & 1.0 @xmath30 1.0 & 3.5 & 325.4 @xmath32 2.7 & 40 @xmath32 9 & this work , b + @xmath0 = 32 ( 12 m ) & 345.796 & 33.19 & b7 & 4.6 & briggs & 1.0 @xmath30 1.0 & 3.0 & 866.8 @xmath32 2.2 & 59 @xmath32 6 & 2 , c + @xmath0 = 65 ( 12 m ) & 691.473 & 116.16 & b9 & 4.6 & briggs & 1.0 @xmath30 1.0 & 8.0 & 826.9 @xmath32 3.2 & 58 @xmath32 10 & 2 , 3 , d
we show the channel map of the @xmath12-matched co ( 10 ) , co ( 21 ) , co ( 32 ) , and co ( 65 ) , and aca - combined co ( 10 ) in figure [ app_cube1 ] , [ app_cube2 ] , and [ app_cube3 ] .
these maps have the same angular resolution of 10 @xmath30 10 , the same velocity resolution of 30 km s@xmath18 , and the same mrs of 45 k@xmath28 ( @xmath29 46 ) , except for the aca - combined co ( 10 ) image .
the lower-@xmath0 three transitions show almost same structures although the co ( 65 ) is only restricted within the central 23 radius . | we present high - resolution ( 10 ) atacama large millimeter / submillimeter array ( alma ) observations of co ( 10 ) and co ( 21 ) rotational transitions toward the nearby ir - luminous merger ngc 1614 supplemented with alma archival data of co ( 32 ) , and co ( 65 ) transitions .
the co ( 65 ) emission arises from the starburst ring ( central 590 pc in radius ) , while the lower-@xmath0 co lines are distributed over the outer disk ( @xmath1 3.3 kpc in radius ) .
radiative transfer and photon dominated region ( pdr ) modeling reveal that the starburst ring has a single warmer gas component with more intense far - ultraviolet radiation field ( @xmath2 @xmath1 10@xmath3 @xmath4 , @xmath5 @xmath1 42 k , and @xmath6 @xmath1 10@xmath7 ) relative to the outer disk ( @xmath2 @xmath1 10@xmath8 @xmath4 , @xmath5 @xmath1 22 k , and @xmath6 @xmath1 10@xmath9 ) .
a two - phase molecular interstellar medium with a warm and cold ( @xmath10 70 k and @xmath1 19 k ) component is also an applicable model for the starburst ring .
a possible source for heating the warm gas component is mechanical heating due to stellar feedback rather than pdr .
furthermore , we find evidence for non - circular motions along the north - south optical bar in the lower-@xmath0 co images , suggesting a cold gas inflow .
we suggest that star formation in the starburst ring is sustained by the bar - driven cold gas inflow , and starburst activities radiatively and mechanically power the co excitation .
the absence of a bright active galactic nucleus can be explained by a scenario that cold gas accumulating on the starburst ring is exhausted as the fuel for star formation , or is launched as an outflow before being able to feed to the nucleus . |
let @xmath3 be a complete discrete valuation ring with finite residue field .
the main result of this note , theorem [ thm:1 ] , is a formula , for each positive integer @xmath0 , for the probability that a random monic polynomial of degree @xmath0 with coefficients in @xmath3 splits over @xmath3 into linear factors . by a standard result ,
@xmath3 is either the ring of formal power series over a finite field , or the ring of integers of a finite extension of the field of @xmath1-adic numbers for some prime @xmath1 . in either case
we write @xmath4 for the residue field of @xmath3 and set @xmath5 .
we encourage the reader to focus on the special cases @xmath6 and @xmath7 $ ] , as the statements and proofs in the general case are essentially the same as in these two special cases . given @xmath8 in @xmath9
, we form the monic ( i.e. leading coefficient 1 ) polynomial @xmath10 .
let @xmath11 be the set of monic degree @xmath0 polynomials with coefficients in @xmath3 , and identify @xmath11 with @xmath9 as above .
let @xmath12 consist of those polynomials that split over @xmath3 into linear factors ( or equivalently , have @xmath0 roots in @xmath3 , counting multiplicity ) .
we identify @xmath13 with a subset of @xmath9 . in section [ secff ] we will similarly write @xmath14 ( resp .
@xmath15 ) for the set of monic polynomials of degree @xmath0 ( resp .
monic polynomials of degree @xmath0 that split completely ) over the residue field @xmath4 .
the compact abelian group @xmath3 has a haar measure @xmath16 of total mass 1 .
give @xmath9 the product topology and product measure @xmath17 ( alternatively , view @xmath9 as a compact abelian group in its own right ) .
one can show that @xmath13 is @xmath17-measurable as follows . given @xmath18
, we can form the polynomial @xmath19 .
the resulting map @xmath20 from @xmath9 to @xmath11 is continuous and has image @xmath13 .
thus , @xmath13 is compact , being the continuous image of a compact set , whence closed and in particular measurable . a main result of this note is a recursive formula for @xmath21 loosely speaking , this is the probability that a random monic polynomial of degree @xmath0 over @xmath3 splits completely over @xmath3 .
let @xmath22 be the order of the residue field of @xmath3 .
the first four values of @xmath23 are : @xmath24 where @xmath25 .
we also give a formula for the probability that a random polynomial that is not necessarily monic of degree @xmath0 splits over @xmath3 .
a student of the first - named author , asher auel @xcite , working independently , has calculated the jacobian determinant of the map @xmath20 , then calculated the resulting @xmath0-fold @xmath1-adic integral .
his approach offers an alternative to the method of this note .
here is our main result .
set @xmath26 . for a sequence @xmath27 of nonnegative integers ,
write @xmath28 for their sum , @xmath29 .
[ thm:1 ] let @xmath3 be a complete discrete valuation ring with finite residue field @xmath4 .
set @xmath30 . for @xmath0 a positive integer , write @xmath23 for the probability that a random degree @xmath0 monic polynomial over @xmath3 splits into linear factors , and set @xmath26 .
if @xmath31 , then @xmath32 note that @xmath23 appears in @xmath22 terms on the right , each time with coefficient @xmath33 but @xmath34 if and only if @xmath35 .
so , as long as @xmath36 , we can solve for @xmath23 in terms of @xmath37 .
here is an approach to @xmath23 using power series . we define @xmath38 and @xmath39 in @xmath40 $ ] by the formulas @xmath41
then theorem [ thm:1 ] gives that @xmath42 we use a method due to euler to get an efficient method for computing @xmath23 .
[ cor:1 ] for @xmath31 , we have @xmath43 in particular , @xmath23 is a rational function in @xmath22 .
for @xmath44 , consider the map @xmath45 that takes the monic degree @xmath0 polynomial @xmath46 to @xmath47 .
thus , @xmath45 maps @xmath48 to itself .
[ lem:0]let @xmath44 .
the map @xmath45 from @xmath48 to itself preserves the measure @xmath17 .
our identification of @xmath11 with @xmath9 makes @xmath45 into an affine map .
for , write @xmath49 with @xmath50
. then @xmath45 is the composition of the linear map @xmath51 and the translation map @xmath52 .
since any translation preserves measure , it is enough to check the linear map @xmath53 .
a basis for @xmath11 is given by the set of monomials @xmath54 for @xmath55 . in terms of this basis , the matrix of @xmath53 is upper triangular with ones on the diagonal .
hence it has jacobian determinant equal to 1 .
we proceed by a series of lemmas .
the basic idea is as follows .
suppose the monic polynomial @xmath46 in @xmath11 splits completely over @xmath3 .
write @xmath56 for the maximal ideal of @xmath3 , and write @xmath57 for the reduction of @xmath58 mod @xmath56 . then @xmath57 splits completely over @xmath59 .
write this factorization as @xmath60 the polynomials @xmath61 are pairwise relatively prime . hence by hensel s lemma and induction , @xmath62 , where @xmath63 in @xmath64 $ ] reduces to @xmath65 in @xmath66 $ ] .
of course , the probability that @xmath58 reduces to @xmath67 ( or to any given degree @xmath0 monic polynomial in @xmath66 $ ] ) is exactly @xmath68 .
the desired result will follow from the following lemma .
[ lem:1 ] 1 .
2 . @xmath70 to prove theorem [ thm:1 ] from the lemma , it suffices to choose a bijection from @xmath4 to the set @xmath71 and to use that @xmath72 and that @xmath73 .
let @xmath74 be a polynomial in @xmath14 .
write @xmath75 for the set of polynomials @xmath58 in @xmath11 whose reduction @xmath57 is @xmath74 . as @xmath75 is a measurable subset of @xmath11 , it inherits a topology from @xmath11 , and also a measure @xmath17 , which we renormalize so that @xmath75 has total measure 1 .
part ( 1 ) of lemma [ lem:1 ] now follows from : [ lem:2](measure theoretic version of hensel s lemma ) .
let @xmath74 and @xmath76 in @xmath66 $ ] be relatively prime monic polynomials of degree @xmath77 and @xmath0 , respectively .
the multiplication map @xmath78 is a bijection , in fact an isomorphism of topological spaces , by the usual hensel s lemma .
then @xmath79 preserves measure , that is , @xmath80 .
the lemma is an immediate consequence of the following , somewhat more general result .
[ lem:3 ] let @xmath81 and @xmath82 be countable products of finite sets .
normalize counting measure so that each of @xmath83 and @xmath84 has total mass 1 , for all @xmath85 , and give @xmath86 and @xmath87 the product measure .
suppose there is a compatible system of bijections @xmath88 between the partial products @xmath89 and @xmath90 .
that is , for each @xmath91 , the map @xmath88 from @xmath92 to @xmath93 is bijective , and for @xmath94 , if @xmath95 in @xmath92 has @xmath96 in @xmath97 as initial string , then @xmath98 has @xmath99 as initial string .
then the map @xmath100 from @xmath86 to @xmath87 is a measure - preserving bijection .
certainly @xmath101 is bijective .
since @xmath101 takes basic open sets to basic open sets of the same volume , it follows that @xmath101 preserves measure .
_ proof of part ( 2 ) of lemma [ lem:1]_. by lemma [ lem:0 ] , it suffices to treat the case @xmath102 .
set @xmath103 we have @xmath58 in @xmath64 $ ] monic with reduction @xmath104 in @xmath66 $ ] .
assume for the moment that @xmath105 does split completely , say as @xmath106 a root of @xmath58 reduces to a root of @xmath57 , so that @xmath83 lies in @xmath56 for each @xmath85 , whence a necessary condition for @xmath58 to split completely is : @xmath107 lies in @xmath108 for each @xmath85 .
the probability that this condition holds is @xmath109 the product of @xmath110 for @xmath55 . set @xmath111
the necessary condition is met if and only if @xmath112 has coefficients in @xmath3 . conditioned on its being met ,
@xmath112 is distributed like a random polynomial in @xmath11 , and the result follows , because @xmath58 splits completely if and only if @xmath112 splits completely .
corollary [ cor:1 ] follows from the following efficient recursive computation of the coefficients of a high power of a known power series which is due to euler @xcite .
the authors are grateful to h. wilf for pointing out lemma [ lem : w ] .
[ lem : w ] let @xmath113 be a field . assume either ( 1 ) @xmath114 and @xmath113 has characteristic 0 , or ( 2 ) @xmath22 is an integer .
let @xmath115 and @xmath116 in @xmath117 $ ] be formal power series with constant term 1 that satisfy @xmath118 then @xmath119 take logarithmic derivatives to get @xmath120 next , cross - multiply to get @xmath121 and equate coefficients of @xmath122 to get @xmath123 there is no harm in including the term @xmath124 in the sum on the left and the term @xmath125 in the sum on the right , since these terms are zero .
now subtract to get equation .
this proves corollary [ cor:1 ] .
recall that @xmath3 is a complete discrete valuation ring with finite residue field @xmath4 , where @xmath30 .
let @xmath0 be a positive integer .
write @xmath126 for the probability that a random monic degree @xmath0 polynomial with coefficients in @xmath4 splits completely .
then @xmath126 , as well as @xmath23 , depends only on @xmath22 and @xmath0 .
our convention is that @xmath127 .
we note some properties of @xmath23 and @xmath128 \1 .
we have @xmath129 . for , if a monic polynomial @xmath58 in @xmath64 $ ] splits completely , then @xmath57 splits completely in @xmath66 $ ] .
we have @xmath130 . for , by the result sometimes called the stars and bars theorem , the number of monic degree @xmath0 polynomials that split completely is @xmath131 .
now use the binomial expansion @xmath132 , and the fact that @xmath133 .
we have @xmath134 .
the second equality follows from observation 2 . for @xmath58 monic in
@xmath64 $ ] , the probability that @xmath57 has a repeated root is at most @xmath135 .
( proof : the set @xmath136 of monic polynomials @xmath58 with constant and linear term in @xmath56 has measure @xmath137 .
hence the union of the sets @xmath138 over a set of lifts @xmath139 of the elements of @xmath4 has measure at most @xmath135 . ) but this tends to zero as @xmath140 .
this proves the first equality .
in this section we drop the assumption that our polynomials are monic .
we identify the set @xmath141 of degree @xmath142 polynomials with coefficients in @xmath3 with @xmath143 .
it gets the measure @xmath144 .
let @xmath145 be those polynomials @xmath46 in @xmath141 that can factored into @xmath0 linear factors , i.e. can be written in the form @xmath146 with @xmath83 , @xmath147 .
one shows that @xmath145 is measurable as follows .
take a pair @xmath148 and @xmath149 of @xmath0-tuples and construct the polynomial @xmath58 as in .
the resulting continuous map from the compact set @xmath150 to @xmath141 has image @xmath145 .
thus @xmath145 is compact , hence closed .
the goal of this section is a formula for @xmath151 .
this may be interpreted as the probability that a random , not necessarily monic , polynomial of degree @xmath142 splits completely into linear factors .
the first four values are : @xmath152 where @xmath153 by convention , we set @xmath154
. we will need the following generalization of lemma [ lem:2 ] .
again it is an immediate consequence of lemma [ lem:3 ] .
let @xmath74 be a ( not necessarily monic ) polynomial in @xmath66 $ ] , with @xmath155 .
write @xmath156 for the set of polynomials @xmath58 in @xmath64 $ ] of degree at most @xmath0 whose reduction @xmath57 is @xmath74 .
the space @xmath156 is a measurable subset of @xmath141 and therefore inherits a topology and a measure .
we renormalize this measure to give @xmath156 total measure 1 . recall that if @xmath76 in @xmath66 $ ] is monic , then we defined @xmath157 to be the set of monic polynomials @xmath58 in @xmath64 $ ] such that @xmath158 .
[ lem:2 nm ] let @xmath74 and @xmath76 in @xmath66 $ ] be relatively prime polynomials of degree @xmath77 and @xmath0 , respectively , and assume that @xmath76 is monic .
let @xmath159 .
then the multiplication map @xmath160 is a measure - preserving homeomorphism .
we will write @xmath161 for the nonzero elements of the residue field @xmath4 .
let @xmath162 be the multiplicity of @xmath163 as a root of @xmath57 , by which we mean that exactly the first @xmath164 leading coefficients of @xmath57 are zero . evidently , @xmath165 .
then if @xmath166 , by hensel s lemma @xmath58 factors uniquely as @xmath167 , where @xmath168 , @xmath169 , @xmath170 reduces mod @xmath56 to a nonzero constant , and @xmath171 is monic .
[ lem:1 nm ] if @xmath172 , then : 1 . @xmath173 @xmath174 2 . @xmath175 .
3 . @xmath176 .
if @xmath177 ( i.e. , if @xmath178 ) , then @xmath179 part ( 1 ) of the lemma follows from lemma [ lem:2 nm ] . after dividing @xmath180 by its leading coefficient , we get a random monic polynomial reducing to @xmath54 . applying part ( 2 ) of lemma [ lem:1 ]
then proves ( 2 ) .
\(3 ) is immediate by definition .
conditioned on @xmath178 , @xmath181 is a random polynomial in @xmath182 .
this proves ( 4 ) .
now write @xmath183
. then @xmath184 is @xmath164 if and only if @xmath185 and ( if @xmath166 ) @xmath186 .
thus , the probability that @xmath184 equals @xmath164 is @xmath187 if @xmath188 , and @xmath189 if @xmath190 .
thus , we have @xmath191 this is easily seen to be equivalent to @xmath192 where @xmath193 here is another equivalent formulation : @xmath194 using euler s method , we get @xmath195 in particular , @xmath196 is a rational function of @xmath22 .
here is a slightly different viewpoint , which leads to a slightly different formulation of the result .
let @xmath197 be the set of lines through the origin in the plane @xmath198 .
such a line is uniquely determined by its slope , which is an element of the set @xmath4 or undefined .
this identifies @xmath197 with @xmath199 ; hence , @xmath197 has cardinality @xmath200 .
assume for the moment that @xmath58 splits completely , and that @xmath57 , its reduction mod @xmath56 , is nonzero .
write @xmath201 then , for each @xmath85 , @xmath202 in @xmath198 is not the origin , so the line through this point and the origin is in @xmath197 . as @xmath85 varies , we get a a set of @xmath0 points in @xmath197 , with multiplicities . for @xmath203 , let @xmath204 be the number of @xmath205 such that @xmath202 lies on @xmath206 .
we have @xmath207 . for any function @xmath208 from @xmath197 to the set @xmath209 of nonnegative integers , set @xmath210 .
reasoning as in lemmas [ lem:1 ] and [ lem:1 nm ] now yields [ nm ] let @xmath3 be a complete discrete valuation ring with finite residue field @xmath4 .
set @xmath30 . for @xmath0 a positive integer , write @xmath211 for the probability that a random degree @xmath142 polynomial over @xmath3 splits into linear factors .
. then @xmath212
in this section we regard @xmath22 as a variable , and define the sequence of rational functions @xmath213 by @xmath214 , and by @xmath215 thus , if we plug in a prime power for @xmath22 we recover the @xmath23 s with their previous meaning .
we note some properties of these rational functions .
[ powerseries ] 1 .
the degree of the numerator of @xmath23 is the degree of the denominator , and @xmath216 .
2 . @xmath23 vanishes at 0 to order @xmath217 .
3 . the only poles of @xmath23 are at roots of unity .
4 . @xmath218 .
5 . if @xmath22 is fixed , @xmath219 and @xmath22 is not an integer then , as @xmath220 , @xmath221 for some complex numbers @xmath222 , @xmath223 .
_ proof of properties ( 1)(4 ) . _
property ( 1 ) follows from observation 3 in section [ secff ] . for ( 4 ) , take the defining equations ( [ def1 ] ) , ( [ def2 ] ) and ( [ def3 ] ) of @xmath23 , and substitute @xmath224 for @xmath225 .
we get @xmath226 which we can rewrite as @xmath227 or as @xmath228 but this is the defining equation for @xmath229 , with @xmath230 substituted for @xmath229 . also , for @xmath231 and @xmath35 , @xmath232
. therefore we must have @xmath233 . for ( 3 ) ,
define @xmath234 by @xmath235 .
we will prove by induction on @xmath0 that the only poles of @xmath236 are at roots of unity .
since @xmath237 , this is clear for @xmath238 . if @xmath239 , multiply corollary [ cor:1 ] by @xmath240 and rewrite it as @xmath241 separating out the terms involving @xmath236 gives @xmath242 where @xmath243 is a polynomial in @xmath22 , @xmath244 , , @xmath245 .
but then @xmath246 , so the result follows from the induction hypothesis .
property ( 2 ) is an immediate consequence of ( 1 ) and ( 4 ) .
this finishes the proof of properties ( 1)(4 ) .
property ( 5 ) will be proved in the remainder of the section . by property
( 3 ) , it makes sense to plug in for @xmath22 any complex number in the complement of the unit disk , @xmath247 . for @xmath248
not a positive integer , we expect that the power series @xmath249 has a complex zero @xmath222 ( necessarily nonzero , as @xmath26 ) which is simple and is the unique zero in the disc @xmath250 ) .
in fact , this is the case if @xmath219 by theorem [ thm : m ] below .
property ( 5 ) follows from this theorem by singularity analysis , with the constant @xmath251 our particular focus is on the asymptotics of @xmath252 , as @xmath253 .
surprisingly , the behavior in @xmath254 is radically different from the behavior in @xmath255 . to explain this , for @xmath256
, we note that the terms do not grow too fast .
[ powseries2 ] for each @xmath256 , there is some positive real @xmath257 such that @xmath258 for all @xmath0 .
since @xmath259 , the series @xmath260 converges , so pick @xmath4 such that @xmath261 , and then pick @xmath262 such that @xmath263 finally , pick @xmath264 large enough so that @xmath265 for @xmath35 ,
, @xmath266 .
we will prove by induction that for all @xmath0 , @xmath267 .
we may assume that @xmath268 .
now for @xmath239 , corollary [ cor:1 ] tells us that @xmath269 however , @xmath270 , so the absolute value of the numerator is bounded by @xmath271 and by our assumption and the induction hypothesis this is less than @xmath272 now the absolute value of the denominator is no smaller than @xmath273 , so to have @xmath267 it will do to have @xmath274 which follows immediately from ( [ eyuni ] ) .
this completes the proof . from this
it follows that the series @xmath249 has an infinite radius of convergence .
if @xmath275 is an integer , then @xmath276 implies that @xmath277 also has an infinite radius of convergence , whence its coefficients are small : @xmath278 for all @xmath279 .
( in fact , much more is true , as we shall see in the next section . ) for @xmath22 nonintegral , on the other hand , property ( 5 ) says that this will not in general be true .
[ powseries3 ] if @xmath280 , @xmath281 for all @xmath0 .
we proceed by induction on @xmath0 . for @xmath231 and @xmath35
the result is trivial .
so let @xmath239 .
corollary [ cor:1 ] then gives the bound @xmath282 and by the induction hypothesis , this is no bigger than @xmath283 so it will do to show that @xmath284 since the left - hand side of this inequality is nonincreasing in @xmath285 , it will do to show that @xmath286 for @xmath287 this can be directly verified .
otherwise , @xmath288 .
now observe that @xmath289 and @xmath290 this completes the proof .
[ thm : m ] if @xmath291 or @xmath219 , @xmath292 has a unique zero in the disc @xmath293 , which is simple .
set @xmath294 .
then @xmath295 has roots at @xmath296 . by looking at the power series for @xmath297 , we can write @xmath298 , where @xmath299 .
plugging in this estimate yields roots @xmath300 and @xmath301 , and since @xmath302 and @xmath303 exactly one of these roots is in @xmath304 . by rouch s theorem , it will do to show that @xmath305 on @xmath306 , where @xmath307 .
let @xmath308 .
we then have @xmath309 now , @xmath310 if @xmath291 , the right - hand side of ( [ eralph ] ) is @xmath311 , and if @xmath219 , it is no more than @xmath312 in any case , it is no more than @xmath313 .
again , we have @xmath314 if @xmath291 , the right - hand side of ( [ eralph2 ] ) is @xmath315 . if @xmath219 , @xmath316 , so it is no more than @xmath317 in any case , it is no more than @xmath318 .
finally , using lemma [ powseries3 ] , we have @xmath319 adding up these inequalities tells us that @xmath320 together with ( [ eralph0 ] ) , this proves the result . as remarked above , property ( 5 )
follows from theorem [ thm : m ] .
we mention some open questions for further research . *
1 . zeroes of the numerator .
* let @xmath321 be the numerator of @xmath23 .
is it true that @xmath321 has no zeroes on the negative real axis ? *
zeroes of @xmath249 .
* let @xmath322 be an integer .
is it true that the zeroes of the function @xmath249 are all real and negative ?
we remark that the zeroes @xmath323 of @xmath249 completely determine @xmath249 as follows . because @xmath39 is an entire function of order zero , i.e.
, for all @xmath324 , @xmath325 as @xmath326 , it is equal up to a constant factor to the weierstrass product @xmath327 .
but then @xmath39 equals this product since @xmath328 . *
3 . asymptotics .
* prove the more accurate asymptotic expansion ( [ zzz ] ) for @xmath329 , for @xmath330 an integer .
in this section @xmath322 is an integer . if @xmath291 , we give the first three terms in the asymptotic expansion of @xmath331 . if @xmath330 , we give the first two terms in the asymptotic expansion of @xmath332 and a guess for its third term .
[ thm : moews ] if @xmath291 , then there is a constant @xmath333 and a continuous function @xmath334 with period 1 such that @xmath335 for all @xmath336 .
( figure [ fig2 ] shows a graph of @xmath334 . )
if @xmath322 is an integer , then there is a constant @xmath337 such that @xmath338 for all @xmath91 , and we conjecture ( conjecture [ c26 ] ) that there is a constant @xmath339 and a continuous function @xmath340 with period 1 such that @xmath341 for all @xmath336 .
the idea of the proof of is this . in the proof of the exact formula for @xmath23
, we looked at a certain infinite tree whose branching was given by the factorization of a polynomial mod the maximal ideal @xmath56 of @xmath3 , mod @xmath342 , etc . in a modification of this method , we use finite trees instead . lemma [ lem : trees ]
expresses @xmath343 as the sum of a certain function @xmath344 over labelled @xmath22-trees ( defined below ) with @xmath0 leaves . by lemma [ blah ] , we will see that for any fixed @xmath22 , the logarithm of the number of labelled @xmath22-trees is @xmath345 .
for example , consider the case @xmath291 . in this case
our labelled @xmath22-trees are the same as binary trees .
the number of binary trees with @xmath0 leaves is given by the catalan number @xmath346 by stirling s formula , this has logarithm @xmath347 if we set @xmath79 to be the maximum value of @xmath348 , we then have @xmath349 thus , to prove , it will be enough to estimate the maximum value of @xmath348 to within @xmath345 .
we will see by lemma [ bal ] that @xmath344 is maximized at a tree we will call the _ well - balanced @xmath22-tree _ , and in subsection [ ss : moews ] we will calculate @xmath344 for the well - balanced @xmath22-tree .
although is a refinement of , its proof is different , and relies on a direct use of , together with an estimate of how much the largest term appearing in contributes to .
a rooted tree is a connected acyclic graph with a distinguished vertex ( the root ) .
we can direct the edges of a rooted tree in a unique way , by directing them away from the root .
the root has in - degree 0 ; all other vertices have in - degree 1 .
the edges emanating from a vertex go to distinct vertices , called the _ children _ of @xmath350 .
a vertex with no children is a _ leaf _ vertex .
a _ @xmath22-tree _ is a rooted tree in which the number of out - edges from each vertex is at most @xmath22 , and is not 1 ( i.e. there is branching at each vertex that is not a leaf . ) for @xmath350 a vertex of the @xmath22-tree @xmath351 , we write @xmath352 for the full subtree whose vertices are @xmath350 and all its descendants .
this is a @xmath22-tree .
a _ labelled @xmath22-tree _ is a @xmath22-tree together with a labelling of its edges with elements of the set @xmath353 so that the out - edges emanating from each vertex have distinct labels .
we list the @xmath22-trees with at most @xmath354 leaves . 1 .
the tree @xmath355 with one vertex and no edges is the unique @xmath22-tree with one leaf . in all other @xmath22-trees ,
a vertex is a leaf if and only if it has total degree 1 . 2 .
the tree @xmath356 with three vertices , a root with two leaf children , is the unique @xmath22-tree with two leaves .
there are @xmath357 ways of labelling this @xmath22-tree .
3 . there are two @xmath22-trees with three leaves ( one if @xmath291 ) .
one , call it @xmath358 , has a root with three leaf children ( if @xmath330 ) . the other ,
@xmath359 , has a root with two children @xmath350 and @xmath360 , with @xmath350 a leaf and @xmath361 .
there are @xmath362 ways of labelling @xmath358 , and @xmath363 ways of labelling @xmath359 .
[ blah ] the number of labelled @xmath22-trees with @xmath364 leaves is at most @xmath365 .
let @xmath366 be the set of @xmath367 symbols : the parentheses @xmath368 and @xmath369 for @xmath370 , plus a dot .
we construct for every labelled @xmath22-tree a distinct string of at most @xmath371 of these symbols , thus constructing an injective map from the set of labelled @xmath22-trees with @xmath364 leaves to @xmath372 we proceed by induction on @xmath364 . for @xmath373 , the tree
@xmath355 corresponds to the dot .
assume @xmath374 , and let @xmath351 be a labelled @xmath22-tree with @xmath364 leaves .
let the root of @xmath351 have @xmath164 children .
for each child @xmath350 of the root of @xmath351 , bracket the sequence corresponding to the subtree @xmath352 with the symbols @xmath368 and @xmath369 , where @xmath375 is the label of the edge from the root to the vertex @xmath350 . by the induction hypothesis
, this uses at most @xmath376 symbols .
the number of labelled @xmath22-trees with @xmath364 leaves is therefore at most @xmath377 equating coefficients of @xmath122 in the defining equations ( [ def1 ] ) , ( [ def2 ] ) and ( [ def3 ] ) of @xmath23 gives us @xmath378 group the @xmath23 terms to get @xmath379 where @xmath380 is the sum over @xmath381 such that @xmath382 for at least two values of @xmath85 .
setting @xmath383 , we have @xmath384 for @xmath385 , set @xmath386 .
we have proved the following .
[ m ] assume @xmath385 .
then , @xmath387 [ remark : t ] we can decompose a @xmath22-tree into its root , plus a subtree @xmath352 for each child @xmath350 of the root of @xmath351 .
this decomposition underlies lemma [ blah ] , and will allow us to interpret lemma [ m ] as a recursion on labelled @xmath22-trees .
write @xmath388 for the number of leaves of the tree @xmath352 . as an easy application of the decomposition in remark [ remark : t ] , we have the following .
let @xmath351 be a @xmath22-tree with more than one vertex .
then the root of @xmath351 is not a leaf , and the number of leaves of @xmath351 equals @xmath389 , where the sum is over the set of children @xmath350 of the root of the tree @xmath351 .
we can now interpret the recursion for @xmath390 in terms of labelled @xmath22-trees .
set @xmath391 .
[ lem : trees]we have @xmath392 where the sum is over labelled @xmath22-trees @xmath351 with @xmath0 leaves , and the product is over all vertices @xmath350 of the tree @xmath351 . for @xmath35
, this follows by our choice of @xmath393 .
the general case follows by remark [ remark : t ] and lemma [ m ] .
we shall write @xmath394 for @xmath395 [ ex : trees ] we recalculate @xmath396 and @xmath397 using lemma [ lem : trees ] . 1 .
we have @xmath398 , so that @xmath399 .
2 . we have @xmath400 .
there are @xmath357 ways of labelling the edges of @xmath356 .
hence @xmath401 .
3 . we have @xmath402 .
there are @xmath362 ways of labelling the edges of @xmath358 ( even if @xmath291 ! ) .
the contribution to @xmath397 from @xmath358 is thus @xmath403 .
we have @xmath404 .
there are @xmath363 ways of labelling the edges of @xmath359 .
hence , the contribution to @xmath397 from @xmath359 is @xmath405 .
add , and this agrees with the formula in section [ intro ] . in subsection [ subsec : rec ] we defined @xmath406 for @xmath385 and @xmath407 .
[ beta]we have @xmath408 as @xmath253 .
the sequence @xmath409 is monotone decreasing .
part ( i ) is clear since @xmath410 tends to 1 as @xmath253 . in order to prove ( ii )
it suffices to show @xmath411 for all @xmath385 .
first consider the case @xmath287 .
we have @xmath412 and this last quantity is equal to @xmath413 next suppose @xmath414 .
then @xmath415 , and we have @xmath416 the last inequality follows since @xmath417 is less than @xmath418 and their sum is @xmath419 . since @xmath420 we have @xmath421 , which implies that @xmath422 for any @xmath423 .
rearranging the terms gives @xmath424 which implies ( taking @xmath425 ) that @xmath426 .
as @xmath427 , the lemma is proved .
let @xmath336 be an integer .
we are interested in @xmath22-tuples of nonnegative integers that sum to @xmath0 , and such that each pair differs by at most one .
we remark that such a @xmath22-tuple @xmath85 exists : write @xmath428 with @xmath429 ( we allow either @xmath430 or @xmath431 ) ; now take @xmath432 and @xmath433 . moreover , if such a @xmath22-tuple contained a value less than @xmath96 ( resp .
larger than @xmath434 ) , all values would be at most @xmath96 ( resp .
at least @xmath96 ) , and the sum would be too small ( resp .
too large ) .
thus all values are @xmath96 or @xmath434 , and the number of each must be @xmath435 ( resp .
@xmath95 ) .
hence the @xmath22-tuple is unique up to order .
[ ilovemevoli ] let @xmath336 be an integer .
there is a unique @xmath22-tree @xmath436 with @xmath0 leaves such that for every vertex @xmath350 of @xmath351 : 1 . if @xmath437 then all children of @xmath350 are leaves .
if @xmath438 then @xmath350 has @xmath22 children and , for any two children @xmath360 and @xmath439 of @xmath350 , @xmath440 and @xmath441 differ by at most 1 . if @xmath35 , then @xmath351 is the unique @xmath22-tree with one ( leaf ) vertex .
if @xmath442 , then @xmath351 is the unique @xmath22-tree with @xmath443 vertices that consists of a root with @xmath0 children all of which are leaves . for @xmath444
, we define @xmath436 by induction on @xmath0 .
write @xmath428 with @xmath445 as in the remarks preceding this lemma . applying property 2 above at the root of @xmath351
, we see that if @xmath351 exists , its root must have children @xmath446 , , @xmath447 such that @xmath448 and @xmath449 .
it now follows from the induction hypothesis that each @xmath450 must equal @xmath451 .
this gives a unique candidate for @xmath351 , namely the @xmath22-tree whose root has precisely @xmath22 children , and such that @xmath452 for @xmath95 of these children @xmath360 and @xmath453 for the remaining @xmath435 children @xmath360 .
it is easy to see that this does indeed satisfy properties 1 and 2 .
we call a @xmath22-tree _ well - balanced _ if it satisfies the two conditions of the lemma . for @xmath351 the well balanced @xmath22-tree with @xmath336 leaves , write @xmath454
. set @xmath455 .
[ balpre ] 1 .
2 . if @xmath239 and we write @xmath428 , @xmath429 , then @xmath457 3 .
the sequence @xmath458 is monotone decreasing .
\(i ) is obvious , and ( ii ) follows immediately from the defining properties of the well - balanced @xmath22-tree and the remarks preceding lemma [ ilovemevoli ] .
we now prove that for all @xmath0 , the sequence @xmath459 is monotone decreasing ; this will prove ( iii ) .
we induce on @xmath0 .
if @xmath460 , the result follows from example [ ex : trees ] , so assume that @xmath461 .
we need to show that @xmath462 . for any @xmath463 , write @xmath464 with @xmath445 ( note that we do not allow @xmath431 ) . by ( ii ) , @xmath465 .
next we calculate @xmath466 .
we have @xmath467 , with @xmath468 , so ( ii ) gives @xmath469 whence @xmath470 we apply this to @xmath471 and @xmath472 .
the desired result now follows from part ( ii ) of lemma [ beta ] , after possibly using the induction hypothesis .
let @xmath473 be the largest tree contribution @xmath474 among @xmath22-trees @xmath351 with @xmath0 leaves .
note that here the labelling of the trees is irrelevant .
we set @xmath475 .
the following lemma says that the well - balanced @xmath22-tree always yields a largest tree contribution .
[ bal ] for all @xmath476 , @xmath477 .
we proceed by induction on @xmath0 . for @xmath231 and @xmath35 ,
the result is obvious , so let @xmath36 .
it will do to prove that @xmath478 .
the contribution from a tree with @xmath0 leaves is at most @xmath479 , where @xmath480 , , @xmath481 are nonnegative integers that sum to @xmath0 , at least two of which are positive .
since at least two @xmath482 s are positive , we have @xmath480 , , @xmath483 ,
so by the induction hypothesis , @xmath484 for all @xmath485 .
it therefore follows that the contribution is no more than @xmath486 now by part ( iii ) of lemma [ balpre ] , @xmath487 let @xmath428 with @xmath429 .
it follows from that is largest when @xmath488 for all @xmath489 , when , by the remarks preceding lemma [ ilovemevoli ] , it equals @xmath490 by lemma [ balpre ] , this equals @xmath491 , so we are done .
[ balcontents ] let @xmath351 be the well - balanced @xmath22-tree with @xmath0 leaves , let @xmath492 , and let @xmath493 , @xmath494 .
then : 1 . if @xmath495 , there are no vertices at distance @xmath4 from the root of @xmath351 .
2 . if @xmath496 , there are @xmath497 vertices at distance @xmath4 from the root of @xmath351 , each of which is a leaf .
3 . if @xmath498 , there are @xmath0 vertices at distance @xmath4 from the root of @xmath351 , each of which is a leaf .
4 . if @xmath499 , there are @xmath500 vertices @xmath501 at distance @xmath4 from the root of @xmath351 , @xmath95 with @xmath502 and @xmath503 with @xmath504 .
we prove the lemma by induction on @xmath4 . if @xmath505 , these results are clear . otherwise : to prove the first claim , suppose that @xmath495
. then by the induction hypothesis , all the vertices at distance @xmath266 from the root of @xmath351 are leaves , and therefore there are no vertices at distance @xmath4 from the root , as desired .
if @xmath496 , then by the induction hypothesis , there are @xmath506 vertices @xmath501 at distance @xmath266 from the root , @xmath507 of which are leaves and @xmath508 of which satisfy @xmath509 , i.e. have 2 children , both leaves ; this gives a total of @xmath497 vertices at distance @xmath4 from the root , all leaves , which is the second claim . if @xmath498 , then write @xmath510 , @xmath511 , @xmath512 . by the induction hypothesis ,
there are @xmath506 vertices @xmath501 at distance @xmath266 from the root , @xmath513 of which satisfy @xmath514 , and @xmath515 of which satisfy @xmath516 . by the properties of the well - balanced @xmath22-tree , the vertices @xmath501 for which @xmath516 must have @xmath517 children , all leaves .
similarly , the vertices @xmath501 for which @xmath514 must have @xmath518 children , all leaves .
this gives @xmath0 vertices at distance @xmath4 from the root , all leaves .
this proves the third claim . finally , if @xmath499 , write @xmath519 , @xmath520 , @xmath521 .
we then have @xmath522 . by the induction hypothesis , therefore , there are @xmath506 vertices @xmath501 at distance @xmath266 from the root , @xmath523 of which satisfy @xmath524 .
since @xmath525 , @xmath526 , so each of these vertices will , by the properties of the well - balanced @xmath22-tree , have @xmath527 children @xmath528 satisfying @xmath529 and @xmath530 children @xmath528 satisfying @xmath531 .
similarly , @xmath532 of the vertices at distance @xmath266 from the root will satisfy @xmath533 and will have @xmath534 children @xmath528 satisfying @xmath529 and @xmath535 children @xmath528 satisfying @xmath531 .
this gives a total of @xmath536 vertices @xmath528 at distance @xmath4 from the root with @xmath529 and @xmath537 vertices @xmath528 at distance @xmath4 from the root with @xmath531 , as desired .
this proves the final claim .
we wish to estimate the contribution from the well - balanced @xmath22-tree . by the previous subsection
, the well - balanced @xmath22-tree gives the largest contribution to the sum .
the bound in theorem [ thm : moews ] will follow from : @xmath538 , the logarithm of the contribution to from the well - balanced @xmath22-tree , is @xmath539 we first treat the contribution @xmath540 to @xmath491 from vertices at distance more than @xmath541 from the root . by lemma [ balcontents ] ,
there are @xmath345 such vertices , and each contributes a factor of @xmath393 to @xmath542 .
thus @xmath543 .
let @xmath4 be an integer such that @xmath544 .
write @xmath545 for the contribution to @xmath491 from the vertices at distance @xmath4 from the root .
write @xmath546 with @xmath547 . by lemma [ balcontents ]
, we have @xmath548 , so that @xmath549 by part ( i ) of lemma [ beta ] .
so , up to terms of order @xmath500 , we have @xmath550\\ & = & -\frac{x+1}2[n+y]\\ & = & -\frac12(n+y + \frac{n - y}{q^k}(n+y))\\ & = & -\frac12(\frac{n^2}{q^k}+n).\\\end{aligned}\ ] ] now sum over @xmath4 such that @xmath544 .
note that @xmath551 is @xmath345 .
we get @xmath552 the lemma is proved .
the bound in theorem [ thm : moews ] is now proved .
we consider a recursion of the form @xmath553 we now show how to solve this recursion if the sequence @xmath554 does not grow too rapidly with @xmath0 .
suppose @xmath555 for @xmath556 .
[ qwerty ] the solution to ( [ recursion3 ] ) has the form @xmath557 where @xmath558 is a continuous function with period 1 .
let @xmath336 .
it is clear from ( [ recursion3 ] ) that , if we set @xmath559 , and let @xmath351 be the well - balanced @xmath22-tree with @xmath0 leaves , then @xmath560 is the sum over all vertices @xmath350 of @xmath351 of @xmath561 . write @xmath562 for the largest integer which is no more than @xmath485 and @xmath563 for @xmath564 .
now define @xmath565 by @xmath566 observe that @xmath567 is continuous and that @xmath568 .
now lemma [ balcontents ] implies that the contribution to @xmath560 from vertices at distance @xmath4 from the root of @xmath351 is @xmath569 , so @xmath570 note that the summand is zero for @xmath571 .
now @xmath572 set @xmath573 or equivalently , @xmath574 it is now clear from ( [ last0 ] ) and ( [ last1 ] ) that @xmath575 .
however , from ( [ last2 ] ) , @xmath576 , and the continuity of @xmath567 , we see that on @xmath577 $ ] , @xmath578 is a sum of continuous functions which are uniformly bounded by a convergent series .
therefore @xmath578 is continuous on @xmath577 $ ] .
the continuity of @xmath558 now follows from the continuity of @xmath578 on @xmath577 $ ] and the fact that , since @xmath579 , @xmath580 .
this concludes the proof .
we can rewrite part ( ii ) of lemma [ balpre ] as @xmath581 empirically , we have observed that a similar recursion appears to hold for @xmath582 .
this is because the main contribution in ( [ eqm ] ) comes when all the @xmath84 s differ by at most 1 . for @xmath291
, we can prove this .
we now prove by induction that @xmath602 for all @xmath0 . for @xmath603
this can be proven by direct computation .
otherwise , fix some @xmath604 , and assume that @xmath605 ,
, @xmath606 . set @xmath607 it is easy to show that ( since @xmath604 ) @xmath608 .
now first suppose that @xmath0 is even , so @xmath609 , @xmath592 .
then from ( [ even1 ] ) , we immediately have @xmath610 we will prove that @xmath611 ; if this is so , then @xmath612 as desired .
however , @xmath613 so we will have @xmath614 provided that @xmath615 i.e. , if @xmath616 since @xmath617 , it will do to have @xmath618 which is true as the expression is positive for @xmath619 and nonincreasing for @xmath620 $ ] .
[ sumit ] fix @xmath22 and some constant @xmath634 , and for nonnegative integers @xmath0 , set @xmath635 then there is some constant @xmath636 such that if @xmath637 , , @xmath638 satisfy @xmath639 , and for @xmath640 , , @xmath22 , we have @xmath641 , then @xmath642 .
it will suffice to prove ( [ bunbun ] ) for large @xmath0 .
take @xmath0 large enough so that @xmath643 .
now @xmath644 summing over @xmath85 , we get @xmath645 by looking at the power series for @xmath646 , we find that @xmath647 , if @xmath648 . by our assumption on @xmath0 , @xmath649 , so @xmath650 .
it follows that the absolute value of the sum on the right - hand side of ( [ ezoe ] ) is bounded , say by @xmath651 , so @xmath652 the result now follows from the definition of @xmath653 and part ( i ) of lemma [ beta ] .
the significance of lemma [ sumit ] is that we can , given a recursion like ( [ recursion1 ] ) or ( [ recursion2 ] ) , subtract off @xmath653 from the unknown to obtain a recursion of the form considered in section [ recursion ] , and then apply theorem [ qwerty ] . applying this to in the @xmath291 case
gives us , and concludes the proof of theorem [ thm : moews ] .
finally , we can not resist remarking that applying lemma [ sumit ] to ( [ recursion1 ] ) and then applying theorem [ qwerty ] to solve the resulting recursion of form ( [ recursion3 ] ) shows that @xmath655 for some continuous @xmath656 of period 1 .
this could be used to give an alternate proof of theorem [ thm : moews ] .
since @xmath491 is the greatest tree contribution to @xmath23 , and since the @xmath657 of the number of labelled @xmath22-trees with @xmath0 leaves is @xmath345 ( lemma [ blah ] ) , it follows that @xmath658 | let @xmath0 be a positive integer and let @xmath1 be a prime .
we calculate the probability that a random monic polynomial of degree @xmath0 with coefficients in the ring @xmath2 of @xmath1-adic integers splits over @xmath2 into linear factors . |
considerable effort is made in the study of the energy loss of hard partons in the hot and dense medium created in ultrarelativistic heavy - ion collisions . in particular medium tomography ( i.e. gaining information about the medium density evolution ) by studying the nuclear suppression factor @xmath0 , the observed transverse momentum spectrum of hard hadrons divided by the scaled expectation from p - p collisions @xcite . however , as argued in @xcite , @xmath0 exhibits very limited sensitivity to the energy loss mechanism or properties of the medium beyond the fact that the quenching is substantial .
if one makes the ansatz that the local quenching power @xmath3 of the medium can be written as a function of the local energy density @xmath4 as @xcite @xmath5^{3/4}$ ] where @xmath6 is evolution proper time , @xmath7 spacetime rapidity and @xmath8 radius of the medium , adjusting the scale parameter @xmath9 parametrizes both the coupling strength of the parton to the medium and uncertainties in the determination of the medium evolution ( see @xcite ) . in @xcite
it was shown that there is about 50% uncertainty in the extraction of the medium quenching power ( given by @xmath9 ) from @xmath0 even for model evolutions which describe bulk soft matter properties ; a 2 + 1d hydrodynamical evolution @xcite , a 3 + 1d hydrodynamical evolution @xcite and a parametrized evolution @xcite .
greater sensitivity to the medium density evolution can hence only be gained in more differential probes such as @xmath10-hadron correlations @xcite , @xmath0 for non - central collisions @xcite or back - to - back hadron correlations @xcite , here we focus on the latter .
we make a monte carlo ( mc ) simulation of hard back - to - back processes in a medium in order to calculate the induced hard hadronic correlation pattern . the model is described in detail in @xcite .
there are three main elements : 1 ) the primary hard pqcd process 2 ) the description of the soft medium and 3 ) the energy loss from hard to soft degrees of freedom coupling 1 ) and 2 ) .
the energy loss for a given parton path is described in a probabilistic language . in order to determine the probability @xmath11 for a hard parton with energy @xmath12 to lose the energy @xmath13 while traversing the medium on its trajectory , we make use of a scaling law @xcite which allows to relate the dynamical scenario to a static equivalent one for each trajectory @xmath14 by calculating @xmath15 as a function of the jet production vertex @xmath16 and its angular orientation @xmath17 .
we set @xmath18 whenever the decoupling temperature of the medium @xmath19 is reached . using the numerical results of @xcite
, we obtain @xmath20 for @xmath21 and @xmath22 as a function of jet production vertex and the angle @xmath17 . from this , the probability distribution of energy loss reflected in single hadron observables ( such as @xmath0 )
can be computed by averaging over all possible production points determined by the nuclear overlap as @xmath23 the averaged energy loss probability underlying dihadron correlations involves a conditional probability of observing a high @xmath1 trigger hadron , hence it is more complicated and must be computed in the full mc simulation .
we do the simulation generating hard pqcd back - to - back events inside the evolving bulk matter model , using eq .
( [ e - omega ] ) to determine @xmath20 for each parton , followed by leading and next to leading fragmentation once the parton emerges from the medium @xcite . in all simulations ,
the scale parameter @xmath9 is adjusted such that the measured @xmath0 is described well , thus we are only interested in information beyond what can be extracted from single hadron observables .
we test six different spacetime evolution scenarios : 1 ) a 2 + 1d bjorken hydrodynamics simulation ( hydrodynamics ) 2 ) a 2 + 1 d bjorken hydrodynamics evolution in which quenching occurs only in the quark - gluon plasma phase ( black core ) , 3 ) a parametrized non - bjorken expansion with box density 4 ) a parametrized non - bjorken expansion with nuclear profile @xmath24 density 5 ) a bjorken expansion with box density and 6 ) a bjorken expansion with @xmath24 density @xcite . in this way
, we represent a selection of different longitudinal and transverse dynamics .
we compare the computed yield of associated hadrons as a function of momentum bin per trigger in the range @xmath25 gev on near and away side with the star data @xcite in fig .
[ f - ypt8 ] .
the near side is described well and there is broad agreement across all different medium models .
the reason is that in all cases @xmath26 is dominated by either transmission without energy loss or complete absorption , the kinematic window to observe shifts in parton energy is very narrow .
however , the probability distribution of experiencing a shift @xmath13 in energy is where tomographical information resides , thus , all scenarios appear similar for a small difference between trigger and associate @xmath1 .
since the away side yield is determined by a different ( conditional ) probability @xmath27 the models exhibit more pronounced differences .
all models fail in the momentum window between 4 and 6 gev to describe the data .
this is presumably due to the fact that pqcd and fragmentation is not the dominant mechanism of hadron production in this momentum regime . within errors ,
all scenarios except the black core one agree with the data above 6 gev associate momentum .
the black core scenario is characterized by a very high quenching power ( 48 gev@xmath28/fm at @xmath6 = 1 fm / c ) in the center and a large halo out of which partons can escape unhindered . on the other hand
, the hydrodynamics scenario shows less quenching power ( 11 gev@xmath28/fm at @xmath6 = 1 fm / c ) but has a larger extension .
we show the origin of dihadrons in each model in fig .
[ f - geo ] .
it is evident that dihadron production in the black core scenario is peripheral while in the hydrodynamics scenario it occurs in the whole medium . from the fact that only one of these scenarios ( hydrodynamics ) describes the data
, we can infer that the produced matter is not characterized by an opaque core and a dilute halo and that dihadron emission is not peripheral but probes indeed the central region of the medium .
while this is some promising start , a larger lever arm in trigger @xmath1 would be needed to be sensitive to shifts in parton energy instead of just absorption and to get more tomographical information @xcite . recently
, elastic energy loss has been suggested as an additional ingredient to describe the nuclear suppression @xcite . to the degree that the magnitude of this contribution is parametrically given by @xmath29 , these results place tight limits on the importance of collisional energy loss : since the integral has virtually no contribution from later times ( long paths ) , elastic energy loss can not resolve spatial structures larger than @xmath30 fm whereas the observed difference between near and away side yield relies on the fact that pathlength differences of the order of 5 fm or more are resolved .
10 m. gyulassy and x. n. wang , nucl . phys .
b * 420 * , ( 1994 ) 583 ; r. baier _ et al .
_ , nucl .
b * 484 * , ( 1997 ) 265 ; b. g. zakharov , jetp lett . *
65 * , ( 1997 ) 615 ; u. a. wiedemann , nucl .
b * 588 * , ( 2000 ) 303 ; m. gyulassy , p. levai and i. vitev , nucl .
b * 594 * , ( 2001 ) 371 ; x. n. wang and x. f. guo , nucl .
a * 696 * , ( 2001 ) 788 . | hard processes are a well calibrated probe to study heavy - ion collisions . however , the information to be gained from the nuclear suppression factor @xmath0 is limited , hene one has to study more differential observables to do medium tomography .
the angular correlations of hadrons associated with a hard trigger appear suitable as they show a rich pattern when going from low @xmath1 to high @xmath1 . of prime interest
is the fate of away side partons with an in - medium pathlength @xmath2(several fm ) . at high
@xmath1 the correlations become dominated by the punchtrough of the away side parton with subsequent fragmentation .
we discuss what information about the medium density can be gained from the data . |
the @xmath0li nucleus is known to have a well developed @xmath5
+ @xmath6 cluster structure and the binding energy is very small @xmath7 mev from the @xmath5-@xmath6 breakup threshold . because of those features , a breakup process of @xmath0li into @xmath5 and @xmath6 is one of the significant mechanisms of the reaction , and the accurate description is required .
as one of the most reliable methods for treating breakup processes , the continuum - discretized coupled channels ( cdcc ) method @xcite has been proposed and successfully applied to analyses of three - body breakup systems , in which the projectile breaks up into two constituents , such as @xmath0li into @xmath5 and @xmath6 .
thus , the cdcc method is expected to be indispensable to analyze @xmath0li breakup reactions . for the cdcc analyses of @xmath0li elastic and inelastic scattering on various targets , sakuragi _ et al . _
@xcite have suggested to reproduce the absolute values of a lot of experimental data of cross sections very well .
they have also resolved the anomalous renormalization problem @xcite in folding model potentials of @xmath0li by showing large dynamical polarization potentials due to breakup processes .
furthermore the cdcc method has been also applied to the sequential coulomb / nuclear breakup via the resonance state of @xmath0li , in which both coulomb and nuclear breakup processes were taken into account consistently @xcite .
the coulomb breakup is of a great interest as the time - reversed reaction of the radiative capture reaction , @xmath5 + @xmath6 @xmath4 @xmath0li + @xmath8 , which is one of a key reaction in the nucleosynthesis in the early universe or during stellar evolution @xcite .
so , it is important that the method could describe not only the nuclear breakup process but also the coulomb breakup one on an equal theoretical footing . in spite of their efforts ,
the cdcc method has not been confirmed yet its applicability to breakup to continuum states of @xmath0li , especially breakup spectra of @xmath0li .
this is a basic interest of the theoretical framework to describe reaction mechanisms relevant to breakup processes . the @xmath1 + @xmath0li reactions are important not only from the basic research interest as shown above but also from the application point of view .
lithium isotopes will be used as a tritium - breeding material in @xmath5-@xmath9 fusion reactors , and accurate nuclear data are required for @xmath1- and @xmath10-induced reactions .
indeed , iaea is organizing a research coordination meeting to prepare nuclear data libraries for advanced fusion devices , fendl-3 , and the maximum incident energy is set at 150 mev to comply fully with the requirements for the ifmif design @xcite .
lithium isotopes are ones of important materials in these libraries . as a matter of fact ,
the neutron interaction with lithium and associated breakup reactions are indispensable in determining the neutron energy spectra in blankets of fusion reactors .
thus the tritium breeding ratio , nuclear heating distributions and radiation damage of structural materials are affected by the @xmath1 + li reactions significantly . in spite of the importance of the @xmath1 + @xmath0li reaction as described above , experimental data leading to @xmath0li continuum breakup processes are extremely rare for the neutron energy region above 20 mev @xcite .
furthermore , the statistical model , used often in evaluation of nuclear data for medium to heavy nuclei , can not be applied to the @xmath0li(@xmath1 , @xmath2 ) reactions . as mentioned in refs .
@xcite , the mechanisms leading to three- or four - body final states , @xmath1 + @xmath5 + @xmath6 and @xmath1 + @xmath1 + @xmath10 + @xmath6 , are important .
therefore more reliable theoretical calculations for the cross sections are highly desirable . in this paper , cross sections for @xmath0li(@xmath1 , @xmath2)@xmath0li@xmath3 @xmath4 @xmath5 + @xmath6 reactions are evaluated by using the microscopic coupled - channels method @xcite , in which we adopt microscopic wave functions of @xmath0li with the @xmath5 + @xmath6 model and the complex jeukenne - lejeune - mahaux effective nucleon - nucleon ( jlm ) interaction @xcite between @xmath1 and @xmath0li .
the calculated elastic , inelastic cross sections and continuum neutron spectra to @xmath0li breakup states show good agreement with the experimental data at low neutron incident energies , where breakup effects of @xmath0li into @xmath5 and @xmath6 are dominant .
furthermore we discuss applicability of the approach used here to the present system for the neutron energy region up to 150 mev .
this paper is organized as follows . in sec .
[ sec : formulation ] , we describe the cdcc method including calculations of @xmath0li wave functions and coupling potentials with the jlm interaction . in sec .
[ sec : results ] and sec .
[ sec : discussion ] , we present the calculated results and discuss the applicability , respectively . finally , we give a summary in sec .
[ sec : summary ] .
the @xmath1 + @xmath0li scattering system is described as a @xmath1 + @xmath5 + @xmath6 three - body model , and the schrdinger equation is written by @xmath11\psi(\mbold{\xi},\mbold{r})&=&0 , \label{hamil-3 } \label{sch - three}\end{aligned}\ ] ] where @xmath12 and @xmath13 represent a coordinate between @xmath1 and the center - of - mass of @xmath0li and a set of internal coordinates of @xmath0li , respectively .
@xmath14 is a kinetic energy operator associated with @xmath12 , @xmath15 is the internal hamiltonian of @xmath0li , and an effective interaction between @xmath16th nucleon in @xmath0li and @xmath1 is represented by @xmath17 .
the total wave function with the total angular momentum @xmath18 and its projection @xmath19 on @xmath20-axis , @xmath21 , is expanded in terms of the orthonormal set of eigenstates of @xmath22 . here
, microscopic wave functions of @xmath0li based on the @xmath5-@xmath6 cluster model are written as @xmath23 where @xmath24 is the relative coordinate between @xmath5 and @xmath6 .
the @xmath5-@xmath6 relative wave function with the angular momentum @xmath25 , total spin @xmath26 and its projection @xmath27 on @xmath20-axis , @xmath28 , can be described as @xmath29_{im},\end{aligned}\ ] ] with the spin of deuteron , @xmath30 .
the internal ground state wave functions of @xmath5 and @xmath6 , @xmath31 and @xmath32 , are assumed as inert core . in the present calculation
, we use the same internal model hamiltonian of the @xmath5-@xmath6 system , @xmath22 , adopted in ref .
@xcite , which are fixed to reproduce the experimental data of the ground state ( -1.47 mev ) and @xmath33-resonance state ( 0.71 mev ) as well as low - energy part of @xmath5-@xmath6 scattering phase shifts .
the ground state is assumed to be a pure @xmath34 state ( @xmath35 ) , and the calculated wave functions well reproduce not only the elastic electron scattering form factor but also the inelastic one to the @xmath33-resonance state @xcite . in this framework , we take into account the totally antisymetrized effect by assuming the orthogonal condition model @xcite . in the cdcc method ,
the relative wave functions of @xmath0li for the ground and continuum states are described as a finite number of discretized states denoted by @xmath36_{im } \,(i=0\mbox{--}n),\end{aligned}\ ] ] whose energies @xmath37 are given by @xmath38 as the discretization approach , we adopt the pseudo - state method @xcite here .
the advantage of the pseudostate method is that if there are resonances in its excitation spectrum , we can describe discretized continuum states with a reasonable number of the basis functions , without distinguishing the resonance states from non - resonant continuous states as mentioned in ref . @xcite . as well known , @xmath0li has three resonances in @xmath39 .
therefore the pseudostate method is very useful for analyses of @xmath0li breakup reactions . in the pseudostate method ,
we diagonalize @xmath15 in a space spanned by a finite number of @xmath40-type basis functions for @xmath41 , @xmath42 , and discretized wave functions for the radial part , @xmath43 are obtained by @xmath44 as the basis functions , the complex - range gaussian basis functions @xcite are adopted , and we include @xmath35 and 2 for the relative angular momentum of @xmath5-@xmath6 : @xmath45 represents the @xmath34-wave state with @xmath46 , and @xmath47 , @xmath48 , and @xmath49 correspond to the @xmath50-wave state with @xmath46 , 2 , and 3 , respectively .
the number of states @xmath51 is decided as involving all open channels for each incident energy .
after the discretization and the truncation of @xmath5-@xmath6 continuum , @xmath21 is reduced to an approximate one , @xmath52 with @xmath53_i \otimes i^l y_{l}(\omega_r)]_{jm}\varphi(\alpha)\varphi(d),\end{aligned}\ ] ] where @xmath54 and @xmath55 represent the elastic channel and breakup channels , respectively . the expansion - coefficient @xmath56 in eq .
( [ appro - expansion ] ) represents the relative motion between @xmath1 and @xmath0li , and @xmath57 is the orbital angular momentum regarding @xmath12 . the relative momentum @xmath58 is determined by the conservation of the total energy , @xmath59 with @xmath60 the reduced mass of the @xmath1 + @xmath0li system .
multiplying eq .
( [ sch - three ] ) by @xmath61^*$ ] from left , one can obtain a set of coupled differential equations for @xmath56 , called the cdcc equation . solving the cdcc equation under the appropriate asymptotic boundary condition
, we can obtain the elastic and discrete breakup @xmath34-matrix elements .
details of the formalism of the cdcc method are shown in ref .
@xcite . for the calculation of diagonal and coupling potentials in the cdcc equation , we use the complex jeukennd - lejeune - mahaux effective nucleon - nucleon ( jlm ) interaction @xcite based on a single folding model .
the jlm interaction can be easily applied to coupled - channels calculations of nucleon - nucleus systems , such as refs .
this interaction has energy and density dependence , and forms a gaussian with both real and imaginary parts ; @xmath62 where @xmath63 is a coordinate of a nucleon in @xmath0li and @xmath1 .
the parameters , @xmath64 , @xmath65 , and @xmath66 , are taken as the same ones used in the original paper @xcite , @xmath67 and @xmath68 .
meanwhile the normalization for the imaginary part , @xmath69 , is optimized to reproduce the elastic cross sections , because the strength corresponding to the loss of flux depends on the model space considered in the calculation .
details for strengths of @xmath70 and @xmath71 are shown in ref .
@xcite . using the jlm interaction , the diagonal and coupling potentials , @xmath72 ,
are obtained by @xmath73 where transition densities , @xmath74 , and averaged matter density , @xmath75 , of @xmath0li between @xmath8 and @xmath76 are defined by @xmath77 and @xmath78 respectively . here
@xmath79 is a coordinate of @xmath16th particle in @xmath0li relative to the center - of - mass of @xmath0li .
-resonance state 0.71 mev above the @xmath5-@xmath6 threshold .
the experimental data are taken from ref .
the data for 14.1 ( 18.0 ) mev are shifted by a factor of 0.1 ( 0.01).,scaledwidth=40.0% ]
figure [ elastic ] shows the differential elastic cross sections of @xmath1 + @xmath0li for incident energies between 7.47 and 24.0 mev .
one sees that the results of the cdcc calculation represented by the solid lines are in good agreement with the experimental data .
the dashed lines represent results of a single channel calculation , in which couplings to breakup states are omitted .
it is found that breakup effects shown by the difference between the dashed and solid lines are significant to reproduce the angular distributions of the elastic scattering . for all incident energies ,
we take @xmath80 to reproduce the data .
the main component of the flux loss from the elastic channel is due to breakups of @xmath0li to @xmath5 and @xmath6 , which can be taken into account in the cdcc calculation directly .
the other effects of the flux loss , such as the excitation of @xmath6 and @xmath5 breakup , are not so large because of low incident energies .
therefore the strength of @xmath69 becomes very small . here
, it should be noted that the single channel calculation can not reproduce the experimental data if any values of @xmath69 are taken .
figure [ inelastic ] shows the angular distributions to the first excited @xmath33 state of @xmath0li for @xmath8111.5 , 14.1 , and 18.0 mev .
the theoretical cross sections are calculated by integrating the breakup cross section to @xmath33-continuum for the resonance energy region .
one sees that the cdcc calculation can also reproduce the inelastic cross section . in fig .
[ neutron - spectrum ] , the calculated neutron spectra are compared with the experimental data at selected angles in the laboratory frame and incident energies .
components of @xmath45 , @xmath47 , @xmath48 , and @xmath49 are represented by the dash - dotted , dashed , dotted and thin solid lines , respectively , and these results are broadened by considering the finite resolution of the experimental apparatus @xcite .
three peaks in the experimental data represent the elastic , inelastic to the @xmath33-resonance , and @xmath82-resonance components , respectively , from higher neutron energies .
the cdcc calculation gives a good agreement with experimental data at the higher neutron energy region .
on the other hand in the low neutron energy region , which corresponds to high exited states of @xmath0li , the calculated cross section underestimates the experimental data for all energies and angles .
the result indicates that experimental data contain contribution from the ( @xmath1 , 2@xmath1 ) reaction corresponding to a four - body breakup reaction @xmath0li(@xmath1 , @xmath83)@xmath6 , as mentioned in ref .
@xcite . in the present calculation ,
the four - body breakup effect is not taken into account directly , and the effect is treated as a absorption effect on the elastic cross section .
+ @xmath0li scattering .
the dotted line represents the experimental data @xcite .
the open circles , squares , and triangles corresponds to results with @xmath80 , 1.0 , and 1.5 , respectively.,scaledwidth=37.0% ]
theoretically the cdcc calculation with the jlm interaction can estimate cross sections of the @xmath1 + @xmath0li scattering for neutron energies up to 160 mev , which is the maximum applicable energy suggested in the original paper @xcite .
however for higher incident energies ( @xmath84 mev ) there is no experimental data of the elastic cross section , which are needed to optimize the normalization factor of the imaginary part .
one of the optimization methods is to fit the neutron total cross section data of @xmath0li , which have been measured up to 560 mev @xcite .
figure [ total ] shows the measured total cross section comparing with the calculated ones with @xmath850.1 , 1.0 , and 1.5 represented by open circles , squares , and triangles , respectively .
one sees that large values for @xmath69 are required with respect to increasing the neutron energy .
although the energy dependence for @xmath69 is rationalized because absorption effects become strong as increasing the energy , the optimized value at 150 mev is much large about @xmath69=1.5 . this normalization problem would be resolved by analyzing @xmath10 + @xmath0li elastic cross sections , which have been measured at higher incident energies , within the present framework .
furthermore for higher incident neutron energies four - body breakup effects of @xmath0li(@xmath1 , @xmath83)@xmath6 are not negligible .
therefore we require the four - body cdcc calculation , which has been successful for analyses of @xmath0he breakup reactions @xcite , in the higher energy region .
we analyze @xmath0li(@xmath1 , @xmath2)@xmath0li@xmath3 @xmath4 @xmath5 + @xmath6 reactions by using the continuum - discretized coupled - channels method with the complex jeukenne - lejeune - mahaux effective nucleon - nucleon interaction
. in the present analysis , it is found that the elastic cross sections for incident energies between 7.47 and 24.0 mev can be reproduced by the present analysis with one normalization parameter for the imaginary part of the jlm interaction ( @xmath80 ) , and breakup effects on the elastic cross section are significant .
furthermore the calculated inelastic cross section to the @xmath33-resonance state and neutron spectra are also good agreement with the experimental data systematically .
thus , the cdcc method with the jlm interaction is expected to be indispensable for the data evaluation of the @xmath0li(@xmath1 , @xmath2 ) reactions , and the advantage is to obtain not only elastic and inelastic cross sections but also neutron spectra within the same framework . for higher neutron incident energies , we have also discussed the applicability of the present microscopic approach .
it is found that the required normalization factor @xmath69 is much large , @xmath86 for 150 mev , from the analyses of the total cross section . in order to investigate the normalization problem
we should analyze @xmath10 + @xmath0li elastic scattering , which have been measured at higher incident energies . furthermore the four - body cdcc calculation is required to describe the four - body breakup process of @xmath0li into @xmath1 + @xmath10 + @xmath6 , which becomes significant for the higher energy .
these results will be reported in a forthcoming paper .
we would like to thank the members of nuclear theory group at hokkaido university .
this work was supported by jsps aa science platform program .
t. fukahori , s. chiba , n. kishida , m. kawai , y. oyama , and a. hasegawa , status of jendl intermediate energy nuclear data , proceedings of the international conference on nuclear data for science and technology , trieste , italy , 1997 ( societa italiana di fisica , ( 1997 ) , p. 899 | we investigate @xmath0li(@xmath1 , @xmath2)@xmath0li@xmath3 @xmath4 @xmath5 + @xmath6 reactions by using the continuum - discretized coupled - channels method with the complex jeukenne - lejeune - mahaux effective nucleon - nucleon interaction . in this study ,
the @xmath0li nucleus is described as a @xmath5 + @xmath6 cluster model .
the calculated elastic cross sections for incident energies between 7.47 and 24.0 mev are good agreement with experimental data .
furthermore , we show the neutron spectra to @xmath0li breakup states measured at selected angular points and incident energies can be also reproduced systematically . |
planetary systems with multiple gas giant planets in long - period circular orbits would mark the long - sought true analogues of our solar system .
the first extrasolar multiple planet system was discovered around upsilon andromedae ( butler et al .
1999 ) and since then , pairs of planets have been detected around hd83443 ( mayor et al .
2000 ) , hd168443 ( marcy et al
. 2001a , udry et al . 2000 ) , gj876 ( marcy et al .
2001b ) , and most recently , 47 uma ( fischer et al 2002 ) .
radial velocity trends among the known planet - bearing stars suggest that additional planetary companions may ultimately be found in more than half of the stars with one detected planet ( fischer et al .
2001 ) .
when first announced in 1996 , thirty radial velocity measurements suggested a single 2.5 m@xmath3 planet in a circular 2.2 au orbit around 47 uma ( butler & marcy 1996 ) .
as described in fischer et al 2002 , 61 subsequent radial velocity measurements ( for a total of 91 velocities ) suggest that a second , less massive planet also exists .
the nominal mass of this second planet is 0.76 m@xmath3 , its eccentricity is small , and its semi - major axis is roughly 3.7 au .
the current best - fit properties of the system ( as reported in fischer et al 2002 ) are shown in table 1 .
of all the multiple systems found to date , 47 uma is by far the most reminiscent of our own solar system .
the star 47 uma is similar in mass , age and rotational velocity to the sun , with @xmath6 , @xmath7 , and @xmath8 ( noyes et al 1984 , baliunas et al .
the accumulated radial velocity measurements suggest that 47 uma bears a startling resemblance to the jupiter - saturn pair in our own solar system .
the orbital period ratios for both systems are quite similar ( 2.38:1 for 47 uma , 2.49:1 for jupiter - saturn ) , and to within the radial velocity measurement errors , the mass ratios are identical ( 3.34:1 for co - planar configurations of 47 uma versus 3.34:1 for jupiter - saturn ) .
the major difference between the two systems is one of overall scale .
the orbital period of 47 uma b is less than one - fourth that of jupiter , and the nominal mass is @xmath9 times greater .
we thus expect that the planets in the 47 uma system will experience more significant mutual perturbations than do jupiter and saturn , and on a shorter timescale .
a dynamical analysis of the properties of the 47 uma system constitutes the main goal of this study .
in particular , we would like to know how dynamical arguments can constrain the parameter space allowed to the system by the radial velocity observations , and we would like to gain a better idea of whether the apparent kinship between the solar system and 47 uma goes beyond a superficial resemblance .
because the outer planet in the 47 uma system has made only two full revolutions since the beginning of the radial velocity observations , the eccentricity , @xmath10 of the outer planet is not well determined .
in fact , @xmath11 provides almost as good a fit to the radial velocity data as does @xmath12 .
furthermore , the inclinations of the orbits to the plane of the sky are unknown , and there are therefore many different configurations of the system which are consistent with the current radial velocity dataset .
one can get an idea of which configurations are allowed for the 47 uma planets by performing numerical integrations of the star 2 planet system , and in particular , we can determine which ranges of mass factors , mutual inclinations and outer planet eccentricities are both consistent with the radial velocity dataset and are also dynamically stable .
this approach has met with considerable success in constraining the properties of the multiple planet system orbiting upsilon andromedae ( see e.g. laughlin & adams 1999 , rivera & lissauer 2000 , lissauer & rivera 2001 , or barnes & quinn 2001 , or gozdziewski et al 2001 ) . in a preliminary dynamical analysis performed by fischer et al 2002 , it was found that during the course of an individual trial integration , the periodic maxima in the osculating eccentricity of the outer planet provide a good running diagnostic of the distance of the system from instability .
once the eccentricity of the outer planet ventures above @xmath13 , the destruction of the system ( via the near - parabolic ejection of the outer planet ) is gererally imminent . ,
@xmath14 , @xmath15 , @xmath16 , @xmath17 , @xmath18 . ]
table 2 shows the outcome of a grid of calculations done for various co - planar configurations of the fit to the system listed in table 1 ( with the co - planar `` mass factor '' given by @xmath19 , where @xmath20 is the inclination of the system with respect to the plane of the sky ) , and with various values for the @xmath21 .
all simulations were started at epoch jd 2451363.5 , the moment of periastron passage for planet c. in each case , the simulation was run for either 100 million years , or until instability occured ( listed as `` unst '' ) .
for stable configurations , we list the largest eccentricity obtained by the outer planet during the course of the simulation .
it is clear that for stability , @xmath22 , but that for small values of @xmath21 , the system planets can be up to six times heavier than their nominal masses , and yet still be dynamically stable .
we note that astrometric limits from the hipparcos satellite rule out mass factors greater than 3 ( fischer et al 2002 ) .
we can compare the results of the integrations with an analytic stability criterion . in the 3-body problem ,
certain values of the energy @xmath23 and angular momentum @xmath24 make it impossible for close encounters to occur between the outer two bodies in the system .
hence in the 47 uma system , close encounters between the two planets can only occur for certain ranges of the initial orbital elements .
in particular , collisions between planets are prevented if @xmath25 where @xmath26 is the total mass of the system , @xmath27 is the mass of the star , @xmath28 , and @xmath29 ( marchal and bozis 1982 , gladman 1993 ) .
when the planets are not in a close encounter , we can express @xmath23 and @xmath24 approximately in terms of the osculating orbital elements of their keplerian orbits about 47 uma .
equation 1 now becomes @xmath30 ^ 2 > s_{\rm crit } \label{eq_stable2}\ ] ] where @xmath31 and @xmath32 .
equation 3 assumes small eccentricities for both planets , and @xmath33 . for the case
in which the planets have coplanar orbits inclined at an angle @xmath20 to the sky , @xmath34 is given by @xmath35 for the values of @xmath36 appropriate to the 47 uma system given in table 1 .
note that the lefthand side of eqn 3 is independent of @xmath20 at this level of approximation .
we have applied the analytic stability criterion to the numerical integrations shown in table 2 .
the simulations above the ragged line are prevented from undergoing planetary close encounters ( that is , they are `` hill stable '' ) by conservation of energy and angular momentum . the simulations below the line are not prevented from having close encounters according to the analytic criterion , although other mechanisms may maintain stability in practice .
the results of the integrations show some interesting disagreements with the analytic criterion .
the threshold for instability tends to occur at lower values of @xmath21 for increasing masses of the planets ( decreasing @xmath37 ) . for low values of @xmath21 , some of the high - mass systems are stable despite the fact that they are not constrained to be so by the analytic criterion .
it is possible that other protection mechanisms are occurring in these cases , but it is also possible that these systems are unstable on very long timescales . as a point of curiosity
, we note that the system represented by the table entry having @xmath38=1.75 , @xmath39 develops instability within several million years .
this point in the phase space of initial conditions is part of a narrow gulley of instability that eats into the table when @xmath40 .
the systems with high @xmath21 are also intriguing .
some of these systems are unstable despite the fact that the stability criterion predicts the planets should not experience close encounters .
a closer examination of these integrations reveals that this prediction is correct .
for example , in an integration with @xmath41 and @xmath42 , the planets never passed within 0.5 au of each other ( equal to about 2.5 hill radii of the inner planet ) , although the outer planet was ejected from the system after 40,000 years . in this case , distant encounters and indirect perturbations ( the effect of the inner planet on the motion of the star ) are responsible for the ejection of the outer planet .
this system and the other unstable cases above the line in table 2 are examples of `` lagrange instability '' in which close encounters can not occur , but escape is still possible .
we note that analytic stability criteria similar to eqn 1 are widely used in the study of hierarchical 3-body systems , but in general these criteria can not establish whether a system is lagrange stable .
as our simulations show , planetary systems that are close to the hill stability limit can be lagrange unstable on short timescales , so one must be careful when applying the hill stability criterion .
it is also worthwhile to point out that the integrations in table 2 do not represent the last word on the stability of the system .
the orbital elements in table 1 all have associated uncertainties , and if alternate initial conditions are drawn from the 1-@xmath43 distributions associated with the parameters , then the stability boundary is subject to change , sometimes with extraordinary sensitivity .
furthermore , the development of instability is a statistical rather than a deterministic process .
if stability is observed over a particular interval for a particular simulation , it does not prove that the configuration is unconditionally stable .
numerical experiments of the sort shown in table 2 must therefore be viewed from a qualitative rather than quantitative standpoint .
figure 1 shows three sets of numerical integrations .
two are of nominal co - planar configurations of the 47 uma system with @xmath44 and @xmath45 respectively .
the third integration is of the jupiter - saturn pair in our own solar system .
the time evolution of the orbital eccentricities and relative longitude of periastron @xmath46 are indicated by the red curves in the figure . in each system , the eccentricities of the planets oscillate in antiphase , as they exchange orbital angular momentum . for 47 uma in the low @xmath21 case ,
the planets lie in a secular resonance in which @xmath47 librates about zero .
is librating .
this should not be confused with the secular resonance in which two of the eigenfrequencies of the laplace - lagrange secular hamiltonian coincide ( kinoshita and nakai 2000 ) .
] this is a stable configuration in which the major axes of each orbit remain roughly aligned , preventing close approaches between the planets . for jupiter and saturn , and for 47 uma in the high @xmath21 case
, @xmath48 circulates through 360 degrees .
this periodically brings the major axes of the orbits into antialignment , at which point the perturbations between the planets are strongest . in these cases shown in figure 1 ,
the system remains stable , but larger initial values of @xmath49 lead to instability ( see table 2 ) .
the black curves in figure 1 , show the results of applying classical laplace - lagrange secular theory to the same systems . in this theory ,
the mutual planetary perturbations are expanded as a fourier series in @xmath49 .
short - period terms , mean - motion resonances , and terms @xmath50 and higher are all neglected , leaving a set of coupled linear differential equations .
following murray and dermott ( 1999 ) , the solution to these equations can be expressed as @xmath51 with @xmath52 , where @xmath53 is the time and @xmath54 , @xmath55 and @xmath56 are constants determined by the masses and initial orbital elements of the planets .
the subscripts 1 and 2 refer to the inner and outer planets , respectively .
the linear secular theory qualitatively reproduces the form of the evolution found in the numerical integrations .
it also predicts the amplitude of the oscillations in @xmath49 , and correctly determines whether or not the planets are librating about the secular resonance .
however , the secular periods predicted by the linear theory are wrong . in the case of jupiter and saturn
, this difference is largely due to the 5:2 near mean - motion resonance between these planets , which is neglected in the secular theory ( murray and dermott 1999 ) .
the discovery of this near - resonance allowed laplace ( 1785 ; see also laskar 1996 ) to identify an additional 900 year periodicity in the orbital elements of jupiter and saturn , which is visible in figure 1 as a high - frequency modulation in saturn s eccentricity .
conversely , for the specific set of osculating orbital elements given in table 1 , the planets in the 47 uma system are not close to any strong resonances ( the nearest is the 7:3 ) , and the shortcomings of the secular theory are probably caused by neglect of higher order terms in the masses and eccentricities . despite its limitations
, we can use the laplace - lagrange theory to understand why the 47 uma planets are in a secular resonance for small @xmath21 but not at larger values , and why jupiter and saturn do not lie in the same resonance . rearranging eqns .
5 , we get @xmath57 since @xmath58 increases linearly with time , the behavior of @xmath48 will depend on a quantity @xmath59 given by @xmath60 when @xmath61 , @xmath48 can take any value and the planets lie outside the secular resonance with @xmath48 circulating . when @xmath62 , @xmath48 is restricted to a particular range of values , and the planets lie in the resonance . in the latter case , librations can take place about 0 ( major axes aligned ) or @xmath63 ( major axes antialigned ) .
figure 2 , shows the extent of the libration regions for the planets in 47 uma , and for jupiter and saturn , as a function of the initial values of @xmath48 and @xmath49 of the outer planet . in the case of 47 uma with low @xmath21
, the planets lie deep inside the secular resonance , with @xmath48 librating about 0 .
however , when @xmath21 is larger ( for the same initial value of @xmath48 ) , the system lies outside the resonance .
jupiter and saturn also lie outside the resonance , but relatively small changes in saturn s eccentricity or @xmath48 would bring the planets into resonance .
( the proximity to the secular resonance is apparent from the form of the @xmath48 curve for jupiter and saturn in figure 1 . ) of the multiple planet systems known to date , two others ( gj 876 and @xmath64 andromedae ) are also likely to be involved in the secular resonance . in both of these systems ,
the libration amplitudes of the outer planets appear to be small ( laughlin & chambers 2001 , chiang , trebachnik , & tremaine 2001 ) , hinting at some sort of dissipative evolution following resonance capture .
by contrast , for the particular parameter values listed in table 1 , 47 uma is undergoing very large librations around the resonance , and jupiter and saturn are out of the resonance .
given the current orbital uncertainties for 47 uma , it is not clear whether this difference is significant , or whether the existence of the possible secular resonance in 47 uma , and the circulation in the jupiter - saturn system is simply a matter of chance .
the doppler radial velocity technique measures only the line - of - sight velocity of the star , and is thus unable to determine the inclinations and nodes of the planetary orbits if the orbits are assumed to be keplerian . in systems such as gj876 ( marcy et al 2001b ) ,
in which the planets are strongly interacting , and where the orbital periods are short enough so that the system can be observed over many cycles , fits to the full n - body motion can potentially determine all of the orbital parameters of the system ( laughlin & chambers 2001 , rivera & lissauer 2001 ) . in the 47 uma system , however , the orbit of the outer planet has been followed with 35 m / s precision for only a single orbital period .
it is presently impossible to obtain the inclinations and nodes from the radial velocity curve itself .
numerical integrations of allowed system configurations can , however , place limits on the system . at a given epoch ,
the orbit of a planet in a multi - planet system is parameterized by six orbital elements ( @xmath65 , @xmath49 , @xmath20 , @xmath66 , @xmath67 , and @xmath68 ) , along with a mass , m. here , the inclination angle @xmath20 is understood to be with respect to the plane of the sky . for a system model containing two planets ,
the overall invariance of the radial velocity curve with respect to system rotations around the line of sight can be accounted for by specifying a parameter @xmath69 corresponding to the relative difference between the lines of nodes for the two planets , yielding , for two planet systems , a total of 13 system parameters .
when fitting to a radial velocity curve , a final parameter corresponding to the velocity zero point is also required .
we generate mutually inclined fits to the 47 uma radial velocity data by randomly selecting fixed values for the inclinations , @xmath70 and @xmath71 , and the angular separation of the lines of nodes , @xmath72 .
the remaining eleven parameters are then obtained by using a levenberg - marquardt minimization scheme driving a three - body integrator to obtain a low @xmath73 fit to the radial velocities ( as described in laughlin & chambers , 2001 ) .
this fitting procedure automatically accounts for the dependence of @xmath74 on viewing geometry ( see chiang , tabachnik , & tremaine , 2001 ) .
each allowed configuration is then integrated for @xmath75 years to check stability . in fischer
et al ( 2002 ) , and in section 2 above , we found that configurations of the 47 uma system in which the osculating eccentricity of the outer planet exceeds @xmath76 are generally dynamically unstable on 100 million year timescales , whereas systems in which @xmath77 are generally stable .
in figure 3 , we plot the results of the orbital integrations of 150 mutually inclined fits to the radial velocity data . the three - dimensional parameter space spanned by @xmath78 , @xmath79 , and @xmath72
is projected onto a two - dimensional space spanned by the mutual inclination , @xmath80 , and the total planetary mass , @xmath81 .
this telescopes the three parameters down to two for easier interpretation .
if a system reaches @xmath82 , it is plotted as a filled circle .
systems that reach @xmath83 are plotted as stars , and stable systems are plotted as open circles .
the figure shows an interesting result that mutual inclination must be less than 45 degrees for the system to be stable , and this result is surprisingly independant of the total mass in the system .
the mass and period ratios of the 47 uma planets are reminiscent of the jupiter - saturn pair in our own solar system .
nevertheless , the 47 uma system packs at least 3.5 jupiter masses into a region which was within the so - called `` snow - line '' of 47 uma s protoplanetary nebula , where temperatures were below 150k , and where water ice grains would have survived evaporation at the time when planetesimals were forming .
it thus seems reasonable to imagine that the 47 uma planets formed in a colder environment , where icy materials were available to build protoplanetary cores , and then migrated inwards .
the phenomenon of planet - disk interaction and migration is now becoming better understood .
in early phases , when the mass of a planet is small , an orbiting planet induces a linear spiral response in its parent disk .
the spiral pattern in the disk material lying inside the planet s orbit exerts a positive torque on the planet , which causes outward migration , whereas the spiral pattern in the outer disk exerts a negative torque on the planet and causes inward migration . if the mass of the planet , @xmath84 , becomes large enough so that the roche radius , @xmath85 is larger than the local scale height @xmath86 of the disk , then the migrating planet can open up a gap if the disk has a sufficiently low viscosity : @xmath87 in a system where two planets have a semi - major - axis ratio of @xmath88 , work by bryden et al ( 2000 ) , and kley ( 2000 ) shows that the ring of gas between the two planets is removed within several hundred orbital periods .
initially , disk material would be present interior to the orbit of the inner planet .
the positive torque exerted by this inner disk is likely to cause outward migration of the inner planet , while the inner disk itself is gradually depleted by accretion onto the central star .
this means that the nascent 47 uma system may have experienced a phase in which the two planets were present with effectively their present masses , and were surrounded by an external disk with little disk mass interior to the inner planet .
we can therefore ask : would the planets then experience interactions which would lead to migration , resonance capture or eccentricity pumping ?
can the present state of the system put interesting limits on the amount of migration that might or might not have occured ? we have modeled a hypothetical forerunner to the 47 uma system using a 2-dimensional hydrodynamical code described by laughlin ( 1994 ) . in the simulation ,
two planets , with masses equal to the nominal masses in table 1 , are placed in circular orbits having an initial semi - major axis ratio , @xmath89 .
the planets are allowed to interact gravitationally with an external circumstellar disk having a total mass of 1 @xmath90 .
most of the mass of the circumstellar disk lies beyond the orbit of the second planet ; the initial surface density profile of the disk has a split - gaussian profile with @xmath91 for @xmath92 , and @xmath93 for @xmath94 , with @xmath95 au , @xmath96 au , and @xmath97 au .
the disk is modeled with a locally isothermal equation of state , in which the scale height of the disk , @xmath98 , is assumed constant .
the inner and outer radial boundaries of the hydrodynamical simulation are located at @xmath99 auand at @xmath100 au .
the initial planetary orbital radii lie at @xmath101 au and @xmath102 au .
although both planets are immersed in the disk , they are not allowed to accrete material which passes within their roche lobes ; the initial disk profile assumes that the outer planet has nearly completed the gap - opening process .
the planets are given a gravitational softening length @xmath103 au in order to minimize erratic local forces arising from individual grid cells .
because of this large softening length , and because we are modeling a phase following gap clearing , material within the roche lobes of the planets is not given special treatment . the planet and the star
are integrated using a 4th order runge - kutta method with a timestep set at 1/20th of the cfl condition which limits the hydrodynamic timestep .
a kinematic viscosity of order @xmath104 is used .
mass which flows across the inner boundary of the hydrodynamical domain is added to the central star .
the simulation employs 256 logarithmically spaced radial zones and 256 evenly spaced azimuthal zones , and was run for 250 orbits of the inner planet .
figure 4 shows an image of the system at the end of the simulation .
the system has settled down to a quasi - steady state configuration in which the outer planet has set up a trailing spiral response in the disk . because the outer lindblad resonance of the inner planet lies in a region of low surface density , there is little coupling between the disk material and the inner planet .
the top panel in figure 5 shows the azimuthal component of the accelerations of the two planets due to the disk material .
after an initial period of readjustment , which is associated with the transient response of the equilibrium disk to the introduction of the two planets , the disk is seen to exert a constant torque on the outer planet , causing it to experience an azimuthal acceleration of @xmath105 .
this causes the outer planet to spiral in on a timescale @xmath106 ( as shown in the second panel from the top of figure 5 ) .
this timescale is consistent with the analytic result that the migration timescale is the viscous timescale of the disk , and this agreement has been previously noted in the simulations of , e.g. , bryden et al ( 2000 ) , and kley ( 2000 ) . the net azimuthal torque on the inner planet averages to near zero .
the inner planet thus shows very little change in semi - major axis ( second panel from the bottom of figure 5 ) . during this independent phase of migration ,
the eccentricities of both planets remain near zero .
the kinematic viscosity @xmath107 used in the simulation corresponds to an `` alpha '' coefficient ( shakura & sunyaev 1973 ) @xmath108 at the radial location of the outer planet .
for the disk under consideration , @xmath109 must be less than @xmath110 for the outer planet to be able to maintain a gap .
the surface density distribution of the disk ( figure 4 ) shows that the outer planet is indeed maintaining a gap in the disk , although some material is lingering in the vicinity of the stable lagrange points @xmath111 and @xmath112 .
the larger mass of the inner planet allows it to easily maintain its gap .
a low mass ring of material between the two planets is replenished by gas flowing through the gap maintained by the outer planet .
the result of the hydrodynamical simulation suggests that migration of the outer planet occurs at a steady rate , and that the disk torque does not directly couple to the inner planet .
we can therefore mimic the effects of the full gas - dynamical calculation by performing much less time - consuming 3-body integrations in which a steady azimuthal acceleration @xmath113 of magnitude suggested by the simulations is applied to the outer planet .
this approach has been applied by lee and peale ( 2002 ) to study the evolution of the gj 876 system , which is now participating in the 2:1 resonance .
additional simulations of the gj 876 configuration by snellgrove , papaloizou & nelson ( 2001 ) have studied resonant capture into the 2:1 resonance using both torqued three body and hydrodynamical approaches .
figures 6 and 7 show the result of one such simulation .
figure 6 is a surface of section in which the ratio of the planets semi - major axes , @xmath114 , is plotted against the orbital phase of the inner planet each time the outer planet reaches aphelion .
mean - motion resonances are visible in this diagram as discrete islands of points in the initial phases of the evolution , the outer planet migrates inward towards the inner planet .
once the planets are locked in resonance , the quantity @xmath114 stops decreasing . in the case shown , the two planets skirt the 5:2 resonance , but are captured into the 7:3 resonance .
once the planets are locked in resonance , they begin to migrate inwards together so as to maintain the 7:3 commensurability .
however , once resonance locking occurs , the amplitude of the secular eccentricity oscillations between the two planets begins to increase , and the eccentricity of the smaller outer planet reaches a value of @xmath115 within 30,000 years . in a simple n - body calculation , which does not include the back reaction of the exterior gas disk , the outer planet eccentricity grows to a point where the system is destabilized .
lee and peale ( 2002 ) have addressed this problem by introducing parameterized eccentricity damping , but fully self - consistent hydrodynamical calculations of the disk response to the resonant migration of the planets need to be done .
the best - fit to the radial velocity data for 47 uma indicates that the system may indeed be in the 7:3 resonance , in which case eccentricity damping is likely to have occured unless the nebula dissipated immediately after the system was captured into resonance .
a longer time baseline of observations will nail down the period of the outer planet more accurately , and further simulations will be done to better elucidate the process of eccentricity damping for planets caught in resonance . in order to gain a better understanding of whether capture into the 7:3 resonance is likely , we have performed a large number of additional three - body integrations in which a constant azimuthal acceleration is applied to the outer planet . for systems in which the planets have initial osculating elements corresponding to those in table 1 , but with @xmath116 , capture into the 3:1 resonance occured in every case we examined .
we have also computed 10,000 simulations in which the outer planet starts with @xmath117 ( i.e. interior to the 3:1 resonance ) , and the initial osculating eccentricity of planet c is chosen randomly to fall between 0.0 and 0.15 .
the results of these simulations are shown in figure 8 .
this figure indicates that capture into the 7:3 resonance is possible , but is much less likely than capture into the 5:2 resonance , which occurs in about 50% of the cases .
( note that many of the symbols in figure 8 overlie one another at the 5:2 resonance . ) in the simulations where the planets avoid being caught in either the 5:2 or the 7:3 , capture into the lower order 2:1 commensurability appears to be certain .
it is clear that evolution into the locked 5:2 or 7:3 commensurabilities depends sensitively on the initial conditions for the integration .
when the eccentricity of the outer planet starts with a small value ( which would be expected from a standard gap - opening scenario ) capture into the 7:3 is observed in 3% of the simulations in which @xmath118 . for large values of the initial eccentricity for the outer planet , capture into higher - order resonances ( e.g. 9:4 , 12:5 , etc ) is occasionally observed .
note that capture into the 7:3 resonance for small values of @xmath21 requires @xmath119 .
when @xmath120 , capture into the 2:1 resonance is certain for @xmath121 .
all of the simulations resulted in some form of resonance capture , which suggests that pairs of resonant planets may be quite common in the galaxy .
although no planetary systems have yet been observed in the 5:2 resonance , this configuration may be particularly ubiquitous .
indeed , lee & peale ( 2002 ) have suggested that jupiter and saturn may once have been participating in a 5:2 resonance , which was subsequently disrupted , perhaps as a result of interactions with a remnant planetesimal or gas disk in the early solar system .
there is still some uncertainty in the period ratio of the 47 uma planets , but the system is clearly interior to the 3:1 resonance .
because capture into 3:1 resonance is virtually certain within our migration scenario , it appears that the planets must have started interior to the 3:1 commensurability .
in the solar system , terrestrial planets formed and survived in a large stable zone between 0 and 2 au from the sun which happens to include the sun s habitable zone ( kasting _ et al . _ 1993 ) . the size and width of this stable zone
is largely determined by the orbits and masses of the giant planets .
most orbits with semi - major axes between 4 and 36 au are unstable due to close encounters with jupiter , saturn , uranus or neptune ( _ e.g. _ gladman and duncan 1990 ) .
many orbits in the asteroid belts ( roughly 24 au and @xmath122 au ) are stable , but these regions contain a number of unstable orbital resonances associated with the giant planets , and it appears that these resonances prevented the formation and/or survival of terrestrial planets in these parts of the solar system ( _ e.g. _ franklin and lecar 2000 , nagasawa _ et al . _ 2000 ,
chambers and wetherill 2001 ) .
the 47 uma system contains 2 giant planets with similar mass and period ratios to jupiter and saturn , and orbits with low eccentricities a characteristic also shared by the giant planets of the solar system .
although the absolute masses and relative inclination of the planets in the 47 uma system are not known at present , it is plausible that these giants could have sculpted a system of asteroid belts and terrestrial planets similar to the solar system , but shifted to smaller semi - major axes reflecting the shorter orbital periods of the giants .
jones _ et al .
_ ( 2001 ) have examined the stability of earth - mass planets in the presence of the inner giant ( _ i.e. _ before the existence of the outer giant had been established ) .
these authors found that for @xmath123 , orbits inside @xmath124 au are stable apart from a narrow unstable zone at the 3:1 resonance at 1.0 au .
for @xmath125 , orbits are stable inside @xmath126 au , and the unstable zone at the 3:1 is wider . in both cases
, stable orbits can exist in a part of the star s habitable zone .
chambers ( 2000 ) made a single simulation of late - stage terrestrial - planet accretion in the presence of the inner giant planet . in this simulation , several small terrestrial planets formed , including a mars - size body moving on a stable orbit at @xmath127 , at the inner edge of the habitable zone .
the presence of the outer giant complicates the dynamics of terrestrial planets . in particular , a strong secular resonance , analogous to the @xmath128 resonance in the solar system , is present at @xmath129 au , and any terrestrial planet forming in this region will have an unstable orbit .
( in the solar system , the @xmath128 resonance marks the inner boundary of the main asteroid belt . )
we have examined the stability of test particles in the presence of the two giant planets using n - body integrations .
the top panel of fig .
9 shows the fates of 280 particles with low-@xmath49 , low-@xmath20 orbits and semi - major axes @xmath130 au , integrated for 20 million years using a hybrid - symplectic algorithm ( chambers 1999 ) . in the figure ,
blue vertical bars indicate the initial semi - major axes of particles that were ejected from the system by close encounters with one of the giant planets .
yellow bars indicate particles that came within 0.1 au of the star , and we assume that these particles would ultimately be removed by colliding with the star .
red bars indicate particles that survived for the length of the integration .
most of the test particles with @xmath131 au are removed , mainly by ejection onto hyperbolic orbits .
this is similar to the result of jones _
et al . _
for the case with just the inner giant planet .
many of the particles with @xmath132 au survived , although there are a number of narrow unstable regions associated with resonances with the giant planets .
the secular resonance at @xmath129 au removed several particles by inducing large orbital eccentricities so that the particles fell into the star .
gaps at the 3:1 and 5:2 mean - motion resonances are also apparent .
two particles survived in the 3:2 resonance at @xmath133 au , although it is unclear whether they would survive on longer timescales . in many ways
the distribution of stable and unstable regions mimics that in the main asteroid belt of the solar system , with kirkwood gaps at the @xmath128 , 3:1 , 5:2 and 2:1 resonances , and the hilda asteroids in a stable region at the 3:2 resonance . most of the test particles with @xmath132 au survived in the simulation
however , this does nt provide a robust indication of whether terrestrial planets can form or survive here .
terrestrial planets have mass and so they modify the strength and location of resonances associated with the giant planets , as earth and venus modify the @xmath134 resonance in the solar system ( namouni & murray 1999 ) . during the late stages of planetary accretion , many planetary embryos with appreciable mass will be present .
these will modify one another s orbits during close encounters as well as causing secular orbital evolution due to more distant perturbations .
in particular , embryos can gravitationally scatter one another into unstable resonance regions , and this greatly enhances the ability of resonances to remove material from a protoplanetary disk ( wetherill 1992 ) .
chambers and wetherill ( 2001 ) have shown that if planetary embryos formed in the asteroid belt of the solar system , there is a two thirds probability that a combination of gravitational scattering and giant - planet resonances would have removed all embryos from the belt .
something similar may have occurred in the region beyond 0.85 au in the 47 uma system . to examine this possibility ,
we have made 4 n - body simulations of terrestrial - planet accretion in the 47 uma system .
we assume that planetary embryos were able to form rapidly by runaway accretion ( wetherill and stewart 1993 ) interior to the orbits of the giant planets .
in addition , since the formation / migration timescales are not known for the giant planets , we make the simplifying assumption that the giants had attained their current orbits and masses at the start of the final stage of accretion of the terrestrial planets . to calculate the orbital and accretional evolution of the embryos we used a hybrid - symplectic n - body integrator ( chambers 1999 ) .
simulation 1 began with 280 lunar - mass bodies ( @xmath135 ) with @xmath136 au and roughly circular , coplanar orbits .
the total initial mass of the embryos is about 2.5 earth masses , which in the solar system is generally sufficient to form a system of 3 or 4 planets with total mass similar to the terrestrial planets ( chambers 2001 ) . in this simulation ,
the giant - planet masses are assumed to have their minimum possible values ( _ i.e. _ we take @xmath137 . )
figure 10 shows 6 snapshots of the evolution in the simulation .
each panel shows the instantaneous values of @xmath65 and @xmath49 for surviving embryos , with symbol radius proportional to the radius of the body . within @xmath138 years
, the inner giant planet has gravitationally ejected most of the embryos with @xmath139 au , and a gap has developed at the secular resonance at @xmath140 au .
after 0.5 million years , the secular resonance is almost empty and embryos are also being dynamically excited and removed at the 3:1 resonance near 1 au . by 2 million years , a large fraction of the initial population of embryos has been lost , while little accretion has taken place . clearly , the timescale to remove material from the unstable parts of the disk is short compared to the accretion timescale . it is also apparent that depletion of material is not confined to the resonance locations , but extends across most of the disk beyond about 0.7 au . as a result
, appreciable accretion is confined to the innermost part of the disk which represents a large contiguous region where resonances are absent . at the end of the simulation ,
this region contains 3 small terrestrial planets , the largest of which has @xmath141 earth masses .
all of these planets lie inside the inner edge of the habitable zone of 47 uma .
a single unaccreted ( lunar mass ) body survives in the habitable zone ( hz ) at @xmath142 au .
the mass of this object simply reflects the initial mass of the embryos used in the calculation , and as such the size of this `` planet '' is not very meaningful .
the middle panel of figure 9 shows the fates of the 280 embryos at the end of simulation 1 , using the same colour coding as the upper panel . in cases where 2 or more embryos merged , each of the embryos
is assigned the same fate as the composite body .
note that the number of surviving objects is much smaller than in the test - particle integration .
in particular , embryos are lost in a wide region with @xmath143 au , and not merely at resonance locations . if planetary embryos did form in the hz of 47 uma , it is unlikely that they would have had a uniform mass .
in fact , runaway accretion of planetesimals , which is thought to lead to the formation of embryos , generally produces a few large objects and many smaller ones ( kokubo and ida 1998 ) . with this in mind
, we made a second accretion simulation which began with 14 mars - sized bodies ( @xmath144 ) and 140 lunar - mass objects .
the total mass and orbital distributions were the same as in simulation 1 .
planetary accretion simulations are notorously stochastic , so we were somewhat surprised that the orbital and accretional evolution in simulation 2 was very similar to the first .
the simulation ended with 3 small planets with @xmath145 au , with masses @xmath146 earth masses , in addition to a single mars - sized embryo with @xmath147 au which had undergone no accretion . in simulations 1 and 2
, the total mass of embryos in the region 02 au is similar to that which would have existed in a solar minimum - mass nebula ( weidenschilling 1977 ) .
however , the giant planets of 47 uma are more massive than jupiter and saturn . in addition
, the giants probably migrated inwards after they formed , which suggests that 47 uma had a more massive protoplanetary disk than the sun . for this reason
, we reran simulations 1 and 2 enhancing the initial masses of the embryos by a factor of 5 ( the giant - planet masses were unchanged ) . in each of these simulations ,
the dynamical evolution progressed through the same stages seen in simulations 1 and 2 .
however , the embryos were typically removed somewhat more quickly since their mutual gravitational perturbations were larger , and they were scattered into unstable resonances more easily . accretion was again confined to the region @xmath145 au , and in these simulations the larger embryos produced fewer and more massive final planets .
simulation 3 ( uniform - mass embryos ) ended with 2 approximately earth - mass planets with @xmath148 and 0.7 au .
the individual fates of the initial bodies in simulation 3 are shown in the bottom panel of figure 9 .
again , objects are removed from a much wider range of orbits than the test - particle case .
simulation 4 ended with a large , 2-earth - mass planet with @xmath149 au , and a second planet with @xmath150 earth masses , and @xmath151 au .
in common with simulations 1 and 2 , simulation 4 ended with a single body in the hz which had undergone no accretion , so that the final mass of this planet reflects the initial masses of the embryos . however , we note that if large embryos did form during the runaway accretion stage , a large unaccreted embryo could exist in the hz of 47 uma , and this body may be large enough to support plate tectonics ( as earth does , but mars does not ) and be considered a habitable planet .
the discovery of a second planet orbiting 47 uma raises a series of very interesting dynamical questions , and can serve to provide important constraints on the formation history of the system .
the 47 uma system is different from the other known multiple - planet extrasolar systems in two important respects .
first , both planets lie at considerable distance from the star .
the period of the inner planet , 47 uma b is five times longer than the inner planet period of any of the other known multiple systems ( currently gj876 , upsilon andromedae , hd 168443 , hd83443 , and hd 82943 ) , and 47 uma c has the longest period of any known extrasolar planet .
second , the orbits of both planets are nearly circular , again in stark contrast to the other known multiple - planet systems .
indeed , 47 uma is currently the best extrasolar analog of our own system .
47 uma b and c are engaged in a secular exchange of eccentricities with a period of roughly 6,000 years .
a first - order theory does a good job of describing this motion , and verifies that the two planets are stabilized by a secular resonance which maintains the periapses of the planets in libration around an aligned configuration .
this secular resonance is also observed in the outer two planets of upsilon andromedae , and is likely a very common state of affairs in the galaxy .
jupiter and saturn currently lie in a different region of secular phase space and do not participate in the secular resonance , although these planets would lie in the resonance for smaller values of @xmath152 or @xmath48 .
the current time baseline of radial velocity data covers only two full periods of 47 uma c , and non - keplerian interactions between the two planets can not be resolved with 35 m / s precision .
the individual and mutual inclinations are therefore unconstrained by the radial velocity measurements .
dynamical calculations show that the system can tolerate mass factors of @xmath153 for co - planar configurations . however , if the system is non - coplanar , stability requires that the mutual inclination be 40 degrees or less , and that this value is quite insensitive to the masses of the two planets .
the small eccentricities of the two planets in the 47 uma system put interesting constraints on the theory of planet - nebula migration .
our results show that when 2-planet systems undergo interaction with an exterior circumstellar disk , the outer planet will migrate inwards until it becomes trapped in a low - order mean motion resonance with the interior planet .
generally , this occurs in the 2:1 resonance , as is the case with gj 876 , but it appears that other resonances , including the 3:1 , 7:3 , and 5:2 can also elicit capture . once two planets are captured in resonance , they migrate inwards together , and their eccentricities increase .
it is not clear how eccentricity damping can occur to the degree that both planets remain on nearly circular orbits .
it seems quite possible that the class of eccentric giant planets discovered to date may be the end result of resonant migration with eccentricity pumping .
once planetary orbits begin to cross , the smaller planet is generally ejected from the system , leaving behind the larger survivor on an eccentric orbit .
this may explain the fact that eccentric giant planets have a larger average mass than the short - period 51 peg type planets , which presumably were able to migrate all the way in without being affected by a second body . a possible scenario for 47 uma might run as follows . the inner planet formed in the outer part of the 47 uma protoplanetary nebula , possibly beyond the ice - line , and then migrated inwards towards its present location at 2 au .
during this migration , a second planet may have formed as a result of density enhancements or vortices beyond the inner edge of the exterior disk .
the growth of this second planet would have led to further gap opening , and an isolation of the inner planet and a halt to its migration .
the remnant disk was then unable to force much further migration of 47 uma c , accounting for its present circular orbit .
had it migrated further , 47 uma c would have been locked in the 2:1 resonance , would have had its eccentricity pumped , and would have escaped , leaving 47 uma b as just another eccentric giant . earth mass planets can survive in the habitable zone of 47 uma , but accretion simulations show that it is unlikely that they could have formed there unless very massive planetary embryos accreted in the habitable zone prior to the formation of the giant planets .
we thank peter bodenheimer , man hoi lee , and stan peale for useful advice .
an anonymous referee provided helpful criticisms .
this material is based upon work supported by nasa under contract no .
rtop 344 - 37 - 22 - 12 issued through the origins of solar systems program .
many of the numerical calculations were performed using the condor cluster run by the nas program at ames research center .
jec acknowledges support from an nrc postdoctoral program and from the nasa origins of solar systems program .
udry , s. , mayor , m. , queloz , d. _ to appear in : _ planetary systems in the universe : observation , formation and evolution .
asp conference series .
ed a. j. penny , p. artymowicz , a. m. lagrange , and s. s. russell .
lrrrr param & 47 uma ( b ) & & 47 uma ( c ) + p ( d ) & 1089.0 ( 2.9 ) & & 2594 ( 90 ) + @xmath154 ( jd ) & 2453622.9 ( 33.6 ) & & 2451363.5 ( 495.3 ) + e & 0.061 ( 0.014 ) & & 0.005 ( 0.115 ) + @xmath66 ( deg ) & 171.8 ( 15.2 ) & & 127.0 ( 55.8 ) + k@xmath155 ( ) & 49.3 ( 1.2 ) & & 11.1 ( 1.1 ) + a ( au ) & 2.09 & & 3.73 + a@xmath156 ( au ) & @xmath157 & & @xmath158 + f@xmath155(m ) ( m@xmath159 ) & @xmath160 & & @xmath161 + m@xmath162 ( m@xmath163 ) & 2.54 & & 0.76 + @xmath164 & & 90 & + rms ( ) & & 7.4 & + @xmath165 & & 1.06 & + | thirteen years of doppler velocity measurements have revealed the presence of two planets orbiting the star 47 ursa majoris on low eccentricity orbits .
a 2-keplerian fit to the radial velocity data suggests that the inner planet has a period @xmath0 d , and a nominal ( @xmath1 ) mass @xmath2= 2.54 m@xmath3 , while the outer planet has a period @xmath4 d , and a mass @xmath2= 0.76 m@xmath3 .
these mass and period ratios suggest a possible kinship to the jupiter - saturn pair in our own solar system .
we explore the current dynamical state of this system with numerical integrations , and compare the results with analytic secular theory .
we find that the planets in the system are likely participating in a secular resonance in which the difference in the longitudes of pericenter librates around zero .
alternately , it is possible that the system is participating in the 7:3 mean motion resonance ( in which case apsidal alignment does not occur ) . using a self - consistent fitting procedure in conjunction with numerical integrations
, we show that stability considerations restrict the mutual inclination between the two planets to @xmath5 40 degrees or less , and that this result is relatively insensitive to the total mass of the two planets .
we present hydrodynamical simulations which measure the torques exerted on the planets by a hypothesized external protoplanetary disk .
we show that planetary migration in response to torques from the disk may have led to capture of the system into a 7:3 mean - motion resonance , although it is unclear how the eccentricities of the planets would have been damped after capture occured .
we show that earth - mass planets can survive for long periods in some regions of the habitable zone of the nominal co - planar system .
a set of planetary accretion calculations , however , shows that it is unlikely that large terrestrial planets can form in the 47 uma habitable zone . |
the authors wish to acknowledge r. m. martin , e. j. mele , p. nelson , m. scheffler , and d. vanderbilt for their comments and valuable suggestions regarding this work , and e. j. walter for his assistance with the numerical calculations .
this work was supported by nsf grant dmr 97 - 02514 and the air force office of scientific research , air force materiel command , usaf , under grant number f49620 - 00 - 1 - 0170 .
amr would like to thank the camille and henry dreyfus foundation for support .
computational support was provided by the national center for supercomputing applications and the san diego supercomputer center .
h. ibach , surf .
* 29 * , 193 ( 1997 ) .
r.j . needs and m.j .
godfrey , phys .
* t19 * , 391 ( 1987 ) .
needs , m.j .
godfrey , and m. mansfield , surf .
sci . * 242 * , 215 ( 1991 ) .
v. fiorentini , m. methfessel , and m. scheffler , phys .
71 * , 1051 ( 1993 ) .
bach , m. giesen , and h. ibach , phys .
lett . * 78 * , 4225 ( 1997 ) .
a. filippetti and v. fiorentini , phys .
b. * 60 * , 14366 ( 1999 ) .
alerhand , d. vanderbilt , r.d .
meade , and j.d .
joannopoulos , phys .
lett . * 61 * , 1973 ( 1988 ) .
s. narasimhan and d. vanderbilt , phys .
lett . * 69 * , 1564 ( 1992 ) . c. mottet , g. trglia , and b. legrand , phys .
b. * 46 * , 16018 ( 1992 ) .
i. meunier , g. trglia , j .-
m . gay , and b. aufray , phys .
b. * 59 * , 10910 ( 1999 ) .
meade and d. vanderbilt , phys .
lett . * 63 * , 1404 ( 1989 ) .
m. mavrikakis , b. hammer , and j.k .
nrskov , phys .
lett . * 81 * , 2819 ( 1998 ) .
e. schrdinger , ann phys .
( leipzig ) * 82 * , 265 ( 1927 ) .
nielsen , and r.m .
martin , phys .
b. * 32 * , 3780 ( 1985 ) .
d. passerone , e. tosatti , g.l .
chiarotti , and f. ercolessi , phys .
b * 59 * , 7687 ( 1999 ) .
a. dal corso , and r. resta , phys .
b. * 50 * , 4327 , ( 1994 ) .
folland , phys .
b. * 34 * , 8296 , ( 1986 ) .
p. ziesche , j. grfenstein , and o.h .
nielsen , phys . rev .
b. * 37 * , 8167 ( 1988 ) .
godfrey , phys .
b. * 37 * , 10176 ( 1988 ) .
p.chen , t. egami , and v. vitek , phys .
b. * 37 * , 2440 ( 1988 ) .
ramer , e.j .
mele , and a.m. rappe , ferroelectrics * 206 - 207 * , 31 ( 1998 ) .
a. filippetti and v. fiorentini , phys .
b. * 61 * , 8433 ( 2000 ) .
n. chetty and r.m .
martin , phys .
b. * 45 * , 6074 ( 1992 ) .
l. mistura , j. chem . phys . * 83 * , 3633 ( 1985 ) ; inter . j. thermophys .
* 8 * , 397 ( 1987 ) .
fung , _ foundations of solid mechanics _ , ( prentice - hall , englewood cliffs , 1965 ) , pp .
landau and e.m .
lifshitz , _ theory of elasticity _ , 3rd ed .
( butterworth - heinemann , oxford , 1986 ) , pp . 1 - 3 .
p. hohenberg and w. kohn , phys . rev . * 136 * , b864 ( 1964 ) .
w. kohn and l. j. sham , phys .
rev . * 140 * , a1133 ( 1965 ) .
mermin , phys . rev . * 137 * , a1441 ( 1965 ) .
h. hellmann , _ einfhrung in die quantenchemie _
( deuticke , leipzig , 1937 ) , pp .
61 and 285 .
feynman , thesis , massachusetts institute of technology , 1939 ( unpublished ) ; phys . rev . *
56 * , 340 ( 1939 ) .
landau and e.m .
lifshitz , _ the classical theory of fields _ , 4th ed .
( pergamon , oxford , 1975 ) , pp .
270 - 273 .
all calculations were performed using the lda functional of perdew and zunger [ j. perdew , a. zunger , phys .
b. * 23 * , 5048 ( 1981 ) ] , a plane - wave basis set with a cutoff of 300 ry , and a 512 k - point sampling of the brillouin - zone .
details regarding efficient implementation of the stress field into plane - wave calculations will be presented elsewhere .
k. nagao , t. takezawa , and h. nagara , phys .
b. * 59 * , 13471 ( 1999 ) .
e. kaxiras , j. broughton , and r.j .
hemley , phys .
* 67 * , 1138 ( 1991 ) .
b. edwards , n.w .
ashcroft , and t. lenosky , europhys .
lett . * 34 * , 519 ( 1996 ) .
calculations for both structures were performed in an orthorhombic unit cell with parameters @xmath55 , @xmath56 , and @xmath57 with @xmath58 and @xmath59 .
all atomic forces were relaxed to @xmath60 5 mev / . | we present a derivation of the stress field for an interacting quantum system within the framework of local density functional theory .
the formulation is geometric in nature and exploits the relationship between the strain tensor field and riemannian metric tensor field .
the resultant expression obtained for the stress field is gauge - invariant with respect to choice of energy density , and therefore provides a unique , well - defined quantity . to illustrate this formalism , we compute the pressure field for two phases of solid molecular hydrogen .
the stress , or the energetic response to deformation or strain , plays an important role in linking the physical properties of a material ( e.g. strength , toughness ) with the behavior of its microstructure .
in addition , the spatial distribution of stress is an invaluable tool for continuum modeling of the response of materials .
the stress concept has been applied at atomistic scales as well . over the last fifteen years , there has been a continuing trend toward understanding various structural and quantum - mechanical phenomena in materials in terms of their response to stress @xcite .
for example , the residual stress at equilibrium has been used to assess the structural stability of systems containing surfaces or strained interfaces .
it has been demonstrated that the desire to minimize surface stress can give rise to reconstructions on high symmetry surfaces @xcite , and the stability of epitaxially grown bimetallic systems has been attributed to the formation of incommensurate overlayers , defects , and dislocations which minimize the stress near the metal - metal interface @xcite .
the stress can have significant effects on chemical reactivity as well .
it has been shown that small molecule chemisorption energies and reaction barriers on certain strained metal and strained semiconductor surfaces are quite different from those on the unstrained surface @xcite .
formally , studies of the above phenomena must include a quantum - mechanical description of the system s electronic degrees of freedom .
therefore , one must consider how a stress is defined quantum mechanically .
methods for calculating the stress in quantum mechanical systems have been developed since the birth of quantum theory itself @xcite .
however , research in developing formalisms for determining the quantum stress in solid - state systems has recently been revitalized .
this is mainly due to ever - increasing opportunities to perform accurate and efficient quantum - mechanical calculations on systems which exhibit stress mediated phenomena .
the stress is a rank - two tensor quantity , usually taken to be symmetric and therefore torque - free .
two useful representations of the stress tensor are the volume - averaged or total stress , @xmath0 , and the spatially varying stress field @xmath1 .
the two representations are related since the total stress for a particular region in a system is the stress field integrated over the volume .
nielsen and martin developed a formalism for calculating the total quantum stress in periodic systems @xcite .
they define the total stress as the variation of the total ground - state energy with respect to a uniform scaling of the entire system .
this uniform scaling corresponds to a homogeneous or averaged strain over the entire system .
they further demonstrate that the total quantum stress is a unique and well - defined physical quantity .
their formulation has been successfully implemented to study a variety of solid state systems @xcite .
other formalisms for determining the total quantum stress have been created as well @xcite . although these formalisms have provided important tools for studying quantum stress , the stress field is a more useful quantity that contains important information regarding the distribution of the stress throughout the system .
a knowledge of the spatial dependence of the quantum stress is vital if one wishes to predict the spatial extent of structural modifications or understand phenomena at interfaces in complex heterogeneous systems .
however , certain definitions of the quantum stress field suggest that it can only be specified up to a gauge .
( this ambiguity manifests itself in classical atomistic models as well . )
it has therefore been asserted that the quantum stress field is not a well - defined physical quantity , even though physical intuition may suggest otherwise .
a traditional way to develop a quantum stress field formalism is to consider the stress field s relationship with the force field . from this perspective
, the stress field can be defined as any rank - two tensor field whose divergence is the force field of the system : @xmath2 ( note that the einstein summation convention for repeated indices is used throughout the letter . ) one can add to @xmath3 a gauge of the form @xmath4 where @xmath5 is any tensor field antisymmetric in @xmath6 and @xmath7 , and recover the same force field , thereby demonstrating the non - uniqueness of this stress field definition .
general formulations for computing non - gauge - invariant stress fields in quantum many - body systems have been derived by nielsen and martin , folland , ziesche and co - workers , and godfrey @xcite .
there have been several attempts to overcome this problem of non - uniqueness .
for example , the stress field formalism of chen and co - workers has been applied to numerous solid state systems to determine the local pressure around a region @xcite . however
, their method assumes that the potential is pair - wise only .
several _ ab - initio _ quantum stress field formulations have been developed , as well .
ramer and co - workers developed a method to calculate the resultant stress field from an induced homogeneous strain @xcite .
they incorporate the additional constraint that the field must be the smoothest fit to the ionic forces .
this method can not be used to calculate the residual stress field at equilibrium , nor can it determine the energy dependence on strains which do not have the periodicity of the unit cell .
filippetti and fiorentini developed a formulation of the stress field based on the energy density formalism of chetty and martin @xcite .
since this formulation is based explicitly on the energy density , which is not gauge - invariant , the resultant stress field is not unique .
mistura succeeded in developing a general gauge - invariant formalism for pressure tensor fields of inhomogeneous fluids within classical statistical models using a riemannian geometric approach @xcite .
this letter extends mistura s work , developing a riemannian geometric formalism for computing gauge - invariant stress fields in quantum systems within the local density approximation ( lda ) of density functional theory ( dft ) .
we show that the response of the total ground - state energy of a quantum system to a local spatially varying strain is a unique and physically meaningful field quantity which can be determined at every point in the system . using a procedure well known in continuum theory
( see for example ref .
@xcite ) , one can formally relate a riemannian metric tensor field @xmath8 with the strain field @xmath9 : @xmath10 , with @xmath11 where @xmath12 @xcite . here
the strain field is defined in terms of a vector displacement field @xmath13 which maps coordinates @xmath14 in the non - deformed system , to the coordinates @xmath15 in the deformed system .
the stress field @xmath16 and strain field obey a virtual work theorem expressing the energy response to variations in the strain : @xmath17 where @xmath18 is the determinant of @xmath19 .
it can be shown that the stress field is related to the functional derivative of the energy with respect to the metric field @xcite : @xmath20 we now derive the quantum stress field of a many - electron system in the presence of a fixed set of classical positive charged ions using local density functional theory @xcite .
the ground state electronic charge density of the system is written as @xmath21 , where @xmath22 are single - particle orthonormal wavefunctions .
for this derivation , we assume orbitals with fixed integer occupation numbers .
the extension to metals with fermi fillings is straightforward , simply necessitating use of the mermin functional instead of the total energy @xcite .
the total charge density of the system can be written as a sum over all ionic charges and @xmath23 : @xmath24 where @xmath25 is the charge of the @xmath26-th ion located at position @xmath27 , and the presence of @xmath28 insures proper normalization of the delta function .
the energy of the system can be written as the following constrained functional : @xmath29 here @xmath30 is the single particle kinetic energy , @xmath31 is the classical coulomb interaction between the total charge density and itself , and @xmath32 is the exchange - correlation energy of the electrons .
the appearance of the last term in eq .
[ totale ] is due to the orthonormality constraint of the orbitals .
( we choose a unitary transformation on @xmath33 which enforces orthogonality . )
one can express @xmath34 as an integral over an energy density @xcite .
the choice of energy density gauge will not affect the derived stress field , since all our results depend only on the total energy @xmath34 . for convenience
, we express the energy terms in eq . [ totale ] as the following : @xmath35 where @xmath36 is the electric field due to the coulomb potential @xmath37 generated by @xmath38 , and @xmath39 is the lda exchange - correlation energy density . to obtain the electronic ground - state energy ,
we require @xmath40 with the additional constraints of a fixed metric ( @xmath41 ) and a fixed ionic charge density ( @xmath42 ) .
this implies that the orbitals must obey the euler - lagrange equations @xmath43 which can be considered the kohn - sham equations in curvilinear coordinates .
also , a least - action principle for @xmath44 requires that @xmath38 and @xmath37 obey the poisson equation : @xmath45 we now vary the total energy with respect to the metric .
it can be proven that we do not need to consider variations in the electronic wavefunctions , charge density and potentials , since all such variations would vanish due to eq .
[ e - l ] and eq.[poisson ] .
this is the same principle used in the derivation of the hellmann - feynman force theorem and the energy - momentum tensor ( the variation of the action with respect to metric ) in general relativity @xcite .
if a different electromagnetic gauge is chosen , variations with respect to the potential @xmath37 would have to be computed explicitly ; this change has no impact on the stress field in eq .
[ stressfield ] . performing the variation of the total energy with respect to the metric
gives the stress field in local density functional theory as : @xmath46 \right ) , \label{stressfield}\end{aligned}\ ] ] where we have used the relation @xmath47 .
using eq .
[ e - l ] , we can rewrite eq .
[ stressfield ] as @xmath48 \right ) \right \ } , \label{stressfield}\end{aligned}\ ] ] when eq .
[ stressfield2 ] is evaluated at the euclidean metric ( @xmath49 ) , it gives the stress field at zero applied strain , and the @xmath50 are then solutions to the standard kohn - sham equations . from here on , we will refer to @xmath51 with an implied evaluation at the euclidean metric .
it is important to note several key features of the form of @xmath51 .
first , the coulombic contribution to the quantum stress field is equivalent to the classical maxwell stress field .
this coulombic term can be obtained by filippetti and fiorentini s formalism in ref .
@xcite if one chooses the maxwell gauge .
also , the contribution of the exchange - correlation energy to our stress field is only in the diagonal ( pressure - like ) terms , which is the proper behavior for local density functionals @xcite and is identical to the exchange - correlation stress derived in ref.@xcite .
however , the kinetic contribution to eq .
[ stressfield2 ] is unique to our derivation .
our stress field contains diagonal terms which are similar to the symmetric and antisymmetric kinetic energy densities . by integrating the stress field over all space ,
we can obtain the total stress @xmath0 : @xmath52 which is identical to the expression derived by nielsen and martin@xcite . in order to demonstrate the utility of our stress field formalism
, we compute within dft the pressure field , @xmath53 , for two phases of solid molecular hydrogen under external hydrostatic pressure of 50 gpa .
@xcite .
both structures consist of stacked two - dimensional triangular lattices of hydrogen molecules , with the molecular axis parallel to the stacking direction and a repeat unit of two layers .
the _
m_-hcp structure has alternating layers shifted so that each hydrogen molecule is directly above triangular hollow sites in the neighboring layers .
the second structure ( belonging to the _ cmca _ space group ) has a different shift , so that each molecule lies directly above midpoints between nearest - neighbor pairs of molecules in adjacent layers @xcite .
the energetics and electronic properties of both structures have been studied extensively from first principles @xcite .
examination of the pressure field permits us to rationalize the energy ordering of these structures .
the _
m_-hcp structure is energetically favored by 60 mev / molecule .
figure [ fig1]a shows a contour plot of the pressure field for the _ cmca _ structure .
the pressure is tensile ( greater than zero ) through the interstitial region , indicating that contraction is locally favorable .
the pressure is greatest in the volume directly above and below each molecule , averaging 3 ev / @xmath54 .
this implies that the system would energetically favor increased intermolecular coordination . in figure
[ fig1]b we show a similar plot for the _ m_-hcp structure .
again the pressure within the interstitial region is tensile .
however , the pressure field has significantly rearranged , and the pressures in the regions above and below molecules have been reduced to approximately 2.25 ev / @xmath54 .
it is also clear from the charge density plots ( figure [ fig1]c and [ fig1]d ) that the reduction in pressure is correlated with an increase in bonding between molecules and increased charge delocalization .
the pressure fields within the molecules are compressive , and they are several orders of magnitude larger than the interstitial features . however , because these large fields are very similar in both structural phases ( and the free molecule ) , they are not important for understanding relative phase stability .
thus , changes in the pressure stress field highlight regions and charge density features that contribute to favorable energetic changes .
we have developed a formulation for unambiguously determining the stress field in an interacting quantum system described by local density functional theory .
the resultant expression for the stress field is gauge - invariant with respect to choice of energy density and can be obtained via a method analogous to the computation of hellmann - feynman forces .
our application of this formalism to solid molecular hydrogen demonstrates that a stress field analysis can associate energetics with particular microscopic structural features of a material .
this stress field formulation has the potential to become an invaluable aid to the understanding of structural phenomena in complex solid - state systems . |
the study of the interaction of spin with inertia and gravity has received a strong impulse from the work of bahram mashhoon @xcite .
his work has stimulated the research on which we report below .
covariant wave equations for scalar and vector bosons , for spin-1/2 fermions @xcite and spin-2 @xcite particles can be solved exactly to first order in the metric deviation .
the background gravitational and inertial fields appear in the solutions as phase factors multiplying the wave function of the corresponding field - free equations .
the phases can be calculated with ease for most metrics .
we summarize the solutions for vector , tensor bosons and fermions in sections ii , iii and iv . in the same sections we also extract the spin - gravity interaction @xcite from the gravity - induced phases .
the optics of the particles is derived in sections v. in sections vi - viii we discuss the relevance of the mashhoon coupling to muon @xmath0 experiments , discrete symmetries and neutrino helicity transitions . the conclusions are contained in section ix .
photons in gravitational fields are described by the maxwell - de rahm equations @xcite @xmath1 which reduce to maxwell equations @xmath2 when @xmath3 , or when , as in lensing , the wavelength @xmath4 of @xmath5 is much smaller than the typical radius of curvature of the gravitational background . in ( [ a ] ) and ( [ b ] )
@xmath6 indicates covariant differentiation .
we use units @xmath7 . in what follows
we consider only ( [ b ] ) and its generalization to massive , charge - less , spin-1 particles @xmath8 in the hope that among the atoms and molecules to be used in interferometry some indeed satisfy ( [ c ] ) @xcite .
following previous work @xcite-@xcite , we show below that equations ( [ b ] ) and ( [ c ] ) can be solved exactly to first order in the metric deviation @xmath9 , where @xmath10 and @xmath11 is the minkowski metric of signature -2 . to first order in @xmath12 , ( [ b ] ) and ( [ c ] ) become @xmath13 where ordinary differentiation of a quantity @xmath14 is equivalently indicated by @xmath15 or @xmath16 . in deriving ( [ d ] ) and
( [ e ] ) , we have used the lanczos - de donder gauge condition @xmath17 in the massless case , the field @xmath18 satisfies the condition @xmath19 it is convenient to impose ( [ g ] ) also in the case of a massive particle .
equations ( [ b ] ) and ( [ c ] ) can be handled simultaneously .
their solution is @xmath20 \nonumber \\ & & + \frac{1}{2}\int_{p}^{x}dz^{\lambda } \gamma_{\alpha\lambda}(z ) \partial^{\alpha}a_{\mu}(x)\nonumber \\ & & -\frac{1}{2}\int_{p}^{x}dz^{\lambda}(\gamma_{\mu\lambda , \sigma}(z)- \gamma_{\sigma\mu,\lambda}(z)-\gamma_{\sigma\lambda,\mu}(z ) ) a^{\sigma}(x)\ , , \label{h}\end{aligned}\ ] ] where @xmath21 satisfies the equation @xmath22 in the case of ( [ d ] ) , and @xmath23 when ( [ h ] ) is a solution of ( [ e ] ) . in ( [ h ] ) @xmath24 is a fixed reference point and @xmath25 a generic point along the particle s worldline .
we can prove that ( [ h ] ) is an exact solution to first order in @xmath12 by straightforward differentiation .
the first two integrals in ( [ h ] ) represent by themselves a solution of the klein - gordon equation @xmath26 .
the additional terms are related to spin .
in fact ( [ h ] ) can be re - written in the form @xmath27 @xmath28 where @xmath29 the rotation matrices @xmath30 satisfy the commutation relation @xmath31=i\epsilon_{ijk}s_k$ ] . by applying stokes theorem to the r.h.s .
of ( [ h ] ) we find @xmath32 where @xmath33 is the total angular momentum of the spin-1 particle , @xmath34 is the linearized riemann tensor , and @xmath35 is the surface bound by the closed path along which the integration is performed .
the weak field approximation @xmath36 does not fix the reference frame completely .
the transformations of coordinates @xmath37 , with @xmath38 also small of first order , are still allowed and lead to the `` gauge '' transformations @xmath39 .
equation ( [ n ] ) therefore indicates that solution ( [ h ] ) is covariant and also gauge invariant .
it also follows from ( [ n ] ) that the term containing @xmath40 in ( [ l ] ) does not contribute to integrations over closed paths , behaves as a gauge term and may therefore be dropped .
the spin - gravity coupling is contained in the third term on the r.h.s . of ( [ l ] ) .
its time integral part is @xmath41 .
since for rotation @xmath42 , one gets @xmath43 , where @xmath44 . in general , one may write @xmath45 , which must now be applied to a solution of the field free equations .
one finds @xmath46 and , for particles polarized parallel or antiparallel to @xmath47 , @xmath48 , as in @xcite.@xmath49 is the energy observed by the co - rotating observer .
for spin-@xmath50 fields , the simplest equation of propagation is derived in @xcite and is given by @xmath51 in lensing , the second term in ( [ mtw ] ) may be neglected when the wavelength @xmath52 associated with @xmath53 is smaller than the typical radius of curvature of the gravitational background @xcite .
we consider here massless or massive spin-@xmath50 particles described by the equation @xcite @xmath54 to first order in @xmath55 , ( [ spin2 ] ) can be written in the form @xmath56 where @xmath57 is the linearized ricci tensor of the background metric and @xmath58 is the corresponding christoffel symbol of the first kind .
it is easy to prove , by direct substitution , that a solution of ( [ wf ] ) , exact to first order in @xmath55 , is represented by @xmath59 \\
\nonumber + \frac{1}{2 } \int_{p}^{x}dz^{\lambda}\gamma_{\alpha\lambda}\left(z\right)\partial^{\alpha}\phi_{\mu\nu}\left(x\right)+ \int_{p}^{x}dz^{\lambda}\gamma_{\mu\lambda,\sigma}\left(z\right)\phi_{\nu}^{\sigma}\left(x\right)+ \int_{p}^{x}dz^{\lambda}\gamma_{\nu\lambda,\sigma}\left(z\right ) \phi_{\mu}^{\sigma}\left(x\right),\end{aligned}\ ] ] where @xmath60 satisfies the field - free equation @xmath61 and the gauge condition ( [ f ] ) has been used . equation ( [ solution ] ) can be written in the form @xmath62\phi_{\mu\nu}+ \\ \nonumber - \frac{i}{2}\int_{p}^{x}dz^{\lambda}\left(\gamma_{\alpha\lambda,\beta}\left(z\right)-\gamma_{\beta\lambda,\alpha}\right)s^{\alpha\beta}\phi_{\mu\nu } \left(x\right ) - \frac{i}{2 } \int_{p}^{x}dz^{\lambda}\gamma_{\beta\sigma,\lambda}\left(z\right)t^{\beta\sigma}\phi_{\mu\nu}\left(x\right)\,,\end{aligned}\ ] ] where @xmath63 from @xmath64 one constructs the rotation matrices @xmath65 that satisfy the commutation relations @xmath66=i \epsilon_{ijk}s_{k}$ ] .
the spin - gravity interaction is therefore contained in the term @xmath67.\ ] ] the solution ( [ solution ] ) is invariant under the gauge transformations @xmath68 . if , in fact , we choose a closed integration path @xmath69 , stokes theorem transforms the first three integrals of ( [ sol ] ) into the gauge invariant result @xmath70 where @xmath71 is the surface bound by @xmath72 , and @xmath73 is the total angular momentum of the particle . for the same path @xmath69 the integral involving @xmath74 in ( [ sol ] )
it behaves like a gauge term and is therefore dropped .
the helicity - rotation coupling for massless , or massive spin-@xmath50 particles follows immediately from the @xmath75 term in ( [ sol ] ) .
in fact , the particle energy is changed by virtue of its spin by an amount given by the time integral of this spin term @xmath76 that must then be applied to a solution of ( [ ff ] ) . for rotation about the @xmath77-axis , @xmath78 ,
we find @xmath79 and the energy of the particle therefore changes by @xmath80 , where the factor @xmath81 refers to the particle s helicity , as discussed by ramos and mashhoon @xcite . equation ( [ hrc ] ) extends their result to any weak gravitational , or inertial field .
the effect of ( [ sg ] ) on @xmath82 can be easily seen in the case of a gravitational wave propagating in the @xmath83-direction and represented by the components @xmath84 $ ] and @xmath85 $ ] .
for an observer rotating about the @xmath86-axis the metric is @xmath87 .
then the two independent polarizations @xmath88 and @xmath89 are transformed by @xmath90 into @xmath91 and @xmath92 . for closed integration paths and vanishing spin , ( [ sol ] ) coincides with the solution of a scalar particle in a gravitational field , as expected .
this proves the frequently quoted statement @xcite that gravitational radiation propagating in a gravitational background is affected by gravitation in the same way that electromagnetic radiation is ( when the photon spin is neglected ) .
the behavior of spin-1/2 particles in the presence of a gravitational field @xmath93 is determined by the covariant dirac equation @xmath94\psi(x)=0\,,\ ] ] where @xmath95 , @xmath96 is the spin connection and the matrices @xmath97 satisfy the relations @xmath98 .
both @xmath99 and @xmath100 can be obtained from the usual constant dirac matrices by using the vierbein fields @xmath101 and the relations @xmath102 where @xmath103 $ ] .
equation ( [ diracequation ] ) can be solved exactly to first order in @xmath104 .
this is achieved by first transforming ( [ diracequation ] ) into the equation @xcite,@xcite,@xcite @xmath105{\tilde \psi}(x)=0\,,\ ] ] where @xmath106 by multiplying ( [ diraceqtrasf ] ) on the left by @xmath107 , we obtain the equation @xmath108 whose solution @xmath109 is exact to first order . the operator @xmath110 is defined as @xmath111\hat{l}^{\alpha\beta}(z)+ \frac{1}{2}\int_p^x dz^\lambda\gamma_{\alpha\lambda}\hat{k}^\alpha\,,\ ] ] @xmath112=\left((x^\alpha - z^\alpha)\hat{k}^\beta- ( x^\beta - z^\beta)\hat{k}^\alpha\right)\psi_0(x)\ , , \qquad [ \hat{k}^\alpha , \psi_0(x)]=i\partial^\alpha\psi_0\,,\ ] ] and @xmath113 satisfies the usual flat space - time dirac equation . @xmath114 and @xmath115 are the angular and linear momentum operators of the particle .
it follows from ( [ exactsolution ] ) and ( [ psitilde ] ) that the solution of ( [ diracequation ] ) can be written in the form @xmath116 and also as @xmath117 where @xmath118 is of first order in @xmath119 .
the factor @xmath120 on the r.h.s . of ( [ psisolution2 ] ) is required by the condition that both sides of the equation agree when the gravitational field vanishes .
it is useful to re - derive some known results from the covariant dirac equation . on multiplying ( [ diracequation ] ) on the left by @xmath121 and using the relations @xmath122=-\frac{1}{4}\sigma^{\alpha\beta}(x)r_{\alpha\beta\mu\nu}\,,\ ] ] and @xmath123=-\frac{i}{4}\ , \sigma^{\alpha\beta}(x)r_{\alpha\beta\mu\nu}\,,\ ] ]
we obtain the equation @xmath124 in ( [ relation2 ] ) and ( [ kgeq+r ] ) @xmath125 $ ] and @xmath126 is the ricci scalar . on applying stokes theorem to a closed space - time path @xmath127 and using ( [ relation1 ] )
, we find that @xmath128 changes by @xmath129 where @xmath130 is the total momentum of the particle . equation ( [ phase ] ) shows that ( [ psisolution2 ] ) is gauge invariant .
the spin - rotation coupling derived by mashhoon by extending the hypothesis of locality can be now derived rigorously from the solution found .
choose a cylindrical coordinate ( @xmath131 , @xmath132 , @xmath133 , @xmath134 ) for an inertial frame @xmath135 .
an observer at rest in a frame @xmath136 rotating with a constant angular velocity @xmath137 relative to @xmath135 will follow the world line @xmath138 . consider an orthogonal tetrad consisting of the observer s four - velocity @xmath139 and the triad @xmath140 ( @xmath141 ) normal to the world line . by using the local tetrad @xcite @xmath142 , where @xmath143 , one can construct a vierbein field @xmath144 along the world line of the observer @xmath145 .
it is then easy to calculate the spinor connection @xmath146 . in calculating the energy ,
only the component @xmath147 is necessary . by using the dirac representation for the @xmath148-matrices , one obtains @xmath149 and from @xmath150 also @xmath151 , where @xmath152 has the usual plane wave form . besides the contribution due to the coupling of the orbital angular momentum to rotation , which gives the sagnac effect @xcite , one obtains the spin - rotation coupling @xmath153 , and also , for spin polarizations parallel or antiparallel to the direction of rotation , one obtains , @xmath154 , as shown by mashhoon .
the present result is exact and follows from the general form of the solution ( [ psisolution2 ] )
. it also agrees with those of hehl and ni @xcite and @xcite .
the mashhoon effect is obviously a prime candidate for experiments with accelerators and will be discussed at length below . according to ( [ phase ] ) , both angular momentum and spin couple to a weak gravitational field in the same way .
this confirms that , unlike the electromagnetic case , the gyro - gravitational ratio of a spin-1/2 particle is 1 , as shown in @xcite . a classical charge @xmath155 moving in a circle with angular momentum @xmath156 forms a current loop of magnetic moment @xmath157 , which gives the gyromagnetic factor @xmath158 .
the magnetic moment of a charged particle depends therefore on the ratio @xmath159 and , for a rotating object , on the space distributions of charge and mass . for a quantum particle
, the dirac equation indicates that @xmath160 .
the corrections to @xmath160 come from quantum electrodynamics where the electron can be pictured at any instant as a bare particle in interaction with a cloud of virtual photons .
qualitatively , if the charge remains associated with the electron , part of the mass energy is carried by the photon cloud resulting in a slight increase for the value @xmath161 of the electron itself . in the gravitational case
, however , the gyro - gravitational ratio of the spin-1/2 particle is @xmath162 .
this suggests , according to @xcite , that the internal distributions of the gravitational mass , associated with the interaction , and of the inertial mass , associated with the angular momentum , equal each other .
in the geometrical optics approximation , valid whenever @xmath163 , where @xmath164 is the momentum of the particle , the interaction between the angular momentum of the source and the particle s spin vanishes .
this interaction is quantum mechanical in origin .
then the geometrical phase @xmath165 is sufficient to reproduce the classical angle of deflection @xcite , as it should , because ( [ phig ] ) coincides with the first two integrals in ( [ h ] ) and ( [ sol ] ) .
we can therefore treat photons , gravitons and fermions simultaneously when spin is neglected .
more detailed calculations involving neutrinos are given in @xcite . if , e.g. , we choose a gravitational background represented by the lense - thirring metric @xcite , @xmath166 , @xmath167 , @xmath168 , and @xmath169 , with @xmath170 , @xmath171 and @xmath172 is related to the angular momentum of the gravitational source . in particular , if the source rotates with angular velocity @xmath173 , then @xmath174 , @xmath175 .
without loss of generality , we assume that the particles are massless and propagate along the @xmath134-direction , hence @xmath176 .
using plane waves for the field free solution , the phase of the wave equation becomes @xmath177 + \frac{1}{2}\int_p^x dz^\lambda \gamma_{\alpha\lambda}(z ) k^\alpha\,.\ ] ] we can define the particle momentum as @xmath178 it is easy to show that @xmath179 satisfies the eikonal equation @xmath180 .
for the lense - thirring metric , @xmath181 is given by @xmath182dz'\right]+k\int_p^q dz ' \phi \\ & & -\frac{k}{2}\int_p^q \left[\left ( ( x - x')(h_{1 , \,z'}-2h_{3,\ , x'})+(y - y')(h_{2 , \,z'}-2h_{3,\ , y'})\right)dz ' \right .
\nonumber \\ & & \left .
\quad -\left((x - x')h_{1,\ , x'}+(y - y')h_{1,\ , y'}\right)dx ' -\left((x - x')h_{2,\ , x'}+(y - y')h_{2,\ , y'}\right)dy'\right ] \nonumber \\ & & + \frac{k}{2}\int_p^q\left[2h_3dz'+h_1dx'+h_2dy'\right]\,,\end{aligned}\ ] ] where @xmath183 is the point at which the particles are generated , and @xmath184 is a generic point along their space - time trajectory .
the components of the momentum are therefore @xmath185 we then have @xmath186 where @xmath187 is the component of the momentum orthogonal to the direction of propagation of the particles .
since only phase differences are physical , it is convenient to choose the space - time path by placing the particle source at distances that are very large relative to the dimensions of the lens , and the generic point is located along the @xmath134 direction .
we therefore replace @xmath184 with @xmath134 , where @xmath188 . using the expression for @xmath189
we find that their contribution is negligible and ( [ k1])-([k3 ] ) simplify to @xmath190 from ( [ k1bis])-([k3bis ] ) we can determine the deflection angle @xmath133 .
let us analyze the case of non - rotating lenses , i.e. @xmath191 .
we get @xmath192 where @xmath193 . by defining the deflection angle as @xmath194
it follows that @xmath195 in the limit @xmath196 we obtain the usual einstein result @xmath197 a general expression for the index of refraction @xmath198 can also be derived from ( [ gph]),([photonmomentum ] ) and @xmath199 .
we now consider the propagation of light and gravitational waves in a background metric represented by @xmath200 , @xmath201 , where @xmath202 is the gravitational potential of the lens and @xmath132 the distance from @xmath203 to the particle .
wave optics effects can be seen by using the type of double slit arrangement indicated in fig.1 .
we will also use a solution @xmath204 in the form of a plane wave @xmath205 and neglect spin effects .
this limits the calculation @xcite of the phase difference to the first two terms in ( [ 9 ] ) and ( [ 10 ] ) .
we also assume for simplicity that @xmath206 , so that propagation is entirely in the @xmath207)-plane and the set - up is planar . the corresponding wave amplitude @xmath208 is therefore @xmath209 + \frac{1}{2}\left[\int_s^x dz^0 \gamma_{00 } \pi^0+\int_s^x dz^2 \gamma_{22 } \pi^2+\int_s^x dz^{3}\gamma_{33}\pi^{3}\right ] \right\}\ , , \label{*}\end{aligned}\ ] ] where @xmath210 , @xmath211 , and we have taken into account the fact that @xmath212 plays no role in the planar arrangement chosen . the phase must now be calculated along the different paths sp+po and sl+lo taking into account the values of @xmath213 in the various intervals .
geometry of a two - image gravitational lens .
the solid lines represent the particle paths between the particle source at @xmath214 and the observer at @xmath215 .
@xmath216 is the spherically symmetric gravitational lens . @xmath217 and the particle paths lie in the same plane.,scaledwidth=90.0% ]
the total change in phase is @xcite @xmath218 all integrations in ( [ * ] ) can be performed exactly and the results can be expressed in terms of physical variables @xmath219 , @xmath220 , @xmath221 , @xmath222 , and @xmath223 or lensing variables @xmath224 , @xmath225 , @xmath226 , @xmath227 , @xmath228 , and @xmath229 .
we find @xmath230 where @xmath231 , @xmath232 , @xmath233 and @xmath234 . as an example , let us consider the simpler case @xmath235 , from which we obtain ( see fig.1 ) @xmath236 .
a simple calculation shows that the probability density of finding a photon , or graviton , at @xmath237 is typically @xmath238
measurements of the interaction of spin with rotation have been carried out in the case of photons using signals from global positioning system satellites @xcite and data published in @xcite can be re - interpreted @xcite as due to the coupling of earth s rotation to the nuclear spins in mercury .
the spin - rotation effect is also consistent with a small depolarization of electrons in storage rings @xcite .
we show below that the same coupling is of particular interest in experiments with storage rings .
it is essential in getting the correct @xmath240-dependence in @xmath241 experiments @xcite and eliminates the need of _ ad hoc _ , phenomenological arguments .
the effect is conceptually important .
it extends to the quantum level the classical coupling of rotation to the intrinsic angular momentum of a body and simplifies the treatment of rotational inertia .
it also yields different potentials for different particles and for different spin states and can not , therefore , be considered universal @xcite . before discussing its connection with @xmath0 experiments , it is useful to briefly recall the usual experimental setup
. the experiment @xcite involves muons in a storage ring a few meters in diameter , in a uniform vertical magnetic field .
muons on equilibrium orbits within a small fraction of the maximum momentum are almost completely polarized with spin vectors pointing in the direction of motion . as the muons decay
, those electrons projected forward in the muon rest frame are detected around the ring .
their angular distribution thence reflects the precession of the muon spin along the cyclotron orbits .
the calculations are performed in the rotating frame of the muon and do not therefore require a relativistic treatment of inertial spin effects @xcite
. then the vierbein formalism yields @xmath242 and @xmath243 where @xmath244 and @xmath245 are the three - acceleration and three - rotation of the observer and , in the chiral representation of the usual dirac matrices , @xmath246=i \left(\matrix { \sigma^i & 0 \cr 0 & -\sigma^i \cr } \right)\,.\ ] ] the second term in ( [ 1 ] ) represents the mashhoon term .
the first term drops out .
in fact , by symmetrization we obtain @xmath247+\frac{i\hbar}{2}\left(\vec{a}\cdot \vec{\alpha}\right).\ ] ] the last term in ( [ simp ] ) and the first term in ( [ 1 ] ) therefore cancel each other .
the remaining first order contributions in @xmath244 and @xmath245 to the dirac hamiltonian add up to @xcite @xmath248 \\ & & -{\vec \omega}\cdot \left({\vec l}+\frac{{\vec \sigma}}{2}\right)\ , .
\nonumber\end{aligned}\ ] ] for simplicity all quantities in @xmath249 are taken to be time - independent .
they are referred to a left - handed tern of axes rotating about the @xmath250-axis in the clockwise direction of motion of the muons .
the @xmath251-axis is tangent to the orbits and in the direction of the muon momentum .
the magnetic field is @xmath252 .
only the mashhoon term and the magnetic moment interaction then couple the helicity states of the muon .
the remaining terms contribute to the overall energy @xmath253 of the states , and we indicate by @xmath254 the corresponding part of the hamiltonian . before decay
the muon states can be represented as @xmath255 where @xmath256 and @xmath257 are the right and left helicity states of the hamiltonian @xmath254 and satisfy the equation @xmath258 the total hamiltonian reduces effectively to @xmath259 , where @xmath260 @xmath261 represents the total magnetic moment of the muon and @xmath262 is the bohr magneton .
electric fields used to stabilize the orbits and stray radial electric fields can also affect the muon spin .
their effects can however be cancelled by choosing an appropriate muon momentum @xcite and will not be considered in what follows .
the coefficients @xmath263 and @xmath264 in ( [ 2 ] ) evolve in time according to @xmath265 where @xmath216 is the matrix @xmath266\ ] ] and @xmath267 represents the width of the muon .
the non - diagonal form of @xmath216 ( when @xmath268 ) implies that rotation does not couple universally to matter .
@xmath216 has eigenvalues @xmath269 and eigenstates @xmath270\ , , \nonumber \\
|\psi_2 > & = & \frac{1}{\sqrt{2}}\,\left[-i|\psi_+>+|\psi_->\right ] \ , .
\nonumber\end{aligned}\ ] ] the muon states that satisfy ( [ 4 ] ) , and the condition @xmath271 at @xmath272 , are @xmath273|\psi_+ > \right . \\
& & \left .
+ \left[e^{-i{\tilde \omega } t } + e^{i{\tilde \omega } t}\right]|\psi_- >
\right\ } \ , , \nonumber\end{aligned}\ ] ] where @xmath274 the spin - flip probability therefore is @xmath275\ , . \nonumber\end{aligned}\ ] ] the @xmath267-term in ( [ 7 ] ) accounts for the observed exponential decrease in electron counts due to the loss of muons by radioactive decay @xcite .
the spin - rotation contribution to @xmath276 is represented by @xmath277 which is the cyclotron angular velocity @xmath278 @xcite .
the spin - flip angular frequency is then @xmath279 which is precisely the observed modulation frequency of the electron counts @xcite .
this result is independent of the value of the anomalous magnetic moment of the particle .
the cancellation of the dirac value of the magnetic moment contribution by the mashhoon term must therefore take place for all spin-1/2 particles in a similar physical set - up .
hence , it is the spin - rotation coupling that generates the correct @xmath280 factor in @xmath137 by exactly cancelling , in @xmath281 , the much larger contribution @xmath262 that fermions with no anomalous magnetic moment produce . the cancellation is made possible by the non - diagonal form of @xmath216 and is therefore a direct consequence of the violation of the equivalence principle .
it has , of course , been argued that this principle does not hold true in the quantum world @xcite .
recently , discrepancies between the experimental and standard model values of @xmath282 have been observed with very high accuracy @xcite .
the most precise data yet give @xmath283 for the negative muon @xcite , and @xmath284 for the positive muon @xcite .
this discrepancy can be used to set upper limits on @xmath183 and @xmath285 invariance violations in spin - rotation coupling @xcite .
the possibility that discrete symmetries in gravitation be not conserved has been discussed in the literature .
attention has in general focused on the potential @xmath286,\ ] ] which applies to a particle of generic spin @xmath287 .
the first term , introduced by leitner and okubo @xcite , violates the conservation of @xmath183 and @xmath285 .
the same authors determined the upper limit @xmath288 from the hyperfine splitting of the ground state of hydrogen .
the upper limit @xmath289 was determined in @xcite from sn 1987a data .
the corresponding potential violates the conservation of @xmath183 and @xmath127 .
conservation of @xmath127 and @xmath285 is violated by the last term , while ( [ 1a ] ) , as a whole , conserves @xmath290 .
there is , as yet , no upper limit on @xmath291 .
these studies are extended here to the mashhoon term . before decay
, the muon states can be represented as in ( [ 2 ] ) where @xmath292 and @xmath293 again are the right and left helicity states of the hamiltonian @xmath294 defined in the previous section .
assume now that the coupling of rotation to @xmath292 differs in strength from that to @xmath293 .
then the mashhoon term can be modified by means of a matrix @xmath295 that reflects the different coupling of rotation to the two helicity states .
the total hamiltonian now is @xmath296 , where @xmath297 a violation of @xmath183 and @xmath285 in ( [ 4a ] ) would arise through @xmath298 . the constants @xmath299 and @xmath300 are assumed to differ from unity by small amounts @xmath301 and @xmath302 . the coefficients @xmath263 and @xmath264 in ( [ 2 ] ) evolve in time according to @xmath303 where @xmath304 and @xmath267 represents , as before , the width of the muon .
the spin - rotation term is off - diagonal in ( [ 6b ] ) and does not therefore couple to matter universally .
it violates hermiticity as shown in @xcite and , in a general way , by scolarici and solombrino @xcite .
it also violates @xmath285 , @xmath183 and @xmath305 , while nothing can be said about @xmath290 conservation which requires @xmath306 to be hermitian . because of the non - hermitian nature of ( [ 4a ] ) , one expects @xmath267 itself to be non - hermitian .
the resulting corrections to the width of the muon are , however , of second order in the @xmath307 s and are neglected .
@xmath216 has eigenvalues @xmath308 where @xmath309 and eigenstates @xmath310 then the muon states ( [ 2 ] ) are @xmath311,\end{aligned}\ ] ] where the condition @xmath312 has been applied .
the spin - flip probability is therefore @xmath313.\end{aligned}\ ] ] when @xmath314 , ( [ 9 ] ) yields @xcite @xmath315,\ ] ] that provides the appropriate description of the spin - rotation contribution to the spin - flip transition probability .
notice that the case @xmath316 ( no spin - rotation coupling ) yields @xmath317\ ] ] and does not therefore agree with the results of the @xmath280 experiments .
hence the necessity of accounting for spin - rotation coupling whose contribution cancels the factor @xmath318 in ( [ a1 ] ) . substituting @xmath319 into ( [ 9 ] )
, one finds @xmath320.\ ] ] one may attribute the discrepancy between @xmath321 and @xmath322 to a violation of the conservation of the discrete symmetries by the spin - rotation coupling term in ( [ 4 ] ) .
the upper limit on the violation of @xmath323 and @xmath305 is derived from ( [ 11 ] ) assuming that the deviation from the current value of @xmath322 is wholly due to @xmath324 .
the upper limit is therefore @xmath325 in the case of negative muons @xcite and of @xmath326 for positive muons @xcite . at the same time
the two values of @xmath327 can be thought of as due to a different coupling strength between rotational inertia and the two helicity states of the muon .
then the values of @xmath328 and @xmath329 can be determined from @xmath330 in ( [ 9 ] ) according to the equations @xmath331 and @xmath332 equations ( [ 11a ] ) and ( [ 11b ] ) have the approximate solutions @xmath333 and @xmath334 more precise , numerical solutions give @xmath335 and @xmath336 .
these values are significant in view of the precision with which @xmath337 have been determined .
it then follows that the coupling of rotation to positive helicity is larger than that to negative helicity , which agrees with the value @xmath338 for both @xmath339 and @xmath340 .
this also means that the violation of @xmath24 and @xmath341 is relatively stronger for positive helicity at a level @xmath342 and that the spin - rotation interaction is an inherent source of @xmath24 and @xmath285 violation .
in this section , it is convenient to write the left and right neutrino wave functions in the form @xmath343 where @xmath344 represents the pauli matrices , @xmath345 are eigenvectors of @xmath346 corresponding to negative and positive helicity and @xmath347 , @xmath348 .
this notation already takes into account the fact that if @xmath349 are the helicity states , then we have @xmath350 for relativistic neutrinos . in general , the spin precesses during the motion of the neutrino .
this can be seen , for instance , from the contribution @xmath150 in @xmath351 .
the expectation value of the contribution of @xmath352 to the effective mechanical momentum can in fact be re - written in the form @xmath353 where @xmath354 .
equation ( [ omega1 ] ) represents the spin - rotation coupling for the lense - thirring metric . here
rotation is provided by the gravitational source , rather than by the particles themselves .
we now study the helicity flip of one flavor neutrinos as they propagate in the gravitational field produced by a rotating mass @xcite .
the neutrino state vector can be written as @xmath355 where @xmath356 and @xmath357 is an affine parameter along the world - line . in order to determine @xmath358 and @xmath359
, we can write ( [ psisolution ] ) as @xmath360 where @xmath361 and @xmath362 is the corresponding solution in minkowski space - time .
the latter can be written as @xmath363\,.\ ] ] strictly speaking , @xmath364 should also be normalized .
however , it can be shown @xcite that @xmath365 is already of @xmath366 , can only produce higher order terms and is therefore unnecessary in this calculation . from ( [ neutrinorl ] ) ,
( [ psisolution3 ] ) and ( [ psizero ] ) we obtain @xmath367 an equation for @xmath368 can be derived in an entirely similar way . if we consider neutrinos which are created in the left - handed state , then @xmath369 , and we obtain @xmath370 where @xmath371 .
as remarked in @xcite , @xmath372 need not be a null vector if we assume that the neutrino moves along an `` average '' trajectory .
we also find , to lowest order , @xmath373 where @xmath374 are the usual christoffel symbols , and @xmath375.\ ] ] in what follows , we compute the probability amplitude ( [ rtl ] ) for neutrinos propagating along the @xmath134 and the @xmath83 directions explicitly . for propagation along the @xmath134-axis
, we have @xmath376 and @xmath377 . as in section iii , we choose @xmath378 .
we get @xmath379 summing up , and neglecting terms of @xmath380 , ( [ rtl ] ) becomes @xmath381 the contributions to @xmath382 vanish . as a consequence @xmath383 and
the probability amplitude for the @xmath384 transition is of @xmath385 , as expected . integrating ( [ dadz ] ) from @xmath386 to @xmath134 ,
yields @xmath387\\ & = & \frac{m}{e}\frac{gm}{2b}\left[1+\frac{z}{r}-i\frac{2\omega r^2b^2}{5r^3}\right]\,.\nonumber\end{aligned}\ ] ] it also follows that @xmath388\,.\ ] ] the first of the two terms in ( [ |alfa - z| ] ) comes from the mass of the gravitational source .
the second from the source s angular momentum and vanishes for @xmath389 because the contribution from @xmath390 to 0 exactly cancels that from 0 to @xmath391 .
in fact , if we consider neutrinos propagating from 0 to @xmath391 , we obtain @xmath392\,.\ ] ] according to semiclassical spin precession equations @xcite , there should be no spin motion when spin and @xmath393 are parallel as in the present case .
this is a hint that rotation of the source , rather than of the particles , should produce a similar effect
. the probabilities ( [ |alfa - z| ] ) and ( [ alfa infinito ] ) mark therefore a departure from expected results .
they are however small of second order .
both expressions vanish for @xmath394 , as it should for a stationary metric . in this case , in fact , helicity
is conserved @xcite .
it is interesting to observe that spin precession also occurs when @xmath395 vanishes @xcite . in the case of ( [ |alfa - z| ] ) the mass contribution
is larger when @xmath396 , which , close to the source , with @xmath397 , becomes @xmath398 and is always satisfied . in the case described by ( [ alfa infinito ] ) , the rotational contribution is larger if @xmath399 which effectively restricts the region of dominance to a narrow strip about the @xmath400-axis in the equatorial plane , if the source is compact and @xmath395 is relatively large . in this case , we take @xmath376 , @xmath401 . as in section iii , the calculation can be simplified by assuming that the motion is in the equatorial plane with @xmath402 , @xmath403 .
we then have @xmath404 summing up , and neglecting terms of @xmath380 , ( [ rtl ] ) becomes @xmath405 the contributions to @xmath382 again vanish and we get @xmath406\sim{\cal o}(m / e)\,.\ ] ] integrating ( [ dadx ] ) from @xmath386 to @xmath83 , we obtain @xmath407 and @xmath408 obviously , the mass contribution is the same as for propagation along the @xmath134-axis .
however , the two cases differ substantially in the behavior of the angular momentum term . in this case , in fact
, this term is even , so it does not vanish for @xmath389 .
if we consider neutrinos generated at @xmath409 and propagating to @xmath410 , we find @xmath411 the mass term is larger when @xmath412 . at the poles @xmath413 and the mass term dominates because the condition @xmath414 is always satisfied .
the angular momentum contribution prevails in proximity of the equatorial plane .
the transition probability vanishes at @xmath415 .
covariant wave equations for massless and massive particles can be solved exactly to first order in @xmath55 .
the solutions are covariant and invariant with respect to the gauge transformations of the electromagnetic field and of @xmath416 and are known when a solution of the free wave equation is known .
the external gravitational field only appears in the phase of the wave function .
we have shown that the coupling of spin to inertia and gravitation follow from the solutions given .
this allows a unified treatment of the interaction of gravity with spin and angular momentum without requiring ad hoc procedures .
according to equations ( [ l ] ) and ( [ sol ] ) , the spin term @xmath417 finds its origin in the skew - symmetric part of the space - time connection . in the case of fermions
, @xmath90 is accounted for by the spinorial connection .
the terms that contain @xmath90 gives rise to the skrotskii effect for both electromagnetic @xcite and gravitational waves @xcite . from the phases we have derived the geometrical optics of the particles and verified that their deflection is that predicted by general relativity .
in addition , the background gravitational field acts as a medium whose index of refraction can be calculated for any metric from ( [ gph ] ) , ( [ photonmomentum ] ) and @xmath418 . a more detailed treatment of the geometrical optics of single flavor neutrinos can be found in @xcite where we also calculate corrections due to the neutrino mass
. for propagation parallel to the axis of rotation of the source , the rotation corrections vanish at infinity . not so for propagation perpendicular to the axis of rotation .
we can finally conclude that the validity of covariant wave equations in an inertial - gravitational context finds support in experimental verifications of some of the effects they predict @xcite , in tests of the general relativistic deflection of light rays and also in the phase wrap - up in global position system measurements @xcite .
we have then asked ourselves the question whether , beside the phase wrap - up in gps , there is a wider role for spin - gravity coupling in physics .
in particular , we have considered muon @xmath280 experiments and helicity transitions in neutrino physics .
we have found that spin - rotation coupling is largely responsible for producing the correct @xmath280 factor in the spin - flip angular frequency @xmath419 .
measurements of this factor already provide the most stringent test yet of einstein s time dilation formula @xcite .
however , muons in storage rings are also rotating quantum gyroscopes and inertia must be an essential ingredient of their description in experiments of high and ever increasing sensitivity . possibly these experiments also concern problems like the violation of the equivalence principle in quantum mechanics and the conservation of discrete symmetries in inertia - gravitation .
we have , in fact , shown that a slightly anomalous inertial contribution to ( [ 11 ] ) at a level of the values @xmath420 and @xmath421 of section vii can produce violations of the discrete symmetries at the same level .
the upper limits on the violations of @xmath24 and @xmath341 that can be reached by @xmath0 experiments are in fact as sensitive as those obtained by other means @xcite , but with an important difference .
the @xmath280 measurements are performed in strictly controlled laboratory conditions rather than in astrophysical situations . in derivations based on the covariant dirac equation
, the coupling of inertia and gravitation to spin is identical to that for orbital angular momentum .
a suggestive interpretation of this result is that the internal distributions of the gravitational mass , associated with the interaction , and of inertial mass , associated with the angular momentum , equal each other .
this is no longer so when @xmath422 .
there is almost a similarity , here , with the electromagnetic case where @xmath423 is required by the dirac equation , but not by quantum electrodynamics .
the deviations of @xmath424 and @xmath425 from unity that are consistent with @xmath241 experiments are both of the order of @xmath282 , or @xmath426 , and differ from each other by @xmath427 . while small values of @xmath428 do not give rise to measurable mass differences in macroscopic objects @xcite , violations of the discrete symmetries can have interesting astrophysical and cosmological implications .
next , we have calculated the helicity transition amplitudes of ultra - relativistic , single flavor neutrinos as they propagate in a lense - thirring field .
these transitions are interesting because at high energies chirality states are predominantly helicity states and right - handed neutrinos do not interact @xcite .
the transition probabilities are of @xmath429 .
two directions of propagation have again been selected and the results contain contributions from both mass and angular momentum of the source
. the transitions also occur in the absence of rotation or with spin parallel to rotation , which is unexpected on semiclassical grounds .
the mass contributions predominate when the neutrinos propagate from @xmath430 to @xmath431 ( and matter effects are neglected ) , provided the impact parameter @xmath432 .
there is , however , a narrow region about the axis of propagation in the equatorial plane where the @xmath433 contribution is larger .
the rotational contribution behaves differently in the two cases .
it vanishes as @xmath434 for propagation along @xmath134 , but not so as @xmath435 in the second case .
in addition , when the neutrinos propagate from @xmath436 to @xmath437 , the mass term dominates in the neighborhood of the poles , while the contribution of @xmath433 is larger close to the equator , with no attenuation at @xmath438 .
in @xcite we have also calculated gravity induced , two - flavor oscillations and derived the relative equation and effective hamiltonian .
the transition probabilities do indeed oscillate for the lense - thirring metric , and the curvature of space - time enters the oscillation probability through the gravitational red - shift of the local energy @xmath439 and the proper distance @xmath440 .
the results presented in this paper can be applied to a number of problems in astroparticle physics and cosmology @xcite .
for instance , an interesting question is whether gravity induced helicity and flavor transitions could effect changes in the ratio @xmath441 of the expected fluxes at earth
. lepton asymmetry in the universe @xcite also is an interesting problem .
it is known that the active - sterile oscillation of neutrinos can generate a discrepancy in the neutrino and antineutrino number densities .
the lepton number of a neutrino of flavor @xmath442 is defined by @xmath443 , where @xmath444 is the number density of neutrinos ( antineutrinos ) and @xmath445 is the number density of photons at temperature @xmath341 .
as noted above , the gravitational field generates transitions from left - handed ( active ) neutrinos to right - handed ( sterile ) neutrinos .
if , in primordial conditions , ( [ alfa infinito ] ) and ( [ pinfinitox ] ) become larger , then helicity transitions may contribute in some measure to lepton asymmetry . finally , we have recently re - examined the behavior of the spin - gravity interaction and found that gravity can distinguish between chirality and helicity @xcite .
we have also found that the spin - gravity interaction can distinguish between dirac and majorana wave packets @xcite . a spin - flip does in fact change a majorana neutrino into an antineutrino and
behaves like a charge conjugation operation .
the spin - gravity couplings discussed in our work make use of the weak field approximation in which gravity enters as a non - dynamical field .
there are , of course , physical situations in which this approach can be trusted and the general agreement between the quasi - classical and the quantum mechanical approaches has been established @xcite .
one could then be tempted to extend our approximation procedure to any order in the metric deviation , as suggested by equations ( [ n ] ) , ( [ phasephi ] ) and ( [ phase ] ) .
this would however lead to inconsistencies that can only be removed , as shown by deser @xcite , by making use of the full non - linear apparatus of general relativity .
furthermore , it is assumed , in calculating the phases induced by gravity , that all possible particle paths reduce , in the average , to the phase integration paths .
this approximation worsens the more `` quantum mechanical '' particles and gravity become .
an additional point concerns the use of the locality hypothesis in replacing non - inertial frames with inertial ones . as shown by mashhoon @xcite , this hypothesis has limitations and important consequences for the measuring process . for standard
accelerated measuring devices , for instance , it entails the introduction of a maximal acceleration @xcite . in the absence of inertial frames , of a complete formulation of spin - gravity couplings in curved space - time and of direct experimental observations , assigning a role to curvature in spin related problems
may then prove difficult , though worthy of investigation .
a final question regards the validity of the equivalence principle in spin - gravity interactions . in @xmath280 experiments
the interaction of spin with gravity depends on the relative direction of spin and rotation of the source . as such it is not universal and
may be regarded as violating those formulations of the equivalence principle that hinge on universality . but a more fundamental violation has been introduced in section vii where the strength of the coupling itself depends on the helicity of the particle @xcite .
this is a violation of the post - newtonian equivalence principle , recently discussed by silenko and teryaev @xcite , by which a particle s spin and angular momenta precession frequencies coincide .
this principle must be the object of rigorous experimental verifications and the authors themselves suggest a number of tests to find even more precise upper limits than those determined by using muon @xmath280 experiments @xcite . in a closely related paper @xcite ,
the same authors show that the spin - gravity dipole coupling term found by obukhov @xcite using the eriksen - korlsrud transformation @xcite does not lead to observable effects .
recently , impressive technical developments in the field of masers @xcite have succeeded in placing an upper limit of @xmath446 on violations of lorentz and @xmath290 symmetries .
higher sensitivities are expected to be reached in the near future .
since the interaction energy for a spin-@xmath447 particle in the rotation field of earth is of the same order of magnitude , these developments bode well for the physics of spin - gravity interactions .
99 b. mashhoon , phys .
* 61 * , 2639 ( 1988 ) .
b. mashhoon , phys .
a * 143 * , 176 ( 1990 ) .
b. mashhoon , phys .
a * 145 * , 147 ( 1990 ) .
b. mashhoon , phys .
lett . * 68 * , 3812 ( 1992 ) .
y. q.cai , g. papini , class .
quantum grav .
* 6 * , 407 ( 1989 ) .
y. q. cai , g. papini , phys .
66 * , 1259 ( 1991 ) ; * 68 * , 3811 ( 1992 ) .
d. singh , g. papini , nuovo cimento b * 115 * , 233 ( 2000 ) .
g. lambiase , g. papini , r. punzi and g. scarpetta , phys .
d * 71 * , 073011 ( 2005 ) .
g. papini , phys .
d * 75*,044022 ( 2007 ) .
hehl , f.w . and lmmerzahl , c. , 1992 , _ relativistic gravity research
_ , adited by j. ehlers and g. schfer , springer , berlin . c. s. misner , k. s. thorne and j.a .
wheeler , _ gravitation _ ( freeman , san francisco , 1973 ) .
riehle , f. , kister , th . ,
witte , a. , helmcke , j. , bord@xmath448,ch . j. , 1991 , _ phys .
lett . _ * 67 * , 177 ; bord@xmath448 , ch.j .
, 1989 , _ phys .
lett . a _ * 140 * , 10 ; bord@xmath448 , ch.j .
lett . a _ * 188 * , 187 ; bord@xmath448 , ch.j . , 1997 , in _ atom interferometry _ , edited by p. berman , academic press , london
r. takahashi , t. nakamura , apj * 595 * , 1039 ( 2003 ) .
j. ramos , b. mashhoon , phys .
d * 73 * , 084003 ( 2006 )
. k. s. thorne in _ three hundred years of gravitation _ , edited by s. w. hawking and w. israel ( cambridge university press , cambridge , 1987 ) , p. 361 .
g. papini in _ relativity in rotating frames _ edited by guido rizzi and matteo luca ruggiero ( kluwer science publishers , dordrecht , 2004 ) , pp .
335 - 359 .
f. w. hehl , and w .- t .
ni , phys . rev .
d , * 42 * , 2045 ( 1990 ) . c. g. de oliveira and j. tiomno , nuovo cimento * 24 * , 672 ( 1962 ) .
j. audretsch , j. phys a : math . gen . * 14 * , 411 ( 1981 ) .
l. kannenberg , ann .
( n.y . ) * 103 * , 64 ( 1977 ) .
g. lambiase , r. punzi , g. scarpetta and g. papini , _ general relativity and gravitational physics , 16th sigrav conference _ , edited by g. esposito , g. lambiase , g. marmo , g. scarpetta , g. vilasi ( american institute of physics , new york , 2005 ) , p. 82 .
j. lense and h. thirring , z. phys .
* 19 * , 156(1918 ) ; ( english translation : b. mashhoon , f.w .
hehl and d.s .
theiss , gen .
* 16 * , 711(1984 ) ) .
neil ashby , living reviews in relativity , 1 ( 2003 ) : in _ gravitation and relativity at the turn of the millennium _ , edited by n. dadhich and j. narlikar ( iuacc , pune , 1998 ) , pp.231 - 258 .
b. j. venema , p. k. majumder , s. k. lamoreaux , b. r. heckel and e. n. fortson , phys .
lett . * 68 * , 135 ( 1992 ) .
b. mashhoon , phys .
lett . a , * 198 * , 9 ( 1995 ) . j. s. bell and j. leinaas , nucl .
b , * 284 * , 488 ( 1987 ) .
g. papini , g. lambiase , phys .
a * 294 * , 175 ( 2002 ) .
b. mashhoon , class . quantum grav . * 17 * , 2399 ( 2000 ) .
j. bailey @xmath449 @xmath450 , nucl . phys . *
b150 * , 1 ( 1979 ) .
farley and e. picasso , advanced series in high - energy physics , vol . 7 , quantum electrodynamics , ed .
t. kinoshita , world scientific , singapore , 1990 , p. 479 .
lewis ryder , j. phys . a : math
* 31 * , 2465 ( 1998 ) .
f. j. farley and e. picasso , ann .
nucl . part .
* 29 * , 243 ( 1979 ) . c. lmmerzahl , gen .
* 28 * , 1043 ( 1996 )
. h. n. brown @xmath449 @xmath450 , muon ( g-2 ) collaboration , phys .
lett . * 86 * , 2227 ( 2001 ) .
w bennett @xmath449 @xmath450 , muon ( g-2 ) collaboration , phys .
lett . * 89 * , 101804 ( 2002 ) ; phys .
lett . * 92 * , 161802 ( 2004 ) .
h. n. brown @xmath449 @xmath450 , muon ( g-2 ) collaboration , phys .
d * 62 * , 091101 ( 2002 ) ; phys .
lett . * 86 * , 2227 ( 2001 ) .
g. lambiase and g. papini , phys . rev.d
* 70 * , 097901 ( 2004 ) . g. lambiase and g. papini , in _ muons : new research _ , edited by j. caparthy ( nova science publishers , new york 2005 ) , pp .
j. leitner and s. okubo , phys . rev . * 136 * , b1542 ( 1964 ) .
almeida , g.e.a .
matsas and a.a .
natale , phys .
d * 39 * , 677 ( 1989 ) .
g. papini , phys .
d * 65 * , 077901 ( 2001 ) .
g. scolarici , l. solombrino , phys .
a * 303 * , 239 ( 2002 ) .
cardall , g.m .
fuller , phys .
d * 55 * , 7960 ( 1997 ) .
bryan w. montague , phys .
rep . * 113 * , 1 ( 1984 ) .
d. singh , n. mobed , and g. papini , j. phys .
a : math . gen . * 37 * , 8329 ( 2004 ) .
bibitemaldov r. aldrovandi , g.e.a .
matsas , s.f .
novaes and d. spehler , phys .
d * 50 * , 2645 ( 1994 ) .
h. casini and r. montemayor , phys .
d * 50 * , 7425 ( 1994 ) .
g. v. skrotskii , dokl .
nauk sssr * 114 * , 73 ( 1957 ) [ sov .
* 2 * , 226 ( 1957 ) ] .
schiff , phys .
* 1 * , 254 ( 1958 ) .
hari dass , phys .
* 36 * , 393 ( 1976 ) ; ann .
phys.(ny ) * 107 * , 337 ( 1977 )
. s. a. werner , j .- l . staudenmann and r. colella , phys .
lett . * 42 * , 1103 ( 1979 ) .
l. a. page , phys .
lett . * 35 * , 543 ( 1975 ) .
r. colella , a.w .
overhauser , s.a .
werner , phys .
lett . * 34 * , 1472 ( 1975 ) .
u. bonse and t. wroblewski , phys .
lett . * 51 * , 1401 ( 1983 ) .
f. combley , f.j.m .
farley , j.h .
field and e. picasso , phys .
lett . * 42 * , 1383 ( 1979 ) .
cooper , m.j .
alguard , r.d .
ehrlich , v.w .
hughes , h. kobayakawa , j.s .
ladish , m.s .
lubell , n. sasao , k.p.schler and p.a .
souder and d.h .
coward , r.h .
miller , c.y .
prescott , d.j .
sherden and c.k .
sinclair and g. baum and w. raith and k. kondo , phys .
* 42 * , 1386 ( 1979 ) .
g. papini , proc . of the @xmath451 canadian conference on general relativity and relativistic astrophysics , edited by r.b .
mann and r.g .
mclenaghan , world scientific , singapore , 1994 , pp .
107 - 119 .
dolgov , phys . rep . * 370 * , 333 ( 2002 ) .
d. singh , n. mobed and g. papini , phys .
lett . * 97 * , 041101 ( 2006 ) ; * 98*,069002 ( 2007 ) .
a. j. silenko and o. v. teryaev , phys .
d * 71 * , 064016 ( 2005 ) .
s. deser , gen .
. grav . * 1 * , 9 ( 1970 ) , also available from gr - qc/0411023 .
b. mashhoon , in _ black holes : theory and observation _ , edited by f. w. hehl , c. kiefer and r. metzler ( springer , berlin , 1998 ) , pp .
269 - 284 .
e. r. caianiello , lettere nuovo cimento * 41 * , 370 ( 1984 ) .
e. r. caianiello , rivista nuovo cimento * 15 * , no . 4 ( 1992 ) .
g. papini in _ imagination and rigor .
essays on eduardo r. caianiello s scientific heritage _ , edited by s. termini ( springer , new york , 2006 ) , pp .
119 - 132 .
g. lambiase , g. papini , r. punzi and g. scarpetta , class .
quantum grav .
* 23 * , 1347 ( 2006 ) .
a. j. silenko and o. v. teryaev , gr - qc/0612103 . n. obukhov , phys .
lett . * 86 * , 192 ( 2001 ) .
e. eriksen and m. korlsrud , nucl .
b * 18 * , 1 ( 1960 ) .
r. l walsworth , _ tests of lorentz symmetry in the spin coupling sector _ , in lecture notes in physics * 702 * , pp .
493 - 505 , springer , berlin , 2006 . | external gravitational fields induce phase factors in the wave functions of particles .
the phases are exact to first order in the background gravitational field , are manifestly covariant and gauge invariant and provide a useful tool for the study of spin - gravity coupling and of the optics of particles in gravitational or inertial fields .
we discuss the role that spin - gravity coupling plays in particular problems . |
calculating the rate of type ia supernovae ( sne ia ) and its dependence on redshift illuminates an important contributor to the metal enrichment history of the universe , the production rate of a specific cross section of stellar systems , and , when linked to star formation history , it can reveal the nature and relative fractions of the progenitor systems .
theory and recent empirical evidence argues that the explosions are derived from degenerate , hydrogen depleted stars , namely white dwarfs @xcite .
the nature of the companion star which must be present to feed mass onto these otherwise stable constructions , however , is more controversial
. it could very well be that there are two or more pathways leading to what we observe as sne ia including events from single degenerate progenitors ( e.g. a white dwarf accreting from a red giant star ; @xcite ) and double degenerate progenitors ( i.e. two white dwarfs ; @xcite ) .
this is a potential obstacle for the further use of sne ia as cosmological probes if the local population used to calibrate the peak magnitude to light curve width relation differs systematically from the distant , cosmologically significant population ( i.e. if high redshift events obey a different relation , cosmology studies may be biased ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the sne ia rate is one tool for resolving this issue .
locally , the quantity of progenitor systems demanded by the sne ia rate may be compared to the actual supply available ( e.g. * ? ? ?
? * ; * ? ? ?
* ) . at greater distances , the delay time distribution ( dtd ) , the distribution of sne ia progenitor systems lifespans from birth to explosion ,
can be compared to the distributions expected form various progenitor models @xcite .
the dtd can be recovered using the sne ia rate and star - formation histories of a targeted population of hosts @xcite , or by comparing sne ia over a range of redshifts to the cosmic star formation history ( e.g. * ? ? ?
* ; * ? ? ?
the results are consistent with a power - law distribution that favors short delays ( for a review , see @xcite ) .
indeed , sne ia appear to occur more frequently in late type galaxies with active star formation than in ellipticals ( e.g. * ? ? ? * ; * ? ? ?
recently , the lick observatory supernova search ( loss ) reported that the rate of sne ia per unit luminosity of the hosts is higher for low luminosity galaxies than luminous ones @xcite .
this so - called `` rate - size '' relation may be connected to a metallicity effect @xcite , since lower luminosity hosts tend to have lower global metallicities @xcite .
white dwarfs that form in low metallicity environments may have a mass distribution that is biased to more massive objects as compared to solar metallicity populations @xcite , and this in turn may lead to a greater supply of sne ia progenitors .
measurement of the sne ia rate was originally performed through searches targeting specific galaxies , and thus the rates derived were in units of the number of sne ia found per time per galaxy @xcite .
this evolved into a rate per galaxy luminosity snub or snuk , which give the number of sne ia per @xmath3 ( in the b or k - bands , respectively ) , per century .
@xcite combined targeted photographic and visual surveys and found a sne ia rate of snub@xmath4 averaged over all galaxy types .
loss finds roughly compatible sne ia snub rates in specific host types if they do not include the rate - size relation @xcite .
they also use an adopted k - band galaxy luminosity function @xcite to convert their measurements into volumetric rates . methods to discover sne ia without preference for host galaxies ( e.g. * ? ? ?
* ; * ? ? ?
* ) , have been widely adopted as the availability of wide field cameras has increased ( e.g. snls , essence , sdss - ii , ptf , panstarrs , skymapper ) .
sne ia rates from such surveys are most readily reported in volumetric units , sne iampc@xmath5yr@xmath6 .
included in this survey category is the sdss - ii , which was designed to bridge the gap between the local sne ia and the high redshift events used as cosmological probes @xcite . over a three year period , the sdss - ii spectroscopically confirmed several hundred sne ia in the range @xmath7 .
@xcite use a sub sample of ( normal ) sne ia discoveries to derive the @xmath8 rate , which is seen to increase roughly as a power - law with redshift , @xmath9 .
sne ia rate studies have been conducted at even higher redshifts , including work done with the supernova legacy survey ( snls ; @xcite ) , _ hst _
@xcite , and subaru @xcite .
the measured sne ia rates increase to @xmath10 , and then level off . for a recent compilation of volumetric sne ia rates from various surveys , see @xcite . in this paper , we determine the volumetric sne ia rate and contributions to this from dwarf and giant host galaxies based on discoveries from the rotse - iiib telescope .
section [ sample ] discusses how our sample was selected .
we discuss our unfiltered magnitude system and how this may relate to sne ia pseudo - absolute magnitude distributions ( pamds ) constructed from filtered observations in [ lf ] .
we also discuss the pamds derived from the loss and sdss - ii in this section .
for the later , we determine an empirical pamd based on a volume limited sub - sample of the sdss - ii after discussing the pamd model that was actually used in the sdss - ii sne ia rate calculation ( and which we adopt for comparison to the sdss - ii ) . in [ de ] we study our ability to detect point sources of various brightness as a function of seeing and limiting magnitudes , and we use this in [ survey_eff ] to determine our overall survey efficiency the probability of discovering sne ia drawn from a given population as a function of distance .
next , we measure multi - band photometry for the host galaxies of our supernovae and calculate rest frame absolute magnitudes in [ hosts ] .
sne ia rates are calculated in [ rates ] , and final conclusions are offered in [ conclusions ] . throughout this paper ,
we assume a flat , @xmath11kms@xmath6mpc@xmath6 , @xmath12 cosmology .
ccccccccccc 2004gu & dec 13 , 2004 & 12:46:24.7 & + 11:56:56 & 0.05 & 17.4 & 6 & @xmath1319.4 & y & y & + 2005bg & mar 28 , 2005 & 12:17:17.1 & + 16:22:17 & 0.02 & 17.1 & 10 & @xmath1319.0 & y & y & + 2005ck & jun 05 , 2005 & 13:02:18.7 & + 28:20:44 & 0.09 & 18.6 & 2 & @xmath1319.9 & n & n & @xmath14 + 2005cr & jun 24 , 2005 & 12:22:17.1 & + 12:23:49 & 0.02 & 16.1 & 10 & @xmath1319.1 & y & y & @xmath15 + 2005hj & oct 30 , 2005 & 01:26:48.3 & @xmath1301:14:17 & 0.06 & 17.6 & 17 & @xmath1319.7 & y & y & @xmath16 + 2005ir & nov 03 , 2005 & 01:16:43.7 & + 00:47:40 & 0.08 & 18.5 & 6 & @xmath1319.4 & y & y & @xmath16 + 2006an & feb 21 , 2006 & 12:14:38.7 & + 12:13:47 & 0.06 & 18.0 & 2 & @xmath1319.4 & n & n & + 2006cj & may 17 , 2006 & 12:59:24.5 & + 28:20:51 & 0.07 & 17.9 & 11 & @xmath1319.5 & y & y & @xmath15 + 2006ct & may 25 , 2006 & 12:09:56.8 & + 47:05:45 & 0.03 & 17.5 & 11 & @xmath1318.6 & y & y & + 2007if & aug 16 , 2007 & 01:10:51.4 & + 15:27:40 & 0.07 & 19.5 & 12 & @xmath1320.6 & y & n & @xmath17 + 2007kh & sep 07 , 2007 & 03:15:12.1 & + 43:10:13 & 0.05 & 18.9 & 10 & @xmath1319.3 & y & y & + 2007op & nov 04 , 2007 & 01:53:12.4 & + 33:44:34 & 0.09 & 18.5 & 6 & @xmath1319.1 & y & y & + 2007qc & oct 27 , 2007 & 11:57:04.7 & + 53:29:55 & 0.04 & 16.8 & 11 & @xmath1319.1 & y & y & + 2007sp & nov 14 , 2007 & 12:04:42.3 & + 49:11:09 & 0.02 & 16.8 & 7 & @xmath1318.2 & y & y & @xmath18 + 2007sw & dec 29 , 2007 & 12:13:36.9 & + 46:29:36 & 0.03 & 16.0 & 2 & @xmath1318.8 & n & n & + 2008e &
jan 04 , 2008 & 11:25:37.0 & + 52:08:26 & 0.03 & 18.2 & 18 & @xmath1318.7 & y & y & + 2008ab & jan 30 , 2008 & 11:34:45.9 & + 53:57:51 & 0.07 & 18.2 & 8 & @xmath1319.7 & y & n & + 2008ac & jan 30 , 2008 & 11:53:45.2 & + 48:25:22 & 0.05 & 17.6 & 15 & @xmath1319.2 & y & y & + 2008ad & jan 30 , 2008 & 12:49:37.2 & + 28:19:47 & 0.05 & 18.0 & 2 & @xmath1319.1 & n & n & + 2008ar & feb 27 , 2008 & 12:24:37.9 & + 10:50:17 & 0.03 & 16.9 & 17 & @xmath1318.8 & y & y & + 2008bg & mar 12 , 2008 & 12:51:11.9 & + 26:17:40 & 0.06 & 18.7 & 9 & @xmath1319.3 & y & y & @xmath15 + 2008by & apr 19 , 2008 & 12:05:21.0 & + 40:56:46 & 0.05 & 17.2 & 2 & @xmath1319.8 & n & n & + 2008bz & apr 22 , 2008 & 12:38:57.7 & + 11:07:46 & 0.06 & 17.7 & 6 & @xmath1319.2 & y & y & + [ table : disco ] our sample is drawn from the background population of supernovae discovered by rotse - iiib in the course of the texas supernova search ( tss ; @xcite ) and the rotse supernova verification project ( rsvp ; @xcite ) .
details of the full rotse - iiib supernova sample will be presented in a forthcoming paper . here
we summarize the key characteristics of the sample selection process .
the tss began survey operations in november of 2004 and was succeeded by rsvp in early 2008 .
our survey instrument , rotse - iiib , has a 0.45 m aperture and a @xmath19 degree field of view .
the search continues to operate and has been expanded to the other rotse - iii telescopes in australia , namibia , and turkey , but for this work we consider only discoveries made by rotse - iiib in texas prior to february 2009 .
we typically targeted galaxy clusters such as virgo , ursa major , coma , and abell clusters for our supernova search , but in a few exceptional cases we chose to target specific galaxies ( e.g. m31 ) . in the present work
, we remove this potential source of bias by selecting only the subset of sne ia found in the background of each field ( @xmath20kms@xmath6 , where @xmath21 is the redshift of the object targeted in a given field ) . during the search period , we detected 76 supernovae in total , and of these 46 were classified as sne ia .
after removing foreground and cluster supernovae , we are left with the 23 sne ia background events listed in table [ table : disco ] .
figures [ fig : winter_spring ] and [ fig : summer_fall ] show our sky coverage with the locations of the discoveries marked .
fields with the highest concentrations of nearby galaxies were typically observed every night as weather and season allowed . over the first few years we observed the less rich fields on alternating nights .
additional time was allocated for the survey in the rsvp era , and fields were added at that time .
but for the summer - fall fields .
] for both tss and rsvp , we employed image subtraction techniques to reveal time variable events .
the tss employed a modified version of the @xcite search code , and the rsvp used the process described in @xcite . for our survey , we typically observe each field multiple times each night . from these
we create 3 co - added frames representing the average of the first half of the nights data ( new1 ) , the average of the second half ( new2 ) , and the average for the whole night ( new ) .
we use filtering to remove particle events and other artifacts during the co - addition .
we then subtract the properly convolved reference template from each of these to generate subtracted frames ( sub1 , sub2 , and sub ) . to remove contamination from solar system bodies and remaining artifacts , we require a candidate to be detected ( @xmath22 ) at consistent celestial coordinates on each of sub1 and sub2 . the candidate is finally required to be detected at the @xmath23 confidence level on the combined subtraction , sub .
the typical fwhm for the survey was 3.2 pixels ( about 10 ) and the @xmath24 limiting magnitude for the nightly coadds was typically around 18.3 mag .
we further vet potential sn candidates against catalogs of known time - variable phenomena .
most of our fields were observed by the sdss as well , so we were able to remove contamination from un - cataloged variable stars by checking for matching `` host '' stars in the sdss data , which is complete to @xmath25 magnitudes below our survey depth .
remaining sources were vetted by human scanners .
all sources judged as possible transients were spectroscopically classified by us or others in the community .
this is a key distinguishing feature of our survey ; other wide - field transient searches generate more candidate supernovae than can be classified with the spectroscopic resources available , so a filtering process such as light curve fitting is typically employed to select only a subset for final spectroscopic confirmation .
table [ table : disco ] also notes the number of times that each discovery was recovered by the automated search pipeline ( @xmath26 ) .
a final analysis of the data adds low significance ( i.e. @xmath27 ) detections that are not passed by the automated selection cuts , and we determine the peak magnitudes for each supernova in our sample by fitting light curve templates to these final data sets .
we use r - band light curve templates for sne ia from loss , which were constructed by interpolating between the faint / fast sn 1999by and the bright / slow sn 1991 t @xcite . see [ lf ] for a discussion of our unfiltered spectral response and a comparison to the r - band .
we use mpfit.pro in idl to determine the best fitting template shape , and subtract off the galactic extinction in the r - band @xcite .
we employ the distance modulus ( based on independent distance indicators when available through ned or otherwise based on the redshift in a flat , @xmath11kms@xmath6mpc@xmath6 , @xmath12 cosmology ) to determine the peak absolute magnitudes shown in table [ table : disco ] .
sn 2004gu was first observed near peak , which adds some uncertainty to the date and magnitude of maximum light ; however , spectroscopic phase information is consistent with the derived date of maximum light , and thus we expect the peak magnitude estimate to be accurate .
according to its spectroscopic age , another event , sn2007sp , was first detected about two months after maximum light , so the light curve fit likely underestimates the peak .
although our sample is small , it contains several sne ia worth of special note . *
* sn 2004gu * is a significant outlier on the hubble diagram , and other authors have grouped it with peculiar events like sn 2006gz and sn 1999ac , which may or may not be linked to super - chandrashekar explosions @xcite . * * sn 2006ct * showed the characteristic features of a type ia supernova , but the blue - shift of the line minima near maximum light was only @xmath28kms@xmath6 @xcite .
the spectra show some similarities to the highly peculiar sn 2002es @xcite , which was also hosted by an early type galaxy , but the photometric decline of sn 2006ct @xcite is slower and more reminiscent of sn 2002cx @xcite and sn 2005hk @xcite . * * sn 2007if * is the most luminous type ia supernova known , and it may be the result of a super - chandrashekar mass explosion @xcite . * * sn 2007qc * was hosted by an extremely low luminosity dwarf galaxy ( see [ hosts ] ) . at @xmath29 ,
the host is among the faintest detected for any sne ia . *
* sn 2008ar * showed spectroscopic evidence for high velocity material and unusually strong absorption in the ir triplet @xcite , which is reminiscent of sn 2007le @xcite .
in addition , at least 7 of our sne ia ( 2004gu , 2005ck , 2005hj , 2006cj , 2007if , 2008ab , 2008by and possibly 2005bg ) show either unusually high luminosities or are spectroscopically classified as sn 1991t/1999aa - like .
this appears to be a high fraction even for a magnitude limited survey when compared to the loss census of sne ia in targeted galaxies @xcite .
our survey is flux limited , and as a result we have a bias against selecting low luminosity sne ia .
this bias is exacerbated by our removal of discoveries made at or below the redshift of the targeted galaxy clusters . to determine the event rate for the sne ia population as a whole
, we must therefore correct the number of events in our observed sample for the fraction we are likely to have missed due to , among other concerns , their lower luminosities ( see [ survey_eff ] ) .
this requires knowledge of the intrinsic luminosity function for all sne ia in our search volume combined with the host galaxy absorption distribution . in this section ,
we consider the pseudo - absolute magnitude distribution ( pamd ) for sne ia , which is the distribution one would obtain by correcting the peak observed magnitudes of a complete sample of sne ia for distance and galactic extinction but not for host galaxy extinction .
this distribution can directly be used to test the selection efficiency of a flux limited survey ( see [ survey_eff ] ) .
first , we consider in [ filtered_v_unfiltered ] how the available sne ia luminosity functions derived ( mostly ) through filtered observations may compare to our unfiltered survey data .
we then discuss the pamd compiled by loss from their targeted , but volume limited sample @xcite in [ losslf ] , and we derive an empirical pamd from a volume limited sub sample of the sdss - ii survey in [ sdsslf ] . we find that the loss and sdss - ii pamds are not consistent , which is discussed in [ loss_v_sdss ] .
we calibrate our instrumental magnitudes against the usno - b1.0 r2 system since some of our search fields are outside of the sdss footprint .
for fields covered by both the sdss and usno - b1.0 , we compared the usno - b1.0 photometry to sdss photometry converted to the r - band using @xmath30 , and we find the two systems agree to about 0.01mag on average and have a field to field dispersion of about 0.14mag .
thus our unfiltered rotse - iiib magnitudes should be , on average , similar to the r - band , but objects with spectral energy distributions that differ significantly from the field stars employed in the calibration may be offset from their true r - band values .
we estimate the potential offset between our unfiltered magnitudes and the true r - band system for sne ia as follows .
we adopt the spectral energy distribution of a g5v star @xcite as representative of a typical field standard used in calibrating the photometric zeropoint of our unfiltered imaging data .
we scale this template to zero magnitude in the r - band , and then measure the synthetic flux using the approximate transmission function of rotse - iiib @xcite , which is plotted in figure [ fig : bandpass ] .
we next scale a template spectrum for a branch - normal sn ia at maximum light to zero magnitude in the r - band and measure its flux in the rotse - iiib bandpass .
the ratio of this and the scaled field star flux ( also measured in the rotse - iiib bandpass ) gives the expected departure of the rotse - iiib magnitudes from the true r - band system .
we measure this offset in the range @xmath31 by redshifting the sn ia template , and find that our unfiltered magnitudes may be 0.05 to 0.10mag brighter than the true r - band system .
we repeat this process for a high luminosity sn ia ( sn 1999aa ; @xcite ) , and a low luminosity event
( sn 1991bg ; @xcite ) .
we find that a sn ia with the same r - band brightness as a field standard will generate a magnitude in the unfiltered system ( calibrated against the r - band ) that is within @xmath32mag of its true r - band magnitude . while this shows that our observed magnitudes should be similar to the true r - band system ,
a more pertinent concern is how the derived absolute magnitudes will compare .
in particular , we have calculated absolute magnitudes from our unfiltered data by simply subtracting off the distance moduli and galactic ( r - band ) extinction terms ; we have not included a `` k - correction '' term to account for the changing rest frame bandpass of our unfiltered system over the modest redshift distribution of our sample .
below , we compare the pseudo absolute magnitudes derived from our data to rest frame ( k - corrected ) r - band magnitudes calculated for the same sources based on filtered photometry to check for an offset .
the cfa3 sample includes mlcs2k2 @xcite fits for four of the supernovae in our sample @xcite , including two that were also in the sdss - ii sample @xcite .
we calculate the pseudo - absolute magnitudes from these fits using the @xmath33 parameters and adding back in the host extinction term . for this comparison
, we calibrate the rotse - iiib magnitudes against the sdss field stars ( converted to the r - band as described above ) to remove the usno - b1.0 vs. sdss - ii offset , which we find to be biased in this small overlapping sample .
our pseudo absolute magnitude estimate for sn 2005hj is about 0.1mag brighter than the fits from the cfa3 and the sdss - ii . for sn 2005ir ,
the sdss - ii and cfa3 differ from each other by about 0.1mag , which is comparable to the combined measurement error .
our absolute magnitude for sn 2005ir is about 0.02mag brighter than the sdss - ii value and 0.09mag fainter than the cfa3 value .
our measurements of sn 2006an and sn 2006cj are 0.03mag brighter and 0.16mag fainter than the cfa3 peak magnitudes , respectively .
additionally , the nearby supernova factory has reported a peak v - band magnitude and @xmath34 color for sn 2007if that implies a peak absolute r - band maximum of about @xmath35mag .
calibrating against the sdss r - band ab magnitudes , @xcite found the rotse - iii observations peaked at @xmath36 .
field stars will typically `` brighten '' by about 0.22mag when converting from the sdss r - band system to cousins r - band ( vega ) magnitudes , which shifts the target photometry by the same amount .
our estimate for the peak of sn 2007if is approximately 0.15mag brighter than the nearby supernova factory value .
to , summarize , of the 7 pseudo - absolute magnitude measurements available in the literature based on ( k - corrected ) filtered photometry , our rotse - iiib estimates are @xmath32mag brighter in three cases , @xmath32mag fainter in 2 cases , and roughly the same in the remaining 2 cases .
on average , our values are 0.02mag brighter .
we conclude that there is no significant systematic offset between our unfiltered , pseudo - absolute magnitude estimates and the ( mostly ) r - band system that defines the luminosity functions considered below ( [ losslf ] and [ sdsslf ] ) , and these should agree to about @xmath32mag for individual events .
we address the impact the possible systematic error in our magnitude system with respect to the loss and sdss - ii pseudo - absolute magnitude distributions will have on our rates in [ conclusions ] .
the lick observatory supernova search ( loss ) has recently published a distribution of pseudo - absolute magnitudes for sne ia from their volume limited search @xcite .
like rotse - iiib , the loss survey engine ( the katzman automatic imaging telescope , or kait ) runs with an unfiltered ccd imager , but follow - up observations are often obtained through standard broadband filters . the sne ia luminosity function reported by @xcite is primarily constructed from r - band data , which the authors describe as the best match to their unfiltered survey data , and also from unfiltered observations as well . in the lower panel of figure [ fig : bandpass ]
we show the response curve for the cousins r - band compared to the typical unfiltered response of an apogee7 camera , which is used by loss .
it is important to note that the loss sne ia luminosity function was drawn from a sample that was highly biased with respect to the host galaxy properties . in particular ,
the loss targeted mainly high luminosity galaxies .
it has been demonstrated that certain sne ia sub - types prefer certain host galaxies @xcite , so the luminosity distribution of the loss sample may be biased with respect to the larger population from which our sample is drawn .
an obvious difference between the loss and rotse - iiib sn ia samples is the presence of high luminosity ( @xmath37 ) events in the latter .
although we could follow previous works and simply discard our most luminous sne ia , we instead chose to calculate an inclusive rate .
we must , therefore , augment the loss luminosity function to account for such events or else our effective survey volume will be under estimated . lacking detailed demographics for this population ,
we simply assume that the population of high luminosity sne ia represents only about 1% of sne ia in a given volume ; if it were much more these would be detected more frequently , and if it were much less we could not expect to find any in our small sample ( see [ survey_eff ] ) .
we assume that sn 2007if was a particularly luminous example of this population ( as noted above , it is the most luminous sn ia known ) and choose a half gaussian with a peak at @xmath38 and @xmath39mag to stand in for the unknown luminosity distribution .
we will use this augmented version of the loss pamd to calculate the sne ia rate in [ rates ] .
for our rate calculations in [ rates ] , we will also use the same bimodal - gaussian luminosity function attenuated by a host galaxy absorption distribution that was assumed by @xcite in deriving the sdss - ii rate .
the model is a revised version of the distribution assumed by @xcite that takes into account the findings of @xcite .
the difference makes for only a negligible change in our rate measurment .
the intrinsic sne ia luminosity function is defined in terms of the mlcs2k2 light - curve shape / luminosity parameter , @xmath33 .
the distribution of @xmath33 is defined as a gaussian with @xmath40 for @xmath41 and @xmath42 for @xmath43 , and it is truncated to lie within the range ( @xmath44 ) .
the final , pseudo - absolute magnitude distribution is achieved by adding host galaxy absorption drawn from a distribution with @xmath45 . with our choice of @xmath46 ,
peak pseudo - absolute magnitudes are then defined as ( cf .
* ) : @xmath47 when applying the sdss - ii pamd model to our rotse - iiib discoveries , it is important to note that the sdss - ii rate is based on a sub - sample of sne ia , and to be consistent with the sdss - ii pamd model we must also remove objects from our sample .
in particular , the sdss - ii includes only the fraction of the population that is well fit by mlcs2k2 ( @xmath48 ) with a light curve width parameter @xmath49 .
such a cut excludes high luminosity sne ia , and in particular , sn 1999aa - like events @xcite .
we note that this is a common choice .
the supernova legacy survey ( snls ) , for example , has presented a volumetric rate for sne ia at @xmath50 , but the spectroscopically confirmed type ia sn 2003fg ( snls-03d3bb ) is not included @xcite .
indeed , the selection cuts that are applied in selecting candidates for spectroscopic follow - up may also pose a bias against high luminosity sne ia . we therefore remove events with peak luminosities brighter that @xmath51 when using the model pseudo - absolute magnitude distribution from the sdss - ii .
we next look at the actual distribution of pseudo - absolute magnitudes from a volume limited sub - sample of the first year sdss - ii sne ia discoveries @xcite , which can be compared to the assumed model as well as the loss pamd . according to @xcite
, the sdss - ii could have detected even a sub - luminous ( sn 1991bg - like ) sn ia out to a redshift of @xmath52 , so we assume the survey to be volume limited out to this limit . there were 29 spectroscopically
confirmed ( or probable ) sne ia from the first year of the sdss - ii search within this redshift limit .
in addition , @xcite lists one probable sn ia ( sdss - ii sn09266 ; based on photometric constraints ) with a host redshift of @xmath53 . this source is not included in the sample of @xcite due to an additional cut on the photometric screening . a second photometrically probable source , sdss - ii sn09739 , with a spectroscopic redshift for its host and a third , photometrically probable sn ia ( sdss - ii sn11092 )
are also given for a total of 32 sne ia in the volume limited sample .
we now derive an empirical sne ia pamd from this volume limited sub sample of the first - year sdss - ii .
we adopt the temporal selection cuts of @xcite and @xcite to remove events with poorly constrained maxima , but unlike these samples , we include peculiar events . for the bulk of the sample ,
we derive the peak absolute r - band magnitudes from the mlcs2k2 @xcite fit parameters given in @xcite .
we add the host extinction estimate back in to retain the pseudo - absolute distribution required for rate studies .
two peculiar sne ia , sn 2005hk and sn 2005gj , were cut from the @xcite sample ; it is still a matter of debate if these events truly belong to the type ia class ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we include them here for completeness . for sn 2005hk , we take the loss measurement for the r - band absolute magnitude @xcite , and for sn 2005gj we take the peak absolute r - band magnitude from @xcite and convert this to r - band .
we similarly use the sdss - ii photometric measurements of @xcite to estimate peak absolute r - band magnitudes for two spectroscopically probable sne ia : sn 2005je and sdss - ii sn06968 .
finally , we choose to include an estimated peak magnitude for sdss - ii sn09266 .
there is no published light curve or peak magnitude for this event , so we assign it an _ ad hoc _ pseudo - absolute magnitude of @xmath54 mag based on the heavy extinction reported ( @xmath55 ; @xcite ) .
we thus have peak magnitudes for 22 of the first year sdss - ii sne ia ; eight of the remaining events have poorly constrained maxima .
the portions of the light curves sampled by the sdss - ii , however , imply a similar distribution of peak magnitudes to the well constrained sample .
photometry is not available for the other two photometrically selected events ( sdss - ii sn09739 and sn11092 ) , so we can not include these .
the peak r - band pseudo - absolute magnitudes ( corrected to @xmath11kms@xmath6mpc@xmath6 ) for the volume limited sdss - ii sample are listed in table [ table : sdss ] .
rlcc 722 & 2005ed & & + 739 & 2005ef & & + 774 & 2005ei & & + 1241 & 2005ff & -18.97 & + 1371 & 2005fh & -19.39 & + 2102 & 2005fn & & + 2561 & 2005fv & -19.03 & + 3256 & 2005hn & -19.11 & + 3592 & 2005 gb & -19.30 & + 3901 & 2005ho & -19.33 & + 4524 & 2005gj & -20.47 & @xmath14 + 5395 & 2005hr & -19.36 & + 5549 & 2005hx & -19.23 & + 5944 & 2005hc & -19.41 & + 6057 & 2005if & -18.95 & + 6295 & 2005js & -17.72 & + 6558 & 2005hj & -19.36 & + 6773 & 2005iu & -19.28 & + 6962 & 2005je & -19.28 & @xmath16 + 6968 & & -19.06 & @xmath16 + 7147 & 2005jh & -18.96 & + 7876 & 2005ir & -19.29 & + 8151 & 2005hk & -18.36 & @xmath15 + 8719 & 2005kp & -19.30 & + 9266 & & -16.00 & @xmath17 + 9739 & & & + 10028 & 2005kt & -19.06 & + 10096 & 2005lj & & + 10434 & 2005lk & & + 10805 & & & + 11067 & 2005ml & & + 11092 & & & + [ table : sdss ] a comparison of the loss sne ia luminosity function and the sdss - ii first year volume limited sample shows a striking result : the samples do not agree ( fig . [ fig : sneia_lf ] )
. about 66% of the loss sne ia are fainter than @xmath56 mag , but only about 27% of the events in the sdss - ii sample are this faint .
we discuss the implications of this disagreement in [ conclusions ] . only sn 2005gj , which appears to draw its power through interaction with a hydrogen shell and could technically be classified as a typeiin , approaches the luminosity of our higher luminosity events , such as sn 2007if @xcite .
also shown in figure [ fig : sneia_lf ] is the r - band pamd model that was assumed in calculating the sdss - ii sne ia rates @xcite , which also appears to differ from the selected sdss - ii population . from a ks test , we find that there is only about a 4% chance that the volume limited sdss - ii sne ia sample is consistent with the attenuated , bimodal - gaussian model of @xcite .
the reddening distribution assumed in the model predicts that it is extremely improbable to have highly extincted events such as sdss - ii sn09266 ( again , this was not spectroscopically confirmed and not included in @xcite ) .
we repeated the ks test using only the members of the volume limited sample that were also used by @xcite in their rate calculation and find the probability that these events were drawn from the assumed model is only about 3% .
except for a few low - luminosity events , the observed sdss - ii sample is systematically biased to higher luminosities than the model predicts .
we discuss the sdss - ii model as applied to the rotse - iiib sample in [ rates ] .
in this section we determine the detectability of sources as a function of observed magnitude relative to the limiting magnitude of a given image .
this curve will be used later to determine the probability of detecting simulated sources with various observed magnitudes , which is needed to calculate our overall survey efficiency . to determine the probability that a source of a given brightness will be detected , we perform a monte carlo simulation .
we add simulated point sources to our co - added images and determine the fractions recovered as a function of magnitude .
this is done relative to the limiting magnitude of the images and coadds recorded in the logs , which is calculated from the 90th percentile magnitude of all objects extracted by sextractor @xcite .
our reference images are typically constructed from dozens of the best individual frames from each field , so they contribute negligibly to the noise on the subtracted frames .
we therefore choose to calculate the detectability of sources directly from blank sky regions of the co - added images without performing the computationally expensive process of image subtraction .
a consequence of this is that we do not account for the effects of contaminating host galaxy light .
other surveys that employ image subtraction have previously tested the effects of host light on detectability and found very little change from bright hosts to blank sky @xcite .
nonetheless , our procedure may overestimate our detection probability for supernovae hosted by bright galaxies ( i.e. the rate may be _
under_estimated ) .
our results do , however , have direct applications for objects with faint hosts or otherwise dark backgrounds . on a given image
, the sensitivity to objects of a given magnitude will vary with location in the field due to variations in the instrumental psf as well as structure in the atmospheric transmissivity ( e.g. passing clouds ) , which can be resolved by our wide field . to account for these effects ,
we locate a set of isolated field standards on each image to use in calculating the local psf and zeropoint .
we divide each image into a grid with as many cells as can be made while maintaining around 100 or more field standards in each ( typically @xmath57 cells ) , and then calculate the average zeropoint and psf for each of these cells .
the psf is found using the daophot routines in idl , and as noted above ( [ filtered_v_unfiltered ] ) , we use the usno - b1.0 r2 system for our zeropoints . to locate appropriate blank sky regions
at which we may add our simulated point sources , we first run sextractor on each unadulterated image with a low detection threshold to identify all objects and artifacts in the field .
we then convolve the object mask image from sextractor with the average psf and select from the remaining , unmasked pixels when placing simulated sources .
we find that even the small perturbations induced by adding a very faint ( @xmath58 ) test source can sometimes elevate a faint ( @xmath59 ) source previously ignored by sextractor into a spurious detection .
it is therefore important to avoid locating test sources near such areas , which is why we use a lower extraction threshold when creating the object mask .
we use the local psf and zeropoint determined above to determine the shape of the test source and scale it to a randomly chosen magnitude . using the dao_value routine
, we create a thumbnail image of the source .
we then add this source to the image one pixel at a time by drawing from a poissonian distribution with the expectation values set by the thumbnail pixel values and accounting for the detector gain .
we finally run sextractor with the same parameter file used in the search . in order to simulate a statistically meaningful number of test objects without biasing the extraction process
, we place only 20 test sources in each cell and repeat the simulation multiple times for each image . the minimum pitch for the test objects
is set to be at least four times the average fwhm , so they do not interfere with each other . for the survey , candidates must be detected at a signal to noise greater than a set limit and they must have a fwhm between half and double the image average to be considered . both of these quantities are effected by the noise in the image and poisson statistics .
we therefore define our detection efficiency from the number of simulated point sources that are detected by sextractor with parameters that pass these selection cuts relative to the total number simulated .
we performed this test on over 7000 co - added images made from 2 or more frames . for comparison ,
our survey includes about 55000 master subtractions .
figure [ fig : de_curves ] shows the average @xmath23 detection efficiency curves for isolated objects . as can be seen , the shape of the curve is robust to changes in the limiting magnitude , but the effect from variations in the fwhm is quite noticeable . .
the top panel shows that the shape of the detection efficiency curve is nearly independent of the limiting magnitude for a fixed fwhm .
the lower panel shows that for a fixed @xmath60 , the shape of the detection efficiency curve depends strongly on the average fwhm of the image . for both plots ,
the 50% completeness is brighter than the logged limit mainly because the simulations require a @xmath23 detection while the logged limits are derived from all extracted sources ( i.e. @xmath24 detections are included ) . ]
accurate limiting magnitudes were not recorded for the co - added images during the first half of our survey , but we do have the limits for the individual images .
using our later data , we determine an empirical relation between the limits of the individual images and the coadds derived from them .
we started by assuming the limiting magnitudes for each individual image is approximately equal to three times the sky noise in the aperture . inverting this
, we can find the sky noise expected on the co - added frame from the sum of the individual images , and turn this into a limiting magnitude guess .
comparing these guesses to the actual values recorded in the logs over the second half of our survey , we find that the agreement is good for the deepest data , but there is a systematic offset for more shallow coadds . we perform a linear fit to these data and use it to derive magnitude limits for the coadds created in the beginning of the survey .
it is not enough to merely detect a source to count it as a discovery ; each such candidate must pass additional screening to be included in our sample .
this includes both machine cuts designed to reject false positives and cuts made by the humans who have the final say in elevating candidates to discoveries .
often it is straight forward to determine the effects of the machine cuts through a monte carlo process : simulated supernovae drawn from the expected population are `` observed '' using the actual survey cadence and limiting magnitudes , and the resulting `` candidates '' are passed through the same machine vetting process as the real data .
the sensitivity of the survey to sources with various apparent brightnesses and light curves can then be directly determined by noting the fraction of the simulated supernova that pass .
the effects humans play in the ultimate selection of candidates could in principle be determined through a similar simulation if the test were performed concurrently with the actual survey to account for learned behavior . to get statistically meaningful results ,
the human scanner would need to be presented with an enormous volume of simulated data , which is far too large of a burden on the researchers to be done in practice , although previous studies have performed limited versions of such human experimentation , which guide our tests .
as stated above , the basic machine cuts applied to the raw candidates are 1 ) the fwhm should be between 0.5 and 2 times the image average , and 2 ) the signal - to - noise should be at least @xmath61 on sub1 and sub2 , and @xmath23 for sub .
to help reduce the number of false positives created by imperfect subtraction of galaxy light , we also required that candidates near the cores of cataloged galaxies show an increase of at least 5% in flux over the reference template ( given the bright limits of our survey and prior coverage of our search fields by galaxy surveys , it is fair to assume that almost all galaxies as bright or brighter than our sne ia sample were cataloged ) .
we are mainly sensitive to sne ia brighter than @xmath62 absolute , so hosts fainter than about @xmath63 are unaffected by this cut . of course , when the host galaxy is resolved , only the fraction of light coincident with the candidate matters in this percentage increase cut . for example
, we recorded a 50% increase for sn 2005bg ( @xmath64 ) , which was discovered near the core of a @xmath65 host .
candidates coincident with sources not listed in the galaxy catalogs ( typically foreground stars ) were required to show a 15% increase .
the remaining candidates located more toward the outskirts of cataloged galaxies or in isolated regions were not required to show a minimum increase in flux .
we do not include this selection criteria in our monte carlo simulations , so our efficiency in discovering supernovae in luminous host galaxies may be overestimated , which again means that our rate in giants may be underestimated .
the basic setup for our efficiency calculation is similar to that used in previous rate studies ( cf . * ? ? ?
we determine the efficiency as a function of luminosity distance by considering a series of thin shells over which the efficiency ( and rate ) can be assumed to be constant . in each of these shells
we simulate a number of sne ia randomly drawn from the pseudo - absolute magnitude distributions discussed in [ lf ] , choose an appropriate light curve template and scale it to the given distance , and select a random date for maximum light . by comparing the expected magnitudes on the dates each simulated supernova
would have been observed by our survey and using the actual survey detection limits on those nights , we determine the fraction of simulated sne ia that could have been detected in each distance bin .
we take into account the detection efficiency curve discussed in [ de ] appropriate to a given observation based on the limiting magnitude and fwhm recorded ( or estimated ) .
for each observation epoch of each simulated supernova , we draw random numbers and compare these to the detection probabilities to determine if a simulated detection is made on each of the sub1 , sub2 , and sub . for each of our 177 fields
, we simulated 100 sne ia in each of 360 distance shells spaced logarithmically from 40 to 1000mpc , which totals to over six million simulations for each pamd assumed . to naturally account for the changing number of survey fields , we allow simulated supernovae to peak anytime in any field during the entire survey period including off season . to assess how accurately these simulations model the actual survey
, we compare the luminosity distance , discovery magnitude , peak absolute magnitude , and detection number distributions predicted by the monte carlo simulations to the observed distributions .
figure [ fig : initial_cdf ] shows that the survey model discussed thus far is a poor representation of the true sample if the augmented loss pamd is adopted ( the same result holds if the sdss - ii model pamd is used ) .
the model predicts the sne ia should be further away than observed , discovered at magnitudes fainter than observed , and the simulated supernovae are typically detected on fewer nights .
this last result is key to understanding the failure of the model thus far .
the model predicts that @xmath66% of the sne ia should be detected by the automatic search selection cuts on exactly one night .
after all , the largest volume element is for the most distant events that are only detectable at their peak .
since we have zero sne ia detected on just one night , it would seem that our survey is biased against such `` one - nighters . ''
a possible explanation for this is the influence of the human scanners who have the final say in selecting targets .
although it was never a formal requirement of the search that events be detected on more than one night ( and in fact , we triggered spectroscopic follow - up more than once on the same night based on a first detection ) , the informal requirement that candidates `` look good '' may have created a bias against the weakest detections .
indeed , @xcite studied the response of human scanners to simulated supernovae and determined that there is such a bias against one - nighters . to account for our apparent selection bias against weak detections , we apply a cut to the simulations and the real sample alike on the number of nights detected .
we find that setting a minimum number of 5 nights , the model and actual distributions agree for both the augmented loss pamd and the sdss - ii model .
this cut preferentially removes the most distant supernovae with the faintest observed magnitudes , so after applying this single correction , all of the model distributions considered agree with the actual observed sample ( fig .
[ fig : final_cdf ] ) .
the minimum number of detections was set by calculating the probability that the observed sample was drawn from the model distribution via a ks test .
we assign a consistency probability , @xmath67 , by randomly drawing from the model distribution and comparing the maximum displacement in the cumulative distributions of the model and each draw . for 4 or fewer detections and the loss pamd
, only 0.3% ( or 0.4% with the sdss - ii pamd model ) or less of the random draws have a displacement as large or larger than the real data compared to the models , so we reject these models . with a minimum of 5 detections ,
the probability rises to 9% ( 5% ) , so we can not reject the null hypothesis that the data were drawn from this model distribution . in this case , the probabilities for the other three distributions are considerably higher , so the model does not appear to be inconsistent with the data .
we also tested the effects of systematically shifting the input pamds in brightness .
if we shift the augmented loss pamd 0.3mag fainter , both the predicted distance and absolute magnitude distributions become incompatible with the observed distributions ( as above we require pipeline detections on 5 or more nights ) . shifting the loss pamd 0.2mag brighter ,
the predicted and observed distance distributions fall into agreement , but the absolute magnitude distributions are not compatible . with the sdss - ii pamd model , a shift of either 0.2mag
fainter or brighter can be ruled out .
this demonstrates that we would be able to detect if the magnitude system of the pamd employed was systematically biased with respect to our unfiltered magnitudes by more that expected from our checks in [ lf ] .
we will use the allowed magnitude offsets to estimate our systematic error in [ conclusions ] . )
compared to the observed sample ( green ) after correction for selection bias . ]
@xcite found that lower luminosity galaxies produced more sne ia per unit luminosity than high luminosity hosts . in order to compare the host luminosity distribution of our sample to the sdss - ii and the expectations of the loss luminosity specific rates coupled with a galaxy luminosity function
, we have measured ugriz photometry for each supernova host .
we make deep stacks from the sdss dr7 data @xcite in each of the 5 bands , excluding data obtained while the supernovae were active ( assumed to be @xmath68d@xmath69d in the rest frame ) , and retain the median using a slightly modified version of the montage package .
cuts on the sky background and psf size are used to reject the worst images ( about a third of the available data ) .
we identify the host galaxies following @xcite .
briefly , we use the sextractor shape parameters ( @xmath70 , @xmath71 , and @xmath72 ) to define an ellipse for each object in the field , and we identify the host as the object that includes the supernova position inside its ellipse with the smallest possible scaling unless the scaling factor , @xmath73 , is greater than 5 . if this is the case , we perform forced photometry at the location of the supernova in a @xmath74 diameter aperture .
otherwise we run sextractor in dual image mode to measure the host galaxy photometry in consistent aperture sizes on the best matching host .
we choose to perform the object detection and aperture definition on the r - band images .
a few host galaxies are worthy of special mention : * 1 ) * the host of sn 2007if is not detected in the co - added sdss data .
we instead use the g - band magnitude of the host reported by @xcite to scale a spectrum of the host galaxy and then measure synthetic photometry in the sdss bands from this . *
2 ) * the loss team assigned the giant galaxy kug1259 + 286 as the host of sn 2005ck , but the sdss redshift for this galaxy is incompatible with the supernova @xcite .
there is a small galaxy nearby , sdssj130218.65 + 282046.8 , for which we find a spectroscopic redshift consistent with sn 2005ck ; however , this galaxy is slightly too far away to meet the host selection criteria described above ( as would be kug1259 + 286 ) , and it is therefore considered hostless . identified or not , the host must be a dwarf , so this should not affect our conclusions . * 3 ) * sn 2007sp is located a considerable distance ( @xmath75 ) from a large galaxy at a similar sdss derived redshift .
although this may yet be the host , for consistent application of the @xcite selection method , we must consider this target to be hostless . *
4 ) * there is a galaxy coincident with the location of sn 2007kh , but spectroscopic follow - up shows this to be a @xmath76 galaxy , which is well in the background . the true host of sn 2007kh is therefore a mystery , so we simply adopt the photometry of the background source , assume the redshift of sn 2007kh , and take this as an upper limit on the host luminosity . *
5 ) * the location of sn 2007op is outside of the sdss dr7 footprint ; however , the host is included in dr8 , so we adopt the photometry from the dr8 pipeline @xcite . in addition , we perform the same analysis on the full set of 516 spectroscopically confirmed , probable , or photometrically probable @xmath8 sne ia from the sdss - ii @xcite .
the larger number of visits to stripe82 over the course of the sdss - ii results in co - added images ( again excluding those contaminated by supernova light ) with limiting magnitudes considerably deeper than the search itself . as a result
, even @xmath77 dwarf galaxies can be detected out to the @xmath78 limit of the sample .
as with the rotse - iiib sample , we manually reviewed the selection of each host galaxy and noted any special circumstances that might produce misleading results ( the potential error rate appears similar for the rotse - iiib and sdss - ii samples ) . as needed , we adjusted the sextractor parameter deblend_mincont to either ensure large galaxies were not inadvertently split into smaller pieces or to attempt to split a likely galaxy blend .
the latter issue was not always resolved , both for the rotse - iiib and sdss - ii hosts .
of particular note , a limited number of sdss - ii hosts appear to be dwarf galaxies heavily blended with ( if not merging into ) larger galaxies , such as sdss - ii sn03592 . when such blends are not separated by sextractor , we perform forced aperture photometry at the location of these supernovae to better reflect the luminosity of the dwarf hosts .
since the host galaxies are distributed over a range of redshifts , we must perform k - corrections to compare the galaxy luminosities in the same rest - frame band passes .
this is accomplished using the kcorrect.pro package , version 4.2 @xcite .
for comparison against the literature , we calculate absolute b- and k - band magnitudes for our hosts in the vega system .
these are recovered from the best fit galaxy template through kcorrect.pro .
we compared our derived k - band magnitudes ( based only on the optical sdss photometry ) for a few low - redshift hosts to detections in the 2mass survey @xcite , and found good agreement .
the four lowest luminosity hosts from the rotse - iiib and nearby sdss - ii samples are shown in figure [ fig : faint_hosts ] and the four highest luminosity hosts from each sample are shown in figure [ fig : bright_hosts ] . in figure
[ fig : hosts ] , we compare the host galaxies of the rotse - iiib sample to the 20 hosts from the sdss - ii sample with redshifts @xmath79 @xcite .
these two samples cover a similar redshift span and can thus be compared directly .
as can be seen , our sample includes relatively few high - luminosity host galaxies and , of particular interest , a number of very low luminosity hosts .
a ks test gives only a 4% chance that these rotse - iiib and low redshift sdss - ii samples are drawn from the same population .
if we assign sdss - ii sn03592 to the larger host , this probability drops to just 1% . the agreement likelihood with the full population of galaxies hosting the sne
ia from the sdss - ii rate calculation is about 0.2% . for this larger sample
, a small fraction of cases that have undetected hosts also have limiting magnitudes brighter than detected rotse - iiib hosts , so the exact distribution below @xmath80 is uncertain .
however , this does not affect our key result : 43% of the rotse - iiib hosts are @xmath81mag dwarf galaxies while only 17% of the full sdss - ii and 2 out of 20 from the @xmath79 sample are hosted by such dwarfs ( including sdss - ii sn03592 ) .
were not detected .
we employ upper limits for these events , but this does not affect our results .
the host galaxies of the rotse - iiib supernovae ( green line ) are typically fainter that hosts from the sdss - ii search in a similar redshift range ( red line ) , and the rotse - iiib sample includes a larger fraction of dwarf hosts than seen or predicted through the other surveys shown . ] figure [ fig : hosts ] also shows the host galaxy distributions from the snls @xcite . as with the sdss data
, we measure the host galaxy photometry using sextractor in dual image mode , this time using the i - band as reference , and we determine the rest frame absolute magnitudes with kcorrect.pro .
we use the megapipe median image stacks provided through cadc @xcite . to check
if these measurements are biased by the inclusion of images with the supernovae detected in the determination of the medians , we compare our i - band observer frame measurements to @xcite .
surprisingly , we find that our measurements are actually systematically fainter than the @xcite values by @xmath82mag .
for example , for the host of snls 03d1ar , we find @xmath83 without correcting for 0.05mag of galactic reddening @xcite .
this agrees with the value in the megapipe catalog , and it is consistent with the sdss dr8 ( @xmath84 ) , but it is fainter than reported in @xcite who find @xmath85 . aside from this
systematic offset , we do not find any evidence that our photometry is biased by the supernova light even in the faintest hosts .
it is interesting that the b - band distributions of the sdss - ii and snls host galaxies appear to diverge toward the bright end , but there is a 12% chance that a random draw from the sdss - ii population would lead to an equal or larger displacement . as with the sdss - ii ,
the deep coadds are sufficiently deep to detect all snls hosts brighter than about @xmath80 , and again the rotse - iiib sne ia prefer fainter hosts . although loss targets specific galaxies , which biases the sample against low luminosity hosts , they adopt the @xcite k - band galaxy luminosity function and factor in their rate - size relation to determine the volumetric sne ia rate @xcite .
we include the k - band host galaxy distribution expected by the loss rate calculation in the lower panel of figure [ fig : hosts ] .
clearly , this expected distribution is different than the observed sdss - ii and snls distributions .
specifically , loss assumes that 80% of sne
ia in a given volume should come from hosts fainter than about @xmath86mag , but the sdss - ii finds only about 40% of its @xmath79 hosts are fainter than this luminosity .
this result holds qualitatively even if the correction for the rate - size relation is removed ; among the giants , the sdss - ii sample shows a preference for the higher luminosity hosts . moving to lower luminosities
, the rotse - iiib sample shows a considerably higher fraction in dwarf galaxies . as explained in
[ survey_eff ] , our selection process for rotse - iiib may have a slight bias against high - luminosity hosts . in the next section
we calculate the volumetric rates of sne ia in dwarfs and giants for comparison against other samples . for convenience
, we will label galaxies fainter than @xmath81 as `` dwarfs , '' and more luminous galaxies as `` giants . '' when discussing the k - band magnitudes of the hosts , we will consider @xmath87 galaxies to be dwarfs , since the @xmath88 colors for many of our @xmath89 hosts are around 2mag .
if our sample s apparent preference of low luminosity hosts is a result of a selection effect , then our sne ia rate in giants should be lower than the actual rate while the rate in dwarfs should be accurate .
we compare our values to published rates as a check .
in this section we calculate the volumetric sne ia rate , @xmath90 , using : @xmath91 [ eqn : rate ] where @xmath92 is the observed number of events , @xmath93 is the _ proper _ time of the survey , @xmath94 is the co - moving volume surveyed , and @xmath95 is the efficiency factor that gives the estimated fraction of events actually discovered to the true number of events in the time and volume covered ( i.e. @xmath96 ) .
we have 18 sne ia appropriate to the augmented loss pamd and 16 appropriate for the sdss - ii model that were detected on 5 or more nights ( see [ lf ] and [ survey_eff ] ) .
we calculated the denominator of equation [ eqn : rate ] in [ survey_eff ] as a sum over a series of logarithmically spaced distance bins to account for the decline in survey efficiency with distance ( see figure [ fig : se ] ) .
the proper time for the survey is calculated by correcting the observer frame span for time dilation in each luminosity distance bin .
the survey period was set from november 1 , 2004 through january 31 , 2009 , with the former date reflecting the month the survey began and the latter set to the end of the standard 2008b astronomical semester .
the final term for the survey volume is calculated for each distance bin by integrating the co - moving volume element of the friedmann - robertson - walker metric with our chosen cosmological parameters ( a flat , @xmath11kms@xmath6mpc@xmath6 , @xmath12 universe ) and factoring in the fraction of sky covered by our survey .
special care must be taken to determine the actual survey area since the galaxy clusters of the survey are covered by an irregular grid of overlapping fields ( see figs .
[ fig : winter_spring ] and [ fig : summer_fall ] ) .
the typical pointing error of rotse - iiib is about @xmath97 degrees , so we assume the reference templates , which represent the intersection of numerous individual images , cover @xmath98 degrees each .
the actual pitch for the main search fields is 1.65 degrees , so this is the effective size for fields bordered on all sides . to determine the total survey area
, we construct an all sky register similar to the images in figures [ fig : winter_spring ] and [ fig : summer_fall ] , and then total all the sky area visited at least once during the survey period .
our total sky coverage is 499.0 square degrees . adopting the loss sne ia pseudo - absolute magnitude distribution with a supplemental population of high luminosity events ( assumed to be 1% of the total population ;
see [ losslf ] ) we find a rate of @xmath99sne iampc@xmath5yr@xmath6@xmath100 ( statistical error only ) . with the sdss - ii model pamd , the volumetric rate drops to the lower ( but statistically consistent ) value of @xmath101sne iampc@xmath5yr@xmath6@xmath100 .
these total rates and the rates specific to dwarf and giant hosts are shown in figure [ fig : rates ] and listed in table [ table : rates ] . lower limits for the complete rotse - iiib samples in each group .
the filled green circle shows the rotse - iiib rates computed assuming the loss sne ia pseudo - absolute magnitude distribution @xcite , and the open circles show the values derived using the same sne ia distribution assumed for the sdss - ii calculation @xcite and removing spectroscopically confirmed sne ia in a similar manner to the sdss - ii vetting process . ]
lcccccc loss & & @xmath102 & & @xmath103 & & @xmath104 + sdss - ii ( @xmath105 ) & 5 & @xmath106 & 74 & @xmath107 & 79 & @xmath108 + * rotse - iiib * ( w/ loss lf ) & 6 & @xmath109 & 12 & @xmath110 & 18 & @xmath111 + * rotse - iiib * ( w/ sdss - ii lf ) & 4 & @xmath112 & 12 & @xmath113 & 16 & @xmath114 + [ table : rates ] for comparison , the sdss - ii found a sne ia rate of @xmath115sne iampc@xmath5yr@xmath6@xmath100 using their @xmath105 sne ia and assuming the rate to be constant out to this limit @xcite . since the sne ia rate increases with redshift and even the lowest sdss - ii redshift bin extends out to larger distances than the rotse - iiib sample , the sdss - ii rate may be slightly inflated with respect to our more nearby rate measurement . in any case
, the sdss - ii rate is below our @xmath24 lower limit derived with the augmented loss pamd and may be incompatible with the sdss - ii model pamd as well at about the @xmath116 level .
we split the samples into dwarf and giant hosts and find that while the rotse - iiib rates in giants are marginally consistent with the sdss - ii results , our rate in dwarfs is significantly higher than the sdss - ii with either the augmented loss or sdss - ii model pamd . to convert the loss host luminosity specific sne ia rate into a volumetric rate
, we follow similar steps as reported in @xcite , but with an important difference .
we start with the rate - size relation and snuk measurement from loss and combine this with the local galaxy k - band luminosity function @xcite . like the loss estimate , the calculation
is done separately for early- and late - type galaxies , and then combined to give the final result .
the published loss rate is determined by multiplying the local galaxy k - band luminosity densities ( @xmath117 and @xmath118 ) by average values of their snuk , which are determined from a weighted average involving the number densities from the galaxy luminosity functions .
we choose to recalculate the loss rate through what we believe to be a more direct approach .
we simply multiply the galaxy luminosity function by the rate - size relation function and integrate . with our approach , we find the volumetric rate implied by the loss sample is @xmath119sne iampc@xmath5yr@xmath6@xmath100 .
this is consistent with , if not slightly lower than , the published loss rate @xcite .
we compute the statistical error through a monte carlo calculation including the uncertainty in the galaxy luminosity function , which gives a somewhat larger confidence interval than reported by loss .
this rate is consistent with the sdss - ii rate , which is an odd result given the apparently different sne ia populations studied ( see [ loss_v_sdss ] ) .
thus , as was the case with the sdss - ii rate , the total rates we derive from our rotse - iiib sample are higher at the @xmath24 level .
we also determine the loss rate in dwarfs and giants by simply changing the integration interval .
again , loss and sdss - ii derive rates that are consistent with each other for dwarfs , giants , and the total host population even though their sne ia populations are apparently distinct ( see further discussion in [ conclusions ] ) .
using either sne ia pseudo - absolute magnitude distribution , our rate in giants is comparable to sdss - ii and loss , so there is no evidence that our selection process is biased against high luminosity hosts .
however , with the loss distribution our rate in dwarf hosts is significantly higher than either sdss - ii or loss , and with the sdss - ii model pamd , our rate is higher at about the @xmath116 level .
we have used the background population of supernovae discovered by the rotse - iiib telescope over a four year period to derive the volumetric rate of sne ia .
our rate is somewhat sensitive to the sne ia pseudo - absolute magnitude distribution ( pamd ) assumed in deriving our overall search efficiency , and we find that recent rates studies have used different , incompatible pamds .
we have calculated our rates using both the loss pamd augmented with a small population of high luminosity events ( 1% of the total ) and the same pamd model ( and similar sample cuts ) employed in the sdss - ii rate measurement to quantify our distance dependent search efficiency . with the augmented loss pamd , we find a rate that is more than double published values in a similar redshift range .
this rate is higher than the sdss - ii or loss rates at @xmath120% confidence ( ignoring systematics for the moment ) .
similarly , performing the calculation on our `` normal - bright '' sne ia in a manner directly comparable to the sdss - ii measurement ( including the same pamd model ) , our rate is still nearly double that found by the sdss - ii out to a slightly higher redshift . in this case ,
our rate is only higher than the sdss - ii value at the @xmath116 confidence level , however .
the loss pamd is utterly incompatible with the sdss - ii model , which , in turn , is incompatible with the actual volume limited sample from the first year sdss - ii .
neither the sdss - ii model nor the loss pamd allow for events as bright as sn2007if , so it would seem that none of the available distributions account for the full population of sne ia .
additionally , the large extinction value reported for the photometrically selected sdss - ii sn09266 suggests that these distributions may also fail to account for the true distribution of host absorptions .
down to about -18.8mag , the loss and sdss - ii model pamds agree , but the loss sample shows almost double the relative contribution from fainter events .
the solid agreement between the loss and sdss - ii rates is perplexing given that they are apparently sampling completely different sne ia populations .
the systematic errors in our rate measurement are dominated by the uncertainty in the sne ia pamd , including how such distributions derived from ( mostly ) filtered photometry apply to our unfiltered data set .
our total rate estimate with the sdss - ii model pamd is 30% lower than our rate with the augmented loss pamd .
part of this is due to the exclusion of sne 2007if and 2008ab from the sdss - ii model rate due to their high luminosities ( both of these events also have dwarf hosts ) .
the larger fraction of low - luminosity events with the loss pamd makes up the additional difference .
our search is mainly sensitive to background supernovae with pseudo - absolute magnitudes brighter than about @xmath121mag , so at least 42% of sne ia are definitely missed by our search if the loss distribution is assumed , but this fraction falls to just 8% if we use the sdss - ii model .
we use our rate based on the sdss - ii model pamd to constrain the lower bound on the sne ia rate and our augmented loss rate for the upper bound . adding in a @xmath122mag systematic offset to account for the allowed uncertainty in our unfiltered magnitude system with respected to the ( filtered ) pamds considered , our final sne ia rate ( including @xmath123 statistical errors ) at a mean redshift of @xmath0 is then between 3.7 and 13.3 in units of @xmath124sne iampc@xmath5yr@xmath6@xmath100
. as our sample is spectroscopically complete , we do not include uncertainty for unclassified candidates .
this can be an important concern for other , non - spectroscopically complete studies .
for example , the sdss has reported 561 spectroscopically confirmed sne ia , but there are at least 1070 photometrically probable sne ia uncovered by the same survey , and presumably many more transient candidates in all including peculiar sne ia , which may not pass the photometric screening @xcite .
there is also the question of whether all objects technically classified as sne ia from their spectra should really be grouped in the same physical category .
for example , some have argued that some peculiar sne ia , such as sn2005hk , may not be thermonuclear supernovae ( cf . * ? ? ?
one event in our sample , sn2006ct , bears some resemblance to this subclass , although we note some differences as well in
[ sample_notes ] . removing
this one event would lower our total rates by about 6% . on the other hand , some events grouped with ccsne
may actually be sne ia in disguise .
in particular , some typeiin may be better suited in the sne ia camp @xcite . over our survey period
, we discovered three background typeiin supernovae ( sne2006db , 2006tf , and 2008am ) , but none of these have been linked to typeia explosions .
in using the augmented loss pamd , we have attempted to measure the total sne ia rate including hubble diagram outliers and other peculiar events .
this should be directly comparable to the rate measured by loss who include sn1991t - like , sn1991bg - like , and sn2002cx - like events in their sample , but this has not always been common practice in the literature .
many of the previous sne ia rate measurements have explicitly excluded events that deviate from the sub - sample of sne ia that are most useful as cosmological probes . for example , the recent sne ia rates measured by the snls @xcite are strictly valid only for sne ia with light curve widths close to the nominal value ( @xmath125 , but see also @xcite ) .
sne ia like sn2003fg ( a.k.a .
snls-03d3bb ) are not discussed .
similarly , the sdss - ii explicitly removed their peculiar sne ia , such as the interacting sn 2005gj , and the selection cut on the mlcs2k2 light - curve fits biases their sample against high luminosity , sn 1999aa - like events @xcite . because of this , the snls and sdss - ii rates only reflect the frequency for a fraction of the larger sne ia population .
as to the high luminosity sne ia excluded by snls and sdss - ii , however , we note that if these constituted more than a few percent of the total population then they would completely dominate our flux limited sample , which is not the case .
although the significance of our rates being higher than the canonical value is low , it is worth discussing how this may have come about .
one point of intrigue is our excess of sne ia in dwarf galaxies over what would be expected from the loss or sdss - ii studies .
if our rate for these hosts is correct , then either the sne ia rate size relation increases for the lowest luminosity hosts as compared to the range studied by loss , or the faint end of the galaxy luminosity function used to convert the loss luminosity specific rate to a volumetric rate may be under - counting the actual supply of dwarf galaxies .
such correction factors are not required for the sdss - ii rate , so the higher frequency of sne ia in low luminosity hosts is puzzling . in the full background sample , 7 out of 23 of the rotse - iiib hosts are fainter than @xmath126mag , as are 5 out of the 18 used for the rate calculation with the augmented loss pamd .
these fractions ( @xmath127% ) are significantly larger than the 1 out of 79 sne ia hosts in the sdss - ii @xmath105 sample that have such low luminosity hosts ( hosts brighter than @xmath128mag can be detected through our analysis out to this redshift limit , but only sdss - ii sn13038 is fainter ) .
if this difference is attributed to a selection bias against giant hosts in the rotse - iiib sample , then we must increase our rate to account for the missing sne ia in giants . but to achieve the same low to high - luminosity host ratio as the sdss - ii , our already high rates must be increased by factors of 4 to 5 !
another possibility to consider is that the sdss - ii may somehow be biased against low luminosity hosts , but this is difficult to imagine .
a remote possibility is that the use of host photo - z s in selecting targets for spectroscopic follow - up may boost the likelihood of classifying sne ia in giants rather than in dwarfs , which were unlikely to have been detected in the reference data available during the search . in that case
, there would still need to be a second bias to reject the sne ia in dwarfs since photometrically probable yet non - spectroscopically confirmed sne ia were included in the sdss - ii rates @xcite
. a possible contributor to the relative infrequency of dwarf hosts in the sdss - ii sample may be the cut on the mlcs2k2 light - curve fits used in selecting photometrically probably sne ia @xcite .
this cut can remove high luminosity sne ia , such as sn 1999aa , which tend to be associated with lower luminosity hosts @xcite .
however , only one of the five sne ia from our rate sample that fall in @xmath126mag hosts can be considered highly luminous : sn 2007if . since
the sdss - ii does contain a number of higher redshift sne ia whose hosts are not detected even in our deep coadds , it does not appear that the sdss - ii had a selection bias against either `` hostless '' sne ia or sne ia in dwarf hosts , and it is not obvious why such hosts are so much more frequent in the rotse - iiib sample .
another questions is why is there such a larger fraction of faint ( pseudo absolute magnitude @xmath129 ) sne ia in the loss sample as compared to the sdss - ii ? despite being targeted , the loss galaxy sample spans a range of masses and colors , and the full loss galaxy sample has absolute b and k - band magnitude distributions that are consistent with the sdss - ii sne ia hosts in the @xmath79 range ( see fig .
[ fig : hosts ] ) . applying the loss rate - size relation only to galaxies brighter than @xmath130 in the @xcite galaxy luminosity function and selecting only the 48% of sne ia fainter than @xmath131 in the loss pseudo - absolute magnitude distribution ,
the _ minimum _ rate of low - luminosity sne ia implied is @xmath132sne iampc@xmath5yr@xmath6@xmath100 . in this case
, the first year sdss - ii @xmath1 sample should contain at least 6 sne ia fainter than @xmath133mag .
even with the peculiar sn2005hk and the photometrically selected sdss - ii sn09266 ( with its heavy host absorption of @xmath134mag ) , there are still only 3 events in the sdss - ii @xmath1 first year sample at such low luminosities .
it is not clear how the sdss - ii could have missed the faint tail of the sne ia distribution ( e.g. sn 1991bg - like events around @xmath135 ) , yet we know of only two such spectroscopically confirmed events discovered in the three years of the sdss - ii ( 2007jh and 2007 mm ; @xcite ) and only one possible sn2002cx - like event ( sn2007ie ; @xcite ) .
additionally , the palomar transient factory @xcite has recently shown that there are perhaps even more varieties of peculiar , low - luminosity objects that may be spectroscopically classified as sne ia ( e.g. * ? ? ?
* ; * ? ? ?
* ) , but these are apparently absent from the sdss - ii as well .
the apparent lack of low - luminosity sne ia could mean that there was an unknown selection bias against such objects and that the total rate of sne ia is higher than found by the sdss - ii . the tension between the loss and sdss - ii pamds poses a potentially worrisome systematic bias for sne ia rate studies .
@xcite argue that the sdss - ii was sensitive to even sub - luminous sne ia below @xmath1 , so even if their model distribution were flawed , this would not have a significant impact on their rate calculation ( but again , the dearth of low - luminosity sne ia as compared to the expectations of loss is puzzling at best and may suggest the sdss - ii has altogether missed some sne ia ) .
however , if there is a low - luminosity population of sne ia that is not accounted for in the sdss - ii model distribution , this will lead to an overestimation of the search efficiency at the higher redshifts considered in @xcite , and thus an underestimation of the actual rate . in general , rate studies that push their samples out to the highest redshifts possible with their surveys have an even stronger malmquist bias and are thus more sensitive to systematic errors in their adopted pseudo - absolute magnitude distributions . these distributions
may also vary with redshift .
it is therefore important to securely determine the true sne ia pseudo - absolute magnitude distribution from complete , volume limited surveys as could potentially be done by the palomar transient factory at low redshifts and at moderate redshifts with the hyper - suprimecam on subaru .
another issue that may lead to bogus rate evolution measurements is the changing definition of the sne ia sub population measured by different surveys .
low to moderate redshift surveys often use light curve fits to select sne ia ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but the highest redshift studies often have far fewer epochs available due to the increased follow - up cost , and must therefore rely more heavily on color selection techniques @xcite .
it is not clear that these different cuts capture identical sub populations so that differences in the rates between different surveys may be taken directly as evidence for evolution in the sne ia rate .
this question may be resolved by performing a rolling search similar to sdss - ii but sensitive to very high redshift sne ia , as will be possible with the hyper - suprimecam survey .
rotse - iii has been supported by nasa grant nnx-08av63 g and nsf grant phy-0801007 .
the research of jcw is supported in part by nsf grant ast1109801 .
this material is based upon work supported in part by the national science foundation under grant no .
1066293 and the hospitality of the aspen center for physics .
this research made use of montage , funded by the national aeronautics and space administration s earth science technology office , computation technologies project , under cooperative agreement number ncc5 - 626 between nasa and the california institute of technology .
montage is maintained by the nasa / ipac infrared science archive . | we present a sample of 23 spectroscopically confirmed type ia supernovae that were discovered in the background of galaxy clusters targeted by rotse - iiib and use up to 18 of these to determine the local ( @xmath0 ) volumetric rate . since our survey is flux limited and thus biased against fainter objects , the pseudo - absolute magnitude distribution ( pamd ) of sne ia in a given volume is an important concern , especially the relative frequency of high to low - luminosity sne ia .
we find that the pamd derived from the volume limited lick observatory supernova search ( loss ) sample is incompatible with the distribution of sne ia in a volume limited ( @xmath1 ) sub sample of the sdss - ii .
the loss sample requires far more low - luminosity sne ia than the sdss - ii can accommodate .
even though loss and sdss - ii have sampled different sne ia populations , their volumetric rates are surprisingly similar . using the same model pamd adopted in the sdss - ii sne ia rate calculation and excluding two high - luminosity sne ia from our sample , we derive a rate that is marginally higher than previous low - redshift determinations . with our full sample and
the loss pamd our rate is more than double the canonical value .
we also find that 5 of our 18 sne ia are hosted by very low - luminosity ( @xmath2 ) galaxies , whereas only 1 out 79 nearby sdss - ii sne ia have such faint hosts .
it is possible that previous works have under - counted either low luminosity sne ia , sne ia in low luminosity hosts , or peculiar sne ia ( sometimes explicitly ) , and the total sne ia rate may be higher than the canonical value . |
at optical wavelengths the spectral energy distribution ( sed ) of elliptical galaxies falls precipitously shortward of the 4000 break , hence the discovery of the uv - rising branch came as one of the most unexpected results of the first uv satellites .
the general prejudice was that ellipticals contained exclusively old and cool stellar populations , similar to metal rich galactic globulars , albeit even more metal rich .
however , a few bright very hot stars were known to be globular cluster members , and the presence of similar objects in ellipticals had been suggested before the discovery of their uv rising branch ( minkowski & osterbrock 1959 ; hills 1971 ) , but this knowledge was not widely spread .
the first uv spectra of ellipticals and bulges of spirals showed instead that shortward of @xmath0 2300 the flux was increasing with decreasing wavelenth . a fundamental contribution to this subject
is due to burstein et al .
( 1988 ) , who collected and organized all the relevant information from iue observations .
the three main results of this study were : 1 ) all studied ellipticals have detectable uv flux , 2 ) their ( 1550v ) color spans a range of @xmath1 2.5 mag , and 3 ) it is strongly correlated with the 2 index .
hence , the ratio of the uv to the optical emission varies by @xmath0 an order of magnitude , and this ratio appears to increase with average metallicity ( @xmath2 ) , assuming that the 2 index traces @xmath2 . the presence of young , massive ( hence hot ) stars in ellipticals was soon entertained in order to account for the observed uv radiation . on the other hand
, low mass stars do evolve through hot evolutionary phases at the end of their life , and some uv radiation _ is _ naturally expected to arise also from purely old stellar populations .
greggio & renzini ( 1990 , hereinafter gr90 ) explored a variety of possible candidates produced by the evolution of low mass stars , both single and in binary systems .
in particular , in gr90 we concentrated on the possibility of producing hot stars in @xmath3 10 gyr old stellar systems , which could account for the observed level of the uv - to - optical flux ratio , and ( qualitatively ) of the correlation with metallicity .
we used simple energetic arguments , based on the fuel consumption theorem ( renzini and buzzoni 1986 ) , to translate the observed level of the uv rising branch into specific requirements for the candidate stars responsible for the uv emission . in this paper
we first summarize the main results of gr90 , and then review and discuss both the observational and the theoretical developments following 1990 .
the argument in gr90 goes as follows .
the ultraviolet sed as measured from iue is consistent with the rayleigh - jeans tail of a black body curve of temperature higher than @xmath025000 k. in order to estimate the ratio of uv to total flux an assumption on the typical temperature of the hot component is necessary .
for example , for ngc 4649 , one of the most powerful ellipticals , @xmath4 is in the range from 0.014 to 0.021 , for @xmath5 k. for single age and single metallicity populations of single stars ( simple stellar populations , ssps ) , the contribution to the total bolometric light of stars in the generic @xmath6-th post - ms evolutionary phase is @xmath7 where @xmath8 is the specific evolutionary flux , i.e. the number of stars evolving through the turnoff and beyond per year per solar luminosity of the parent population ( in units of @xmath9 ) , and @xmath10 is the amount of fuel burned during the phase @xmath6 by stars of initial mass equal to the turn - off mass ( ) at the age @xmath11 of the population . the fuel @xmath10 is expressed in of equivalent hydrogen , i.e. @xmath12 , where @xmath13 and @xmath14 are respectively the mass of hydrogen and helium burned during the phase @xmath6 . for old ssps @xmath15 stars
@xmath16 yr@xmath17 , almost independent of composition and age ( cf .
fig . 1 in renzini 1998 ) .
going one step further , equation ( 1 ) can be generalized to a collection of ssps ( composite stellar population , csp ) , e.g. one exhibiting a narrow range of ages but a wide metallicity distribution , that gr90 adopted as a fair description of the stellar content of ellipticals . to this end , suffice to substitute @xmath18 with the fuel averaged over the metallicity distribution @xmath19 .
thus , for old stellar populations with a metallicity distribution the following simple relation holds : @xmath20 this equation was the main tool used in gr90 to evaluate various kinds of stars as potential contributors to the uv rising branch in ellipticals
. from the observational requirement @xmath21 , equation ( 2 ) immediately indicates that the hot stars responsible for the uv emission from giant elliptical galaxies should burn at least @xmath0 0.01 @xmath22 of equivalent hydrogen . as already mentioned , the range in @xmath23 colors spanned by ellipticals is consistent with @xmath4 varying by one order of magnitude .
accordingly , the fuel burned by the candidate hot stars , averaged over the metallicity distribution , should increase from @xmath24 0.001 to @xmath24 0.01 when the average metallicity of the csp increases .
gr90 listed four candidates , which are naturally produced in the advanced stages of the evolution of single stars : \(i ) post agb stars ( p agb ) , i.e. stars which leave the agb after the first thermal pulse , and reach @xmath25 k before approaching the white dwarf ( wd ) cooling sequence .
this is certainly the most common channel to reach the finale wd stage .
typical luminosity @xmath26 .
\(ii ) post early agb stars ( p eagb ) , i.e. stars leaving the agb before the first thermal pulse , as most of their hydrogen envelope is lost before .
typical luminosity @xmath27 .
\(iii ) hot hb stars ( hhb ) , sometimes also called extreme hb ( ehb ) stars , i.e. stars which spend the core helium burning phase at high temperatures , and whose subsequent evolution ( shell helium burning phase ) also takes place at high temperature ( agb manqu )
. typical luminosity @xmath28 for hhb , few @xmath29 for agb manqu .
\(iv ) post rgb stars ( p rgb ) , i.e. stars which fail helium ignition because they loose their envelope while climbing on the rgb .
typical luminosity @xmath27 .
the first three channels eventually produce carbon - oxygen wds , the last one helium wds .
1 shows schematically the evolutionary paths corresponding to channels ( i ) , ( ii ) and ( iii ) .
also shown are the limiting magnitudes for objects in m31 reached with 1.5 h exposures with wfpc2 in two wood s filters .
it appears evident how difficult it is to detect individual hhb stars at this distance .
a stellar population of given metallicity will certainly produce stars evolving through channels ii ) , iii ) and iv ) provided it becomes sufficiently old .
however , the age at which this happens can not be accurately predicted . a model star of given initial
mass will evolve through one of the four channels above depending on the wind mass loss rate efficiency ( @xmath30 ) .
for @xmath30 below a critical value it will go through the p
agb , and for larger and larger values of @xmath30 it will switch to the p
eagb , hhb+agb - manq , and finally to the p rgb track .
as illustrated in gr90 , this whole range of possibilities is realized by varying the mass loss rate parameter @xmath30 by just @xmath0 a factor of two , i.e. by an amount vastly smaller than any observational uncertainty affecting empirical rgb and agb mass loss rates .
this leaves ample freedom to theoreticians .
all available evolutionary calculations indicated ( and still indicate ) that p agb stars burn less than @xmath31 , and p
rgb objects even less .
this allowed gr90 to conclude that stars in these evolutionary phases could play only a minor role in the production of the uv upturn .
more promising appeared the p - eagb , with @xmath32 up to @xmath33 , and the hhb and their agb manqu progeny , burning in total @xmath0 0.5 of helium ( equivalent to @xmath0 0.05 of hydrogen ) .
if all stars were to go through the hhb+agb manq channel @xmath0 5 times more uv radiation would be produced than needed to account for the @xmath34 of the total luminosity emitted in the uv , as in the @xmath23 bluest galaxies .
thus , a relatively small fraction ( @xmath1 20 @xmath35 ) of the stellar population in ellipticals needed to evolve through channel ( iii ) in order to fit the observations .
the trend of the ( 1550v ) color increasing with @xmath2 could then be understood if a larger fraction of the population was evolving through channel ( iii ) at higher average metallicity .
such a trend could be accomplished in either of two ways : 1 ) with a modest increase with metallicity of the mass loss rate parameter @xmath30 , or 2 ) with the helium abundance ( @xmath36 ) increasing with metallicity ( @xmath2 ) .
indeed , at fixed age and metallicity , a larger @xmath36 corresponds to a smaller envelope mass for the star evolving along the rgb , so that it is easier to produce objects ( ii ) to ( iv ) .
moreover , higher helium _ per se _ favors higher effective temperatures , e.g. during the hb phase ( cf .
sweigart & gross 1976 ) .
in essence , which of the 4 channels is realized depends on how the mass loss rate and @xmath36 scale with @xmath2 , both parameters @xmath37 and being poorly known observationally . at the same time , hosting stellar populations with a metallicity spread , ellipticals should be inhabited by all four kind of objects , though in different proportions .
the main conclusions in gr90 can be summarized as follows : \(i ) p agb stars , these hot low mass objects necessarily present in ellipticals do not provide enough uv flux to account for the level of the uv rising branch in the most powerful galaxies .
\(ii ) the uv upturn in old stellar populations could be accounted for only by the presence of p eagb and hhb stars , with up to @xmath38 of all evolving stars venturing through these channels .
\(iii ) the production of these stars at high @xmath2 , as seemed implied by the @xmath23 2 correlation , was possible within the uncertainties affecting the empirical determination of @xmath37 and , which are inputs to stellar evolution theory .
\(iv ) whatever mechanism is responsible for the production of these stars ( i.e. mass loss or ) , one expects that all possibile candidates are present in ellipticals , though in different proportions .
\(v ) if the hot stars responsible for the uv emission in the most powerful ellipticals were p eagb and hhb objects , the uv rising branch should fade away rapidly with increasing redshift , possibly already disappearing at redshifts as low as @xmath39 .
this was a direct consequence of the sensitivity of the effective temperature of hb stars on the envelope mass at helium ignition .
gr90 paid less attention at the temperature distribution of the stars evolving trough the various channels , but pointed out the overwhelming difficulty of predicting such distribution , that from the theoretical point of view depends on a number of arbitrary functions .
we felt indeed that detailed spectral synthesis modelling was not worth the effort , while the only possible firm conclusions could be reached with very simple arguments .
however , it was mentioned that the @xmath40 dip in the sed requires a gap in the temperature distribution between hhb stars and the remaining , cool hb stars .
no explanation for the existence of this gap was given in gr90 . moreover ,
if a continuity exists between hhb and p rgb stars , then one may expect hhb stars to extend all the way to the helium main sequence , hence to fairly high effective temperatures ( @xmath41 k ) .
in other words , this scenario would most naturally produce a fairly wide temperature distribution of uv emitters .
it was also speculated that a sizable fraction of hhb and their agb manqu progeny could be helium stars , suffice indeed fairly modest mass loss rates ( @xmath42 ) for these stars to lose completely their hydrogen envelope .
this possibility was predicted to be subject to observational test , as in this case the uv upturn should have exhibited some wolf - rayet , wn - like features , such as low or absent civ and lyman lines , and strong heii and nv .
in 1990 much of the available information on the uv upturn came from iue , and it has been organized by burstein et al .
later , most of the observational novelties came from uit , hut , and hst .
direct uv imaging became possible , as well as spectroscopy down to the lyman continuum .
uit and hst imaging have definitely ruled out massive stars as the origin of the uv upturn in m32 , the bulge of m31 , as well as in ngc 1399 , one of the uv most powerful ellitpicals ( oconnell et al .
1992 ; king et al . 1992 ; bertola et al .
1995a ; cole et al .
1998 ) .
extremely blue , low mass stars have been directly imaged in the bulge of m31 and in m32 by hst ( king et al . 1992 ; bertola et al .
1995a ; brown et al . 1998 ) .
although none of these objects is a massive elliptical , they follow the @xmath23 @xmath43 2 correlation which characterizes all _ quiescent _ es . the king et al . and bertola et al .
data were taken with the pre - costar foc .
the first group obtained images of a central field in m31 through a filter centered at @xmath44 = 1750 , resolving more than 100 objects . based on their measure of the uv magnitudes , and on an upper limit to their @xmath45 color
, the authours concluded that the resolved stars in the f175w images are p agb stars ; and , by comparing with the iue flux from the same area , that these stars account for only a fraction ( @xmath46 ) of the total flux at 1750 .
bertola et al .
( 1995a ) obtained images of m31 , m32 and ngc 205 through the combined uv filters f150w and f130lp , resolving 81 , 10 and 78 stars in the three objects respectively .
the point like sources in ngc 205 were interpreted as young ob stars ( as already known from ground - based observations ) , while the luminosity of the sources in m31 and m32 suggests that these are p agb stars . by comparing with iue data ,
the authors conclude that the resolved p
agb objects can account for the total uv flux in the case of m32 , while for m31 @xmath47 of the uv flux comes from an unresolved background .
therefore , both groups conclude that the uv light in the bulge of m31 likely comes from the combination of p agb stars and fainter objects , which appear as an unresolved diffuse background on the hst image .
the different value derived by the two groups for the contribution of the resolved sources results from the different assumptions on the sensitivity calibration of the pre - costar foc , and on the uncertainties on the red leak through the uv filters . at any rate , the conclusions from both groups confirmed the prediction of gr90 , i.e. the population of hot stars in old stellar systems is composite , with contribution from p agb stars bright enough to be individually detected in nearby spheroids , and fainter sources such as p eagb and hhb+agb manqu stars as faint as @xmath28 ( hence below detection threshold ) .
concerning m32 , the low level of its uv upturn is in agreement with the notion that the uv sources in this galaxy should just be p
agb stars , with very few stars if any going through the ( ii)-(iii ) channels .
uv color gradients have also been detected in a few objects ( o connell et al .
1992 ) . with the exception of m32 ,
uv colors become redder with increasing radius , probably tracing the 2 gradients .
the uv light appears diffuse , but more concentrated than the visual light , in agreement with the expectation that it is produced by the higher @xmath2 stars , preferentially found in the central regions ( see also brown et al .
1998 ) .
post - costar foc photometry of m32 and m31 has been recently obtained by brown et al .
( 1998 ) in two uv filters , namely f275w and f175w . again
, many point like sources are resolved in these images : 433 stars in m31 and 138 in m32 down to @xmath48 mag and @xmath49 mag .
brown et al . ( 1998 ) show that the pre - costar foc calibrations were likely in severe error , basically leading to an overestimate of the intrinsic uv flux from the sources . as a result , the resolved stars in brown et al .
are interpreted as agb
manqu objects , the bright progeny of hhb stars .
again , the cumulative flux from the resolved stars accounts for only a fraction ( @xmath50 ) of the total iue flux .
although still affected by some uncertainty , in particular a possible systematic underestimate of the flux in the f275w filter at the 0.3 mag level , the photometry by brown et al .
( 1998 ) is in reasonable agreement with the expectations from iue and hut spectra .
the interpretation of the nature of the resolved sources in brown et al .
( 1998 ) rests essentially upon the characteristic of the luminosity functions ( lf ) in the two uv filters .
there appears to be an increasing number of objects towards fainter magnitudes , a trend which is not present in the p
agb tracks of vassiliadis & wood ( 1994 ) to which the empirical lf was compared .
these tracks peak instead at magnitudes for which there are virtually no stars observed at all .
brown et al .
conclude that the bulk ( @xmath51 ) of all stars do indeed go through the p
agb channel , but the mass of the p agb stars is in excess of @xmath52 , for which the p
agb timescale is so short to be consistent with the observed lf .
however , such a high value of the p agb mass would imply the existence of a prominent and very bright population of agb stars , for which there is no evidence in the bulge of m31 ( renzini 1998 ) .
moreover , this population would produce an enourmous amount of energy ( @xmath53 of fuel would be burned on the agb ) , hence leading to optical
infrared colors at variance with the observed ones . in our opinion , the brown et al .
lf demonstrates that the vassiliadis & wood p agb tracks are inapplicable to the case of the m31 bulge .
these tracks are based on the assumption that the transition from the agb to the planetary nebula stage takes place on a nuclear time scale , being controlled by the burning of the residual envelope mass .
for the low values of the p agb mass expected in an old stellar population ( @xmath54 ) this transition time is indeed very long ( @xmath55 yr ) , and a sizable number of hot p agb stars would have been observed , lying along the nearly horizontal track in the upper part of fig .
however , one knows from galactic globular clusters that the transition is instead much faster , taking place either on a mass - loss time scale , or even more probably on a thermal time scale ( kaeufl , renzini , & stanghellini 1993 , and references therein ) .
we conclude that the observed lf is likely due to the combination of the low - mass p
agb channel ( @xmath51 of the total stellar evolutionary flux ) , with the transition to high temperatures taking place on a thermal time scale , plus p
eagb and/or the hhb+agb manqu objects for the residual stellar evolutioanry flux ( @xmath56 of the total ) .
an apparently puzzling result of the brown et al .
study is that the lfs of the uv stars in m31 and m32 look similar in shape , in spite of the strong difference in the level of the uv upturn in these two galaxies . according to brown et al .
the fraction of the total evolutionary flux that has to go through the non p agb channel is 2 @xmath35 and 0.5 @xmath35 respectively in the bulge of m31 and in m32 .
hence , the similarity of the two lfs comes from both being dominated by p
agb stars that do _ not _ evolve on a nuclear time scale through their transition from the agb to high temperatures . in 1993
foc imaging in four uv bands was obtained of the central regions of the ellipticals ngc 1399 and ngc 4552 and of the bulge of the ngc 2681 spiral ( pi f. bertola ) .
the aim was to study the spacial structure of the uv emission , checking for color gradients and if any patcheness existed due to star formation .
no such patcheness was found neither strong color gradients , but instead ngc 4552 and ngc 2681 showed a central , unresolved , point - like source ( bertola et al .
to our surprize , we found that the point - like source in ngc 4552 had changed its brightness by a factor @xmath57 in the f342w band , compared to a previous foc image taken in 1991 : a central _ flare _ had been discovered , possibly due to a red giant having been tidally stripped by a massive central black hole ( renzini et al .
subsequent , post - costar foc imaging and fos spectroscopy confirmed that the central source is a variable mini - agn , possibly the faintest known agn , with broad ( fwhm@xmath58 km s@xmath17 ) emission lines ( cappellari et al .
1998 ) . while trying to better understand the uv upturn , we had serendipitously found yet another way of gathering information on the central black hole demography in galaxies . extending the observed spectral range down to the lyman limit hut
has detected the maximum in the uv spectral energy distribution ( ferguson et al .
1991 ) . to date ,
hut data for 8 early type objects , including the bulge of m31 , have been collected ( ferguson @xmath59 davidsen 1993 ; brown , ferguson @xmath59 davidsen 1995 ; brown et al .
1997 ) . in all studied objects
, the uv rising branch appears to have a turn - over at @xmath60 1000 , which indicates that the bulk of the radiation comes from moderately hot stars , with temperatures in the range 2000025000 k ( brown et al .
1995 ) . assuming that this is the characteristic spectral energy distribution in the uv for giant ellipticals , like ngc 4649
, we obtain a better estimate for the ratio @xmath4 of @xmath61 , which translates into @xmath62 for the hot stars inhabiting the most powerful ellipticals .
since the uv sed has a minimum around 2300 , a large contribution from stars with intermediate temperatures , say @xmath1 10000 k , is excluded .
thus the bulk of the uv emission comes from stars in a narrow range of temperatures .
this is an important constraint for the astrophysically plausible evolutionary paths that can account for the uv rising branch phenomenon .
for example , an even distribution of stars on the hb like in the globular cluster m3 , corresponds to a spectrum flatter than observed in ellipticals , due to the similar contribution from stars in the wide effective temperature range ( see e.g. nesci @xmath59 perola 1985 , ferguson 1995 ) .
another important characteristic of the hut spectra of early type systems is the fact that they are composite : when removing from the observed spectrum the theoretical contribution of hhb star , according to their complete evolution from the zahb to the wd final stage , some residual flux at the shorter wavelenghts is left ( ferguson and davidsen 1993 ; brown et al .
the best fits to the sed of all the studied objects are obtained with contributions from both hhb and p agb evolutionary tracks . based on their detailed modelling , brown et al .
( 1997 ) conclude that approximately 10 @xmath35 of the total stellar population should go through the hhb channel of evolution in ngc 1399 , one of the strongest uv emitters .
this is in very good agreement with the predictions of the fuel consumption theorem : for a two components csp , with 90@xmath35 of the stars evolving through the p
agb channel and the remaining 10@xmath35 going through the hhb evolution , the average fuel burned in the hot evolutionary phases is : @xmath63 adopting @xmath64 = 0.003 and @xmath65 = 0.05 ( see sect .
2 ) one gets @xmath66 = 0.0077 , close indeed to the estimate above .
all of the 8 objects in the brown et al
. sample seem to require some contribution from hhb stars , in different proportions .
this does not come unexpected , since their 2 indices range from 0.31 to 0.36 , which puts them among the high average metallicity objects .
brown et al . (
1995 ) also claimed that , within their sample , a larger fraction of stars evolve through the hhb channel in the stronger 2 galaxies . modelling the uv spectra
, they derive the stellar evolutionary flux through the hhb plus agb manqu track which is needed to account for the observed uv emission .
the ratio beteween this evolutionary flux and that of the total stellar population sampled by the hut aperture appears to be nicely correlated with the 2 index of the parent galaxy .
the value of this ratio is model dependent , and somewhat different figures are obtained in the more detailed computations in brown et al . (
1997 ) . nevertheless , judging from their tables , the general trend seems confirmed .
thus , it can be concluded that galaxies with 2 indices in excess of @xmath0 0.3 very likely host hhb stars in their nuclei .
only a small fraction of the population need to go through this extreme evolutionary channel to account for the observed uv fluxes , varying from @xmath67 to @xmath68 for galaxies with 2 ranging from 0.3 to 0.36 .
these results are clearly consistent with the expectations from gr90 .
the low s /
n in the hut spectra prevents accurate determinations of abundances .
the absorption features seem however to indicate a low metallicity in the stellar atmospheres of the stars mainly contributing the uv emission : @xmath69 ( brown et al 1997 ) .
this would imply that the uv rising branch phenomenon is not directly related to the presence of high @xmath2 stars , and the correlation between the @xmath23 color and 2 index has to find a different explanation ( park & lee 1997 , see below ) from what proposed in gr90 .
alternatively , @xmath70 is not representative of the true metallicity of the hot stars in ellipticals , as heavy elements may diffuse out of the atmospheres of hhb stars ( brown et al .
1997 ) . a more firm result of the abundance analysis is the lack of civ as would be expected if massive stars were responsible for the uv emission .
however , hut spectra show also strong ly@xmath71 and ly@xmath72 lines , indicating that the vast majority of hhb and agb manqu do _ not _ lose their hydrogen envelope , and do not become wn - like helium stars .
this means that average mass loss rates during these phases must be lower than @xmath73 and @xmath74 , respectively .
one crucial prediction of gr90 concerned the evolution with redshift of the uv upturn . if due to a combination of p - eagb and hhb+agb manqu stars , the uv upturn should fade away already at fairly low redshift , see for example the realization by barbaro , bertola , & burstein ( 1992 ) . to check this prediction two cycle - i
hst projects were implemented ( pis r. windhorst and a. renzini , respectively ) .
fos spectra of @xmath75 elliptical galaxies selected for being either weak radiogalaxies and/or cluster members were obtained .
they all showed a strong uv upturn , which at first sight appeared to be in clear conflict with the prediction .
however , it soon turned out that a similarly strong uv upturn was also shown by the fos spectrum of an innocent g2v star , which certainly did not have it on its own . while searching for the vanishing uv upturn of ellipticals the red - scatterd light problem of fos
was discovered instead ( windhorst et al . 1994 ) .
this lead to a novel approach to the calibration of fos and lately of eso instruments for the vlt which makes more use of first physical principles , and less recourse to least square fits ( bushouse , rosa , & mller 1995 ; rosa 1997 ; ballester & rosa 1997 ) .
the fos scattered light problem had the effect of reducing dramatically the s / n ratio for uv observations of high-@xmath76 ellipticals , and therefore attempts at detecting the vanishing upturn effect moved to wfpc2 , now equipped with wood s filters .
a first attempt was made by a group including r. gilmozzi , e. held , r. viezzer and ourselves .
wfpc2 images of the cluster abell 895 ( @xmath77 ) were obtained through the wood s filter f218w , and through the f555w and f814w filters .
no detectable flux from cluster ellipticals was found in a coadded 10,000 second integration through the f218w filter .
the result is shown in fig . 2 for the brightest cluster member ( reproduced from renzini 1996 ) , with the 3-@xmath78 upper limit falling disappointingly on top of the expected upturn if such galaxies had the same rest frame @xmath23 color of the local elliptical ngc 4649 .
similarly disappointing was the result of an analogous experiment by buson et al .
( 1998 ) , who imaged the abell 851 cluster ( @xmath79 through the f218w and f702w filters ( corresponding to @xmath0 the rest frame @xmath23 color ) . again , the f218w data are not deep enough to detect the cluster ellipticals even if they were to maintain the same rest frame @xmath23 color of the bluest ellipticals at zero redshift .
the failures of these attempts has to be ascribed to the low sensitivity of wfpc2 when used in conjunction with wood s filters ( that indeed we nicknamed wood s filters ) .
an alternative approach has been recently pursued by brown et al .
( 1998 ) for a sample of ellipticals in the @xmath80 cluster a370 .
the combination of two long - pass filters of foc ( f130lp and f370lp ) has allowed them to isolate the contribution of the emission shortward of @xmath0 2700 in the rest frame , hence sampling the uv upturn .
surprizingly , no appreciable evolution compared to nearby ellipticals has been detected , and brown et al .
conclude that this result excludes some models of the upturn , while others are still acceptable provided that the bulk of stars in these galaxies formed at @xmath81 .
more observations are needed to study in detail the evolution of the uv rising branch with increasing redshift , and to derive informations on the nature and the age of the uv bright stars .
since fos and ghrs have been removed from hst , stis may now offer a better chance to detect the vanishing uv upturn effect .
in 1990 only a handful of p agb , p eagb , and hhb+agb manqu evolutionary sequences existed in the literature . in the last decade a large effort has been devoted to construct extensive sets of evolutionary tracks , primarily with the aim of understanding the uv upturn phenomenon .
hundreds of stellar evolutionary sequences for low mass stars , with up to super solar metallicities and helium abundances have been computed to isolate the range of parameteres which produce p eagb and hhb objects ( e.g. castellani & tornamb 1991 ; horch , demarque & pinsonneault 1992 ; castellani , limongi & tornamb 1992 , 1995 ; dorman , rood and o connell 1993 , hereinafter dro93 ; fagotto et al .
1994a , b , c ; yi , demarque & kim 1997a ) . basically , the overall evolutionary picture illustrated in gr90 has been confirmed . the average temperature at which the helium burning occurs is essentially controlled by the envelope mass of the stars at helium ignition , being hotter the lower the envelope mass .
high values of the helium abundance favor the production of hot helium burners , and widen the range of envelope masses for which this condition is satisfied . according to dro93 , for @xmath2= stellar models with hb envelope masses 0 @xmath82
0.05 evolve either as p eagb or as hhb and agb manqu .
this critical value for the envelope mass increases with the helium abundance ( see also yi et al .
1997a ) , reaching values as high as 0.15 for @xmath83 = ( 0.45,0.06 ) .
it follows that at high @xmath83 the condition on the envelope mass necessary to produce hot stars is more easily met .
two interesting aspects of the evolution of low mass stars have been disclosed , which were not considered in gr90 : 1 ) for high @xmath2 and especially @xmath36 , the evolution of hb stars presents a pronounced dichotomy , with some stars starting the evolution on the red side of the hb , spending there a fraction of their hb phase , and then after zipping to high temperatures where they burn the rest of their fuel ( horch et al .
1992 , but see also sweigart & gross 1976 ) ; and 2 ) for high assumed mass loss rates some stars _ peel off _ the rgb and experience their core helium flash at high effective temperatures ( castellani & castellani 1993 ; dcruz et al
. 1996 ) .
the systematics with @xmath36 and @xmath2 of the post rgb evolution can be appreciated in dro93 and yi et al .
( 1997a ) : up to @xmath1 the dependence of the zahb on 0 is relatively mild , and a flat distribution of envelope masses maps into an even distribution in log @xmath84 of stars on the hb .
subsequent evolution remains confined in the red ( in the blue ) for the more massive ( less massive ) hb objects , while the intermediate hb ( ihb ) stars evolve along wide redward / blueward loops , thereby providing intermediate temperature objects .
correspondingly , for @xmath85 it is relatively difficult to produce the 2300 minimum in the sed .
however , as the metallicity increases , the evolution of the hb objects tends to become more skewed either towards the red , or to the blue .
the effect is very strong for large values of the parameter . at ( @xmath86)= ( 0.46,0.06 )
the ihb objects virtually disappear , and the bulk of stars are either redder than log @xmath87 3.7 or bluer than log @xmath87 4.2 ( see figures 2 and 3 in dro93 ) .
this behavior may help producing the 2300 minimum in the sed . as illustrated in the introduction , when the mass loss parameter @xmath30 is sufficiently large , the evolution on the rgb is aborted before the core mass has grown enough to trigger the central helium flash . in gr90
it was assumed that in this case further evolution would just take the model star to the ( helium ) white dwarf stage .
actual computations in this @xmath30 range show , instead , that there are models which succeed in igniting helium after departing from the rgb , either while crossing the hrd , or during the subsequent cooling phase towards the wd stage ( castellani & castellani 1993 ) .
thus , there is a mass range ( or a @xmath30 range ) for which the helium core flash occurs in the hot region of the hrd ( hot helium flashers , in dcruz et al .
1996 nomenclature ) .
subsequent evolution of these objects ( hereinafter hhef ) is the same as for hhb star , with a very low envelope mass .
if they exist , the hhef have the minimum envelope mass that hb stars can have , hence naturally defining the hot end of the horizontal branch
. stars with more massive envelopes will ignite helium at the tip of the rgb , to appear on the zahb with lower .
stars less massive than the hhef will fail helium ignition , thus becoming helium wds . at super solar metallicites ,
hhef are produced for @xmath88 , which is @xmath0 2 times larger than the value which fits the properties of the hb in globular clusters ( dcruz et al .
after helium ignition , the hhef are found on the zahb at log @xmath84 @xmath0 4.4 for supersolar @xmath2 , which seems to be the maximum possible temperature for hb models ( castellani , deglinnocenti and pulone 1995 ) . inspired by the @xmath23 2 correlation
most authors have explored under which conditions hhb and related stars are produced in metal rich and super metal rich populations .
dorman , oconnell and rood ( 1995 , hereinafter dor95 ) assume the mass loss on the rgb as principal actor in originating the uv rising branch .
two are the main points which support this picture : ( 1 ) the presence of extended hb in the cmd of globular clusters , which require a spread in rgb mass loss by @xmath0 3 0@xmath35 among stars within the same cluster ; and ( 2 ) the population of hot subdwarfs in the solar vicinity , which shows that at @xmath0 solar metallicity hhb stars and their progeny can occasionally be produced .
thus , in dor95 view , the hot stars in ellipticals are ( moderately ) old , @xmath0 solar metallicity objects which happen to loose 2 - 3 times more mass than the average reimers rate . questioning the real significance of the 2 index as a metallicity indicator
, dor95 generically ascribe the origin of the correlation @xmath23 2 to either an age or a metallicity spread among ellipticals .
however , no attempt is made to eplore the effect of a metallicity distribution on the uv sed of ellipticals .
if a large dispersion of the mass loss rate applies to all @xmath2 components , it seems difficult to avoid a sizeable contribution of ihb stars in the uv spectral range . in tantalo
1996 ) models hhb stars are produced at high @xmath2 basically because a large is assumed .
these authors construct self consistent chemo - spectro - photometric models for the ellipticals , which thus contain a metallicity distribution as computed from the chemical evolution .
the final integrated properties of the models depend not only on the assumptions on the parameters governing the stellar evolution ( e.g. the mass loss ) , but also on those important for the chemical evolution ( i.e. star formation rate , imf , stellar yield , depth of the galactic potential , supernova feedback , galactic winds etc . ) .
tantalo et al .
most massive galaxy models present a strong uv upturn developing as early as 5.6 gyr .
this value of the age is extremely sensitive to the specific choice of the parameters @xmath30 = 0.45 and = 2.5 , which cause the ssp model at @xmath89 to produce hhb stars already at 5.6 gyr . on the other hand , all other ssp models in tantalo et al .
grid ( with @xmath90 ) produce hhb stars only for ages in excess of @xmath0 12 gyr .
therefore , the uv properties of the composite model also depend critically on the precise population of the highest metallicity bin ( see also yi , demarque & oemler 1998 ) . finally , yi , demarque & oemler ( 1997b,1998 ) propose a model in which all the relavant parameters are allowed to vary while searching for a best fit .
they finally favour a positive ( but moderate ) ( = 2 - 3 ) , a modest trend of @xmath30 with metallicity ( ranging from @xmath0 0.5 to @xmath0 0.71 for @xmath2 ranging from 0.02 to @xmath3 ) ; a mass dispersion of the hb of @xmath0 0.06 ( calibrated on gcs properties ) ; and a metallicity distribution , as suggested by chemical evolution models . with these prescriptions ,
yi et al . (
1998 ) reach a reasonable fit with the observations , at ages @xmath3 10 gyr .
the fit is better when adopting the @xmath2 distribution from infall models , as opposed to closed box models .
indeed , in the latter case too much flux is produced in the mid - uv , due to the broad distribution of hb temperatures of metal poor stars .
while in all the above attempts the uv emission arises from stars in the high-@xmath2 tail of the metallicity distribution , lee ( 1994 ) and park and lee ( 1997 ) maintain that the uv flux originates from the emission of metal poor stars . considering that the stellar populations in galaxies are characterized by a metallicity distribution
, they explore the possibility that the optical light comes from the high @xmath2 and the uv light come from the low @xmath2 components .
the relatively low strength of absorption features in the uv - rising branch mentioned in section 3.2 is in agreement with this picture .
the trend of increasing @xmath91 color with increasing 2 ( hence @xmath92 ) would have an indirect origin , resulting from the brighter ellipticals ( with stronger 2 indices ) being older than the fainter ones , as it may be expected in some cosmological simulations .
however , in the park and lee model , ages as old as of @xmath020 gyr are needed to produce a uv output such as that of giant ellipticals , which looks uncomfortably too old .
another problem comes from the strongest uv upturns being shown by galaxies with 2@xmath3 0.3 , which can not have a major metal poor component , as the metallicity distribution shoud be trimmed below z@xmath1 0.5 z@xmath93 ( greggio 1997 ) .
thus the low @xmath2 tail may not be present in the central regions of the most powerful ellipticals ( see also chiosi , vallenari & bressan 1997 ; yi et al .
moreover , as already noticed , it is more difficult to low-@xmath2 populations to produce the 2300 minimum in the sed , as the @xmath94 of hb stars is a mild function of the envelope mass .
indeed , the @xmath95 color of the uv rising branch in ellipticals is systematically bluer than in metal poor globular clusters ( dor95 ) .
this shows that the average temperature of the hot stars in ellipticals is higher than the average temperature of hb stars in gcs .
only high ( @xmath86 ) models seem to reach blue enough colors ( dor95 , yi et al .
1995 , 1998 ) .
finally , we notice that two metal rich globular clusters in the galactic bulge have been found to host a sizable population of hhb stars ( rich et al .
this has shown that it is indeed possible to produce numerous hhb stars at @xmath96 , although this may be due to some yet unidentified dynamical process in these particular clusters . to summarize ,
most of the theoretical work has just shown quantitatively , in a detailed level , that the options are equivalent : hot stars are produced at high metallicities in old stellar populations if is large and/or if the mass loss rate moderately increases with metallicity .
which combination of the two effects is at work remains unclear . as we will discuss later
, high @xmath36 in combination with high @xmath2 could be a necessary ingredient in order to avoid too much light in the mid - uv from ihb stars .
in summary , the current observational evidences that need to be explained include : * in those ellipticals with the strongest uv upturns @xmath97 0.015 . *
the uv sed requires a small temperature range for the hot stars in the nuclei of ellipticals , peaking at @xmath98 k. * among ellipticals with 2 @xmath3 0.3 , the fraction of hot evolved objects going throuh the hhb channel of evolution varies sizeably , possibly ranging from @xmath99 to @xmath100 for 2 increasing from 0.3 to 0.36 . *
the shape of the lf of the uv bright stars is similar in m32 and in the bulge of m31 . in both galaxies
resolved stars account for only of a small fraction of the uv flux .
thanks to the hut spectra reaching shorter wavelengths compared to iue , a smaller @xmath101 is derived , hence a smaller average fuel consumption @xmath102 is required for the hot stars in the most powerful ellipticals , compared to gr90 estimates .
however the average fuel consumption remains much larger than what provided by p
agb stars , and the best candidates hot stars remain p
eagb and/or hhb+agb manqu objects .
hhb stars and their progeny tend to be favored essentially because their @xmath103 is larger than the required @xmath19 , so only a ( small ) fraction of the stellar evolutionary flux has to go through a very hot helium burning stage . besides , p eagb stars are likely to be distributed over the whole range from @xmath0 5000 to @xmath0 70000 k ( see tracks in castellani and tornamb 1991 ) , thus providing too much flux both in the mid - uv and in the most extreme uv spectral range .
if the bulk of hot stars in giant es are hhb + agb manqu , some @xmath68 of the evolving population in the uv brightest ellipticals has to evolve through channel ( iii ) ( see section 2 ) .
this constrains the combination of the parameters @xmath30 and , plus all the parameters which play a role in determining the metallicity distribution in the central regions of the most powerful ellipticals , in particular its high metallicity tail .
can we learn something on these parameters from the uv sed ?
as repeatedly noticed here , the observations require that the hhb stars are characterized by a narrow range of 0 . in this respect ,
stellar populations of high metallicities , possibly coupled with large helium abundances , are favored .
this stems from an observational argument , since the seds of galactic ( low @xmath2 ) globulars with extended hbs tend to be flatter than those typical of giant ellipticals . at the same @xmath104 color ( i.e. for the same average @xmath103 ) , the @xmath105 color of gcs are redder than those of giant ellipticals , with only high @xmath2 ssp models matching the @xmath106 color ( dor95 , yi et al .
1995 , 1998 ) .
there is also a theoretical argument in favor of the high @xmath2 hypothesis , based on the shape of the relation between and 0 of hb stars . as 0 decreases keeps low untill a threshold value
is reached , after which the relation becomes becomes extremely steep ( cf .
fig . 3.1 in renzini 1977 ) . as a consequence
there is a very narrow range of 0 which corresponds to intermediate effective temperatures .
perhaps more importantly , at high metallicites hb stars appears to exhibit a bimodal behavior , with most of the hb ( as well as the subsequent shell helium burning stage ) being spent either at high or low , virtually avoiding the intermediate regime ( see tracks in dro93 ) .
this tendency is reinforced when high @xmath36 combines with high @xmath2 ( yi et al .
1997b ) . in conclusion ,
the presence of the 2300 minimum and the relatively steep slope of the uv sed in ellipticals speak in favor of high @xmath2 and @xmath36 hhb stars and their progeny . at this point
we notice that the strong 2 index in the nuclei of giant es requires that the @xmath2 distribution has a small ( if any ) component at @xmath107 ( casuso et al .
1996 , greggio 1997 ) .
the sed in the uv offers another argument in support of this picture ( see also bressan , chiosi & fagotto 1994 ; tantalo et al .
1996 ) .
turning now to the question of producing hhb stars at @xmath108 in less than a hubble time , an enhancement of the mass loss rate parameter over the value which fits the gcs properties ( @xmath109 ) seems difficult to avoid . at low dcruz et al .
( 1996 ) require an enhancement of a factor 23 ; at = 23 yi et al .
require an enhancement of a factor @xmath110 .
it s interesting to notice that , due to this large values of @xmath30 , a small mass loss dispersion easily produces the hot helium flashers ( see section 3 ) .
this class of objects naturally provides an upper limit to the distribution on the hb ( castellani & castellani 1993 ) , offering an elegant solution to the problem of why the sed of giant ellipticals shows the turnover at @xmath111 .
indeed , if the hb were populated down to the helium ms , stars would be distributed all the way up to @xmath0 50000 and beyond , hence producing a hard uv spectrum shortward of ly@xmath112 , which is not observed .
thus , the hot stars in the nuclei of giant es are likely to be objects in the helium burning phase which happened to undergo a particularly heavy mass loss while on the rgb .
their large @xmath2 and @xmath36 would produce an evolution confined in a narrow range of effective temperatures .
the hot edge of this range would be populated by objects which ( due to mass loss ) failed helium ignition on the rgb , but succeeded later during the evolution towards the wd stage .
these stars would belong to the high-@xmath2 tail of the distribution in the gr90 picture , or , alternatively , to the high mass loss tail of the distribution of stars around @xmath0 ( dor95 ) .
in the first case , the mass loss parameter @xmath30 should increase with the metallicity ; in the second a large dispersion of @xmath30 at z@xmath0 is needed .
the first option more naturally accounts for the @xmath23 vs 2 correlation , which we are going to consider next .
finally , we attach a great significance to the fact that the bulk of the uv emission in m31 bulge and m32 comes from objects which are fainter than the detection threshold with foc . indeed , this leaves little alternative to hhb stars as the main uv producers ( cf .
1 ) . among the various possibilities ,
the csp in the nuclei of es with 2@xmath3 0.3 can be modeled by a family of closed box models , provided they are pre - enriched to @xmath96 ( greggio 1997 ) . in these models ,
the metallicity distribution is @xmath113 , with @xmath2 varying between a minimum value @xmath114 ( @xmath0 0.5 ) , and a maximum value @xmath115 .
here @xmath116 is the yield as defined by tinsley ( 1980 ) . since 2
is measured in the optical , where low @xmath2 stars have more weight , its value is very sensitive to @xmath114
. the uv flux , instead , would be more sensitive to @xmath115 , if generated by stars in the high-@xmath2 tail of the distribution .
therefore , the @xmath91 2 correlation requires that @xmath114 and @xmath115 are well correlated , e.g. they both increase with galaxy mass ( luminosity ) , which seems plausible .
if hhb stars are produced only above a threshold metallicity at the present age of the stellar populations in ellipticals , then the @xmath91 2 correlation can result from the metallicity distribution shifting to higher and higher values in galaxies with higher and higher 2 ( gr90 ) .
as for the galaxies with 2@xmath117 , there are very few of them in the burstein et al sample , and they define a correlation with a different slope . from the 2 index one
expects their metallicity distribution to be shifted to lower values , and thus it would be interesting to know whether their uv sed allows for a larger contribution from low @xmath2 stars ( i.e. with intermediate temperatures , hence leading to flatter uv upturns ) . to date
this problem has not been quantitatively investigated .
the detection of the redshift evolution of the uv upturn remains perhaps the most attractive opportunity for the future . by detecting the effect we could in fact catch two birds with one stone .
if indeed the uv upturn fades away at @xmath118 , this will represent the decisive test for the hhb+agb manqu origin of the upturn in @xmath119 ellipticals .
moreover , the empirical determination of the derivative @xmath120 ( hence of @xmath121 ) for galaxies of given value of the central velocity dispersion @xmath122 , could be used to set constraints on the age dispersion among local ellipticals that would possibly be much tighter than those set by either optical colors or the fundamental plane relations .
the approach would be the same that bower , lucey & ellis ( 1992 ) have pioneered to set such constraints using the small dispersion about the average @xmath123 relation of local cluster ellipticals , with one advantage .
indeed , @xmath124 evolves very slowly in old populations , i.e. , by 0.02 - 0.03 mag / gyr , while e.g. @xmath23 should evolve 10 , perhaps 20 times faster . in principle
, rest frame @xmath125 colors could set @xmath126 times tighter constraints to age dispersions .
however , the time derivative of @xmath23 as determined from synthetic populations is extremely model dependent , which therefore makes extremely attractive its direct , empirical determination .
we speculate that extensive studies of the uv upturn for cluster vs. field ellipticals up to @xmath127 could greatly help tightening current constraints on the star formation history of early - type galaxies .
windhorst , r.a . ,
pascarelle , s.m . ,
keel , w.c .
, bertola , f. , mccarthy , p.j . , oconnell , r.w . , renzini , a. , & spinrad , h. 1994 , in frontiers of space and ground - based astronomy , ed .
w. wansteker et al .
( dordrecht : kluwer ) , p. 663 | observational as well as theoretical developments subsequent to 1990 concerning the origin of the ultraviolet emission of elliptical galaxies are reviewed and discussed . in particular , successes and failures of an extensive set of predictions contained in greggio & renzini ( 1990 )
are discussed in the light of these development .
strong support to the preferred hot horizontal branch and agb manqu scenario for the origin of the uv emission has come from uit , hut and hst observations .
however , the uv sensitivity of such space telescopes has proved insufficient to detect the predicted fade away effect of the uv upturn with redshift , that we consider the ultimate check for this scenario .
some serendipitous discoveries prompted by studies of the uv upturn are also mentioned .
# 1 # 2 _ mem .
soc .
astron .
it . _ * # 1 * , # 2 # 1 # 2 _ the messenger _ * # 1 * , # 2 # 1 # 2 _ astron .
nach . _ * # 1 * , # 2 # 1 # 2 _ astron .
astrophys .
_ * # 1 * , # 2 # 1 # 2 _ astron .
astrophys .
lett . _ * # 1 * , l#2 # 1 # 2 _ astron .
astrophys . rev . _ * # 1 * , # 2 # 1 # 2 _ astron .
astrophys .
suppl .
ser . _
* # 1 * , # 2 # 1 # 2 _ astron .
j. _ * # 1 * , # 2 # 1 # 2 _ ann .
rev .
astron .
astrophys . _ * # 1 * , # 2 # 1 # 2 _ astrophys .
j. _ * # 1 * , # 2 # 1 # 2 _ astrophys .
j. lett .
_ * # 1 * , l#2 # 1 # 2 _ astrophys .
j. suppl .
_ * # 1 * , # 2 # 1 # 2 _ astrophys .
space sci .
_ * # 1 * , # 2 # 1 # 2 _ adv .
space res .
_ * # 1 * , # 2 # 1 # 2 _ bull .
astron .
inst .
czechosl . _ * # 1 * , # 2 # 1 # 2 _ j. quant .
spectrosc .
. transfer _ * # 1 * , # 2 # 1 # 2 _ mon . not .
r. astr .
soc . _ * # 1 * , # 2 # 1 # 2 _ mem .
r. astr .
soc . _ * # 1 * , # 2 # 1 # 2 _ phys .
lett .
rev . _ * # 1 * , # 2 # 1 # 2 _ publ .
astron .
. japan _ * # 1 * , # 2 # 1 # 2 _ publ .
astr .
soc .
pacific _ * # 1 * , # 2 # 1 # 2 _ nature _ * # 1 * , # 2 = 11 versim#1#22.9truept = 12 2 0 -1 epsf.sty |
a current speeding up universe was discovered almost two decades ago , by using 42 type ia supernovae as standard indicators @xcite .
further data have definitively forecasted this experimental evidence , confirming that the universe is accelerated at late times @xcite .
particularly , after a particular transition epoch @xcite , the universe started unexpectedly to accelerate , showing that only standard matter can not be responsible for its present dynamics @xcite .
the most accepted interpretation includes dark matter and dark energy , respectively responsible for structure formation and repulsive dynamics @xcite . in particular , the first component , i.e. dark matter , permits structures to cluster and form during early phases of universe s evolution , whereas dark energy seems to influence current dynamics , enabling the universe to positively accelerate at small redshift @xcite .
the most accredited paradigm is named _ the @xmath4cdm model _ and assumes a non - evolving vacuum energy cosmological constant @xmath4 , which dominates over baryons and dark matter at our epoch @xcite .
moreover , present data also prospect a spatially flat universe , characterized by baryons and cold dark matter which accounts for almost the @xmath6 of the whole energy content , while the @xmath4 density for about the @xmath7 @xcite .
quite surprisingly , the corresponding dark energy equation of state for the pressure is negative , leading to a non - clear physical interpretation of the main components of the energy momentum tensor @xcite .
thus , the observed speed up description is so far theoretically uncomplete and leads cosmologists to suppose that general relativity breaks down at precise energy scales @xcite .
this fact would provide the expected repulsive effects , capable of accelerating the universe today @xcite .
hence , a self consistent enlargement of general relativity reviews both dark matter and energy effects as fundamental properties of a single theory .
those frameworks represent the well - known _
modified theories of gravity _ , consisting of many models trying to include higher curvature terms , the torsion scalar , additional scalar fields , novel particles derived from supersymmetry and so forth ( see @xcite for details ) . among several possibilities ,
a feasible extension of general relativity , which reproduces the dark energy effects , is the class of @xmath0 models .
those paradigms manage to replace the ricci scalar @xmath8 in terms of a generic analytic function @xmath0 , modifying correspondingly the einstein - hilbert action @xmath9 as @xmath10 where @xmath11 is a coupling constant related to the gravitational constant @xmath12 , @xmath13 the spacetime involved and @xmath14 the matter action .
since @xmath0 is not known _ a priori _ , there exist many formulations of modified @xmath0 theories @xcite passing almost all the available experimental tests at the solar system regime .
hence , there exists a strong degeneracy among cosmological models derived from @xmath0 approaches and the correct form of the @xmath0 function is still object of debate .
however , any proposed @xmath0 represents a _ ad hoc _ formulation for describing the universe dynamics @xcite .
frequently , in fact , the particular choice of @xmath0 is physically unmotivated and so the need of puzzling physical @xmath0 represents the main issue related to the @xmath0 picture . the main purpose of our paper is to discriminate , among different sets of @xmath0 possibilities , the ones passing the local bounds offered by cosmography and to propose a model independent reconstruction of cosmographic @xmath0 functions . to perform our numerical analyses ,
we need to numerically solve the modified friedmann equations .
the modified friedmann equations explicitly depend upon the hubble rate @xmath15 and derivatives , the ricci scalar @xmath8 and the @xmath0 function and derivatives . hence , to express the friedmann equations in terms of a single variable , we rewrite them in function of the redshift @xmath16 . to fix the evolution of @xmath15 and then to calibrate the corresponding hubble rates , entering the friedmann equations , we assume that the curvature dark energy fluid is mimicked in terms of three consolidate dark energy candidates . in particular , we approximate the hubble evolution using the @xmath4cdm model , the chevallier - polarski - linder ( cpl ) parametrization and finally a phenomenological dark energy approximation ( starobinski @xcite ) .
afterwards , we reconstruct the behavior of the @xmath0 function by numerically solving the above cited cosmological equations , fixing the free parameters to agree with the cosmographic bounds . in particular , we consider a numerical approach to determine the behaviors of @xmath0 , passing through the corresponding @xmath5 function , which is biunivocally determined by considering the dependence upon the redshift of the ricci scalar @xmath8 . hence , we rewrite the @xmath0 in terms of a redshift function only , i.e. @xmath5 , which consists in replacing the time dependence of @xmath8 by means of the more feasible variable @xmath16 . employing cosmological data around
@xmath1 $ ] , in which the most of data is concentrated , we depict the shapes of @xmath5 , @xmath17 , @xmath18 and @xmath19 , i.e. the curvature equation of state of the fluid responsible for the universe speeding up in function of @xmath16 only .
afterwards , inverting the numerical equations , i.e. passing from the redshift @xmath16 to @xmath8 , we will be able to build up the numerical behavior of @xmath0 in function of @xmath8 as well .
further , we demonstrate that our method involves the smallest number of impositions possible , since to set _ model independent _ initial conditions on the involved friedmann equations , we make use of the above cited _ cosmography _ of the universe .
indeed , cosmography represents a method to describe the universe dynamics , without using any assumptions _
a priori_. the strategy of cosmography precisely assesses numerical settings in the redshift domain @xmath1 $ ] and it represents an accurate way to bound @xmath20 , the universe acceleration @xmath21 and its variation @xmath22 at present time . to permit cosmography to act on the initial settings of the friedmann equations , we propose the strategy of matching @xmath5 and derivatives with the cosmographic coefficients
moreover , we assume that the ricci scalar and the corresponding derivatives may be related to the cosmographic coefficients as well .
thus , we finally compare the cosmological results with the solar system constraints and we infer a class of @xmath0 functions which better behave as the universe expands . as a consequence ,
we propose new viable @xmath0 functions , which are not postulated but inferred from numerically solve the friedmann equations .
our treatment lies on a direct reconstruction built up in a model independent way , by using the universe cosmography .
once our @xmath0 function is known , we turn back solving the friedmann equations , showing the behavior of our proposal in functions of the redshift parameter @xmath16 .
thus , our methodology enables to highlight the correct form of the @xmath0 function , suggesting which one is really suitable for describing current dynamics .
the paper is thus structured : in sec .
ii , we focus on the main features of @xmath0 cosmology , highlighting the basic requirements that every @xmath0 paradigm should provide . in sec .
iii , we describe the basic demands of cosmography as a tool to fix initial settings on the numerics we are going to show . in particular , we describe the role of every cosmographic coefficients and how cosmography can be framed in the context of @xmath0 gravity . in sec .
iv , we match cosmography with @xmath0 theories and we show which coefficients are of particular interest in our procedures , emphasizing the peculiar property of cosmography to fix limits on @xmath5 model independently . besides we proceed to numerical reconstruct the @xmath5 function , calibrating the hubble rate in the interval @xmath1 $ ] using three different calibrating @xmath23 , i.e. the @xmath4cdm model , the cpl parametrization and a phenomenological reconstruction of dark energy .
once the function is determined , we analyze the principal properties and we discuss their implications in modern cosmology . in sec .
v , we describe the @xmath0 function and we calculate the corresponding properties in the time domain .
we show that the function is compatible with some other paradigms already presented in the literature , although it actually differs from them for some other properties .
finally , sec .
vi is devoted to conclusion and final perspectives of our work @xcite .
to determine the dynamical equations in @xmath0 cosmology , one varies eq .
( [ actionmetric ] ) with respect to the metric @xmath24 .
this leads to fourth order field equations of the form @xcite
@xmath25 f'(r)= \kappa \,\tilde{t}_{\mu\nu}^{(m)}\,,\ ] ] where , baptizing @xmath14 the matter action , we write down @xmath26 with @xmath27 and it will be imposed equal to 1 hereafter .
equivalently , we explicitly split the matter counterpart from curvature dark energy , i.e. @xmath28 where @xmath29 /2 + \right .
\nonumber \\ ~ & ~ & \nonumber \\ ~ & + & \left .
f'(r)^ { ; \alpha \beta } \left ( g_{\alpha \mu } g_{\beta \nu } - g_{\alpha \beta } g_{\mu \nu } \right ) \right \}\ , , \label{eq : curvstress}\end{aligned}\ ] ] the tensor responsible for curvature corrections , usually named the _ curvature energy - momentum tensor_. note that the prime , i.e. @xmath30 , denotes the derivative with respect to @xmath8 , while the semicolon , i.e. @xmath31 , denotes the covariant derivative . this term represents a source for the whole energy - momentum tensor and permits to reproduce the universe dynamics by means of curvature corrections . clearly , by looking at eq .
( [ set ] ) , the matter tensor is coupled to the curvature tensor as @xmath32 and the corresponding modified friedmann equations become @xcite @xmath33\,,\ ] ] and @xmath34 in these equations we neglected scalar curvature terms , i.e. @xmath35 , according to recent observations @xcite and we define the curvature density as @xmath36 the barotropic pressure as @xmath37 and we consider @xmath38 because we study universe expansion in the matter dominated phase . for our purposes
, we assume the pressure to be barotropic since it does not explicitly depend upon the universe entropy @xmath9 @xcite .
the well - known effective curvature barotropic factor is therefore given by @xmath39 } { \left [ f(r ) - r f'(r ) \right ] /2 - 3 h \dot{r } f''(r)}\ , , \label{eq : wcurv}\ ] ] and represents the effective equation of state for the curvature term .
finally , the corresponding ricci scalar @xmath8 can be framed in terms of the hubble parameter as @xcite @xmath40 where we assume @xmath41 , for @xmath42 to hold . in our calculations
, @xmath43 represents the ricci scalar at our time @xmath44 , which corresponds to the redshift @xmath45 . later on ,
we take into account the above cosmological equations and we describe their evolutions by numerically solving eq .
( [ eq : fried1 ] ) .
in particular , numerical outcomes derived from eq .
( [ eq : fried1 ] ) enable us to frame the @xmath0 corrections and to obtain viable @xmath0 candidates , which would represent effective classes of cosmological @xmath0 models .
the problem of fixing the initial conditions to use is here overcome by the use of cosmography . in the next paragraph , we report the basic demands offered by cosmography to set the initial conditions on @xmath0 functions , showing that it is possible to rewrite @xmath0 functions in terms of the redshift @xmath16 .
we therefore highlight the basic requirements that each classes of @xmath0 functions must satisfy and we suggest which functions better work than others .
in this section , we briefly describe the principal aspects of cosmography , characterizing its implications in the context of @xmath0 cosmology .
to do so , we show how to determine numerical initial settings on the @xmath0 function , in order to reconstruct the @xmath0 shapes , in the interval @xmath3 .
first developments towards a complete cosmographic theory has been discussed in @xcite , whereas the modern interpretation of cosmography has been firstly discussed by weinberg @xcite , who noticed that all cosmological observables may be expanded in taylor series .
afterwards , it was soon clear that those series may be directly compared with data , without the need of postulating any cosmological model _ a priori _ @xcite .
the first step is to expand the scale factor @xmath46 around present time @xmath47 @xcite .
thus , the corresponding power series coefficients are referred to as the _ cosmographic series _ ( cs ) , if computed at @xmath47 .
cosmography is essentially based on two simple assumptions , summarized as follows @xcite : * cosmography requests the cosmological principle to hold .
so , cosmography represents a model independent technique to fix cosmic bounds once spatial curvature is somehow fixed @xcite .
present - time universe appears to be spatially flat , i.e. @xmath35 , and several confirmations , coming from this assumptions , are easily determined @xcite .
hence , cosmography becomes at most a model independent approach to limit cosmological models at our epoch . since cosmography does not involve any cosmological model _ a priori _ , it is sometimes referred to as _
cosmokinetics _ @xcite ;
* current time description of our universe passes through a direct composition of the whole energy content , by means of a non - specified number of cosmological fluids , e.g. dark energy , neutrinos , radiations , baryons , etc .
thus , the net pressure becomes a direct sum of different species , i.e. @xmath48 , which corresponds to separate the energy densities as : @xmath49 @xcite . in our work , we circumscribe our attention to barotropic fluids only , in which the pressure becomes an energy function only . from the above assumptions
, we immediately get how to model independently define the cs . indeed ,
expanding @xmath46 , we easily get @xmath50 and so , for our purposes : @xmath51 where we truncated @xmath46 at the second order of taylor expansion . for guaranteeing causality among physical processes ,
_ e.g. _ photon emissions from a source , we considered @xmath52 and we finally set @xmath53 , i.e. the scale factor today @xmath54 , without losing generality .
the cosmographic coefficients are easily defined by : [ pinza ] @xmath55 c [ cols="^ " , ] + + , width=328,height=170 ]
we get three expressions for @xmath5 and @xmath0 , for a total of six analyses .
all the numerical results , performed by using the three hubble rates , i.e. @xmath4cdm , cpl , and the phenomenological one , generate close outcomes which converge to the approximate solution @xmath56 which represents an effective form of @xmath0 inferred from reconstructing @xmath0 from our numerical curves . for guaranteeing that eq .
( [ fdir ] ) accurately reproduces the universe dynamics , we set the three free constant around the following values : [ costanti ] @xmath57 while @xmath58 represents the cosmological constant and @xmath43 is the present value of the ricci scalar , which becomes @xmath59 using the cosmographic results .
moreover , eq . ( [ fdir ] ) accurately reproduces the gravity lagrangian in the range @xmath60 $ ] .
all curves of @xmath5 and @xmath0 , approximating to @xmath61 , provide the relevant fact that eq .
( [ fdir ] ) reduces to the hilbert - einstein action @xmath62 , recovering general relativity with the addition of a cosmological constant .
rephrasing it differently , the @xmath4cdm model persists to be favorite at small redshift domains , indicating however the cosmological constant as a limiting case of a more general paradigm .
another significative result is offered by reconstructing the effective hubble rate from the first friedmann equation eq .
( [ eq : fried1 ] ) , as due to the @xmath63 corrections of eq .
( [ fdir ] ) .
we find , in particular , that a viable approximation is the one plotted in fig .
( [ dmpioneers ] ) .
the approximation may be framed in terms of @xmath64 where [ coeffffff ] @xmath65 that are in agreement with experimental cosmological prediction . here ,
@xmath66 is fixed in terms of the planck results @xcite .
the behaviors of @xmath67 and @xmath19 provide similar contributions in all the three models investigated and their numerical difference are proportional to the percent difference settled by the three different hubble rate .
this suggests that our numerical outcomes do not strongly depend upon the initial conditions on @xmath15 and on its evolution .
in this paper , we described a method to frame the correct @xmath0 , basing our attention on numerically solving the modified friedmann equations . among several possibilities to feature the @xmath0 by postulating its shape according to observations , we propose a new technique consisting in assuming initial settings on the modified friedmann equations offered by cosmography .
the procedure shows the advantage to be model independently fixed in the redshift interval @xmath1 $ ] , leading to a @xmath0 reconstruction which well fits the experimental outcomes provided by present data .
thus , we managed to alleviate the degeneracy among the different choices of @xmath0 functions , by initially reconstructing the dynamic equations in terms of the single variable @xmath16 . indeed , employing the correspondence @xmath68 and showing its invertibility in the redshift range @xmath69 $ ]
, one can shift all variables in terms of the redshift @xmath16 , showing that all observable quantities of interest can be framed analogously .
the differential equation we got has been derived by featuring the first friedmann equation in terms of @xmath16 .
the initial conditions on its dynamics have been bounded by means of cosmography , i.e. a numerical technique which determines cosmic bounds at late times , by simply comparing taylor expansions directly to data .
the powerful of cosmography , i.e. the fact of being completely model independent , set up the initial conditions on the @xmath0 functions and enables one to reproduce universe s expansion history as the knowledge of @xmath23 is somehow known .
we proposed three hubble rates which actually provide the numerics in the interval @xmath3 , i.e. the @xmath4cdm , the cpl and the phenomenological approach .
we aimed to choose those frameworks since are likely the simplest frameworks which better adapt their shapes to frame the universe evolution .
correspondingly , we got a model independent reconstruction of @xmath5 , i.e. the function @xmath0 in terms of the redshift @xmath16 .
we reported the shapes of @xmath5 for each model involved in our numerical analyses and we showed that our outcomes are basically similar but not equivalent , obtaining consequently different classes of plausible @xmath5 functions .
in so doing , we passed through a numerical description of @xmath0 functions , by means of the auxiliary @xmath5 .
in fact , we got a defined class of viable functions which integrate the numeric friedmann equations , providing integration constants accurately fine - tuned by means of cosmography itself .
the final outcome lies on viable forms of @xmath0 functions , which extend the standard approach of general relativity .
the terms here involved are compatible with recent developments on modified @xmath0 functions and permit to conclude that the universe is accurately featured by @xmath8 corrections to the standard hilbert - einstein action . besides , in order to extend @xmath0 effective expression till to the dynamics of the early universe , we are trying to apply our numerical procedure , here described , for redshift parameter major than unity , i.e. @xmath70
. this topic will be the subject of a future work .
the author wishes to express his gratitude to g. lambiase for discussions on the topic of present work and a. strumia for his support .
the author is finantially by infn section of pisa .
a particular thanks to e. vicari for his support and useful discussions .
a. zitrin , m. redlich , t. broadhurst .
j. * 789 * , 51 , ( 2014 ) ; r. i. thompson , c. j. a. p. martins , p. e. vielzeuf , mon . not .
soc . , * 428 * , 3 , 2232 , ( 2013 ) ; r. bean , e. e. flanagan , i. laszlo , m. trodden , phys .
d , * 78 * , 123514 , ( 2008 ) ; s. basilakos , m. plionis , astrophys . j. * 650 * , l1 , ( 2006 ) ; o. farooq , b. ratra , astroph . jour .
, * 764 * , 2 , ( 2013 ) .
r. rebolo , et al . , mon . not .
soc . , * 353 * , 747 , ( 2004 ) ; s. dodelson , et al . ,
apj , * 572 * , 140 , ( 2002 ) ; w. j. percival , et al . ,
mon . not .
soc . , * 337 * , 1068 , ( 2002 ) ; a. s. szalay , et al . , apj , * 591 * , 1 , ( 2003 ) ; e. hawkins , et al . , mon . not .
, * 346 * , 78 , ( 2003 ) ; a. c. pope , et al . ,
apj , * 607 * , 655 , ( 2004 ) .
s. capozziello , g. lambiase , arxiv[gr - qc]:1304.5640 , ( 2013 ) ; m. j. mortonson , d. h. weinberg , m. white , arxiv[astro - ph]:1401.0046 , ( 2014 ) ; m. li , x. d. li , s. wang , y. wang , front .
phys . , * 8 * , 6 , 828 , ( 2013 ) ; a. h. g. peter , arxiv[astro - ph]:1201.3942 , ( 2012 ) .
s. capozziello , v. faraoni , _ beyond einstein gravity _ , springer , ( 2011 ) ; s. capozziello , m. de laurentis , v. faraoni , arxiv:0909.4672v2[gr - qc ] , ( 2009 ) ; g. lambiase , s. mohanty , l. pizza , gen . relativ . and gravit . , * 45 * , 1771 , ( 2013 ) . after the first e - print publication of this paper we have been aware of other analogous approaches that try to reconstruct @xmath0 : s. nojiri , s. d. odintsov , phys
d , * 74 * , 086005 , ( 2006 ) ; s. nojiri , d. odintsov , j. phys .
* 66 * 012005 , ( 2007 ) ; s. nojiri , s. d. odintsov , d. saez - gomez , phys .
b * 681 * , 74 , ( 2009 ) ; p. k. s. dunsby , e. elizalde , r. goswami , s. odintsov , d. saez - gomez , phys .
d * 82 * , 023519 , ( 2010 ) ; m. rinaldi , g. cognola , l. vanzo , s. zerbini , arxiv:1406.1096 [ gr - qc ] .
a. aviles , a. bravetti , s. capozziello , o. luongo , phys . rev .
d , * 90 * , 043531 , ( 2014 ) ; c. gruber , o. luongo , phys .
d , * 89 * , 103506 , ( 2014 ) ; h. wei , x. p. yan , y. n. zhou , jcap , 1401 , 045 , ( 2014 ) ; j. liu , h. wei , arxiv[astro - ph.co]:1410.3960 , ( 2014 ) .
s. capozziello , m. de laurentis , o. luongo , annal .
, * 526 * , 309 - 317 , ( 2014 ) ; a. aviles , a. bravetti , s. capozziello , o. luongo , phys .
d , * 90 * , 043531 , ( 2014 ) ; b. bochner , d. pappas , m. dong , arxiv [ astro-ph.co]:1308.6050 , ( 2013 ) ; m. demianski , e. piedipalumbo , c. rubano , p. scudellaro , mont . not .
soc . , * 426 * , 1396 - 1415 , ( 2012 ) ; j. c. carvalho , j. s. alcaniz , mont . not .
* 418 * , 1873 - 1877 , ( 2011 ) ; l. xu , y. wang , phys .
b , * 702 * , 114 - 120 , ( 2011 ) .
* appendix a * + in this appendix we show the analytical expression for the barotropic density in function of z , i.e. @xmath71 and barotropic curvature factor @xmath19 in function of z : @xmath72 where @xmath73 and @xmath74.\end{gathered}\ ] ] these expressions have been used in this manuscript to compute the correct expression of @xmath75 and @xmath19 from @xmath5 . | the challenging issue of determining the correct @xmath0 among several possibilities is here revised by means of numerical reconstructions of the modified friedmann equations around the redshift interval @xmath1 $ ] .
frequently , a severe degeneracy between @xmath0 approaches occurs , since different paradigms correctly explain present time dynamics . to set the initial conditions on the @xmath0 functions , we involve the use of the so called cosmography of the universe , i.e. the technique of fixing constraints on the observable universe by comparing expanded observables with current data .
this powerful approach is essentially model independent and correspondingly we got a model independent reconstruction of @xmath2 classes within the interval @xmath1 $ ] . to allow the hubble rate to evolve around @xmath3 , we considered three relevant frameworks of effective cosmological dynamics , i.e. the @xmath4cdm model , the cpl parametrization and a polynomial approach to dark energy . finally cumbersome algebra permits to pass from @xmath5 to @xmath0 and the general outcome of our work
is the determination of a viable @xmath0 function , that effectively describes the observed universe dynamics . |
a broad - line region ( blr ) is present in all agns accreting at moderate- to high - eddington ratios .
blrs are important both because they are our best probe of how agns work and because of their potential for readily providing masses of supermassive black holes ( smbhs ) back to the earliest times of galaxy formation . however , in order to be able to use blrs to reliably estimate the masses of smbhs it is essential to understand the structure and kinematics of blrs . over the last four decades
there have been wide - ranging and , not infrequently , mutually contradictory views of the nature of the blr ( see reviews by @xcite , @xcite , and @xcite ) .
however , i believe that the situation is improving .
i review here what i consider to be the clearest pointers to the underlying structure and kinematics of the blr and i argue that , while there are certainly many interesting problems remaining , the basic picture is now becoming fairly secure .
i furthermore believe that this picture applies to _ all _ blrs because blr equivalent widths and line ratios are remarkably similar , especially in the ultraviolet .
the two most basic questions about the blr are `` what does it look like ? '' and `` how is it moving ? '' the traditional picture of the blr of an agn for over 40 years ( and one which is widely depicted in cartoons of agns ) has been that there is a central source emitting ionizing radiation roughly spherically , and that it is surrounded by a roughly spherical mist of cloudlets .
this is depicted in the left - hand - panel of fig .
1 . each individual cloud
, if it is big enough , will have a structure as shown in the right - hand - panel of fig .
it will be highly ionized on the front , and if it has a high - enough column density , it will be mostly neutral on the back .
the front emits high - ionization lines such as heii , hei , ovi , nv , and civ , while the back emits low - ionization lines such as mgii , caii , oi , and feii .
all these lines are well - known in agns .
the emissivity of each line as a function of distance from the front of the cloud can be calculated with the photoionization code cloudy @xcite .
2 shows emissivities for some well - known lines .
@xcite showed that the sum of contributions from clouds with a distribution of cloud properties ( densities and distances from the center ) will automatically produce a total spectrum similar to what is observed from agns .
this is the so - called loc model .
this was important because it showed that no `` fine - tuning '' of cloud conditions was needed to explain agn spectra .
despite the success of the traditional picture in general , and the loc model in particular , in explaining the overall spectrum of an agn , the problem with this picture ( see @xcite ) is that to explain the strengths of the blr lines the covering factor has to be large ( 50% or so ) , yet if this is so , and if the cloudlets are covering the central source uniformly , we ought to see lyman continuum absorption by the blr clouds .
in fact lyman continuum absorption due to the blr is _ never _ convincingly seen ( @xcite - see discussion in @xcite and @xcite ) .
we believe , as proposed by s. phinney ( see @xcite ) , that the need for a high covering factor plus the lack of lyman continuum absorption _ requires the blr to have a flattened distribution and requires us to be viewing it through a hole . _
this conclusion is supported by recovery of what is called the `` transfer function '' of some lines ( the transfer function is the temporal response of a line to a delta - function event in continuum light curve ) .
transfer functions for low - ionization lines have always implied that there is little or no gas along the line of sight @xcite , and thus that at least the low - ionization gas in the blr has a flattened distribution . having a high overall covering factor but
a flattened distribution means that near the equatorial plane there will be a close to a 100% chance that any path will intersect a blr cloud .
the clouds will thus be self - shielding .
radiation from the central source can freely escape near the axis of symmetry , but is strongly diluted in the equatorial plane .
this is schematically illustrated in fig .
it is easy to calculate the average radial dependence of the ionization and the emissivities of all the lines coming from cloudlets with a distribution such as in fig .
the ionization structure of a single cloud in cloudy is now spread out in radius as illustrated schematically in fig .
4 . the horizontal axis in fig .
2 can now be read as distance into the blr rather than distance into an individual cloud .
our model is in fact very similar to the old `` filling factor '' model of @xcite .
the earliest reverberation mapping of multiple lines @xcite showed that the high - ionization lines were coming from smaller radii than the low - ionization lines .
high - ionization lines were also wider ( e.g. , @xcite ) .
the radial ionization stratification of the blr has been well confirmed by later reverberation mapping .
the best reverberation - mapped agn is ngc 5548 .
the horizontal axis of fig .
5 ( taken from @xcite ) shows the reverberation - mapping time lags ( i.e. , the effective radii ) for lines of a variety of ions from @xcite , @xcite , and @xcite .
( see table in @xcite for details ) .
the observed lags cover an order of magnitude in radius .
ngc5548 is not unique in this regard : an identical range of radii has also been found for mrk 110 by @xcite .
the vertical axis shows the lags predicted for the same lines by the loc model @xcite .
it can be seen that while there is a correlation , these predicted lags cover a much smaller range of radii .
the reason for this can be appreciated in fig .
4 . in the loc model ( top part of the figure )
every cloud has a highly - ionized front part ; the clouds just differ in the degree of ionization . the self - shielding model @xcite solves the problem of why the ionization stratification is so strong . in the self - shielded model ( bottom of fig .
4 ) there is a clear spatial separation of the differing ionizations .
it can be seen in fig .
5 that the self - shielding model gives good agreement with the observed lags .
@xcite made the important suggestion that the outer edge of the blr coincided with the dust sublimation radius of the torus .
reverberation mapping observations show that the low - ionization gas in the blr indeed extends out to the dusty torus @xcite .
the covering factor of the torus can be calculated statistically from the ratio of type-1 ( face - on ) to type-2 ( edge - on ) agns , and directly for individual objects from the strength of the thermal emission ( e.g. , @xcite ) .
we argue @xcite that the covering factors of the blr and torus have to be the same .
this is because if the _ torus _ has a lower covering factor than the blr we would see the blr in absorption against the central continuum source in some objects near the type-2 viewing position .
this is never seen . on the other hand , if the _ blr _ has a lower covering factor , some of the dusty torus will see direct radiation from the central source
. this can not be the case for much of the torus because it would then be unable to exist as close in as is seen .
the overall picture we get of the torus and blr is indicated schematically in the cartoon in fig . 6 and in the computer generated renditions shown in fig .
the best description of the appearance is to say that the blr and torus look like a bird s nest .
this picture is identical to that favored for totally independent reasons in an unfortunately almost totally overlooked paper by @xcite .
they inferred a `` bird s nest '' geometry from a combined analysis of line profiles and transfer functions in ngc 5548 .
the positions of masers in ngc 1068 also provides support for a thick blr - torus @xcite .
the kinematics of the blr have been a long - standing problem .
it has been known from the earliest days of agn studies that the lines are very broad ( for a review of the earliest literature see @xcite ) , but doppler shifts only tell us the motion of gas along the line of sight . to know
whether the gas is inflowing , outflowing , moving in random virialized orbits , or in more planar keplerian orbits in a disc we need to know the line - of - sight velocity as a function of position relative to the black hole .
the discovery of narrow intrinsic absorption in ngc 4151 @xcite and of broad absorption lines ( bals ) in phl 5200 @xcite proved that _
some _ gas was outflowing from agns .
however , bals commonly extend to velocities several times higher than those observed for the blr in the same objects ( see , for example , @xcite ) , so it is not clear that there is necessarily any connection between bals and blrs . the case for an outflowing blr
was strengthened though when @xcite and @xcite showed that a radiatively - accelerated outflow could reproduce the observed line profiles well in some objects .
however , @xcite showed that other models could provide comparably good fits to broad - line profiles , and so demonstrated that fits to individual line profiles alone could not uniquely determine the kinematics .
more progress was made by comparing lines of differing ionizations .
@xcite discovered that high - ionization broad lines were blueshifted with respect to low - ionization lines , and pointed out that this requires there to be _ radial _ motions plus some source of opacity .
this blueshifting has now been widely confirmed .
@xcite suggested that the blueshifting could be the result of a `` disk - wind '' model where the high - ionization lines arise in a wind outflowing above the accretion disc .
@xcite pointed out a problem with any purely radial motion : the profiles of c iv and lyman @xmath0 were observed to be very similar , yet , for optically - thick clouds , lyman @xmath0 is emitted very anisotropically . to satisfy this constraint the clouds either had to be optically thin , or not moving purely radially .
obviously the question of the direction of motion could be settled if it could be determined which gas was on which side of the black holes . the best way of doing this is through velocity - resolved reverberation mapping @xcite .
how this works is illustrated in fig .
surprisingly , velocity - resolved reverberation mapping results @xcite strongly ruled out significant outflow of both high- and low - ionization lines ( see example in fig .
ruling out significant outflow of the blr was important not just because of what it said about how agns work , but also because it meant that the blr motions were _ gravitationally dominated_. the blr could thus be used for determining the masses of the central black holes .
this permitted the first reverberation mapping determinations of black hole masses and eddington ratios @xcite .
note that while gravity dominates blr motions , the simple fact that radiation pressure was at one time considered to be driving blr motions ( e.g. , @xcite ) should warn us that radiation pressure might not be negligible @xcite .
while the velocity - resolved reverberation mapping results were good news for the new black - hole - mass - determination industry , they created a problem for the generally accepted `` disk - wind '' explanation of the blueshiftings of high - ionization lines .
disk - wind models are very theoretically appealing , and strong blueshiftings have been taken as signs of strong winds ( e.g. , @xcite ) . however , at the same time , people working on black hole mass determinations were firmly believing that they were using virialized lines !
this has almost caused agn observers to suffer from multiple - personality disorder!@xmath1 it is very difficult to finesse a disk - wind model to fit the velocity - resolved reverberation mapping constraint , all the more so since outflow was first excluded for the _ high_-ionization c iv line @xcite .
we believe , however , that there is a simple solution to the problem : the opacity needed to cause the blueshifting is not primarily absorption but _ scattering _ @xcite .
electron scattering had in fact been considered in the late 1960s to be a significant source of line broadening in agns @xcite , but the idea fell out of favor with the success of the @xcite radiative acceleration model in fitting profiles .
it has , however , long been well known ( e.g. , @xcite ) that scattering off regions with a net radial motion produces line shifts . for an infalling scattering medium ,
photons gain energy .
this is explained in fig .
the process is similar to the well - known fermi acceleration process .
the effect of scattering off radially moving material in agns was considered by @xcite and @xcite . as can be seen in fig .
9 , velocity - resolved reverberation mapping not only excludes outflow , but it also shows that there is a slight _ inflow_. initially this result only had @xmath290% confidence for any one line in one object , but it has been found repeatedly for many lines in many objects now and thus , as pointed out by @xcite , the overall significance is high .
early examples included @xcite .
more recent examples can be found in @xcite , @xcite , @xcite , @xcite , @xcite , @xcite @xcite .
important independent evidence for inflow comes from high - resolution spectropolarimetry ( e.g. , @xcite ) .
the systematic change in polarization as a function of velocity across the balmer lines requires a net inflow of a scattering region somewhat exterior to the balmer lines .
polarization reverberation mapping @xcite can reveal the location of scattering regions .
we have used the _ stokes _ monte carlo radiative transfer code @xcite to model the effects on line profiles of scattering off an inflowing external medium .
the two geometries considered are shown in fig .
one is an infalling spherical distribution of scatterers and the other an infalling cylindrical distribution . in fig
12 we show a comparison of observed profiles of two low- and high- ionization lines in pks 0304 - 392 with various models with .
we adopted an infall velocity of @xmath3 km s@xmath4 based on velocity - dependent reverberation mapping , spectropolarimetry , and the observed mean blueshift ( see @xcite for details ) .
it can be seen that both spherically and cylindrically symmetric models readily reproduce the blueshifting .
an additional advantage of having significant scattering in the blr is that it solves the `` smoothness problem '' for blr line profiles @xcite .
the intrinsic line broadening in an individual blr cloudlet is only of the order of the sound speed ( @xmath2 15 km s@xmath4 ) , yet the velocity broadening of the blr as a whole is hundreds of times greater .
this requires the number of clouds to be very high @xcite .
the limit on the number of discrete clouds has now been pushed up to @xmath5 @xcite .
this constraint is relaxed if there is broadening by scattering . for a typical agn , several independent lines of evidence ( the blueshifting , velocity - resolved reverberation mapping , and spectropolarimetry ) all point to the inflow velocity being of the order of @xmath3 km s@xmath4 .
as has been mentioned , velocity - resolved reverberation mapping ( see , for example , fig .
9 ) implies that the _ dominant _ motion is _ not _ radial , but keplerian or random .
the observed widths of broad lines are indeed several times higher than the inflow velocity , and , of course , the predominant motion for a flattened distribution must be keplerian . as is clear from fig .
6 and 7 , when we observe the blr ( i.e. , in type-1 objects ) we are always seeing it close to face - on .
the keplerian component of velocity must be reduced by @xmath6 , where @xmath7 is the angle between the axis of rotational symmetry and the line of sight .
the statistics of line profiles in the sdss @xcite suggest that for the vast majority of objects @xmath7 is @xmath8 ) .
as was realized by @xcite , the statistics of line widths imply that , in addition to keplerian motion , there has to be a substantial additional component of velocity _ perpendicular _ to the orbital plane .
osterbrock appropriately called this `` turbulence '' .
the vertical component is also necessary for the reconciling the structure of the blr with its kinematics .
as @xcite showed , for ngc 5548 the combined constraints of reverberation mapping and time - averaged line profile and favor the sort of `` bird s nest '' blr distribution shown in figs .
6 and 7 . in summary , i believe that all the evidence points to the blr having a nest - like appearance and having velocity components : @xmath9 where the keplerian velocity , @xmath10 , of an emission line is a couple of times larger that the turbulent velocity , @xmath11 , which is in turn somewhat bigger than the inflow velocity , @xmath12 . the ratios of blr height to radius and of @xmath13 to @xmath11 are similar to those deduced by @xcite .
the only change to the osterbrock model is recognizing that there is also a significant inflow .
it is well established from radio properties ( see @xcite ) that core - dominated agns are simply lobe - dominated agns viewed from near the jet axis ( i.e. , near face - on ) .
@xcite showed from a comparison of continuum shapes and line ratios that core - dominated and lobe - dominated agns have the same underlying optical - to - uv continuum shape and that the sed differences are just due to increased reddening in the lobe - dominated agns .
we thus have every reason to expect the blrs of core - dominated and lobe - dominated agns to be the same on average .
lobe - dominated radio - loud agns should therefore be an excellent laboratory for studying how orientation affects the appearance of the blr .
@xcite found that lobe - dominated agns preferentially had broader and more irregular line profiles .
@xcite discovered that the fwhm of h@xmath14 increases as we see agns more edge - on .
this provided strong support for a flattened blr .
agns with the peaks of their broad balmer lines blueshifted or redshifted from the systemic velocity have long been known @xcite .
it was proposed @xcite that these peaks might represent separate blrs each associated with a member of a supermassive black hole binary , but line profile variability observations on long and short timescales have delivered two fatal blows to this hypothesis .
firstly , although for a while it looked like the expected binary orbital motion was being seen in long - term profile variations in 3c390.3 @xcite , further observations showed that the radial velocity changes were completely inconsistent with a binary black hole @xcite but were instead consistent with orbital motion of concentrations of blr gas orbiting in a disk .
the second fatal blow was that velocity - resolved reverberation mapping of 3c390.3 strongly ruled out the binary blr hypothesis because the redshifted and blueshifted peaks varied simultaneously on a light - crossing timescale @xcite .
this demonstrated conclusively that the double - peaked profiles arose from an inclined disk , as had been widely proposed ( see references and discussion in @xcite ) . despite these double fatal blows to the idea that displaced
broad - line peaks might be due to supermassive binaries , the topic of what signs there might be of sub - parsec supermassive binaries nonetheless remains one of considerable current interest ( see review by tamara bogdanovi in these proceedings ) .
a subsequent comprehensive survey of radio - galaxies by @xcite revealed many disk - like balmer line profiles .
they found the fwhms of the balmer lines to be approximately double those of agns with single - peaked balmer lines . as is shown in fig .
14 , a factor of two reduction in line width is sufficient to make displaced peaks disappear .
@xcite , @xcite , and @xcite have argued that a disk - like emission line contribution is probably present in all blrs but simply hard to recognize because , as illustrated in fig .
13 , the classic double peaks become hard to see when the disk is near to face - on .
it is straight forward to estimate the inclinations of the blrs from broad disk - like line profiles .
these can often be estimated to within a few degrees . @xcite
get inclinations which predominantly have @xmath15 .
their fits to the disk profiles also provide important confirmation that a significant turbulent velocity is needed and give the turbulent velocity for each object . without the turbulent velocity component the peaks of the line profiles would be much too sharp .
the turbulent velocities are fairly well determined from the line profile fits ( to @xmath16 km s@xmath4 ) .
the average blr turbulent velocity needed is 1300 km s@xmath4 .
this is roughly what would be expected from the height of the blr / torus .
the 1-@xmath17 scatter in the derived turbulent velocities is only @xmath18 km s@xmath4 , which is only slightly greater than the average formal uncertainty in the estimates .
@xcite has estimated inclinations for single - peaked agns . for these
we mostly see disks with inclinations of @xmath19 ( see also bon et al . in these proceedings ) .
the difference in @xmath6 between the displaced - blr - peak agns and single - peaked agns is thus about a factor of two .
this agrees with the ratios of fwhms for the two samples .
the component of velocity perpendicular to the equatorial plane is _ vital _ for agn black hole mass determinations ! without this strong turbulent velocity component , variations in @xmath6 would introduce _ substantial _
scatter into agn black hole mass estimates , especially since type-1 agns are observed close to face - on .
there is recent evidence that there is remarkably little scatter in agn mass estimates .
firstly , it has become apparent @xcite that the two main methods of estimating black hole masses from the blr agree surprisingly well .
the dibai single - epoch - spectrum method @xcite and reverberation mapping methods agree to within the expected errors . @xcite has shown furthermore that a simple refinement of the method produces even better agreement .
the agreements mean that such methods are estimating the effective radii of the blr correctly .
as @xcite discuss , the success of the dibai method means that the inner regions of agns are very similar .
in particular : 1 .
the spectral energy distribution ( sed ) from the optical to the far uv must be very similar in all type-1 agns because the optical region where the flux is measured is far removed in energy from the far uv which is photoionizing the gas . although agn seds _
look _ different , @xcite and @xcite have already argued that the apparent variation is not real but is primarily caused by reddening .
there is a simple scaling relationship between the luminosity and the effective radius .
this is supported by reverberation mapping estimates of the effective radii of blrs @xcite .
both the dibai and reverberation - mapping methods of estimating black hole masses depend on observed blr line widths , so geometric differences and orientation effects will affect _ both _ methods .
an important _ external _ check on the accuracy of agn black hole mass estimates is provided by the tightness of the relationship between black hole mass , @xmath20 , and luminosity , @xmath21 , of the bulge of the host galaxy .
@xcite have recently shown that estimating @xmath20 by the dibai method and @xmath21 from the fraction of starlight in sdss spectra gives a scatter of @xmath22 dex in @xmath23 ( see fig .
@xcite have estimated @xmath21 completely independently for a different set of agns using hst photometry and published reverberation mapping mass estimates .
they get a scatter in @xmath23 of @xmath24 dex . both of these scatters in the agn @xmath25
log@xmath21 relationships are smaller than the @xmath26 dex scatter @xcite and others find when @xmath20 is determined by stellar dynamical methods , but they are still greater than the @xmath27 dex scatter in the @xmath20@xmath28 relationship for pure bulge ( i.e. , barless ) galaxies @xcite . the dibai method and
the method proposed by @xcite seems to give particularly tight @xmath20@xmath28 and @xmath20@xmath29 relationships for the most massive elliptical galaxies @xcite .
this is probably because they have the least intrinsic scatter in the @xmath20@xmath28 relationship .
these comparisons with predictions from host galaxy properties imply that black hole mass determinations from the blr are surprisingly accurate as accurate as the best stellar - dynamical estimates .
this accuracy of black hole mass estimates made using the blr provides strong support for all type-1 agns being very similar as far as the structure and kinematics of the blr goes , and for orientation effects being minimal .
the accuracy of agn black hole mass estimates is thus consistent with there being a substantial turbulent blr velocity component and type-1 agns being seen close to pole - on .
the circular component of motion of blr clouds is a simple consequence of gravity , but the turbulent vertical component of motion has to be maintained against dissipation losses , and an outward transfer of angular momentum is necessary to get inflow .
we now know that the viscosity needed to drive the outward flow of angular momentum in accretion discs , and hence the inward flow of matter , is the magneto - rotation instability ( mri ) @xcite
. over the last decade increases in computing power and the development of more sophisticated programs by several groups have allowed increasingly detailed magneto - hydrodynamic ( mhd ) simulations of accretion flows ( e.g. , @xcite , and references therein ) . in these models attention
has been focused a lot on the low - density outflows . because emissivity goes as the square of density ,
the emission is dominated by the high - density . to my mind
what is impressive about every single one of these models is that , despite the different modeling approaches , _ the velocity fields of the high - density material all match the velocity field inferred for the blr_. i.e. , the dominant motion is keplerian , but there is substantial turbulence , and a significant inflow .
this is very clear when one watches movies different groups make of their simulations .
i believe that these simulations give us a physical basis for what we have deduced from blr observations : the blr _ is _ the material accreting onto the black holes . it was indeed noted a long time ago that if the blr is inflowing it can provide the necessary mass flux for powering the agn @xcite .
i think we now have a fairly good emerging picture of what the blr is like and what role it plays in the life of an agn .
interstellar material approaching the nucleus settles into a flattened distribution , the thick torus .
material loses angular momentum because of mri turbulence and gradually spirals inwards . when the material of the torus gets within the dust sublimation radius , the dust evaporates and we have the blr .
the turbulent blr continues to spiral inwards towards the black hole where it is eventually accreted .
the degree of ionization increases as the gas gets closer in .
the optically - thick material , which will tend to be concentrated towards the mid - plane , produces continuum emission ; the more optically - thin material produces the blr .
not all of the blr is accreted .
some of it is driven off the surface of the blr / torus in a high - velocity , low - density wind , as is found in all the mhd simulations and as is observed .
although i believe we are getting a clear overall picture of the blr , there is still plenty to work on both observationally and theoretically ! for a start , the picture discussed above needs to be thoroughly tested to verify that it works for all objects and not just a few well - observed objects such as ngc5548 .
more work needs to be done to see whether a disk - wind model ( the leading rival to the model presented here ) could also explain everything .
we know that there is outflowing gas as well as gas accreting onto the black hole .
the question is : how much of this is also contributing to the broad - line profiles ? ( especially to the high - ionization lines ) .
@xcite , for example , have shown that outflows can match some observed blr profiles , so determining the relative contribution of an outflow to broad - line profiles from line - profile fitting alone is difficult .
i think that reverberation mapping and spectropolarimetry ( see , for example , @xcite ) are going to provide the best answers .
@xcite found marginal evidence for some outflow in mrk 110 , and @xcite have found a clearer signature of an apparent outflow component in velocity - resolved reverberation mapping of ngc3227 .
since the @xcite results are from a single short observing campaign , i do not think that ngc3227 presents a major problem yet for the general picture presented here.@xmath28 relationship . ]
if follow - up observing campaigns confirm the signature of outflow in ngc3227 then this would be a significant challenge to the model favored here .
nevertheless , the @xcite result does caution us that agns might not all be identical in the relative dominance of inflow and outflow .
even if we are right about the basic structure of the blr and torus of agns , there are still a lot of interesting and potentially important details in need of further investigation .
although in this review i have been emphasizing the similarities among agns and what they imply , there are some significant _ differences _ in the blrs too ( see , for example , @xcite ) . if our basic framework of how an agn works is correct , then the differences need to be explicable within the framework too .
space here only permits a brief mention of some of these problems , but fortunately many of them are reviewed and discussed elsewhere in these proceedings ( see , for example , the reviews by mike eracleous and jack sulentic ) .
it has been known for over three decades now that object - to - object differences are correlated with each other , and one of main drivers of the correlated differences is the eddington ratio ( see @xcite and these proceedings ) .
since we now have reliable agn black holes masses , we also have reliable eddington ratios , so there is a lot that can be done in investigating the dependence of blr properties on accretion rate .
i think there is a lot that needs explaining here .
the biggest object - to - object difference in optical spectra is optical feii emission @xcite .
understanding how the very strong optical feii emission seen in agns is produced has been a long - standing problem ( see @xcite for recent discussions ) . in the blr model discussed here , optical feii emission arises in the outer part of the blr just inside the torus ( and quite likely overlapping with it ) , but this does not readily explain why optical feii is so much stronger in some objects than others .
another mystery of the correlated object - to - object differences is the strength of the _ narrow_-line region ( nlr ) emission .
this is the other strong object - to - object difference and it is mysteriously strongly anti - correlated with feii emission @xcite .
a complete model of agns needs to explain why the nlr and blr know about each other .
although i have argued that the basic properties of agns with broad disk - like balmer line profiles are consistent with the picture presented here , these objects , and especially the variability of their profiles , present some special challenges , as is discussed in mike eracleous s review .
there is also a lot more to be learned with orientation effects . in summary ,
i think that although our overall picture of the blr and the role it plays in the agn phenomenon is becoming clearer , many mysteries remain , there is still a lot to learn , and there are probably surprises in store .
i am grateful to luka popovi , milan dimitrijevi , and the other members of the scientific organizing committee of the 7th serbian conference on spectral line shapes for inviting me to speak on this topic .
i would like to express my appreciation to dragana ili and all the members of the local organizing committee for providing a very pleasant and stimulating experience throughout the conference , both culturally and scientifically .
i also have to thank my collaborators and former graduate students for all their contributions and discussions over the years , and the anonymous referee for useful comments .
this research has been supported in part by us national science foundation grant ast 08 - 03883 .
anderson , k. s. , & kraft , r. p. 1969
, apj , 158 , 859 anninos , p. , fragile , p. c. , & salmonson , j. d. 2005 , apj , 635 , 723 antonucci , r. 1993 , ara&a , 31 , 473 antonucci , r. r. j. , kinney , a. l. , & ford , h. c. 1989 , apj , 342 , 64 arav , n. , barlow , t. a. , laor , a. , sargent , w. l. w. , & blandford , r. d. 1998 , mnras , 297 , 990 atwood , b. , baldwin , j. a. , & carswell , r. f. 1982 , apj , 257 , 559 auer , l. h. , & van blerkom , d. 1972 , apj , 178 , 175 axon , d. j. , robinson , a. , young , s. , smith , j. e. , & hough , j. h. 2008 , memorie della societa astronomica italiana , 79 , 1213 balbus , s. a. , & hawley , j. f. 1991 , apj , 376 , 214 baldwin , j. a. 1975 , apj , 201 , 26 baldwin , j. a. , ferland , g. j. , korista , k. t. , hamann , f. , & lacluyz , a. 2004 , apj , 615 , 610 baldwin , j. , ferland , g. , korista , k. , & verner , d. 1995 , apj .
lett . , 455 , l119 bentz , m. c. , peterson , b. m. , pogge , r. w. , vestergaard , m. , & onken , c. a. 2006 , apj , 644 , 133 bentz , m. c. , et al .
2008 , apj . lett . , 689 , l21 bentz , m. c. , peterson , b. m. , netzer , h. , pogge , r. w. , & vestergaard , m. 2009a , apj letts . , 697 , 160 bentz , m. c. , peterson , b. m. , pogge , r. w. , & vestergaard , m. 2009b , apj .
, , 694 , l166 bentz , m. c. et al
. 2009c , apj submitted [ arxiv:0908.0003 ] blumenthal , g. r. , & mathews , w. g. 1975 , apj , 198 , 517 bochkarev , n. g. , & gaskell , c. m. 2009 , ast .
, 35 , 287 bon , e. 2008 , serbian astron .
j. , 177 , 9 bon , e. , popovi , l. . , ili , d. , & mediavilla , e. 2006 , new astronomy review , 50 , 716 boroson , t. a. , & green , r. f. 1992 , apjs , 80 , 109 boroson , t. a. , & oke , j. b. 1984 , apj , 281 , 535 bottorff , m. c. , baldwin , j. a. , ferland , g. j. , ferguson , j. w. , & korista , k. t.2002 , apj , 581 , 932 capriotti , e. , foltz , c. , & byard , p. 1980 , apj , 241 , 903 capriotti , e. , foltz , c. , & byard , p. 1980 ,
apj , 245 , 396 clavel , j. , et al .
1991 , apj , 366 , 64 crenshaw , d. m. , & blackwell , j. h. , jr . 1990 ,
letts . , 358 , l37 denney , k. d. , et al . 2009a , apj , 702 , 1353 denney , k. d. , et al . 2009b , apj letts . ,
704 , l80 dibai , .
a. 1977 , soviet astron .
, 3 , 1 dietrich , m. , et al.1998 , apjs , 115 , 185 dietrich , m. , wagner , s. j. , courvoisier , t. j .- l . , bock , h. , & north , p. 1999 ,
a&ap , 351 , 31 done , c. , & krolik , j. h. 1996 , apj , 463 , 144 dong , x. , wang , j. , wang , t. , wang , h. , fan , x. , zhou , h. , & yuan , w. 2009 , arxiv:0903.5020 doroshenko , v. t. , sergeev , s. g. , & pronik , v. i. 2008 , astron . rep . , 52 , 442 edmonds , f. n. , jr .
1950 , apj , 112 , 307 eracleous , m. , & halpern , j. p. 1994 , apjs , 90 , 1 eracleous , m. , & halpern , j. p. 2003 ,
apj , 599 , 886 eracleous , m. , halpern , j. p. , gilbert , a. m. , newman , j. a. , & filippenko , a. v. 1997 , apj , 490 , 216 ferland , g. j. , korista , k. t. , verner , d. a. , ferguson , j. w. , kingdon , j. b. , & verner , e. m. 1998 , pasp , 110 , 761 ferrara , a. , & pietrini , p.1993 , apj , 405 , 130 gaskell , c. m. 1982 , apj , 263 , 79 gaskell , c. m. 1983 , liege international astrophys .
colloq . , 24 , 473 gaskell , c. m. 1987 , nato asic proc . 208 : astrophysical jets and their engines , 29 gaskell , c. m. 1988 , apj , 325 , 114 gaskell , c. m. 1996 , apj .
, 464 , l107 gaskell , c. m. 2009a , apj . , submitted [ arxiv:0908.0328 ] gaskell , c. m. 2009b , in preparation gaskell , c. m. , goosmann , r. w. , & klimek , e. s. 2008 , memorie della societa astronomica italiana , 79 , 1090 gaskell , c. m. , goosmann , r. w. , merkulova , n. i. , shakhovskoy , n. m. , & shoji , m. 2008a , apj , submitted [ arxiv:0711.1019 ] gaskell , c. m. , klimek , e. s. , & nazarova , l. s. 2007 , apj , submitted [ arxiv:0711.1025 ] gaskell , c. m. & kormendy , j. 2009 , in galaxy evolution : emerging insights and new challenges , ed . s. jogee , lei hao , g. blanc , and i. marinov , asp conference series , in press [ arxiv:0907.1652 ] gaskell , c. m. , & snedden , s. a. 1997 , baas , 29 , 1252 gaskell , c. m. , & snedden , s. a. 1999 , in structure and kinematics of quasar broad line regions , ed . c. m. gaskell , w. n. brandt , m. dietrich , d. dultzin - hacyan , & m. eracleous , asp conf .
ser , 175 , 157 gaskell , c. m. & sparke , l. s. 1986 , apj , 305 , 175 goosmann , r. w. , & gaskell , c. m.2007 , a&ap , 465 , 129 graham , a. w. 2008 , apj , 680 , 143 greenhill , l. j. , gwinn , c. r. , antonucci , r. , & barvainis , r. 1996 , apj lett , 472 , l21 gltekin , k. et al .
2009 , apj , 698 , 198 hawley , j. f. , & krolik , j. h. 2001 , apj , 548 , 348 horne , k. , welsh , w. f. , & peterson , b. m. 1991 , apj letts , 367 , l5 hu , c. , wang , j .- m . , ho , l. c. , chen , y .- m . , zhang , h .- t . , bian , w .- h .
, & xue , s .- j .
2008 , apj , 687 , 78 ili , d. , popovi , l. .
, len - tavares , j. , lobanov , a. p. , shapovalova , a. i. , & chavushyan , v. h. 2008 , memorie della societa astronomica italiana , 79 , 1105 isobe , t. , feigelson , e. d. , akritas , m. g. , babu , g. j. , 1990 , apj , 364 , 104 joly , m. , vron - cetty , m. , & vron , p. 2008 , revista mexicana de astronomia y astrofisica conference series , 32 , 59 kallman , t. r. , & krolik , j. h.1986 , apj , 308 , 805 kaneko , n. , & ohtani , h. 1968 , aj , 73 , 899 kaspi , s. , maoz , d. , netzer , h. , peterson , b. m. , vestergaard , m. , & jannuzi , b. t. 2005 , apj , 629 , 61 kaspi , s. , smith , p. s. , netzer , h. , maoz , d. , jannuzi , b. t. , & giveon , u. 2000 , apj , 533 , 631 kollatschny , w. 2003 , a&ap , 407 , 461 koratkar , a. p. & gaskell , c. m. 1989 , apj , 345 , 637 koratkar , a. p. , & gaskell , c. m. 1991a , apj , 375 , 85 koratkar , a. p. , & gaskell , c. m. 1991b , apjs , 75 , 719 koratkar , a. p. , & gaskell , c. m. 1991c , apj . letts . , 370 , l61 korista , k. t. , et al . 1995 ,
apjs , 97 , 285 korista , k. t. , & goad , m. r. 2000 , apj , 536 , 284 krolik , j. h. , horne , k. , kallman , t. r. , malkan , m. a. , edelson , r. a. , & kriss , g. a. 1991 , apj , 371 , 541 kuehn , c. a. , baldwin , j. a. , peterson , b. m. , & korista , k. t. 2008 , apj , 673 , 69 la mura , g. , di mille , f. , ciroi , s. , popovi , l. . , & rafanelli , p. 2009 , apj , 693 , 1437 leighly , k. m. , & moore , j. r. 2004 , apj , 611 , 107 lynds , c. r. 1967 , apj , 147 , 396 lynds , c. r. 1968 , aj , 73 , 888 macalpine , g. m. 1972 , apj , 175 , 11 macalpine , g. m. 2003 , revista mexicana de astronomia y astrofisica conference series , 18 , 63 maiolino , r. , shemmer , o. , imanishi , m. , netzer , h. , oliva , e. , lutz , d. , & sturm , e. 2007 , a&ap , 468 , 979 mannucci , f. , salvati , m. , & stanga , r. m. 1992 , apj , 394 , 98 mathews , w. g. , & capriotti , e. r. 1985 , astrophysics of active galaxies and quasi - stellar objects , 185 marconi , a. , axon , d. j. , maiolino , r. , nagao , t. , pastorini , g. , pietrini , p. , robinson , a. , & torricelli , g. 2008 , apj , 678 , 693 marziani , p. , sulentic , j. w. , dultzin - hacyan , d. , calvani , m. , & moles , m. 1996 , apjs , 104 , 37 mathews , w. g. , & wampler , e. j. 1985 , pasp , 97 , 966 mathis , j. s. 1970 , apj , 162 , 761 mayall , n. u. 1934 , pasp , 46 , 134 miley , g. k. , & miller , j. s. 1979 , apj letts . , 228 , l55 netzer , h. , & laor , a. 1993 , apj lett , 404 , l51 obrien , p. t. , et al .
1998 , apj , 509 , 163 ohsuga , k. , mineshige , s. , mori , m. , & kato , y. 2009 , pasj , 61 , l7 osterbrock , d. e. 1977 , apj , 215 , 733 osterbrock , d. e. 1978 , proceedings of the national academy of science , 75 , 540 osterbrock , d. e. , & mathews , w. g. 1986 , ara&a , 24 , 171 padovani , p. , & rafanelli , p. 1988 ,
a&ap , 205 , 53 peterson , b. m. et al .
1991 , apj , 338 , 119 pijpers , f. p. , & wanders , i. 1994 , mnras , 271 , 183 popovi , l. . ,
mediavilla , e. , bon , e. , & ili , d. 2004 , a&ap , 423 , 909 proga , d. 2003 , apj , 585 , 406 sergeev , s. g. , pronik , v. i. , sergeeva , e. a. , & malkov , yu .
f. 1999 , apjs , 121 , 159 seyfert , c. k. 1943 , apj , 97 , 28 shafee , r. , mckinney , j. c. , narayan , r. , tchekhovskoy , a. , gammie , c. f. , & mcclintock , j. e. 2008 , apj .
letts . , 687 , l25 shuder , j. m. 1982 , apj , 259 , 48 smith , j. e. , robinson , a. , young , s. , axon , d. j. , & corbett , e. a. 2005 , mnras , 359 , 846 steiner , j. e. 1981 , apj , 250 , 469 suganuma , m. , et al . 2006 , apj , 639 , 46 sulentic , j. w. , marziani , p. , & dultzin - hacyan , d. 2000 , ara&a , 38 , 521 turnshek , d. a. , grillmair , c. j. , foltz , c. b. , & weymann , r. j. 1988 , apj , 325 , 651 ulrich , m .- h . , & horne , k. 1996 , mnras , 283 , 748 verner , e. , bruhweiler , f. , johansson , s. , & peterson , b. 2009 , physica scripta volume t , 134 , 014006 welsh , w. f. , martino , d. l. , kawaguchi , g. , & kollatschny , w. 2007 , the central engine of active galactic nuclei , 373 , 29 weymann , r. j. 1970 , apj , 160 , 31 wilkes , b. j. 1984 , mnras , 207 , 73 wilkes , b. j. , & carswell , r. f. 1982 , mnras , 201 , 645 wills , b. j. , & browne , i. w. a.1986 , apj , 302 , 56 | i review progress made in understanding the nature of the broad - line region ( blr ) of active galactic nuclei ( agns ) and the role blrs play in the agn phenomenon .
the high equivalent widths of the lines imply a high blr covering factor , and the absence of clear evidence for absorption by the blr means that the blr has a flattened distribution and that we always view it near pole - on .
the blr gas is strongly self - shielding near the equatorial plane .
velocity - resolved reverberation mapping has long strongly excluded significant outflow of the blr and shows instead that the predominant motions are keplerian with large turbulence and a significant net inflow .
the rotation and turbulence are consistent with the inferred geometry .
the blueshifting of high - ionization lines is a consequence of scattering off inflowing material rather than the result of an outflowing wind .
the rate of inflow of the blr is sufficient to provide the accretion rate needed to power the agn . because the motions of the blr are gravitationally dominated , and the blr structure is very similar in most agns , consistent black hole masses can be determined .
the good correlation between these estimates and masses predicted from the bulge luminosities of host galaxies provides strong support for the similarity of agn continuum shapes and the correctness of the blr picture presented .
it is concluded that although many mysteries remain about the details of how agns work , a general overall picture of the torus and blr is becoming clear . active galactic nuclei , broad - line region , accretion processes , dust |
the recent iso mission has produced a number of deep mid - infrared extragalactic surveys [ 1,2,3,4 ] many of which are presented elsewhere in these proceedings . in order to accurately evaluate the apparent source evolution found in these surveys it is essential to have a stable and exact local infrared picture that is compatible with iso surveys .
there have been a number of previous iras 12 @xmath0 m extragalactic samples produced from either the point source catalog , hereafter psc , or the faint source catalog , hereafter fsc .
the fsc was constructed by co - adding the individual psc scans and is consequently deeper at 12 @xmath0 m by approximately one mag ; the fsc is considered complete to @[email protected] jy . due to the greater depth of the fsc only those samples constructed from it will be considered here [ 5,6 hereafter rms and fsxh ] .
in essence none of these samples are truly compatible with the deeper iso surveys because they apply longer wavelength iras colour selection criteria and do not objectively classify galaxies .
some of these samples additionally suffer from inaccurate source flux estimation , no correction for the overdensity due to large scale structure and inaccurate k - correction . due to the lack of space here
these latter two points are not considered although we refer the interested reader to [ 6,14 ] for excellent coverage of these problems .
selecting objects at 12 @xmath0 m without colour selection will produce an abundance of stars over galaxies due to the jeans tail of stellar emission . without exception every extragalactic 12 @xmath0 m sample to date
has had ( the majority of ) stars removed by applying longer wavelength iras colour criteria .
this technique is clearly incompatible with iso surveys where no colour criterium is applied and will cause a bias towards dusty galaxies .
this also enforces that every galaxy must have a longer wavelength flux , producing incompleteness even within the selection boundaries .
for example , in rms the primary selection is @[email protected] jy but every galaxy must also have @[email protected]@xmath1 or @xmath4@xmath2@xmath1 .
however due to the completeness of fsc ( @[email protected] jy and @[email protected] jy ) this sample can not be complete for @[email protected] jy or @[email protected] respectively . to produce accurate extragalactic luminosity functions and understand the galaxy contributions to fainter source counts it is necessary to classify galaxies in an objective way , the most common technique is with optical line ratios [ 7,8 ] .
to date the only classified 12 @xmath0 m sample is rms although their classification was taken from various catalogues which differ in the definition of extragalactic type and completeness . as a comparison to this classification
we have obtained line ratios from the literature for 349 of the 483 rms galaxies with @xmath6@xmath20 degrees .
this gives completenesses of 72% , 78% and 93% for objects
@[email protected] , 0.3 and 0.5 jy respectively .
due to the spectroscopic incompleteness at lower fluxes and the colour selection incompleteness we only consider those of @[email protected] jy here , see table 1 ; our classification follows that of [ 7,9 ] .
.extragalactic classification [ cols= " < , < , < , < , < , < " , ] [ tab1a ] all galaxies are found to show h@xmath7 emission although for some galaxies w@xmath8(h@xmath7)@xmath51 angstrom and they would appear as absorption line objects in lower resolution / signal to noise spectra . of the hii galaxies , 50% show evidence for significant star formation ( w@xmath8(h@xmath7)@xmath210 angstroms ) .
rms classified an object as an agn if it is present in an agn catalogue .
we find a good agreement in classification for this object class , the principal reason for the construction of the rms sample .
liners were classified in the rms sample if they were present in an agn catalgoue and consequently this sample is thought to be incomplete .
we confirm this here .
rms did not classify hii galaxies although they considered those galaxies not classified as a liner or agn , but with high infrared luminosities , to be starburst galaxies and all other objects to be normal galaxies .
the fsc detection algorithm has been optimised for unresolved sources therefore fluxes for extended sources need to be calculated from the coaddition of scans ( the addscan technique nowadays accessed via xscanpi ) [ 10 ] .
however , if a source is unresolved , the flux calculation using this technique leads erroneously to a larger flux than the fsc flux [ 10 ] . whilst fsxh carefully calculate the fluxes of extended and unresolved sources
seperately , rms treat all sources as extended and consequently overestimate the fluxes of unresolved galaxies ; due to the large beamsize of iras a large number of galaxies can be overestimated .
the magnitude of this effect can be estimated from isocam observations .
unfortunately only a few observations are available for the lw10 filter ( the closest to the iras 12 @xmath0 m band ) therefore in order to predict the iras 12 @xmath0 m fluxes we have used observations in the lw2 ( 6.7 @xmath0 m ) and lw3 ( 14.3 @xmath0 m ) bands and a spectral decomposition technique similar to that described in [ 11 ] . in this simple model the mid - infrared emission
is produced by two components : hii regions ( using m17 [ 12 ] ) and photo - dissociation regions ( pdr ) ( using ngc7023 [ 13 ] ) .
the ratio of hii to pdr is calculated from the ratio of lw2 to lw3 fluxes , a synthetic spectra is produced and the iras 12 @xmath0 m flux is calculated .
this technique will be somewhat imprecise due to the uncertainty of the cam photometry and the ability of the model to reproduce the galactic spectrum .
overall we estimate an uncertainty of @xmath920% , roughly equal to the worst error in the fsc photometry . in figure 1
we plot our predicted fluxes against the fsc and rms fluxes divided by the predicted flux ( i.e. the relative errors in flux ) .
m flux comparison ] only those galaxies with z@xmath22,500 kms@xmath10 have been plotted as they should all be unresolved in the iras beam . in general a good agreement between the predicted flux and fsc flux is found leading us to believe that the true fluxes for these objects are close to the fsc flux .
consequently rms may be overpredicting the source flux for @xmath950% of their objects .
this will have the effect of shifting galaxies from faint flux / luminosity bins to higher bins causing unpredictable effects in the source count and luminosity function determinations .
our new iras 12 @xmath0 m sample aims to address these problems and create a local infrared sample that is compatible with iso surveys .
the sample definition is close to that of rms : sources are taken from the fsc for those objects with @xmath11b@xmath11@xmath225 degrees , @[email protected] jy and a moderate or good quality detection ( s / n@xmath23 ) .
this selection results in 31002 objects , by comparison the rms colour selection bias produces @xmath91100 objects .
our sample provides an interesting complement to the psc - z extragalactic survey [ 15 ] which selects those objects from the psc with @[email protected] jy . as with our sample
they do not apply colour selection criteria , stars are identified on schmidt plates . in terms of the extragalactic objects we would expect many galaxies in common
although our sample should have a higher fraction of quasars and early type galaxies . by not applying colour selection to distinguish galaxies from stars we have had to devise an alternative technique . the key assumption in our technique
is that stars have , in the majority , predictable properties and therefore with the large and complete optical stellar databases currently available ( in particular the guide star catalogue [ 16 ] , hereafter gsc ) it is possible to find the majority of stars in our sample through positional cross correlation .
an important factor here is in determining the completeness of an optical stellar catalgoue in an infrared sample , requiring an understanding of the general optical and infrared characteristics of stars . to determine these characteristics for our stars we have used simbad , which provides stellar classifications , and produced a large sample ( @xmath99000 objects ) of classified stars . when cross correlationing to optical positions we consider a star correlated if its optical position falls within 5@xmath12 of the iras position ( the mean major and minor error ellipse axes are 16.9 `` and 2.0 '' respectively ) .
the properties of our classified stellar sample are shown in figure 2 .
we have only plotted those stars for which there are at least 10 objects in a classified class and only a subset of these classes are shown here for clarity . a clear correlation between stellar type and flux ratio is found .
the predicted black body colours for the different stellar types follows a straight line passing close to the a1v to k5iii points .
the interesting deviation observed for stars beyond type moiii is _ possibly _ due to an increasing amount of stellar absorption in the v band .
these optical - infrared flux correlations provide an essential tool , the ability to predict the iras 12 @xmath0 m flux of a star for a given spectral type or b - v colour . in the cases where we find an iras source associated with both a galaxy and a star we will be able to predict the flux from the star and
therefore estimate the flux from the galaxy . the gsc is considered complete for 6@xmath5v@xmath515 mags over the whole sky . based on our stellar infrared diagnostic correlation
this corresponds to a depth of @[email protected] jy and therefore any stars earlier than a type m6iii in our sample should be in the gsc .
however , to accurately cross correlate the positions between the gsc and our sample requires accounting for proper motion .
the schmidt plates for the gsc were taken in 1975 and 1982 , by comparison the mean iras observation epoch is 1983.5 . by analysing a sub - sample of hipparcos stars ( @xmath94000 objects , selected by sky area )
we find a mean proper motion of 0.03``yr@xmath10 , with a one @xmath12 maximum of 0.13''yr@xmath10 ( the maximum in the whole catalogue is 6"yr@xmath10 ) .
taking into account the different observation epochs and the mean iras error ellipse , virtually all our stars should fall within 5@xmath12 of their gsc position .
we initially searched for associations by selecting all objects within 2 of the iras position ( 2 corresponds to @xmath95@xmath12 of the mean positional uncertainty in the major axis direction ) although we only consider a star cross - correlated if it falls within 5@xmath12 of an iras source . from this cross correlation
we find that over that 29000 of our iras sources are stars .
of these a small ( @xmath90.5% ) , but significant , fraction show a considerable infrared excess ( @xmath13@xmath2@xmath1 ) and warrant further study .
our extragalactic cross correlation is performed in a similar manner although the problems associated with correlating galaxies are somewhat different .
proper motion is not a problem although the extended size of galaxies is as the peak mid - infrared position can vary from the peak optical position and therefore some positional uncertainty must be included when trying to correlate an infrared galaxy to an optical source .
although a 5@xmath12 positional uncertainty can correspond to 2 if the galaxy lies along the major axis of the iras error ellipse it can also correspond to just 10 " if the galaxy lies perpendicular to this direction .
therefore we only consider a galaxy cross correlated if it falls within 10@xmath12 of an iras source .
due to the often unknown incompleteness of extragalactic catalogues it is not possible to accurately determine the completeness of our sample in extragalactic catalogues and therefore we have simply used the largest ( and most appropriate ) databases . in the cross correlation presented here
we have used the qigc sample [ 17 ] , which has just become available in electronic form , ned , simbad and the fsc 25 @xmath0 m sample [ 18 ] . using these databases
we find that over 700 of our sources , not confirmed as stars , fall within our 10@xmath12 threshold . from these confirmed galaxies we find @xmath950 are elliptical or s0 systems ( not all are active galaxies ) and @xmath9150 have @xmath1@xmath2@xmath13 and are most probably pdr dominated galaxies .
a number of galaxies have @xmath1@xmath22@xmath3 and would therefore not be picked up by rms .
one of these galaxies is ngc 3115 , a nearby bulge dominated galaxy of hubble type s0 .
this galaxy is only detected at 12 @xmath0 m , the 25 , 60 and 100 @xmath0 m fluxes are upper limits , although with a v band mag of @xmath98.9 it is bright optical galaxy . in terms of it
b - v ( 1.0 ) and b/12 ( 2.5 ) colours it corresponds to a k0iii star . as a comparison we have plotted a sample of our _ normal _ infrared galaxies , see figure 3 .
although our initial cross correlation has been successful we have @xmath9500 objects for which we do not have an optical identification .
these objects do not have iras cirrus / confused flags or low 12 @xmath0 m fluxes although approximately 75% have upper limit 60 @xmath0 m iras fluxes .
visual inspection of a number of these objects with the digital sky survey shows them to be nearby bright stars , suggesting incompleteness in the gsc at bright fluxes .
however there is also probably a substantial population of stars not yet accounted for : those with v@xmath215 mags ( e.g. m7iii and m8iii stars ) , dark molecular clouds and planetary nebulae .
we are currently compiling a list of additional stellar objects to cross correlate to our sample to allow us to produce a definitive list of unidentified extragalactic sources .
our primary aim is to construct complete 12 @xmath0 m fsc selected samples of galaxies and stars . from this we will create accurate extragalactic source counts and classified luminosity functions with optical slit spectroscopy . as many of our extragalactic objects will be extended we also intend to obtain integrated spectra to provide a classification which is compatible with the distant objects found in iso surveys where the observed slit spectra will be produced by the majority of the galaxy . with our stellar sample
we wish to create a complete list of high galactic latitude stars and sources to help constrain galactic models and provide further diagnostics in distinguishing stars from galaxies in faint surveys ( e.g. the hubble deep field [ 3 ] ) .
both our galactic and stellar samples show objects that deviate from the norm ( i.e. galaxies with stellar colours and stars with galactic colours ) and warrant further study in their own right .
+ + * acknowledgements * we thank i.matute for assistance in producing the initial catalogue and m.moshir for fruitful discussions .
we are grateful to o.laurent and h.roussel for providing the isocam fluxes for some of the rms sample sources .
we acknowledge and thank the ec tmr extragalactic networks for postdoctoral grant support . this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propoulsion laboratory , california institute of technology , under contact with nasa .
saunders , w. , oliver , s. , keeble , o. , rowan - robinson , m. , canavezes , a. , et al .
( 1997 ) in maddox , s. , aragon - salamanca , a. , wide field spectroscopy and the distant universe , world scientific press , singapore | we present new techniques to produce iras 12 @xmath0 m samples of galaxies and stars .
we show that previous iras 12 @xmath0 m samples are incompatible for detailed comparison with iso surveys and review their problems . we provide a stellar infrared diagnostic diagram to distinguish galaxies from stars without using longer wavelength iras colour criteria and produce complete 12 @xmath0 m samples of galaxies and stars .
this new technique allows us to estimate the contribution of non - dusty galaxies to the iras 12 @xmath0 m counts and produce a true local mid - infrared extragalactic sample compatible with iso surveys .
we present our initial analysis and results . |
observations show that the fraction of potential binary open clusters is not negligible .
this finding is not surprising as open clusters are born in star complexes ( efremov 1978 , 1995 , 2010 ) with several of these objects being formed nearly at the same time and in close proximity ( for a recent review on this topic see , e.g. , de la fuente marcos & de la fuente marcos 2008 , 2009a ) . an early attempt to study this topic by rozhavskii et al .
( 1976 ) concluded that the fraction of multiple systems among open clusters was @xmath020% .
almost two decades later and using a larger sample , subramaniam et al .
( 1995 ) found that about 8% of open clusters may be genuine binaries .
loktin ( 1997 ) further strengthened the idea that multiple open clusters are not uncommon by providing a catalogue of 31 probable multiple systems .
the question of the possible existence of a sizeable fraction of candidate binary clusters in the galactic disk has been revisited again by de la fuente marcos & de la fuente marcos ( 2009b ) . using complete , volume - limited samples , they have found consistent and statistically robust figures : at the solar circle , at least 12% of all open clusters could be part of potential binary systems .
this number agrees well with the ones previously found by various authors for the magellanic clouds ( see dieball et al .
2002 for details ) .
the cluster binary fraction is not only numerically similar ; both in the clouds and milky way the pair separation histogram shows a conspicuous bimodal distribution .
yet another common characteristic is the practical absence of close , almost coeval pairs among objects older than about 100 myr .
all this body of solid observational evidence strongly indicates that , if primordial binary star clusters do form , they appear not to be able to survive as such for long .
obvious evolutionary paths include merging , tidal disruption and ionization but , under what conditions merging is favored ? what is the characteristic merging timescale ?
what is the ( dynamic / kinematic ) signature of an open cluster formed by merging ?
how does tidal disruption proceed ?
what happens to ionized primordial pairs ? is the initial number of stars in the cluster a factor to consider in the final outcome ?
can we recover the initial dynamical properties of primordial pairs ?
are long term stable binary clusters possible at all ? in this work
we examine the evolution of simulated primordial binary clusters in an attempt to provide convincing answers to the above questions .
this paper is organized as follows : in 2 we present the simulations .
we show our results in 3 . in
4 we discuss our results and our conclusions are presented in 5 .
the goal of these simulations is to understand how primordial binary clusters evolve , how they merge when they do , and how the resulting merger remnants look like .
we use the direct @xmath1-body code nbody6.gpu ( aarseth & nitadori 2009 ) .
this @xmath1-body code is an improved parallelized version of the standard scalar code nbody6 ( aarseth 2003 ) for use with multi - core machines and nvidia cuda - capable devices .
this code has been recently applied to study the dynamics of globular clusters ( heggie & giersz 2009 ; trenti et al .
2010 ) and it is publicly available from the ioa web site.sverre/web/pages/nbody.htm ] the initial conditions for each simulation ( see details in table [ results ] ) were generated by randomly selecting stellar positions and velocities according to the plummer distribution ( plummer 1911 ) .
the plummer density profile is given by @xmath2 where @xmath3 is the radial coordinate , @xmath4 is the total mass of the cluster and @xmath5 is the plummer radius , a scale parameter which sets the characteristic size of the central regions of the cluster .
it is related to the half - mass radius by @xmath6 ( aarseth & fall 1980 ; heggie & hut 2003 ) . for details on how to set up a numerical plummer model
see kroupa ( 2008 ) , also aarseth et al .
( 1974 ) . in the simulations , the actual input parameter that defines the initial physical size of the cluster is the virial radius , given by @xmath7 where @xmath8 is
the gravitational constant and @xmath9 is the total energy of the system . for a plummer model
the potential energy is ( e.g. , binney & tremaine 2008 ) @xmath10 and @xmath9 = @xmath11/2 .
therefore , @xmath12 each simulation follows the evolution of two plummer models in an initially bound orbit .
input parameters are the initial separation ( apoclustron distance , @xmath13 = 10 , 20 , and 30 pc ) and orbital eccentricity ( @xmath14 = 0.0 , 0.2 , 0.4 , 0.6 , 0.8 , and 0.95 ) .
the apoclustron is function of both semi - major axis and eccentricity , @xmath15 . therefore and if we assume that the two clusters are point - like masses moving in keplerian orbits , the initial orbital period of the system can be approximated by @xmath16 where @xmath17 and @xmath18 are the cluster masses , the apoclustron is in pc and the masses in @xmath19 ( see actual values in table [ results ] ) .
stellar masses in the range [ 0.17 , 10.0 ] @xmath19 are drawn from a salpeterian imf with an average stellar mass of 0.5 @xmath19 .
salpeter ( 1955 ) used the observed luminosity function for the solar neighborhood and theoretical evolution times to derive an initial mass function ( imf ) which may be approximated by a power - law : @xmath20 where @xmath21 is the number of stars per unit mass interval .
the value of @xmath22 is 2.35 for masses between 0.4 and 10.0 @xmath19 . the imf used ( a single power - law ) and the mass range of stars ( [ 0.17 , 10.0 ] @xmath19 ) are equivalent to the realistic canonical two - part power - law over the mass range [ 0.08 , 10.0 ] @xmath19 as described in ( e.g. ) kroupa ( 2008 ) .
neglecting stars more massive than 10 @xmath19 is a good approximation since such stars are rare ( see , e.g. , the @xmath23-star cluster mass relation of weidner et al .
all stars are started on a zero - age main sequence with a uniform composition of hydrogen , @xmath24 = 0.7 , helium , @xmath25 = 0.28 , and metallicity , @xmath26 = 0.02 .
stellar evolution is computed according to the algorithm described by aarseth ( 2003 ) .
primordial binaries were not included but binary and multiple system formation was allowed and observed .
external perturbations were represented by a fixed galactic tidal field .
the tidal radius is given by the expression ( e.g. , aarseth 2003 ) @xmath27 where @xmath28 and @xmath29 are the oort s constants of galactic rotation .
for a star located at that distance from the cluster , the central attraction of the cluster is balanced by the galactic tidal force .
the oort s constants are chosen to be @xmath28 = 14.4 , @xmath29 = -12.0 km s@xmath30 kpc@xmath30 ( binney & tremaine 2008 ) . the cluster pair is assumed to move in a circular orbit at the solar circle with no passing molecular clouds ( see aarseth 2003 , pp .
127 - 129 , for additional details ) .
no escaping stars were removed from the calculations .
the simulations presented here have been performed on a dell precision t5500 + nvidia tesla s1070 system .
the t5500 has 2 quad - core intel xeon e5540 processors at 2.53 ghz , the s1070 includes 4 gpu cores each with 240 stream processors running at 1.44 ghz for a total of 960 processing units .
the tesla parallel computing architecture is relatively new ( lindholm et al . 2008 ) but it is already in use for astrophysical applications ( e.g. thompson et al . 2010 ; jonsson & primack 2010 ; schive et al . 2010 ; trenti et al .
here we describe the main results for the four different sets of models : 18 models with two equal clusters ( @xmath31 = @xmath32 = 4096 , @xmath33 , @xmath34 , @xmath35 = 1 ) , 36 models with @xmath31 = 4096 , @xmath32 = 2048 , @xmath35 = 0.5 , 18 models with @xmath31 = 4096 , @xmath32 = 1024 , @xmath35 = 0.25 , and 4 additional control models for single clusters submitted to the same tidal field but with @xmath1 = 8192 , 6144 , 5120 , and 4096 , respectively .
calculations were stopped after a simulated time of 210 myr or one galactic rotation at the solar circle ; therefore , all the most interesting phases of the binary cluster evolution could be studied .
the state of the system at the end of the simulation is summarized in table [ results ] .
the merging timescale has been computed by assuming that a merger takes place if the pair separation becomes less than the average value of the core radius for observed open clusters .
the average value of core , half - mass , and tidal radii for real open clusters are 1/2/10 pc ( binney & tremaine 2008 ) , respectively .
this rather conservative merging criterion may appear arbitrary but it responds to what is observed in our calculations : if the pair separation goes under 1 pc , the two clusters quickly merge ( see figure [ trajectory ] ) . from a strictly observational point of view
, star clusters separated by less than twice the core radius are unlikely to be identified as separated objects ; therefore , 2 pc could also be a valid ( and less conservative ) observational criterion for merging .
figure [ evolequal ] shows the evolution of the orbital separation for models with two equal orbiting clusters and different initial apoclustron and eccentricity .
the initial virial radius for the two plummer models is 1 pc with a mean density of 489 @xmath19 pc@xmath37 .
our numerical results show that two outcomes are possible : merging or a special type of soft ionization . long term binary stability is not observed .
merging is found in all initially close pairs .
less eccentric pairs take longer to merge , with the circular case taking almost four full initial orbital periods to complete the merging process . for wider pairs ,
a consistent behavior is also found ; clusters gradually increase their separation with the rate of increase slowing down over time .
this rate depends on the initial value of the eccentricity of the pair and it is slower for higher eccentricity .
an example of merging appears in figures [ merging ] , [ trajectory ] ( model with two clusters in an initially circular orbit and apoclustron 10 pc ) and the outcome of ionization after 210 myr is displayed in figure [ separatedtwins ] ( this figure is for a non - equal mass case but the outcome is similar for the equal @xmath1 case ) .
the ionized clusters that we call `` separated twins '' , after theis ( 2001 ) , are not completely unrelated as a rather conspicuous tidal bridge is always observed .
formation of separated twins is driven by mass loss induced by stellar evolution , two - body relaxation , and ( less important ) mutual tidal disruption . in relative terms , separated twins retain a proportionally higher fraction of stars , which is not surprising : following innanen et al .
( 1972 ) , for two clusters separated by a distance larger than three times the outer radius of each cluster , the amount of mutual disruption is rather negligible .
mergers are less rich in relative terms , as they have lost a fraction of their original stars due to mutual disruption and instabilities during the merging process .
the longer the merging process takes , the larger the amount of mutual disruption . in the @xmath38 plane
, merger remnants appear significantly more elliptical than their isolated counterparts ( see figure [ merging ] ) .
the merger process imparts rotation to the resulting stellar system in the same sense as the orbital motion ( alladin et al .
1985 ) due to a tidal torque .
young candidate binary open clusters appear to show a clear trend ; in general , the two members of the pair have rather different radii ( de la fuente marcos & de la fuente marcos 2009c ) .
this could be a property intrinsic to open cluster formation in complexes , the result of dynamical interaction , or caused by an observational selection effect .
if intrinsic to open cluster formation , primordial binary clusters may ordinarily sport a massive primary and a less massive secondary .
on the other hand , not having the same initial population is likely to have an impact on the tidal evolution of the pair .
this group of simulations is aimed at studying this case .
the primary cluster is as described in the previous section . here
we explore the evolution of primordial cluster pairs in which the secondary of the system has half the population of the primary , @xmath40 .
two different sub - cases are investigated : i ) both clusters have the same stellar density and ii ) the secondary is half the size of the primary , i.e. the secondary is denser than the primary .
the secondary cluster has a virial radius of 0.79 pc so the mean density is the same for both primary and secondary clusters .
the behavior of the orbital separation is similar to that found in the previous section but the soft ionization process is slightly slower ( see figure [ evolhalfsd ] ) .
the merging timescale is also affected : mergers for models with @xmath41 0.5 take longer now but those for less eccentric pairs are faster .
merging is only observed for models with an initial apoclustron distance of 10.0 pc .
cores of merger remnants are more extended than those observed in the previous case .
the secondary cluster has a virial radius of 0.5 pc with a mean density of 1956 @xmath19 pc@xmath37 .
the behavior of the orbital separation is similar to that found in the previous section ( see figure [ evolhalfdd ] ) .
this is to be expected : a study by sensui et al .
( 2000 ) showed that the internal structure of galaxies does not play a role in the merging time - scales ( inside a galaxy cluster ) only the distribution of galaxies inside the cluster matters .
this argument should also hold for star clusters in a star cluster complex ( fellhauer et al .
2002 , 2009 ) .
the single main difference appears for highly eccentric models in which merging is observed in two cases ( as in @xmath31 = @xmath32 ) : @xmath13 = 10.0 and 20.0 pc .
this set of simulations is designed to study the impact of enhanced tidal forces on the secondary cluster . with a virial radius of 0.63 pc
, it still has the same reference mean density used throughout this study .
the dynamical behavior of the pair is now substantially different ( see figure [ evolfourth ] ) .
the tidal disruption timescale is much shorter than that for merging .
eventual destruction of the less massive cluster is observed in all cases , including close pairs . in some cases ,
the secondary cluster appears extremely distorted and elongated with no clearly identifiable core ( see figure [ spaghetti ] ) : in other words , the secondary cluster gets torn apart in a relatively short timescale . technically speaking , shredded clusters are different from typical open cluster remnants , they look more like stellar streams and they may be rather young .
remnants of shredded clusters may show up in kinematic studies even if they can not be detected as stellar overdensities .
tidal shredding is a form of shearing by differential rotation .
the secondary cluster in proximity to the most massive primary becomes stretched out by tidal forces .
the secondary distends and flattens in the direction of the primary evolving as to minimize its gravitational potential energy becoming an ovoid stretched along the axis connecting the two bodies .
figure [ earlytidal ] shows the early evolution of the model displayed in figure [ spaghetti ] .
the secondary cluster undergoes early rapid expansion mainly due to stellar mass loss .
in addition , the characteristic timescale for equipartition of kinetic energy is much shorter for the secondary cluster contributing to its overall faster internal dynamical evolution .
coupling of these two processes effectively drives the cluster away from equilibrium as the total potential energy per unit mass quickly decreases .
this behavior is , in principle , similar to the one described in portegies zwart & rusli ( 2007 ) : the secondary expands quickly initiating mass transfer to the primary .
there is , however , a major difference : extreme tidal distortion is not observed in their models ( see comments below ) . in our calculations and as the massive companion s
gravitational pull is stronger on the overextended secondary cluster s near side than on the far side , the secondary cluster is literally shredded with the help of the galactic tidal field . for the purpose of comparison
, we computed four single plummer models with consistent tidal field : the virial radii were 1.26 , 1.14 , 1.08 and 1.00 for a mean density of 489 @xmath19 pc@xmath37 .
the rest of the details of the simulation are common to the binary simulations .
after 210 myr the values of the core , half - mass , and tidal radii were 0.095/1.92/18.0 pc ( @xmath1 = 8192 ) , 0.106/2.23/18.0 pc ( @xmath1 = 6144 ) , 0.37/2.62/18.0 pc ( @xmath1 = 5120 ) and 0.34/2.61/18.0 pc ( @xmath1 = 4096 ) .
the average values of the same magnitudes for the merged clusters were : 0.29/4.17/22.6 pc ( @xmath35 = 1.0 ) and 4.35/6.37/20.5 pc ( @xmath35 = 0.5 ) .
the central regions of merger remnants are significantly more extended than those of equivalent single models .
merger remnants are less populated than an equivalent single cluster of the same age as a result of mutual disruption of the pair during the merging process . typical densities of candidate bound open clusters younger than 13 myr are in the range 4 - 400 @xmath42 ( wolff et al .
2007 ) with the upper limit represented by @xmath43 and @xmath44 persei .
slesnick et al .
( 2002 ) estimated that the total mass of the stars with @xmath45 in @xmath44 and @xmath43 persei is 3700 and 2800 @xmath19 , respectively , although these are , probably , lower limits for the masses of these clusters .
therefore , the characteristics of the simulated clusters match well those of the most massive classical open clusters .
these are the ones expected to be able to survive for several galactic rotations and , therefore , statistically speaking more likely to be observed .
their values are , however , far from those typical of starburst clusters with densities in the range @xmath46 to several 10@xmath47 @xmath42 and @xmath1 = several 10@xmath48 - 10@xmath49 ( see , e.g. , pfalzner 2009 ) .
similar trends have been found for star clusters in other galaxies ( pfalzner & eckart 2009 ) .
the above results can only be properly understood within the context of the galactic tidal field ; i.e. cluster tidal radii and separations play a major role in the outcome of primordial binary cluster evolution .
the strength of the cluster - cluster interaction is maximum when the intercluster separation becomes smaller than the tidal radii .
this interpretation was proposed in de la fuente marcos & de la fuente marcos ( 2009c ) in the form of a classification scheme ( [ bincri ] ) based only in the values of separation ( @xmath50 ) and tidal radii ( @xmath51 , @xmath52 = 1 , 2 ) : @xmath53 the scheme was implemented by considering the available observational evidence . for open cluster pairs of similar mass ratio in the detached and weakly interacting categories , ionization into the separated twins state
is the observed evolutionary path .
semi - detached and in - contact pairs merge in a timescale that depends strongly on the orbital eccentricity .
very eccentric pairs merge in a timescale of nearly 10 myr and , therefore , the actual observation of the merging process may be very difficult as it may happen even before the embedded phase ends . if the mass ratio is appreciably different , extreme tidal distortion followed by actual destruction of the smallest cluster is always observed .
the above interpretation has strong implications on what should be observed within the milky way at different galactocentric distances and in other galaxies .
weaker galactic tidal fields increase the probability of observing semi - detached or in - contact cluster pairs and , eventually , mergers .
the tidal force gradient determines the cluster tidal radius ( e.g. elmegreen & hunter 2010 ) .
even if the fraction of primordial binary clusters may well be similar across different galaxies , the preferential evolutionary path could be rather different : ionization being dominant in massive galaxies and the central regions of galaxies in general .
figures [ evolequal ] , [ evolhalfsd ] , [ evolhalfdd ] and [ evolfourth ] , also show probable primordial pairs ( age difference @xmath5430 myr ) with pair age @xmath54 210 myr .
following de la fuente marcos & de la fuente marcos ( 2009b ) , the pair age is assumed to be that of the youngest cluster in the pair .
the agreement is very significant .
most pairs , if primordial , appear to be evolving towards the separated twins state . as expected , the observed number of close pairs is consistent with short merging timescale .
only seven pairs in the figures ( see also figure [ realmerg ] ) appear to be in a bound state : ngc 1976/ngc 1981 , ascc 20/ascc 16 , collinder 197/ascc 50 , ngc 6250/lynga 14 , ngc 3324/ngc 3293 , ngc 6613/ngc 6618 , and trumpler 22/ngc 5617 . another candidate ( not in the figure )
to bound pair is loden 1171/loden 1194 ( see de la fuente marcos & de la fuente marcos 2009b for details ) .
both trumpler 22/ngc 5617 and loden 1171/loden 1194 seem to be the evolved state of primordial pairs with almost circular initial orbits and original separation @xmath54 20 pc .
the possible triple cluster ngc 1981/ngc 1976/collinder 70 is a singular object with the inner pair probably undergoing merging .
some of these binary candidates are displayed in figure [ real ] .
out of the clusters in this figure , the pairs ngc 3324/ngc 3293 and trumpler 22/ngc 5617 may be actual bound clusters .
in contrast , the pair ngc 659/ngc 663 is likely evolving into the separated twins state .
another example of bound clusters could be the cluster pair ngc 3590/hogg 12 .
piatti et al .
( 2010 ) have pointed out this pair as a strong open cluster binary system candidate .
both clusters have similar ages ( 30 myr ) , reddenings and metallicities .
they appear to be located at 2 kpc from the sun and at that distance the pair separation is a mere 3.6 pc . in this case
( see figure [ real2 ] ) , one of the clusters ( hogg 12 ) seems to be undergoing extreme tidal distortion .
a dramatic example of ongoing merger candidate is the case of the partially embedded massive young cluster ngc 2244 described by li ( 2005 ) where two structures may be separated by just @xmath0 7 pc .
regarding the remarkable double - peaked pair separation distribution observed for young open cluster pairs ( de la fuente marcos & de la fuente marcos 2009b ) , the vast majority of observed candidates evolve into separated twins which translates into the dominant peak ( 25 - 30 pc ) in the pair separation distribution .
the secondary peak at smaller separations ( 10 - 15 pc ) can be explained as due to close pairs in almost circular orbits or undergoing merging .
the average tidal radius of open clusters at the solar circle is nearly 20 pc .
therefore , cluster pairs with original separations smaller than the cluster tidal radius are likely to merge but those born with wider separations are bound to evolve into the separated twins state .
our calculations show that only equal - mass pairs formed with originally small separations and nearly circular orbits are likely to be observed in close proximity for more than 100 myr and actually pose as genuine binary open clusters . in summary ,
the bimodal distribution observed in the pair separation histogram is the result of evolutionary effects and not of different formation channels .
separated twins are , in principle , easy to identify ; just think about the double cluster in perseus ( see caption in figure [ separatedtwins ] ) .
but , what about merger remnants ? one easy to implement test is based on star counts .
figure [ radialprofile ] shows the surface number density profile for models with @xmath35 = 1 at 210 myr . two single models with @xmath1 = 8192 and 4096 evolved in a similar tidal field and a king profile ( king 1962 ) are also included as reference .
the outer regions of merger remnants are similar to those of an equivalent single cluster having twice the population of the individual clusters but the number density of the inner regions is rather different . in general ,
merger remnants are expected to be fainter .
however , some models ending in merger show clear central cusps which are absent from single models .
merger models characterized by higher central concentration are preferentially associated to pairs with high initial eccentricity and , therefore , shorter merger timescale .
they are also the pairs in which the mutual tidal disruption has been the weakest as they have been interacting for a shorter period of time .
this explains why their cores are nearly 25 - 50% denser than that of an equivalent king model .
in contrast , mergers of low eccentricity pairs have lower central densities .
this is the result of longer merger timescale and enhanced mutual tidal disruption .
this lower density translates into higher production of hierarchical systems in the form of temporarily stable triple and quadruple stellar systems with respect to single models .
this is to be expected as higher stellar density increases the probability of relatively close - range gravitational interactions that quickly destroy multiple systems . in general ,
the central regions of merger remnants are distinctively different from those of clusters evolving without companions but haloes look very similar . regarding the kinematic signature of merging , it is also obvious in the central regions .
data in the figures discussed here are referred to the center of masses of the merged / single cluster . in this analysis
we use the root mean square velocity defined as the square root of the average velocity - squared of the stars in the cluster , their radial velocity and transverse velocity .
the @xmath55 is always greater than or equal to the average as it includes the standard deviation ( @xmath56 ) .
the radial velocity is given by @xmath57 where @xmath58 is the position of the star and @xmath59 its velocity .
finally , the transverse velocity is defined by @xmath60 the transverse velocity is also the product of the angular speed @xmath61 and the magnitude of the position vector . therefore and for a given star cluster region , transverse velocity and angular speed exhibit the same behavior : both increase or decrease concurrently .
the modulus of the specific angular momentum is the magnitude of the position vector times the transverse velocity .
figure [ vrms ] shows the rms velocity for concentric shells similar to those in figure [ radialprofile ] for merged models and the two reference single models already presented .
the value has been summed over all the stars in the shell considered and divided by the number of particles inside that shell . in general , the outcome of a typical merged model shows a value of the rms velocity in the core intermediate to those of the reference models and it gets closer to that of the most massive reference model in the outer parts of the cluster .
the behavior beyond the half - mass or effective radius is similar for all the models , the rms velocity decreases gradually outwards .
it is , however , not so clear how this peculiar velocity distribution can be used to identify merger remnants .
the behavior of the radial velocity profile in figure [ vr ] is , however , easier to interpret . in mergers ,
radial velocities tend to be lower or even negative in the central regions of the cluster ( see fig .
3 , top panel in baumgardt et al . 2003 ) .
this trend is observed in several of the studied cases and it may point out to collapse of the central regions induced by the gravogyro instability ( see below ) or to statistical fluctuations .
the transverse velocity profile is displayed in figure [ vt ] . in general and
up to 10 pc from the center of the merger these profiles are very smooth .
the value of the transverse velocity in the central pc of the merger remains remarkably constant in clear contrast with what is observed for single cluster models .
this can be interpreted as evidence for a moderate amount of global rotation . on theoretical grounds ( king 1961 )
, rotation causes a slight increase in the rate at which a cluster is losing stars .
this is consistent with the relatively smaller final population observed in merged models .
single self - gravitating and globally rotating @xmath1-body systems are affected by a little known instability that is playing a role here : the gravogyro instability .
initially proposed by inagaki & hachisu ( 1978 ) and hachisu ( 1979 , 1982 ) and further studied by akiyama & sugimoto ( 1989 ) using numerical techniques , the gravogyro instability or gravogyro catastrophe is triggered when specific angular momentum is removed ( by escaping stars ) from the cluster and a deficit in the supporting centrifugal force is induced .
in other words , the escape of stars from the core to the outer regions transports angular momentum from the inner regions of the cluster to the outer parts . as a result ,
the inner regions react to compensate the loss of centrifugal force contracting in order to increase rotation .
faster rotation induces additional mass loss .
the gravogyro effect increases the average angular speed of the central regions of the cluster decreasing the average value of the local angular momentum .
the overall result is faster dynamical evolution of the star cluster and higher mass loss .
the gravogyro instability leads to further mass loss through the galactic tidal boundary .
this process causes the contraction of the core of the star cluster as observed in figure [ radialprofile ] ( models with @xmath14 = 0.2 and 0.6 , mainly ) .
but before the onset of the gravogyro instability , a merger must take place in order to have a rotating system .
this is the result of yet another instability : the synchronization instability .
the numerical / theoretical work of sugimoto & makino ( 1989 ) and sugimoto et al . ( 1991 ) showed that binary star clusters undergo an instability at a critical separation and this process is able to trigger rapid merging of star clusters .
the orbit of the bound cluster pair circularizes and shrinks due to the loss of angular momentum carried away by escaping particles . as the separation of the binary system decreases , the orbital period of the binary becomes shorter and the mutual tidal effects stronger .
concurrently , the spin of each cluster increases following the orbital motion due to the induced tidal torque attempting to synchronize cluster rotation and orbital period .
this transformation of orbital angular momentum into rotational angular momentum leads to an instability at a critical separation when the exchanged orbital angular momentum becomes insufficient to supply the necessary angular momentum for spin synchronization . at that moment
merging takes place as the orbital angular momentum is no longer capable of balancing the gravitational attraction between the two star clusters .
figures [ evolequal ] , [ evolhalfsd ] and [ evolhalfdd ] clearly indicate that in our models and after the pair separation becomes @xmath54 2 pc , merging proceeds very quickly .
the gyration radius of a self - gravitating system of stars can to first - order be approximated by @xmath62 ( e.g. , portegies zwart & rusli 2007 ) ; therefore , merger occurs when the intercluster separation falls below twice the gyration radius .
this process is similar to the tidal locking effect observed in planetary systems .
the main difference here is that in planet - satellite interactions both conversion of orbital angular momentum into spin and vice versa are possible . on the other hand and in absence of dissipation ( drag forces ) , the outcome of tidal locking is always a stable binary configuration not a merger as in the case of star clusters .
the main reason for the lack of stable outcome in the case of star clusters is that the total angular momentum is not actually preserved as stellar evolution , rotation - induced escapees ( those resulting from the gravogyro effect ) , and two - body relaxation are concurrently and constantly removing angular momentum from the system .
the total angular momentum removal rate depends on many factors , the age of the cluster being one of them .
other facts to consider here are the masses of the clusters , the mass ratio , and the galactic tidal field .
the gravogyro and synchronization instabilities are somehow related as they involve rotation and transfer of angular momentum but they are not the same process . for originally
non - rotating , merging clusters , synchronization occurs first and then , eventually and after global cluster rotation is induced , the gravogyro instability starts on the resulting merger .
the final phases of the synchronization instability and the onset of the gravogyro instability overlap ; in fact , and as described above , the enhanced escape rate caused by the gravogyro instability helps to deplete the remaining orbital angular momentum .
the most dramatic example is observed for the slowest merging model , the one with @xmath14 = 0.0 .
single star clusters formed out of unstable disks might exhibit some degree of primordial rotation .
ernst et al .
( 2007 ) studied @xmath1-body models of rotating globular clusters to confirm that the gravogyro instability takes place in monocomponent clusters ( clusters with equal - mass stars ) .
they found that the @xmath63-component of the specific angular momentum decreases over time for the inner regions of the clusters increasing , in return , the average angular speed of the same areas ( and therefore the transverse velocity ) ; see their figs . 3 and 4 .
for these models the effect is very important with a significant amount of angular momentum being transfered outwards .
the induced deficit of centrifugal force triggers the collapse of the core of the star cluster .
if the galactic tidal field is included in the calculations , rotation increases the escape rate dramatically . in sharp contrast , their models of systems with two - mass components ( two mass groups for a very simple imf ) show that the effect of rotation is rather negligible . in this case and
if rotation is primordial , two concurrent processes are at work : mass segregation and rotation .
both compete to accelerate the collapse of the cluster core but mass segregation clearly dominates . in summary
, they found that if energy equipartition is at work the role of rotation is somewhat secondary but in absence of energy equipartition , rotation through the gravogyro effect speeds up and controls the evolution of the cluster .
this is what we observe in our models .
figures [ vtevol ] and [ vtevol2 ] show the evolution of the average value of the transverse velocity for relevant concentric cluster shells ( see the caption for details ) of two representative models : the ones with the longest and shortest , respectively , merging timescale . by the time ( @xmath0 150 myr )
the clusters in the slowest merging model actually merge , energy equipartition has been achieved and the gravogyro effect is the strongest .
conversely , the fastest merging model shows that rotation is quickly lost from the merger remnant through energy equipartition ; the time evolution of the average transverse velocity being similar to that of a single cluster model with @xmath1 = 4096 . as for the time evolution of the @xmath63-component of the specific angular momentum in our models ,
representative results are displayed in figs .
[ jzevol ] and [ jzevol2 ] . for the slowest merging model and after merging , systematic loss of angular momentum is observed with the outer regions sporting the largest share of angular momentum . in the case of an early merger ,
angular momentum also decreases over time but the magnitude of this angular momentum is much higher and positive as expected for a rotating system .
unfortunately , little attention has been devoted to the topic of binary cluster evolution . in a ground breaking paper , sugimoto & makino ( 1989 ) used @xmath1-body simulations to study the merging timescale of two identical star clusters ( @xmath1 = 2048 ) .
they found that the interaction provoked circularization of the orbits of the clusters and synchronization of the spin of each cluster with its orbital revolution .
the loss of orbital angular momentum caused the eventual demise of the pair .
the entire process was described by them as the synchronization instability .
this phenomenon is also responsible for merging in our models .
their work was continued in makino et al .
( 1991 ) for non - equal clusters .
merging of two stellar systems usually gives surface density profiles @xmath64 ( sugimoto & makino 1989 ; makino et al .
1990 ; okumura et al .
1991 ) . in our models
this is only true for the outskirts of merger remnants ; the central regions can be better described by @xmath65 or @xmath66 ( see figure [ radialprofile ] )
. using analytical arguments , ballabh & alladin ( 2000 ) showed that merging always occurs if the distance of closest approach is about twice the sum of the dynamical radii of the clusters .
this is confirmed by our calculations .
de oliveira et al .
( 2002 ) used @xmath1-body simulations to study the dynamical status of the cluster pair ngc 1912/ngc 1907 .
their simulations found the formation of stellar bridges similar to the ones found in our calculations .
faster encounters produce weaker tidal debris in the bridge area .
our simulations are also consistent with this result
. the most realistic simulations of binary cluster evolution so far have been performed by portegies zwart & rusli ( 2007 ) using the kira integrator of the starlab simulation environment .
their calculations were aimed at understanding the nature of the cluster pair ngc 2136/ngc 2137 , in the large magellanic cloud .
their models are , therefore , more massive ; their primary clusters have @xmath1 in the range [ 9,000 , 24,855 ] with smaller secondaries moving in circular orbits .
they found that cluster pairs with initial separation smaller than 15 - 20 pc merge in @xmath5460 myr .
pairs with larger initial separation tend to become even more widely separated over time . in spite of the different code and initial conditions used , our results are , in general , fully compatible with theirs .
there is , however , a major difference induced by the fact that they do not take into account the background galactic tidal field .
they did not observe the extreme tidal distortion displayed in figure [ spaghetti ] .
besides , they only follow their models for about 100 myr .
their main models have @xmath35 = 0.167 ; this value triggers catastrophic destruction in our models .
the soft ionization widely found in our simulations was originally described in a series of little known but very interesting papers ( theis 2001 , 2002a , b ) . within the context of globular cluster formation
, theis simulations found that identical twin clusters may merge or evolve into well separated twins sharing a common galactic orbit .
our calculations clearly show that the impact of merging on the evolution of close and therefore young open cluster pairs is all but negligible .
the same can likely be said about young star cluster pairs in other galaxies .
on the other hand , the rapid decrease in star cluster numbers for ages older than 20 myr ( _ infant mortality _ ) observed in our galaxy and others and usually attributed exclusively to the catastrophic gas ejection mechanism originally proposed by hills ( 1980 ) can also be the result of merging or tidal disruption in close primordial pairs .
merging and tidal disruption of the less massive companion may easily halve ( at least ) the initial population of relatively close open cluster pairs . merging , disruption and infant mortality , concurrently , can efficiently reduce the number of observed young star clusters and accelerate dramatically the transition of stars born in clusters to the field populations .
it may be argued that the loss of clusters through merging can only be significant if the fraction of binary clusters at birth is high , but is the primordial semi - major distance distribution supportive of this ? if the orion nebula star forming complex could be considered as representative of the kind of environment in which most open clusters actually form , the answer may be in the affirmative .
current available evidence ( de la fuente marcos & de la fuente marcos 2009b ) suggests that the group ngc 1981/ngc 1976/ collinder 70/@xmath67 ori may form several close pairs , all of them with separations @xmath54 30 pc and few myr old .
there is also a statistical tool that may help to understand the timeline of the relative importance of these processes : the generalized fractal dimension or multifractal analysis .
de la fuente marcos & de la fuente marcos ( 2006 ) showed that the generalized fractal dimension changes very significantly for clusters younger than 40 myr but remains almost constant for older clusters .
catastrophic gas ejection is an essentially self - driven process and operates on all scales ; merging or extreme tidal distortion only operates on small length - scales within star - forming complexes where the typical intercluster distance is @xmath54 30 pc .
processes operating on all scales keep the value of the generalized fractal dimension almost constant across the multifractal spectrum .
the opposite is observed when processes operate only for objects in close proximity .
our numerical results are consistent with the fractal dimension results if most pairs are born in originally very eccentric orbits ( @xmath68 0.5 ) and/or with very different masses .
those are the ones merging ( or being tidally destroyed ) within 40 myr of forming .
using available observational evidence , de la fuente marcos & de la fuente marcos ( 2009b ) have demonstrated that the population of binary open clusters is statistically significant and that the fraction of candidate binary clusters in the milky way disk is comparable to that in the magellanic clouds , @xmath0 10% . out of this population ,
nearly 40% of them can be classified as genuine primordial binary open clusters although only a relatively small fraction ( @xmath017% ) appear to be able to survive as conspicuous pairs for more than 25 myr .
the distribution of open cluster separations exhibits an apparent peak at 10 - 15 pc , analogous to the one observed in both lmc and smc . here
, we have used @xmath1-body simulations in an attempt to understand how primordial binary open clusters evolve and what initial orbital elements are required in order to explain their observed properties .
vzquez et al .
( 2010 ) concluded that available observational evidence indicates that double cluster lifetimes are short and that is what simulations confirm .
our main conclusions can be summarized as follows . 1 .
long term stability of binary open clusters appears not to be possible .
primordial binary open clusters seem to be inherently transient objects , at least for the range of cluster parameters explored here .
the results of our simulations interpreted within the context of the available observational data clearly indicate that the vast majority of primordial binary open clusters gradually evolve into well separated objects .
after one galactic rotation the separation is in the range 200 - 500 pc but they exhibit relatively prominent tidal bridges .
formation of separated twins is driven by mass loss induced by stellar evolution , two - body relaxation , and mutual tidal disruption .
close primordial binary open clusters quickly merge into a single object .
the merging timescale depends on the orbital and physical characteristics of the pair .
most close pairs merge within 100 myr of formation . for eccentric pairs
the merging scale is even shorter , about 50 myr .
our numerical results and the short characteristic observational survival time for candidate primordial pairs , 25 myr , strongly suggest that nearly 80% of primordial binary open clusters are born in systems with orbital eccentricities @xmath69 0.5 .
merging is driven by the synchronization instability .
4 . for clusters
pairs of appreciably different masses extreme tidal distortion or shredding of the secondary is observed , causing the complete disruption of the less massive cluster within one galactic orbit .
remnants of shredded clusters may show up in kinematic studies even if they can not be detected as stellar overdensities .
the gravogyro instability shapes the spatial and kinematic structure of merger remnants but the effect is only dominant for primordial pairs in almost circular orbits .
the observed candidate pair separation histogram shows a bimodal distribution . in the light of the present results ,
the two peaks are mainly the result of ionization and merging .
they do not appear to be the result of different formation channels but different evolutionary paths .
the sharp decline in open cluster numbers observed for objects older than about 20 myr can be explained by three different processes operating concurrently : catastrophic gas expulsion , tidal disruption and merging .
this can efficiently reduce the number of observed young star clusters and accelerate dramatically the transition of stars born in clusters to the field populations .
it has often been assumed that binary open cluster formation in the milky way is uncommon .
in contrast , our results indicate that the lives of primordial binary clusters are violent and hazardous .
close pairs ( if formed ) merge in a short timescale , being the shortest for very eccentric pairs , secondaries in low mass ratio pairs are rapidly destroyed , and wide pairs quickly ionize in the background tidal field due to mass loss . as open clusters are actually born in close proximity ( complexes ) , these appear to be the genuine reasons behind the apparent lack of binary open clusters in our galaxy .
star cluster binarity is , therefore , a transient phenomenon .
the effects and trends observed in the present set of simulations are robust for the range of open cluster parameters studied . as usual , it is potentially dangerous to make unwarranted extrapolations to larger / smaller or denser clusters
. it could be possible that for much larger and denser star clusters the merging timescale is longer .
nevertheless , the absence of binary globular clusters in the milky way appears to indicate that long - term binary cluster stability is , in fact , unlikely .
the role of the gravogyro effect of the evolution of merger remnants appears to be well documented in our present work but larger simulations are needed to better understand the statistical strength of this process .
the authors would like to thank s. aarseth and k. nitadori for providing the code used in this research .
the authors also would like to acknowledge the help of j. l. mazo fras , n. pedone , and p. chan of dell computer with the t5500+s1070 system .
in preparation of this paper , we made use of the nasa astrophysics data system and the astro - ph e - print server .
this research has made use of the webda database operated at the institute of astronomy of the university of vienna , austria .
this work also made use of the aladin , simbad and vizier databases , operated at the cds , strasbourg , france .
aarseth , s.j .
2003 , gravitational @xmath1-body simulations ( cambridge : cambridge university press ) aarseth , s.j . , &
fall , s.m .
1980 , , 236 , 43 aarseth , s.j . , & nitadori , k. 2009 , nbody6.gpu , http://www.ast.cam.ac.uk/@xmath0sverre/web/pages/nbody.htm aarseth , s.j . ,
hnon , m. , & wielen , r. 1974 , , 37 , 183 akiyama , k. , & sugimoto , d. 1989 , pasj , 41 , 991 alladin , s.m . , ramamani , n. , & meinya singh , i. 1985 , j. astrophys .
, 6 , 5 ballabh , g.m . , & alladin , s.m .
2000 , basi , 28 , 261 baumgardt , h. , makino , j. , hut , p. , mcmillan , s. , & portegies zwart , s. 2003 , , 589 , l25 binney , j. , & tremaine , s. 2008 , galactic dynamics , 2nd ed .
( princeton : princeton university press ) dieball , a. , mller , h. , & grebel , e.k .
2002 , , 391 , 547 efremov , y.n . 1978 , sov .
lett . , 4 , 66 efremov , y.n .
1995 , , 110 , 2757 efremov , y.n .
2010 , , 405 , 1531 elmegreen , b.g . , & hunter , d.a .
2010 , , 712 , 604 ernst , a. , glaschke , p. , fiestas , j. , just , a. , & spurzem , r. 2007 , , 377 , 465 fellhauer , m. , wilkinson , m.i . , & kroupa , p. 2009
, , 397 , 954 fellhauer , m. , baumgardt , h. , kroupa , p. , & spurzem ,
r. 2002 , celest .
astron . , 82 , 113 de la fuente marcos , r. , & de la fuente marcos , c. 2006 , , 452 , 163 de la fuente marcos , r. , & de la fuente marcos , c. 2008 , , 672 , 342 de la fuente marcos , r. , & de la fuente marcos , c. 2009a , , 14 , 180 de la fuente marcos , r. , & de la fuente marcos , c. 2009b , , 500 , l13 de la fuente marcos , r. , & de la fuente marcos , c. 2009c , , 700 , 436 hachisu , i. 1979 , pasj , 31 , 523 hachisu , i. 1982 , pasj , 34 , 313 heggie , d.c . , & hut , p. 2003
, the gravitational million - body problem ( cambridge : cambridge university press ) heggie , d.c . , & giersz , m. 2009 , , 397 , l46 hills , j.g . 1980 , , 235 , 986 inagaki , s. , & hachisu , i. 1978 , pasj , 30 , 39 innanen , k.a . , wright , a.e . , house , f.c . , & keenan , d. 1972 , , 160 , 249 jonsson , p. , & primack , j.r
2010 , , 15 , 509 king , i. 1961 , , 66 , 68 king , i. 1962 , , 67 , 471 kroupa , p. 2008 , in lecture notes in physics 760 , the cambridge n - body lectures , eds .
aarseth , c.a .
tout , & r.a .
mardling ( berlin heidelberg , springer ) , 181 li , j.z .
2005 , , 625 , 242 lindholm , e. , nickolls , j. , oberman , s. , & montrym , j. 2008 , ieee micro , 28 , 39 loktin , a.v .
1997 , a&at , 14 , 181 makino , j. , akiyama , k. , & sugimoto , d. 1990 , pasj , 42 , 205 makino , j. , akiyama , k. , & sugimoto , d. 1991 , , 185 , 63 okumura , s.k . ,
ebisuzaki , t. , & makino , j. 1991 , pasj , 43 , 781 de oliveira , m.r . ,
fausti , a. , bica , e. , & dottori , h. 2002 , , 390 , 103 pfalzner , s. 2009 , , 498 , l37 pfalzner , s. , & eckart , a. 2009 , , 508 , l5 piatti , a.e . ,
clari , j.j . , & ahumada , a.u .
2010 , , 122 , 516 plummer , h.c . 1911 , , 71 , 460 portegies zwart , s.f . , &
rusli , s.p .
2007 , , 374 , 931 rozhavskii , f.g .
, kuzmina , v.a . , & vasilevskii , a.e .
1976 , astrophysics , 12 , 204 salpeter , e.e .
1955 , , 121 , 161 schive , h .- y . ,
tsai , y .- c .
, & chiueh , t. 2010 , , 186 , 457 sensui , t. , funato , y. , & makino , j. 2000 , pasj , 51 , 943 slesnick , c.l . ,
hillenbrand , l.a . , & massey , p. 2002
, , 576 , 880 subramaniam , a. , gorti , u. , sagar , r. , & bhatt , h.c .
1995 , , 302 , 86 sugimoto , d. , & makino , j. 1989 , pasj , 41 , 1117 sugimoto , d. , ebisuzaki , t. , & makino , j. 1991 , in iau symposium , no .
148 , the magellanic clouds , eds .
r. haynes , & d. milne ( dordrecht , kluwer ) , 217 theis , c. 2001 , in asp conf .
228 , dynamics of star clusters and the milky way , eds .
s. deiters , b. fuchs , a. just , r. spurzem , & r. wielen ( san francisco , ca : asp ) , 347 theis , c. 2002a , in iau symposium series , v. 207 , extragalactic star clusters , eds .
d. geisler , e.k .
grebel , & d. minniti , 681 theis , c. 2002b , , 281 , 97 thompson , a.c . , fluke , c.j . , barnes , d.g . , & barsdell , b.r .
2010 , , 15 , 16 trenti , m. , vesperini , e. , & pasquato , m. 2010 , , 708 , 1598 vzquez , r.a . , moitinho , a. , carraro , g. , & dias , w.s .
2010 , , 511 , a38 weidner , c. , kroupa , p. , & bonnell , i.a.d .
2010 , , 401 , 275 wolff , s.c .
, strom , s.e .
, dror , d. , & venn , k. 2007 , , 133 , 1092 plane ( top ) and the @xmath38 plane ( bottom ) of the evolution of the model with @xmath35 = 1.0 , @xmath13 = 10.0 pc ( initial separation , apoclustron distance ) , and @xmath14 = 0.0 .
color is related to the mass of the simulated star in @xmath19 according to the key provided .
merging for this model is the slowest of all the models ending in merging . once the two clusters get closer than 2 pc merging proceeds very quickly .
the @xmath24 axis points towards the galaxy center , @xmath25 is tangent to the galactocentric pair motion , and @xmath26 is perpendicular to the galaxy plane . for real open clusters
, we do not have access to the @xmath70 view . in the @xmath38 plane , the merger remnant looks significantly more elliptical than the single / individual cluster . by the end of the simulation
all the stars more massive than about 3.8 @xmath19 have already evolved away from the main sequence .
more massive objects are all stellar remnants or ( less frequently ) stellar mergers . no objects were removed from the calculations .
the initial orbital period was nearly 46 myr . [ merging ] ] plane ( top ) and the @xmath38 plane ( bottom ) of the evolution of the model with @xmath35 = 0.25 , @xmath13 = 10.0 pc , and @xmath14 = 0.95 .
colors are like in figure [ merging ] .
extreme tidal distortion of the less massive cluster and subsequent separation is observed for all pairs with appreciably different mass ratio . the identifiable cluster at 133 , 210 myr is the primary .
the secondary cluster gets shredded in a relatively short timescale ( see fig . [ earlytidal ] ) .
the @xmath24 axis points towards the galaxy center , @xmath25 is tangent to the galactocentric pair motion , and @xmath26 is perpendicular to the galaxy plane . for real open clusters , we do not have access to the @xmath70 view .
[ spaghetti ] ] .
this behavior is similar to the one identified by portegies zwart & rusli ( 2007 ) : the initially less massive cluster expands quickly initiating mass transfer to the more massive cluster .
fast expansion of the secondary cluster is the result of stellar mass loss ( see the text for details ) .
as for observed open clusters , de la fuente marcos & de la fuente marcos ( 2009c ) concluded that binary cluster candidates in the galactic disk appear to show a tendency to have components of different physical size .
our results clearly indicate that this effect may well be the result of dynamical interactions .
[ earlytidal ] ] ccccccc & & & & & & + @xmath13(pc ) & ( myr)/(pc ) & ( myr)/(pc ) & ( myr)/(pc ) & ( myr)/(pc ) & ( myr)/(pc ) & ( myr)/(pc ) + & ( myr ) & ( myr ) & ( myr ) & ( myr ) & ( myr ) & ( myr ) + + + 10 & 175/0.44@xmath71 & 95/0.21 & 72/0.22 & 39/0.16 & 32/0.23 & 37/0.14 + & 46 & 35 & 28 & 23 & 19 & 17 + 20 & @xmath72/369 & @xmath72/343 & @xmath72/325 & @xmath72/299 & @xmath72/263 & 100/0.19 + & 130 & 99 & 78 & 64 & 54 & 48 + 30 & @xmath72/497 & @xmath72/485 & @xmath72/470 & @xmath72/448 & @xmath72/421 & @xmath72/391 + & 239 & 182 & 144 & 118 & 99 & 88 + + + + + 10 & 78/0.18 & 64/0.10 & 53/0.55 & 61/0.22 & 59/0.32 & 51/0.17 + & 53 & 40 & 32 & 26 & 22 & 19 + 20 & @xmath72/332 & @xmath72/313 & @xmath72/321 & @xmath72/272 & @xmath72/246 & @xmath72/230 + & 150 & 114 & 91 & 74 & 62 & 55 + 30 & @xmath72/459 & @xmath72/445 & @xmath72/437 & @xmath72/421 & @xmath72/394 & @xmath72/384 + & 276 & 210 & 166 & 136 & 114 & 101 + + + + + 10 & 78/0.52 & 43/0.38 & 67/0.46 & 56/0.50 & 64/0.40 & 64/0.24 + & 53 & 40 & 32 & 26 & 22 & 19 + 20 & @xmath72/301 & @xmath72/308 & @xmath72/317 & @xmath72/273 & @xmath72/254 & 159/0.48 + & 150 & 114 & 91 & 74 & 62 & 55 + 30 & @xmath72/443 & @xmath72/428 & @xmath72/429 & @xmath72/404 & @xmath72/384 & @xmath72/370 + & 276 & 210 & 166 & 136 & 114 & 101 + + + + + 10 & @xmath72/134@xmath73 & @xmath72/116 & @xmath72/96 & @xmath72/91 & @xmath72/77 & @xmath72/46 + & 58 & 44 & 35 & 29 & 24 & 21 + 20 & @xmath72/282 & @xmath72/276 & @xmath72/262 & @xmath72/233 & @xmath72/212 & @xmath72/178 + & 164 & 125 & 99 & 81 & 68 & 60 + 30 & @xmath72/415 & @xmath72/401 & @xmath72/405 & @xmath72/380 & @xmath72/358 & @xmath72/351 + & 302 & 230 & 182 & 149 & 125 & 111 + | the basic properties of the candidate binary star cluster population in the magellanic clouds and milky way are similar .
the fraction of candidate binary systems is @xmath010% and the pair separation histogram exhibits a bimodal distribution commonly attributed to the transient nature of these clusters .
however , if primordial pairs can not survive for long as recognizable bound systems , how are they ending up ? here , we use simulations to confirm that merging , extreme tidal distortion and ionization are possible depending on the initial orbital elements and mass ratio of the cluster pair .
the nature of the dominant evolutionary path largely depends on the strength of the local galactic tidal field .
merging is observed for initially close primordial binary clusters but also for wider pairs in nearly parabolic orbits .
its characteristic timescale depends on the initial orbital semi - major axis , eccentricity , and cluster pair mass ratio , becoming shorter for closer , more eccentric equal mass pairs .
shredding or extreme tidal distortion of the less massive cluster and subsequent separation is observed in all pairs with appreciably different masses .
wide pairs steadily evolve into the separated twins state characterized by the presence of tidal bridges and cluster separations of 200 - 500 pc after one galactic orbit . in the galaxy ,
the vast majority of observed binary system candidates appear to be following this evolutionary path which translates into the dominant peak ( 25 - 30 pc ) in the observed pair separation distribution .
the secondary peak at smaller separations ( 10 - 15 pc ) can be explained as due to close pairs in almost circular orbits and/or undergoing merging .
merged clusters exhibit both peculiar radial density and velocity dispersion profiles shaped by synchronization and gravogyro instabilities .
both simulations and observations show that , for the range of open cluster parameters studied here , long term binary cluster stability in the milky way disk is highly unlikely . |
much like their long - duration cousins , the discovery of the first x - ray @xcite , optical @xcite , and radio @xcite afterglows of short - duration ( @xmath4s ) gamma - ray bursts ( grbs ) has driven great progress in recent years .
based primarily on the discovery of the first four afterglows , grb050509b ( @xmath5 ; @xcite ) , grb050709 ( @xmath6 ; @xcite ) , grb050724 ( @xmath7 ; @xcite ) , and grb051221 ( @xmath8 ; @xcite ) , a general picture of short - duration grb hosts quickly emerged . unlike long - duration events , short - duration grbs occur in both early- and late - type galaxies , all at relatively low redshift ( compared to @xmath9 for long - duration _ swift _ grbs ; @xcite ) . if , as is widely believed , short - duration grbs result from coalescence of a compact binary ( neutron star neutron star [ ns - ns ] or neutron star
black hole [ ns - bh ] ; @xcite ) , these nearby events imply a long progenitor delay time ( @xmath10gyr ; @xcite ) , inconsistent with the properties of such systems observed in the milky way . in a recent work , @xcite
have searched the afterglow error circles of nine additional short - duration bursts without an established redshift , identifying a population of faint ( @xmath11 26mag ) putative host galaxies .
spectroscopy of the brightest four of these galaxies indicate they lie at @xmath12 , while a comparison with field galaxies indicates the fainter hosts likely reside at @xmath13 .
the implications of the existence of a high - redshift ( @xmath3 ) population of short - duration bursts would be profound , in terms of the implied isotropic energy release ( @xmath14 @xmath15erg ) , the progenitor delay times ( both shorter and a broader distribution ) , and the possibility of detection with ground - based gravitational wave detectors . in this _
letter _ , we present spectra of the host galaxies of two recent grbs : the short - duration grb070429b and the likely short - duration with extended emission grb070714b .
both events have detected optical afterglows , providing both a secure host association and a secure redshift , thereby overcoming the primary limitation of our previous statistical study .
both hosts lie at @xmath1 , providing strong evidence in favor of a substantial population of high - redshift short - duration bursts .
we conclude by discussing the impact of this population on short - duration burst progenitor models .
the burst alert telescope ( bat ; @xcite ) on board _ swift _ detected and localized grb070429b at 03:09:04 ut @xcite .
the duration of the prompt emission was measured to be @xmath16s , placing the event firmly in the short - duration category @xcite . fitting a power - law with index @xmath17 , @xcite measure a prompt fluence of @xmath18erg @xmath19 in the 15 150kev band .
the fading x - ray afterglow was identified by the x - ray telescope ( xrt ; @xcite ) on board _ swift _ in ground analysis at location @xmath20 , @xmath21 ( j2000.0 ) with a 90%-confidence uncertainty of @xmath22 @xcite .
we imaged the position of the x - ray afterglow with the gemini multi - object spectrograph ( gmos ; @xcite ) mounted on the 8-m gemini south telescope beginning at utc 07:59:55 29 april ( @xmath23h after the trigger ) .
we identified a single faint , extended object inside the xrt error circle at coordinates @xmath24 , @xmath25 ( j2000.0 ) , with an uncertainty of @xmath26 ( fig .
[ fig:0429_finder ] ; @xcite ) . an additional ,
fainter source inside the xrt error circle was identified by @xcite at coordinates @xmath27 , @xmath28 ( j2000.0 ) .
the field was observed by several other facilities in the following months , yet no variability was detected in either source ( e.g. @xcite ) .
subsequent re - analysis of the initial observations from the _ swift _ uv - optical telescope ( uvot ; @xcite ) beginning 230s after the burst trigger revealed a marginal source in the @xmath29 , @xmath30 , and @xmath31 filters at @xmath32 , @xmath33 ( j2000.0 ) which subsequently faded away @xcite . while none of the individual uvot detections was particularly significant , summing all seven uvot filters from 591 to 2661s after the trigger results in a 3.9-@xmath34 detection @xcite .
we therefore accept that this source is the optical afterglow of grb070429b .
the uvot afterglow falls only @xmath35 from the centroid of the candidate host first proposed by @xcite ( fig .
[ fig:0429_finder ] ) . because the angular extent of the galaxy in our gemini image is about 1 , the uvot afterglow position is directly coincident with this galaxy . using the cumulative galaxy counts in the hudf and goods surveys
@xcite , we estimate the probability of an @xmath36mag galaxy falling within @xmath35 of the afterglow location to be @xmath37 . on the other hand ,
the chance alignment of the other object identified in the xrt error circle by @xcite is as high as 0.3 .
it is therefore likely that this source is the host galaxy of grb070429b .
we obtained a spectrum of this object with the low resolution imaging spectrograph ( lris ; @xcite ) mounted on the 10-m keck i telescope on the night of 12 august 2007 .
two 1800 s exposures were taken at high air mass ( @xmath38 ) with the 07 slit .
we employed a dichroic at 5600 to split the beam in two .
the 400/8500 grating on the red side provides coverage from 5600 8500 , with a dispersion of 1.86 pixel@xmath39 .
the 400/3400 grism was responsible for the dispersion on the blue end , resulting in coverage from the atmospheric cut - off up to @xmath40 5600 and a dispersion of 1.09 pixel@xmath39 .
spectra were reduced in the iraf environment using standard routines .
dithered spectra were subtracted to remove residual sky lines .
cosmic rays were removed using the la cosmic routine @xcite .
the resulting spectrum was extracted optimally @xcite and wavelength calibration was performed first relative to calibration lamps and then tweaked based on night sky lines .
both air - to - vacuum and heliocentric corrections were then applied to all spectra .
extracted spectra were divided through by a smoothed flux standard to remove narrow band ( @xmath41 ) instrumental effects @xcite .
flux calibration was performed relative to the spectrophotometric standard feige 11 @xcite . to account for slit losses
, we scaled the resulting spectrum to match the measured photometric @xmath42- and @xmath43-band magnitudes @xcite .
finally , we de - reddened the spectrum to account for the modest galactic extinction : @xmath44 @xcite .
the resulting spectrum of the host galaxy of grb070429b is shown in figure [ fig:0429_spec ] .
we find a single faint emission line at @xmath45 .
due to the lack of blueward features , we identify this line as [ ] @xmath0 3727 at a redshift of @xmath46 .
fitting the line with a gaussian profile , we measure a flux of @xmath47erg @xmath19 s@xmath39 . adopting the most recent @xmath48cdm parameters for a flat universe from wmap ( @xmath49 km s@xmath39 mpc@xmath39 ; @xmath50 , @xmath51 ; @xcite ) , and using the transformation from @xcite
, we infer a star formation rate of @xmath52m@xmath53 yr@xmath39 . correcting for the rest - frame @xmath30-band luminosity
, we find a specific star formation rate of @xmath54m@xmath53 yr@xmath39 l@xmath55 .
we caution , however , these values are only lower limits , as they do not incorporate any extinction native to the host .
these results are in overall agreement with the findings in the gcn circular of @xcite .
grb070714b was detected by the _
swift_-bat at 04:59:29 ut on 17 july 2007 @xcite . using the best - fit power - law spectrum with index @xmath56 , the 15 150kev fluence in the prompt emission was measured to be @xmath57erg @xmath19 . while the duration of the prompt emission is significantly longer than the canonical long - short divide ( @xmath58s ) , the burst light curve is comprised of several short spikes ( duration @xmath59s ) with a long , soft tail @xcite .
the spectral lag is consistent with zero @xcite , a property typically observed in short - duration bursts @xcite . like grbs 050709 and 050724 ,
grb070714b therefore appears likely to be a member of the short - duration bursts with extended emission ( e.g. @xcite ) .
however , together with the uncertainty regarding the duration , we caution that a small number of long - duration ( i.e. massive star progenitor ) _ swift _ events have been observed with zero lag @xcite .
a fading optical afterglow was discovered inside the xrt error circle @xcite by several groups @xcite .
the latest flux measurements and most precise astrometric position were reported by @xcite , who find an @xmath43-band magnitude of 25.5mag and location @xmath60 , @xmath61 ( @xmath62 ; j2000.0 ) .
the probability of a chance association for such a precise localization ( @xmath63 ) is quite small , @xmath64 @xcite .
we obtained two 2100s dithered spectra of the host galaxy of grb070714b with lris on the night of 13 august 2007 .
these spectra were reduced in an identical manner as those obtained for grb070429b ( [ sec:070429 ] ) .
flux calibration was performed relative to the spectrophotometric standard star bd+17@xmath654708 @xcite .
the resulting spectrum is shown in figure [ fig:0714_spec ] .
we find a single bright emission line at @xmath66 , which we again identify as [ ] @xmath0 3727 at @xmath67 .
this confirms the spectroscopic redshift first proposed by @xcite in the gcn circulars .
we measure a flux of @xmath68erg @xmath19 s@xmath39 for this line , corresponding to a lower limit on the star formation rate of @xmath69m@xmath53 yr@xmath39 @xcite .
the corresponding specific star formation rate for grb070714b is @xmath70m@xmath53 yr@xmath39 l@xmath55 .
in summary , we present the redshifts of the host galaxies of the short - duration grb070429b ( @xmath71 ) and the short - duration with extended emission grb070714b ( @xmath72 ) .
these are the two highest redshifts for short - duration grbs with detected optical afterglows , making the grb - host associations robust .
our results lend credence to the primary conclusions of our previous study of short - duration grb hosts and their redshifts , which was based in part on xrt positions ( @xmath73 ) , and in some cases on probabilistic redshifts from host magnitudes .
we now attempt to estimate the contribution of high - redshift galaxies to the observed short - duration population .
we first consider only those short - duration events with optical afterglows ( and hence secure host identifications ) and secure host redshifts ( from spectroscopy ) : grbs 050709 ( @xmath74 ) , 050724 ( @xmath75 ) , 051221a ( @xmath76 ) , 061006 ( @xmath77 ; @xcite ) , and the two hosts presented here .
high redshift ( @xmath3 ) hosts constitute 33% of this sample .
however , because these events are biased towards the brightest ( and therefore , probabilistically , most nearby ) hosts ( see @xcite , fig . 8) , we also consider three additional events with secure hosts but no host redshifts ( due primarily to faintness ) : grbs 060313 @xcite , 060121 @xcite , and 051227 @xcite . if , as is likely from their faintness , all three events lie at @xmath3 , the percentage of high redshift short - duration grbs may be as high as 56% .
we note that this result does not change dramatically if we include xrt - host associations ( e.g. grbs 050509b and 060502b ) .
the range of one- to two - thirds is in line with our previous analysis .
however , we caution that the errors on these estimates are quite large due to small number statistics , and we do not consider here any selections effects associated with the above samples . at @xmath1 ,
the isotropic prompt energy release from grb070429b is @xmath78erg , while the corresponding value for grb070714b is @xmath79erg .
these values are several orders of magnitude larger than most of the previously established short - duration grbs at @xmath80 , though still well short of typical values for long - duration events ( @xmath81erg ; @xcite ) .
the x - ray luminosity at @xmath82d , @xmath83 , extrapolated from the online _ swift _ x - ray light curve repository @xcite , is consistent with the correlation found between @xmath83 and @xmath84 found for the majority of short - duration events @xcite .
like long - duration events , collimation corrections may play an important role in the total energy budget for short - duration bursts ( e.g. grb051221a ; @xcite ) .
however , if a full 56% of short - duration events reside at @xmath3 , the energy release of short - duration events may approach the canonical long - duration value of @xmath85erg @xcite .
energies in excess of @xmath15erg are difficult to explain in the context of @xmath86 annihilation models @xcite , and may instead favor energy extraction via magneto - hydrodynamic processes ( e.g. @xcite ) .
much as with energetics , the specific star formation rate of grb070429b falls intermediate between the nearby short - duration events ( @xmath87m@xmath53 yr@xmath39 l@xmath55 ; @xcite and references therein ) and long - duration grbs ( @xmath88m@xmath53 yr@xmath39 l@xmath55 ; @xcite ) .
the specific star formation rate of grb070714b , however , is more consistent with the long - duration population , although large errors preclude any firm conclusions from being drawn .
the detection of short - duration grbs in late - type galaxies at @xmath89 is more consistent with our current understanding of compact binary coalescence models ( e.g. @xcite ) .
progenitor models with long delay times ( @xmath90gyr ) and narrow , log - normal distributions ( @xmath91 ) were consistent with previous analyses based solely on the most nearby events .
our results here favor models with a shorter characteristic delay time ( @xmath92gyr ) , but much broader distribution ( @xmath93 ; @xcite ) .
further studies with a significantly expanded sample will be required for a detailed comparison of the properties of short - duration hosts with those predicted by compact binary merger models ( e.g. @xcite ) . based in part on observations obtained at the gemini observatory ( program i d : gs-2007a - q-6 ) , which is operated by the association of universities for research in astronomy , inc . , under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the science and technology facilities council ( united kingdom ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , cnpq ( brazil ) and secyt ( argentina )
. s. b. c. is supported by a nasa graduate student research fellowship .
m. m. k. acknowledges funding from the gordon and betty moore foundation .
grb research at caltech is supported by nasa . , i. _ et al .
_ 2003 , in instrument design and performance for optical / infrared ground - based telescopes .
edited by iye , masanori ; moorwood , alan f. m. proceedings of the spie , volume 4841 , pp .
1645 - 1656 ( 2003 ) . ,
m. iye & a. f. m. moorwood , 16451656 | we present optical spectra of the host galaxies of the short - duration gamma - ray burst grb070429b and the likely short - duration with extended emission grb070714b . in both cases
, we find a single emission line that we identify as [ ] @xmath0 3727 at @xmath1 .
both events are more distant than any previous short - duration grb with a secure host association from the sub - arcsecond position of an optical afterglow . grbs070429b and 070714b
provide strong evidence in support of our previous claims in berger et al .
that a significant fraction of short - duration hosts ( @xmath2 ) reside at @xmath3 .
we discuss the implications of the existence this population on the energetics of short - duration grbs , as well as on progenitor models . in the context of the degenerate binary merger scenario , such events require progenitor systems with a range of lifetimes and disfavor progenitor models with a long , narrow lifetime distribution . |
in order to obtain reliable empirical information on the strength and geometry of astrophysical magnetic fields we need to develop and apply suitable diagnostic tools within the framework of the quantum theory of spectral line polarization ( e.g. , the recent monograph by * ? ? ?
. a nave knowledge and/or an unjustified neglect of some of the physical mechanisms that produce polarization in spectral lines may lead to significant errors in our inferences and/or to missing the exciting opportunity of proposing new practical ideas for magnetic field diagnostics . in this respect ,
the main aim of this paper is to help clarify further the physics of the observed polarization in the spectral lines of the he i 10830 multiplet , because they contain precious information on the dynamical and magnetic properties of a variety of solar chromospheric and coronal structures .
the most important mechanisms that induce and modify polarization signatures in the spectral lines that originate in the atmospheres of the sun and of other stars are the zeeman effect , anisotropic radiation pumping and the hanle effect ( e.g. , the recent review by * ? ? ?
the zeeman effect requires the presence of a magnetic field , which causes the atomic and molecular energy levels to split into different magnetic sublevels characterized by their magnetic quantum number @xmath3 . as a result ,
the wavelength positions of the @xmath4 ( @xmath5 ) , @xmath6 ( @xmath7 ) and @xmath8 ( @xmath9 ) transitions do not coincide and , therefore , their respective polarization signals do not cancel out .
the zeeman effect is most sensitive in _ circular _ polarization ( quantified by the stokes @xmath10 parameter ) , with a magnitude that for not too strong fields scales with the ratio between the zeeman splitting and the width of the spectral line ( which is very much larger than the natural width of the atomic levels ! ) .
this so - called _ longitudinal _ zeeman effect responds to the line - of - sight component of the magnetic field . in contrast , the _ transverse _ zeeman effect responds to the component of the magnetic field perpendicular to the line of sight , producing instead _ linear _ polarization signals ( quantified by the stokes @xmath1 and @xmath2 parameters ) that are normally below the noise level of present observational possibilities for intrinsically weak fields ( typically b@xmath11 gauss for solar spectropolarimetry ) . the anisotropic illumination of the atoms in the outer regions of a stellar atmosphere may produce atomic level polarization ( that is , population imbalances and/or quantum coherences between the magnetic sublevels pertaining to the upper and/or lower level of the line transition under consideration ) , in such a way that the populations of substates with different values of @xmath12 are different .
this is termed _ atomic level alignment_. under such circumstances , the polarization signals of the @xmath4 and @xmath13 transitions do not cancel out , even in the absence of a magnetic field , simply because the population imbalances among the magnetic sublevels imply more or less @xmath4 transitions , per unit volume and time , than @xmath13 transitions .
interestingly , this atomic level polarization is modified by the presence of a magnetic field _ inclined _ with respect to the symmetry axis of the pumping radiation field ( i.e. , the hanle effect ; e.g. , the review by * ? ? ?
the magnetic field intensity ( measured in gauss ) that is sufficient to produce a sizable change in the atomic polarization of a given level is @xmath14 where @xmath15 and @xmath16 are , respectively , the level s lifetime ( in seconds ) and its land factor .
as discussed by @xcite a notable example of a multiplet whose spectral lines are sensitive to both effects ( that is , to the zeeman splitting and to the atomic level polarization ) is the he i 10830 multiplet , whose stokes profiles can be observed in a variety of plasma structures of the solar chromosphere and corona , such as sunspots @xcite , coronal filaments @xcite , prominences @xcite , emerging flux regions @xcite , chromospheric spicules @xcite , active region filaments ( martnez pillet et al .
2006 ; in preparation ) and flaring regions ( sasso et al .
2006 ) .
the he i 10830 multiplet originates between a lower term ( @xmath17 ) and an upper term ( @xmath18 ) .
therefore , it comprises three spectral lines : a ` blue ' component at 10829.09 ( with @xmath19 and @xmath20 ) , and two ` red ' components at 10830.25 ( with @xmath21 ) and at 10830.34 with ( @xmath22 ) which appear blended at solar atmospheric temperatures .
although the circular polarization of the he i 10830 lines is dominated by the longitudinal zeeman effect , we expect their linear polarization to be the result of the joint action of the transverse zeeman effect and of the atomic polarization that anisotropic radiation pumping processes induce in the helium levels .
actually , as we shall show in this paper , the degree of anisotropy of the solar continuum radiation at the wavelengths of the relevant helium transitions is sufficiently important so as to produce a significant amount of atomic polarization in the lower and upper levels of the he i 10830 multiplet , even at relatively low atmospheric heights ( e.g. , @xmath231000 km ) .
elastic collisions with neutral hydrogen atoms are unable to destroy this atomic polarization that the anisotropy of the solar radiation field can ( in principle ) induce in the helium levels .
obviously , for sufficiently weak magnetic fields ( e.g. , for @xmath24 g ) the linear polarization is dominated by the _ selective emission _ and _ selective absorption _ of polarization components that result from the atomic level polarization @xcite .
however , it would not be correct to give for granted that for relatively strong fields ( say , for @xmath25 g ) the linear polarization of the he i 10830 multiplet is going to be necessarily dominated by the transverse zeeman effect . unfortunately , with very few exceptions @xcite
, the modeling of spectropolarimetric observations of the he i 10830 multiplet has been carried out without taking into account that , in principle , the observed linear polarization may be actually the result of the joined action of the transverse zeeman effect and of the atomic level polarization . to neglect _ a priori
_ the influence of atomic level polarization simply because the considered point of the observed field of view is significantly magnetized is not justified .
this is a reasonable approximation when the plasma diagnostics is being done via the analysis of only the stokes @xmath26 and @xmath10 profiles , such as those measured in sunspot umbrae ( e.g. , * ? ? ?
* ) , but it might not be so adequate in the case that stokes @xmath1 and @xmath2 are also considered , such as those observed by @xcite and @xcite in emerging flux regions .
a recent investigation has already pointed out that , even when only the contribution of the zeeman effect is accounted for , the wavelength positions and the strengths of the @xmath4 and @xmath13 components should be calculated in the incomplete paschen - back effect regime , given that the linear zeeman effect theory _ overestimates _ the amplitudes of the emergent linear and circular polarization that is , it _ underestimates _ the inferred magnetic field strength @xcite . in this paper
we focus instead on clarifying the role played by the presence of atomic level polarization on the emergent stokes profiles of the he i 10830 multiplet for increasing values of the strength of the assumed magnetic field .
as we shall see below , the influence of atomic level polarization on the stokes @xmath1 and @xmath2 profiles of the he i 10830 multiplet turns out to be significant , even for magnetic field strengths as large as 1000 g. in fact , as pointed out below , some of the stokes profiles of emerging flux regions that @xcite and @xcite have interpreted in terms of the linear zeeman effect theory neglecting atomic level polarization show , however , clear observational signatures of the presence of this physical ingredient . on the other hand , if any stokes profiles observation of a moderately magnetized solar atmospheric region turns out to show no hint at all of the presence of atomic level polarization ( such as it seems to be the case with the active region filament observations reported by v. martnez pillet and collaborators during the fourth international workshop on solar polarization ) the reason , in our opinion , is to be found in the lower degree of anisotropy of the radiation field that pumps the helium atoms inside such elongated , optically - thick plasma structures .
the outline of this paper is as follows .
the formulation of the problem is presented in section 2 , where we inform about the equations we have solved for calculating the emergent stokes profiles from an optically thick slab located at a height of about 2000 km above the solar visible surface " .
section 3 shows several illustrative examples of the generated spectral line polarization when neglecting or taking into account the atomic level polarization induced by the photospheric continuum radiation .
section 4 investigates the extent to which the anisotropy of the photospheric radiation field may be modified by the radiation generated by the assumed plasma structure itself .
finally , in section 5 we summarize the main conclusions of our work and insist on the scientific interest for spectropolarimetry from space .
as illustrated in fig . 1 , we consider a constant - property slab of magnetized chromospheric material located at a height of only 3 arc seconds ( @xmath27 ) km above the solar visible " surface . the magnetic field ,
whose strength we will vary at will , is assumed to be horizontal ( i.e. , parallel to the solar surface " ) .
the slab s optical thickness at the wavelength and line of sight under consideration is @xmath28 , while the symbol @xmath29 denotes the optical thickness of the slab along its normal direction at the line center of the red blended component .
we have chosen @xmath30 , which is just at the transition limit between an optically thin and an optically thick medium .
all the atoms inside this slab are assumed to be illuminated by the unpolarized and limb - darkened photospheric radiation field whose center - to - limb variation has been tabulated by @xcite .
this implies that , for the moment , we are neglecting the influence of possible radiative transfer effects inside this @xmath30 slab on the anisotropy factor , whose definition is @xcite @xmath31 where @xmath32 and @xmath33 with @xmath34 ( @xmath35 being the angle between the ray under consideration and the solar local vertical ) .
therefore , in a stellar atmosphere the possible values of the anisotropy factor vary between @xmath36 for the limiting case of illumination by a purely horizontal radiation field without any azimuthal dependence , and @xmath37 for purely vertical illumination .
two main factors contribute to the value of @xmath38 at a given height in the _ quiet _ solar atmosphere : ( 1 ) the center to limb variation of the solar radiation field at the wavelength under consideration and ( 2 ) the geometrical effect due to the fact that , the larger the atmospheric height , the smaller the solid angle subtended by the solar visible sphere .
we point out that , at a height of 3 arc seconds in the quiet solar atmosphere , @xmath39 at 10830 , while @xmath40 if only the above - mentioned geometrical effect is taken into account ( that is , when the contribution due to the center - to - limb variation is disregarded ) .
the radiative transitions caused by the anisotropic illumination of the slab s helium atoms induce population imbalances and quantum coherences among the magnetic substates of the energy levels ( that is , atomic level polarization ) , which we quantify by solving the statistical equilibrium equations for the multipole components , @xmath41 , of the atomic density matrix ( see section 7.6a in * ? ? ?
we do this using a realistic model atom that includes the 5 lowest terms of the triplet system of neutral helium , which implies 11 @xmath42-levels and 6 transitions between the terms ( see fig . 13.9 in landi deglinnocenti & landolfi 2004 ) .
such equations take fully into account the hanle and zeeman effects produced by the assumed horizontal magnetic field .
we point out that we calculate the wavelength positions and the strengths of the zeeman components in the incomplete paschen - back effect regime . from the calculated density - matrix elements it is then possible to compute the coefficients @xmath43 and @xmath44 ( with @xmath45 ) of the emission vector and the coefficients @xmath46 , @xmath47 , and @xmath48 of the @xmath49 propagation matrix of the stokes vector transfer equation for a wavelength interval covering the 10830 multiplet ( see sections 7.6b in * ? ? ?
the emergent stokes vector @xmath50 ( with @xmath51=transpose , @xmath52 the frequency and @xmath53 the line of sight ) is given by the following expression , which can be easily obtained as a particular case of eq .
( 27 ) of @xcite : @xmath54^{-1 } \left [ \left ( e^{-\tau } \mathbf{1 } - \psi_m \mathbf{k } ' \right ) \mathbf{i}_{\rm sun } + ( \psi_m+\psi_0 ) \mathbf{s } \right ] , \label{eq : shortcar}\ ] ] where @xmath55 is the identity matrix and @xmath56 the stokes vector that illuminates the slab s boundary that is most distant from the observer , while @xmath57 and @xmath58 are given by @xmath59 in these expressions @xmath60 is the propagation matrix of the stokes vector transfer equation ( whose elements are @xmath61 , @xmath62 and @xmath63 ; @xmath64 being @xmath65 or @xmath10 ) , while @xmath66 is the emission vector .
the coefficients @xmath67 and @xmath68 depend only on the optical depth of the slab at the frequency and line - of - sight under consideration and their expressions are : @xmath69 it is of interest to note that when the anomalous dispersion terms are neglected in eq .
( [ eq : shortcar ] ) ( i.e. , the @xmath48 terms are taken equal to zero ) we obtain @xmath70 @xmath71 these approximate formulae for the emergent stokes parameters coincide with those proposed by @xcite for modeling the hanle and zeeman effects in solar chromospheric spicules and coronal filaments , which provide a very good approximation whenever @xmath72 and @xmath73 .
although such inequalities are often met in solar spectropolarimetry ( e.g. , * ? ? ?
* ) , all the calculations of this paper have however been carried out using the exact analytical solution given by eq .
( [ eq : shortcar ] ) , because they allow us to provide accurate results also for relatively high field strength values .
it is of interest to point out that eqs .
( [ eq : stokesi ] ) and ( [ eq : stokesx ] ) simplify as follows for the optically thin and optically thick limiting cases : \(1 ) _ optically thin case _ ( @xmath74 ) @xmath75 @xmath76 \(2 ) _ optically thick case _ ( @xmath77 ) @xmath78 @xmath79 where @xmath80 .
this section presents results for the two line of sights illustrated in fig . 1 : @xmath81 and @xmath82 , where @xmath34 ( @xmath35 being the angle between the solar radius vector through the observed point and the line of sight )
. therefore , @xmath81 corresponds to the case of an off - limb observation ( @xmath0 scattering geometry ) , while @xmath82 to that of a solar disk center observation ( forward scattering geometry ) .
we point out that the boundary conditions are @xmath83 for the @xmath0 scattering case , but @xmath84 and @xmath85 for the forward scattering case ( with @xmath86 taken from @xcite ) .
figure [ fig : fig2 ] shows the emergent stokes profiles corresponding to the two lines of sight illustrated in fig .
the left panels of fig .
2 concern the @xmath0 scattering or @xmath81 case , which is typical of any off - limb observation .
the right panels consider the forward scattering or @xmath82 case , which is typical of an on - disk observation at or close to the center of the solar disk . in both cases
we have assumed a constant - property slab with @xmath30 located at a height of only 3 arc seconds above the visible solar surface . as mentioned above
, the magnetic field is assumed to be horizontal ( i.e. , parallel to the solar surface " ) and oriented as indicated in fig .
1 that is , in a way such that also for the off - limb case the magnetic field vector is perpendicular to the line of sight . from top to bottom fig .
[ fig : fig2 ] shows the emergent stokes profiles for increasing values of the magnetic strength and for the following three calculations of increasing realism : \(1 ) the dotted lines indicate the case without atomic level polarization . here
the only mechanism responsible for the emergent polarization is the zeeman splitting of the upper and/or lower energy levels , which produces wavelength shifts between the @xmath4 and @xmath13 components , whose positions and strengths have been calculated in the incomplete paschen - back effect regime .
therefore , zero polarization is found for @xmath87 g. \(2 ) the dashed lines correspond to the case in which we have taken into account the influence of the atomic polarization of the two upper levels of the he i 10830 multiplet that can carry atomic level polarization -that is , those with @xmath88 and @xmath89 .
therefore , in addition to the above - mentioned zeeman effect contribution , we have here the possibility of a selective emission of polarization components , even for the zero field case . for example , this is the reason why the dashed line of the upper left panel of fig .
[ fig : fig2 ] shows a non - zero linear polarization signal for the off - limb zero field case .
\(3 ) the solid lines correspond to the most general situation in which the influence of the atomic polarization of the lower level is also taken into consideration , in addition to that of the upper levels and to the zeeman effect .
the consideration of lower - level polarization has two consequences .
first , the amount of upper level polarization and the ensuing selective emission of polarization components is modified .
second , we can also have a selective absorption of polarization components .
for instance , this is the reason why the blue line of the he i 10830 multiplet shows a non - zero linear polarization signal in the @xmath90 g right panel of fig .
[ fig : fig2 ] . as seen in fig .
[ fig : fig2 ] , for magnetic strengths @xmath24 gauss there is no significant contribution of the transverse zeeman effect . in this _ weak field _
regime the emergent linear polarization is completely dominated by the atomic polarization of the lower and upper levels of the three line transitions involved .
we emphasize that upper - level polarization leads to a _ selective emission _ of polarization components , while lower - level polarization to a _ selective absorption _ of polarization components ( see fig .
as pointed out by @xcite , selective absorption is the only mechanism that can produce linear polarization in the blue component of the he i 10830 multiplet .
this is because the total angular momentum of its upper level is @xmath91 , which implies that , in the weak field regime under consideration , @xmath92 .
the reason why the off - limb panels of fig .
2 with @xmath87 and @xmath90 gauss show no linear polarization in that blue line is that our slab with @xmath93 is optically thin at that blue wavelength , and also because eqs .
( 11 ) and ( 12 ) with @xmath83 imply @xmath94 which in the weak field regime is zero for the blue component of the helium multiplet ( because its upper level , with @xmath91 , is intrinsically unpolarizable and @xmath92 ) .
interestingly , the same eqs .
( 11 ) and ( 12 ) indicate that , if the boundary intensity @xmath95 were non zero , then there should be a significant @xmath96 signal .
in fact , this is the situation we have in all the right panels of fig .
[ fig : fig2 ] , which correspond to simulated observations at solar disk center . obviously , due to symmetry reasons , forward scattering processes in the absence of a magnetic field produce zero polarization , as seen in the @xmath87 gauss right panel of fig .
2 . the same applies if there is a microturbulent magnetic field , or if the magnetic field vector is parallel to the symmetry axis of the radiation field that illuminates the slab s helium atoms .
however , as shown in the @xmath90 gauss right panel of fig .
2 , forward scattering processes in the presence of an inclined magnetic field do produce linear polarization in the lines of the he i 10830 multiplet . here ,
the linear polarization is actually _ created _ by the hanle effect .
this is easy to understand by reasoning within the framework of the oscillator model for the hanle effect in a triplet - type transition with @xmath97 and @xmath21 ( e.g. , * ? ? ?
for the more general case of a line transition between two isolated levels having any possible @xmath98 and @xmath99 values , it is first necessary to recall ( e.g. , * ? ? ?
* ) that the hanle effect tends to destroy the quantum coherences between the magnetic sublevels pertaining to each @xmath42-level , _ without modifying the population imbalances_-level under consideration , which are commonly denoted by the symbol @xmath100 . ] . in the magnetic field reference system , the quantum coherences ( i.e.
, the @xmath101 values with @xmath102 ) tend to vanish for magnetic strengths @xmath103 ( that is , in the saturation regime of the hanle effect ) , so that in practice in this regime we are only left with the population imbalances among the magnetic sublevels pertaining to each @xmath42-level ( i.e. , with the @xmath101 values with @xmath104 ) .
therefore , in the presence of a magnetic field inclined with respect to the symmetry axis of the pumping radiation field , the population imbalances of the upper and/or lower levels can produce linear polarization parallel or perpendicular to the horizontal component of the magnetic field vector , simply because of the ensuing selective emission and selective absorption of polarization components .
the amplitude of the resulting stokes @xmath1 signal can be easily estimated , in the optically thin limit of eq .
( 12 ) and/or in the optically thick limit of eq .
( 14 ) , by introducing into such expressions the following approximate formulae ( see trujillo bueno 2003b ) : @xmath105 @xmath106 where @xmath107 quantifies the degree of population imbalance of the @xmath42-level under consideration , while @xmath108 and @xmath109 are numerical coefficients that depend on the @xmath98 and @xmath99 values and @xmath110 , with @xmath111 given by eq . ( 10.12 ) of @xcite for @xmath112 . ] ( e.g. , @xmath113 and @xmath114 for a line transition with @xmath19 and @xmath20 , @xmath115 and @xmath116 for a line transition with @xmath97 and @xmath21 , and @xmath117 for a line transition with @xmath118 ) .
it is very important to note that in eqs .
( [ eq : qia ] ) and ( [ eq : qib ] ) @xmath119 , where @xmath120 is _ the angle between the magnetic field vector and the line of sight_. consider the case of our slab of chromospheric plasma at a given height above the visible solar surface " and permeated by a horizontal magnetic field ( see fig .
1 ) . a magnetic field of 100 gauss is more than sufficient to seriously reduce the quantum coherences ( as quantified in the magnetic field reference frame ) , but is still sufficiently weak so as to be sure that the contribution from the transverse zeeman effect is negligible . for the he i
multiplet this happens for @xmath121 g , approximately . under such circumstances the above - mentioned expressions should provide a reasonable approximation for estimating the emergent @xmath96 at the line center of a significantly strong spectral line . in agreement with our detailed numerical calculations
such approximate formulae predict linear polarization for a disk center observation , be it for a line transition with @xmath19 and @xmath20 , or for one with @xmath97 and @xmath21 , or for one with non - zero @xmath98 and @xmath99 values ( except the particular case of a spectral line with @xmath122 , because a level with @xmath123 can not be aligned ) .
we now turn our attention to discussing the cases of fig .
[ fig : fig2 ] for which the transverse zeeman effect plays a significant role -that is , those with @xmath124 gauss .
as seen in the corresponding panels , the influence of atomic level polarization on the emergent linear polarization is significant , mainly for the red blended component ( which results from two transitions whose upper levels can be polarized ) .
as expected , the stronger the field the smaller the difference between the solid - line profiles ( which show the joint effect of all physical ingredients ) and the dotted profiles ( which neglect the influence of atomic level polarization ) .
the dashed profiles assume that the lower level is unpolarized , but take into account the selective emission processes that result from the presence of upper - level atomic polarization . for a slab with @xmath30 ( which implies a smaller optical thickness at the wavelength of the blue line ! ) the influence of lower - level polarization is significant for the @xmath125 gauss disk center case , but insignificant for sensibly stronger fields .
however , upper - level polarization plays an important role on the linear polarization of the red blended component , even for magnetic strengths as large as 1000 gauss , for both the off - limb and on - disk cases .
for example , for the @xmath125 gauss case ( which is representative of the strengths found in emerging flux regions ) the accurately computed emergent stokes @xmath1 profile is very different from that obtained taking into account only the influence of the transverse zeeman effect within the framework of the paschen - back effect theory .
interestingly , the above - mentioned observational signature of the presence of atomic polarization in the helium levels is clearly seen in many of the stokes profiles observed in emerging flux regions , such as those of fig . 2 of @xcite .
the solid lines of our fig . [ fig : fig3 ] show a good theoretical fit to such observations ( see fig . 2 of lagg et al .
2004 ) , which we have obtained by taking into account the influence of atomic level polarization , in addition to that of the zeeman effect .
the dotted lines of fig . [ fig : fig3 ] neglect , however , the influence of atomic level polarization .
a comparison of such theoretical stokes profiles with those observed by @xcite indicates the presence of atomic level polarization in a relatively strong field region ( @xmath231000 gauss ) .
it also shows that neglecting the influence of atomic level polarization on the emergent stokes profiles is a suitable approximation for interpreting the circular polarization , but an unsuitable one for modeling the observed linear polarization .
we conclude that the stokes @xmath1 and @xmath2 profiles observed by @xcite in emerging flux regions are strongly modified by the presence of atomic level polarization , even at those atmospheric points of the observed field of view for which field strengths as large as 1000 gauss are inferred . in any case
, it may be tranquilizing to point out that an inversion of the observed profiles neglecting atomic level polarization yields a similar magnetic field vector , in spite of the fact that the corresponding theoretical fit is very poor .
a much better theoretical fit is automatically obtained when the influence of atomic level polarization is properly taken into account , which is important for the reliability of the stokes inversion results ( see asensio ramos & trujillo bueno 2006 , in preparation , for a detailed description of our forward modeling and inversion codes we have applied in this investigation ) .
at the fourth international workshop on solar polarization ( spw4 ) that took place in boulder ( usa ) during september 2005 , v. martnez pillet and collaborators reported on spectropolarimetric observations of ( low - lying ) active region filaments in the he i 10830 multiplet , taken with the new version ( tip-2 ) of the tenerife infrared polarimeter ( martnez pillet et al .
2006 ; in preparation ) .
in particular , in his spw4 talk v. martnez pillet pointed out that the stokes profiles of the observed active region filaments had the typical shape of polarization profiles produced by the zeeman effect , without showing any hint at all of the observational signature of the hanle effect in forward scattering that @xcite had found in solar coronal filaments located at relatively large heights above quiet regions of the solar surface " .
in other words , the typical shapes of the linear polarization profiles observed by martnez pillet et al . in active region filaments were similar to those of the dotted lines of fig .
he also reported that hanle - effect signals similar to those found by @xcite were however present in the quiet regions of the observed field of view , outside the filament plasma . which could be the explanation of this enigmatic observational finding ?
one possibility is that elastic collisions with neutral hydrogen atoms completely destroy the polarization of the helium levels , but this is very unlikely because the typical densities of solar plasma structures are too low to affect the ( short - lived ) upper levels .
in fact , simple estimates based on eq .
( 7.108 ) of landi deglinnocenti & landolfi ( 2004 ) suggest that at a height of 2000 km in the fal - c semi - empirical model of fontenla et al .
( 1993 ) the upper - level rates of depolarizing collisions are about four orders of magnitude smaller than the corresponding einstein @xmath126-coefficient .
therefore , collisional depolarization seems to be indeed negligible , even if the hydrogen number density of active region filaments were three orders of magnitude larger than those of the fal - c model at chromospheric heights .
another depolarizing possibility could be the bound - free transitions caused by the uv ionizing radiation coming downwards from the corona , but as is well known most of the ionizations take place from the singlet states of he i ( instead of from the triplet states of the 10830 multiplet ) and , in any case , the intensity of the ionizing radiation seems to be too low so as to produce any significant effect ( casini & manso sainz 2006 ; private communication ) . in our opinion , what is happening here is that the radiation field generated by the active region filament itself ( which is not an optically - thin structure ! ) makes a negative contribution to the anisotropy factor , so that the anisotropy of the true radiation field that illuminates the slab s helium atoms is negligible . in order to investigate this possibility
we have calculated the anisotropy factor at each point inside a slab of total optical thickness @xmath127 in a way similar to that followed by asensio ramos , landi deglinnocenti & trujillo bueno ( 2005 ) , but taking into account the center - to - limb variation of the photospheric radiation field .
if the center - to - limb variation is parameterized by the following functional form @xcite @xmath128 the resulting expression for the anisotropy factor is . @xmath129 where @xmath28 is measured starting at the lower boundary .
the quantities @xmath130 , @xmath131 , @xmath132 and @xmath133 are given by : @xmath134u + \big [ { \mathrm{e}_{2}(\tau ) } - 4{\mathrm{e}_{4}(\tau ) } + 3{\mathrm{e}_{6}(\tau)}\big]v \nonumber \\ a ' & = & 2- { \mathrm{e}_{2}(\tau ) } - { \mathrm{e}_{2}(\tau_\mathrm{tot } - \tau ) } \nonumber \\ b ' & = & { \mathrm{e}_{2}(\tau ) } - \big [ { \mathrm{e}_{2}(\tau ) } - { \mathrm{e}_{3}(\tau ) } \big]u - \big [ { \mathrm{e}_{2}(\tau ) } - { \mathrm{e}_{4}(\tau ) } \big]v.\end{aligned}\ ] ] the quantities @xmath135 and @xmath136 are the coefficients of the center - to - limb variation given by @xcite , while @xmath137 is the exponential integral of order @xmath138 .
it is also interesting to point out that for the case of an isolated slab located at any given height above the solar visible " surface the previous expression simplifies considerably , since it is given by @xmath139 figure [ fig : fig4 ] shows how the anisotropy factor varies inside a slab of optical thickness @xmath140 , which is located at only 3 arc seconds above the visible surface " and is illuminated from below by the continuum photospheric radiation .
the results are shown for increasing values of the ratio @xmath141 ( where @xmath142 is the intensity of the vertical ray coming from the underlying photosphere , while @xmath143 is the slab s source function ) .
the dotted horizontal line indicates the value of the anisotropy factor , @xmath38 , that we would have if the contribution of the radiation field generated by the slab itself were neglected .
the remaining curves show the anisotropy factor for a range of @xmath141 values .
as we can see , there is a range of @xmath141 values around @xmath144 ( i.e. , the expected values when the atomic excitation is dominated by radiative transitions ) for which @xmath145 at many points inside the slab . in our opinion
, these are precisely the physical conditions inside the active region filaments observed by v. martnez pillet et al . , which are elongated plasma structures located at relatively low heights above the visible solar surface " .
the illumination conditions in the emerging flux regions observed by @xcite are clearly different . here
the illumination is due to the typical radiation field of a stratified atmosphere , whose anisotropy is dominated by the positive contribution caused by the limb darkening of the outgoing radiation ( e.g. , trujillo bueno 2001 ) . on the contrary , active region filaments are elongated plasma structures located at relatively low heights above the solar visible surface " .
such plasma structures levitating " in the solar atmosphere have a significant optical thickness , which implies that we have to take into account the negative contribution to @xmath38 caused by the ( mainly ) horizontal radiation field generated by the filament itself . if this reduction of the radiation field s anisotropy is not taken into account when doing inversions of the stokes @xmath1 and @xmath2 profiles observed in such type of optically - thick structures ( that is , when only the positive anisotropy of the photospheric radiation field is considered ) , the inversion algorithm may artificially select magnetic field vectors with inclinations around the van vleck angle ( @xmath146 ) , because it is for this particular inclination that the contribution of atomic level polarization is minimized .
fortunately , the observed stokes @xmath10 profile is often available , but it is very unlikely that the observed linear polarization is automatically reproduced after choosing a magnetic strength that fits the observed circular polarization assuming that the magnetic field inclination is close to that of the van vleck angle . obviously , when a spectropolarimetric observation of the he i 10830 multiplet does not show any hint at all of the presence of atomic level polarization , the best one can do for inferring the magnetic field vector is to apply an inversion code which takes only into account the zeeman effect , but with the positions and strengths of the @xmath4 and @xmath13 components calculated within the framework of the paschen - back effect theory .
probably , one of our most interesting conclusions is that the modeling of the emergent stokes @xmath1 and @xmath2 profiles of the he i 10830 multiplet should not be done by neglecting the possible presence of atomic level polarization . actually , the degree of anisotropy of the solar continuum radiation at 10830 is sufficiently important so as to produce a sizable amount of atomic alignment in the lower and upper levels of the he i 10830 multiplet , even at relatively low atmospheric heights such as 1000 km . for weak magnetic fields ( e.g. , @xmath24 g )
the emergent linear polarization is fully dominated by the _ selective emission _ and _ selective absorption _ of polarization components that result from this atomic level polarization . for stronger magnetic fields ,
the contribution of the transverse zeeman effect can not be neglected .
however , the emergent linear polarization may still show an important contribution caused by the presence of atomic level polarization , even for magnetic strengths as large as 1000 g ( see fig .
2 ) . in emerging magnetic flux regions the anisotropic illumination of the helium atoms
is expected to be more or less similar to that corresponding to a stratified stellar atmosphere , with the ( mainly vertical ) outgoing radiation contributing with positive values to the anisotropy factor , @xmath38 , and with the ( mainly horizontal ) incoming radiation contributing with negative values . as a result ,
@xmath147 at a height of 3 arc seconds in the quiet solar atmosphere and a significant amount of lower - level and upper - level polarization is induced by the ensuing radiative transitions .
interestingly , the observational signature of this atomic level polarization is clearly seen in many of the stokes @xmath1 and @xmath2 profiles observed by lagg et al .
( 2004 ) , even at points of the observed field of view for which magnetic strengths as large as 1000 g are inferred .
the illumination conditions are however different in optically thick plasma structures embedded in the solar atmosphere , such as those encountered in ( low lying ) active regions filaments . here , in addition to the positive contribution to @xmath38 due to the continuum radiation field coming from the underlying solar photosphere , we have to take into account the negative contribution caused by the radiation field generated by the optically - thick plasma structure itself . as a result ,
the anisotropy factor of the true radiation field that illuminates the helium atoms inside the filament plasma may be very different , as illustrated in fig . 4 for the case of a constant - property slab of total optical thickness unity .
interestingly , for @xmath148 we find that @xmath145 at many points inside the slab ( with @xmath95 the intensity of the vertical ray of the photospheric radiation and @xmath149 the slab s source function ) . under such circumstances
the ensuing atomic level polarization turns out to be negligible and the emergent stokes @xmath1 and @xmath2 profiles are dominated by the transverse zeeman effect . in our opinion
, this is the main reason that explains the absence of any observational signature of the presence of atomic level polarization in the linear polarization profiles observed by martnez pillet and collaborators in active region filaments . finally , it is interesting to point out that for the case of isolated plasma structures located at any given height above the solar visible surface " the anisotropy factor may reach a significant value ( see the thin solid - line curve of fig .
4 , which corresponds to @xmath150 ) .
this suggests that a variety of plasma structures , like loops embedded in the @xmath151 k solar corona , should produce measurable linear polarization even at spectral line wavelengths for which the underlying solar disk is seen completely dark ( e.g. , at the euv and x - ray spectral regions ) .
this last point emphasizes further the relevance of the scientific case for spectropolarimetry from space , so that space agencies worldwide should indeed try to open soon this new diagnostic window on the universe .
we would like to thank roberto casini ( hao ) , rafael manso sainz ( iac ) and egidio landi deglinnocenti ( universit di firenze ) for stimulating scientific discussions and for their careful reading of this paper .
the content of section 4 was motivated by the interesting observational results on active region filaments presented by v. martnez pillet and collaborators during the fourth international workshop on solar polarization , which took place in boulder ( usa ) during september 2005 .
this paper has been completed during the summer of 2006 while jtb was holding the _ gauss - professur _ granted by the _ akademie der wissenschaften zu gttingen_. he wishes to thank franz kneer and the rest of the colleagues of the _ institut fr astrophysik _ for their hospitality and interest on this type of investigations on the physics of scattering polarization .
this research has been partially funded by the spanish ministerio de educacin y ciencia through project aya2004 - 05792 . , c. , lagg , a. , & solanki , s. 2006 , in highlights of astronomy , vol .
14 , part ii.3 : solar active regions and 3d magnetic structure , eds .
debi prasad choudhary and michal sobotka , cambridge university press , in press , j. , landi deglinnocenti , e. , casini , r. , & martnez pillet , v. 2005 , in @xmath152 eslab symposium on trends in space science and cosmic vision 2020 ( 19 - 21 april 2005 ; noordwijk , the netherlands ) , eds .
f. favata et al . ,
esa publications division , vol .
sp-588 , 203 profiles of the he i 10830 multiplet calculated for the two line of sights illustrated in fig . 1
: @xmath0 scattering ( left panels , where the stokes @xmath1 profile is normalized to the maximum line - core intensity of the stokes @xmath26 emission profile of the ` red ' line ) and forward scattering ( right panels , where the stokes @xmath1 profile is normalized to the maximum line - core depression of the stokes @xmath26 absorption profile of the ` red ' line ) ) . each panel shows the results of three possible calculations for a given strength of the assumed horizontal magnetic field , whose orientation is as shown in fig .
1 that is , such that the magnetic field vector is always perpendicular to the line of sight .
while the dotted lines neglect the influence of atomic level polarization , the solid lines take it fully into account .
the dashed lines show what happens when only the lower level of the multiplet is assumed to be completely unpolarized .
the calculations have been carried out for a thermal velocity @xmath153 . the positive reference direction for stokes
@xmath1 is along the direction of the horizontal magnetic field.,height=604 ] | the polarization observed in the spectral lines of the he i 10830 multiplet carries valuable information on the dynamical and magnetic properties of plasma structures in the solar chromosphere and corona , such as spicules , prominences , filaments , emerging magnetic flux regions , etc .
therefore , it is crucial to have a good physical understanding of the sensitivity of the observed spectral line polarization to the various competing physical mechanisms .
here we focus on investigating the influence of atomic level polarization on the emergent stokes profiles for a broad range of magnetic field strengths , in both @xmath0 and forward scattering geometry .
we show that , contrary to a widespread belief , the selective emission and absorption processes caused by the presence of atomic level polarization may have an important influence on the emergent linear polarization , even for magnetic field strengths as large as 1000 g. consequently , the modeling of the stokes @xmath1 and @xmath2 profiles should not be done by taking only into account the contribution of the transverse zeeman effect within the framework of the paschen - back effect theory , unless the magnetic field intensity of the observed plasma structure is sensibly larger than 1000 g. we point out also that in low - lying optically thick plasma structures , such as those of active region filaments , the ( horizontal ) radiation field generated by the structure itself may substantially reduce the positive contribution to the anisotropy factor caused by the ( vertical ) radiation field coming from the underlying solar photosphere , so that the amount of atomic level polarization may turn out to be negligible . only under such circumstances
may the emergent linear polarization of the he i 10830 multiplet in such regions of the solar atmosphere be dominated by the contribution caused by the transverse zeeman effect . |
in this letter , we consider the numerical integration of the partial differential equation ( pde ) in the form @xmath0 where subscripts @xmath1 or @xmath2 denote the partial differentiation with respect to @xmath1 or @xmath2 , and @xmath3 is the variational derivative of @xmath4 .
we assume the periodic boundary condition @xmath5 , where @xmath6 is a constant .
when the derivatives of @xmath7 do not appear in @xmath8 , the equation is called the ( nonlinear ) klein gordon equation in light - cone coordinates .
moreover , the class of pdes in the form is closely related to the ostrovsky equation @xcite , the short pulse equation @xcite , etc . for their numerical treatments , due to the possible indefiniteness caused by the spatial derivative in the left - hand side , it seems a systematic numerical framework for is yet to be investigated , though a few exceptions for specific cases can be found ( see , e.g. , @xcite ) . in this letter , we focus on a certain class of conservative methods . under the periodic boundary condition , the target equation has the conserved quantity @xmath8 : @xmath9^l_0 - \int^l_0 u_t u_{tx } { \mathrm{d}}x = - \int^l_0 u_t u_{tx } { \mathrm{d}}x = 0 .
\hspace{6pt } \label{eq_conservation}\end{aligned}\ ] ] note that the skew - symmetry of the differential operator @xmath10 is crucial here .
a numerical scheme is called conservative when it replicates such a conservation property ( see , e.g. , @xcite ) .
the numerical solutions obtained by such schemes are often more stable than those of general - purpose methods . there , the crucial point for the discrete conservation law is the skew - symmetry of difference operator , which corresponds to that of the differential operator ; when one tries to construct a conservative finite - difference scheme for the equation , the differential operator @xmath11 in left - hand side must be replaced by one of the skew - symmetric difference operators , for example , the central difference operators , the compact finite difference operators ( see , e.g. , kanazawa matsuo yaghchi @xcite ) , and the fourier - spectral difference operator ( see , e.g. , fornberg @xcite ) .
this is intrinsically indispensable , at least to the best of the present authors knowledge .
this , however , at the same time , leads to an undesirable side effect that the numerical solutions tend to suffer from spatial oscillations . in this letter , to work around this technical difficulty , we propose a novel `` average - difference method , '' which is tough against such undesirable spatial oscillations .
a similar method has been , in fact , already investigated by nagisa @xcite . however , he used this method for advection - type equations , and concluded the method was unfortunately not more advantageous than existing methods .
in this letter , we instead construct an average - difference method for the pde , and combine it with the idea of conservation mentioned above .
then we compare the proposed and existing methods in the case of the linear klein gordon equation , which is the simplest case with @xmath12 . as a result ,
the average - difference type method is successfully superior to the existing methods in view of the phase speed of each frequency component .
the conservative scheme for the pde can be constructed in the spirit of discrete variational derivative method ( dvdm ) ( see , the monograph @xcite for details ) . there , one utilizes the concept of the `` discrete variational derivative '' and skew - symmetric difference operators . the symbol @xmath13 denotes the approximation @xmath14 , where @xmath15 and @xmath16 are the temporal and spatial mesh sizes , respectively . here
, we assume the discrete periodic boundary condition @xmath17 , and thus , we use the notation @xmath18 .
let us introduce the spatial central difference operator @xmath19 and the temporal forward difference operator @xmath20 : @xmath21 the discrete counterpart @xmath22 of the functional @xmath23 can be defined as @xmath24 where @xmath25 is an appropriate approximation of @xmath26 .
then , the discrete variational derivative @xmath27 is defined as a function satisfying @xmath28 for the construction of such one , see @xcite . by using the discrete variational derivative
, we can construct a conservative scheme @xmath29 as stated in the introduction , the key ingredient here is the skew - symmetry of the central difference operator .
suppose the numerical scheme has a solution @xmath13 under the periodic boundary condition .
then , it satisfies @xmath30 .
_ thanks to the definition of the discrete variational derivative , we can follow the line of the discussion as follows : @xmath31 whose right - hand side vanishes due to the skew - symmetry of the central difference operator @xmath19 : @xmath32 holds for any @xmath33 .
_ the discrete conservation law can also be proved similarly for the other skew - symmetric difference operators .
in this section , we propose the novel method . there , instead of the single skew - symmetric difference operator , we employ the pair of the forward difference and average operators : @xmath34 the average - difference method for the equation can be written in the form @xmath35 the name `` average - difference '' comes from the idea of approximating @xmath11 with the pair of @xmath36 ; this makes sense for more general pdes , and thus is independent of any conservation properties .
still , in this letter we focus on and .
although it is constructed in the spirit of dvdm , now the forward difference operator @xmath37 loses the apparent skew - symmetry , and accordingly , the proof of the discrete conservation law becomes unobvious .
a similar proof can be found in nagisa @xcite .
suppose the average - difference method has a solution @xmath13 under the periodic boundary condition .
then , it satisfies @xmath30 .
[ thm_dvdm_ad ] _ by using the definition of the discrete variational derivative , we see that @xmath38 here , for brevity , we introduce the notation @xmath39 note that the equation implies the relation @xmath40 . by using the identity @xmath41 which holds for any @xmath42
, we see @xmath43 which proves the theorem . _
in order to conduct a detailed analysis , we consider the simplest case , the linear klein gordon equation @xmath44 under the periodic domain with the period @xmath45 .
the exact solution of the linear klein gordon equation can be formally written in the form @xmath46 where @xmath47 is the imaginary unit , and @xmath48 is determined by the initial condition @xmath49 : @xmath50 in view of the superposition principle , we consider the single component @xmath51 ( @xmath52 ) . in order to clarify the difference between the standard conservative method and proposed method , we consider the following three semi - discretizations @xmath53 where @xmath54 for @xmath55 ( @xmath56 ) . here , @xmath57 denotes the fourier - spectral difference operator , i.e. , @xmath58 where @xmath59 is obtained by the discrete fourier transform : @xmath60 note that , the implicit midpoint method for the semi - discretizations above coincide with the numerical schemes constructed in the previous sections .
we consider the solution of the semi - discretizations above in the form @xmath61 ( @xmath62 ) for each @xmath63 , which gives an exact solutions of , , and with appropriate choices of @xmath64 .
for the central difference scheme , we see @xmath65 if we employ the fourier - spectral difference operator instead of the central difference , we see @xmath66 and @xmath67 holds for any @xmath63 . for the average - difference scheme ,
we obtain @xmath68 the phase speeds @xmath69 corresponding to each numerical scheme are summarized in fig .
[ fig_speed ] ( @xmath70 ) . as shown in fig .
[ fig_speed ] , the phase speed of the central difference scheme are falsely too fast for high frequency components ( @xmath71 ) . on the other hand ,
the error of the phase speeds of the average - difference method are much smaller .
( @xmath72 ) .
the red circles , blue crosses , and green triangles correspond to the average - difference , fourier - spectral difference , and central difference schemes , respectively . ] in this section , we conduct a numerical experiment under the periodic boundary condition @xmath73 with the initial condition @xmath74 the corresponding solution can be formally written in @xmath75 figures [ fig_step_cd ] , [ fig_step_ps ] , and [ fig_step_ad ] show the numerical solutions of the central difference scheme , the fourier - spectral difference scheme , and the average - difference method , respectively ( the temporal discretization : implicit midpoint method ) . as shown in fig .
[ fig_step_cd ] , the central difference scheme suffers from the spatial oscillation , whereas the other schemes reproduce the smooth profiles until @xmath76 .
the cause of this difference lies on the discrepancy in phase speeds of high frequency components ( fig .
[ fig_speed ] ) . , @xmath77 ) .
] , @xmath77 ) .
] , @xmath77 ) .
] however , as shown in fig .
[ fig_comp ] , which shows the numerical solutions of each schemes at @xmath78 , the fourier - spectral scheme also suffers from the undesirable spatial oscillation , whereas the proposed method , average - difference method reproduces a better profile .
moreover , the values of error @xmath79 at @xmath80 ( i.e. , @xmath81 ) for the numerical solutions of central difference scheme , fourier - spectral difference scheme , and the proposed method are @xmath82 , @xmath83 , and @xmath84 , respectively . this could be attributed to the fact that the fourier - spectral difference can be regarded as a higher - order central difference , and thus should share the same property to a certain extent .
( @xmath85 , @xmath77 ) .
the black dotted line represents the exact solution .
the green , blue , and red solid lines denote the numerical solution of the central difference scheme , the fourier - spectral difference scheme , and the average - difference method , respectively . ]
the results above can be extended in several ways . first , instead of the cumbersome proof in theorem [ thm_dvdm_ad ]
, we can introduce the concept of generalized skew - symmetry , by which a more sophisticated `` average - difference '' version of the dvdm could be given .
second , we should try more general pdes to see to which extent the new dvdm is advantageous .
finally and ultimately , we hope to construct a systematic numerical framework for ( 1 ) , based on the above observations .
the authors have already got some results on these issues , which will be reported somewhere soon .
this work was partly supported by jsps kakenhi grant numbers 25287030 , 26390126 , and 15h03635 , and by crest , jst .
the second author is supported by the jsps research fellowship for young scientists .
99 l. a. ostrovsky , nonlinear internal waves in the rotating ocean , okeanologia * 18 * ( 1978 ) , 181191 .
t. schfer and c. e. wayne , propagation of ultra - short optical pulses in cubic nonlinear media , phys .
d , * 196 * ( 2004 ) , 90105 .
t. yaguchi , t. matsuo and m. sugihara , conservative numerical schemes for the ostrovsky equation , j. comput .
, * 234 * ( 2010 ) , 10361048 . y. miyatake , t. yaguchi and t. matsuo , numerical integration of the ostrovsky equation based on its geometric structures , j. comput . phys . ,
* 231 * ( 2012 ) , 45424559 .
d. furihata and t. matsuo , discrete variational derivative method : a structure - preserving numerical method for partial differential equations , crc press , boca raton , 2011 .
e. celledoni , v. grimm , r. i. mclachlan , d. i. mclaren , d. oneale , b. owren and g. r. w. quispel , preserving energy resp .
dissipation in numerical pdes using the `` average vector field '' method , j. comput .
* 231 * ( 2012 ) , 67706789 .
h. kanazawa , t. matsuo and t. yaguchi , a conservative compact finite difference scheme for the kdv equation , jsiam letters * 4 * ( 2012 ) , 58 .
b. fornberg , a practical guide to pseudospectral methods , cambridge university press , cambridge , 1996 .
y. nagisa , finite difference schemes for pdes using shift operators ( in japanese ) , the university of tokyo , master s thesis , 2014 . | we consider structure - preserving methods for conservative systems , which rigorously replicate the conservation property yielding better numerical solutions . there , corresponding to the skew - symmetry of the differential operator , that of difference operators is essential to the discrete conservation law .
unfortunately , however , when we employ the standard central difference operator , the simplest one , the numerical solutions often suffer from undesirable spatial oscillations . in this letter
, we propose a novel `` average - difference method , '' which is tougher against such oscillations , and combine it with an existing conservative method .
theoretical and numerical analysis in the linear case show the superiority of the proposed method . |
in the last few decades several aspects of cuprate superconductors ( scs ) have been studied and many efforts have been spent to understand the unique and complex phase diagram exhibited by this class of materials ; see , e.g. , ref . .
superconductivity in cuprates appears to be due to a mechanism analogous to the bcs one in ordinary superconductors . however , superconductivity is only one of the characteristic features of these materials .
there are many other new properties that require more complex mechanisms and can be understood only if the interplay between bcs and additional order parameters is considered .
for instance , at @xmath3 , @xmath4 at very low doping @xmath5 is an insulator with long - range magnetic order .
increasing @xmath5 , at @xmath6 an insulator - superconductor first - order transition takes place , giving rise to a superconducting state in which spins are still magnetically ordered @xcite . at @xmath7
another phase transition occurs , and , for @xmath8 , the material shows no magnetic order
it is paramagnetic but is still superconducting .
neutron - scattering experiments@xcite suggested that this transition is continuous .
moreover , in the ordered phase @xmath9 , they revealed the presence of collinearly polarized spin - density waves ( sdws ) with wavevectors @xmath10 where @xmath11 is a function of the doping concentration and @xmath12 is the lattice spacing .
the wave vectors @xmath13 are two - dimensional since cuprates are supposed to be made of weakly interacting planes and thus behave approximately as two - dimensional systems .
following ref . , we assume that superconductivity is not relevant at the transition which is instead driven by the interaction among the sdw degrees of freedom
. since @xmath3 one should take into account the quantum nature of the system .
quantum phase transitions can be studied by introducing a supplementary dimension parametrized by an imaginary time variable @xmath14 .
the relevant order parameter is the spin field which is parametrized as @xmath15,\ ] ] where @xmath16 are complex amplitudes .
there are two interesting limiting cases .
the first one is when the order parameter can be written as @xmath17 which corresponds to collinearly polarized sdws .
the second one is when @xmath18 , with @xmath19 and @xmath20 , which corresponds to circularly polarized sdws . in cuprates experiments
indicate that the ground state shows a collinear behavior.@xcite the standard strategy for writing down an effective hamiltonian for a given physical system consists in considering all polynomials of the order parameter of order less than or equal to four that are compatible with the expected symmetries . in the sdw - sc
to sc phase transition the order parameter is the complex field @xmath21 , with @xmath22 and @xmath23 .
the corresponding symmetries are the following : ( i ) so(3 ) spin rotations : @xmath24 ; ( ii ) translational symmetry of the spin waves : @xmath25 ; ( iii ) spatial inversion : @xmath26 ;
( iv ) interchange of the @xmath27 and @xmath28 axes : @xmath29 and @xmath30 .
the most general hamiltonian with these symmetries is@xcite @xmath31 where @xmath32 and @xmath33 are parameters called sdw velocities .
terms such as @xmath34 are forbidden by spatial inversion symmetry and terms like @xmath35 , even if permitted by all symmetries , can be eliminated by redefining the fields as @xmath36 .
hamiltonian ( [ lgwhor ] ) admits several different ground states depending on the values of the parameters .
they are classified in app .
in particular , there is the possibility that both fields correspond to collinearly polarized sdws as observed in experiments : @xmath37 and @xmath38 , where the vectors @xmath39 and @xmath40 satisfy either @xmath41 or @xmath42 . in this paper
we investigate the nature of the fixed points ( fps ) of the renormalization - group ( rg ) flow of the effective hamiltonian ( [ lgwhor ] ) .
if a stable fp exists and its attraction domain includes systems with collinearly polarized sdws , then the sdw - sc to sc transition may be continuous .
otherwise , it must be of first order . in our study
, we consider only the case @xmath43 that simplifies the analysis and allows us to perform a high - order perturbative analysis .
therefore , we consider the theory @xmath44 where the field @xmath16 is a complex @xmath45 matrix , @xmath22 , @xmath46 .
the physically relevant case is @xmath47 .
we first perform a standard analysis close to four dimensions,@xcite computing the rg functions in powers of @xmath48 .
a one - loop analysis indicates that a stable fp exists only for @xmath49 .
apparently , this result casts doubts on the existence of a stable fp in three dimensions .
however , in three dimensions there may exist fps that are absent for @xmath50 .
this is indeed what happens in the ginzburg - landau model of superconductors , in which a complex scalar field couples to a gauge field@xcite and in o(2)@xmath51o(@xmath52 ) symmetric models.@xcite thus , a more careful investigation of the rg flow in three dimensions calls for strictly three - dimensional perturbative schemes . for this purpose
we consider two field - theoretical perturbative approaches : the minimal - subtraction scheme without @xmath53 expansion@xcite ( in the following we will indicate it as @xmath54-@xmath55 scheme ) and the massive zero - momentum ( mzm ) renormalization scheme.@xcite the use of two different schemes is crucial , since the comparison of the corresponding results provides a nontrivial check on the reliability of our conclusions . in the @xmath54-@xmath55 scheme
one considers the massless ( critical ) theory in dimensional regularization,@xcite determines the rg functions from the divergences appearing in the perturbative expansion of the correlation functions , and finally sets @xmath56 without expanding in powers of @xmath53 ( this scheme therefore differs from the standard @xmath53 expansion@xcite ) . in the mzm scheme
one considers instead the three - dimensional massive theory in the disordered ( high - temperature ) phase .
we compute the @xmath57 functions to five loops in the @xmath54-@xmath55 scheme and to six loops in the mzm scheme .
we use a symbolic manipulation program that generates the diagrams ( approximately one thousand at six loops ) and computes their symmetry and group factors , and the compilation of feynman integrals of refs . .
the series are available on request .
the perturbative expansions are then resummed using the known large - order behavior .
the perturbative analysis of the rg flow in the full theory is not sufficiently stable to provide reliable results .
therefore , we have focused on the stability of the fps that occur in specific submodels of hamiltonian ( [ lgwh ] ) .
the analysis of the perturbative series indicates the stability of the @xmath0 collinear fp that occurs in the model with @xmath58 and @xmath59 .
moreover , its basin of attraction includes systems with collinear sdws .
therefore , we expect the continuous transition observed experimentally in cuprates to belong to this universality class .
this implies an effective enlargement of the symmetry at the transition point
. the corresponding critical exponents would be @xmath60 the paper is organized as follows . in sec .
[ sec2 ] we discuss the possible ordered phases that occur in model ( [ lgwhor ] ) in the mean - field approximation .
details are given in app .
[ appa ] . in sec .
[ 4drgflow ] we discuss the fp structure close to four dimensions in the standard @xmath53 expansion .
[ substa ] contains the main results of this work .
we consider three different submodels ( sec .
[ submodels ] ) and then investigate the stability properties of the fps occuring in each of them ( sections [ stabo2on ] , [ stabo4on ] , and [ stabmn ] ) .
conclusions are presented in sec .
[ conclusions ] . in app .
[ appo4on ] and [ appmn ] we give some technical details .
the phase diagram of hamiltonian ( [ lgwhor ] ) can be studied in the mean - field approximation . due to the large number of couplings
the analysis is quite complex .
we have limited our considerations to the case @xmath61 .
we summarize here the results that are derived in app .
[ appa ] . for @xmath62
the system is disordered and @xmath63 . for @xmath64
a continuous phase transition occurs followed by a magnetized phase with @xmath65 .
the nature of the ordered phase depends on the values of the quartic parameters .
the analysis reported in app .
[ appa ] shows that there are seven possibilities : * @xmath66 is a collinear sdw ( @xmath67 , @xmath68 real ) while @xmath69 .
* @xmath66 is a circularly polarized sdw ( @xmath70 , @xmath39 and @xmath40 real , @xmath71 , @xmath42 ) while @xmath69 . *
@xmath66 and @xmath72 correspond to collinear sdws with the same axis and amplitude : @xmath67 , @xmath73 , @xmath74 real .
* @xmath66 and @xmath72 correspond to collinear sdws with orthogonal axes and same amplitude : @xmath37 , @xmath38 , @xmath75 real , @xmath42 , @xmath71 .
* @xmath66 and @xmath72 are circularly polarized sdws with the same rotation plane and amplitude : @xmath70 , @xmath76 , @xmath39 and @xmath40 real , @xmath71 , @xmath42 .
* @xmath66 is a collinear sdw and @xmath72 is a circularly polarized sdw .
the rotation plane of @xmath72 is orthogonal to the axis of @xmath66 .
explicitly : @xmath37 , @xmath77 , @xmath78 real , @xmath79 , @xmath80 . * @xmath66 and @xmath72 are elliptically polarized sdws with different rotation planes but with the same amplitude , @xmath81 . for cuprates the relevant solutions are ( 3 ) and ( 4 ) .
necessary conditions to obtain ( 3 ) are @xmath82 and @xmath83 , \label{cond21}\ ] ] while ( 4 ) requires @xmath84 and @xmath85 . \label{cond22}\ ] ] these conditions are not sufficient , since for some values of the parameters satisfying eqs .
( [ cond21 ] ) or ( [ cond22 ] ) the ordered phase is given by solutions ( 6 ) or ( 7 ) .
note that the sign of @xmath86 is not the relevant parameter that selects the collinear sdws among all possible solutions .
it is interesting to note that the mean - field solution predicts either @xmath87 or @xmath88 in the case of collinear sdws .
this result is easy to understand .
if both fields correspond to collinear sdws , then one can take @xmath66 and @xmath72 real . in this case the only term of the hamiltonian that contains a scalar product of the two fields is @xmath89 that forces the two fields to be either parallel or orthogonal , depending on the sign of @xmath90 .
note that this also holds if we add additional higher - order terms to the hamiltonian , as long as the transition is continuous .
indeed , for a continuous transition @xmath91 at the transition ( @xmath92 in the disordered phase ) and thus higher - order terms do not play any role . on the other hand , this relation may not be valid if the transition is of first order .
also the coupling to the charge - density waves ( cdws ) that are present in cuprates@xcite does not change this conclusion , since they couple to the scalars @xmath93 , @xmath94 .
solutions ( 3 ) and ( 4 ) also satisfy @xmath81 . this property does not necessarily hold if we take into account the cdws ( see refs . for an extensive discussion ) .
indeed , let @xmath95 and @xmath96 be the complex amplitudes of the cdws coupled respectively to @xmath97 and @xmath98 . in the absence of the cdw - sdw coupling , for some values of the cdw hamiltonian parameters
, the ordered solution corresponds to @xmath99 , @xmath100 .
if now the cdw - sdw coupling is included , one may obtain a ground state with @xmath101 and @xmath102 .
the rg flow close to four dimensions can be investigated perturbatively in @xmath103 . in the minimal - subtraction ( @xmath55 )
the one - loop @xmath57 functions are : @xmath104 where @xmath105 are the renormalized quartic couplings corresponding to the quartic hamiltonian parameters @xmath106 , @xmath107 .
they are normalized so that , at tree level , @xmath108 , where @xmath109 and @xmath110 label the renormalized and hamiltonian parameters respectively and @xmath111 .
the fps of the rg flow are the common zeroes of the @xmath57 functions . for @xmath47 there are 4 fps while for @xmath112 there are 7 fps : they are all unstable . only for @xmath49
does a stable fp exist .
it has @xmath113 ( for @xmath114 we obtain @xmath115 , @xmath116 ) , so that at the fp the symmetry becomes @xmath117 .
this fp is the chiral fp that occurs in @xmath118 in the large-@xmath119 limit.@xcite in order to determine the behavior in three dimensions , one should extend the computation to higher order in @xmath53 and determine the function @xmath120 such that the chiral fp point identified above exists for @xmath121 and is no longer present for smaller values of @xmath119 .
we have not pursued this approach for several reasons .
first , the analogous five - loop computation that was performed in the @xmath122 model@xcite was not able to explain the correct physics of these models for @xmath123 ( see sec .
ii.d in ref . ) .
moreover , this calculation is only concerned with the stable fp that is present for @xmath124 ( in the present case the chiral @xmath125 fp ) , while in @xmath126 the stable fp may be different , an unstable or even a new fp .
the analysis that will be presented in the next section favors this last possibility .
the three - dimensional properties of the rg flow are determined by its fps .
some of them can be identified by considering particular cases in which some of the quartic parameters vanish .
the corresponding fps are also fps of the general theory . in this section ,
we identify some of them , and then determine their stability with respect to the complete theory . for particular values of the couplings hamiltonian ( [ lgwh ]
) reduces to that of simpler models .
three cases have already been extensively studied in the literature:@xcite * for @xmath127 there is no interaction between the two sdws and hamiltonian ( [ lgwh ] ) reduces to that of two identical decoupled o(2)@xmath128-symmetric models .
the general @xmath129-symmetric model is defined by the hamiltonian density@xcite @xmath130 + { g_{1,0}\over 4 ! }
\bigl ( \sum_{ai } \phi_{ai}^2\bigr)^2 \nonumber \\ & & \qquad\qquad + { g_{2,0}\over 4 ! } \left [ \sum_{i , j } \bigl ( \sum_a \phi_{ai } \phi_{aj}\bigr)^2 - \bigl(\sum_{ai } \phi_{ai}^2 \bigr)^2 \right ] , \label{hch}\end{aligned}\ ] ] where @xmath131 is a real @xmath132 matrix field ( @xmath133 and @xmath134 ) .
hamiltonian ( [ hch ] ) is obtained from eq .
( [ lgwh ] ) by setting @xmath135 and @xmath136 the properties of @xmath137 models are reviewed in refs . .
in three dimensions perturbative calculations within the mzm scheme@xcite and within the @xmath54-@xmath55 scheme@xcite indicate the presence of a stable chiral fp with attraction domain in the region @xmath138 for all values of @xmath119 ( only for @xmath139 the evidence is less clear , since the mzm analysis does not apparently support it ) . for @xmath112 , these conclusions have been recently confirmed by a monte carlo simulation.@xcite a stable collinear fp for @xmath140 exists for @xmath141.@xcite apart from the collinear fp for @xmath112 , these fps do not exist close to four dimensions . for @xmath112
the collinear fp is equivalent to an xy fp and corresponds to @xmath142 , @xmath143 , where @xmath144 is the fp value of the renormalized coupling in the o(2 ) @xmath145 model .
* for @xmath146 and @xmath147 , hamiltonian ( [ lgwh ] ) reduces to ( [ hch ] ) with @xmath148 and @xmath149 .
the correspondence is given by @xmath150 where @xmath151 is a @xmath152 matrix , and @xmath153 we have already discussed the fps of the @xmath154 theory . the @xmath155 theory does not present stable fps for @xmath156.@xcite analyses of the available six - loop series in the mzm scheme and five - loop series in the 3-@xmath157 @xmath55 scheme indicate the presence of a stable collinear fp for @xmath140.@xcite this fp does not exist close to four dimensions . * for @xmath158
we obtain the @xmath159 model with @xmath160 and @xmath161 .
the so - called @xmath159 model is defined by the hamiltonian density@xcite @xmath162 + { g_{1,0}\over 4 ! } \bigl ( \sum_{ai } \phi_{ai}^2\bigr)^2 + { g_{2,0}\over 4 ! } \sum_{aij } \phi_{ai}^2 \phi^2_{aj } , \label{hmn}\end{aligned}\ ] ] where @xmath131 is a real @xmath132 matrix , i.e. , @xmath133 and @xmath134 .
the correspondence is obtained by setting @xmath163 a stable fp is the o@xmath164 fp with @xmath165 and @xmath166 , where @xmath167 is the fp value of the renormalized coupling in the @xmath168-symmetric vector model . in three dimensions the analysis of five- and six - loop series@xcite indicates the presence of a second stable fp with @xmath169 for @xmath170 and @xmath171 , 3 , and 4 . beside these three models ,
there are two other submodels for which no results are available : * for @xmath172 we obtain two chiral models coupled by an energy - energy term .
note that in this model the rg flow does not cross the planes @xmath173 and @xmath174 . * for @xmath175
we obtain a model with an additional u(1 ) symmetry : @xmath176 , @xmath177 . in this model
the rg flow does not cross the plane @xmath178 . finally , note an additional symmetry of hamiltonian ( [ lgwh ] ) .
it is invariant under @xmath176 , @xmath179 , @xmath180 and @xmath181 , while the other couplings are unchanged .
this implies that the rg flow in the space of renormalized couplings does not cross the plane @xmath182 and that , for any fp with @xmath183 there is an equivalent one with @xmath184 .
in particular , we can limit our considerations to @xmath185 . in order to study
the rg flow of the theory one can start by discussing the stability in the full theory of the fps of the models ( 1 ) , ( 2 ) , and ( 3 ) discussed above . for @xmath112 and @xmath47 ,
the only cases we consider , model ( 1 ) has two fps : * the chiral fp , in which @xmath186 and @xmath187 ; correspondingly @xmath188 , @xmath189 , @xmath190 ; * the collinear fp , in which @xmath191 and @xmath192 ; correspondingly @xmath193 , @xmath194 , @xmath190 . here @xmath195 and @xmath196 are the chiral and collinear fps of the o(2)@xmath51o(@xmath119 ) theory .
the analogous fps are present in model ( 2 ) : * the chiral fp , in which @xmath186 and @xmath187 ; correspondingly @xmath188 , @xmath197 , @xmath198 , @xmath199 ; it does not exist for @xmath200 .
this is the fp that is relevant for @xmath201 close to four dimensions ; * the collinear fp , in which @xmath191 and @xmath192 ; correspondingly @xmath193 , @xmath194 , @xmath198 , @xmath199 .
it exists for both @xmath112 and @xmath47 . here
@xmath195 and @xmath196 are the chiral and collinear fps of the o(4)@xmath51o(@xmath119 ) theory . finally , the @xmath159 theory gives two fps : * the o(2@xmath119 ) fp .
this is unstable in the full theory , being already unstable in model ( 1 ) ; * the @xmath159 fp @xmath202 , @xmath203 ; correspondingly @xmath204 , @xmath205 , @xmath206 .
it exists only for @xmath112 . in the following
we study the stability of these fps in the complete theory ( [ lgwh ] ) .
for this purpose , using the @xmath57 functions of the general theory we have computed the stability matrices of the fps at six and five loops respectively in the mzm and @xmath54-@xmath55 schemes .
the perturbative series have been resummed by using the conformal - mapping method described , e.g. , in ref . .
for a fp belonging to a submodel , the large - order behavior needed for the conformal - mapping summation is the same as that characterizing all series of that submodel .
for all submodels we consider , the large - order behavior is already known.@xcite we want to establish the stability properties of the decoupled @xmath137 fps ( 1a ) and ( 1b ) in the complete theory ( [ lgwh ] ) . for this purpose
we need the rg dimensions of the operators present in hamiltonian ( [ lgwh ] ) that break the symmetry of model ( 1 ) , i.e. , of the operators associated with the quartic couplings @xmath207 .
it is useful to rewrite them as @xmath208 where @xmath209 @xmath210 and , using the correspondence ( [ corre ] ) , @xmath211 the quadratic operator @xmath212 transform as a spin-@xmath213 and a spin-@xmath214 representation with respect to the @xmath215 and @xmath216 groups , respectively .
since @xmath217 , @xmath218 , and @xmath219 belong to different irreducible representations , they do not mix under rg transformations at the decoupled @xmath137 fps .
their rg dimensions @xmath220 can be derived from the rg dimensions @xmath221 of the quadratic operators
@xmath212 at the @xmath137 fp , using the relation @xmath222 the quadratic term @xmath223 corresponds to the energy operator and thus @xmath224 and @xmath225 , where @xmath226 and @xmath227 are the specific - heat and correlation - length critical exponents of the given @xmath137 fp .
the rg dimensions @xmath228 and @xmath229 were computed in ref .
( there , they are named @xmath230 and @xmath231 respectively ) . at the chiral fp ( 1a )
we obtain:@xcite @xmath232 at the collinear fp ( 1b ) we obtain:@xcite @xmath233 these results show that the decoupled @xmath137 fps are unstable in the complete theory ( [ lgwh ] ) for both @xmath234 .
it is also interesting to discuss submodels ( a ) and ( b ) mentioned in sec .
[ submodels ] . in model ( a ) one should only consider @xmath217 . the numerical results apparently indicate that the fps are always unstable ( but , with the present errors , we can not really exclude the opposite possibility ) , except in one case . for @xmath112 ,
the collinear fp is stable . in model ( b )
one should consider @xmath217 and @xmath219 .
for @xmath123 , all fps are unstable . here
we investigate the stability of fps ( 2a ) ( it does not exist for @xmath47 ) and ( 2b ) .
for this purpose we must compute the rg dimensions of the perturbations of the @xmath117 model appearing in the complete theory .
this is done in app .
[ appo4on ] .
there are two relevant operators with rg dimensions @xmath235 and @xmath236 .
the corresponding perturbative series are reported in app .
[ appo4on ] .
they are analyzed using the conformal mapping method.@xcite the errors we will report takes into account the variation of the estimates when changing the resummation parameters @xmath237 defined in ref .
we use @xmath238 and @xmath239and the uncertainty of the fp coordinates .
the analysis of the six - loop series in the mzm scheme and of the five - loop 3@xmath157-@xmath55 series gives the following results at the collinear fp ( 2b):@xcite @xmath240 for @xmath47 the mzm and @xmath54-@xmath55 results are consistent and apparently indicate that @xmath235 and @xmath236 are negative , though with somewhat large errors .
a better understanding of the relevance of the two operators can be obtained from fig .
[ fig : distribution ] , where we give the distributions of the estimates of @xmath235 and @xmath236 obtained by varying the parameters @xmath226 and @xmath241 . for @xmath235 low - order calculations
predict @xmath242 .
however , as the number of loops increases , @xmath235 decreases .
the six - loop mzm results indicate that @xmath243 , a result that is also supported by the trend observed in the @xmath54-@xmath55 results . as for @xmath236 ,
the mzm results clearly indicate @xmath244 .
however , this is not fully confirmed by the @xmath54-@xmath55 results .
though they give @xmath244 , there is a trend towards larger values of @xmath236 .
overall , these results support the stability of the collinear fp ( 2b ) in the complete theory for @xmath200 .
similar conclusions hold for @xmath112 . for completeness , we report here the corresponding critical exponents:@xcite @xmath245 for @xmath112 we also study the stability of the chiral fp ( 2a ) . using the results of refs . for the fp we have:@xcite @xmath246
the chiral fp is clearly unstable .
finally , note that the same discussion also applies to submodel ( b ) , since the stability of the fp depends on the same operators with rg dimensions @xmath235 and @xmath236 . for submodel
( a ) one should only consider @xmath235 . in this case
also the chiral fp ( 2b ) might be stable .
here we investigate the stability of fp ( 3b ) for @xmath112 ( it does not exist for @xmath47 ) .
for this purpose we must compute the rg dimensions of the perturbations of the @xmath159 fp appearing in the complete theory .
this is done in app .
[ appmn ] .
there are two relevant operators with rg dimensions @xmath235 and @xmath236 .
the analysis of the perturbative series in the mzm scheme gives @xmath247 the results in the @xmath54-@xmath248 scheme are very imprecise , although negative values for @xmath235 and @xmath236 seem to be favored .
there results indicate , although with limited confidence , that the @xmath159 fp present for @xmath112 may be stable in the complete theory .
in this paper we have studied the quantum phase transition that occurs in two - dimensional systems that exhibit an ordered phase with sdw order .
the effective hamiltonian of the relevant critical modes @xmath16 is given in eq .
( [ lgwhor ] ) .
a detailed mean - field analysis shows that in some parameter region hamiltonian ( [ lgwhor ] ) has a continuous transition separating a spin disordered phase from an ordered phase characterized by two collinearly polarized sdws .
there are two different possibilities for the the axes of these sdws : either @xmath87 or @xmath88 .
we have then investigated the role of fluctuations in a simplified model in which the two sdws have the same velocity . for this purpose
we have generated six - loop perturbative series in the mzm scheme and five - loop series in dimensional regularization with minimal subtraction .
close to four dimensions , an analytic @xmath53-expansion calculation shows no presence of stable fps .
however , past experience indicates that a fp may exist in three dimensions and be absent for @xmath249 .
therefore , we have considered two strictly three - dimensional schemes .
we have analyzed the stability of some fps that belong to known submodels .
the analysis of the perturbative series supports the stability of the @xmath250 collinear fp .
the analyses of the mzm and @xmath54-@xmath55 expansions do not provide sufficiently stable results for the rg flow in the full theory , i.e. in the general space of its five quartic couplings .
in particular , they do not allow us to draw any definite conclusion on the existence of other stable fps . in any case , even without the analysis of the full flow , simple considerations ( reported in app .
[ appa ] ) show that systems with collinear sdws with the same axis ( the mean - field solution ( 3 ) reported in sec .
[ sec2 ] ) are in the attraction domain of the @xmath250 collinear fp .
it should be remarked that our rg analysis is only valid for @xmath43 . in order to extend the results to the generic case @xmath251 one
should also consider the operator @xmath252 and determine its rg dimension @xmath253 at the @xmath0 collinear fp . if @xmath254 the previous conclusions are unchanged . on the other hand , if @xmath255 the @xmath0 collinear fp is unstable with respect to the perturbation @xmath256 . in this case
the transition may be of first order or continuous depending on the existence and attraction domain of a stable fp with @xmath257 .
note that , from a practical point of view , our results are of interest even if @xmath255 .
indeed , one expects the sdw velocities @xmath32 and @xmath33 to be close in magnitude , of the order of the spin - wave velocity of the nel state of the undoped insulator.@xcite therefore , the rg flow always starts very close to the stable fp of the theory with @xmath43 , and thus the critical behavior is controlled by this fp except in a narrow interval around the critical doping .
experiments indicate that the sdw - sc to sc transition is continuous and is associated with collinear sdws .
it is thus natural to conjecture that its critical behavior is controlled by the @xmath0 collinear fp , since this fp is stable in model ( [ lgwh ] ) and its basin of attraction includes systems with collinear sdws .
the corresponding critical exponents are then predicted to be @xmath1 , @xmath258 .
mdp acknowledges that part of this work was done at the university of roma tre as part of his phd thesis .
in this appendix we compute the possible ground - state configurations of hamiltonian ( [ lgwh ] ) , that allow us to identify the possible symmetry - breaking patterns .
we consider translation - invariant configurations and the space - independent hamiltonian density @xmath259 where @xmath260 is the part of the hamiltonian that is quartic in the fields .
since @xmath261 for stability , for @xmath62 the ground state always corresponds to @xmath63 . for @xmath65 , @xmath63 is a local maximum of @xmath262 and thus the ground state is nontrivial .
the value @xmath64 corresponds to a second - order transition point in the mean - field approximation . in order to determine the ground states for @xmath65
, we will first determine all stationary points of @xmath262 ; the ground state is the one with the lowest energy .
note that , if @xmath263 , @xmath264 is a stationary point , then @xmath265 this relation is quite general .
indeed , assume @xmath262 to be of the form @xmath266 then @xmath267 on a stationary solution , the derivative vanishes , proving eq .
( [ hstat ] ) .
the calculation of the ground states also allows us to determine the stability domain of the hamiltonian .
indeed , a point in the coupling space does not belong to the stability domain if there is a field such that @xmath268 .
being @xmath260 homogeneous , it is not restrictive to consider only fields such that @xmath269 . thus , the determination of the minima of @xmath260 is equivalent to the determination of the minima of @xmath270 where @xmath271 is now interpreted as a lagrange multiplier .
( [ hstat ] ) shows that @xmath260 can be negative only for @xmath62 .
thus , the stability domain of @xmath260 is obtained by determining the stationary points of @xmath262 for @xmath271 _
positive_. in order to determine the minima , we can use the symmetry of the hamiltonian . using the o(@xmath119 ) symmetry
we can always write @xmath272 then , by using the u(1 ) symmetry we can also fix @xmath273 .
indeed , we first perform an o(2 ) rotation on the first two components : @xmath274 where @xmath275 is either @xmath276 or @xmath277 . then , we apply a u(1 ) rotation , @xmath278 .
if we choose @xmath279 the transformed field has the form ( [ phi1-gauge ] ) with @xmath273 .
once @xmath66 has been fixed we can use o@xmath280 and u(1 ) rotations to write @xmath281 if @xmath282 , one can use u(1 ) rotations to set @xmath283 .
the analysis of the minima is nontrivial due to the complexity of the stationarity equations .
we have only consider the case @xmath282 that is relevant experimentally .
other ground states are present for @xmath284 .
we found seven relevant minima ( only five of them occur for @xmath285 ) : * @xmath286 , @xmath287 .
* @xmath288 , @xmath289 . *
@xmath290 , @xmath291 .
* @xmath292 , @xmath293 . *
@xmath294 , @xmath295 . *
@xmath296 , @xmath297 , @xmath298 , @xmath299 . *
@xmath300 , @xmath301 , @xmath302 , @xmath303 , with energy @xmath304/\delta_7 $ ] , with @xmath305 .
alternatively , if we define the four vectors @xmath306 , @xmath307 , @xmath308 , @xmath309 , and @xmath310 , the solution can be characterized more geometrically as follows : @xmath311 , @xmath312 , @xmath313 , and @xmath314 , with @xmath315 whenever a component is not explicitly written , it vanishes . moreover , we defined @xmath316 , @xmath317 , @xmath318 and we simplified the notation writing @xmath319 instead of @xmath320 , etc . beside the seven solutions reported above , for @xmath321 we also found stationary points with @xmath322 and @xmath323 a numerical analysis indicates that they are never absolute minima of the hamiltonian and thus they are never relevant for the ground - state calculation .
for this reason , these solutions have not been included above .
the computation of all stationary points is quite straightforward , except for solution 7 .
we shall now briefly sketch how it is derived .
assume that @xmath324 and @xmath325 and define @xmath326 , etc .
then , @xmath327 using this relation , we can rewrite @xmath328 , @xmath329 , @xmath330 , and @xmath331 as linear equations in @xmath332 , @xmath333 , @xmath334 , @xmath335 , and @xmath336 . considering also @xmath337 that follows from eq .
( [ eq1:appa ] ) , we obtain a system of equations that allows us to determine all components .
given the list of solutions , we can determine the stability domain of the hamiltonian .
using solutions 1 - 5 , we obtain the necessary conditions @xmath338 these conditions are sufficient for @xmath112 . for @xmath339 we must also consider solutions 6 and 7 .
solution 6 gives the necessary condition @xmath340 numerically , we find that solution 7 is also relevant for stability , although we have not been able to write down an easy condition . for cuprates the relevant solutions are 3 and 4 .
in view of the possibility that the @xmath341 fp is stable it is important to understand to which ground state of the @xmath341 hamiltonian ( [ hch ] ) they correspond . for generic @xmath52 and @xmath213 , @xmath342 , model ( [ hch ] ) is stable for @xmath343 and @xmath344 and has two ground states depending on the sign of @xmath345 : for @xmath138 the ground state is chiral , while for @xmath346 the ground state is collinear.@xcite the corresponding energies are : @xmath347 } & \qquad\qquad \hbox{(chiral ) } ; \\
\vphantom{a } & \\ h = - \displaystyle{3 r^2 \over 2 g_{1,0 } } & \qquad\qquad \hbox{(collinear)}. \end{array}\ ] ] using @xmath348 , @xmath349 , @xmath350 , @xmath351 , @xmath352 , we immediately see that for @xmath148 and @xmath160 and @xmath353 solutions 1 and 3 correspond to the collinear case . for @xmath353 solutions 6 and 7
correspond to the chiral case , while solutions 2 , 4 , and 5 correspond to a stationary state that is never a ground state in the chiral theory .
for @xmath160 instead , solutions 2 , 4 , and 5 are those corresponding to the chiral case .
this result is relevant to identify the attraction domain of the @xmath341 collinear fp in the full theory .
indeed , it shows that the attraction domain of this fp includes systems whose ground state is given by solutions 1 and 3 ( and therefore two collinearly polarized sdws ) .
nothing can be said on the other solutions : in this case an analysis of the rg flow of the full theory is needed .
we need to classify the operators that break @xmath354 this is essentially discussed in ref . .
there are , however , two differences : first , we have only so(2 ) symmetry , instead of o(2 ) symmetry ; second , there is an additional exchange symmetry that forbids the appearance of spin-2 operators . in the notations of ref
( @xmath355 and @xmath119 of ref .
correspond to @xmath119 and 4 respectively ) we define @xmath356 where @xmath151 is the real field defined in eq .
( [ o4oncorr ] ) .
note that @xmath357 would be forbidden if we had o(2 ) invariance instead of so(2 ) invariance .
moreover , @xmath358 and @xmath357 correspond to different components of the same operator , so that they have the same rg dimension .
hamiltonian ( [ lgwh ] ) can then be written as @xmath359 + t_1 ( \sum_a \phi_{a}^2 ) ^2 + t_2\sum_{a , b } \left [ ( \phi_a \cdot \phi_b)^2 -\phi_a^2 \phi_b^2 \right ] \nonumber \\ & & + t_3 p_1 + t_4 p_2 + t_5 p_3,\end{aligned}\ ] ] where @xmath360 since all operators are irreducible with respect to @xmath125 transformations , if the couplings belong to the @xmath125 theory , the stability matrix defined with respect to the couplings @xmath361 has the form @xmath362 here @xmath363 is the rg dimension of @xmath364 and @xmath365 is the rg dimension of @xmath366 .
in the mzm scheme for @xmath112 we find : @xmath367 @xmath368 for @xmath47 we obtain : @xmath369 @xmath370 in the @xmath54-@xmath55 we find for @xmath112 : @xmath371 @xmath372 and finally for @xmath47 : @xmath373 @xmath374
the analysis of the perturbations at the @xmath159 fp is quite simple . in our case @xmath376 ,
@xmath170 and the relevant symmetry group is o(@xmath377 ) , which is broken by the terms proportional to @xmath378 , @xmath379 , and @xmath380 . if @xmath131 is the field defined in eq .
( [ corr - mn ] ) , @xmath22 , @xmath381 , we define the following spin-2 and spin-4 operators that transform irreducibly under o(@xmath377 ) : @xmath382 where @xmath383 . then , the relevant operators are : @xmath384 where @xmath385 and @xmath386 .
these operators give rise to different breakings of o(@xmath387 ) : @xmath388\oplus[{\rm o}(n)\otimes{\rm o}(2 ) ] \stackrel{p_2}{\longrightarrow}{\rm o}(n)\otimes[{\rm o}(2)\oplus{\rm o}(2 ) ] \nonumber \\ & \stackrel{p_3}{\longrightarrow}&{\rm o}(n)\otimes{\rm s}[{\rm o}(2)\oplus{\rm o}(2)].\end{aligned}\ ] ] in terms of @xmath389 , @xmath358 , and @xmath357 hamiltonian ( [ lgwh ] ) can then be written as @xmath359 + t_1 \sum_a ( \phi_{a}^2)^2 + t_2 \phi_1 ^ 2 \phi_2 ^ 2 \nonumber \\ & & + t_3 p_1 + t_4 p_2 + t_5 p_3,\end{aligned}\ ] ] where @xmath390 since all operators are irreducible with respect to @xmath391 transformations , if the couplings belong to the @xmath159 theory , the stability matrix defined with respect to the couplings @xmath361 has the form @xmath362 note that two eigenvalues are degenerate , since @xmath358 and @xmath357 are different components of the same irreducible operator @xmath392 .
the corresponding rg dimensions are @xmath363 , @xmath365 .
s. sachdev , rev .
* 75 * , 913 ( 2003 ) .
s. wakimoto , g. shirane , y. endoh , k. hirota , s. ueki , k. yamada , r.j .
birgeneau , m.a .
kastner , y.s .
lee , p.m. gehring , and s.h .
lee , phys . rev .
b * 60 * , r769 ( 1999 ) ; s. wakimoto , r.j .
birgeneau , y.s .
lee , and g. shirane , phys .
b * 63 * , 172501 ( 2001 ) .
g. aeppli , t.e .
mason , s.m .
hayden , h.a .
mook , and j. kulda , science * 278 * , 1432 ( 1997 ) .
y. zhang , e. demler , and s. sachdev , phys .
b * 66 * , 094501 ( 2002 ) .
lee , r.j .
birgeneau , m.a .
kastner , y. endoh , s. wakimoto , k. yamada , r.w .
erwin , s .- h .
lee , and g. shirane , phys .
b * 60 * , 3643 ( 1999 ) .
wilson and m.e .
fisher , phys .
* 28 * , 240 ( 1972 ) .
see , for example , s. mo , j. hove , and a. sudb , phys .
b * 65 * , 104501 ( 2002 ) , and references therein .
p. calabrese , p. parruccini , a. pelissetto , and e. vicari , phys .
b * 70 * , 174439 ( 2004 ) . m. de prato , a. pelissetto , and e. vicari , phys . rev .
b * 70 * , 214519 ( 2004 ) .
r. schloms and v. dohm , nucl .
b * 328 * , 639 ( 1989 ) .
g. parisi , _ field - theoretic approach to second - order phase transitions in two- and three - dimensional systems _
, cargse lectures ( 1973 ) , j. stat . phys .
* 23 * , 49 ( 1980 ) . g. t hooft and m.j.g .
veltman , nucl .
b * 44 * , 189 ( 1972 ) .
nickel , d.i .
meiron , and g.a .
baker , jr .
, _ compilation of 2-pt and 4-pt graphs for continuum spin models _ , guelph university report , 1977 , unpublished .
o. zachar , s.a .
kivelson , and v.j .
emery , phys .
b * 57 * , 1422 ( 1998 ) .
a. pelissetto , p. rossi , and e. vicari , nucl .
b * 607 * , 605 ( 2001 ) .
h. kawamura , j. phys .
: condens .
matter * 10 * , 4707 ( 1998 ) .
p. calabrese and p. parruccini , nucl .
b * 679 * , 568 ( 2004 ) .
a. pelissetto and e. vicari , phys
. rep . * 368 * , 549 ( 2002 ) .
b. delamotte , d. mouhanna , and m. tissier , phys .
b * 69 * , 134413 ( 2004 ) .
a. pelissetto , p. rossi , and e. vicari , phys .
b * 63 * , 140414(r ) ( 2001 ) .
p. calabrese , p. parruccini , and a.i .
sokolov , phys .
b * 66 * , 180403(r ) ( 2002 ) ; phys . rev .
b * 68 * , 094415 ( 2003 ) . p. calabrese , a. pelissetto , and e. vicari , nucl .
b * 709 * , 550 ( 2005 ) .
p. parruccini , phys .
b * 68 * , 104415 ( 2003 ) . for @xmath112
the collinear fp is ( ref . )
@xmath393 and @xmath394 ( mzm ) , @xmath395 and @xmath396 ( 3@xmath157-@xmath55 ) . for @xmath47
the collinear fp is : @xmath397 and @xmath398 ( mzm ) , @xmath399 and @xmath400 ( 3@xmath157-@xmath55 ) .
a. aharony , in _ phase transitions and critical phenomena _ ,
vol . 6 , edited by c. domb and m.s .
green ( new york , academic , 1976 ) .
a. pelissetto and e. vicari , phys .
b * 62 * , 6393 ( 2000 ) . at the chiral fp we use ( refs . ): @xmath401 , @xmath402 , @xmath403 ( @xmath47 ) ; @xmath404 , @xmath405 , @xmath406 ( @xmath112 ) . at the collinear fp , for @xmath47 we use ( refs . ):
@xmath407 , @xmath408 , @xmath409 .
for @xmath112 , we use the mapping with the xy model ( refs . ) and the results of m. campostrini , m. hasenbusch , a. pelissetto , p. rossi , and e. vicari , phys .
b * 63 * , 214503 ( 2001 ) , @xmath410 , @xmath411 , @xmath412 .
j. m. carmona , a. pelissetto , and e. vicari , phys .
b * 61 * , 15136 ( 2000 ) .
we use @xmath413 and @xmath414 ( mzm ) , @xmath415 and @xmath416 ( 3@xmath157-@xmath55 ) .
it is interesting to note that these exponents are also relevant for quantum chromodynamics ( the theory of strong interactions ) with two quarks at finite temperature if the anomaly contribution at the critical point is small [ f. basile , a. pelissetto , and e. vicari , proceedings of the symposium on lattice field theory 2005 , pos ( lat2005 ) 199 , hep - lat/0509018 , and j. high energy phys .
* 02 * , 044 ( 2005 ) ; a. butti , a. pelissetto , and e. vicari , j. high energy phys .
* 08 * , 029 ( 2003 ) ] . | we study the nature of the two - dimensional quantum critical point separating two phases with and without long - range spin - density - wave order , which has been recently observed in cuprate superconductors .
we consider the landau - ginzburg - wilson hamiltonian associated with the spin - density critical modes , perform a mean - field analysis of the phase diagram , and study the corresponding renormalization - group flow in two different perturbative schemes at five and six loops , respectively .
the analysis supports the existence of a stable fixed point in the full theory whose basin of attraction includes systems with collinear spin - density - wave order , as observed in experiments .
the stable fixed point is characterized by an enlarged @xmath0 symmetry .
the continuous transition observed in experiments is expected to belong to this universality class .
the corresponding critical exponents are @xmath1 and @xmath2 . |
low - dimensional spin systems with spin - singlet ground states have attracted considerable study over the past 15 years .
these systems are interesting for a number of reasons , including possible relevance to high temperature superconductivity and the existence of novel ground states .
most of the systems studied to date have been limited to quasi - one - dimension .
these include spin ladder systems ( e.g. srcu@xmath1o@xmath2@xcite ) , spin - peierls systems , ( e.g. cugeo@xmath2@xcite ) and haldane gap systems ( e.g. y@xmath1banio@xmath8@xcite ) .
these materials all have a spin gap in the energy spectrum , which is associated with the spin singlet ground states .
theoretical studies of spin ladders@xcite have predicted d - wave superconductivity at low concentrations of hole doping in these systems .
in fact , the mixed ladder - chain system sr@xmath9ca@xmath10cu@xmath11o@xmath12 has been found to be a superconductor under pressure , specifically in the cases of [email protected]@xcite and [email protected],@xcite with the superconductivity thought to originate in the ladder layers .
_ cuprates have d - wave superconducting states and a partial ( or pseudo ) spin gap , which provides further evidence for strong similiarities between doped ladders and doped cuo@xmath1 planes and motivates study of the effects of perturbing spin gap systems .
two dimensional systems with spin singlet ground states are comparatively rare and often involve lattices where the active spins have a small number of nearest neighbours , such as cav@xmath15o@xmath16,@xcite where the depleted square lattice consists of plaquettes of four spins , coupled to neighbouring plaquettes through one vertex .
srcu@xmath1(bo@xmath2)@xmath1 is a second example of a quasi - two - dimensional spin system with a spin - singlet ground state .
@xcite the unit cell is tetragonal and the structure is characterized by layers consisting of cu@xmath17 , o@xmath18 and b@xmath19 running perpendicular to the c - axis .
these layers are separated from each other by planes composed of sr@xmath17 ions .
all cu@xmath17 sites have localized spin s = 1/2 moments and one nearest neighbour with which they form dimeric units , while each of these units is surrounded by four orthogonal dimers .
this configuration of spins corresponds to the two - dimensional model that shastry and sutherland first introduced in 1981.@xcite this model is topologically equivalent to a two - dimensional heisenberg model with nearest - neighbour coupling j and next - nearest - neighbour coupling j@xmath20 . with decreasing j@xmath20/j , a phase transition from a four - sublattice neel state to a gapped spin singlet state was shown to take place at a critical value of j@xmath20/j @xmath21 0.7 .
the ratio of j@xmath20/j in srcu@xmath1(bo@xmath2)@xmath1 is about 0.68 , a little bit smaller than the critical value.@xcite high resolution inelastic neutron measurements@xcite have identified one , two , and three triplet excitations .
these measurements clearly show the appearance of the energy gap in the spectrum of excitations with decreasing temperature , and from them the energy gap is evaluated to be @xmath21 34 k. the one and two triplet excitations both display a very large drop - off in inelastic intensity with increasing temperature .
this temperature dependence is generally not seen in a system undergoing a phase transition , and no evidence of symmetry breaking has been associated with the appearance of the spin singlet ground state . another interesting feature of this system is the magnetization plateaus that are observed at 1/8 , 1/4,@xcite and 1/3@xcite of the saturation magnetization in high magnetic fields ( @xmath22 20 t ) .
this is the first 2d system where these plateaus have been observed .
while superconductivity has been predicted and observed in some quasi - one - dimensional systems with a spin singlet ground state , it has also been predicted in the doped shastry - sutherland model.@xcite liu et .
al have recently investigated a series of polycrystalline srcu@xmath1(bo@xmath2)@xmath1 samples , with dopants added both in@xcite and out@xcite of the cubo@xmath2 planes .
they found that the spin gap behaviour was strongly affected by doping , but did not find any evidence for superconductivity .
the resistivity of the pure sample at room temperature was found to be 41.73 m@xmath23m,@xcite and the resistivity of all doped samples was found to be only one or two orders of magnitude smaller .
this implies that all their samples were insulators and that the carriers introduced as a result of doping were not very mobile , at least at room temperature .
however , the theoretical predictions discussed above make it important to better understand shastry - sutherland systems such as srcu@xmath1(bo@xmath2)@xmath1 and their response to perturbations .
muon spin relaxation measurements of a spin singlet state give a signal characteristic of a non - magnetic state .
zero field measurements@xcite of the spin - peierls system cugeo@xmath2 gave essentially a temperature - independent signal , with weak relaxation being ascribed to isolated defects associated with impurities . in the case of srcu@xmath1(bo@xmath2)@xmath1 , fukaya _
et al._@xcite reported zero field ( zf ) and longitudinal field ( lf ) @xmath0sr measurements of ceramic specimens .
in contrast to their expectations of seeing no spin relaxation associated with this state , the authors reported the onset of a large , roughly constant relaxation for temperatures of 3 k and below .
they attributed this relaxation to dilute defects or to @xmath0sr detecting weak intrinsic magnetism that other probes could not see .
the authors also reported that dynamic fluctuations persisted in the system down to base temperature ( 20 mk ) and that these fluctuations likely have a quantum origin .
lappas _ et al .
_ also investigated ceramic samples of srcu@xmath1(bo@xmath2)@xmath1 using the @xmath0sr technique.@xcite the authors reported the same anomalous relaxation in zf as observed by fukaya _
_ using a different batch of samples .
they attributed this relaxation to an unusual spin freezing process .
the authors went on to perform 0.6 t frequency shift measurements on their samples .
they concluded that implanted muons may liberate spin density in the spin gap regime that spin - freezes at very low temperatures .
we grew single crystals of srcu@xmath1(bo@xmath2)@xmath1 , srmg@xmath3cu@xmath4(bo@xmath2)@xmath1 , sr@xmath5la@xmath6cu@xmath1(bo@xmath2)@xmath1 , and sr@xmath7na@xmath3cu@xmath1(bo@xmath2)@xmath1 in floating zone image furnaces at tokyo@xcite and mcmaster university@xcite following slightly different procedures .
neutron scattering measurements have been reported by haravifard et al.@xcite on the system srmg@xmath3cu@xmath4(bo@xmath2)@xmath1 . aside from the original crystal growth article@xcite , this seems to be the only paper describing doped single crystals of this system as they have proven to be quite difficult to grow .
dc susceptibility measurements of a mcmaster crystal of srcu@xmath1(bo@xmath2)@xmath1 in a field of 5 t parallel to the c - axis are shown in fig .
[ scbo_susceptibility ] ; results for the tokyo crystals are essentially identical .
there is a sharp decrease in the susceptibility below 15 k which provides evidence for the spin singlet ground state .
a small upturn is seen in the susceptibility curve below 4 k , which likely arises due to magnetic impurities and/or defects of cu@xmath17 ions in srcu@xmath1(bo@xmath2)@xmath1 .
the susceptibility data below 7.5 k was fit to the following functional form : @xmath24 the first curie term is due to the contribution of magnetic impurities , the second term is to account for the spin gap , and the third temperature - independent term accounts for van - vleck paramagnetism and core diamagnetism . the spin gap @xmath25 was found to be @xmath21 20.2(7 ) k in our field of 5 t , which is in line with other reported estimates .
if we ascribe the curie term to free cu spins , then this concentration would be only 0.12(4 ) % .
we performed 5 t transverse field ( tf ) @xmath0sr measurements on the pure and doped srcu@xmath1(bo@xmath2)@xmath1 samples to investigate the presumed non - magnetic spin singlet ground states in these systems .
these measurements were conducted on the m15 surface muon channel at triumf in vancouver , canada , using a helium gas flow cryostat in the temperature range 2 k @xmath26 t @xmath26 150 k. the samples were mounted in a low - background spectrometer with the c - axis parallel to the incoming muon momentum ( and the external field ) . in this high field experiment , the measurements of the muon precession frequency signal in our samples and a reference material ( caco@xmath2 ) were taken simultaneously to allow for a precise determination of the frequency shift .
in addition , we performed zf-@xmath0sr and low tf-@xmath0sr measurements ( h @xmath26 1 t ) on the pure sample . after observing the anomalous relaxation previously seen by fukaya _
et al._@xcite in the powder sample , we performed lf-@xmath0sr on our single crystal to characterize the nature of this relaxation .
for these low temperature measurements ( between 20 mk and 10 k ) , we used an oxford instruments dilution refrigerator on the m15 beam line .
the sample was mounted on a silver sample holder , with the c - axis parallel to the incoming muon beam , as in the case of the high tf measurements .
implanted positive muons reside at sites which are minima in the electrostatic potential . generally in oxides ,
muons are hydrogen - bonded to o@xmath18 ions , roughly 1 away .
[ pure_fft ] shows a fast fourier transform of the transverse field muon precession signal measured in an applied magnetic field h = 5 t @xmath27 at t = 20 k. two peaks are clearly visible which correspond to two magnetically - inequivalent muon sites in srcu@xmath1(bo@xmath2)@xmath1 ; we associate these with the two magnetically - inequivalent oxygen sites in the system . to take these two muon sites into account ,
this 5 t tf data was fit to the sum of two exponentially - damped cosine functions .
our exact fitting function is given by : @xmath28 where @xmath29 and @xmath30 are the average muon precession frequencies , @xmath31 and @xmath32 are the muon relaxation rates for the two sites , and @xmath33 and @xmath34 represent the relative occupancy of the two unique muon sites . @xmath33 and @xmath34 are both temperature independent and were found to have a ratio of roughly 3:2 .
the ratio of the two inequivalent o sites in the unit cell is 2:1 , with fewer o sites lying in the cu - o - cu superexchange path .
however , the probability of occupying a specific site depends not just on the relative number of such sites , but also on the exact energies of those sites .
generally , there should only be one true stable muon site , but a metastable site could be long - lived on the time range of our experiment , leading to the site - occupation ratio that we observe here .
the frequency shift is the fractional difference between the applied field and the local field at the muon site , and it is proportional to the local spin susceptibility .
we see that for one of the muon sites the frequency shift roughly follows a curie - weiss temperature dependence for t @xmath22 4 k ( represented by the crosses in fig . [ 5t_knightshift]a ) , with the precise value of the curie constant depending on the temperature range being fit .
this is in contrast to @xmath35b - nmr frequency shift measurements.@xcite in that case , the authors applied a field of 7 t perpendicular to the [ 1 -1 0 ] direction and observed a decrease in the frequency shift below the singlet formation temperature of 15 k. the lack of a similar decrease in our @xmath36 frequency shift for this site indicates the presence of a larger local field than expected in the singlet regime .
the fundamental difference between the two experiments is that in our case there is a muon present in the system , which means the larger local field is likely the result of a muon perturbation effect .
more specifically , it seems that the muon has locally broken a spin singlet bond and created at least two quasi - free spins .
these spins are then polarized by the applied field which leads to our observed frequency shift .
as this appears to be a strong effect , it then seems reasonable to associate this muon site with the oxygen site in the cu - o - cu superexchange path .
the shift at the other muon site is considerably smaller and negative ( see inset of fig . [
5t_knightshift]a ) . in this case
, the frequency shift follows the temperature - dependence of the bulk susceptibility reasonably well , indicating the lack of any strong muon perturbation . for this reason
, this site likely corresponds to the o@xmath18 ion out of the cu - o - cu superexchange path .
muon perturbation effects have been observed in other materials . in the system prni@xmath8
, it has been shown that an implanted muon locally alters the crystal field levels.@xcite muon - induced singlet - breaking , similar to what is being described in this work , has been observed in the spin-1/2 ladder system kcucl@xmath2.@xcite an implanted muon has also been reported to perturb a link in the af chain of the quasi 1-d systems kcuf@xmath2@xcite and dichlorobis ( pyridine ) copper ( ii)@xcite , where the observed local spin susceptibility agreed well with theoretical calculations.@xcite our measurement of the local susceptibility at two different sites in srcu@xmath1(bo@xmath2)@xmath1 should serve to stimulate similiar calculations for the two - dimensional shastry - sutherland system . motivated by the discovery of the two muon sites in srcu@xmath1(bo@xmath2)@xmath1 , we fit the zf-@xmath0sr data to the sum of two terms , each taken to be the product of an exponential and a static gaussian kubo - toyabe relaxation function .
the gaussian kubo - toyabe part represents relaxation due to nuclear moments that are oriented randomly in zf .
this is generally a temperature - independent contribution to the relaxation and this was assumed in this case by fixing the gaussian kubo - toyabe relaxation rates .
the ratio of the asymmetries of the two sites was fixed to the value obtained when fitting the high tf data .
4 shows the exponential relaxation of one site is roughly independent of temperature for t @xmath22 4 k , then rises sharply at 4 k and saturates for temperatures of 3 k and below .
the other site has a very small , almost temperature - independent relaxation . the large increase and saturation of the exponential relaxation for the one muon site at low temperatures may imply slowing down or freezing of the spins liberated by the muon perturbation effect as previously suggested by lappas _ et al._.@xcite [ cols="^,^,^,^,^",options="header " , ] to see if this was a suitable description , it was necessary to examine longitudinal field ( lf ) @xmath0sr data taken below the saturation temperature , specifically at 20 mk in this case ( fig .
[ lf_lineshapes ] ) .
values of @xmath21 5 g and 80 g were estimated for the local fields at the two muon sites from the zf time spectrum at t @xmath13 20 mk using the relation b @xmath37 ( @xmath38 mhz / t ) with @xmath39 representing the muon relaxation rate as before .
assuming these local fields are static , a comparable lf ( up to one order of magnitude larger ) should be enough to completely decouple them . however , we are still seeing significant relaxation even in the case of lf @xmath13 2 kg .
this implies that the spins are dynamic in the low temperature regime and their behaviour can not be described by spin freezing .
these spin fluctuations are fast for t @xmath22 4 k and then slow down substantially below t @xmath21 4 k. the saturation of the relaxation at the low temperatures suggests that these fluctuations are not thermal but instead have a quantum origin .
this is in agreement with what fukaya _ et al .
_ reported in the ceramic samples.@xcite if the lf spectra are examined closely at early times , it becomes clear that they have a distinct gaussian nature .
in general , dynamic spin fluctuations can not be described by a gaussian lineshape .
this appears to be one of the rare cases of undecouplable gaussian behaviour in @xmath0sr , which is still poorly understood .
such behaviour has been previously observed in frustrated and/or low - dimensional systems.@xcite fig .
[ mg_fft ] shows a fast fourier transform of the transverse field muon precession signal measured in an applied magnetic field h = 5 t @xmath27 at t = 20 k for srmg@xmath3cu@xmath4(bo@xmath2)@xmath1 .
similar ffts were obtained for the la and na - doped samples .
there are clearly three frequency peaks in this case as opposed to the two that were seen in the fft of the pure sample .
for this reason , the data was fit to the same functional form as given by eq .
( 2 ) , but three exponentially - damped cosine terms were included instead of two .
the frequency shifts calculated from this fit are given in fig .
[ 5t_knightshift]b - d ; the shifts for all three doped samples examined appear to be essentially equivalent .
the upper and lower branches are remarkably similar to the two branches in the pure case , but the middle branch is new and so it seems is associated with the dopants . in one situation , we are substituting non - magnetic mg in for the cu sites .
it is well - known that this leads to the direct breaking of spin singlets , as some cu spins no longer have a nearest neighbour cu with which to form a dimer . in the other two cases
, we are substituting la or na in for the sr sites .
these are both forms of charge doping , with each la introducing one electron into our system and each na introducing one hole .
each electron or hole introduced by doping is expected to break a singlet pair and make a localized spin-1/2 .
resistivity measurements performed by liu _
et al._@xcite on la and na - doped polycrystalline samples verify the localized nature of these carriers .
the breaking of singlets by the different dopants is supported by dc susceptibility measurements performed on our single crystals with an applied field h = 1 kg @xmath40 ab plane.@xcite all three doped samples had much larger curie tails than the pure sample , indicating the presence of many more free spin-1/2 cu s in the singlet regime .
we have calculated the spin gap , curie constant and the resulting fraction of free spins by fitting this data to eq .
( 1 ) ; the results are shown in table i. in all the doped cases , the spin gap has been supressed .
also , the fraction of free spins is at least five times as great as in the pure case , but still lower than what we expect considering the concentration of our starting materials .
this is likely due to difficulties involved in doping this system and suggests that our doping concentrations should be viewed as nominal values .
our @xmath0sr results complement the susceptibility measurements by providing microscopic evidence that singlets are broken in this system when it is doped both in and out of the cubo@xmath2 planes .
it appears there are three distinct possibilities for muons in the doped samples .
the muon can either stop at the sites in or out of the superexchange path as before .
if the muon stops at the site in the superexchange path , it will break the singlet and create two quasi - free spins in close proximity .
however , in some cases it seems the singlet will already be broken due to the dopants and then the muon will have no singlet to break . in this case
, the muon will only be in close proximity to one quasi - free spin and so a smaller frequency shift will be measured with decreasing temperature , giving us the middle branch .
in fact , figure [ 5t_knightshift]b - d shows that this middle branch gives about half the shift of the upper branch .
we determined there are two different muon sites in srcu@xmath1(bo@xmath2)@xmath1 , which we associate with muons located adjacent to the two inequivalent o sites in the system .
one site exhibits a large increase in the frequency shift with decreasing temperature that is unaffected by singlet formation , indicating that the muon sits in the cu - o - cu superexchange path and has broken the corresponding dimer .
the frequency shift of the other site is considerably smaller and negative .
it roughly scales with the dc bulk susceptibility , indicating the lack of a perturbation effect in this case .
we also observe a large relaxation increase for the site in the superexchange path in the singlet regime , which is in contrast to the weak , practically temperature - independent relaxation expected in such a state .
analysis of lf-@xmath0sr data has shown that this relaxation is likely due to slow but dynamic fluctuations of the liberated spins that persist down to our base temperature of 20 mk . finally , all three doped samples studied have remarkably similar frequency shifts . in each case , there are now three branches at low temperatures instead of two as observed in the pure sample . the upper and lower branches map on to the two branches in the pure sample very well , and so the third branch is attributed to the dopants .
specifically , this third branch represents the case when the muon sits at a site in the superexchange path lacking a corresponding singlet due to it already being broken as a result of doping .
our results demonstrate that electron or hole doping of srcu@xmath1(bo@xmath2)@xmath1 by substituting on the sr site gives the same result as substitution of cu by non - magnetic mg : namely , the breaking of a localized singlet and the liberation of a single cu spin .
we find that the number of free spins is significantly less than expected from the nominal dopant concentration and that reaching doping levels sufficient for metallic ( and superconducting ) behavior may be difficult in this system .
a. a. aczel and g. j. macdougall are supported by nserc cgs fellowships .
research at mcmaster university is supported by nserc and the canadian institute for advanced research quantum materials program .
research at columbia university is supported by nsf dmr-05 - 02706 , nsf dmr-01 - 02752 , and che-01 - 11752 .
m. azuma , z. hiroi , m. takano , k. ishida and y. kitaoka , phys .
lett . * 73 * , 3463 ( 1994 ) .
k. hirota , d.e .
cox , j.e .
lorenzo , g. shirane , j.m .
tranquada , m. hase , k. uchinokura , h. kojima , y. shibuya and i. tanaka , phys .
73 * , 736 ( 1994 ) .
j. darriet and j.p .
regnault , solid state commun .
* 86 * , 409 ( 1993 ) .
e. dagotto , j. riera , and d. scalapino , phys .
b * 45 * , 5744 ( 1992 ) .
m. uehara , t. nagata , j. akimitsu , h. takahashi , n. mori , and k. kinoshita , j. phys .
japan * 65 * , 2764 ( 1996 ) .
t. nagata , m. uehara , j. goto , j. akimitsu , n. motoyama , h. eisaki , s. uchida , h. takahashi , t. nakanishi , and n. mori , phys .
* 81 * , 1090 ( 1998 ) .
s. taniguchi , t. nishikawa , y. yasui , y. kobayashi , m. satoh , t. nishioka , m. kontani , and k. sano , j. phys .
japan * 64 * , 2758 ( 1995 ) .
h. kageyama , k. yoshimura , r. stern , n.v .
mushnikov , k. onizuka , m. kato , k. kosuge , c.p .
slichter , t. goto , and y. ueda , phys .
lett . * 82 * , 3168 ( 1999 ) .
s. shastry and b. sutherland , physica b * 108 * , 1069 ( 1981 ) .
liu , j.l .
luo , n.l .
wang , x.n .
jing , d. jin , t. xiang , and z.h .
wu , phys .
b * 71 * , 014441 ( 2005 ) .
gaulin , s.h .
lee , s. haravifard , j.p .
castellan , a.j .
berlinsky , h.a .
dabkowska , y. qiu , and j.r.d .
copley , phys .
* 93 * , 267202 ( 2004 ) .
k. onizuka , h. kageyama , y. narumi , k. kindo , y. ueda , and t. goto , j. phys .
japan * 69 * , 1016 ( 2000 ) .
shastry and b. kumar , prog .
* 145 * , 1 ( 2002 ) .
c.h . chung and y.b .
kim , phys .
lett . * 93 * , 207004 ( 2004 ) .
t. kimura , k. kuroki , r. arita , and h. aoki , phys .
b * 69 * , 054501 ( 2004 ) . g.t .
liu , j.l .
luo , y.q .
guo , s.k .
su , p. zheng , n.l .
wang , d. jin , and t. xiang , phys .
b * 73 * , 014414 ( 2006 ) .
kojima , y. fudamoto , m. larkin , g.m .
luke , j. merrin , b. nachumi , y.j .
uemura , m. hase , y. sasago , k. uchinokura , y. ajiro , a. revcolevschi , and j .-
renard , phys .
lett . * 79 * , 503 ( 1997 ) .
a. fukaya , y. fudamoto , i.m .
gat , t. ito , m.i .
larkin , a.t .
savici , y.j .
uemura , p.p .
kyriakou , g.m .
luke , m.t .
rovers , h. kageyama , and y. ueda , physica b * 326 * , 446 ( 2003 ) .
a. lappas , a. schenck , and k. prassides , physica b * 326 * , 431 ( 2003 ) .
h. kageyama , k. onizuka , t. yamauchi , and y. ueda , j. crystal growth * 206 * , 65 ( 1999 ) .
dabkowska , a.b .
dabkowski , b.d .
gaulin , s.r .
dunsiger , g.m .
luke , s. haravifard , and m. cecchinel , j. crystal growth ( to be published ) s. haravifard , s.r .
dunsiger , s. el shawish , b.d .
gaulin , h.a .
dabkowska , m.t.f . telling , t.g .
perring , and j. bonca , phys .
* 97 * , 247206 ( 2006 ) .
k. kodama , k. arai , m. takigawa , h. kageyama , and y. ueda , journal of magnetism and magnetic materials * 272 - 276 * , 491 ( 2004 ) .
r. feyerherm , a. amato , a. grayevsky , f.n .
gygax , n. kaplan , and a. schenck , z. phys .
b * 99 * , 3 - 13 ( 1995 ) .
d. andreica , n. cavadini , h.u .
gudel , f.n .
gygax , k. kramer , m. pinkpank , and a. schenck , physica b * 289 - 290 * , 176 - 180 ( 2000 ) .
chakhalian , r.f .
kiefl , r. miller , s.r .
dunsiger , g. morris , s. kreitzman , w.a .
macfarlane , j. sonier , s. eggert , i. affleck , and i. yamada , physica b * 326 * , 422 - 426 ( 2003 ) .
chakhalian , r.f .
kiefl , r. miller , j. brewer , s.r .
dunsiger , g. morris , w.a .
macfarlane , j.e .
sonier , s. eggert , i. affleck , a. keren , and m. verdaguer , phys .
* 91 * , 027202 ( 2003 ) .
s. eggert and i. affleck , phys .
lett . * 75 * , 934 ( 1995 ) .
s. eggert and i. affleck , phys .
b * 46 * , 10866 ( 1992 ) . y.j .
uemura , a. keren , k. kojima , l.p .
le , g.m .
luke , w.d .
wu , y. ajiro , t.asano , y. kuriyama , m. mekata , h. kikuchi , and k. kakurai , phys .
* 73 * , 3306 ( 1994 ) .
a. fukaya , y. fudamoto , i.m .
gat , t. ito , m.i .
larkin , a.t .
savici , y.j .
uemura , p.p .
kyriakou , g.m .
luke , m.t .
rovers , k.m .
kojima , a. keren , m. hanawa , and z. hiroi , phys .
lett . * 91 * , 207603 ( 2003 ) . | we have performed @xmath0sr studies on single crystals of srcu@xmath1(bo@xmath2)@xmath1 , a quasi - two - dimensional spin system with a spin singlet ground state .
we observe two different muon sites which we associate with muons located adjacent to the two inequivalent o sites .
one site , presumed to be located in the cu - o - cu superexchange path , exhibits a large increase in the frequency shift with decreasing temperature which is unaffected by the singlet formation , indicating that the muon has locally broken the singlet bond .
we have also performed @xmath0sr on single crystals of srmg@xmath3cu@xmath4(bo@xmath2)@xmath1 , sr@xmath5la@xmath6cu@xmath1(bo@xmath2)@xmath1 , and sr@xmath7na@xmath3cu@xmath1(bo@xmath2)@xmath1 .
we have found that the frequency shifts of these doped samples are equivalent and contain three branches at low temperatures .
two of these branches map on to the branches observed in the pure sample reasonably well and so we attribute the third branch to effect of the dopants . specifically
, this third branch represents the case when the muon sits at a site in the superexchange path lacking a corresponding singlet , due to it already being broken as a result of doping .
this then leads to the conclusion that singlets are broken in this system when it is doped both in and out of the cubo@xmath2 planes . |
quantum boltzmann time - correlation functions play a central role in chemical physics , and are ( usually ) impossible to calculate exactly .
one promising approach is to treat the statistics quantally and the dynamics classically .
the standard way to do this is to use the linearized semi - classical initial value representation ( lsc - ivr or ` classical wigner approximation ' ) , @xcite but this has the drawback of not satisfying detailed balance .
recently,@xcite however , we found that a single change to the lsc - ivr derivation gives a classical dynamics which does satisfy detailed balance .
this modified version of lsc - ivr is called ` matsubara dynamics '
. we can summarise matsubara dynamics as follows : at initial time , the quantum statistics gives rise to delocalized distributions in position which are smooth functions of imaginary time . if we constrain the lsc - ivr dynamics to conserve this smoothness ( by including only the smooth ` matsubara ' modes see sec .
ii ) we find that it satisfies detailed balance , and gives better agreement than lsc - ivr with the exact quantum result.@xcite we suspect ( but have not yet proved ) that matsubara dynamics reproduces the time - dependence of the exact kubo - transformed time - correlation function up to order @xmath0 , and is thus the correct theory for describing quantum statistics and classical dynamics .
matsubara dynamics suffers from the sign problem and is hence impractical , but the findings just described suggest that it should be the starting point from which to make further approximations if one wishes to devise practical methods that combine quantum statistics with classical dynamics .
numerical tests in ref .
( see also fig .
1 ) showed that the popular centroid molecular dynamics@xcite ( cmd ) and ring - polymer molecular dynamics@xcite ( rpmd ) methods appear to be two such approximations . here
we confirm this , by deriving the terms that cmd and rpmd leave out from the matsubara dynamics .
matsubara dynamics approximates the quantum kubo - transformed time - correlation function@xcite @xmath1 \label{kubo}\end{aligned}\ ] ] by @xmath2}(t ) \end{aligned}\ ] ] where @xmath3}(t ) = { \alpha_m\over 2\pi\hbar}\int d{\bf \widetilde p } & \int d{\bf \widetilde q}\ a({\bf \widetilde q})e^{-\beta[{{\widetilde}h}_m({\bf { \widetilde}p},{\bf { \widetilde}q})-i\theta_m({\bf \widetilde p},{\bf \widetilde q})]}{\nonumber}\\ & \times e^{{\cal l}_{m}t}b({\bf \widetilde q})\label{mattsc}\end{aligned}\ ] ] and @xmath4!^2 $ ] .
the position coordinates @xmath5 , with @xmath6 , are the @xmath7 matsubara modes , which describe closed paths @xmath8 that are smooth functions of the imaginary time @xmath9 @xmath10 , where @xmath11 is the centroid coordinate ( see the appendix ) ; @xmath12 , and @xmath13 are similarly defined for momentum .
the functions @xmath14 and @xmath15 are obtained from the operators @xmath16 and @xmath17 ( see the appendix ) , such that @xmath18 gives @xmath19 .
@xcite the propagator @xmath20 contains the matsubara liouvillian @xmath21 in which the potential energy @xmath22 is given in the appendix . the quantum boltzmann distribution in eq .
( [ mattsc ] ) is complex , and contains the matsubara hamiltonian @xmath23 and the phase @xmath24 where @xmath25 are the matsubara frequencies @xmath26 .
matsubara dynamics is inherently classical ( meaning that terms @xmath27 disappear from the quantum liouvillian on decoupling the matsubara modes , leaving @xmath28 ) , and conserves the hamiltonian @xmath29 _ and _ the phase @xmath30 , and thus satisfies detailed balance .
clearly eq .
( [ mattsc ] ) suffers from the sign problem because of the phase @xmath30 .
let us make the coordinate transformation @xmath31 .
this gives @xmath3}(t ) = & { \alpha_m\over 2\pi\hbar}\left[\prod_{n=-(m-1)/2}^{(m-1)/2}\int_{-\infty - im{{\widetilde}\omega}_n{{\widetilde}q}_{-n}}^{\infty - im{{\widetilde}\omega}_n{{\widetilde}q}_{-n } } d{\overline p}_n\right ] { \nonumber}\\ & \times \int d{\bf \widetilde q}\ a({\bf \widetilde q})e^{-\beta { \widetilde}r_m ( { \bf \overline p},{\bf { \widetilde}q})}e^{{\cal l}_{m}t}b({\bf \widetilde q } ) \label{trans}\end{aligned}\ ] ] where @xmath32 is the ` ring - polymer ' hamiltonian familiar from quantum statistics.@xcite equation ( [ trans ] ) is simply eq .
( [ mattsc ] ) in disguise , but at @xmath33 , we can use a standard contour - integration trick@xcite to shift @xmath34 onto the real axis , giving @xmath3}(0 ) = { \alpha_m\over 2\pi\hbar}\int d{\bf \overline p } & \int d{\bf \widetilde q}\ a({\bf \widetilde q})b({\bf \widetilde q})e^{-\beta { \widetilde}r_m ( { \bf \overline p},{\bf { \widetilde}q})}\end{aligned}\ ] ] which now contains the ( real ) ring - polymer distribution,@xcite and hence no longer suffers from the sign problem .
unfortunately , this trick does not work for @xmath35 ( see sec .
iv ) , so we are stuck with eq .
( [ mattsc ] ) , which motivates us to find approximations to matsubara dynamics .
this approximation can be made if @xmath36 is a function of just the centroid @xmath37 ( or @xmath38),@xcite in which case we need only the matsubara dynamics of the centroid reduced density @xmath39}{\nonumber}\\ & \times e^{{\cal l}_{m}t}b({\bf \widetilde q})\end{aligned}\ ] ] where the primes denote integration over all modes except @xmath38 and @xmath37 .
differentiation with respect to @xmath40 , application of eq .
( [ matliou ] ) , and integration by parts gives @xmath41}{\nonumber}\\ & \times { \cal l}_0e^{{\cal l}_{m}t}b({\bf \widetilde q})\label{dotb}\end{aligned}\ ] ] where @xmath42 , with mass @xmath43 , at temperature @xmath44 ( in atomic units ) .
the position autocorrelation functions in ( a ) are taken from ref . . the position - squared autocorrelation functions in ( b )
were calculated numerically using the procedure described in ref .
, with @xmath45 matsubara modes.@xcite the differences between the matsubara and exact quantum results show the importance of real - time quantum coherence in this model system , the neglect of which ( in the matsubara calculations ) has blue - shifted and broadened the spectrum . ] in the usual way of mean - field dynamics,@xcite we can split the force on the centroid into @xmath46 where @xmath47 is the mean - field force @xmath48 } { \partial { \widetilde u}_m({\bf \widetilde q})\over\partial { \widetilde q}_0}{\nonumber}\\ = & -{1\over z(q_0)}\int d{\bf \overline p } ' \int d{\bf \widetilde q}'\ { \nonumber}\\ & \times e^{-\beta { \widetilde}r_m({\bf \overline p},{\bf { \widetilde}q})}\,{\partial { \widetilde u}_m({\bf \widetilde q})\over\partial { \widetilde q}_0}\end{aligned}\ ] ] ( and we have used the @xmath33 contour - integration trick to get to the second line ) , @xmath49 and @xmath50 is the fluctuation force ( defined by eq .
( [ fluct ] ) as the difference between the exact force and @xmath47 ) .
equation ( [ dotb ] ) then splits into @xmath51b({\widetilde}q_0,{\widetilde}p_0,t){\nonumber}\\ & + \int d{\bf \widetilde p } ' \int d{\bf \widetilde q}'\ e^{-\beta[{{\widetilde}h}_m({\bf { \widetilde}p},{\bf { \widetilde}q})-i\theta_m({\bf \widetilde p},{\bf \widetilde q } ) ] } { \nonumber}\\ & \times f_{\rm fluct}({\bf { \widetilde}q}){\partial \over \partial { \widetilde p}_0 } e^{{\cal l}_{m}t}b({\bf \widetilde q})\label{mfluc}\end{aligned}\ ] ] this type of expression is encountered in coarse - graining , where the integral is sometimes approximated by a generalized langevin term.@xcite it is an exact rewriting of eq .
( [ dotb ] ) .
neglect of the integral term gives the mean - field approximation @xmath52b({\widetilde}q_0,{\widetilde}p_0,t)\end{aligned}\ ] ] which is cmd.@xcite thus cmd corresponds to approximating matsubara dynamics by leaving out the fluctuation term in eq .
( [ mfluc ] ) .
this result is not a surprise , and is consistent with previous numerical findings@xcite that cmd causes errors through neglect of fluctuations ( see sec .
what is new is that eq .
( [ mfluc ] ) gives an explicit formula for these fluctuations , in the case that the quantum dynamics can be approximated by matsubara dynamics .
we now return to eq .
( [ trans ] ) , which is just eq .
( [ mattsc ] ) rewritten in terms of @xmath53 . expressing @xmath28 in terms of these coordinates gives @xmath54}+i{\cal l}_m^{[\rm i]}\label{liouliou}\end{aligned}\ ] ] where @xmath55}= \!\!\!\ !
\sum_{n=-(m-1)/2}^{(m-1)/2 } \ ! { { \overline p}_n\over m}{\partial \over \partial { \widetilde q}_n } \ ! - \ ! \left[m{\widetilde}\omega_n^2{\widetilde}q_n \
! + \!{\partial { \widetilde u}_m({\bf \widetilde q})\over\partial { \widetilde q}_n}\right]\!{\partial \over\partial { \overline p}_n}\end{aligned}\ ] ] is the rpmd liouvillian ( corresponding to the ring - polymer hamiltonian @xmath56 ) and @xmath57}=\sum_{n=-(m-1)/2}^{(m-1)/2 } { \widetilde}\omega_n\left(\overline p_n{\partial \over\partial { \overline p}_{-n}}-{\widetilde}q_{n}{\partial \over\partial { { \widetilde}q}_{-n}}\right)\end{aligned}\ ] ] note that the complete liouvillian @xmath28 does not correspond to a hamiltonian in @xmath53 ( because the transformation from @xmath58 to @xmath59 is non - canonical ) , and that any resemblance to rpmd@xcite is at this stage illusory , since the imaginary parts of @xmath60 , @xmath61 , contribute terms that cancel the spring terms in @xmath62}$ ]
. l8.5 cm c|c l8.5 cm & & & + satisfies detailed balance , because the centroid mean - field force is decoupled from the matsubara fluctuations & & & satisfies detailed balance , because @xmath62}$ ] and @xmath63}$ ] [ in eq .
( [ liouliou ] ) ] independently satisfy detailed balance + is the centroid mean - field approximation to matsubara dynamics & & & has the same centroid mean - field approximation as matsubara dynamics , namely cmd + is exact for linear tcfs in the harmonic limit , since the centroid mean - field force is then equal to the matsubara force & & & is exact for linear tcfs in the harmonic limit , since the neglected term @xmath63}$ ] does not act on the centroid + gives the exact centroid - averaged matsubara liouvillian dynamics at @xmath33 & & & gives the exact matsubara liouvillian dynamics at @xmath33 + suffers from the _ curvature problem _ in vibrational spectra because of the neglect of the matsubara fluctuations & & & suffers from _ spurious resonances _ in vibrational spectra because the neglect of @xmath63}$ ] shifts the fluctuation frequencies + gives the mean - field - averaged matsubara force on the centroid & & & gives the exact matsubara force on the centroid + breaks down completely for non - linear @xmath64 and @xmath65 ( see fig .
1b ) because @xmath66 and @xmath67 depend on non - centroid modes & & & breaks down more rapidly for non - linear ( than for linear ) @xmath64 and @xmath65 ( see fig .
1b ) because the neglected term @xmath63}$ ] acts directly on the non - centroid modes + if we now try to shift @xmath60 , @xmath61 , onto the real axis , we find that the dynamics generated by @xmath28 propagates an initial distribution of real phase - space points into the complex plane , along unstable trajectories
. we do not know whether the contour - integration trick remains valid for such trajectories ; even if it does , they appear to be at least as difficult to treat numerically as the sign problem in eq .
( [ mattsc ] ) . however , it is possible@xcite to follow a path along which one gradually moves @xmath60 , @xmath61 , towards the real axis whilst gradually discarding @xmath63}$ ] , such that the dynamics remains stable ( and the contour - integration trick remains valid ) at every point along the path . at the end of the path , @xmath63}$ ] has been completely discarded , and @xmath60 has reached the real axis .
this results in the approximation , @xmath3}(t ) \simeq{\alpha_m\over 2\pi\hbar}\int \!\ !
d{\bf \overline p } & \int \!\ !
d{\bf \widetilde q}\ a({\bf \widetilde q})e^{-\beta { \widetilde}r_m ( { \bf \overline p},{\bf { \widetilde}q})}e^{{\cal l}_m^{[\rm rp]}t}b({\bf \widetilde q})\end{aligned}\ ] ] which is rpmd.@xcite a harmonic analysis@xcite shows that the main effect of discarding @xmath63}$ ] is erroneously to shift the matsubara fluctuation frequencies to the ring - polymer frequencies . since @xmath63}$ ]
does not act directly on @xmath11 , it follows that an rpmd time - correlation function involving linear operators ( for which @xmath68 or @xmath69 ) will agree initially with the matsubara result , but will then lose accuracy as the errors in the fluctuation dynamics couple to the centroid through the anharmonicity in @xmath22 .
this result is not a surprise , as the ring - polymer frequencies are known to interfere with the centroid dynamics.@xcite what is new is that we have identified the approximation made by rpmd , namely the neglect of @xmath63}$ ] .
we have shown that both cmd and rpmd are approximations to matsubara dynamics , which , as mentioned in the introduction , is probably the correct way to describe quantum statistics and classical dynamics .
cmd neglects the matsubara fluctuation term ; rpmd neglects part of the matsubara liouvillian .
so far as we can tell , there is no direct physical justification that can be given for either of these approximations .
cmd and rpmd are useful because , as has long been known,@xcite they preserve detailed balance , and satisfy a number of important limits .
these properties@xcite ( and a few others ) can be rederived from matsubara dynamics , and are listed in table i. note also that cmd and rpmd give the same @xmath33 leading - order error terms when compared with matsubara dynamics as with the exact quantum dynamics .
@xcite one new finding , less drastic than it first appears , is that both cmd and rpmd give qualitatively wrong fluctuation dynamics at barriers . in matsubara dynamics , some of the distributions in @xmath8 stretch indefinitely over the barrier top , such that a proportion of the distribution ends up on either side . in cmd and rpmd ,
all of the distribution ends up on one side of the barrier ( because cmd decouples the fluctuation modes necessary for stretching over the barrier , and rpmd shifts the frequencies of these modes from imaginary to real@xcite ) .
however , cmd and rpmd are still powerful tools for estimating quantum reaction rates , as the exact @xmath33 behaviour of these methods ( see table i ) ensures that classical rate theory ( in the mean - field centroid or ring - polymer space ) gives lower bound estimates of the @xmath70 quantum transition - state theory rate,@xcite for the special case of a centroid dividing - surface ( cmd ) , and for the general case ( rpmd ) .
the main new result of this work is that , in relating cmd and rpmd to matsubara dynamics , we have obtained explicit formulae for the terms that are left out , which may lead to improvements in these methods .
for example , it might be possible to include approximately the matsubara fluctuation term of eq .
( [ mfluc ] ) which is missing in cmd , or to exploit the property that rpmd gives the correct matsubara force on the centroid.@xcite tjhh , mjw and sca acknowledge funding from the u.k .
engineering and physical sciences research council .
am acknowledges the european lifelong learning programme ( llp ) for an erasmus student placement scholarship .
tjhh also acknowledges a research fellowship from jesus college , cambridge and helpful discussions with dr adam harper .
the set of @xmath7 matsubara modes @xmath71 is defined as@xcite @xmath72 where @xmath7 is odd@xcite and satisfies @xmath73 ; @xmath74 , are a set of discrete path - integral coordinates distributed at equally spaced intervals @xmath75 of imaginary time , and @xmath76 the momentum coordinates @xmath77 are similarly defined in terms of @xmath78 .
@xmath11 and @xmath69 are the position and momentum centroid coordinates . we define the associated matsubara frequencies @xmath79 such that they include the sign of @xmath80 , which gives @xmath30 the simple form of eq .
( [ thet ] ) . the functions @xmath66 and @xmath67 in eq .
( [ mattsc ] ) are obtained by making the substitutions @xmath81 into the functions @xmath82 the matsubara potential @xmath22 is obtained similarly , by subsituting for @xmath83 in the ring - polymer potential @xmath84 we emphasise that the formulae above and in sec .
ii result from just one approximation , namely decoupling the matsubara modes from the non - matsubara modes in the exact quantum liouvillian ( which causes all liouvillian terms @xmath27 to vanish).@xcite this article considers only dynamics on a single born - oppenheimer potential energy surface ; it might be possible to extend the analysis to multi - surface methods using the meyer - miller representation ; see : h.d .
meyer and w.h .
miller , j. chem .
phys . * 70 * , 3214 ( 1979 ) ; n. ananth , j. chem .
phys . * 139 * , 124102 ( 2013 ) ; j.o .
richardson and m. thoss , j. chem . phys .
* 139 * , 031102 ( 2013 ) . to simplify the algebra , we assume that the system is one - dimensional , and that @xmath16 and @xmath17 are functions of @xmath88 .
matsubara dynamics generalizes straightforwardly to multi - dimensions and to operators that are functions of @xmath89 ; see ref . .
note that this distribution corresponds to the matsubara ( i.e. smoothed ) version of the ring - polymer distribution of ref . which converges in the large @xmath7 limit to the same distribution as the more commonly used discrete form of ref .
see also : q. shi and e. geva , j. chem .
phys . * 118 * , 8173 ( 2003 ) , which obtains cmd by decoupling the centroid from lsc - ivr dynamics ( which is equivalent to filtering out the non - matsubara modes from lsc - ivr , then following the steps above ) .
the only property that can not be derived from matsubara dynamics is that cmd is equivalent to a minimum - energy wave - packet treatment in the limit @xmath90 ( where , clearly , one does not expect a description based on quantum statistics and classical dynamics to be valid ) ; see : r. ram ' irez and t. l ' opez - ciudad , j. chem .
phys . * 111 * , 3339 ( 1999 ) .
note that two of the rpmd normal - mode frequencies become imaginary on cooling below the instanton cross - over temperature of @xmath91 ; see : j.o .
richardson and s.c .
althorpe , j. chem .
phys . * 131 * , 214106 ( 2009 ) . | we recently obtained a quantum - boltzmann - conserving classical dynamics by making a single change to the derivation of the ` classical wigner ' approximation . here , we show that the further approximation of this ` matsubara dynamics ' gives rise to two popular heuristic methods for treating quantum boltzmann time - correlation functions : centroid molecular dynamics ( cmd ) and ring - polymer molecular dynamics ( rpmd ) .
we show that cmd is a mean - field approximation to matsubara dynamics , obtained by discarding ( classical ) fluctuations around the centroid , and that rpmd is the result of discarding a term in the matsubara liouvillian which shifts the frequencies of these fluctuations .
these findings are consistent with previous numerical results , and give explicit formulae for the terms that cmd and rpmd leave out . _
copyright ( 2015 ) american institute of physics .
this article may be downloaded for personal use only .
any other use requires prior permission of the author and the american institute of physics .
the following article appeared in the journal of chemical physics , 142 , 191101 ( 2015 ) and may be found at http://dx.doi.org/10.1063/1.4921234_ |
below , we use standard notions from the theory of finite order knot invariants ; see , e.g. @xcite or @xcite .
a _ chord diagram _ of order @xmath5 is an oriented circle endowed with @xmath6 pairwise distinct points split into @xmath5 disjoint pairs , considered up to orientation - preserving diffeomorphisms of the circle .
a _ weight system _ is a function on chord diagrams satisfying the @xmath7-term relation ; see fig .
[ fourtermrelation ] . for a chord diagram @xmath8 with two chords @xmath9 and @xmath10 having neighboring ends , as in fig .
[ fourtermrelation ] , we will write this relation as @xmath11 . in figures , the outer circle of the chord diagram is always assumed to be oriented counterclockwise .
dashed arcs may contain ends of arbitrary sets of chords , same for all the four terms in the picture . to each weight system
, a finite order knot invariant can be associated in a canonical way , which makes studying weight systems an important part of knot theory .
there is a number of approaches to constructing weight systems .
in particular , weight systems can be constructed from semisimple lie algebras , although the result is complicated .
the present paper has been motivated by an aspiration for understanding the weight system corresponding to the simplest nontrivial case of the lie algebra @xmath4 . the _ intersection graph _
@xmath12 of a chord diagram @xmath8 is the simple graph whose vertices are in one - to - one correspondence with the chords in @xmath8 , and two vertices are connected by an edge iff the corresponding chords intersect one another .
the _ @xmath7-term relation for graphs _ , introduced in @xcite , is the graph counterpart of the @xmath7-term relation for chord diagrams .
it is defined as follows .
denote by @xmath13 the set of vertices of a graph @xmath14 and by @xmath15 the set of its edges .
let us associate to each ordered pair of ( distinct ) vertices @xmath16 of a graph @xmath14 two other graphs @xmath17 and @xmath18 .
the graph @xmath17 is obtained from @xmath14 by erasing the edge @xmath19 in the case that this edge exists , and by adding the edge otherwise . in other words , we simply change the adjacency between the vertices @xmath9 and @xmath10 in @xmath20 .
this operation is an analogue of edge deletion , but we prefer to formulate it in a slightly more symmetric way .
the graph @xmath18 is obtained from @xmath14 in the following way . for any vertex @xmath21 we switch its adjacency with @xmath9 to the opposite one if @xmath22 is joined with @xmath10 , and we do nothing otherwise .
all other edges do not change .
note that the graph @xmath18 depends not only on the pair @xmath23 , but on the order of vertices in the pair as well .
a graph invariant @xmath24 is a _
@xmath7-invariant _ if it satisfies the _
@xmath7-term relation _
@xmath25 for each graph @xmath14 and for any pair @xmath16 of its vertices .
figure [ fourtermrelationfor graphs ] shows the @xmath7-term relation for intersection graphs corresponding to a sample @xmath7-term relation for chord diagrams .
( 200,100)(60,0 ) ( 12,72)(90,0)4 ( 55,72)(180,0)2@xmath26 ( 145,72)@xmath27 ( 5,91)(90,0)4(-1,-3)10 ( 0,94)(90,0)4e ( -3,84)(90,0)4(1,0)31 ( -9,87)(90,0)4d ( -8,72)(90,0)4(3,-1)36 ( -14,73)(90,0)4f ( 32,72)(90,0)4(-5,-2)35 ( 34,73)(90,0)4b ( 12,92)(90,0)4(1,-5)7.5 ( 10,95)(90,0)4c ( 18,91)(-1,-5)7.6 ( 18,94)(90,0)4a ( 108,91)(-2,-3)20.9 ( 198,91)(1,-2)13 ( 288,91)(1,-1)12.3 ( 2,30)(90,0)4 ( 22,30)(90,0)4 ( 2,-10)(90,0)4 ( 22,-10)(90,0)4 ( -8,10)(90,0)4 ( 32,10)(90,0)4 ( 55,10)(180,0)2@xmath26 ( 145,10)@xmath27 ( -6,-16)(90,0)4a ( 24,-16)(90,0)4b ( -6,32)(90,0)4e ( 22,32)(90,0)4d ( -14,12)(90,0)4f ( 34,12)(90,0)4c ( -8,10)(90,0)4(1,0)39 ( -8,10)(90,0)4(1,2)10 ( -8,10)(90,0)2(1,-2)10 ( -8,10)(90,0)4(3,-2)30 ( 32,10)(90,0)2(-3,-2)30 ( 32,10)(90,0)4(-1,2)10 ( 32,10)(90,0)4(-1,-2)10 ( 2,30)(90,0)4(1,0)20 ( 2,-10)(180,0)2(1,0)20 ( 2,-10)(90,0)4(1,2)20 any graph invariant satisfying the @xmath7-term relation ( that is , a @xmath7-invariant ) determines a weight system @xcite : the value of this weight system on a chord diagram is set to be the value of the @xmath7-invariant on the intersection graph of the diagram . let @xmath8 be a chord diagram
. denote by @xmath28 the number of circuits of length @xmath29 ( @xmath27 edge @xmath29-gons ) in the intersection graph of @xmath8 .
it is well known ( see @xcite , exercise 6.4.10 ) that , for @xmath30 , the parity of the number @xmath28 is a weight system with values in @xmath31 .
however , no integer - valued weight system with the same parity has been known . in this paper , we construct , for each @xmath1 , an integer - valued weight system @xmath0 whose value on each chord diagram has the same parity as @xmath32 , @xmath33 for any chord diagram @xmath8 .
this weight system counts circuits of length @xmath2 in the intersection graph of the diagram with signs depending on the mutual position of the corresponding chords . for completeness ,
let us recall the proof of the fact mentioned above .
it is worth to be compared with the proof for the integer - valued analogue below .
[ pmod2 ] the value @xmath34 is a weight system for each @xmath30 . *
the value @xmath28 obviously depends on the intersection graph of a chord diagram rather than on the diagram itself .
let us extend the function @xmath35 to arbitrary graphs ( that are not necessarily intersection graphs ) in the obvious way : let @xmath36 be the number of edge @xmath29-gons in @xmath20 having @xmath29 pairwise distinct vertices .
we are going to prove that @xmath37 satisfies the @xmath7-term relation for graphs ; the proposition then follows .
suppose @xmath20 contains an edge @xmath38 .
then for the two terms on the left - hand side of the @xmath7-term relation we have that @xmath39 is the number of edge @xmath29-gons in @xmath20 passing through the edge @xmath38 .
similarly , for the right - hand side , @xmath40 is the number of edge @xmath29-gons in @xmath41 passing through @xmath38 .
all the @xmath29-gons in @xmath20 passing through @xmath38 contain a chain @xmath42 and split into three disjoint classes according to the adjacency of the vertices @xmath22 and @xmath43 to @xmath9 and @xmath10 : 1 .
the vertex @xmath22 is adjacent to @xmath10 and @xmath43 is adjacent to @xmath9 ; 2 .
the vertex @xmath22 is adjacent to @xmath10 , but @xmath43 is not adjacent to @xmath9 ; 3 .
the vertex c is not adjacent to b. all edge @xmath29-gons in @xmath41 passing through the four points @xmath44 admit a similar classification .
the three edge @xmath7-gons in the leftmost graph in fig . [ fourtermrelationfor graphs ] passing through the edge @xmath38 split in the three classes in the following way : * the edge quadrangles @xmath45 and @xmath46 belong to the first class ; * there are no edge quadrangles belonging to the second class ; * the edge quadrangle @xmath47 belongs to the third class .
now , the @xmath29-gons in @xmath20 belonging to the second class are in one - to - one correspondence with the @xmath29-gons in @xmath41 containing the path @xmath48 .
the edge @xmath29-gons of the third kind are the same in both graphs @xmath20 and @xmath41 . and ,
finally , the edge @xmath29-gons of the first kind in each of the two graphs come in pairs : the chain @xmath42 can be replaced with the chain @xmath48 .
hence , the number of edge @xmath29-gons of the first type is even for each of the two graphs , and the required assertion follows .
the authors are grateful to the participants of the seminar `` combinatorics of finite order knot invariants '' at the department of mathematics nru hse for useful discussions .
the authors would also like to express their gratitude to the referees for careful proofreading and valuable suggestions .
in order to define the weight system @xmath0 , let us take a chord diagram @xmath8 and choose an arbitrary orientation of its chords .
this orientation induces an orientation of the edges of the intersection graph @xmath12 in the following way .
we orient an edge @xmath38 from @xmath9 to @xmath10 if the beginning of the chord @xmath10 belongs to the arc of the outer circle of @xmath8 which starts at the beginning of @xmath9 and goes in the positive direction to the end of @xmath9 ; see fig .
we denote the directed intersection graph of an oriented chord diagram @xmath8 by @xmath12 , like in the case of ordinary intersection graph of a non - oriented chord diagram , since this convention causes no misunderstanding . oriented edges in a directed graph
will also be called _
we say that a circuit of even length @xmath49 in a directed graph is _ positively oriented _
( or its sign is @xmath50 ) if the number of arrows in this circuit oriented in either direction is even ; otherwise the circuit is _ negatively oriented _ ( or has the sign @xmath26 ) . since the total number of arrows is even , the sign is independent of the choice of the direction of the circuit .
for example , all the arrows in the @xmath7-circuit @xmath51 in the directed intersection graph in fig .
[ feo ] are oriented in the same direction , meaning it is positive .
the function @xmath0 takes a chord diagram @xmath8 to the difference between the number of positively and negatively oriented @xmath2-gons in the directed intersection graph @xmath12 for arbitrarily chosen orientation of the chords in @xmath8 .
now we are going to show that this definition makes sense .
the function @xmath0 is well defined , that is , its value on a chord diagram does not depend on the chosen orientation of the chords .
* proof . *
it suffices to check that if we switch the direction of a single chord in a chord diagram @xmath8 to the opposite one , then the sign of each circuit remains the same .
changing the direction of a chord @xmath9 in @xmath8 leads to switching directions of all the arrows in @xmath12 incident to the vertex @xmath9 , preserving the directions of all other arrows .
hence , the sign of any circuit not containing the vertex @xmath9 remains the same .
any circuit passing through @xmath9 contains exactly two arrows incident to @xmath9 . after switching the direction of both of them
, the sign also remains the same .
let us compute the value of @xmath52 on the chord diagram @xmath8 shown in fig .
[ fer2c ] . due to the proposition above
, an arbitrary orientation of the chords can be chosen .
for the one in the figure , the directed intersection graph @xmath12 is shown in fig .
[ fer2c ] .
figure [ fer2c ] also shows all the oriented @xmath7-gons in the intersection graph .
the difference between the number of positively and negatively oriented @xmath7-gons is @xmath53 .
( 200,40)(50,0 ) ( 12,12 ) ( 45,10)@xmath54 ( -2,25)(1,-1)28 ( -2,-1)(1,1)28 ( 12,32)(0,-1)40 ( -8,12)(1,0)40 ( 9,34)a ( -9,27)b ( -13,-4)d ( -16,10)c ( 70,-5 ) ( 70,25 ) ( 100,-5 ) ( 100,25 ) ( 70,-5)(1,1)30 ( 70,-5)(1,0)30 ( 70,-5)(0,1)30 ( 100,-5)(0,1)30 ( 100,-5)(-1,1)30 ( 100,25)(-1,0)30 ( 62,-10)a ( 102,-10)b ( 62,27)d ( 102,27)c ( 170,-5 ) ( 170,25 ) ( 200,-5 ) ( 200,25 ) ( 170,-5)(1,0)30 ( 170,-5)(0,1)30 ( 200,-5)(0,1)30 ( 200,25)(-1,0)30 ( 240,-5 ) ( 240,25 ) ( 270,-5 ) ( 270,25 ) ( 240,-5)(1,1)30 ( 240,-5)(1,0)30 ( 270,-5)(-1,1)30 ( 270,25)(-1,0)30 ( 310,-5 ) ( 310,25 ) ( 340,-5 ) ( 340,25 ) ( 310,-5)(1,1)30 ( 310,-5)(0,1)30 ( 340,-5)(0,1)30 ( 340,-5)(-1,1)30 ( 180,-15)@xmath26 ( 250,-15)@xmath50 ( 320,-15)@xmath50 the function @xmath0 on chord diagrams is indeed a weight system , that is , it satisfies the @xmath7-term relation . * proof . *
the proof is similar to that of proposition [ pmod2 ] , but is slightly more complicated .
let @xmath8 be a chord diagram , and let @xmath9 and @xmath10 be a pair of chords in @xmath8 having neighboring ends and intersecting one another .
pick an orientation of the chords in @xmath8 such that the arrow @xmath38 in the directed graph @xmath12 of @xmath8 is oriented from @xmath9 to @xmath10 .
the difference @xmath55 counts the number of signed circuits of length @xmath2 in @xmath12 containing the edge @xmath38 .
each such circuit contains a sequence of vertices @xmath42 .
let us split the set of such circuits into the following three groups : 1 .
the vertex @xmath22 is adjacent to @xmath10 and @xmath43 is adjacent to @xmath9 ; 2 .
the vertex @xmath22 is adjacent to @xmath10 , but @xmath43 is not adjacent to @xmath9 ; 3 .
the vertex c is not adjacent to b. all edge @xmath2-gons in @xmath56 passing through the four points @xmath44 admit a similar classification .
now , the @xmath2-gons in @xmath12 belonging to the second class are in one - to - one correspondence with the @xmath2-gons in @xmath56 containing the path @xmath48 : just preserve the other edges in each @xmath2-gon . in order to show that the signs of the corresponding circuits in both directed graphs coincide
, it suffices to consider all possible mutual positions of the four chords @xmath44 in the chord diagram @xmath8 .
there are essentially two different such positions , depending on whether the chords @xmath22 and @xmath43 intersect one another . in both cases , for a chosen orientation of the chords , we easily check coincidence of the signs of the corresponding @xmath2-gons .
the edge @xmath2-gons of the third kind are the same in both graphs @xmath12 and @xmath57 .
finally , the edge @xmath2-gons of the first kind in each of the two directed intersection graphs come in pairs : the chain @xmath42 can be replaced with the chain @xmath48 . in order to prove that the signs of the two @xmath2-gons in a pair are opposite , it suffices to consider all possible mutual positions of the four chords @xmath44 in the chord diagram @xmath8 .
hence , the required assertion follows .
different chord diagrams can have one and the same intersection graph .
below , we show that the value of the weight system @xmath0 depends on the intersection graph of the chord diagram rather than on the diagram itself .
a natural question then arises , whether @xmath0 can be extended to a @xmath7-invariant of graphs .
we show that this is true for @xmath58 and @xmath59 .
the case of arbitrary @xmath60 is discussed in sec .
[ s5 ] .
[ tig ] the weight system @xmath0 acquires the same value on any two chord diagrams with coinciding intersection graphs .
since weight systems taking the same values on chord diagrams with coinciding intersection graphs are in one - to one correspondence with finite order knot invariants not distinguishing mutant knots @xcite , we conclude the following : the canonical knot invariant associated to the weight system @xmath0 does not distinguish mutant knots .
the proof of the theorem is based on a statement in @xcite giving a complete description of the situations where two chord diagrams have the same intersection graph .
we start with the definition of a share .
a _ share _ is a part of a chord diagram consisting of two arcs of the outer circle possessing the following property : each chord one of whose ends belongs to these arcs has both ends on these arcs .
the complement of a share also is a share .
the whole chord diagram is its own share whose complement contains no chords .
a _ mutation of a chord diagram _ is another chord diagram obtained by a rotation / reflection of a share ; see fig .
[ fm ] . obviously , mutations preserve the intersection graphs of chord diagrams .
we call the subgraph of the intersection graph induced by the vertices corresponding to chords forming a given share also a share in the intersection graph
. @xcite [ cd - mut - theorem ] two chord diagrams have the same intersection graph if and only if they are related by a sequence of mutations . * proof * of theorem [ tig ] .
we are going to prove that a mutation of a chord diagram does not change the value of the weight system @xmath0 , for arbitrary @xmath1 .
it suffices to consider only reflection and rotation of a share ; for the composition of reflection and rotation , the result will follow automatically .
pick an arbitrary orientation of chords in a chord diagram .
a mutation of a chord diagram with oriented chords produces a chord diagram with oriented chords in a natural way . for a given share in the intersection graph , its reflection results in inverting the orientation of all the arrows belonging to the corresponding share in the intersection graph . similarly , rotation of a share results in inverting all the arrows between the two shares in the intersection graph .
denote the two sets of vertices belonging to the complementary shares in the intersection graph by @xmath61 and @xmath62 , respectively .
the set @xmath61 contains a subset @xmath63 and the set @xmath62 contains a subset @xmath64 such that * any vertex from @xmath65 is connected to all vertices of @xmath66 ( and vice versa ) ; * each edge connecting vertices from @xmath61 and @xmath62 connects , in fact , a vertex from @xmath65 to a vertex from @xmath66 .
denote the set of all edges in the intersection graph connecting a vertex from @xmath65 to a vertex from @xmath66 by @xmath67 .
the graph with the set of vertices @xmath68 and the set of edges @xmath67 is the complete bipartite graph with the parts @xmath69 .
any circuit in the intersection graph contains an even number of edges from @xmath67 .
indeed , any path starting in @xmath61 switches between @xmath61 and @xmath62 after each passing through an edge in @xmath67 .
since a circuit returns to the original vertex , the number of such passings must be even .
now , the rotation mutation changes the orientation of all arrows in @xmath67 , whence of an even number of arrows in any circuit .
in particular , it changes orientation of an even number of arrows in any @xmath2-gon , hence preserving its sign .
pick an arbitrary orientation of the chord diagram such that all the arrows in @xmath67 are oriented from @xmath61 to @xmath62 .
such an orientation always exists .
indeed , if the share @xmath61 consists of two arcs @xmath70 and the share @xmath62 consists of two arcs @xmath71 following along the positive direction of the outer circle in the alternating order , @xmath72 , see fig .
[ forient ] , then it suffices to orient all the chords connecting @xmath70 and @xmath71 from @xmath73 to @xmath74 and from @xmath75 to @xmath76 , respectively . what happens if we reflect the share @xmath62 ?
if the @xmath2-gon we consider has even number of arrows in the share @xmath62 , then its sign remains the same .
the circuits with odd number of arrows in @xmath62 split into pairs in the following way .
the circuit intersects the set @xmath67 by even number of arrows .
let us split these arrows into pairs : two edges belong to one and the same pair iff their ends in the share @xmath62 are connected by a path that is a part of the circuit totally contained in @xmath62 .
for such pair of arrows , let the first one be @xmath77 and the second one be @xmath78 , @xmath79 , @xmath80 , @xmath81 ( it may well happen that either @xmath82 or @xmath83 , but not both ) . by _ switching pairs _ we mean replacing each pair of edges @xmath77 , @xmath78 , with the pair @xmath84 , @xmath85 ; see fig .
[ fsp ] . switching all pairs for a circuit
produces a new circuit .
this transformation is an involution : switching all pairs in this new circuit restores the original one .
thus , all circuits are split into pairs .
we are going to show that any two circuits of even length @xmath2 having odd number of arrows in @xmath62 and belonging to the same pair have opposite signs . indeed , if we choose directions in both circuits that coincide inside the share @xmath61 , then they are opposite inside the share @xmath62 .
both circuits have the same number of edges inside @xmath67 in either direction .
therefore , the only difference in the orientation is inside the share @xmath62 , where each arrow has opposite directions in the two circuits . since the number of arrows in @xmath62 is odd , the signs of the two circuits are opposite . thus , both kinds of mutation preserve the value of @xmath0 .
theorem [ tig ] above shows that each weight system @xmath0 defines a function on intersection graphs ; we denote this function by @xmath0 as well . in this section
we prove the following special case of a general theorem proved in sec .
[ s5 ] .
[ pge ] the functions @xmath52 and @xmath86 can be extended to @xmath7-invariants of graphs .
* proof . * in order to prove that @xmath0 can be extended to a @xmath7-invariant of graphs , it suffices to prove this for graphs with exactly @xmath2 vertices .
indeed , denote by @xmath87 the function whose values on graphs with @xmath2 vertices coincides with the constructed extension , and which vanishes on all other graphs
. then the convolution @xmath88 provides us with the desired extension to arbitrary graphs ; here @xmath61 is the function whose value on the arbitrary graph is equal to @xmath89 .
it is obvious that when restricted to intersection graphs , @xmath88 coincides with @xmath0 . here
we make use of the fact that the vector space spanned by graphs carries a natural graded bialgebra structure : the product of two graphs is given by their disjoint union , and the coproduct @xmath90 of a graph @xmath14 is given by @xmath91 where @xmath92 is the subgraph of @xmath14 induced by a subset @xmath93 of its vertices .
the convolution product , denoted by @xmath94 , is induced on the dual space of graph invariants from the coproduct on the space of graphs @xcite : @xmath95 for arbitrary graph invariants @xmath24 and @xmath96 .
these operations naturally descend to the @xmath7-bialgebra of graphs and @xmath7-invariants , respectively .
if @xmath58 , then each graph with @xmath97 vertices is an intersection graph , and any @xmath7-term relation for intersection graphs has a chord diagram counterpart , meaning we are done . for @xmath59
, there are two graphs with 6 vertices that are not intersection graphs , namely , the @xmath98-wheel and the @xmath99-prism ; see fig .
there are three ways to express the @xmath98-wheel as a linear combination of intersection graphs through the @xmath7-term relation , and for the @xmath99-prism there are two such ways ; see fig .
a direct computation shows that in both cases all the different representations provide the same value , which one can admit for the value of the extended graph invariant @xmath86 : for the @xmath98-wheel it is @xmath100 , and for the @xmath99-prism it is @xmath101 .
now , a direct verification , taking into account those @xmath7-term relations for graphs that do not have chord diagram counterparts , ensures that this extended invariant of graphs with @xmath102 vertices is indeed a @xmath7-invariant .
( 300,450)(10,20 ) ( 55,551 ) ( 55,575 ) ( 34,562 ) ( 76,562 ) ( 40,535 ) ( 70,535 ) ( 55,550)(0,1)24 ( 55,551)(2,1)21 ( 55,551)(-2,1)21 ( 55,550)(1,-1)16 ( 55,550)(-1,-1)16 ( 40,535)(1,0)32 ( 55,575)(2,-1)21 ( 55,575)(-2,-1)21 ( 40,535)(-1,4)7 ( 70,535)(1,4)7 ( 90,551)@xmath26 ( 28,532)@xmath9 ( 72,532)@xmath10 ( 135,551 ) ( 135,575 ) ( 114,562 ) ( 156,562 ) ( 120,535 ) ( 150,535 ) ( 135,550)(0,1)24 ( 135,551)(2,1)21 ( 135,551)(-2,1)21 ( 135,550)(1,-1)16 ( 135,550)(-1,-1)16 ( 135,575)(2,-1)21 ( 135,575)(-2,-1)21 ( 120,535)(-1,4)7 ( 150,535)(1,4)7 ( 170,551)@xmath27 ( 108,532)@xmath9 ( 152,532)@xmath10 ( 215,551 ) ( 215,575 ) ( 194,562 ) ( 236,562 ) ( 200,535 ) ( 230,535 ) ( 215,550)(0,1)24 ( 215,551)(2,1)21 ( 215,551)(-2,1)21 ( 215,550)(1,-1)16 ( 200,536)(3,2)36 ( 200,535)(1,0)32 ( 215,575)(2,-1)21 ( 215,575)(-2,-1)21 ( 200,535)(-1,4)7 ( 230,535)(1,4)7 ( 250,551)@xmath26 ( 188,532)@xmath9 ( 232,532)@xmath10 ( 295,551 ) ( 295,575 ) ( 274,562 ) ( 316,562 ) ( 280,535 ) ( 310,535 ) ( 295,550)(0,1)24 ( 295,551)(2,1)21 ( 295,551)(-2,1)21 ( 295,550)(1,-1)16 ( 280,536)(3,2)36 ( 295,575)(2,-1)21 ( 295,575)(-2,-1)21 ( 280,535)(-1,4)7 ( 310,535)(1,4)7 ( 268,532)@xmath9 ( 312,532)@xmath10 ( 128,515)@xmath101 ( 205,515)@xmath100 ( 285,515)@xmath101 ( 55,451 ) ( 55,475 ) ( 34,462 ) ( 76,462 ) ( 40,435 ) ( 70,435 ) ( 55,450)(0,1)24 ( 55,451)(2,1)21 ( 55,451)(-2,1)21 ( 55,450)(1,-1)16 ( 55,450)(-1,-1)16 ( 40,435)(1,0)32 ( 55,475)(2,-1)21 ( 55,475)(-2,-1)21 ( 40,435)(-1,4)7 ( 70,435)(1,4)7 ( 90,451)@xmath26 ( 28,432)@xmath9 ( 55,457)@xmath10 ( 135,451 ) ( 135,475 ) ( 114,462 ) ( 156,462 ) ( 120,435 ) ( 150,435 ) ( 135,450)(0,1)24 ( 135,451)(2,1)21 ( 135,451)(-2,1)21 ( 135,450)(1,-1)16 ( 120,435)(1,0)32 ( 135,475)(2,-1)21 ( 135,475)(-2,-1)21 ( 120,435)(-1,4)7 ( 150,435)(1,4)7 ( 170,451)@xmath27 ( 108,432)@xmath9 ( 135,457)@xmath10 ( 215,451 ) ( 215,475 ) ( 194,462 ) ( 236,462 ) ( 200,435 ) ( 230,435 ) ( 215,450)(0,1)24 ( 215,451)(2,1)21 ( 215,451)(-2,1)21 ( 215,450)(1,-1)16 ( 215,450)(-1,-1)16 ( 201,436)(1,3)13 ( 200,436)(3,2)36 ( 215,475)(2,-1)21 ( 215,475)(-2,-1)21 ( 230,435)(1,4)7 ( 250,451)@xmath26 ( 188,432)@xmath9 ( 215,457)@xmath10 ( 295,451 ) ( 295,475 ) ( 274,462 ) ( 316,462 ) ( 280,435 ) ( 310,435 ) ( 295,450)(0,1)24 ( 295,451)(2,1)21 ( 295,451)(-2,1)21 ( 295,450)(1,-1)16 ( 280,436)(3,2)36 ( 281,436)(1,3)13 ( 295,475)(2,-1)21 ( 295,475)(-2,-1)21 ( 310,435)(1,4)7 ( 268,432)@xmath9 ( 295,457)@xmath10 ( 128,415)@xmath101 ( 205,415)@xmath101 ( 292,415)@xmath89 ( 55,351 ) ( 55,375 ) ( 34,362 ) ( 76,362 ) ( 40,335 ) ( 70,335 ) ( 55,350)(0,1)24 ( 55,351)(2,1)21 ( 55,351)(-2,1)21 ( 55,350)(1,-1)16 ( 55,350)(-1,-1)16 ( 40,335)(1,0)32 ( 55,375)(2,-1)21 ( 55,375)(-2,-1)21 ( 40,335)(-1,4)7 ( 70,335)(1,4)7 ( 90,351)@xmath26 ( 28,332)@xmath10 ( 55,357)@xmath9 ( 135,351 ) ( 135,375 ) ( 114,362 ) ( 156,362 ) ( 120,335 ) ( 150,335 ) ( 135,350)(0,1)24 ( 135,351)(2,1)21 ( 135,351)(-2,1)21 ( 135,350)(1,-1)16 ( 120,335)(1,0)32 ( 135,375)(2,-1)21 ( 135,375)(-2,-1)21 ( 120,335)(-1,4)7 ( 150,335)(1,4)7 ( 170,351)@xmath27 ( 108,332)@xmath10 ( 135,357)@xmath9 ( 215,351 ) ( 215,375 ) ( 194,362 ) ( 236,362 ) ( 200,335 ) ( 230,335 ) ( 215,350)(0,1)24 ( 215,351)(2,1)21 ( 215,350)(-1,-1)16 ( 200,335)(1,0)32 ( 215,375)(2,-1)21 ( 215,375)(-2,-1)21 ( 200,335)(-1,4)7 ( 230,335)(1,4)7 ( 250,351)@xmath26 ( 188,332)@xmath10 ( 215,357)@xmath9 ( 295,351 ) ( 295,375 ) ( 274,362 ) ( 316,362 ) ( 280,335 ) ( 310,335 ) ( 295,350)(0,1)24 ( 295,351)(2,1)21 ( 280,335)(1,0)32 ( 295,375)(2,-1)21 ( 295,375)(-2,-1)21 ( 280,335)(-1,4)7 ( 310,335)(1,4)7 ( 268,332)@xmath10 ( 295,357)@xmath9 ( 128,315)@xmath101 ( 205,315)@xmath101 ( 292,315)@xmath89 ( 55,270 ) ( 55,240 ) ( 35,280 ) ( 75,280 ) ( 35,230 ) ( 75,230 ) ( 35,230)(0,1)50 ( 35,230)(1,0)40 ( 35,280)(1,0)40 ( 75,230)(0,1)50 ( 55,240)(0,1)30 ( 35,230)(2,1)20 ( 35,280)(2,-1)20 ( 75,280)(-2,-1)20 ( 75,230)(-2,1)20 ( 90,251)@xmath26 ( 43,240)@xmath9 ( 58,265)@xmath10 ( 135,270 ) ( 135,240 ) ( 115,280 ) ( 155,280 ) ( 115,230 ) ( 155,230 ) ( 115,230)(0,1)50 ( 115,230)(1,0)40 ( 115,280)(1,0)40 ( 155,230)(0,1)50 ( 115,230)(2,1)20 ( 115,280)(2,-1)20 ( 155,280)(-2,-1)20 ( 155,230)(-2,1)20 ( 170,251)@xmath27 ( 123,240)@xmath9 ( 138,265)@xmath10 ( 215,270 ) ( 215,240 ) ( 195,280 ) ( 235,280 ) ( 195,230 ) ( 235,230 ) ( 195,230)(0,1)50 ( 195,230)(1,0)40 ( 195,280)(1,0)40 ( 235,230)(0,1)50 ( 215,240)(0,1)30 ( 195,230)(2,1)20 ( 195,280)(2,-1)20 ( 235,280)(-2,-1)20 ( 235,230)(-2,1)20 ( 195,280)(1,-2)20 ( 235,280)(-1,-2)20 ( 250,251)@xmath26 ( 203,240)@xmath9 ( 218,265)@xmath10 ( 295,270 ) ( 295,240 ) ( 275,280 ) ( 315,280 ) ( 275,230 ) ( 315,230 ) ( 275,230)(0,1)50 ( 275,230)(1,0)40 ( 275,280)(1,0)40 ( 315,230)(0,1)50 ( 275,230)(2,1)20 ( 275,280)(2,-1)20 ( 315,280)(-2,-1)20 ( 315,230)(-2,1)20 ( 275,280)(1,-2)20 ( 315,280)(-1,-2)20 ( 283,240)@xmath9 ( 298,265)@xmath10 ( 132,215)@xmath89 ( 205,215)@xmath100 ( 285,215)@xmath101 ( 55,170 ) ( 55,140 ) ( 35,180 ) ( 75,180 ) ( 35,130 ) ( 75,130 ) ( 35,130)(0,1)50 ( 35,130)(1,0)40 ( 35,180)(1,0)40 ( 75,130)(0,1)50 ( 55,140)(0,1)30 ( 35,130)(2,1)20 ( 35,180)(2,-1)20 ( 75,180)(-2,-1)20 ( 75,130)(-2,1)20 ( 90,151)@xmath26 ( 43,140)@xmath9 ( 78,125)@xmath10 ( 135,170 ) ( 135,140 ) ( 115,180 ) ( 155,180 ) ( 115,130 ) ( 155,130 ) ( 115,130)(0,1)50 ( 115,130)(1,0)40 ( 115,180)(1,0)40 ( 155,130)(0,1)50 ( 135,140)(0,1)30 ( 115,130)(2,1)20 ( 115,180)(2,-1)20 ( 155,180)(-2,-1)20 ( 170,151)@xmath27 ( 123,140)@xmath9 ( 158,125)@xmath10 ( 215,170 ) ( 215,140 ) ( 195,180 ) ( 235,180 ) ( 195,130 ) ( 235,130 ) ( 195,130)(0,1)50 ( 195,130)(1,0)40 ( 195,180)(1,0)40 ( 235,130)(0,1)50 ( 215,140)(0,1)30 ( 195,180)(2,-1)20 ( 235,180)(-2,-1)20 ( 235,130)(-2,1)20 ( 235,180)(-1,-2)20 ( 235,180)(-1,-2)20 ( 250,151)@xmath26 ( 203,140)@xmath9 ( 238,125)@xmath10 ( 295,170 ) ( 295,140 ) ( 275,180 ) ( 315,180 ) ( 275,130 ) ( 315,130 ) ( 275,130)(0,1)50 ( 275,130)(1,0)40 ( 275,180)(1,0)40 ( 315,130)(0,1)50 ( 295,140)(0,1)30 ( 275,180)(2,-1)20 ( 315,180)(-2,-1)20 ( 315,180)(-1,-2)20 ( 283,140)@xmath9 ( 318,125)@xmath10 ( 130,115)@xmath101 ( 205,115)@xmath101 ( 285,115)@xmath101
it is well known that for an arbitrary lie algebra @xmath103 endowed with a nondegenerate invariant scalar product @xmath104 one can construct a weight system @xmath105 with values in the center @xmath106 of the universal enveloping algebra @xmath107 .
invariancy of the scalar product means that @xmath108)=([x , y],z)$ ] for any three elements @xmath109 .
the @xmath99-dimensional lie algebra @xmath4 ( taken , for definiteness , over the field @xmath110 of complex numbers ) provides the first nontrivial example of this construction .
since the center @xmath111 of the universal enveloping algebra @xmath112 is isomorphic to the ring of polynomials in the casimir element @xmath113 of @xmath4 , we obtain a weight system with values in the ring @xmath114 $ ] of polynomials in a single variable @xmath113 ; see @xcite and @xcite .
we denote this weight system by @xmath115 $ ] . here
@xmath116 denotes the graded vector space spanned by chord diagrams modulo the @xmath7-term relations ; @xmath117 where the finite dimensional vector space @xmath118 is spanned by chord diagrams with @xmath5 chords , modulo the @xmath7-term relations .
the space @xmath116 is a graded commutative cocommutative hopf algebra , with a multiplication @xmath119 , and a comultiplication @xmath120 .
the weight system @xmath121 is an algebra homomorphism .
it takes an arbitrary chord diagram with @xmath5 chords to a polynomial of degree @xmath5 in @xmath113 with the leading coefficient @xmath89 .
the invariant nondegenerate scalar product @xmath104 on @xmath4 can be chosen in a unique way up to a nonzero multiplicative constant .
the choice of the multiplicative constant affects the non - leading coefficients of the value of @xmath121 on a chord diagram .
below , we use the normalization chosen in @xcite , which differs from that in @xcite and @xcite .
being a graded commutative cocommutative hopf algebra , @xmath116 is generated by its primitive elements .
recall that an element @xmath122 of a hopf algebra is said to be _ primitive _ if @xmath123 . primitive elements
form a graded vector subspace @xmath124 of @xmath116 , @xmath125 for each @xmath126 , there exists a natural projection @xmath127 of the space of chord diagrams of order @xmath5 to the subspace @xmath128 of primitive elements along the subspace @xmath129 spanned by decomposable chord diagrams . a chord diagram with @xmath5 chords
is said to be _
decomposable _ if it can be represented as a product of two chord diagrams , each with less than @xmath5 chords .
the following statement is an immediate consequence of a result obtained in @xcite , theorem 3 , for an arbitrary simple lie algebra .
@xcite the value of the weight system @xmath121 on an arbitrary element in @xmath128 is a polynomial of degree at most @xmath130 in @xmath113 .
now we are able to formulate the conjecture mentioned in the introduction , which relates the weight systems @xmath0 and @xmath121 .
[ conj - sl2-hc ] for any chord diagram @xmath8 with @xmath2 chords , the coefficient of @xmath3 in the value @xmath131 is @xmath132 .
we did not manage to prove the conjecture , but the computer experiments show that it is true up to @xmath133 .
below , we present table 1 , which contains the values of the weight systems @xmath134 and @xmath135 for some chord diagrams with @xmath2 chords .
the original definition of a weight system @xmath136 is too cumbersome from the computational point of view , which makes it difficult to be computed even for chord diagrams with very few chords .
the computations we made used the recurrence relation for @xmath121 obtained in @xcite , theorem 2 : if a chord diagram contains a leaf , that is , a chord intersecting only one other chord , then the value of the @xmath4 universal weight system on the diagram is @xmath137 times its value on the result of deleting the leaf . in addition , the value of the weight system @xmath121 on a chord diagram satisfies the recurrence relations shown in fig .
[ 7fig:31 ] .
# 1 # 1#2#3#4 ( -0.26,0.97)(0,1.035)(0.26,0.97 ) ( 0.26,0.97)(0.52,0.9)(0.71,0.71 ) ( 0.71,0.71)(0.9,0.52)(0.97,0.26 ) ( 0.97,0.26)(1.035,0)(0.97,-0.26 ) ( 0.97,-0.26)(0.9,-0.52)(0.71,-0.71 ) ( 0.71,-0.71)(0.52,-0.9)(0.26,-0.97 ) ( 0.26,-0.97)(0,-1.035)(-0.26,-0.97 ) ( -0.26,-0.97)(-0.52,-0.9)(-0.71,-0.71 ) ( -0.71,-0.71)(-0.9,-0.52)(-0.97,-0.26 ) ( -0.97,-0.26)(-1.035,0)(-0.97,0.26 ) ( -0.97,0.26)(-0.9,0.52)(-0.71,0.71 ) ( -0.71,0.71)(-0.52,0.9)(-0.26,0.97 ) # 1[20pt][23pt]#1 # 1w ( [ 20pt][30pt]#1 ) # 1w ( ) # 1#2#3#4#5#6#7 @xmath138 { } & \\
\hfill = 2 \wdi{#5 } - 2 \wdi{#6 } # 7 & \end{aligned}\ ] ] # 1[20pt][23pt]#1 # 1b^(2)([20pt][30pt]#1 ) # 1w ( ) # 1#2#3#4#5#6#7 @xmath139 { } \wdi{#1}-\wdi{#2}-\wdi{#3}+\wdi{#4 } & \\
\hfill = \,\,\ , \wdi{#5 } - \wdi{#6 } # 7 \phantom{verylongstring } & \end{aligned}\ ] ] ; .
the projection to the subspace of primitive elements along the subspace of decomposable elements has the following form .
@xcite the projection @xmath140 takes a chord diagram @xmath8 with @xmath5 chords to the linear combination @xmath141 where the summations run over all unordered partitions of the set @xmath142 of the chords in @xmath8 into disjoint union of nonempty subsets , and @xmath143 is the chord subdiagram induced from @xmath8 by the subset @xmath144 of chords .
@xmath145 & @xmath12 & @xmath146 + & & @xmath147 & & + @xmath148 & ( 20,30)(0,-2 ) ( 0,10 ) ( 0,-2)(0,1)24 ( -8,2)(1,1)16 ( 8,2)(-1,1)16 ( -12,12)(2,-3)9 & @xmath149 & ( 20,30)(0,0 ) ( 0,10 ) ( 10,0 ) ( 10,20 ) ( 20,20 ) ( 0,10)(1,-1)10 ( 0,10)(1,1)10 ( 10,20)(0,-1)20 ( 10,20)(1,0)10 & @xmath150 + & & @xmath151 & & + @xmath148 & ( 20,30)(0,-2 ) ( 0,10 ) ( -10,4)(1,0)20 ( -8,2)(0,1)16 ( -10,16)(1,0)20 ( 8,2)(0,1)16 & @xmath152 & ( 20,30)(0,0 ) ( 0,0 ) ( 20,0 ) ( 0,20 ) ( 20,20 ) ( 0,0)(1,0)20 ( 0,0)(0,1)20 ( 20,20)(0,-1)20 ( 20,20)(-1,0)20 & @xmath89 + & & @xmath153 & & + @xmath148 & ( 20,30)(0,-2 ) ( 0,10 ) ( -10,4)(1,0)20 ( -8,2)(1,1)16 ( -10,16)(1,0)20 ( 8,2)(-1,1)16 & @xmath154 & ( 20,30)(0,0 ) ( 0,0 ) ( 20,0 ) ( 0,20 ) ( 20,20 ) ( 0,0)(1,0)20 ( 0,0)(0,1)20 ( 0,0)(1,1)20 ( 20,20)(0,-1)20 ( 20,20)(-1,0)20 & @xmath89 + & & @xmath155 & & + @xmath148 & ( 20,30)(0,-2 ) ( 0,10 ) ( 0,-2)(0,1)24 ( -12,10)(1,0)24 ( -8,2)(1,1)16 ( 8,2)(-1,1)16 & @xmath156 & ( 20,30)(0,0 ) ( 0,0 ) ( 20,0 ) ( 0,20 ) ( 20,20 ) ( 0,0)(1,0)20 ( 0,0)(0,1)20 ( 0,0)(1,1)20 ( 20,20)(0,-1)20 ( 20,20)(-1,0)20 ( 0,20)(1,-1)20 & @xmath89 + & & @xmath157 & & + @xmath99 & ( 20,30)(0,-2 ) ( 0,10 ) ( -10,16)(1,0)20 ( -10,4)(1,0)20 ( -12,8)(2,3)9 ( -12,12)(2,-3)9 ( 12,8)(-2,3)9 ( 12,12)(-2,-3)9 & @xmath158 & ( 20,30)(0,0 ) ( 3,0 ) ( 21,0 ) ( 3,24 ) ( 21,24 ) ( -3,12 ) ( 27,12 ) ( 3,0)(1,0)18 ( 3,24)(1,0)18 ( -3,12)(1,2)6 ( -3,12)(1,-2)6 ( 27,12)(-1,2)6 ( 27,12)(-1,-2)6 & @xmath89 + & & @xmath159 & & + @xmath99 & ( 20,30)(0,-2 ) ( 0,10 ) ( -10,16)(1,0)20 ( -10,4)(1,0)20 ( -12,8)(2,3)9 ( -12,12)(2,-3)9 ( 2,-2)(1,3)7 ( 2,22)(1,-3)7 & @xmath160 & ( 20,30)(0,0 ) ( 3,0 ) ( 21,0 ) ( 3,24 ) ( 21,24 ) ( -3,12 ) ( 27,12 ) ( 3,0)(1,0)18 ( 3,24)(1,0)18 ( -3,12)(1,2)6 ( -3,12)(1,-2)6 ( 27,12)(-1,2)6 ( 27,12)(-1,-2)6 ( -3,13)(5,2)24 ( 27,13)(-5,2)24 & @xmath150 + & & @xmath161 & & + @xmath99 & ( 20,30)(0,-2 ) ( 0,10 ) ( 0,-2)(0,1)24 ( -12,10)(1,0)24 ( -10,16)(2,-1)21 ( -6,21)(1,-2)11 ( 10,16)(-2,-1)21 ( 6,21)(-1,-2)11 & @xmath162 & ( 20,30)(0,0 ) ( 3,0 ) ( 21,0 ) ( 3,24 ) ( 21,24 ) ( -3,12 ) ( 27,12 ) ( 3,0)(1,0)18 ( 3,24)(1,0)18 ( -3,12)(1,2)6 ( -3,12)(1,-2)6 ( 27,12)(-1,2)6 ( 27,12)(-1,-2)6 ( -3,13)(5,2)24 ( 27,13)(-5,2)24 ( 3,0)(0,1)24 ( 3,1)(4,5)19 ( 3,1)(5,2)24 ( 21,0)(0,1)24 ( 21,1)(-4,5)19 ( 21,1)(-5,2)24 ( -3,12)(1,0)30 & @xmath163 + & & @xmath164 & & + here is more evidence supporting the conjecture : 1 .
the weight system @xmath121 , similarly to the weight system @xmath0 , depends on the intersection graphs of the chord diagram rather than on the diagram itself see @xcite ; 2 .
the value @xmath165 on a product of two nontrivial diagrams with @xmath2 chords in total is @xmath150 , since there are no @xmath2-gons in the intersection graph ; on the other hand , @xmath166 , since the diagram @xmath167 is decomposable ; 3 . for a chord diagram @xmath8 with @xmath2 chords having a leaf ( a chord intersecting only one other chord ) , @xmath168
; it can also be easily proved that the degree of the polynomial @xmath169 is less than @xmath60 .
the same argument as in the proof of proposition [ pge ] allows one to extend the value of the @xmath121 weight system to a @xmath7-invariant of graphs with up to @xmath102 edges .
namely , we can set the value of this invariant on the @xmath98-wheel to be equal to @xmath170 , with the projection to the subspace of primitive elements @xmath171 , and on the @xmath99-prism to be equal to @xmath172 , with the projection to the subspace of primitive elements @xmath173 .
after the first version of the present text has been spread as a preprint and submitted for publication , a paper by bar - natan and vo @xcite appeared . in this paper , conjecture [ conj - sl2-hc ]
is shown to be true , and now it can be stated as a theorem : one of the key ingredients in the proof is proposition 3.13 in @xcite which reads as follows .
denote by @xmath174 the @xmath175-valued function on chord diagrams equal to @xmath89 if the adjacency matrix of the intersection graph of the diagram is nondegenerate over @xmath31 , and equal to @xmath150 otherwise .
the function @xmath174 is extended to @xmath116 by linearity .
it is easy to show that @xmath174 satisfies the so - called @xmath148-term relation , @xmath176 , for any chord diagram @xmath8 and any pair of chords @xmath9 and @xmath10 with neighboring ends in it @xcite ; the @xmath7-term relation is an obvious corollary of the @xmath148-term one . indeed , the function @xmath174 is known to be extendable to a multiplicative @xmath175-valued @xmath7-invariant of graphs , see @xcite or @xcite : set @xmath178 to be equal to @xmath89 if the adjacency matrix of @xmath14 is nondegenerate over @xmath31 , and equal to @xmath150 otherwise .
it satisfies obviously the @xmath148-term relation for graphs , @xmath179 , for any graph @xmath14 and any pair of vertices @xmath180 in it . for intersection graphs with @xmath2 vertices ,
the projection of this graph invariant to primitive elements coincides with the invariant @xmath181 .
hence , the projection to primitive elements @xmath182 of this @xmath7-invariant on arbitrary graph @xmath14 with @xmath2 vertices is a @xmath7-invariant that coincides with @xmath183 if @xmath14 is an intersection graph . | we introduce a new series @xmath0 , @xmath1 , of integer valued weight systems .
the value of the weight system @xmath0 on a chord diagram is a signed number of cycles of even length @xmath2 in the intersection graph of the diagram .
we show that this value depends on the intersection graph only .
we check that for small orders of the diagrams , the value of the weight system @xmath0 on a diagram of order exactly @xmath2 coincides with the coefficient of @xmath3 in the value of the @xmath4-weight system on the projection of the diagram to primitive elements . |
an outstanding achievement of modern cosmology has been the detection of anisotropies in the cosmic microwave background @xcite .
these anisotropies provide a snapshot of the universe about 400,000 years after the hot big bang , just as the universe became transparent to electromagnetic radiation . despite this observational triumph , as yet it has been insufficient to differentiate between competing paradigms for galaxy formation , nor has it shed much light on cosmological processes taking place before photon decoupling .
there are , however , other types of radiation , such as gravitational radiation , which penetrate through this electromagnetic surface of last scattering , travelling virtually unaffected since their emission , even when this occurred in the first few fractions of a second .
of course , for gravitons this remarkable transparency is due to their very weak interactions with ordinary matter which , in turn , makes them difficult to observe .
however , pioneering experiments have been proposed which could detect stochastic backgrounds of gravitational waves generated in the early universe over a range of frequencies . gravitational radiation detected with a particular frequency
today would have been created at a characteristic time ( if generated by classical , causal mechanisms ) .
assuming that the radiation is emitted at a time @xmath0 before equal matter - radiation ( @xmath1 years ) and that it is created with a wavelength comparable to the horizon @xmath2 , then its frequency today is given by @xmath3 where the redshift @xmath4 and ` little @xmath5 ' is the rescaled hubble constant lying in the range @xmath6 .
the quantity that is measured by a particular experiment is the dimensionless amplitude @xmath7 where @xmath8 is the gravitational radiation density contribution to the universe s density per octave at a frequency @xmath9 , and @xmath10 is the critical energy density .
clearly , from ( [ eqnone ] ) gravitational radiation with higher frequencies will be more difficult to detect .
there are currently four frequency bands available for studying gravitational radiation @xcite .
first , there are the microwave background anisotropies created by tensor modes which may have been detected by the cobe - dmr experiment . in inflationary scenarios ,
the precise signal ratio due to tensor versus scalar modes is sensitively model - dependent , but it could be as high as 50% .
assuming that the entire cobe signal is due to gravitational waves implies a weak upper bound of @xmath11 at a frequency of @xmath12 .
a second upper limit on @xmath13 comes from pulsars , since a stochastic background would lead to timing noise in measurements of the incoming periodic electromagnetic signal .
a recent limit is @xmath14 at @xmath15 @xcite , though the statistical veracity of this result has been questioned with re - analyses of the same data suggesting both stronger and weaker bounds @xcite .
thirdly , proposed ground based detectors will study frequencies around 100hz with a maximum sensitivity of about @xmath16 for the first generation of ligo detectors and @xmath17 for the next generation ligo and virgo detectors . finally , the proposed space - based interferometer lisa will have a sensitivity of @xmath17 at about @xmath18 .
the inflationary paradigm @xcite is strongly motivated because it resolves a number of shortcomings of the standard hot big bang cosmology , including the horizon , flatness and monopole problems . however , possibly the most significant testable prediction of inflationary scenarios is a nearly scale - invariant spectrum of adiabatic density perturbations produced through quantum mechanical effects when the very early universe is dominated by the potential energy of some scalar inflaton field @xmath19 .
an important by - product of this process is the excitation of tensor modes resulting in a background of gravitational waves .
the spectra for scalar and tensor modes created in a slow - roll inflationary model are the following , @xmath20 where @xmath21 is the hubble parameter during inflation , @xmath22 is the value of the inflaton field when the mode with comoving wavenumber @xmath23 leaves the hubble radius @xmath24 , and the slow - roll parameter is @xmath25 with the prime @xmath26 denoting differentiation with respect @xmath19 .
here , @xmath27 is related to the speed at which @xmath19 rolls down the potential @xmath28 ( less than unity during inflation ) .
once a particular mode is created , it is driven outside the hubble radius ( or ` horizon ' ) by the rapid expansion and it effectively ` freezes ' until it returns inside the horizon during the subsequent radiation and matter dominated eras .
lengthscales corresponding to the size of the observed universe went outside the horizon early when the inflaton field had a value denoted by @xmath29 ; after this there were about 60 e - foldings of expansion before the end of inflation when @xmath30
. for a specific inflaton potential @xmath28 , the dynamics of the hubble parameter in the hamilton - jacobi formalism is given by @xcite @xmath31 the hubble parameter monotonically decreases during slow - roll inflation , so from ( [ spectrum ] ) the largest contribution to the gravitational wave spectrum is due to modes that were driven outside the horizon early in inflation and have just come back inside the horizon at the present day .
these scales correspond to those probed by the cobe - dmr experiment and the observed anisotropies can be used with ( [ spectrum ] ) to normalize @xmath32 . given this initial condition , ( [ eom ] ) can be used to calculate the hubble parameter during inflation and hence the resulting spectrum of gravitational waves .
paradoxically , as we shall see , the larger the contribution of a model to the cobe signal , the smaller the signal will be in the frequency bands corresponding to lisa and ligo / virgo ; the hubble parameter must decrease more rapidly in order to maintain the condition that there be 60 e - foldings before the end of inflation @xcite . in order to illustrate this point , we shall calculate the spectrum of gravitational radiation produced by an inflationary model similar to polynomial chaotic inflation , but slightly modified to make analytic calculations tractable @xcite . using a simple method to obtain the cobe normalization @xcite , the potential for this model
is given by @xmath33 where @xmath34 and we require @xmath35 .
for this model , the slow - roll and hubble parameters are given by @xmath36 for @xmath37 , substituting into ( [ spectrum ] ) and appropriately redshifting yields @xmath38^{\alpha/2}\,.\ ] ] for @xmath39 , that is , scales which come inside the horizon later during the matter era , we have less redshifting and ( [ chaotic ] ) must be multiplied by an additional factor @xmath40 .
to delineate the predictions of different slow - roll inflation models , one can use different values of @xmath41 ; the obvious limiting cases being @xmath42 , the simplest chaotic model , and @xmath43 ( say ) for which almost all the cobe signal is due to gravitational waves ( this case corresponds closely to power law inflation with an exponential potential ) .
the results of these cases are plotted in figure 1 .
the @xmath42 model has a low contribution to cobe but the spectrum for @xmath37 decreases slowly down to an amplitude of @xmath44 at a few hz .
in contrast , the model with @xmath43 has a much larger contribution to cobe , but has a very small amplitude for scales @xmath37 ; in the lisa frequency band this has already fallen to @xmath45 , eight orders of magnitude smaller than for @xmath42 .
nonetheless , both these models produce signals well below the sensitivity of proposed interferometers by several orders of magnitude .
the largest possible cobe normalization provides a weak upper limit on the amplitude of gravitational waves for all scales @xmath37 which must be below the level @xmath46 .
it may be possible to carefully construct a slow - roll potential which would almost achieve this limit in the lisa and ligo / virgo frequency bands , but undoubtedly the potential would look rather contrived . in any case , it is clear that the prospects for a detectable slow - roll signal are bleak , so we shall have to consider less orthodox inflationary models . in the next section ,
we consider extended and hybrid inflation , but there are also more speculative superstring - inspired models which can give a larger gravitational wave background ( for example see @xcite ) . 0.5in5.0 in
cosmological phase transitions in which symmetries are spontaneously broken are now an integral part of modern cosmology . in a first - order phase transition
the field becomes trapped in a metastable local minimum of the potential the false vacuum .
the transition to the true vacuum takes place by the nucleation of bubbles and their subsequent rapid growth .
when these vacuum bubbles collide copious amounts of gravitational radiation can be emitted , particularly if the relative velocity of the walls is relativistic as in strongly first - order phase transitions .
it has been estimated that the maximum contribution to the gravitational wave background from such a transition would be @xcite @xmath47 at a frequency of @xmath48 where @xmath49 is the nucleation rate of bubbles , @xmath50 is the temperature of the phase transition , @xmath51 is the relevant hubble parameter and @xmath52 is the number of relativistic degrees of freedom . assuming that the electroweak phase transition is strongly first - order implies @xmath53 at a frequency of @xmath54 , inside the lisa band with a detectable amplitude , as illustrated in figure 1 .
however , the minimal standard model is currently believed to have only a weakly first - order transition , if not second - order , with the bubble walls reaching velocities well below the speed of light . in this case , @xmath55 has been estimated to be around @xmath56 which is well outside lisa s sensitivity @xcite .
nevertheless , there are a number of well - motivated extensions to the standard model ( like supersymmetry ) which could entail symmetry - breaking just above the electroweak scale ; if strongly first - order , such phase transitions would create a distinctive and detectable lisa signal . fortuitously from this point of view , lisa has a very interesting frequency response range .
while the prospects for detecting gravitational waves from slow - roll inflation seem poor , there are a number of inflationary models which do not end in the standard re - heating scenario .
these include extended inflation @xcite and hybrid inflation @xcite which exit through a phase transition ; this provides an extra , potentially more powerful source of gravitational waves if this final phase transition is strongly first - order @xcite .
the radiation could be detected by the advanced ligo and virgo detectors if the re - heat temperature of the universe was close to @xmath57gev or @xmath58gev scales which are far from ruled out by current observations .
we have included a sample spectra for such a model in figure 1 .
cosmic strings are line - like topological defects which may form during phase transitions in the early universe ( see , for example , @xcite ) . due to the large string mass per unit length ( typically about @xmath59 for gut scale strings ) , they may have acted as the initial seeds for the formation of large - scale structure @xcite . in general , a network of strings will evolve towards a self - similar scaling solution by the production of loops and the subsequent emission of radiation into a preferred channel . for local strings ,
this channel is gravitational radiation , thus creating a stochastic background of radiation over frequencies from @xmath60hz to @xmath61hz .
scale - invariant string evolution implies that the density of the string network @xmath62 remains constant relative to the background density .
statistically , the network has the same properties at any two times , except for a universal scaling with respect to the growing horizon size , that is , there exists a characteristic length scale @xmath63 .
this scaling regime has been shown to exist numerically @xcite and various parameters have been estimated , notably @xmath64 during the radiation era .
the radiative dynamics of the strings is a crucial factor determining the precise contributions at various frequencies .
the power radiated by a string loop is normally written as a sum over loop harmonics @xcite , that is , @xmath65 where @xmath66 @xcite is a constant parametrized solely by the particular loop trajectory and not its length , @xmath67 is the string mass per unit length and @xmath68 is some cut - off introduced by the effects of radiation backreaction @xcite ( see also @xcite ) . assuming that @xmath69 is constant and for reasonable values of @xmath70
, one can deduce that the spectrum of gravitational radiation due to loops created during the radiation era is independent of frequency and only very weakly dependent on the loop radiation spectrum .
the amplitude of this flat spectrum is given by @xmath71 where @xmath72 , @xmath73 , @xmath74 and @xmath41 is the constant loop production size with respect to the horizon .
in contrast , the radiation background today due to loops produced in the matter era is highly dependent on the loop radiation spectrum .
assuming that @xmath75 , one can deduce that @xmath76=a+(1-q)\log_{10}[\omega]$ ] .
therefore , the amount of radiation produced in the matter era which feeds into higher frequencies is extremely sensitive to @xmath77 and @xmath70 .
it is precisely these frequencies that are relevant for the pulsar timing experiment @xcite .
one can investigate these effects more fully as in ref .
@xcite using a numerical algorithm which calculates the spectrum of radiation for given spectra and cosmological parameters @xcite .
the parameters that were used are @xmath78 , @xmath79 and the electroweak model particle spectrum with three lepton generations .
it was found that the contribution to the pulsar timing frequency was the same for all @xmath80 irrespective of the value of @xmath70 , while for @xmath81 it was sensitive to the value of @xmath70 . in particular , for @xmath82 and @xmath83 the amplitude of the spectrum will be in conflict with the constraint from pulsar timings for the string energy density @xmath84 normalised to cobe @xcite .
however , if @xmath82 and @xmath85 then the contribution to the pulsar timing frequency will be the same as for @xmath86 and @xmath83 .
the spectrum of radiation emitted by loops is likely to be cut off by the effects of backreaction as shown numerically for goldstone boson radiation from global cosmic strings @xcite , therefore it seems sensible to exclude the @xmath82 , large @xmath70 spectra on physical grounds ( as we discuss in ref .
@xcite ) . using well - motivated values of @xmath77 and @xmath70
, one can use the pulsar timing constraint to deduce that @xmath87 @xcite . using @xmath88
also reduces the contribution to the pulsar timing frequency from matter era loops due to curvature domination at @xmath89 , though we expect this effect to be small .
the contribution from cosmic strings to the lisa and ligo / virgo frequency ranges is likely to be slightly lower than that relevant for pulsar timing because particle mass thresholds cause entropy transfers which reduce the relative density contribution of any pre - existing decoupled radiation .
the cosmic string spectrum shown in figure 1 illustrates this effect with a gradual rise in the otherwise flat spectra at frequencies around @xmath90hz .
this particular rise is caused by particle annihilation near the qcd and electroweak phase transitions . if the particle physics model has more degrees of freedom at higher energies there could be other steps related to other phase transitions .
however , the dependence on @xmath69 is reasonably weak and therefore we can conservatively estimate that @xmath91 at @xmath92hz and @xmath93 at 100hz for @xmath94 .
we note also that very precise determinations of the stochastic background from cosmic strings at different frequencies would measure @xmath95 in different cosmological epochs , providing fascinating insight into the particle content of the early universe ( at times as early as @xmath96 for ligo and virgo ) .
finally , we should note that there are other types of cosmic strings which do not produce a large background of gravitational radiation , such as those formed when global symmetries are broken .
the dynamics of global strings is dominated by the production of goldstone bosons and a simple calculation suggests that gravitational waves will be suppressed by approximately four orders of magnitude ; the global string spectrum is also illustrated in figure 1 .
gravitational waves potentially provide a unique and unparallelled probe of the very early universe .
the proposed interferometers , both terrestrial and space - based , could rule out or severely constrain a number of viable theoretical models , most notably the cosmic string scenario . on the other hand
, the detection of a primordial background of gravitational waves would have a profound impact on our understanding of high energy physics and cosmology , providing an unprecedented view of the earliest moments after the creation of the universe .
we would like to thank rob caldwell for his collaboration in reference @xcite and we would like to acknowledge helpful discussions with bruce allen and alex vilenkin .
rab is supported by pparc personal fellowship grant gr / k94799 and eps is currently a pparc advanced fellow .
this work has also been supported by pparc rolling grant gr / k29272 .
# 1#2#3#4#5#6 # 1#2#3#4 # 1#2#3#4#5#6#1 [ # 2 ] , in _ # 4 _ , # 5 , eds .
( # 6 ) . #
1#2#3#4 # 1#2#3#4#5#6#7#8 | the next generation of gravitational wave detectors holds out the prospect of detecting a stochastic gravitational wave background generated in the very early universe . in this article , we review the various cosmological processes which can lead to such a background , including quantum fluctuations during inflation , bubble collisions in a first - order phase transition and the decay of a network of cosmic strings . we conclude that signals from strongly first - order phase transitions , possibly at the end of inflation , and networks of local cosmic strings are within the sensitivity of proposed detectors .
however , backgrounds from standard slow - roll inflation and the electroweak phase transition are too weak . epsf.tex |
in the case of spanish language for several thresholds , after logarithmic binning .
( inset panel ) non - collapsed distribution , for different thresholds . ]
* gutenberg - richter law . *
the energy @xmath14 released during voice events is a direct measure of the vocal fold response function under air pressure perturbations , and its distribution @xmath15 could in principle depend both on the threshold @xmath3 and on the language under study . in the inset of figure [ fig : figure_colapso_e_integrada ]
we observe that @xmath15 is power law distributed over about six decases , saturated by an exponential cut - off .
this distribution has been interpreted before as the analogue of a gutenberg - richter law in voice , as the precise shape of energy release fluctuations during voice production parallel those occurring in earthquakes @xcite . as increasing @xmath3 induces a flow in rg space , systems which lie close to a critical point ( unstable fixed point in rg space )
show scale invariance under @xmath3 and hence the distributions can be collapsed into a @xmath3-independent shape , thereby eliminating the trivial dependence on @xmath3 .
this has been shown to be the case for human voice and accordingly ( technical details can be found in the si ) we can express the collapsed energy distribution as @xmath16 for @xmath17 , where @xmath18 is the lower limit beyond which this law is fulfilled , @xmath19 is a scaling function and the relevant variable is @xmath20 , the scaling exponent . in order to collapse every curve @xmath15 the theory predicts to rescale @xmath21 . in the outset of figure [ fig : figure_colapso_e_integrada ] we show the result of this analysis for the case of spanish language , where we find that @xmath22 , this exponent being approximately language - independent ( see table [ table : tablas_slopes ] for other languages and si for additional details ) . interestingly ,
these exponents are compatible with those found in rainfall , another natural system that has been shown to be compatible with soc dynamics @xcite , and can not be explained by simple null models @xcite . in what follows
we explore the emergence of classical linguistic laws in these acoustic signals .
+ .summary of scaling exponents associated to the energy release distribution ( @xmath20 ) , zipf s law ( @xmath23 ) , heaps law ( @xmath24 ) and brevity law ( @xmath25 ) for the six different languages .
power law fits are performed using maximum likelihood estimation ( mle ) following clauset @xcite and goodness - of - fit test and confidence interval are based on kolmogorov - smirnov ( ks ) tests . in all cases ,
ks are greater than 0.99 .
exponents associated to energy release are compatible with those found in rainfall @xcite .
results are compatible with the hypothesis of language - independence . [ cols="^,^,^,^,^",options="header " , ] * zipf s law . * .
the inset panel shows the raw , threshold - dependent distributions and the outer panel zipf s law has been collapsed ( see the text for details ) . ]
the illustrious george kingsley zipf formulated a statistical observation which is popularly known as zipf s law @xcite . in its original formulation @xcite , it establishes that in a sizable sample of language the number of different words ( vocabulary ) @xmath26 which occur exactly @xmath27 times decays as @xmath28 , where the exponent @xmath23 varies from text to text @xcite but is usually close to @xmath29 .
an alternative and perhaps more common formulation of this law @xcite is defined in terms of the rank , such that if words are ranked in decreasing order by their frequency of appearance then the number of occurrences of words with a given rank @xmath30 goes like @xmath31 , where it is easy to see that both exponents are related via @xmath32 and thus @xmath33 approximates to 1 @xcite . here for convenience we make use of the former and explore @xmath34 applied to the statistics of types .
again @xmath34 could in principle depend on the threshold but assuming that the signal complies with the scale - invariance mentioned above , one can collapse all threshold - dependent curves into a universal shape and thus remove any dependence on this parameter by rescaling @xmath35 where @xmath36 is the total number of different types present in the signal and @xmath37 as the total number of tokens ( see si for technical details ) .
results are shown in the case of basque language in figure [ fig : figure_zipf ] , where a clear threshold - independent decaying power law emerges with a scaling exponent @xmath38 .
analogous results with compatible exponents for other languages can be found in si and table [ table : tablas_slopes ] .
null models systematically deviate from these results , and neither display the characteristic power law decay nor any invariance under variation of the energy threshold ( si ) . +
* heaps law .
* is related with the original ( see the text ) and leads to @xmath39 . results for other languages are found in table [ table : tablas_slopes ] . ] together with zipf s law and connected mathematically ( see @xcite and references therein ) , the second classical linguistic law is heaps law , the sublinear growth of the number of different words @xmath36 in a text with text size @xmath37 ( measured in total number of words ) : @xmath40 @xcite ( a constant rate for appearance of new words leads to @xmath41 ) . here
the vocabulary @xmath36 is defined as the total number of different types that appear in the signal , whereas @xmath37 is defined as the total number of tokens found for a given threshold .
results are shown for a specific language in figure [ fig : figure_heap ] ( see si for the rest ) . in the outset panel
we present the collapsed , threshold - independent curves , where again we find a scaling law with an effective exponent @xmath42 related to the original exponent @xmath43 . in this case equivalent computation on the null model yield a heaps law with the trivial exponent @xmath44 ( si ) .
+ these results are quantitatively consistent with previous results on written texts @xcite .
in particular , several authors @xcite point out that , at least asymptotically , the relation @xmath45 holds with good approximation , and this is on reasonably good agreement with our findings in human voice as well .
interestingly , a recent work @xcite has found that , as opposed to indoeuropean ( alphabetically based ) languages , zipf s law breaks down and heaps law reduces to the trivial case for written texts in chinese , japanese , korean and other logosyllabic languages . applying our methodology in a database of logosyllabic languages
could thus evaluate to which extent those differences arise also in human voice . + that describes the relative frequency of a type of mean duration @xmath46 . in every case
we find a monotonically decreasing curve which yields a brevity law . in the outset panel
we present the collapsed , threshold - independent curve @xmath47 , that evidences an initial power law decay with an exponent @xmath48 . ]
* brevity law . *
the tendency of more frequent words to be shorter @xcite can be generalized as the tendency of more frequent elements to be shorter or smaller , and its origin has been suggested to be related to optimization and information compression arguments in connection with other linguistic laws @xcite . in acoustics ,
spontaneous speech indeed tends to obey this law after text segmentation @xcite , and has been found also in other non - human primates @xcite .
here we can test brevity law in essentially two different ways .
first , note that voice events ( tokens ) map into types according to the logarithmic binning of their associated energy , hence voice events with different duration might yield the same type as previously noted .
thus for each type we can compute its mean duration @xmath46 averaging over all voice events that fall within that type , and then plot the histogram @xmath49 that describes frequency of each type versus its mean duration .
brevity law would require @xmath49 to be a monotonically decreasing function .
these results are shown for a particular language in log - log scales in figure [ fig : figure_brevity ] , finding initially a power law decaying relation which is indicative of a brevity law ( results are again found to be language independent , see si for additional results ) .
the inset provide the threshold - dependent distributions and the outset panel provides the collapsed , threshold - independent shape @xmath47 ( see table [ table : tablas_slopes ] for scaling exponents ) . again in this case results in null models deviate from such behavior ( si ) and are clearly different from the random typing @xcite .
alternatively , one can also directly observe the duration frequency at the level of voice events , finding similar results ( see si ) .
in this work we have explored the equivalent of linguistic laws directly in acoustic signals .
we have found that human voice -which actually complies to soc dynamics with critical exponents compatible with those found in rainfall @xcite- manifests the analog of classical linguistic laws found in written texts ( zipf s law , heaps law and the brevity law or law of abbreviation ) .
these laws are found to be invariant under variation of the energy threshold @xmath3 , and can be collapsed under universal functions accordingly . as @xmath3 is the only free parameter of the method , this invariance determines that the results are not afflicted by ambiguities associated to arbitrarily defining unit boundaries .
results appear to be robust across languages and timescales ( ranging six different indoeuropean languages and different scales extending all the way into the intraphoneme range , and invariant under energy threshold variation ) .
interestingly , an equivalent analysis performed on null models defined by randomizing the signal @xmath13 ( yielding white noise with the same instantaneous energy distribution of the original signal ) fail to reproduce this phenomenology ( si ) .
the concrete range of exponents found for both zipf and heaps laws are compatible between each other and somewhat similar -but not identical- to the typical ones observed in the literature for written texts @xcite , whereas to the best of our knowledge this is the first observation of scaling behavior with a clear exponent in the case of brevity law in speech .
actually , our finding of a power law in brevity law differs from the case of random typing where a power law does nt conform @xcite .
+ the specific and complex alternation of air stops ( silences ) intertwined with voice production are at the core of the microscopic voice fluctuations . during voice production , acoustic communication
is governed by the so - called biphasic cycle ( breath and glottal cycle , see @xcite for a review ) that together with some other acoustic considerations ( pitch period , voice onset time , the relation between duration , stress and syllabic structure @xcite ) determines the microscopic structure of human voice , including silence stops . however , these timescales are in general very large : as previously stated , this current study focus and scans voice properties even at intraphonemic timescales , where the statistical laws of language emerge directly from the physical magnitudes that govern acoustic communication .
our results therefore open the possibility of speculating whether the fact that these laws have been found in upper levels of human communication might be a result of a scaling process and a byproduct of the physics rather than derived from the choice of the typical units of study on the analysis of written corpus ( phonemes , syllabus , words , ... ) , like differences between analysis of indoeuropean and logosyllabic languages demonstrates @xcite . as a matter of fact , in a previous work human voice
has been framed within self - organized criticality ( soc ) , speculating that the fractal structure of lungs drives human voice close to a critical state @xcite , this mechanism being ultimately responsible for the microscopic self - similar fluctuations of the signal .
this constitutes a new example of the emergence of soc in a physiological system , different in principle from the classical one found in neuronal activity @xcite .
one could thus speculate that the emergence of these complex patterns is just a consequence of the presence of soc , what in turn would support the physical origin of linguistic laws . from an evolutionary viewpoint , under this latter perspective human voice , understood as a communication system which has been optimized under evolutionary pressures , would constitute an example where complexity ( described in terms of robust linguistic laws ) emerges when a system is driven close to criticality , something reminiscent of the celebrated edge of chaos hypothesis @xcite .
+ more generally , the method used and proposed here also addresses the longstansing problem of signal segmentation .
it has been acknowledged that there is no such thing as a correct segmentation strategy @xcite . in
written corpus white space is usually taken as an easy marker of the separation between words , however this is far from evident in continuous speech where separation between words or concepts is technologically harder to detect , conceptually vague and probably ill - defined .
few exceptions that used oral corpus for animal communication still require to define ad hoc segmentation algorithms @xcite , or manual segmentation strategies which usually give an arbitrary or overestimated segmenting times @xcite , what might even raise epistemological questions .
as such , this segmentation problem unfortunately has prevented wider , comparative studies in areas such as animal communication or the search for signs of possible extraterrestrial intelligence from radio signals ( in this line only few proposals have been made @xcite ) . by varying the energy threshold the method presented here automatically partitions and symbolizes the signal at various energy scales , providing a recipe to establish an automatic , general and systematic way of segmenting and thus enabling comparison of across acoustic signals of arbitrary origion for which we may lack the syntax , code or exact size of its constituents .
+ to round off , we hope that this work paves the way for new research avenues in comparative studies .
open questions that deserve further work abound ; just to name a few : in the light of this new method , what can we say about the acoustic structure in other animal communication systems ?
can we find evidence of universal traits in communication that do not depend on a particular species but are only physically and physiologically constrained , or on the other hand are linguistic universals a myth @xcite ?
how these laws evolve with aging @xcite ? are they affected by cognitive or pulmonary diseases ? what is the precise relation between soc and linguistic laws in this context ? and
in particular , can we find mathematical evidence of a minimal , analytically tractable soc model that produce these patterns ? these and other questions are interesting avenues for future work .
ha , l. q. , sicilia - garcia , e. i. , ming , j. , and smith , f. j. ( 2002 ) .
extension of zipf s law to words and phrases . in proceedings of the 19th international conference on computational linguistics - volume 1 , pages 1 - 6 .
association for computational linguistics ferrer i cancho , r. , riordan , o. , and bollobs , b. ( 2005 ) .
the consequences of zipf s law for syntax and symbolic reference .
proceedings of the royal society of london b : biological sciences , 272(1562):561 - 565 aylett , m. and turk , a. ( 2006 ) .
language redundancy predicts syllabic duration and the spectral characteristics of vocalic syllable nuclei .
the journal of the acoustical society of america , 119(5):3048 - 3058 tomaschek , f. , wieling , m. , arnold , d. , and baayen , r. h. ( 2013 ) .
word frequency , vowel length and vowel quality in speech production : an ema study of the importance of experience . in interspeech ,
pages 1302 - 1306 ferrer - i cancho , r. , hernndez - fernndez , a. , lusseau , d. , agoramoorthy , g. , hsu , m. j. , and semple , s. ( 2013 ) .
compression as a universal principle of animal behavior . cognitive science , 37(8):1565 - 1578 ferrer
i cancho , r. and hernndez - fernndez , a. ( 2013 ) . the failure of the law of brevity in two new world primates .
statistical caveats .
glottotheory : international journal of theoretical linguistics , 4(1):45 - 55 kello , c. t. , brown , g. d. , ferrer - i cancho , r. , holden , j. g. , linkenkaer - hansen , k. , rhodes , t. , and van orden , g. c. ( 2010 ) .
scaling laws in cognitive sciences .
trends in cognitive sciences , 14(5):223 - 232 kuhl , p. k. , conboy , b. t. , coffey - corina , s. , padden , d. , rivera - gaxiola , m. , and nelson , t. ( 2008 ) .
phonetic learning as a pathway to language : new data and native language magnet theory expanded ( nlm - e ) .
philosophical transactions of the royal society of london b : biological sciences , 363(1493):979 - 1000 mccowan , b. , hanser , s. f. , and doyle , l. r. ( 1999 ) . quantitative tools for comparing animal communication systems : information theory applied to bottlenose dolphin whistle repertoires .
animal behaviour , 57(2 ) gustison , m. l. , semple , s. , ferrer
i cancho , r. , and bergman , t. j. ( 2016 ) .
gelada vocal sequences follow menzerath s linguistic law .
proceedings of the national academy of sciences of the usa , 113(19):e2750-e2758 | linguistic laws constitute one of the quantitative cornerstones of modern cognitive sciences and have been routinely investigated in written corpora , or in the _ equivalent _ transcription of oral corpora .
this means that inferences of statistical patterns of language in acoustics are biased by the arbitrary , language - dependent segmentation of the signal , and virtually precludes the possibility of making comparative studies between human voice and other animal communication systems .
here we bridge this gap by proposing a method that allows to measure such patterns in acoustic signals of arbitrary origin , without needs to have access to the language corpus underneath .
the method has been applied to six different human languages , recovering successfully some well - known laws of human communication at timescales even below the phoneme and finding yet another link between complexity and criticality in a biological system .
these methods further pave the way for new comparative studies in animal communication or the analysis of signals of unknown code .
the main objective of quantitative linguistics is to explore the emergence of statistical patterns ( often called linguistic laws ) in language and general communication systems ( see @xcite and @xcite for a review ) .
the most celebrated of such regularities is zipf s law describing the uneven abundance of word frequencies @xcite .
this law presents many variations in human language @xcite but also shows ubiquity @xcite in many linguistic scales @xcite , has been claimed to be universal @xcite and has consequences for syntax and symbolic reference @xcite . on the other hand heaps law , also called herdan s law @xcite states that the vocabulary of a text grows allometrically with the text length @xcite , and is mathematically connected with zipf s law @xcite , the scaling exponent being dependent on both zipf law and the vocabulary size .
finally the zipf s law of abbreviation ( or brevity law for short ) is the statistical tendency of more frequent elements in communication systems to be shorter or smaller @xcite and has been recently claimed as an universal trend derived from fundamental principles of information processing @xcite . as such , this statistical regularity holds also phonetically @xcite and implies that the higher the frequency of a word , the the shorter its length or duration , probably caused by a principle of compression @xcite , although this is a general pattern that can change depending on other acoustical factors like noise @xcite , pressure to communicate at long distances calls @xcite or communicative efficiency @xcite and energetic constraints @xcite . + linguistic laws extend beyond written language and have been shown to hold for different biological data @xcite .
according to some authors @xcite , the presence of scaling laws in communication is indicative of the existence of processes taking place across different cognitive scales . interpreting linguistic laws as scaling
laws which emerged in communication systems @xcite actually opens the door for speculating on the existence of underlying scale - invariant physical laws operating underneath @xcite .
of course , in order to explore the presence or absence of such patterns one needs to directly study the acoustic corpus , i.e. human voice , as every linguistic phenomenon or candidate for language law can be camouflaged or diluted after the transcription process . notwithstanding
the deep relevance of linguistic laws reported in written texts , we still wonder up to which of these laws found in written corpora are related or derive from more fundamental properties of the acoustics of language -and are thus candidates for full - fledged linguistic laws- , or emerge as an artifact of scripture codification .
+ * linguistic laws in acoustics ? * acoustic communication is fully determined by three physical magnitudes extracted from the signals : frequency , energy and time @xcite . the configuration space induced by these magnitudes results from an intrinsic evolutionary relationship between the production and perception of sound systems @xcite that further shapes the range of hearing and producing sounds for a variety of life forms @xcite .
since animals use their acoustic abilities both to monitor their environment and to communicate we should expect that natural selection has in some sense optimized these sensorial capabilities , but , despite the great differences that evolution has involved for different animals that communicate acoustically , there are many similarities between their mechanisms of sound production and hearing @xcite . focusing on primates ,
traditionally language has been distinguished from vocalizations of nonhuman primates only at a qualitative , semantic level @xcite .
interestingly , it is well known that children use statistical cues to segment the input @xcite and probably share with non - human primates some of these mechanisms , albeit with some differences @xcite .
the discovery of these statistical learning abilities has boosted a new approach to the study of language @xcite and suggests that statistical learning of language could be based on patterns or more generally linguistic laws @xcite : research on language acquisition shows that higher frequency facilitates learning , and zipf s law tell us that vocabulary learning is easier than expected a priori given the skewness of the distribution , for instance .
we advance that , as will be shown below , these patterns are already present in the physical sound waves produced by human voice even at levels below those generally considered linguistically significant , i.e. , below the phoneme timescale . + empirical evidence of robust linguistic laws holding in written texts across different human languages has been reported many times ( see @xcite and references therein ) , and it has been shown that these laws are not fully observed in random texts @xcite .
studies with oral corpus are however much less abundant , and they systematically imply a transcription of the acoustical waves into words in the case of human speech or some ill - defined analog of words in the case of animal communication , as the main segments to analyze @xcite . a few current efforts take a different road and consider other possible written units such as lemmas @xcite or compare written and oral production for some linguistic patterns @xcite , in general showing that frequencies of elements in written corpora can be taken as a rough estimate for their frequency in spoken language @xcite .
all in all , the exploration of linguistic laws in oral corpora is scarce .
in fact , all linguistic studies in oral and written corpora are influenced by our segmentation decisions and our definition of word , intimately biased by an inherently anthropocentric perspective and , of course , by our linguistic tradition .
the idea that speech is produced like writing , as a linear sequence of vowels and consonants may indeed be a relic of our scripture technology . as a matter of fact
, it is well - known in linguistics that both vowels and consonants are produced linearly but also depend on their surrounding elements : this is the traditional and well - studied concept of coarticulation @xcite .
the boundaries between acoustical elements are therefore difficult to identify if we are not native speakers of a language , and that s yet a crucial problem of phonotactics and speech segmentation and recognition @xcite with differences across languages @xcite .
precisely because of this , classical signal segmentation based on the concept of `` word '' inherited from writing has lead some researchers to search how to transform artificially written corpora into phoneme or syllabic chains with different objectives @xcite , at the time it involves two major problems in communication studies , namely ( i ) the impossibility of performing fully objective comparative studies between human and non - human signals @xcite , where signals could be physical events , behaviours or structures to which receivers respond @xcite .
this problem leads researchers sometimes to manually segment acoustic signals guided only by their expertise , and prevents to explore signals of unknown origin including for instance the search for possible extraterrestrial intelligence @xcite . and
( ii ) a rather arbitrary definition of the units of study guided by ortographic conventions already produces non - negligible epistemological problems at the core of linguistics @xcite .
+ in this work we explore the acoustic analog of classical linguistic laws in the statistics of speech waveform from extensive real databases @xcite that for the first time extend all the way into the intraphoneme range ( @xmath0s ) @xcite . in order to do so in a systematic way , in what follows we will present a methodology that enables the direct linguistic analysis of any acoustical waveform without needs to invoke to _ ad hoc _ codes or assume any concrete formal communication system underneath ( materials and methods , see figure [ fig : figure1 ] for an illustration ) .
this method only makes use of physical magnitudes of the signal such as energy or time duration which therefore allow for a nonambiguous definition of our units .
speech is indeed a physical phenomenon and as such in this work we interpret it as a concatenation of events of energy release , and propose a mathematically well - defined way for segmenting this orchestrated suite of energy release avalanches @xcite .
we find clear evidence of robust zipf , heaps and brevity laws emerging in this context and speculate that this might be due to the fact that human voice seems to be operating close to criticality , hence finding an example of a biological system that , driven by evolution , has linked complexity and criticality .
we expect that this methodology can open a fresh line of research in communication systems where a direct exploration of underlying statistical patterns in acoustic signals is possible without needs to predetermine any of the aforementioned non - physical concepts , and hope that this will allow researchers to develop comparative studies between human language and other acoustical communication systems or even to unravel whether
if a generic signal shares these patterns @xcite . + * data . * for this work we have used a tv broadcast speech database named kalaka-2 @xcite . originally designed for language recognition evaluation purposes , it consists of wide - band tv broadcast speech recordings ( 4h per language sampled using 2 bytes at a rate of 16khz ) ranging six different languages : basque , catalan , galician , spanish , portuguese and english and encompassing both planned and spontaneous speech throughout diverse environmental conditions , such as studio or outside journalist reports but excluding telephonic channels . + * the method . *
the objects under study , speech waveforms or otherwise any generic acoustic signal , are fully described by an amplitude time series @xmath1 ( see figure [ fig : figure1 ] for an illustration of the method ) . in order to unambiguously extract a sequence of symbols
-the equivalent to words and phrases- from such signal without the need to perform _ ad hoc _ segmentation , we start by considering the semi - definite positive magnitude @xmath2 which , dropping irrelevant constants , has physical units of energy per time ( si for additional details on speech waveform statistics ) . by defining an energy threshold @xmath3 in this signal @xcite we will unambiguously separate voice events ( above threshold ) from silence events ( below threshold ) . more concretely , @xmath3 is defined as a relative percentage and its actual value in energy units depends on the signal variability range : for example @xmath4 means that @xmath5 of the data falls under this energy level .
it has been shown that @xmath3 decimates the signal similarly to a real - space renormalization group ( rg ) transformation @xcite , in such a way that increasing @xmath3 induces a flow in rg space .
systems operating close to a critical state lie in an unstable fixed point of this rg flow and its associated signal statistics are therefore shown to be threshold - invariant .
now , @xmath3 not only works as an energy threshold that filters out background or environmental noise ( noise filtering being a key aspect that species have learned to perform @xcite ) but , as previously stated , enable us to unambiguously define what we call a _ token _ or voice event , that is , a sequence of consecutive measurements @xmath6 , from a silence event of duration @xmath7 .
each token is in turn characterized by a duple @xmath8 where @xmath9 is the duration of the event and @xmath10 corresponds to the total energy released during that event @xmath11 obtained summing up the instantaneous energy over the duration of the event .
accordingly , the signal is effectively transformed to an ordered sequence of tokens @xmath12 , each of these being separated by silence events of highly heterogeneous durations @xmath7 which , incidentally , are known to be power law distributed @xcite .
finally , by logarithmically binning the scale of integrated energies we can assign an energy label ( the bin ) to each token , hence mapping the initial acoustic signal into a symbolic sequence of fundamental units which we call _ types_. the logarithmic binning is justified here invoking fechner - weber s law @xcite .
note that two tokens whose integrated energy fall in the same energy range are mapped to the same type even if their duration can be different , so in principle several tokens could map into the same type ( see si for a table of type / token ratios ) . by default we define as many bins as voice events
such that the set of bins can be understood as an abstraction of a universal language vocabulary , accordingly some bins might be empty and in general each bin will occur with uneven frequencies .
as such , types can be understood as acoustically - based universal abstractions of a fundamental unit , an abstract version of words or phonemes that appear intertwined in a signal with characteristic patterns .
+ to summarize , with this methodology we are able to map an arbitrary acoustic signal into a sequence of types separated by silence events ( figure 1 ) .
standard linguistic laws can then be directly explored in acoustic signals without needs to have an _ a priori _ knowledge neither of the signal code nor of the adequate segmentation process or the particular syntax of the language underlying the signal .
this protocol is thus independent of the communication system and can be used to make unbiased comparisons across different systems and signals .
needless to say , results could in principle depend on the particular value of @xmath3 , as this scans the signals at different energy thresholds .
however human voice has been recently shown to be invariant under changes in @xmath3 -an evidence of self - organized criticality ( soc ) @xcite in this system- and , accordingly , parameter - free laws can be extracted using a proper collapse theory as it will be shown in the results section . finally , in order to guarantee that the emergence of linguistic laws is only due to the structure and correlations of the signal and not due to the process of symbolization we will compare the results obtained from speech signals to properly defined null models which randomize the signal @xmath13 .
these null models thus maintain the marginal instantaneous energy distribution and remove any other correlation structure , yielding non - gaussian white noise with a fat - tailed marginal distribution . |
let @xmath0 be a compact riemannian manifold with boundary .
we consider two elliptic self - adjoint operators defined on @xmath1 .
the first is the usual laplace - beltrami operator @xmath2 on @xmath3 with respect to the induced metric @xmath4 .
since @xmath1 is compact , the spectrum of @xmath5 is discrete and consists of eigenvalues which we denote by @xmath6 where eigenvalues are counted with multiplicities .
note that @xmath1 is not necessarily connected , so eigenvalue @xmath7 might not be simple .
that is the reason why we start our numeration with @xmath8 as opposed to @xmath9 .
the second operator is dirichlet - to - neumann operator @xmath10 , which is defined in the following way . for any @xmath11
, there is a unique harmonic extension @xmath12 , i.e. there is a unique @xmath13 such that @xmath14 and @xmath15 .
then one defines @xmath16 , where @xmath17 is an outward unit normal vector to @xmath1 .
similarly to the laplacian the operator @xmath10 is elliptic self - adjoint with discrete spectrum .
we denote the corresponding sequence of eigenvalues by @xmath18 where eigenvalues are counted with multiplicities .
note again that the numeration starts with @xmath19 as opposed to @xmath20 .
the numbers @xmath21 and the sequence @xmath22 are sometimes refered to as steklov eigenvalues of @xmath23 and steklov spectrum of @xmath23 respectively .
the study of steklov spectrum has become rather popular in the recent years , see e.g. @xcite and references therein .
the aim of this paper is to show that under certain curvature restrictions on @xmath23 and @xmath1 the sequence @xmath22 is majorated by the sequence @xmath24 .
the precise statement is the following .
[ maintheorem ] let @xmath0 be a riemannian manifold of dimension @xmath25 with boundary .
suppose that the weitzenbck curvature on @xmath26-forms @xmath27}$ ] is nonnegative and the lowest @xmath28-curvature @xmath29 of @xmath1 is bounded from below by a positive constant @xmath30 .
then , if @xmath31 , for all @xmath32 one has the inequality @xmath33 moreover , if @xmath34 then for all @xmath35 @xmath36 in the following sections we define the weitzenbck curvature on @xmath37-forms and the lowest @xmath37-curvatures of the boundary for all @xmath38 .
for now , let us mention that convexity of @xmath1 is enough for boundedness of @xmath39 for all @xmath37 , whereas nonnegativity of the curvature operator suffices for nonnegativity of @xmath40}$ ] for all @xmath37
. bound ( [ maininequality ] ) is sharp for @xmath41 on a euclidean ball @xmath42 of radius @xmath43 . since both @xmath44 and @xmath45 on the sphere @xmath46
have multiplicity @xmath17 , we have that inequality ( [ maininequality ] ) is sharp for @xmath47 . moreover , as it follows from the remark after theorem [ thrs ] , the equality in inequality ( [ maininequality ] ) can only be achieved on a ball .
hence , for @xmath48 inequality ( [ maininequality ] ) is strict . for a fixed manifold @xmath23 bound of the type ( [ maininequality ] ) , i.e. inequality bounding @xmath49 in terms of a linear function of @xmath50 ,
can not possibly be sharp for all @xmath51 .
indeed , according to weyl s law ( see e.g. @xcite ) @xmath52 , whereas @xmath53 .
a curious feature of the proof is that it uses the extension of the steklov operator to the differential forms of higher degree defined by raulot and savo @xcite , whereas the final statement does not refer to differential forms . in case @xmath23
is a surface we have the following theorem . [ th2 ] let @xmath0 be a connected oriented riemannian surface .
then one has @xmath54 where @xmath55 is the length of @xmath1 , @xmath51 is the number of boundary components and @xmath56 is the genus of @xmath23 . in particular , when @xmath57 one obtains , @xmath58 theorem [ th2 ] is a consequence of theorem [ thyy ] below . as it is explained in @xcite theorem [ thyy ] can be seen as a generalization of hersh - schiffer - payne inequality @xcite , which is inequality ( [ steklov2 ] ) for simply connected planar domains , i. e. @xmath59 , @xmath60 .
theorem [ th2 ] makes the connection more evident . a particular case @xmath59 of inequality ( [ steklov2 ] )
have already been pointed out in @xcite .
a similar inequality was obtained by girouard and polterovich in @xcite @xmath61 however , one can easily see that inequality ( [ steklov2 ] ) yields a better upper bound unless either @xmath60 , @xmath59 or @xmath62 , when both inequalities yield the same bound .
let us also mention a remarkable series of papers by fraser and schoen @xcite , where the authors investigate the optimal upper bound of the form ( [ steklov ] ) for @xmath63 .
note that in the mentioned papers the enumeration of steklov eigenvalues starts with 0 , which is compensated by the fact that we have @xmath64 and @xmath65 in the left hand side of ( [ steklov2 ] ) .
bound ( [ steklov ] ) has an advantage of being linear in all the parameters involved .
in fact , hassanezhad @xcite proved that there exists a bound of the form @xmath66 with implicit universal constants @xmath67 and @xmath68 , which are independent of @xmath51 .
thus , by introducing linear dependence on @xmath51 we are able to make the constants in the above bound explicit .
let us also mention that for @xmath63 , kokarev proved in @xcite an upper bound @xmath69.\ ] ] the paper is organized as follows .
section [ background ] contains necessary background from differential geometry . in section [ dnoperator ]
we describe the extension of the operator @xmath10 to differential forms due to raulot and savo .
theorem [ th2 ] is proved in section [ digresion ] .
we prove theorem [ maintheorem ] for orientable manifolds in section [ proof ] and in the subsequent section we show how to generalise this proof to nonorientable manifolds .
finally in the last section we compare our results to a similar result due to wang and xia @xcite .
let @xmath70 denote the levi - civita connection on @xmath23 associated to the metric @xmath71 .
the operator @xmath70 and the metric @xmath71 have natural extensions to the bundle @xmath72 of differential @xmath37-forms on @xmath23 for all @xmath38 , which will be denoted by the same letters .
let @xmath73 be the formal adjoint of @xmath70 .
then the weitzenbck curvature @xmath40}$ ] on @xmath37-forms is defined by the bochner formula @xmath74 } \omega,\ ] ] where @xmath75 and @xmath5 is the hodge laplacian on @xmath76 .
according to @xcite the condition @xmath27}\geqslant 0 $ ] is equivalent to requiring non - negativity of the following expressions : * _ second ricci curvature : _
@xmath77 for any orthonormal triplet @xmath78 . * _ isotropic curvature : _
@xmath79 for any orthonormal quadruplet @xmath80
. note that by summing inequality ( [ 2ndrc ] ) over @xmath81 from some orthonormal basis of @xmath82 in the tangent space of @xmath23 , one obtains @xmath83 , i. e. @xmath27}\geqslant 0 $ ] implies @xmath84 .
note that the non - negativity of the curvature operator implies non - negativity of @xmath40}$ ] for all @xmath37 .
the principal curvatures @xmath85 of @xmath86 are defined as eigenvalues of the second fundamental form of @xmath1 .
then the lowest @xmath37-curvature @xmath39 of @xmath1 is defined as @xmath87 , where @xmath88 range over all @xmath37-element subsets of @xmath89 . this way
@xmath90 is equivalent to the convexity of @xmath1 and @xmath91 is proportional to the mean curvature of @xmath1 in @xmath23 .
there are several definitions of dirichlet - to - neumann operator on forms , see e.g. @xcite . here
we discuss the one due to raulot and savo .
for @xmath92 there exists a unique differential form @xmath93 , see e.g. @xcite , such that @xmath94 then the dirichlet - to - neumann operator @xmath95 on @xmath96 is defined as @xmath97 . as it was pointed out in @xcite , @xmath98 , @xmath37-th betti number of @xmath23 .
raulot and savo proved that @xmath95 is a positive elliptic self - adjoint operator and therefore its spectrum consists of a sequence of eigenvalues @xmath99 where eigenvalues are written with multiplicities . since @xmath100
, we sometimes interchange notations @xmath101 and @xmath21 for the @xmath102-th eigenvalue of the dirichlet - to - neumann operator on @xmath3 . as in the case of @xmath10 ,
the numbers @xmath103 are sometimes referred to as steklov eigenvalues of @xmath23 on differential @xmath37-forms .
this particular definition of @xmath95 is of interest to us due to the following two theorems .
[ thyy ] let @xmath0 be a compact oriented @xmath17-dimensional riemannian manifold with nonempty boundary .
let @xmath104 be the @xmath51-th steklov eigenvalue on differential @xmath105-forms of @xmath23 and @xmath50 be the @xmath51-th eigenvalue for the laplacian operator on @xmath1 .
then for any two positive integers @xmath37 and @xmath106 , one has @xmath107 where @xmath108 is the @xmath51-th betti number of @xmath23 .
[ thrs ] let @xmath0 be a compact @xmath17-dimensional riemannian manifold with nonempty boundary .
let @xmath109 .
assume that @xmath23 satisfies @xmath40}\geqslant 0 $ ] and that @xmath110 .
then if @xmath111 one has the inequality @xmath112 if @xmath113 one has @xmath114 in the same paper @xcite it is also proved that equality in ( [ thrseq2 ] ) implies that @xmath23 is a euclidean ball of radius @xmath43 .
for a connected oriented surface @xmath115 inequality ( [ thyyeq ] ) reads @xmath116 where @xmath51 is the number of boundary components of @xmath23 and @xmath56 is the genus of @xmath23 . in this section
we would like to obtain an upper bound for @xmath49 independent of the spectrum of the boundary .
for simplicity of exposition let us only prove inequality ( [ steklov ] ) .
as one could see , the proof presented below can be easily modified to cover inequality ( [ steklov2 ] ) .
assume that @xmath117 , then inequality ( [ corollary ] ) takes the following form @xmath118 let us introduce some notation .
for each vector @xmath119 we introduce the sequence @xmath120 in the following way . for each @xmath121 consider an arithmetic progression @xmath122 .
then @xmath120 is a nondecreasing sequence obtained by taking the union of sequences @xmath123 and reordering the entries .
the sequence @xmath120 in the context of steklov problem first appeared in the paper @xcite .
it is easy to see that the right hand side of ( [ corollary2 ] ) is @xmath124 ^ 2 $ ] , where @xmath125 is the length of the @xmath102-th boundary component of @xmath23 and @xmath126 $ ] denotes the @xmath17-th entry of a sequence @xmath67 . denoting by @xmath127 the @xmath128 norm of @xmath129 , i. e. @xmath130 , inequality ( [ corollary2 ] ) implies @xmath131)^2.\ ] ] thus , in order to obtain a universal upper bound on @xmath132 it is enough to evaluate the right hand side of inequality ( [ corollary3 ] ) .
[ algebra ] one has @xmath133 = d+k-1.\ ] ] first , let us note that the supremum is achieved .
indeed , for any positive constant @xmath134 , one has @xmath135 , therefore the function @xmath136 $ ] is scale - invariant for any @xmath137 .
since it is also continuous as a function of @xmath129 , the supremum is achieved .
hence , we can fix @xmath138 one of maximizers and let us denote by @xmath23 the corresponding maximal value .
* claim 1 . * for each @xmath121 there is @xmath139 such that @xmath140 .
indeed , let @xmath139 be the smallest integer , such that @xmath141 .
assume that @xmath142 is strictly greater than @xmath143 .
we will show that in that case @xmath23 is not a maximum .
indeed , by definition of @xmath139 it is easy to see that @xmath144 = s_{(l_1,\ldots,\frac{d_i|l|}{m},\ldots , l_k)}[d].\ ] ] at the same time , @xmath145 , therefore the replacement @xmath146 increases the value of the function @xmath147 $ ] .
we arrive at a contradiction .
* claim 2 . *
@xmath148 > s_{(l_1,\ldots , l_k)}[d-1]$ ] .
the proof of this claim is similar to the previous one . indeed , by previous claim @xmath149 .
then , arguing by contradiction , one has @xmath150 = s_l[d-1]= s_{(l'_1,\ldots , l_k)}[d]$ ] , where @xmath151 . once again @xmath152 and we arrive at a contradiction . according to the first claim , @xmath153 .
hence , for each @xmath102 the sequence @xmath123 has exactly @xmath154 entries less than @xmath155 .
therefore , the sequence @xmath120 contains @xmath156 entries less than @xmath143 .
since @xmath157 = \frac{m}{|l|}$ ] , claim 2 implies @xmath158 at the same time , @xmath159 , which implies @xmath160 . substituting this into equality ( [ sum ] ) one obtains the statement of proposition .
note that the above proposition is purely algebraic .
nevertheless , as a corollary we obtain theorem [ th2 ] .
the proof is obtained by an easy combination of theorem [ thyy ] and theorem [ thrs ] .
first of all , let us make a couple of observations .
* observation 1 .
* formula ( [ bochner ] ) implies that @xmath161 } * = * w^{[p]}$ ] .
therefore nonnegativity of @xmath40}$ ] is equivalent to nonnegativity of @xmath161}$ ] . * observation 2 .
* application of theorem [ thrs ] for @xmath162 now yields that under the conditions of theorem [ maintheorem ] one has the inequality @xmath163 if @xmath31 and @xmath164 if @xmath34 . in particular , that yields @xmath165 .
since @xmath27}\geqslant 0 $ ] implies @xmath166}\geqslant 0 $ ] and @xmath167 implies @xmath168 , we can apply theorem [ thrs ] for @xmath169 and conclude that @xmath170 .
taking into account @xmath171 , the application of theorem [ thyy ] for @xmath172 yields the inequality @xmath173 combining this inequality with inequlities from observation 2 , one completes the proof of theorem [ maintheorem ] for orientable manifolds .
it is possible to generalize theorem [ thyy ] to include the case of non - orientable manifolds in a way that allows us to apply the arguments of the previous section .
below we give the statement and the short outline of the argument .
let us pass to an orientable double cover @xmath174 and endow it with the metric @xmath175 such that the involution @xmath176 exchanging the leaves of @xmath177 is an isometry .
then @xmath176 induces the decomposition of differential forms on @xmath178 into even and odd with respect to @xmath176 which is compatible with hodge - morrey decomposition ( for details on hodge - morrey decomposition see @xcite ) .
similarly , the eigenvalues of dirichlet - to - neumann and laplace operators are divided into those corresponding to odd and even eigenforms respectively .
if we denote by @xmath179 the @xmath102-th even eigenvalue of @xmath5 on @xmath180 and similarly by @xmath181 the @xmath182-th odd steklov eigenvalue on @xmath183 , then inequality ( [ thyyeq ] ) for @xmath178 becomes @xmath184 in order to prove ( [ nonorientable ] ) one follows the proof of theorem [ thyy ] accounting for the presence of @xmath176 .
the key points are the following : * differential @xmath137 is compatible with the decomposition into odd and even forms ; * hodge star operator sends odd forms to even and vice versa ; * since the kernel of @xmath95 consists of neumann harmonic @xmath37-fields and even neumann harmonic fields on @xmath178 are the pullbacks of neumann harmonic fields on @xmath23 , then @xmath185 , @xmath186 . to complete the proof of theorem [ maintheorem ] one makes the following two observations .
first , local curvature conditions on @xmath23 pass to @xmath178 , therefore betti numbers disappear from the inequality ( [ nonorientable ] ) .
second , there is an obvious inequality @xmath187 . since even eigenvalues of @xmath178 coincide with the eigenvalues of @xmath23 , the same arguments as in section [ proof ] conclude the proof
in this secton we compare results of theorem [ maintheorem ] with the following statement .
[ thwx ] let @xmath0 be an @xmath17-dimensional compact connected riemannian manifold with non - negative ricci curvature and boundary @xmath1 .
assume that the principal curvatures of @xmath1 are bounded from below by a positive constant @xmath30 .
then one has the inequality @xmath188 we start our comparison with the case @xmath34 , where conditions of our theorem coincide with conditions of theorem [ thwx ] ( since @xmath189 } = * w^{[2 ] } * = ric$ ] ) . theorem [ maintheorem ] yields for @xmath63 and @xmath34 @xmath190 while inequality ( [ thwxeq ] ) for @xmath34 becomes @xmath191 let @xmath193 .
then @xmath194 and @xmath23 is convex , i.e. it satisfies the conditions of the theorem .
if @xmath195 the lower bound for principal curvatures is @xmath196 while the lower bound @xmath197 for gaussian curvature of @xmath1 is @xmath198 . using the classical bound @xmath199 ( see e. g. p. 186 of @xcite ) one sees that the bound given by theorem [ maintheorem ] is better once @xmath200 , i.e. once the ellipsoid is oblong enough . in case @xmath31
our condition @xmath27}\geqslant 0 $ ] implies condition @xmath201 of theorem [ thwx ] . at the same time , the assumption `` principle curvatures are bounded from below by @xmath202 '' of theorem [ thwx ] implies the the lowest @xmath28-curvature is bounded from below by @xmath203 .
thus , while our condition on the interior curvature is stronger , the condition on the curvature of @xmath1 is weaker . now , suppose that @xmath31 and @xmath0 satisfies the conditions of both theorem [ maintheorem ] and theorem [ thwx ] , i.e. @xmath27}\geqslant 0 $ ] and all principle curvatures of @xmath1 are bounded from below by @xmath202 , then the lowest @xmath28-curvature is bounded from below by @xmath203 .
an application of theorem [ thwx ] yields inequality ( [ thwxeq ] ) , which can be rewritten as @xmath204 while theorem [ maintheorem ] for @xmath63 gives @xmath205 which is clearly stronger . the right hand sides of those inequalities differ by an expression , which is equal to zero iff @xmath206 .
according to the results of xia @xcite the latter happens only if @xmath23 is a euclidean ball of radius @xmath43 .
therefore , our estimate yields a better bound for any convex bounded domain of the euclidean space other than a ball .
moreover , our estimate can also be applied to nonconvex domains , since our condition on the curvature of @xmath1 is only @xmath167 .
the author is grateful to iosif polterovich for fruitful discussions and comments on the initial versions of the manuscript .
the author also thanks spiro karigiannis for providing the reference @xcite .
99 m. belishev , v. sharafutdinov , _ dirichlet to neumann operator on differential forms .
sci . math .
* 132*:2 ( 2008 ) , 128 - 145 . i. chavel , _ riemannian geometry a modern introduction , 2nd edition _ cambridge university press , new york , 2006 .
a. fraser , r. schoen , _ the first steklov eigenvalue , conformal geometry , and minimal surfaces .
, * 226*:5 ( 2011 ) , 4011 - 4030 .
a. fraser , r. schoen , _ sharp eigenvalue bounds and minimal surfaces in the ball .
_ preprint arxiv:1209.3789 .
a. girouard , l. parnovski , i. polterovich , d. sher .
, _ the steklov spectrum of surfaces : asymptotics and invariants
. soc . * 157*:3 ( 2014 ) , 379 - 389 .
preprint arxiv:1311.5533 .
a. girouard , i. polterovich , _ spectral geometry of the steklov problem .
_ preprint arxiv:1411.6567 .
a. girouard , i. polterovich , _ upper bounds for steklov eigenvalues on surfaces .
_ electron .
, * 19 * ( 2012 ) , 77 - 85 .
a. hassannezhad , _ conformal upper bounds for the eigenvalues of the laplacian and steklov problem .
_ journal of functional analysis , * 261*:12 ( 2011 ) , 3419 - 3436 .
j hersch , l. payne , m. schiffer , _ some inequalities for stekloff eigenvalues .
rational mech .
, * 57 * ( 1975 ) , 99 - 114 . m. s. joshi , w. r. b. lionheart , _ an inverse boundary value problem for harmonic differential forms .
_ asymptot .
* 41*:2 ( 2005 ) , 93 - 106 .
g. kokarev , _ variational aspects of laplace eigenvalues on riemannian surfaces .
_ adv . math .
* 258 * ( 2014 ) , 191 - 239 .
preprint arxiv:1103.2448 .
p. petersen , _ demystifying the weitzenbck curvature operator_. preprint is available on the homepage of the author at http://www.math.ucla.edu/~petersen/. s. raulot , a. savo , _ on the first eigenvalue of the dirichlet - to - neumann operator on forms . _ j. of functional analysis 262 ( 2012 ) , 889 - 914 .
g. schwarz , _ hodge decomposition
a method for solving boundary value problems . _ lecture notes in math . , springer , 1995 .
v. sharafutdinov , c. shonkwiler , _ the complete dirichlet - to - neumann map for differential forms . _ j. geom
. anal . * 23*:4 ( 2013 ) ,
2063 - 2080 . q. wang , c. xia , _
sharp bounds for the first non - zero stekloff eigenvalues . _ j. of functional analysis 257 ( 2009 ) , 2635 - 2644 .
c. xia , _
rigidity for compact manifolds with boundary and non - negative ricci curvature .
soc . 125 ( 1997 ) , 1801 - 1806 .
l. yang , c. yu , _ a higher dimensional generalization of hersch - payne - schiffer inequality for steklov eigenvalues_. preprint arxiv:1508.06026 . | consider a compact riemannian manifold with boundary . in this short note
we prove that under certain positive curvature assumptions on the manifold and its boundary the steklov eigenvalues of the manifold are controlled by the laplace eigenvalues of the boundary .
additionally , in two dimension we obtain an upper bound for these steklov eigenvalues in terms of topology of the surface without any curvature restrictions . |
cusp forms are the building blocks of the modern theory of automorphic forms .
they are elusive objects and even the existence of the everywhere unramified forms on @xmath2 @xmath3 that we study in this paper , is known only by indirect means ( selberg 1956 using the trace formula ) . the first reliable numerical computations of such forms was carried out by hejhal and rackner @xcite .
they computed some high frequency maass cusp forms ( both even and odd ) for @xmath4 .
let @xmath5 be the modular surface .
it has an orientation reversing isometry @xmath6 induced from the reflection of @xmath7 , given by @xmath8 . the even maass cusp forms ( which will be the central objects in this paper ) are functions @xmath9 satisfying * @xmath10 , @xmath11 , * @xmath12 , * @xmath13 , * @xmath14 , normalized so that @xmath15 , where @xmath16 is the hyperbolic area measure . such a @xmath17 is automatically a cusp form @xcite and we will assume that it is also an eigenfunction of all the hecke operators @xmath18 , @xmath19 .
the corresponding automorphic cusp forms on @xmath20 are the everywhere unramified forms and their @xmath21-functions @xmath22 are the @xmath23 analogues of the riemann zeta function .
we will use the parameter @xmath24 to parametrize the eigenvalue by @xmath25 and we let @xmath26 be the numbering of such @xmath17 s when ordered by @xmath27 .
selberg @xcite proved a weyl law for these @xmath17 s ( thus demonstrating their existence ) ; @xmath28 let @xmath29 denote the nodal line of @xmath17 , that is the set @xmath30 .
it consists of a finite union of real analytic curves .
the connected components of @xmath31 are the nodal domains of @xmath17 and we denote their number by @xmath32 . more generally , for any subset @xmath33 we denote by @xmath34 the number of such domains that meet @xmath35 . from @xcite
we have in figure 1 , a picture of the nodal lines for four such @xmath17 s with @xmath36 and @xmath37 ( note that @xmath24 is denoted by @xmath38 in @xcite ) .
this paper is concerned with answering some questions that have been raised in connection with such pictures .
the most basic is to determine the number of nodal domains , either in totality or in subsets of a fundamental domain .
another question , which appears particularly hard to address analytically , is whether @xmath39 is nonsingular ( and hence is self - avoiding ) . using heuristic arguments relating the zero sets of random waves to an exactly solvable bond percolation model , as well as further heuristics relating eigenfunctions of quantizations of chaotic hamiltonians to random waves , bogomolny and schmit @xcite
suggest that in such settings @xmath32 has an asymptotic law ! in our setting of even maass cusp forms for @xmath40 , it reads + * bogomolny - schmit conjecture * @xmath41 there is also a localized version for @xmath34 for @xmath35 a subset of @xmath40 like the windows in figure 1 and the prediction is remarkably accurate for these four forms .
based on this and other experimentation in different settings ( @xcite , @xcite,@xcite,@xcite ) it seems that this prediction is probably correct , with either the above constant or one very close to it . according to courant s general nodal domain theorem @xcite and , one has asymptotically @xmath42
however , even showing that @xmath32 ( or more generally @xmath34 with @xmath35 a compact piece of @xmath40 ) goes to infinity with @xmath26 is problematic . at issue
is the fact that @xmath32 is a global topological invariant and hence is difficult to study from local considerations . in particular , it is not true for a general riemannian surface that @xmath32 must increase with @xmath26 , so that the problem is truly a global one ( see @xcite , @xcite and @xcite where it is shown that there are infinite sequences of eigenfunctions each with few nodal domains ) .
our main result is a conditional proof ( assuming something much weaker then a generalized riemann hypothesis ) that @xmath43 goes to infinity with @xmath26 .
we turn to a precise formulation of our results and methods .
the isometry @xmath6 of @xmath40 maps the nodal domains of @xmath17 bijectively to themselves . in this way
, the nodal domains are partitioned into those fixed by @xmath6 , which we call _ inert _ , and those paired off by @xmath6 which we call _ split _ ( see figure 2 from @xcite , where the black and red ones are inert and the green ones are split ) . correspondingly we write @xmath44 where @xmath45 is the number of inert domains and @xmath46 , which is even , is the number of split domains .
the nodal domains for @xmath17 that we will produce are all inert . according to the estimates given below ,
if the bogomolny - schmit conjecture is true , then most of the nodal domains for @xmath17 are split ( when @xmath47 ) ; however we have no proof that there are any split nodal domains !
our analysis is based on intersecting @xmath39 with various analytic arcs in @xmath40 .
if @xmath48 is such a simple arc then clearly @xmath49 denote by @xmath50 the number of sign changes of @xmath17 when traversing @xmath48 .
a key role is played by the arc @xmath51 which is fixed by @xmath6 .
this arc is composed of three piecewise analytic geodesics , denoted by @xmath52 , @xmath53 and @xmath54 in figure 3 . a simple topological analysis ( see section 2.1 ) shows that @xmath55 and more generally for @xmath48 a nonempty segment of @xmath56 @xmath57 the complexification technique of toth and zelditch ( using a form of jensen s lemma ) may be applied to our @xmath17 s whose normal derivative vanishes along @xmath56 , to yield latexmath:[\[\label{eq : oh8 } inert nodal domains is much smaller than what is predicted by the bogomolny - schmit conjecture . to give a lower bound for @xmath45 we will use and and
hence we seek sign changes in @xmath59 along a given curve . to this end
, we examine the behavior of @xmath59 when restricted to certain curves .
these restriction questions were first examined in @xcite and @xcite , and sharp versions are discussed in @xcite .
let @xmath60 be a nonempty simple analytic segment .
we seek sharp upper and lower bounds for the @xmath61-restrictions of @xmath59 .
it is plausible that for any such fixed @xmath48 and large @xmath24 ( recall that @xmath17 will always be @xmath61 normalized over @xmath40 ) @xmath62 where throughout the paper @xmath63 denotes the hyperbolic arclength measure @xmath64 for @xmath65 .
clearly , this implies that @xmath48 is not part of the nodal line @xmath39 for @xmath24 large enough , which itself is a very strong statement .
if @xmath48 is a geodesic or horocycle segment , then one can show without difficulty that @xmath48 is not a part of @xmath39 ( see section 2.2 ) .
if @xmath48 is a closed horocycle or a long enough ( compact ) subsegment of @xmath56 , we establish the sharp bounds in .
one of the key ingredients in the proof of the lower bound is the quantum unique ergodicity ( que ) theorem of lindenstrauss @xcite and soundararajan @xcite .
it asserts that the restriction of @xmath66 to a nice @xmath67 ( say a geodesic ball or more generally micro local lifts to @xmath68 ) satisfies @xmath69 [ theoremi1 ] let @xmath70 be a closed horocycle in @xmath40 .
then , for @xmath71 @xmath72 using probabilistic arguments , hejhal and rackner ( @xcite ( 6.12 ) ) conjecture that the restricted mean square of @xmath17 to a closed horocycle is asymptotic to @xmath73 .
this follows from the conjectured equation in appendix a. [ theoremi2 ] if @xmath48 is a long enough but fixed compact subsegment in @xmath52 or @xmath53 , then for @xmath74 @xmath75 in section 6 ( see theorem [ theorem5.1 ] ) we first prove the lower bound in theorem [ theoremi2 ] in non - localised form , namely the inequality between the first and third terms in this theorem . for this ,
little of the arithmetic features of @xmath40 ( other than que ) are used .
the localized version is technically much more involved , requiring a delicate asymptotic analysis of the off - diagonal terms , with both arithmetic and analytic input being crucial .
it is in this analysis that we are forced to make @xmath48 long ( see section 6.2 and appendix a ) .
the sharp lower bounds in the restriction theorems above can be combined with strong upper bounds for @xmath76 for @xmath77 an arbitrary segment of @xmath70 or @xmath48 , and subconvexity bounds for @xmath78 on @xmath40 ( see @xcite ) to give lower bounds for @xmath79 and @xmath80 ( see the end of section 5 and section 6.3 ) .
[ theoremi4 ] let @xmath81 be a fixed closed horocycle in @xmath40 .
then for @xmath71 @xmath82 the upper bound above can be established without recourse to arithmetic , that is it holds for any hyperbolic surface @xmath83 and any closed horocycle on @xmath83 , as was shown recently by jung @xcite .
the lower bound can be improved to @xmath84 if one were to assume that the @xmath85-norm of @xmath17 restricted to compact sets is bounded by @xmath86 .
the random wave model for the @xmath17 s predicts that @xmath87 as @xmath88 ( see appendix b ) .
our primary interest is in sign changes of @xmath17 along @xmath56 .
the lower bound in the next theorem is conditional on the lindelof hypothesis for the @xmath21-functions @xmath22 associated to the form @xmath17 .
this conjecture asserts that @xmath89 and its truth follows from the corresponding riemann hypothesis . for our purpose , what we need is a bound of the form @xmath90 for some finite @xmath91 but with @xmath92 .
the best subconvex bounds for @xmath22 known give @xmath93 ( see @xcite,@xcite ) which is far short of @xmath94 .
for this reason we simply assume the full lindelof hypothesis .
[ theoremi5 ] fix @xmath95 a sufficiently long compact geodesic segment on @xmath52 or @xmath53 and assume the lindelof hypothesis for the @xmath21-functions @xmath22
. then @xmath96 as noted in , general complexification techniques @xcite yield @xmath97 .
theorem [ theoremi5 ] together with leads to our main result .
[ theoremi6 ] with the same assumptions as in theorem [ theoremi5 ] @xmath98 and in particular that @xmath99 goes to infinity as @xmath26 goes to infinity .
finally , we examine the behavior of @xmath17 in the shrinking regions near the cusp of @xmath40 . for @xmath100 in the standard fundamental domain for @xmath101 , and when @xmath102 , @xmath17 is approximated very well by one term in its fourier expansion . from this
, one can execute a much finer analysis of @xmath39 even into a slightly larger region away from the cusp .
it appears that all the nodal domains in these regions are inert domains , see section 4 ( this being the analogue of the suggestion in @xcite that all the zeros of a holomorphic hecke eigenform of weight @xmath103 in the region of the standard fundamental domain with @xmath104 are real , that is they lie on the boundary @xmath56 ) . in any case , we show that there are at least @xmath105 inert domains in the region @xmath106 . in particular , @xmath107 other finer aspects of the geometry and topology of @xmath39 near the cusp ( here @xmath108 with @xmath109 any positive constant ) are discussed in section 4 .
our results concerning these even maass cusp forms can be extended to the eisenstein series @xmath110 and to odd cusp forms on @xmath40 .
the eisenstein series are even with respect to the symmetry @xmath6 and other than renormalizing the @xmath61-norms by restricting to a compact subset of @xmath40 and invoking ( @xcite , @xcite ) the proofs go through in the same way .
the odd forms vanish identically on @xmath56 and the analysis of the nodal domains proceeds by working with the normal derivative of @xmath17 restricted to @xmath56 .
the analogue @xmath0-restriction lower bounds and @xmath85 upper bounds can be established with small modifications as indicated in appendix c. in particular , one concludes that theorem [ theoremi6 ] is valid for the odd cusp forms as well .
one can also extend these results on nodal domains to congruence subgroups as long as they carry reflection symmetries .
for example if @xmath111 is a surface of genus @xmath112 , where @xmath113 is the usual ` hecke ' congruence group , then our reflection @xmath114 induces an isometry @xmath115 of @xmath83 .
the fixed set of @xmath115 consists of a finite union of geodesics running between cusps .
we can diagonalize the space of newforms ( the oldforms being taken care of at a lower level ) with respect to @xmath115 , yielding @xmath115-even and @xmath115-odd forms @xmath17 .
correspondingly , @xmath17 has @xmath115 inert and split nodal domains . the main result , theorem [ theoremi6 ] , extends to this setting , namely if @xmath83 is fixed ( as is @xmath112 ) and @xmath17 is as above , then the lower bound in theorem [ theoremi6 ] for the number of inert nodal domains for @xmath17 holds as @xmath88 ( see the remark after theorem [ theoremnodal ] ) . in particular , the number of nodal domains for @xmath17 goes to infinity with @xmath24 . the techniques used to prove our main results rely heavily on the fourier expansion of @xmath17 and especially the arithmetic interpretation of its fourier coefficients , so that it is not clear if the results extend to compact arithmetic surfaces @xmath83 . in a follow - up paper @xcite
we show that indeed similar results hold for nodal domains of eigenfunctions on certain such @xmath83 s .
again , the key is that @xmath83 carry a reflection symmetry . to produce such arithmetic surfaces with a symmetry
we take @xmath83 to be @xmath116 , where @xmath101 corresponds to the unit group of an anisotropic rational quadratic form @xmath117 of signature @xmath118 . to ensure that @xmath83 has a reflective symmetry @xmath115 , one requires that there be an integral root vector of @xmath119 of length @xmath120 (
the sign is such that this gives a hyperbolic reflection ) . in this compact setting the proof of the critical theorem [ theoremi2 ] for @xmath115 even eigenfunctions as well as various of the other ingredients , are deduced using more representation theoretical tools .
we leave the precise statements and details to part ii .
0.2 in + ar and ps thank the bsf grant .
all three authors have been partially supported by grant nsfdms-0554345 . + ar thanks the isf center of excellency grant and the veblen fund at ias .
he was also supported by the erc grant 291612 .
+ ag thanks the institute for advanced study and princeton university for making possible visits during part of the years 2010 - 2012 when much of this work took place .
he also acknowledges support from the ellentuck fund at ias and the vaughn fund at osu .
+ the software mathematica@xmath121 was used on a dual - core pc to generate some of the images , using data from @xcite . 0.2 in
we give in this section a discussion related to equations to in the introduction and a proof that segments of horocycles and geodesics can not be contained in nodal lines .
let @xmath5 be the modular surface and @xmath122 be the isometry of @xmath40 ( as a hyperbolic surface ) induced by the reflection of @xmath7 given by @xmath123 in what follows , @xmath124 will be an even hecke - maass cusp form , so that @xmath125 let @xmath56 be the set of points on @xmath40 fixed by @xmath6 so that @xmath126 as shown below ( here we take the standard fundamental domain @xmath127 ) .
the nodal line of @xmath17 is the set @xmath128 while the nodal domains of @xmath17 are the ( finitely many ) connected components of @xmath31 .
we denote them by @xmath129 where @xmath130 is their number . from , it is clear that if @xmath131 is a nodal domain ( of @xmath17 ) then so is @xmath132 .
moreover , since the @xmath133 s are disjoint , they fall into two types : * inert domains @xmath134 where @xmath135 for some @xmath136 and @xmath137 ; * split domains @xmath138 , @xmath139 where @xmath140 for some @xmath136 + and @xmath141 . thus the list of @xmath131 s is @xmath142 so that @xmath143 geometrically , we can recognize the two types of nodal domains according as @xmath144 is empty or not .
if @xmath144 is nonempty , then it is an open nonempty subset of @xmath56 , so that the invariance of @xmath56 under @xmath6 ensures that @xmath145 is nonempty and so @xmath146 . if on the other hand @xmath144 is empty , then being open and connected , @xmath133 lies entirely in @xmath147 or @xmath148 as indicated in figure 1 , so that @xmath133 does not meet @xmath149 since @xmath150 .
thus , inert domains are exactly the domains meeting @xmath56 and can be thought of as boundary nodal domains for @xmath148 .
they are the analogues of what are called real zeros for holomorphic forms in @xcite .
topologically , @xmath56 is a circle when compactified at the cusp @xmath151 .
the vanishing of @xmath17 at the cusp and the normalization of the first fourier coefficient ( see section 3 ) ensures that @xmath17 ( restricted to @xmath56 ) is positive in a neighborhood of @xmath151 .
let @xmath152 with @xmath153 , denote the zeros of @xmath17 arranged cyclically on @xmath56 . denote by @xmath154 the subset @xmath155 of the @xmath156 s at which @xmath17 changes sign .
clearly @xmath157 is even and @xmath158 ( since @xmath159 ) .
we call @xmath160 the number of zeros of @xmath17 on @xmath56 and @xmath26 the number of sign changes . each inert nodal domain @xmath161 meets @xmath162 and the intersection consists of a disjoint set of segments of the form @xmath163 .
moreover , different inert domains give rise to different disjoint sets of such segments .
it follows that the number of inert @xmath161 s is at most the number of intervals @xmath163 so that we have the upper bound @xmath164 . to obtain a lower bound for @xmath165
, we first assume that the nodal line @xmath39 is nonsingular ( that is @xmath17 and @xmath166 never vanish together ) so that @xmath167 . in this case
, @xmath39 has no self - crossings and consists of say @xmath168 simple closed curves @xmath169 , ... , @xmath170 (
each an embedded circle in @xmath171 .
if @xmath172 , then @xmath173 passes through @xmath174 for some @xmath175 and being an embedded circle must cross @xmath56 at some other point @xmath176 .
the resulting reflected curve ( by @xmath6 ) yields a closed curve in @xmath40 which is one of the components of @xmath39 and which we denote by @xmath177 .
these curves meet @xmath56 at exactly two distinct points and every point of @xmath154 appears once .
thus we get exactly @xmath178 simple closed curves in @xmath40 corresponding to the sign changes of @xmath17 on @xmath56 .
the remaining @xmath179 closed curves come in pairs and do not meet @xmath56 .
it follows from this , by induction on @xmath168 and @xmath180 , that there are exactly @xmath181 nodal domains whose boundary contains one of the @xmath177 s . in other words , under the nonsingular assumption , there are @xmath181 inert domains and @xmath182 split domains . in general , @xmath39 consists of piecewise analytic arcs with singularities formed by crossings ( cusps ) and self - intersections .
however these may be removed locally one at a time at the cost of possibly decreasing the number of inert nodal domains but not changing @xmath154 or at worst not increasing the quantity @xmath183 . in more detail , suppose that locally ( away from @xmath56 ) , @xmath39 looks like the figure on the left below with @xmath184 denoting the sign of @xmath17 near the singularity and @xmath185 the corresponding pieces of the complement . deforming the curves near the singularity ( as shown in the figure on the right above ) would possibly decrease the number of connected components by one and keeps everything away from the singularity the same .
if the singularity looks like that below , with @xmath186 and @xmath187 joined as shown , then we remove the arc joining them which may again decrease the number of domains without changing anything else . next , consider the figure on the left below where @xmath39 meets the boundary @xmath56 . here ,
@xmath186 and @xmath187 are sign changes of @xmath17 on @xmath56 . to get the figure on the right , we deform the arcs containing @xmath186 and @xmath187 into each other so that @xmath186 and @xmath187 are replaced by the point @xmath188 which is not a sign change . here
we reduce the value of @xmath26 by two and lose exactly one inert nodal domain , so that @xmath189 is unchanged and the singularity @xmath190 is removed .
last , we consider the case when one encounters a possible cusp singularity on the boundary as shown below . here , there are two possible deformations .
suppose that @xmath17 is positive on @xmath191 and @xmath192 , and negative on @xmath193 .
then , we may remove the singularity @xmath190 which is not a sign change , as shown in the middle figure . in this way
, the number of inert nodal domains may have possibly decreased while the number of sign changes remains unchanged .
the same conclusion holds if @xmath17 is positive on @xmath191 and @xmath193 , and negative on @xmath192 .
now we may remove the arc between @xmath191 and @xmath193 as shown on the right . in this way
, we remove all crossings and singularities so that when completed , @xmath39 will be a nonsingular curve and the quantity @xmath183 will not have increased .
we then have the following [ theoremnodal ] let @xmath165 denote the number of inert nodal domains of @xmath17 .
suppose @xmath160 denotes the number of zeros of @xmath17 on @xmath56 , and @xmath26 the number of sign changes .
then @xmath194 if @xmath39 is nonsingular , then there are exactly @xmath195 inert nodal domains of @xmath17 .
an analog of this lower bound also holds for other arithmetic surfaces as mentioned in the introduction .
however , one must take into account the genus @xmath112 of the surface .
then , the lower bound takes the shape @xmath196 so that for a fixed surface , @xmath197 .
we end this section by showing that the restriction of @xmath17 to a geodesic segment or a horocycle segment can not be identically zero .
that is to say that @xmath39 can not contain a geodesic segment or a horocycle segment .
we begin with a horocycle segment @xmath198 . if @xmath199 then by analytic continuation @xmath200 , where @xmath201 is the complete horocycle extending @xmath81 .
it is well known @xcite from the theory of the horocycle flow , that the projection of @xmath202 in @xmath40 is either a closed horocycle or it is dense in @xmath40 . in case of the latter , clearly @xmath203 in @xmath40 , which is a contradiction . on the other hand , if the projection of @xmath202 in @xmath40 is a closed horocycle @xmath204 , then theorem [ theoremi1 ] is applicable ( see section 5 and theorem 5.1 ) so that we have an essentially sharp , and in particular a non - zero , lower bound for the @xmath61-restriction of @xmath17 to @xmath204 , which contradicts that @xmath200 .
one can also use the ( exponentially small ) lower bound of @xcite which has the advantage of not making any arithmetic assumptions .
however , we give an simpler argument as follows : suppose @xmath204 is the set @xmath205 with a fixed @xmath206
. then for any integer @xmath207 , @xmath208 where @xmath209 are the hecke fourier coefficients and @xmath174 is the bessel - macdonald function ( see sections 3 and 5 for details and notation ) .
it is known that the bessel function above does not vanish if we choose @xmath210 . moreover
, the hecke relations ensures that @xmath211 for all @xmath180 chosen to be either a prime or its square .
thus the @xmath61-norm above does not vanish .
this completes the proof of the claim for horocyclic segments .
for geodesic segments @xmath48 , assume that @xmath212 .
first , if @xmath213 then along @xmath48 we have both @xmath212 and @xmath214 so that by the reflection principle ( or carleman estimates ) @xmath17 vanishes identically on @xmath40 .
hence we may assume that @xmath48 is disjoint from @xmath56 .
let @xmath215 be the lifted complete geodesic in @xmath7 , so that again from analytic continuation @xmath216 .
hence by the reflection principle , as an eigenfunction on @xmath7 , @xmath17 satisfies @xmath217 , where @xmath218 is the reflection of @xmath7 in @xmath215 .
moreover , we also have @xmath219 where @xmath220 is the reflection in @xmath221 , for @xmath222 and @xmath223 .
thus , @xmath224 is invariant under the subgroup @xmath225 of the isometries of @xmath7 generated by @xmath226 and @xmath227 .
the reflection group @xmath228 is known to be a maximal discrete subgroup of the isometry group of @xmath7 . since @xmath229 (
since we are assuming here that @xmath230 is not contained in @xmath56 ) , we conclude that @xmath225 is not a discrete subgroup of the isometry group of @xmath7 .
hence the same is true of the subgroup of words of even length in @xmath226 and @xmath227 , as a subgroup of @xmath231 .
in particular , the closure of this last group contains a 1-parameter connected subgroup of @xmath231 which we denote by @xmath232 .
so @xmath233 is invariant under the action of @xmath232 on @xmath7 . by connectedness , @xmath17 is also invariant under @xmath232 so that @xmath17 is an eigenfunction that lives on @xmath234 .
such an eigenfunction is a function of one variable ( satisfying an ode ) and unless it is identically zero can not be invariant under @xmath101 .
thus @xmath203 on @xmath7 , a contradiction , proving that @xmath212 is impossible .
if @xmath124 is a maass eigenform associated with the eigenvalue @xmath235 for the laplace operator on the modular surface @xmath236 , then recall that it has a fourier expansion of the type @xmath237 where @xmath238 , @xmath239 the macdonald - bessel function and @xmath240 . in
what follows , we will always assume that our eigenform is @xmath61-normalized , that it is a hecke - eigenform and that @xmath241 .
it is then known that ( see iwaniec @xcite and hoffstein - lockhart @xcite ) that for @xmath71 @xmath242 it will be convenient to use the notation @xmath243 one has @xmath244 where @xmath209 , with @xmath245 , are the hecke eigenvalues which by @xcite are known to satisfy @xmath246 with @xmath247 . and are expected to satisfy the ramanujan - selberg conjecture @xmath248 + we recall here the asymptotic behavior of @xmath249 for @xmath250 with large @xmath251 and all @xmath252 .
[ lemmaone ] let @xmath253 be a sufficiently large positive constant and @xmath251 . * if @xmath254 , then @xmath255 * if @xmath256 , then @xmath257 * if @xmath258 , then @xmath259 where @xmath260 is the airy function .
the implied constants are absolute . here
the non - negative function @xmath261 ( shown in fig.8 ) is given by @xmath262 for @xmath263 , it is a decreasing function , while for @xmath264 it is increasing . near @xmath265 , it has order of magnitude @xmath266 and near @xmath267 it has a logarithmic singularity .
the lemma follows readily from asymptotic expansions for the airy function , and from uniform asymptotic expansions for the k - bessel function by balogh @xcite , @xcite ( based on earlier work of olver ) .
[ corlone ] uniformly in @xmath252 and @xmath268 sufficiently large , one has @xmath269
define @xmath270 [ propone ] let @xmath271 . then for any @xmath272 with any fixed @xmath273
, we have * @xmath274 * for any @xmath275 @xmath276 * @xmath277 @xmath278 where the implied constants depend at most on @xmath279 and @xmath280 ( and not on @xmath281 , @xmath48 and @xmath282 ) .
this follows by decomposing the sum in into three sums with @xmath180 s in the range @xmath283 , @xmath284 and @xmath285 so that we may apply lemma [ lemmaone ] .
the analysis is straightforward using , , together with the bound @xmath286 as @xmath287 for the airy function .
the middle _ transitional range _ is responsible for the error terms indicated together with the error terms given by lemma [ lemmaone ] , while the infinite range gives a term with exponential decay .
we next provide a good approximation to @xmath288 for a sequence of large values of @xmath282 ( this being the analog of theorem 3.1 in @xcite , where a similar result was shown for holomorphic hecke cusp forms ) .
define , for integers @xmath289 @xmath290 [ proptwo ] for any small @xmath74 let @xmath291 with @xmath292 .
assume that @xmath17 is an even maass cusp form .
then , for any @xmath293 , writing @xmath294 we have @xmath295 moreover , if @xmath296 , then for any @xmath282 satisfying @xmath297 @xmath298 where the implied constants depend at most on @xmath280 .
the main terms above come from the _ transitional range _ , where the @xmath180 s are restricted to satisfy @xmath299 so that the contribution here comes only from @xmath300 in the first case , with a similar analysis for the second .
one uses lemma [ lemmaone ] , and the fact that the cusp form is an even function .
the error term comes from the rapid exponential decay of the tail end together with the error from the main term .
we end this subsection with the following elementary lemma that will be used often .
[ lemmazero ] let @xmath74 be fixed .
then for all @xmath301 and @xmath302 @xmath303 in this subsection all the statements are valid for maass forms @xmath17 without the assumption that it is even .
[ lemmatwo ] for @xmath304 , and @xmath305 @xmath306 the case @xmath307 is due to iwaniec @xcite . for @xmath308 ,
one follows kim s proof @xcite that @xmath309 corresponds to an automorphic cusp form on @xmath310 , and then applies molteni s extension @xcite to @xmath311 of iwaniec s method to bound sums of fourier coefficients of cusp forms ( see also @xcite , and @xcite where a similar result is proved for @xmath312 ) . for the proofs of some of the theorems in later sections , we will need stronger versions of the above lemma for the case @xmath307 , with sharp cutoff functions in suitable intervals .
it is clear that one need only prove the case when @xmath17 is either even or odd .
[ shortsum ] for any fixed @xmath313 @xmath314 and @xmath315 as we are assuming @xmath17 is even or odd , there is no loss in assuming @xmath19 when convenient .
it is also clear that the second part follows easily from the first .
we apply que with an incomplete eisenstein series , that is with @xmath316 with a function @xmath317 to be chosen , applied to the integral @xmath318 unfolding , this tells us @xmath319 let @xmath320 and set @xmath321 ( while @xmath112 is not smooth , it has compact support and is a legitimate choice ) so that @xmath322 ( i ) . .
[ propone ] and we have @xmath323 the error terms being negligible because @xmath279 is sufficiently small . interchanging the order of summation and integration and simplifying gives us @xmath324 the integral does not exceed @xmath325 , so that the statement in the proposition follows on putting @xmath326 .
\(ii ) . + the integral in can be investigated asymptotically to sufficient accuracy .
let @xmath327 and recall that @xmath328 so that @xmath329 applying stirling s formula as @xmath330 ( see section 6.1 for a similar treatment ) , we have to good accuracy that @xmath331 where @xmath332 where @xmath333 .
then for @xmath334 we have @xmath335 , where @xmath336 is ( essentially ) the abel transform @xmath337 in particular , if @xmath338 , then so is @xmath339 .
moreover , we see that @xmath340 so that now reads @xmath341 then , with the choice of @xmath342 as before @xmath343 also , if @xmath344 , then @xmath345 , say .
hence , using positivity and the above , as @xmath330 @xmath346 then , from , and the above , we have @xmath347 which proves the proposition . [ shortsumcorl ]
let @xmath348 with @xmath349 and @xmath350 .
there exist small positive constants @xmath351 and @xmath352 such that if @xmath353 , with @xmath354 fixed , then @xmath355 and @xmath356 ( here @xmath357 suffices ) .
we will use the following : there is a positive number @xmath358 such that * if @xmath359 , then @xmath360 for all @xmath361 ; * if @xmath362 , the same is true for those @xmath180 not of the form @xmath363 with @xmath364 odd . otherwise , the cosine vanishes .
note that this implies that if @xmath365 , then @xmath360 for all odd numbers @xmath180 .
thus the lower bounds in the corollary follow from prop .
[ shortsum ] with @xmath366 when @xmath359 . for the remaining cases we consider first the case when @xmath365 .
write @xmath367 with @xmath368 and @xmath369 odd .
put @xmath370 then , @xmath371 say , where @xmath372 where the @xmath373 denotes a sum over odd numbers . if @xmath374 , then @xmath375 so that @xmath376 , and we are done . in the other case , we first observe that @xmath377 substituting into gives @xmath378 since @xmath379 is bounded above independently of @xmath17 , the conclusion follows .
last , we consider the case when @xmath380 so that we write @xmath381 .
first we have @xmath382 say .
suppose @xmath383 .
then it follows that @xmath384 , so that @xmath385 . substituting this into gives the result
next , if @xmath386 , then the hecke relations imply @xmath387.then we rewrite as @xmath388 from , we also have @xmath389 . combining with
the above we conclude that @xmath390 completing the proof .
the second conclusion follows from the first .
in this section , we investigate the structure of the nodal domains as we move away from the cusp .
as is expected , the nodal domains are quite predictable very near the cusp but soon lose this feature as we move into the fundamental domain .
let @xmath165 denote the number of inert nodal domains of @xmath17 in @xmath40 , and let @xmath39 denote the nodal line . more generally , let @xmath281 denote an analytic arc segment in @xmath40 and let @xmath391 be the number of nodal domains of @xmath17 whose boundary meets @xmath281 , so that @xmath392 .
our first result gives a lower bound for @xmath165 ( and so for @xmath393 ) by counting the number of sign changes of @xmath17 on the boundary @xmath56 , since it follows from theorem [ theoremnodal ] that the number of nodal domains must exceed half the number of such sign changes .
[ theorem-3.1 ] if @xmath17 is an even hecke - maass cusp form , then @xmath394 this follows from the following proposition with @xmath395 ; [ prop-3.2 ] for any fixed @xmath396 , there are at least @xmath397 sign changes @xmath398 of @xmath124 in the segment @xmath399 , with a positive constant @xmath400
. * proof .
* for large @xmath282 , we expect that only the terms @xmath401 would contribute significantly in the sum .
since @xmath17 is even , the zeros of @xmath17 are the same as those of @xmath402 using lemma [ lemmaone ] , we see that the sum above is exponentially small and that the first term is highly oscillatory due to the behavior of the bessel function .
more precisely , let @xmath71 be a sufficiently small fixed number and choose @xmath282 so that @xmath403
. then , we have @xmath404 for any positive integer @xmath364 , let @xmath405 be the unique solution to @xmath406 so that if @xmath407 ( or not close to it ) , we have @xmath408 due to the restriction imposed on @xmath282 , it follows that one may take @xmath409 , so that we get at least @xmath397 sign changes of @xmath17 .
it remains to consider the lines @xmath410 ( and those @xmath411 close by ) .
the term @xmath412 is now the main term unless @xmath413 is very small .
so first suppose that @xmath414 , so that @xmath415 the analysis is now the same as that given above for the case @xmath416 , with @xmath282 replaced with @xmath417 , so that we assume @xmath418 and the same conclusion follows .
next , if @xmath419 , the hecke relations imply that @xmath420 , so that this time the term involving @xmath421 can be absorbed into the error term so that we obtain @xmath422 the conclusion then follows if we restrict @xmath282 so that @xmath423 . to describe the nodal domains further , we will need some results regarding the distribution of zeros of the functions @xmath424 .
it is known that ( see @xcite , @xcite ) @xmath424 has infinitely many positive simple zeros in the interval @xmath425 , and has no zeros for @xmath426 .
we denote the zeros by @xmath427 with @xmath428 such that @xmath429 . for @xmath268
fixed and @xmath430 , we have @xmath431 . on the other hand for any fixed @xmath432 , with @xmath268 growing , one has @xmath433 where @xmath434 denotes the ( negative ) zeros of the airy function @xmath260 , with @xmath435@xmath436 .
it is known that @xcite @xmath437 as @xmath438 .
the first few zeros of the airy function are approximately @xmath439 , @xmath440 , @xmath441 etc . .
we recall the definition of @xmath442 from , so that if @xmath443 , @xmath444 is nonzero ( the first zero occurs in the transitional region )
. then by lemma [ lemmaone ] , corollary [ corlone ] and , we have @xmath445 for some small @xmath71 .
it thus follows that the nodal line @xmath39 must lie in an exponentially narrow neighborhood of the lines @xmath446 .
if , as @xmath282 varies in this range away from @xmath447 , @xmath39 does not intersect these vertical lines , then there are only two possible configurations for the nodal domains here , illustrated by figure 9 .
the question of which of the two occurs is completely determined by the sign of the second fourier coefficient @xmath421 , assuming it is not zero ( because the bessel functions are of constant sign ) .
thus , if @xmath448 , we have figure 9(a ) and figure 9(b ) if @xmath449 ( the black dots in the figures are the points @xmath450 with @xmath451 and @xmath120 ) . under these conditions
the nodal domains here would be inert .
it is conceivable that @xmath39 may intersect the line @xmath446 but that would occur only if @xmath421 is non - zero but exponentially small as @xmath24 grows , which is probably unlikely ( at any event , it is easily shown that there can not be more than one such intersection in this region ) .
the analysis above can be extended to the region @xmath452 except for neighborhoods of those @xmath282 s close to the points @xmath450 ( because in this range , the terms in the sum in have exponential decay ) . in other words , @xmath39 is exponentially close to the vertical lines @xmath446 or the appropriate horizontal lines @xmath453 .
now suppose there is a split nodal domain within this region . then necessarily
this domain must be contained in a band of exponentially small thickness that weaves about the aforementioned lines .
since the length of @xmath39 is @xmath454 ( see @xcite ) the area @xmath91 of such a band is exponentially small .
however , by the faber - krahn theorem ( see prop .
[ theorem-3.4 ] below ) it is necessarily the case that @xmath455 ( here we assumed that @xmath39 is a smooth curve ) so that all nodal domains for this range of @xmath282 are inert .
we illustrate this analysis , using data from @xcite , with @xmath456 and @xmath457 in fig .
10 ( showing only half of the fundamental domain ) , where @xmath458 and @xmath459 respectively . here , @xmath460 , @xmath461 , @xmath462 and @xmath463 .
the shaded regions represent the points where @xmath464 is positive , so that the boundaries represent @xmath39 .
the topmost regions extend to infinity without any change and the first horizontal cut occurs near @xmath465 .
the situation changes rather dramatically as we move into the region @xmath466 since now @xmath467 so that the zeros of the main term above form a curve @xmath468 close to @xmath39 . to determine the points on @xmath468 , it is not hard to verify that for some values of @xmath282 there is a uniquely determined @xmath469 and for the rest , there are two choices of @xmath469 not close to each other in general ( we assume here , for simplicity that @xmath470 )
moreover , there are no @xmath282 s which give rise to points in the interior for which @xmath468 has a horizontal tangent except for those @xmath282 s that satisfy the simultaneous condition that @xmath444 and @xmath471 are both small .
since split nodal domains must have points ( not on the boundary @xmath56 ) with horizontal tangents , we expect that there are few ( if any ) such for this range of @xmath282 .
in particular , we expect that most nodal domains here are inert .
figure 11(a ) illustrates the domains for @xmath456 with 11(b ) a close - up view , while figure 12 does the same with @xmath457 . here
@xmath472 and @xmath473 and we observe the influence of the vertical lines @xmath474 and @xmath475 . the close - up view also shows the lack of horizontal tangents except on the boundary .
while we are unable to be more precise when we consider extended ranges of @xmath282 , the following theorem shows that there is some form of regularity ( locally ) in the neighborhoods of many special values of @xmath282 : [ theorem-3.2 ] let @xmath74 be sufficiently small and @xmath17 an even maass cusp form .
let @xmath289 be an integer satisfying @xmath476 and assume that @xmath477 .
put @xmath442 as in so that @xmath478 . then in the bounded region in the strip @xmath479 with @xmath297 , the nodal line @xmath39 is located in narrow neighborhoods ( going to zero with @xmath188 ) of the lines @xmath480 with @xmath481 odd integers .
we use the second part of proposition [ proptwo ] . by the cauchy - schwarz inequality ,
the sum is bounded by @xmath482 by lemma [ lemmatwo ] and lemma [ lemmazero ] .
the conditions on @xmath282 ensures that @xmath483 so that @xmath484 provided @xmath485 .
moreover , the airy function is asymptotically close to @xmath486 so that the assumption on the size of @xmath487 ensures that the sign changes of @xmath17 are determined by those of @xmath488 , from which the theorem follows .
we end this section by giving an extension of the courant nodal domain theorem for regions close to the cusp .
[ theorem-3.3 ] suppose @xmath39 is a smooth curve .
then , for any @xmath489 sufficiently small , there are at most @xmath490 nodal domains of @xmath17 in the region @xmath491 .
this theorem , taken together with theorem [ theorem-3.1 ] suggest the possibility that almost all the nodal domains near the cusp are inert . for @xmath492 , put @xmath493 where we put @xmath494 .
the theorem is a consequence of the following [ theorem-3.4 ] suppose @xmath39 is a smooth curve .
* let @xmath495 be integers such that @xmath496 and @xmath497 both exceed @xmath498 for some @xmath499 sufficiently small .
then , the number of nodal domains of @xmath17 in @xmath500 is @xmath501 .
* suppose @xmath502 and @xmath477 .
then the number of nodal domains of @xmath17 in the region @xmath503 is @xmath504 .
let @xmath505 be the nodal domains of @xmath17 wholly contained in @xmath500 ( so that they do not intersect the horizontal boundaries ) . since @xmath17 is an eigenfunction of @xmath506 with eigenvalue @xmath507 , on each @xmath508 , @xmath507 is the least non - zero eigenvalue of the associated dirichlet boundary value problem .
we appeal to the faber - krahn theorem for bounded domains in hyperbolic space ( see @xcite , chap .
4.2 ) which states that if the boundary of @xmath508 is smooth ( which we will assume ) , there is a constant @xmath509 such that @xmath510 summing over all the @xmath508 s gives us @xmath511 since @xmath512 .
the remaining nodal domains necessarily intersect the horizontal boundaries @xmath442 or @xmath513 so that the number of such nodal domains is bounded by the number of zeros of @xmath17 on these closed horocycles .
applying the formula with @xmath514 and with @xmath515 , we have @xmath516 for some constant @xmath517 and for any small @xmath518 .
now suppose that @xmath519 .
it follows that if @xmath281 is a zero of @xmath520 , then @xmath281 is near @xmath521 for some odd integer @xmath180 .
this zero is necessarily simple since prop .
[ proptwo ] can be extended to @xmath522 so that the main term is not small at @xmath281 .
we now consider the complexification of @xmath523 , with @xmath469 a complex variable .
the estimates for the fourier coefficients from , , and the bessel function show that @xmath524 is a holomorphic function of @xmath525 for @xmath526 , for any real number @xmath281 . applying jensen s lemma ,
in each ball of radius @xmath527 centered at @xmath281 there are @xmath528 zeros so that @xmath523 has at most @xmath529 real zeros for @xmath530 .
part one of the proposition then follows .
the second part follows using @xmath531 since @xmath245 . to complete the proof of theorem [ theorem-3.3 ]
, we have to choose @xmath289 in prop . [ theorem-3.4 ]
such that our condition @xmath532 is satisfied . by the hecke relations , for any prime number @xmath186 , this condition is satisfied with @xmath533 or @xmath534 .
choosing @xmath535 and @xmath186 sufficiently large but independent of @xmath188 gives us our conclusion .
for @xmath536 , denote by @xmath537 the closed horocycle on @xmath40 given by the points @xmath538 , with @xmath539 .
we are interested in the restriction of @xmath17 to @xmath537 .
we first give a sharp bound for the @xmath61-restriction in [ theoremone ] * for any fixed @xmath206 and any @xmath74 @xmath540 where the implied constants depend on @xmath280 and @xmath541 .
* for all @xmath542 with any fixed @xmath273 we have @xmath543 using this , we then prove [ theoremtwo ] for @xmath74 and fixed @xmath544 @xmath545 it is easy to see that if @xmath546 for some sufficiently large constant @xmath91 , then there are two zeros of @xmath124 on the horocycle @xmath537 .
this follows from the exponential decay of the bessel function so that the term @xmath547 dominates @xmath548 .
moreover , since there are no real zeros of @xmath549 with @xmath550 , the cosine for even maass forms ( or sine for odd forms ) provides the sign change as @xmath469 varies .
jung @xcite has recently established a lower bound of @xmath551 for @xmath552 in theorem [ theoremone ] without having to assume that @xmath17 is a hecke eigenform ( such a bound follows trivially from the argument given in section 2.2 for hecke eigenforms ) .
this , together with results of toth and zelditch @xcite allows him to establish the upper bound in theorem [ theoremtwo ] of @xmath454 for such @xmath17 s and , in fact , for any hyperbolic surface with finite area ( without any arithmetic assumptions ) . in appendix
a , we obtain the analog of the above theorems as @xmath541 moves towards the cusp .
the lower bound above depends on standard que and is responsible for our assumption that @xmath541 be bounded .
it is not sufficient for our purposes as @xmath541 grows with @xmath188 for which we appeal to a conjectured quantitative version of que ( see also the introduction for precise statements ) . while the theorems above hold for fixed @xmath541 , we show in theorem [ theoremthree ] below that as we approach the cusp with @xmath553 , there are some horocycles with many zeros that are evenly spaced out on the horocycle ( see also theorem [ theorem-3.2 ] ) .
[ theoremthree ] let @xmath74 be sufficiently small and @xmath17 an even maass cusp form .
there are @xmath554 integers @xmath289 satisfying @xmath476 such that there are @xmath555 zeros of @xmath17 on the horocycle @xmath556 , with @xmath557 .
thus , there are a growing number of horocycles @xmath537 such that each have @xmath558 zeros of @xmath17
. the asymptotic behavior of @xmath559 in theorem [ theoremtwo ] should follow that predicted by a random wave model . in appendix
b we show that such a model predicts that for @xmath560 @xmath561 which is consistent with the result in the above theorem .
[ [ section ] ] * proof of theorem [ theoremone ] . * for @xmath541 and @xmath506 as in prop .
[ propone ] , we have @xmath562 with the error term uniform in @xmath541 .
it is easily seen that the error term is @xmath563 so that applying the cauchy - schwarz inequality and using , we have @xmath564 it follows from lemma [ lemmatwo ] that we obtain the desired upper - bound ( for @xmath541 not fixed ) , since the right hand side above is @xmath565 for the lower bound , using lemma [ lemmaone ] and discarding some terms from the resulting sum by positivity , we have for any @xmath206 ( still not fixed ) @xmath566 with @xmath567 as given in .
let @xmath568 be small and define the set @xmath569 then , @xmath570 is contained in the set @xmath571 the pair of integers @xmath572 as above are lattice points in @xmath573 which are close to the real analytic curve @xmath574 with @xmath469 in the appropriate range .
since this curve has no line segments , it is known since van der corput ( see huxley @xcite ) that the number of such lattice points satisfies @xmath575 .
we choose @xmath576 with @xmath577 small and fixed .
the contribution to the sum in from those @xmath180 in @xmath154 is @xmath578 using and lemma [ lemmatwo ] . on the other hand for @xmath180 not in @xmath154 ,
the contribution is @xmath579 so that on combining with the previous analysis and , we get for @xmath580 @xmath581 which gives us our lower bound on applying prop .
[ shortsum ] , for any fixed @xmath541 .
[ [ section-1 ] ] * proof of theorem [ theoremtwo ] . *
the upper bound is a consequence of the complexification argument of toth - zelditch ( @xcite , theorem 6 ) where the lower bound obtained in theorem [ theoremone ] above ( or more generally the weaker lower bound of @xcite ) is used to satisfy the notion of a good curve " in @xcite .
it remains to prove the lower bound in the statement of our theorem .
let @xmath582 be the number of sign - changes of @xmath124 on @xmath537 , say at the points @xmath583 with @xmath584 ( we put @xmath585 and @xmath586 ) .
then , @xmath587 using the @xmath85 estimate @xmath588 in @xcite and theorem [ theoremone ] , we have @xmath589 on the other hand , by prop .
[ propone ] , the second integral in is bounded by @xmath590 which is then bounded by @xmath591 after using the cauchy - schwarz inequality , lemma [ lemmazero ] and lemma [ lemmatwo ] .
then from we conclude @xmath592 from which our lower bound for @xmath582 follows .
if one were to use the conjectural @xmath85 estimate @xmath593 , valid for @xmath100 in a compact part of @xmath7 , then one would obtain @xmath594 .
[ [ section-2 ] ] * proof of theorem [ theoremthree ] . *
the proof is the same as in theorem [ theorem-3.2 ] so that @xmath595 thus , if @xmath596 , the error terms are negligible provided @xmath487 is not too small .
we use the hecke relations @xmath597 valid for all prime numbers @xmath598 .
it follows that either @xmath599 or if not , then @xmath600 .
we then choose @xmath289 accordingly , so that now we can find @xmath601 values of @xmath289 in the interval @xmath602 with the main term dominating in .
for each such fixed @xmath289 , we then see that there are @xmath555 sign changes of @xmath524 for @xmath539 .
in this section , which is the heart of the paper , we prove theorem [ theoremi2 ] and theorem [ theoremi5 ] . however , in section 6.1 , we first give bounds for the @xmath61-restriction of @xmath17 to any complete geodesic joining two cusps . the rest of the section
is then devoted to proving the main result which gives a sharp lower bound for this @xmath61-restriction to a sufficiently long but fixed segment @xmath48 of @xmath52 ( and which carries over without change to @xmath53 ) .
this gives a localization of theorem [ theorem5.1 ] below ; the proof involves choosing a similar auxiliary function @xmath603 but the analysis is more intricate . in @xcite
we develop the analogue of theorem [ theorem5.1 ] but for a non - split closed geodesic about which reflection defines a global symmetry of a compact hyperbolic surface ( so that for this result no arithmetic tools are used ) .
[ theorem5.1 ] let @xmath81 denote any geodesic joining two cusps .
then @xmath604 by assuming a suitable version of the quantitative que ( see appendix a ) , one can get an asymptotic formula for the integral in the theorem of the type @xmath605 .
we provide a proof of this theorem when @xmath81 is @xmath52 ( see the remarks at the end for the general case ) .
we will denote @xmath188 by @xmath24 to distinguish it from the imaginary part of the complex variable @xmath606 .
recall that @xmath607 where for @xmath608 @xmath609 by parseval s formula , it then follows that @xmath610 + + ( i ) .
+ using stirling s formula , the product of gamma functions above is bounded by @xmath611 we may then neglect the contribution from say , @xmath612 in for it gives us an estimate of exponential decay for that part of the integral , after using and the fact that the @xmath21-function has at most polynomial growth in @xmath188 . for the remaining , we then have @xmath613 if we were now to assume the lindelof hypothesis , that is @xmath614 , our required upper bound for theorem [ theoremi2 ] would follow easily . to obtain an unconditional result
, we replace @xmath615 with dirichlet polynomials using the approximate functional equation . to this end
, we use theorem 2.5 from @xcite which implies that for @xmath112 a smooth real function satisfying @xmath616 , we have @xmath617 where @xmath618 is the analytic conductor of @xmath22 given by @xmath619 the contribution from the error term above to is seen to be @xmath620 .
we choose @xmath621 such that it is zero for @xmath622 , is decreasing smoothly for @xmath623 and is non - negative otherwise , so that @xmath624 in the sum above .
we break the integral in into the two ranges @xmath625 and @xmath626 , so that after suitable changes of variables , we finally have to estimate @xmath627 where @xmath628 and @xmath629 . expanding the sum in the integral above
, we see that the diagonal contribution is @xmath563 . for the non - diagonal terms
, we break the integral into @xmath528 dyadic subintervals of the type @xmath630 $ ] so that we have to estimate @xmath631 where @xmath373 indicates that the variables are restricted to satisfy @xmath632 .
since the product of the @xmath112-functions in the integral are monotonic in the variable @xmath188 , integration by parts shows that the integral is bounded by @xmath633 so that after a suitable change of variables , we are left to estimate @xmath634 so that we obtain the required bound of @xmath620 after applying and the restriction imposed above on @xmath364 and @xmath180 ( and so also on @xmath635 ) .
+ + ( ii ) . + the method here is similar to the one above but more intricate . from , applying stirling s formula and discarding a portion of the integral , we may rewrite as implying @xmath636 where @xmath637 is a fixed , non - negative @xmath638-function supported in the interval @xmath639 , with @xmath640 , and where we define @xmath641 as in , we may now write the approximate functional equation in the form @xmath642 since the conductor has size @xmath643 .
the phase function @xmath644 satisfies @xmath645 to estimate the lower bound for the second integral in , we use cauchy - schwarz as follows : fix @xmath646 ( which will be chosen small ) and set @xmath647 consider @xmath648 this is @xmath649 using standard mean - value theorems for dirichlet polynomials and theorem [ shortsum ] . to evaluate the integral in , we expand the sums and use the approximate functional equation ( discarding the error term ) to get @xmath650 the diagonal term @xmath651 above gives , using theorem [ shortsum ] , a lower bound @xmath652 with an absolute constant . for the integral in the first sum in , for the non - diagonal terms , repeated integration by parts
gives an estimate of the type @xmath653 , for any integer @xmath654 . substituting this and estimating the resulting sums using prop .
[ shortsum ] gives an estimate of the type @xmath655 .
for the rest of the sums , we observe that the oscillatory integrals have stationary points only when @xmath656 ; however , if @xmath109 is chosen small enough , these are avoided . moreover ,
the derivative of the phase is monotonic , so that the integral is bounded by @xmath657 . by cauchy - schwarz , and prop .
[ shortsum ] , the resulting sum in is @xmath658 .
thus the diagonal term dominates and combining the estimates from and gives us the lower bound in the theorem . for the case of arbitrary geodesics with two cusps as endpoints , it suffices to consider the vertical half - lines with real part @xmath659 , for @xmath660 and with @xmath661 and @xmath187 coprime . following the reasoning above ,
@xmath22 is replaced with @xmath662 , which is the @xmath21-function twisted with @xmath663 and similarly for @xmath603 , but with a modification for the range of summation .
this modified l - function has a approximate functional equation ( which may be derived by the methods in @xcite ) of the type @xmath664 where @xmath665 is the multiplicative inverse of @xmath661 modulo @xmath187 .
the upper bound follows immediately . for the lower bound , the incooperation of terms depending on @xmath187 present no new difficulties as @xmath187 is fixed , so that the conclusions above still hold , with the exception of the nature of the diagonal term which is now of the form @xmath666 by corollary [ shortsumcorl ] with @xmath667 .
we are now ready to prove the main result , namely the lower bound in theorem [ theoremi2 ] for @xmath48 a sufficiently long ( but fixed ) compact segment of @xmath56 .
the analysis is delicate and it would be more direct and simpler in terms of showing the off - diagonal terms are smaller than the diagonal , if we had a more flexible version of que as in ( indeed , one would then be able to take @xmath48 to be an arbitrary fixed segment on @xmath56 ) . since this last is not available to us , we make do with standard que and a cruder , more involved analysis in which we are able to show that the diagonal dominates the off - diagonal ( but with both the same order of magnitude in @xmath24 ) , if @xmath48 is long enough .
we begin by setting up the test function to be used .
let @xmath661 be a positive number independent of @xmath188 to be chosen ( @xmath661 will be suitably large ) and let @xmath668 ( again to be chosen later ) .
let @xmath669 be an even function satisfying the following properties for @xmath670 : * @xmath671 for @xmath672 , * @xmath673 for @xmath674\cup[2a+b,\infty)$ ] , * @xmath675 for @xmath676 and * @xmath677 for some constant @xmath678 with any @xmath679 .
the support of @xmath103 determines the corresponding segment on @xmath52 which we denote by @xmath48 .
the size of @xmath680 is determined by the size of @xmath661 and any constant that depends on either will generically be denoted by @xmath681 .
similarly , we will let @xmath253 denote a generic positive constant independent of @xmath661 and @xmath188 but which will change .
the parameter @xmath289 will be chosen suitably large later on but will be independent of @xmath661 and @xmath188 . at the end of the proof , we will choose @xmath661 suitably and then let @xmath24 become arbitrarily large so that the dependence of @xmath681 on @xmath661 will be unimportant
. in places where the value of @xmath661 plays an important role we will be precise and indicate this dependence .
finally , the symbol @xmath682 will mean an implied constant independent of @xmath661 and @xmath24 . for any real function @xmath683
, we have @xmath684 by , we write @xmath685 by , and we have @xmath686 .
put @xmath687 with @xmath688 to be chosen later and define @xmath689 then , we write in , @xmath690 .
we provide the analysis for @xmath451 and @xmath120 together , so that we write @xmath691 where we define @xmath692 and with @xmath693 while @xmath694 does the same except it vanishes for @xmath695 .
we decompose the sum above into three pieces : @xmath696 denoting the part of the sum with @xmath697 which is the diagonal , @xmath698 denoting the sum with small spacing @xmath699 and @xmath700 for the large spacings .
we first deal with @xmath700 and prove the following [ prop - e10 ] for any integer @xmath703 @xmath704 as @xmath88 .
the proof of this proposition relies heavily on the asymptotics of @xmath705 which is determined with the following [ lemma - e1 ] * for @xmath706 , @xmath707 and @xmath400 real ( see @xcite ) @xmath708 * if @xmath709 and @xmath710 , then ( @xcite ) @xmath711 * if @xmath712 , then ( @xcite ) @xmath713 let @xmath714 be the fourier transform .
then , applying integration by parts multiple times shows that @xmath715 as @xmath716 for any integer @xmath654 so that @xmath717 for all @xmath469 .
moreover , if @xmath718 , we have first that @xmath719 if @xmath720 and since @xmath721 we have also @xmath722 , with @xmath289 chosen suitably large .
we first note that in the contribution to @xmath700 from either @xmath364 or @xmath180 exceeding @xmath723 is exponentially small due to the decay of the bessel functions ( since @xmath724 $ ] ) .
thus we may assume both @xmath725 in what follows ( denoted by an asterisk in the summation ) so that using lemma [ lemma - e1 ] @xmath726 for some constant @xmath273 .
we put @xmath727 so that @xmath728 and in lemma [ lemma - e1 ] we put @xmath729 then if @xmath730 is small , @xmath731 , while if @xmath732 , then @xmath733
. thus since @xmath734 and @xmath735 , in either case we have @xmath736 is sufficiently large so that we may use part ( 3 ) of lemma [ lemma - e1 ] in part ( 2 ) to get @xmath737 we first break the integral in with @xmath738 $ ] and @xmath739 . for @xmath740 ,
using we have @xmath741 we consider two cases : \(i ) if @xmath742 , then @xmath743 , so that the contribution to the integral in is @xmath744 the contribution to @xmath700 is then @xmath745 for @xmath289 chosen large .
this contribution will be negligible and will be ignored .
next the case \(ii ) if @xmath746 . now
@xmath747 so that @xmath748 now , the contribution to @xmath700 is @xmath749 we break the sum into two pieces : first with @xmath750 .
then , @xmath751 so that the contribution from the sum is negligible if @xmath289 is chosen suitably large . for the rest , @xmath752 , so that the sum is @xmath753 for the remaining range of @xmath469 with @xmath754 we see that @xmath755 .
we first consider the contribution from the error - terms in . if @xmath756 then @xmath757 so that @xmath758
. then @xmath759 the contribution to @xmath700 in is @xmath760 which is negligible .
if on the other hand @xmath761 , the error term in gives us @xmath762 .
then , as in the case above , @xmath763 the contribution to @xmath700 is @xmath764 we now consider the contribution from the main term in to @xmath700 .
denoting it by @xmath765 and using integration by parts , we estimate @xmath766 since @xmath767 and @xmath768 , the contribution from these terms to @xmath700 is seen to be the same as in .
next , it is easily seen that @xmath769 so the contribution to is ( following the remarks made after lemma [ lemma - e1 ] ) @xmath770 thus , the contribution to is then @xmath771 we now estimate the sum in by considering three cases : @xmath772 , @xmath773 and @xmath774 . in the first case ,
the main term in is @xmath775 for the second case , we see that @xmath776 so that in we find the main term is , by lemma [ lemmatwo ] @xmath777 which is negligible .
finally , for the last case , we have @xmath778 so that both @xmath364 and @xmath180 exceed @xmath779 so that we may reverse their roles and get the estimate obtained in the first case , that is . collecting the estimates from , , and
, we see that @xmath780 as @xmath781 , for @xmath782 and @xmath120 .
we first remark that in our application at hand , we will not need to deal with small spacings since we are able to choose @xmath783 .
this is due to the fact that we choose our segment @xmath48 to be sufficiently long .
however , we include the analysis here to indicate that que can be used to control such sums .
it suffices to deal here with @xmath784 .
this is because when @xmath38 is bounded , @xmath785 is a sum where @xmath364 and @xmath180 both run through a bounded interval ( depending on @xmath38 ) so that the contribution from them is negligible as @xmath330 and if @xmath786 as @xmath661 grows ( we use here the bounds for the fourier coefficients and the estimates for the bessel function away from the @xmath787 in corollary [ corlone ] ) . [ prop - e11 ] for bounded @xmath38 , @xmath788 for the proof , we use the notation from the subsection above and rewrite @xmath789 we denote the above as @xmath790 .
we will show that @xmath791 is small using que directly , while @xmath792 will be bounded in a manner similar to @xmath793 in the previous subsection .
we will first consider @xmath794 .
we assume @xmath795 so that @xmath796 , and since @xmath635 is relatively small , the integral in is @xmath797 by lemma [ lemmaone ] . thus @xmath798 it may appear that the estimate above for the integral is rather crude .
a careful analysis shows that there is not sufficient oscillation in the integral to extract futher savings since @xmath799 .
we next consider @xmath791 .
let @xmath800 with the support contained in the interval @xmath801 with @xmath802 .
for any @xmath803 and @xmath804 we have the incomplete poincare series @xmath805 the sum is finite and @xmath806 is @xmath101-invariant . moving @xmath100 into the standard fundamental domain
we see that @xmath807 uniformly in @xmath100 .
since @xmath808 for @xmath809 , it follows from que that there is a function @xmath810 such that @xmath811 unfolding the integral above and using the fact that @xmath17 is even , we see that the integral in is equal to ( up to to a multiplicative constant ) the sum over @xmath364 in @xmath791 with @xmath812 , so that @xmath813 which is negligible if @xmath38 is bounded and @xmath814
. hence from , and , we have @xmath815 it is worth pointing out that a stronger version of prop .
[ prop - e11 ] holds with @xmath816 with @xmath188 if we assume a quantitative form of que . to obtain this we have to estimate the sobolev norm of @xmath817 uniformly in @xmath635 . since @xmath818 where @xmath819 it follows that for @xmath820 , @xmath821
then for @xmath112 fixed and @xmath809 @xmath822 hence , for @xmath103 fixed , the sobolev norm satisfies ( for @xmath809 ) @xmath823 then , qque ( see appendix a1 ) implies that there are constants @xmath103 and @xmath824 such that equation takes the form @xmath825 so that @xmath826 the error term here is now @xmath827 for the larger range @xmath828 .
noting that @xmath829 , from , we rewrite the diagonal term as @xmath830 where the error term comes from @xmath831 .
the main term here is @xmath832 where @xmath833 . by que
this is @xmath834 collecting all the estimates from prop .
[ prop - e10 ] , prop .
[ prop - e11 ] and the remarks made above , we conclude that @xmath835 so that there are positive absolute constants @xmath836 such that @xmath837 to estimate these sums , we appeal to prop .
[ shortsum ] , which by partial summation gives us @xmath838 on choosing @xmath839 , @xmath783 , @xmath840 and @xmath841 , say and letting @xmath661 be sufficient large .
this completes the proof of the lower bound .
the method here is exactly the same as that used in the proof of theorem [ theoremtwo ] , where one dealt with closed horocycles . for this
we obtain an upper bound for the analogue of the integral appearing in so that we have to prove @xmath842 using the discussion in sec .
6.1 , it is easy to see that parseval s theorem applied to the product of @xmath843 and the characteristic function for the subinterval @xmath844 $ ] shows that the integral in is @xmath845 in exactly the same way as was done in sec .
6.1 leading to , it then follows that the integral in is @xmath846 for some constant @xmath273 .
then the lindelof hypothesis for @xmath615 gives us the estimate in .
the usual formulation of que on the configuration space @xmath40 was given in . in our analysis , we have used the fourier expansion of @xmath17 repeatedly . we examine here the formulation of que in terms of the fourier coefficients .
for holomorphic cusp forms of even weight @xmath103 ( @xmath61-normalized with respect to the petersson inner - product ) it is shown in @xcite that que is equivalent to : for @xmath847 and @xmath848 fixed @xmath849 as @xmath850 with @xmath851 where @xmath852 are the fourier coefficients of @xmath853 .
similarly , we expect that for maass forms we have : for @xmath847 and @xmath848 fixed @xmath854 as @xmath855 . while and show that follows from que for certain @xmath635 s , there is a basic difficulty in realizing all @xmath635 s of compact support in this way .
the problem comes from the nonlocal transform @xmath339 of @xmath112 in ( and it is more complicated when @xmath856 ) , which is a feature of maass forms and is not present in the holomorphic case . as a consequence we have that implies que ( which is straightforward ) but we do nt know the converse implication .
the flexible form , in as much as it allows one to truncate the @xmath180-sum into a compact dyadic piece would allow for a much more direct treatment treatment of section 6.2 .
in fact , it allows one to establish theorem [ theoremi2 ] for @xmath48 a fixed but arbitrary small segment in @xmath56 .
this in turn would yield theorem [ theoremi6 ] for such a @xmath48 .
both forms of que above can be quantified with rates in obvious ways , and again the quantitative version of will imply the quantitative que which we call qque : [ conjecture2]*qque * + there is a @xmath857 and a @xmath858 such that for any @xmath859 with @xmath860 we have @xmath861 where @xmath862 is the sobolev @xmath103-norm on @xmath40 .
while que is known thanks to lindenstrauss @xcite and soundararajan @xcite , it is not known with any rate .
there is little doubt ( in our minds ) about the truth of conjecture [ conjecture2 ] as it follows from a subconvex estimate for the triple product @xmath21-function @xmath863 , where @xmath864 is another maass form ( fixed ie independent of @xmath17 ) or a unitary eisenstein series .
is itself a unitary eisenstein series or a dihedral cusp form , this subconvexity is known @xcite , @xcite . ]
this implication is a consequence of watson s explicit formula @xcite .
the subconvex estimate is in turn a weak form of the sharp estimate for critical values of such @xmath21-functions , known as the lindelof hypothesis .
it is well known that the latter is a consequence of the riemann hypothesis for @xmath863 so that conjecture [ conjecture2 ] has a very firm basis . for our application , we need a more explicit version of qque stated as follows : let @xmath865 [ qque2 ] there is a number @xmath857 such that for @xmath206 , as @xmath330 @xmath866 consequently , @xmath867 for @xmath868 .
the lindelof hypothesis for triple product @xmath21-functions would allow the exponent @xmath869 , in which case the second lower bound would be valid for almost the full range of @xmath541 .
this conjecture is a more precise version of the `` escape of mass '' proposition 2 of @xcite .
we now apply these quantitative versions to the analysis of the zeros of @xmath17 for short closed horocycles nearer the cusp . in the proof of theorem [ theoremone ]
the upper bound was valid for closed horocycles @xmath537 with @xmath870 , while the lower bound , which uses que was then restricted to bounded values of @xmath541 .
this restriction to bounded @xmath541 carries over to theorem [ theoremtwo ] . to extend these theorems to closed horocycles located higher up in the fundamental domain
, we need the analog of prop .
[ shortsum ] for shorter sums . for such sums , we appeal conjecture [ qque2 ] .
[ lemmaathree ] assume conjecture [ qque2 ] .
then for any @xmath871 @xmath872 the proof is identical to that in prop .
[ shortsum ] with @xmath873 and replacing with the conjecture .
using prop .
[ lemmaathree ] we have the following : let @xmath874
. then [ theoremaone ] assume conjecture [ qque2 ] . for any @xmath74 and @xmath875 @xmath876 we then have [ theorematwo ] assume conjecture [ qque2 ] . for @xmath74 and @xmath877 @xmath878
theorem [ theoremthree ] suggests that the upper bound here may be close to best possible .
the proof of theorem [ theoremaone ] follows directly from the proof of theorem [ theoremone ] since implies our result on using the analog of prop .
[ shortsum ] given in prop .
[ lemmaathree ] . + * proof of theorem [ theorematwo ] .
* it is clear that the above conditional results provide a proof of the lower bound using the analysis following . for the upper bound , we give a direct proof ( suggested by the complexification idea in @xcite ) as follows : for any @xmath206 , define for @xmath879 @xmath880 the estimates for the fourier coefficients from , , and the bessel function
show that @xmath881 is a holomorphic function in the horizontal strip given by @xmath882 .
moreover , it has at most a polynomial rate of growth in @xmath24 in such a strip .
the lower bound in theorem [ theoremaone ] shows that @xmath881 is not identically zero .
then , by jensen s lemma , it follows that @xmath881 has at most @xmath528 zeros in a disc of radius at most @xmath883 centered at a point @xmath469 real .
hence , @xmath881 has at most @xmath884 real zeros in the interval @xmath885 which then implies the upper bound for the zeros in the theorem .
in @xcite we used the random wave model in edelman - kostlan @xcite to obtain predictions on the distribution of zeros of the real part of holomorphic cusp forms on vertical lines .
we apply the same methods here to even cusp forms but on horizontal line segments . from ( 7 )
we write @xmath886 for any fixed @xmath541 , which may vary with @xmath24 , we consider the vectors @xmath887 where @xmath888 denotes the unit vector with the @xmath889 in the nth- + coordinate .
the probability density function for the real zeros of the the associated random wave function is given by @xmath890 where @xmath891 is the derivative with respect to @xmath469 and with the standard inner - product .
this is the expected number of real zeros of the associated wave function per unit length at the point @xmath469 .
unlike the situation in @xcite , we do not expect the term @xmath892 to contribute to the main term , in which case , we should have @xmath893 define , for @xmath251 @xmath894 so that with @xmath895 we have @xmath896 for @xmath469 not close to a half - integer , we would not expect the terms above involving @xmath469 to contribute due to cancellation . using lemma [ lemmaone ]
, we write the sum @xmath897 for @xmath180 s restricted by the ranges as indicated there . using corollary [ corlone ] , we see that @xmath898 is negligible . similarly , @xmath899 is bounded by @xmath900 if @xmath901 . in
what follows , we will restrict @xmath541 to satisfy @xmath902 , so that @xmath903 .
put @xmath904 and @xmath905 so that @xmath906 where we denote by @xmath907 any function of the type @xmath908 , with @xmath253 a constant .
we show that the second sum in above is @xmath827 for all @xmath469 , while the same is true for the first sum except for those @xmath469 near integers . for @xmath909 ,
the first sum is asymptotic to @xmath910 and it is easy to show it is @xmath827 if @xmath911 . to deal with the second sum , we use the following lemma from @xcite ( pg .
306 ) : for those @xmath917 , with @xmath577 sufficiently small , the sum is trivially @xmath827 . for @xmath918 , we subdivide the sum into subsums with @xmath919 with @xmath920 . for @xmath180 in such a subsum @xmath921 moreover , we may take @xmath922 so that the lemma gives an estimate for each subsum of @xmath923 simplifying and summing over @xmath432 then gives us the estimate of @xmath827 if @xmath924 , as is the case .
we thus have @xmath925 except for those @xmath469 with @xmath926 . in exactly the same way and with the same restriction , @xmath927 while @xmath928 .
this then gives us @xmath929 uniformly for all @xmath469 except those satisfying @xmath930 , with @xmath931 ( the last condition may probably be relaxed ) . integrating over all @xmath932 , using the estimate @xmath933 for the exceptional @xmath469 s , we obtain the expected number @xmath934 of zeros of @xmath124 on such a horocycle ( compare with theorem [ theoremthree ] and theorem [ theorematwo ] ) . in this appendix , we sketch the proof of the remark made in the introduction regarding the nodal domains of an odd maass form
. let @xmath17 be an odd , real hecke - maass eigenform , so that @xmath17 satisfies @xmath935 .
hence @xmath17 vanishes identically on @xmath56 ,
which is therefore part of the nodal line of @xmath17 , and so we consider the nodal domains within @xmath148 in figure 3 . to prove the analogous lower bound for the number of nodal domains for @xmath17 which meet @xmath95 ( these being the ` inert ' nodal domains ) we seek sign - changes of the normal derivative @xmath936 of @xmath17 along @xmath56 ( and consequently those of @xmath937 along @xmath52 and @xmath53 ) .
each point in @xmath48 at which there is such a sign - change must be the end point of an interior nodal line and a topological argument similar to the one in section 2 shows that the number of associated nodal domains is at least half of the number of such sign - changes . as before in order to produce sign changes in @xmath936 we compare its @xmath1-norm , its @xmath0-norm and also its meanvalue all restricted to @xmath48 .
write @xmath941 and for any @xmath942 , put @xmath943 for fixed @xmath282 , we apply bernstein s inequality to get @xmath944 then choose @xmath945 so that @xmath946 and all its derivatives are well approximated by @xmath17 and its derivatives respectively , with an error that decays exponentially in @xmath188 so that the lemma follows .
next , we need to prove @xmath947 which is the analogue of .
this is obtained in exactly the same way as the discussion there , where now becomes @xmath948 then , stirling s formula together with the lindelof hypothesis gives the required result .
finally , we need to prove , as in section 6.2 @xmath949 the method is the same , with the non - diagonal terms offering no new problems . however , the diagonal term can not be evaluated as in section 6.2.3 because @xmath950 is not automorphic so that the diagonal is not an inner - product involving an incomplete eisenstein series .
instead we obtain a lower - bound for the diagonal using positivity , asymptotic estimates for the bessel functions and finally que by using prop .
[ shortsum ] .
the constant obtained this way is much smaller than that obtained for the even case , but does not affect the outcome if @xmath661 is chosen large enough . to obtain the analogue of section 6.2.2 we could have proceeded by using the gradient of @xmath17 instead of @xmath951 as follows @xmath952 for small values of @xmath635
one may rewrite the last integral , after integrating by parts , as an inner product with incomplete eisenstein and poincare series against @xmath233 , with a small error , so that que can be applied . for the large values of @xmath635
, one has to expand the sums and analyse the resulting expression
. this device will prove useful if were to hold .
i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series , and products _ , translated from the russian .
6th edition .
translation edited and with a preface by alan jeffrey and daniel zwillinger . academic press , inc .
, san diego , ca , 2000 . | this paper deals with some questions that have received a lot of attention since they were raised by hejhal and rackner in their 1992 numerical computations of maass forms .
we establish sharp upper and lower bounds for the @xmath0-restrictions of these forms to certain curves on the modular surface .
these results , together with the lindelof hypothesis and known subconvex @xmath1-bounds are applied to prove that locally the number of nodal domains of such a form goes to infinity with its eigenvalue .
= 1.0 cm |
the @xmath0 ising model @xcite is a standard theoretical laboratory to study the effects of quenched disorder and frustration on the critical behavior of spin systems .
we consider the two - dimensional ( 2d ) @xmath0 ising model defined on a square lattice by the hamiltonian @xmath12 where @xmath13 , the sum is over all pairs of lattice nearest - neighbor sites , and the exchange interactions @xmath14 are uncorrelated quenched random variables , taking values @xmath0 with probability distribution @xmath15 for @xmath2
we recover the standard ising model , while for @xmath3 we obtain the so - called bimodal ising glass model . for @xmath16 , the disorder average of the couplings
is given by @xmath17=j(2p-1)\neq 0 $ ] and ferromagnetic ( or antiferromagnetic ) configurations are energetically favored . note that its thermodynamic behavior is symmetric under @xmath18 . in the following we set @xmath19 without loss of generality .
ising model in the @xmath20-@xmath8 plane .
the phase diagram is symmetric under @xmath18 . ]
the 2d @xmath0 ising model has been extensively investigated ; see , e.g. , refs .
@xcite for recent reviews . as sketched in fig .
[ phasedia ] , at finite temperature it presents a paramagnetic and a ferromagnetic phase .
they are separated by a transition line , which starts at the pure ising transition point , at @xmath2 and @xmath21 , and ends at the disorder - driven ferromagnetic @xmath9 transition , at @xcite @xmath7 .
the point where this transition line meets the so - called nishimori ( n ) line @xcite , at @xcite @xmath22 and @xmath23 ( see also refs . @xcite for analytical estimates of @xmath24 ) is a multicritical point ( mnp ) @xcite .
the mnp divides the paramagnetic - ferromagnetic ( pf ) transition line in two parts .
the pf transition line from the ising point at @xmath2 to the mnp is controlled by the ising fixed point . here
disorder gives only rise to ( universal ) logarithmic corrections to the standard ising critical behavior ; see , e.g. , ref .
@xcite and references therein .
the slightly reentrant low - temperature pf transition for @xmath25 belongs instead to a different strong - disorder ( sdi ) universality class @xcite .
several studies have discussed the possible existence of a glassy transition , considering in most of the cases models with symmetric disorder distributions such that @xmath17=0 $ ] . at variance with the three - dimensional case ,
no finite - temperature glassy phase occurs and a critical behavior is only observed at @xmath9 .
moreover , recent results for the bimodal ising model @xcite have provided compelling evidence that the bimodal ising model and other models with symmetric continuous disorder distributions , for instance the gaussian distribution , undergo a zero - temperature glassy transition in the same universality class . in all cases , for @xmath26 the correlation length increases as @xmath27 with @xcite @xmath28 . even though the asymptotic critical behavior is the same , for small values of the temperature and in finite volume the bimodal ising glass model , and in general any model with a discrete disorder distribution , presents a peculiar phenomenon related to the discreteness of the hamiltonian spectrum .
this implies that , although the bimodal and gaussian ising glass have the same critical behavior in thermodynamic limit @xcite , i.e. , if one takes @xmath29 before @xmath26 , much more care has to be exercized in the finite - size scaling ( fss ) limit .
in particular , as we shall see , in the bimodal ising glass model one can not observe the standard fss limit in terms of the scaling variable @xmath10 .
an appropriate variable is instead the renormalization - group ( rg ) invariant quantity @xmath11 . in the case of the 2d @xmath0 ising model
, a natural scenario is that a zero - temperature glassy transition occurs for any @xmath8 in the range @xmath6 , and that the glassy critical behavior is independent of @xmath8 .
this implies that a nonzero @xmath17 $ ] is irrelevant for the critical behavior , as found in mean - field models @xcite and in the 3d @xmath0 ising model @xcite . in this paper , which completes a series of papers @xcite devoted to the study of the phase diagram and critical behavior of the 2d @xmath0 ising model
, we investigate the glassy behavior for @xmath6 . for this purpose
, we present monte carlo ( mc ) simulations at @xmath4 , which is relatively close to the low - temperature pf line ( @xmath30 ) , and at @xmath3 , up to lattice sizes @xmath31 and for @xmath32 .
as we shall see , our results provide a strong evidence of the universality of the glassy zero - temperature critical behavior and thus provide strong support to the scenario of a universal glassy critical line for @xmath9 and @xmath6 .
moreover , we provide evidence that the magnetic and glassy behaviors at the @xmath9 multicritical glassy point ( mgp ) , where the low - temperature pf and the @xmath9 glassy transition lines meet , at @xmath7 , see fig .
[ phasedia ] , are decoupled .
finally , we discuss the critical behavior of the overlap quantities along the pf line that connects the mnp to the @xmath9 mgp : we observe an apparently @xmath20-dependent critical behavior .
the paper is organized as follows . in sec .
[ sec : mc ] we define the quantities we have considered in the mc simulations . in sec .
[ sec : results ] we discuss the behavior at the @xmath9 glassy transition . in particular , we discuss the freezing phenomenon observed in a finite volume at very small temperatures due to the discreteness of the hamiltonian spectrum , the universality of the glassy critical behavior , and the critical behavior of the overlap susceptibility . in sec .
[ sec : resultspf ] we discuss the critical behavior of the overlap correlations along the low - temperature paramagnetic - ferromagnetic transition line , below the mnp . finally , in sec .
[ sec : conclusions ] we present our conclusions .
the critical modes at the glassy transition are those related to the overlap variable @xmath33 , where the spins @xmath34 belong to two independent replicas with the same disorder realization @xmath35 . in our mc simulations we measure the overlap susceptibility @xmath36 and the second - moment correlation length @xmath37 defined from the correlation function @xmath38 = [ \langle \sigma_0 \,\sigma_x \rangle^2]$ ] , where the angular and the square brackets indicate the thermal average and the quenched average over disorder , respectively .
we define @xmath39 and @xmath40 where @xmath41 , @xmath42 , and @xmath43 is the fourier transform of @xmath44 .
we also consider some quantities that are invariant under rg transformations in the critical limit , which we call renormalized couplings .
we consider the ratio @xmath11 and the quartic cumulants @xmath45\over [ \rho_2]^{2 } } , \quad u_{22 } \equiv { [ \rho_2 ^ 2 ] -[\rho_2]^2 \over [ \rho_2]^2 } , \label{uc}\ ] ] where @xmath46 . in the case of a @xmath9 transition with a nondegenerate ground state ,
as expected in the 2d ising glass model with a gaussian disorder distribution , we have @xmath47 for @xmath9 , hence the corresponding overlap - susceptibility exponent @xmath48 vanishes , @xmath49 , and @xmath50 for @xmath26 .
in particular , @xmath51 indicates the self - averaging of the ground - state distribution , as already suggested by the results of ref .
moreover , since @xmath49 , it is natural to conjecture that the two - point overlap function becomes essentially gaussian in the limit @xmath52 .
if this occurs , we also have @xmath53
. as we shall see , the results for the @xmath0 ising model are consistent with these predictions .
we also consider magnetic quantities .
we define the magnetic susceptibility @xmath54 and the second - moment correlation length @xmath55 in terms of the magnetic two - point function @xmath56.\ ] ] for symmetric disorder distributions we have @xcite @xmath57 and @xmath58 for any @xmath20 . for other values of @xmath8 , we expect them to converge to a finite nonuniversal value .
we also consider the four - point magnetic susceptibility @xmath59 defined by @xmath60/l^2 , \\ & & \mu_{k } \equiv \ < \ ; ( \sum_x \sigma_x\ ; ) ^k \>.\end{aligned}\ ] ] for symmetric disorder distributions we have @xcite @xmath61 assuming universality we expect @xmath62 also for nonsymmetric disorder distributions , i.e. , for any @xmath63 .
, @xmath64 , and @xmath65 versus @xmath20 at @xmath4.,title="fig : " ] , @xmath64 , and @xmath65 versus @xmath20 at @xmath4.,title="fig : " ] , @xmath64 , and @xmath65 versus @xmath20 at @xmath4.,title="fig : " ] , @xmath64 , and @xmath65 versus @xmath20 at @xmath3.,title="fig : " ] , @xmath64 , and @xmath65 versus @xmath20 at @xmath3.,title="fig : " ] , @xmath64 , and @xmath65 versus @xmath20 at @xmath3.,title="fig : " ] we perform mc simulations of the square - lattice @xmath0 ising model with periodic boundary conditions for @xmath4 and @xmath3 and for several values of the lattice size @xmath66 , with @xmath67 .
we employ the metropolis algorithm , the random - exchange method @xcite , and multispin coding .
furthermore , for the largest lattices we use the cluster algorithm described in ref . @xcite . for each lattice size
we collect data in the range @xmath68 . at @xmath4
we take @xmath69 for @xmath70 , @xmath71 for @xmath72 , @xmath73 for @xmath31 . at @xmath3
we take @xmath74 for @xmath75 and @xmath76 for @xmath77 .
typically , we consider @xmath78 disorder samples for each @xmath20 and @xmath8 . in a few cases
, we consider @xmath79 disorder samples . in the following ,
we first discuss the freezing regime , which occurs for sufficiently low temperatures in any finite system , then we prove the universality of the glassy transition by considering the fss behavior of the renormalized couplings , and finally discuss the behavior of the overlap susceptibility and of the magnetic quantities . in figs .
[ rvst0.8 ] and [ rvst0.5 ] we show the mc estimates of @xmath11 , @xmath64 , and @xmath65 .
we note that the data corresponding to different lattice sizes cross each other around @xmath80 and are mostly independent of @xmath20 for @xmath81 .
similar results for @xmath64 using the bimodal distribution were also reported in ref .
@xcite . usually , a crossing point corresponds to a transition point . instead
, in the present case in which the disorder variables are discrete , the crossing is due to a nonuniversal phenomenon which is related to the discreteness of the hamiltonian spectrum @xcite . to review the argument ,
let us consider the states corresponding to the two lowest energy values for a given lattice size @xmath66 .
their energies differ by @xmath82 and their degeneracies are given by @xmath83 and @xmath84 , respectively .
numerical studies @xcite have shown that @xmath85 . at sufficiently low temperatures only the states with the lowest energy contribute to the thermodynamics .
this occurs when @xmath86 , i.e. , for @xmath87 in this regime the observed behavior is independent of @xmath20 . in the opposite limit , i.e. , when @xmath88 , the presence of the gap is negligible and the system is expected to have the same behavior as models with continuous distributions .
the crossover from one regime to the other occurs at a @xmath66-dependent freezing temperature @xmath89 which scales as @xmath90 .
it is natural to define @xmath89 by requiring @xmath91 , but this definition is somewhat unpractical . in practice , @xmath89 can be estimated from the data by identifying it with the temperature that marks the onset of the @xmath20-independent behavior of the different observables . for @xmath4 , the estimates of @xmath11 and @xmath64 reported in fig .
[ rvst0.8 ] allow us to estimate @xmath92 for @xmath93 , respectively .
slightly larger results are obtained by using @xmath65 .
the estimates of @xmath89 for @xmath3 are close to those obtained for @xmath4 , showing that @xmath89 is little dependent on @xmath8 .
consistently with the above - reported argument , the freezing temperature @xmath89 approximately decreases as @xmath90 , see fig . [ freezing ] .
a fit to @xmath94 gives @xmath95 . in the frozen region the estimates of the renormalized couplings should be very close to the corresponding @xmath9 estimates , since they are essentially determined by the lowest - energy configurations .
the data are consistent with this prediction .
indeed , see fig . [ freezing ] , for both @xmath4 and @xmath3 , @xmath64 and @xmath65 below @xmath96
slowly approach the values @xmath97 and @xmath98 , respectively , for @xmath29 .
in particular , @xmath65 apparently vanishes as @xmath99 .
below @xmath96 the ratio @xmath11 increases as @xmath66 , see fig .
[ frozenrxi ] , indicating that @xmath100 at @xmath9 .
these results imply that , at @xmath9 , the large-@xmath66 limit of the renormalized couplings is identical to that observed in models with continuous distributions ; in this case , as we discussed before , we predict @xmath51 , @xmath101 , and @xmath102 .
this equality should not be taken as an obvious fact .
for instance , the stiffness exponent is different in the two cases . as estimated from the onset of the @xmath20-independent behavior for @xmath26 , and estimates of @xmath65 and @xmath64 in the frozen region ( we indicate them by @xmath103 and @xmath104 ) .
results for @xmath105 and at the temperature @xmath106 , which is well within the frozen region for the lattice sizes considered .
the dotted lines are drawn to guide the eye . ] , where @xmath107 is the value of @xmath37 in the frozen region . ]
the presence of freezing for @xmath108 makes the study of the @xmath9 glassy critical behavior quite hard . indeed , in order to observe the glassy critical behavior in ising glass models with a discrete distribution
, one must approach @xmath9 by keeping @xmath109 for each lattice size .
this makes a standard fss analysis impossible .
indeed , in the fss limit a rg invariant quantity @xmath110 should scale as @xmath111 the condition @xmath109 implies that this scaling behavior can only observed for @xmath112 for @xmath29 , the ratio @xmath113 diverges and thus this makes the range of values of @xmath114 which are accessible smaller and smaller as @xmath66 increases .
this implies that the standard fss limit , @xmath26 , @xmath29 at fixed @xmath10 does not exist .
however , as we shall now discuss , one can still study fss if one uses the ratio @xmath11 as basic fss variable , i.e. , if one considers the scaling form @xmath115 usually , expressions ( [ fss1 ] ) and ( [ fss2 ] ) are equivalent .
this is not the case here : only the fss scaling form ( [ fss2 ] ) holds in the presence of freezing . as is clear from figs .
[ rvst0.8 ] and [ rvst0.5 ] , the ratio @xmath11 at fixed @xmath66 increases as @xmath20 decreases .
hence , the condition @xmath109 translates into @xmath116 where @xmath107 is the value of @xmath37 in the frozen region . for @xmath117 @xmath118 diverges and thus , by increasing @xmath66 , one has access to the whole fss region .
thus , in the presence of freezing fss can still be used but only in the form ( [ fss2 ] ) . since @xmath119 does not appear in eq .
( [ fss2 ] ) , fss can not be used to determine @xmath119 ( this exponent can only be determined by using infinite - volume data ) , though it can still be used to check universality .
note that the presence of freezing and the limitations in the use of fss are always expected in models with a @xmath9 transition and discrete hamiltonian spectrum .
in particular , these phenomena should also be considered in the 3d diluted @xmath0 ising model close to the percolation point , where the glassy transition temperature vanishes @xcite . and @xmath65 versus @xmath11 for @xmath4 and @xmath3 . we only plot data satisfying @xmath120 for clarity . ,
title="fig : " ] and @xmath65 versus @xmath11 for @xmath4 and @xmath3 .
we only plot data satisfying @xmath120 for clarity .
, title="fig : " ] versus @xmath11 for @xmath4 .
we plot data for all available values of @xmath66 .
below , we only show the results for @xmath121 .
, title="fig : " ] versus @xmath11 for @xmath4 .
we plot data for all available values of @xmath66 .
below , we only show the results for @xmath121 .
, title="fig : " ] in order to verify universality , we consider the quartic cumulants @xmath64 and @xmath65 , which should scale according to eq .
( [ fss2 ] ) .
the function @xmath122 should be universal , hence @xmath8-independent .
universality is nicely supported by the data shown in fig .
[ scaling ] .
as expected , we find that @xmath123 and @xmath51 for @xmath124 , and @xmath125 and @xmath126 for @xmath102 . all data for @xmath64 fall onto a single curve with small scaling corrections , supporting the existence of the finite - size scaling limit in terms of @xmath11 , as discussed in the previous section .
the convergence to a single curve is also clear in the case of @xmath65 , although corrections are quite evident .
note that , in the region around the peak , for @xmath127 , the data at @xmath4 and @xmath3 converge from opposite sides . in these universality checks
there are no free parameters to be adjusted , and thus these comparisons provide strong support to the hypothesis that these models belong to the same universality class .
an analogous universality check was performed in ref .
@xcite . by studying the fss behavior of @xmath64 versus @xmath11
, it provided numerical evidence that the @xmath0 ising model with a bimodal distribution ( our model with @xmath3 ) and that with a gaussian distribution of the couplings belong to the same universality class .
it is interesting to note that while the data for @xmath64 scale nicely up to @xmath128 , for @xmath65 significant deviations are observed at smaller values of @xmath11 , see fig .
[ scaling0p8 ] . for @xmath4 and @xmath129 the data show significant deviations close to the peak , then approach the common curve and then show again a significant deviation the data turn up for @xmath130 .
a similar phenomenon occurs for @xmath131 . for @xmath132 deviations close to the peak
are quite small , but again the data begin to turn up as @xmath133 .
for @xmath134 fss holds quite nicely , at least up to @xmath135 .
the value @xmath136 marks the onset of the crossover region between the critical regime where fss holds and the freezing regime that sets in at @xmath137 . note that @xmath137 is significantly larger than @xmath136 , indicating that the breaking of fss occurs before freezing .
for @xmath3 the conclusions are similar , although , for a given @xmath66 , @xmath136 is significantly smaller than the corresponding value for @xmath4 . for instance , for @xmath132
, we have @xmath138 for @xmath3 and @xmath139 for @xmath4 .
this clearly reflects the fact that @xmath107 for @xmath3 is smaller than for @xmath4 , see fig .
[ frozenrxi ] . finally , we investigated the critical behavior of the overlap susceptibility .
as discussed in ref .
@xcite it should behave in the fss limit as @xmath140 where @xmath141 is an analytic function of @xmath20 which is related to the overlap - magnetic scaling field . in order to determine @xmath48
we have performed fits of the data with @xmath4 to @xmath142 where @xmath143 and @xmath144 are polynomials in @xmath145 with @xmath146 . to avoid any bias from the presence of the freezing region , we have only used the data satisfying @xmath147 , where , for @xmath148
, @xmath136 is the value determined before from the analysis of @xmath65 .
for @xmath149 we used somewhat arbitrarily @xmath150 , while for @xmath151 we used all our data which in any case satisfy @xmath152 . to identify scaling corrections we only considered data satisfying @xmath153 and @xmath154 for several values of @xmath155 and @xmath156 .
the results depend strongly on these parameters .
for @xmath157 we obtain @xmath158 , 0.33(3 ) for @xmath159 . for @xmath160
, we obtain @xmath161 for @xmath162 .
apparently the results always decrease as @xmath155 decreases and @xmath156 increases .
it is impossible to estimate reliably @xmath48 from these results and thus we only quote an upper bound : @xmath163 in principle , the same analysis can be performed for @xmath3 . however , in this case the estimates of @xmath136 are smaller , so that fewer data can be used in the fit . in practice
, no estimates of @xmath48 can be obtained .
as we mentioned at the beginning there are strong theoretical reasons to expect @xmath164 .
thus , we tried to verify whether our results are consistent with this hypothesis . for this purpose
we considered the data with @xmath165 and @xmath166 and we fitted them to eq .
( [ fitchi ] ) setting @xmath164 .
we find that the data satisfying @xmath147 are reasonably described by ansatz ( [ ansatz - chi ] ) . in fig .
[ plot - chiresc ] we report @xmath167 where @xmath168 is the polynomial determined in the fit ( [ fitchi ] ) .
the agreement is quite good up to @xmath169 .
it should be noted that , if we also include data with @xmath170 , the fits become much less dependent on @xmath155 and @xmath156 , exclude @xmath171 , and give the estimate @xmath172 : apparently the data that belong to the region @xmath173 are well described by eq .
( [ ansatz - chi ] ) with @xmath174 , while they can not be fitted by taking @xmath164 .
this is evident from fig .
[ plot - chiresc2 ] where we report @xmath175 $ ] : in this figure essentially all data fall on a single rescaled curve .
note that the quality of the collapse is only marginally better than that given in fig .
[ plot - chiresc ] the @xmath176 of the fit ( dof is the number of degrees of freedom of the fit ) is similar in the two cases although this is not evident from the figures since they have a completely different vertical scale .
defined in eq .
( [ chiresc ] ) for @xmath4 and @xmath164 . ]
defined in eq .
( [ chiresc ] ) for @xmath4 and @xmath174 . ] as a final check we verify if the data at @xmath177 are also consistent with @xmath164 .
the results are reported in fig .
[ plot - chiresc3 ] .
also in this case the data with @xmath178 are consistent with @xmath164 . in the plot
we have also multiplied @xmath179 by a constant in such a way that @xmath179 assumes the value @xmath180 for @xmath181 as it does for @xmath182 . with this choice
the curves for @xmath179 should be the same for both values of @xmath8 .
as it can be seen the shape of the two curves is indeed the same .
quantitatively , the two curves are the same up to @xmath183 , while they differ significantly for @xmath184 .
this is not surprising . as we have explained , for @xmath3 , the data such that @xmath185 are probably already in the crossover region before the onset of freezing .
defined in eq .
( [ chiresc ] ) for @xmath3 and @xmath164 . ]
the magnetic variables do not become critical in limit @xmath26 .
indeed , the magnetic susceptibility @xmath54 and second moment correlation length @xmath55 are finite in the limit @xmath26 .
for @xmath3 we have @xmath57 and @xmath58 for any @xmath20 . for other values of @xmath8 ,
they converge to nonuniversal values such that @xmath186 and @xmath187 . for @xmath4
we find @xmath188 and @xmath189 in the limit @xmath26 .
we also consider the four - point magnetic susceptibility @xmath59 which should scale as @xmath36 in the critical limit . in fig .
[ g4 ] we plot the ratio @xmath190 , where @xmath191 .
this quantity shows smaller scaling corrections than @xmath192 and clearly converges to an @xmath66 independent constant .
the asymptotic behavior sets in for @xmath193 , before the freezing region @xmath194 . at @xmath4 . ]
we now investigate the behavior of the overlap correlations along the low - temperature pf transition line , see fig .
[ phasedia ] , from the mnp to the @xmath9 axis . the critical behavior of the magnetic correlations was numerically studied in ref .
@xcite by varying @xmath8 for two values of @xmath20 : @xmath195 and @xmath196 .
it was found that the critical behavior was universal , controlled by a single strong - disorder fixed point , with critical exponents @xmath197 and @xmath198 .
the ferromagnetic @xmath9 transition point at @xmath199 , where the low - temperature pf transition line ends , is a multicritical glassy point ( mgp ) , because it is connected to three phases and it is the intersection of two different transition lines , the pf line at @xmath200 and the glassy line at @xmath9 . at @xmath9 the critical point at @xmath201 separates a ferromagnetic phase from a @xmath9 glassy phase , while for @xmath202 the transition line separates a ferromagnetic from a paramagnetic phase .
therefore , on general grounds , the critical behavior at @xmath9 and @xmath201 should differ both from that observed along the pf line and that observed along the glassy line @xmath9 , @xmath203 , unless the magnetic and glassy critical modes are effectively decoupled .
such a decoupling is apparently supported by the numerical results of refs .
@xcite . indeed , the estimates of magnetic critical exponents at @xmath9 are quite close and substantially consistent with those found along the pf transition line below the mnp .
all results are therefore consistent with a single magnetic fixed point that controls the magnetic critical behavior both at @xmath202 and at @xmath9 .
the analysis of the overlap correlation functions also supports the decoupling of the critical modes .
indeed , the arguments we gave in sec .
[ sec : mc ] on the behavior of the ( overlap ) renormalized couplings and correlation functions should hold at @xmath9 for any value of @xmath8 hence , in the ferromagnetic phase , at the mgp as well as along the glassy transition line since they only rely on the assumption of a nondegenerate ground state . therefore , also at the @xmath9 mgp we expect @xmath204 , @xmath205 , @xmath206 , and @xmath164 .
the behavior of the overlap correlations is therefore identical for @xmath203 , for @xmath207 , and at the mgp , in agreement with the decoupling scenario .
we wish now to understand the critical behavior of the overlap correlations along the pf line . in order to investigate this issue
, we perform mc simulations at two critical points along the low - temperature pf line , at @xmath208 $ ] and @xmath209 $ ] , as determined in ref .
@xcite , for lattice sizes up to @xmath72 .
the fss analysis of the data of the overlap susceptibility @xmath36 shows that @xmath210 with quite small but nonzero values of @xmath48 .
fits of the data satisfying @xmath120 which take into account the nonanalytic scaling corrections give @xmath211 at @xmath195 and @xmath212 at @xmath196 .
note that these values are much smaller than the pure ising value @xmath213 ( which is simply twice the value of the magnetic exponent @xmath214 ) holding along the pf line from the pure ising point to the mnp , and also much smaller than the value @xmath215 at the mnp point .
the fss analyses of the renormalized couplings @xmath11 , @xmath64 and @xmath65 lead to apparently @xmath20-dependent critical values : @xmath216 , @xmath217 , @xmath218 at @xmath195 , and @xmath219 , @xmath220 , @xmath221 at @xmath196 ( where the errors are essentially due to the uncertainty on @xmath222 and on the scaling correction exponent ) . note that these values are very close to the values at the @xmath9 mgp ; still we consider unlikely that the overlap behavior is the same as that occurring at the mgp , mainly because this would require @xmath223 along the whole pf line between the mnp and the mgp . indeed
, the condition @xmath223 is quite unlikely for a finite - temperature transition .
these results can be better explained by a @xmath20-dependent asymptotic critical behavior which , with decreasing @xmath20 , approaches the @xmath9 glassy behavior characterized by the values @xmath49 , @xmath97 , @xmath98 , @xmath223 .
in this paper we consider the two - dimensional @xmath0 ising model , focusing mainly on the @xmath9 glassy transition occurring for @xmath224 .
the main results are the following .
* we first discuss the freezing phenomenon that occurs on any finite lattice at sufficiently low temperatures .
we investigate the behavior of several quantities in this regime , verifying explicitly the expected logarithmic dependence on the lattice size @xmath66 .
we also show that the presence of this regime makes it impossible to use the standard form of fss : the fss limit @xmath26 , @xmath117 at fixed @xmath10 does not exist .
fss can be formulated only if one considers @xmath11 as basic fss variable .
* we study the fss behavior of the renormalized couplings @xmath225 and @xmath65 for @xmath4 and @xmath3 as a function of @xmath11 .
we find that they have the same fss curves , a clear indication that the critical behavior for @xmath4 and @xmath3 is the same .
this allows us to conjecture that the critical behavior is independent of @xmath8 in the interval @xmath224 , and therefore , together with the results of ref .
@xcite , that there exists a single universality class for 2d ising glassy transitions .
we also investigate in detail the critical behavior of the overlap susceptibility , showing that the numerical data are consistent with @xmath164 , if one discards data that are close to the region where freezing occurs . * finally , we discuss the critical behavior of the overlap variables along the pf line .
an analysis of the numerical data available for the @xmath9 mgp indicates that at this point glassy and magnetic modes are decoupled .
for @xmath200 we observe an apparent @xmath20-dependent critical behavior .
note that a similar phenomenon was also observed in the xy model with random shifts @xcite : along the critical line that starts at the xy pure point and ends at the nishimori multicritical point , magnetic quantities show a universal behavior while overlap variables show a disorder - dependent critical behavior
. the most accurate estimates of @xmath119 have been obtained by computing the stiffness exponent @xmath227 at @xmath9 in models with continuous distributions .
we mention @xmath228 ( ref .
@xcite ) , @xmath229 ( ref .
@xcite ) , @xmath230 ( ref .
@xcite ) , @xmath231 ( ref .
@xcite ) obtained by using the ising glass model with a gaussian distribution for the couplings , and @xmath232 [ f. liers , j. lukic , e. marinari , a. pelissetto , and e. vicari , phys . rev .
b * 76 * , 174423 ( 2007 ) ] obtained in the random - anisotropy model in the strong - anisotropy limit , whose glassy critical behavior is in the same universality class . the stiffness exponent is instead not related to @xmath119 in the case of discrete distributions @xcite .
finite - temperature estimates of @xmath119 are reported in refs .
@xcite . c. j. geyer , in _ computer science and statistics : proc .
of the 23rd symposium on the interface _ , edited by e. m. keramidas ( interface foundation , fairfax station , 1991 ) , p. 156 ; k. hukushima , k. nemoto , j. phys .
* 65 * , 1604 ( 1996 ) ; d. j. earl , m. w. deem , phys .
* 7 * , 3910 ( 2005 ) . | we investigate the zero - temperature glassy transitions in the square - lattice @xmath0 ising model , with bond distribution @xmath1 ; @xmath2 and @xmath3 correspond to the pure ising model and to the ising spin glass with symmetric bimodal distribution , respectively .
we present finite - temperature monte carlo simulations at @xmath4 , which is close to the low - temperature paramagnetic - ferromagnetic transition line located at @xmath5 , and at @xmath3 .
their comparison provides a strong evidence that the glassy critical behavior that occurs for @xmath6 , @xmath7 , is universal , i.e. , independent of @xmath8 .
moreover , we show that glassy and magnetic modes are not coupled at the multicritical zero - temperature point where the paramagnetic - ferromagnetic transition line and the @xmath9 glassy transition line meet . on the theoretical side
we discuss the validity of finite - size scaling in glassy systems with a zero - temperature transition and a discrete hamiltonian spectrum .
because of a freezing phenomenon which occurs in a finite volume at sufficiently low temperatures , the standard finite - size scaling limit in terms of @xmath10 does not exist : the renormalization - group invariant quantity @xmath11 should be used instead as basic variable . |
the derivation of the conditional mean and variance of the number of photons in the signal field conditioned on @xmath30 photons being detected in the idler field is described in this appendix . to calculate the conditional mean and variance
, we need to evaluate the sums defined by @xmath72 where @xmath73 and @xmath74 are the probability distributions given in the text in eq .
( [ eq - pngn ] ) and eq .
( [ eq - pmu ] ) respectively .
the simplest case is that of @xmath75 . performing the summation over @xmath4 by recognizing the negative binomial series
, we find @xmath76 where we have defined @xmath77 and @xmath78 for simplicity .
next , using the identity for gauss hypergeometric series , @xmath79 , where @xmath80 is the beta function , we express @xmath81 as @xmath82 where @xmath83 is shorthand for the @xmath84th - order partial derivative with respect to @xmath85 .
we follow an analogous procedure for the case when @xmath86 , where the primary difference is that the summations over @xmath28 are now derivatives of the negative binomial series .
the resulting sum @xmath87 is given by @xmath88,\ ] ] with @xmath89 .
the sum @xmath90 is similarly given by @xmath91 , \end{split}\ ] ] where @xmath92 is defined for simplicity . the conditional mean and variance can be expressed in terms of these sums as @xmath93 and @xmath94 . substituting into these equations and simplifying , we find the general form for the conditional mean and variance to be @xmath95 + where @xmath96 . by setting @xmath97
it is not difficult to recover the single - mode results of eq .
( [ eq - condmean ] ) and eq .
( [ eq - condvar ] ) shown in the text .
furthermore , the single - mode and multimode expressions for @xmath53 can be easily found from these results . | a technique for conditionally creating single- or multimode photon - number states is analyzed using bayesian theory .
we consider the heralded @xmath0-photon states created from the photons produced by an unseeded optical parametric amplifier when the heralding detector is the time - multiplexed photon - number - resolving detector recently demonstrated by fitch , _ et al_. [ phys .
rev .
a * 68 * , 043814 ( 2003 ) . ] and simultaneously by achilles , _
et al_. [ opt
. lett .
* 28 * , 2387 ( 2003 ) . ] .
we find that even with significant loss in the heralding detector , fields with sub - poissonian photon - number distributions can be created .
we also show that heralded multimode fields created using this technique are more robust against detector loss than are single - mode fields .
a fock state of the electromagnetic field contains a definite number of photons and hence displays no intensity fluctuations .
in contrast , the distribution of the intensities of a classical field of light has a variance at least as large as its mean @xcite .
thus , fock states and other states with sub - poissonian photon - number statistics are of interest not only because of their fundamentally nonclassical nature but also because of their potential to make radiometric measurements with greater accuracy than possible using classical states of light .
research in fock state generation is also motivated by its applications in quantum information science @xcite including the areas of quantum cryptography @xcite and quantum computing @xcite .
unfortunately , it can be experimentally challenging to produce light fields that contain definite numbers of photons .
several theoretical proposals aimed at creating fock states `` on demand '' have been offered @xcite , but they involve trapped atoms in high - q cavities and to the best of our knowledge only single - photon states have been created in the laboratory @xcite .
several stochastic sources of one- and two - photon states have been experimentally demonstrated using various processes including atomic and molecular fluorescence @xcite , coulomb blockade for electrons @xcite and action in a micromaser @xcite .
in addition , a particularly simple method of generating single - photon states using the photon pairs created in the process of spontaneous parametric down - conversion was first demonstrated by hong and mandel @xcite . in their experiment
, they used the single - photon detection of one photon of the pair to `` herald '' the presence of the other photon of the pair .
this heralding scheme is generalizable to create multiphoton states of light . in an unseeded optical parametric amplifier ( opa ) , pump photons interact nonlinearly with a noncentrosymmetric crystal to produce multiple pairs of less energetic photons termed the signal and idler through the process of parametric down - conversion ( pdc ) .
in general , the number of photons found in the signal and idler fields is random and known to obey thermal statistics , which can be characterized by the gain @xmath1 of the opa .
but since the emission of every signal photon requires the simultaneous emission of an idler photon , if one knows the number of photons in the idler field then the number in the signal field is determined with complete certainty .
thus , in principle , by allowing transmission of the signal field only when the idler field is found to contain @xmath2 photons , the signal field will be guaranteed to also contain exactly @xmath2 photons .
this idea was considered by holmes _
et al . _
@xcite for ideal photon - number - resolving detection .
however , the photon - number - resolving detection needed to create heralded multiphoton states is not trivial to implement and is an area of active research .
photon - number - resolving detection has been achieved using superconducting transition - edge sensors @xcite , visible light counters @xcite , charge integration photon detectors @xcite , and by time - multiplexed detection involving only commercially available apds @xcite . in the present work ,
we show that , in particular , the photon - number resolution achieved using time - multiplexed detectors ( tmds ) , as first conceived by fitch _ et al . _
@xcite and independently by achilles _
et al . _
@xcite , is well - suited for creating heralded fields with near - definite numbers of photons even in the presence of significant detector loss .
this technique has been demonstrated experimentally for two - photon states @xcite but has not yet been analyzed in a comprehensive fashion for states containing greater numbers of photons .
( a ) scheme to produce heralded fields containing a definite number of photons . an unseeded optical parametric amplifier ( opa ) produces photon pairs through the process of parametric down - conversion ( pdc ) .
when a time - multiplexed photon - number - resolving detector ( shown in ( b ) ) detects @xmath3 photons in the idler field , it heralds the presence of @xmath4 photons in the signal field with probability @xmath5 . ]
a schematic of the proposed heralding technique is shown in fig .
[ fig - setup ] . as mentioned above , the source of photon pairs is the output of an unseeded opa and the heralding detector is a tmd .
the tmd provides photon - number resolution by first randomly splitting the incoming state into many pathways that are separated in time from one another , i.e. they are separated into time bins .
the number of photons in the incoming field can be estimated to be equal to the total number of photons detected by two avalanche photodiodes ( apds ) .
any errors in this estimation will introduce uncertainty into the number of photons contained in the heralded state .
undercounting is the main source of errors in tmds .
the two primary mechanisms for undercounting errors are losses and inadequate splitting of the incoming beam .
losses in the detection system stemming both from the beam splitters and from inefficiencies in detectors themselves can be quite significant and must be included in the analysis .
inadequate splitting occurs because of the finite number of times the incoming beam is split .
hence , a chance exists that two or more photons will occupy the same time bin resulting in an undercounting error at the apds , which can perform only on / off detection .
more quantitatively , a system containing @xmath6 beam splitters will have @xmath7 different time bins .
if the number of incoming photons @xmath0 is much less than @xmath8 , the probability of undercounting due to imperfect splitting will be small .
however , when @xmath0 is comparable to @xmath8 , we must account for splitting errors . using a standard urn model @xcite
, we can calculate the conditional probability of detecting @xmath2 photons given that the incoming field has @xmath0 photons in it .
this probability is given by the expression @xmath9^n,\ ] ] where @xmath10 is the single - photon detection efficiency .
it is not difficult to verify that in the limit @xmath11 and @xmath12 detection errors due to imperfect splitting are negligible and that @xmath13 , which is the expected distribution for an ideal photon - number - resolving detector subject to a loss of @xmath14 per photon .
we note that detector dark counts are a source of overcounting errors .
however , in the present analysis , we do not include the effect of detector dark counts , since often they can be rendered neglible through detector gating or by use of detectors with intrinsically low dark count rates ( commercially available apd dark count rates can be as low as 5 dark counts per second ) .
the response of the detection system is shown in fig .
[ fig - pnn ] for a system comprised of five beam splitters ( m=32 ) for three different scenarios . for no loss in the tmd ( @xmath15 ) and few enough incident photons
, the detector will accurately measure the number of photons in the field , i.e. the initial slope of the mean of the probability distribution is linear with unity slope .
once the number of incident photons increases and becomes comparable to the maximum number of photons @xmath8 the system can count , undercounting will become very likely and the detection system will saturate . as the loss in the system increases ( @xmath10 decreases ) ,
two effects are seen : ( 1 ) the initial slope decreases and is given roughly by @xmath10 and ( 2 ) the error in the estimate increases as seen by the increase in the vertical spread of the distribution in the plot .
regions where there is a significant probability of detecting @xmath2 photons conditioned on @xmath0 photons being incident on the tmd for three different single - photon detection efficiencies ( @xmath16 1.0 , 0.66 , 0.33 ) .
the mean number of photons detected given @xmath0 photons input into the detection system are plotted using the solid ( @xmath17 ) , dotted ( @xmath18 ) and dashed ( @xmath19 ) lines .
the shaded regions are within one standard deviation of the means . ] as we already mentioned , in a heralding system any error in the detection of the number of photons in the heralding field will directly lead to an uncertainty in the number of the photons in the heralded field .
if this uncertainty , as determined by the variance of the number photons estimated to be in the heralding field , becomes larger than the mean number of photons in the heralded field , the state can , in principle , be produced by classical means .
however , since the source of the photon pairs is an opa , we can use this _ a priori _ information about its photon emission statistics to improve our estimation through the use of bayesian theory .
the probability that an opa characterized by gain @xmath1 will emit @xmath20 pairs of photons is given by @xmath21 , where @xmath22 is the state containing @xmath20 photons in both the signal and idler fields and @xmath23 is the state emitted by the opa , which is given by @xmath24 @xcite . as a notational device
, we will identify the signal field with the heralded field and the idler field with the heralding field .
it can be shown that the resulting photon - number distribution for either the signal or idler field obey bose - einstein statistics and is given by @xmath25 we can calculate the resulting mixed state of the signal field given that @xmath3 photons are detected in the idler field .
the density matrix is given by @xmath26 where @xmath27 is a fock state of the signal field with @xmath28 photons in it .
the function @xmath29 is the probability that the signal field has @xmath28 photons in it given that @xmath30 photons are detected in the idler field . according to bayes theorem , it is given by the expression @xmath31
we estimate the number of photons in the signal field using maximum - likelihood estimation .
the maximum - likelihood estimate of the number of photons in the signal field given @xmath30 photons detected in the idler field is found using @xmath32 where the @xmath33 function returns the value of @xmath34 that maximizes @xmath29 .
comparison of the photon - number statistics for a typical heralded field with @xmath35 ( in blue ) and a poisson - distributed field with a mean photon number of 5 ( in gray ) .
the heralded field , for the case in which @xmath18 and @xmath36 , is conditioned on the detection of 4 photons in the idler field by the tmd . ] in fig .
[ fig - psig4 ] we show the photon - number statistics for a heralded field estimated to have @xmath35 photons in it and compare it to a classical coherent state that has poisson statistics with a mean number of photons equal to 5 . from the figure it is clear that the photon - number distribution of the heralded state is much narrower than the distribution for the comparable classical coherent state .
the width of the heralded field distribution is exactly the error in the maximum - likelihood estimate and is found using the mean squared error , i.e. @xmath37 .
figure [ fig - relerr ] compares the ratio of the spreads of the heralded fields to their classical counterparts for different numbers of photons in the heralding field for various values of the single - photon detection efficiency @xmath10 and the opa gain @xmath1 .
when this ratio is equal to unity , the heralded field has as much uncertainty in its number of photons as a poissonian field .
the smaller the ratio is , the closer the heralded field is to a fock state .
the quality of the heralded state degrades as the loss increases and the gain of the opa increases , as can be seen by noting that the ratio increases in these cases .
uncertainty in the number of photons in a heralded field when @xmath30 photons are detected in the heralding field normalized to the uncertainty in a comparable poisson - distributed field with mean number of photons @xmath38 for various detection efficiencies ( @xmath39 ) and opa gains ( @xmath40 ) . ] to gain further insight into the behavior of the heralded state , we consider the mean number of photons in the heralded field .
although the mean number of photons can not be an estimator of the number of photons in the heralded field , since it can take non - integer values , it is useful parameter that characterizes the field and can be analytically calculated .
the mean number of photons in the signal field conditioned on @xmath30 photons being detected in the idler field can be expressed in the following remarkably simple form ( see appendix ) @xmath41 similarly , the conditional variance of the number of photons in the signal field can be expressed by @xmath42 where in both equations we have introduced the quantity @xmath43\right\}^{-1}$ ] for notational convenience .
an ideal system would produce fock states with definite photon number so that @xmath44 would vanish .
we note that @xmath45 can not vanish since that requires @xmath10 to vanish , which is not a physically meaningful situation .
thus , @xmath44 goes to zero only in the limit in which @xmath46 .
also in this limit the conditional mean @xmath47 , which is consistent with the expected behavior of an ideal heralding system .
two cases exist in which @xmath45 approaches unity .
the first case occurs in the low gain limit .
irrespective of the system loss @xmath10 , @xmath45 can be made arbitrarily close to one for small enough gain since @xmath48 .
intuitively , this limit is easily understood . when the gain is small enough
, it is much more likely for the source to produce @xmath28 photon pairs and the detector to count @xmath30 of them than it is for the source to produce @xmath49 pairs and the detector to still detect only @xmath30 .
more quantitatively , there exists a gain such that the ratio @xmath50 .
this statement is easily verified by noting that @xmath51 and can be made large for small enough @xmath1 .
the second case in which @xmath45 goes to one occurs when @xmath52 and @xmath17 . in this limit
the tmd behaves like an ideal photon number resolving detector and the system achieves perfect heralding .
although these limiting cases are useful in providing insight into the behavior of the system , they are often not viable experimentally .
if the gain is too low , the rate at which the heralded states are produced becomes impractically small .
furthermore in many cases @xmath8 can not be too large , if for no other reason than that the loss incurred by adding splitters can become excessive . to treat the problem more generally , we calculate mandel s @xmath53 parameter defined by @xmath54 @xcite .
mandel s
@xmath53 parameter is useful for characterizing the nature of the photon - number distribution of a field .
the @xmath53 parameter is bounded below by @xmath55 . in the limit in which @xmath56 ,
the heralded field is in a fock state . for @xmath57 ,
the field has poissonian photon - number statistics ; for positive and negative @xmath53 the field has super- and sub - poissonian statistics respectively .
we use this parameter as a merit function describing the quality of the heralded states produced using the technique .
mandel s @xmath53 parameter versus the gain of the opa and the single - photon detection efficiency of the time - multiplexed detector ( tmd ) for the case where @xmath58 , i.e. , the heralded field is a single - mode field estimated to contain 5 photons .
the colored regions are labeled with the number of detected idler photons needed to herald the presence of 5 photons in the signal field .
the white contours are lines of constant @xmath59 , the probability of creating the heralded state .
the tmd is assumed to contain 5 beam splitters ( m=32 ) ] in practice , it is useful to know the tmd and opa parameters necessary to produce a state that contains an estimated @xmath38 photons and a certain value of the @xmath53 parameter . to this end , we calculate @xmath53 at constant @xmath38 versus the parameters @xmath10 and @xmath1 .
figure [ fig - q5 ] shows the result of the calculation for a heralding detector comprised of five beam splitters for the case when @xmath58 . to calculate @xmath53 for constant @xmath38
it is necessary to invert eq ( [ eq - nml ] ) to find the appropriate value of @xmath30 to be substituted into eqs ( [ eq - condmean ] ) and ( [ eq - condvar ] ) . for certain values of the gain and detector efficiencies ,
eq ( [ eq - nml ] ) can not be inverted , i.e. there is no integer number of detected photons @xmath3 that correspond to a signal field in which @xmath58 . in these instances , a signal field estimated to contain five photons is not possible to produce and is marked by the gray regions in the plot .
also , the gray bands divide the space into regions corresponding to the number of detected idler photons that herald the presence of 5 photons in the signal field .
the figure confirms that the losses as well as the opa gain should be kept as small as possible .
in addition , the efficiency with which the heralded states are created is important .
the white contour lines overlaid on the plot show the probability with which the heralded state is successfully created , which is denoted by @xmath60 .
this probability is calculated from the sum @xmath61 where @xmath3 is obtained from the inversion of eq .
( [ eq - nml ] ) for different values of @xmath1 and @xmath10 .
the discrete jumps seen in the contour lines are due to the fact the @xmath3 can take only integer values .
we see that it is possible to create nearly ideal fock states even with significant detector loss ; however , the efficiency of the creation of the heralded states is sacrificed . up to this point , we have considered only single - mode states produced by an opa .
although the single - mode limit can be reached , e.g. by both spatially and frequency filtering @xcite , the photons emitted by a free - space opa are generally multimodal in nature and can be approximately described by a multimode thermal distribution @xcite .
the probability that the opa will emit @xmath20 pairs of photons distributed among any @xmath62 modes is given by @xmath63 where @xmath1 is the parameter characterizing the gain per mode of the opa .
we have assumed that each mode is equally likely to be occupied .
the multimode thermal distribution has a variance of @xmath64 .
since when @xmath65 is held constant , the variance decreases with increasing number of modes , it is reasonable to expect that we may be able to create multimode fields containing near - definite numbers of photons under more relaxed conditions than by restricting ourselves to single - mode states .
mandel s @xmath53 parameter versus gain of the opa and the single - photon detection efficiency of the tmd for the case where @xmath35 and @xmath66 , i.e. the signal field is estimated to contain 5 photons spread between 5 modes .
the colored regions are labeled with the number of detected idler photons needed to herald the presence of 5 photons in the signal field .
the white contours are lines of constant @xmath67 , the probability of creating the heralding state . ] in fig .
[ fig - q5a ] we plot the @xmath53 parameter under the same conditions as in fig .
[ fig - q5 ] but now assuming that we are interested in the multimode output of the opa ( see appendix for calculation details ) .
we take @xmath68 as an example , which is typical value for a free space opa @xcite . when comparing fig .
[ fig - q5 ] and fig . [ fig - q5a ] , we see that the multimode case is more tolerant of loss in the heralding detector than the single mode case . for example , if we require a heralded state preparation rate of @xmath69 , the heralded output of the opa is sub - poissonian so long as @xmath70 , but in the single - mode case we have the more stringent requirement that @xmath71 . in conclusion
, we have shown that it is possible to create multiphoton heralded states with highly non - classical photon - number distributions using a tmd as the heralding detector .
we calculated mandel s @xmath53 parameter as a function of the detector loss and the opa gain to characterize the quality of the heralded state . even in the presence of significant loss
, we found that it is possible to produce states with highly sub - poissonian photon - number distributions at reasonable efficiencies .
in addition , for situations where a single - mode output is not required , we showed that the full multimode output of an opa can provide less uncertainty in the number of photons in the field than the single - mode output .
since the tmds involve only commercially available apds and fibers , the technique we have presented here can be a particularly simple way to create states of light containing nearly definite numbers of photons .
the authors would like to thank christine silberhorn and gerd leuchs for helpful discussions .
this work was supported by the us army research office through a muri grant .
kwc thanks the support from the croucher foundation . |
we consider a tight - binding hamiltonian to model the graphene - adatom system .
the graphene part of the hamiltonian , denoted @xmath6 , is written in terms of the creation and annihilation operators for electrons of spin @xmath7 on a site in the @xmath8 primitive cell @xcite . denoting the creation ( annihilation ) operators for the @xmath9 and @xmath10 sub - lattices by @xmath11 ( @xmath12 ) and @xmath13 ( @xmath14 ) , we write @xmath6 for nearest - neighbor hopping on graphene as @xmath15 , where @xmath16 and @xmath17 .
we note that here and in subsequent equations the notation of ref . is used . in eq .
( [ eq : ccham ] ) , @xmath18 $ ] , where @xmath19 is the nearest - neighbor bond length for graphene , and @xmath20 is the hopping energy between nearest neighbor carbon atoms ( for graphene @xmath21 ) @xcite . the extra part of the hamiltonian due to an adatom at a @xmath22 site may be written in real space as @xmath23 where @xmath24 and @xmath25 are creation and annihilation operators for an electron of spin @xmath7 at the site of the adatom , @xmath26 is the on - site energy of an electron on that site ( relative to the dirac point of the pure graphene band structure ) , and @xmath27 is the energy for an electron to hop between the adatom and the carbon atom at the site @xmath28 of the @xmath9 sub - lattice . in terms of bloch eigenstates of @xmath6 , @xmath29.\end{aligned}\ ] ] here @xmath30 ( i=1,2 ) is the annihilation operator for a bloch electron of wave vector @xmath31 , spin @xmath7 , and within the @xmath32 band .
@xmath33 is a phase angle given in ref . and for a hydrogen adatom we take @xmath34 and @xmath35 @xcite . to calculate the magnetic moment
we add a hubbard term @xmath36 , which acts only on the adatom . within the mean - field approximation @xmath36
is given by @xmath37.\ ] ] where @xmath38 is the average number of electrons with spin @xmath7 on a @xmath22 site of the @xmath9 sub - lattice .
for a hydrogen adatom we take @xmath39 to be the difference between the ionization potential and the electron affinity , which gives @xmath40 @xcite .
the total mean - field hamiltonian consists of the sum of eqs .
( [ eq : ccham ] ) , ( [ himp ] ) , and ( [ mft_u ] ) , which is quadratic in the electron creation and annihilation operators and therefore readily diagonalized .
the total spin - dependent densities of states @xmath41 can then be calculated , given the values of @xmath38 .
the result is @xcite @xmath42\right)\right]_{z = e + i0^+},\ ] ] where @xmath43 is the density of states per spin and per primitive cell of pure graphene , @xmath44 is the number of primitive cells , and @xmath45 the corresponding green s function , as defined in ref . .
the quantities @xmath38 are obtained by integrating the local spin - dependent density of states on the adatom , @xmath46 , from the bottom of the valence band up to the fermi energy .
@xmath46 is , in turn , given by @xmath47 the total magnetic moment @xmath0 can be calculated as a function of the fermi energy @xmath1 from the expression @xmath48de,\ ] ] where @xmath49 is the bohr magneton .
using these results , we can numerically calculate @xmath50 for various choices of tight binding parameters .
this is done iteratively , for a given choice of @xmath1 , as follows .
first , we make initial guesses for @xmath38 .
next , we integrate the spin - dependent local density of states up to the fermi energy using eq .
( [ eq : local_dos ] ) and the initial guesses for @xmath38 .
this gives the next iteration of @xmath38 .
we iterate until the changes in @xmath38 in two successive cycles are less than @xmath51 . finally , using the converged values we calculate the total magnetic moment from eq .
( [ eq : magmom ] ) . ) . in each case , all the parameters but one are appropriate to h on graphene , as given in the text . in ( a ) each curve represents a different on - site energy @xmath26 ( given in ev in the legend ) . in ( b ) , each curve represents to a different hopping energy @xmath52 ( in ev ) , and in ( c ) each curve represents a different hubbard energy @xmath39 ( in ev ) .
the maximum calculated magnetic moment in each figure is noted within the text . ,
title="fig:",scaledwidth=80.0% ] ) . in each case , all the parameters but one are appropriate to h on graphene , as given in the text . in ( a )
each curve represents a different on - site energy @xmath26 ( given in ev in the legend ) . in ( b ) , each curve represents to a different hopping energy @xmath52 ( in ev ) , and in ( c ) each curve represents a different hubbard energy @xmath39 ( in ev ) .
the maximum calculated magnetic moment in each figure is noted within the text .
, title="fig:",scaledwidth=80.0% ] ) . in each case , all the parameters but one are appropriate to h on graphene , as given in the text . in (
a ) each curve represents a different on - site energy @xmath26 ( given in ev in the legend ) . in ( b ) , each curve represents to a different hopping energy @xmath52 ( in ev ) , and in ( c ) each curve represents a different hubbard energy @xmath39 ( in ev ) .
the maximum calculated magnetic moment in each figure is noted within the text .
, title="fig:",scaledwidth=80.0% ]
we have carried out these calculations as a function of @xmath1 for a variety of values of @xmath26 , @xmath53 , and @xmath39 . in every case , we find that the total magnetic moment is non - zero only when @xmath1 lies within a limited range .
we write this condition as @xmath54 where @xmath55 and @xmath56 are the lower and upper energies within which @xmath57 . within this range ,
the magnitude of @xmath0 is controlled by varying the fermi energy @xmath1 , typically by applying a gate voltage to the sample which interacts with the conduction electrons via the electric field effect @xcite .
if the graphene - adatom system is neutral and no voltage is applied , @xmath1 will be constrained to have a particular value controlled by the charge neutrality condition .
introducing a gate voltage will shift @xmath1 from this neutral value ( denoted @xmath58 ) and hence change the magnetic moment . as an illustration of this picture
, we show in fig .
[ fig1 ] the total magnetic moment @xmath50 under various conditions . in each case , we assume all parameters but one are those thought to describe h on graphene , and vary the remaining parameter @xcite .
we assume a single h adatom is placed on a graphene sheet containing @xmath59 carbon primitive cells , giving 1 h atom per 1000 c atoms . in fig .
[ fig1](a ) , we assume that @xmath20 and @xmath53 are those of the h - graphene system , while each curve represents a different value of the on - site energy @xmath26 .
we find that , when @xmath60 and we allow @xmath26 to become much less then @xmath61 that the onset energy @xmath62 , whereas the upper cutoff energy @xmath56 shows only a minimal dependence on @xmath1 . in fig .
[ fig1](b ) we plot @xmath50 versus @xmath1 for several values of @xmath53 , with other parameters the same as the h - graphene system .
as @xmath53 is reduced , the upper energy cutoff , @xmath56 , also decreases , whereas the lower energy onset remains approximately unchanged at @xmath63 , independent of @xmath53 . in fig .
[ fig1](c ) , we plot @xmath50 versus @xmath1 for several values of the electron - electron energy @xmath39 , assuming the other parameters same as the h - graphene system .
as @xmath39 increases , so do both @xmath55 and @xmath56 . as @xmath64 , @xmath65
as well , i. e. , @xmath0 persists no matter how large @xmath1 in this case . for all parameters we have considered
, we find that the magnitude of the magnetic moment @xmath66 . in fig .
[ fig1](a ) .
the maximum value of @xmath67 , corresponding to @xmath68 and @xmath69 . in fig .
[ fig1](b ) the maximum value of @xmath70 for @xmath71 , while in fig . [ fig1]c a maximum value of @xmath72 for @xmath73 . in previous work
@xcite , it has been shown that @xmath74 , @xmath75 .
the numerical results of fig . [ fig1 ] can be qualitatively understood as follows .
if the system of graphene plus adatom has a net magnetic moment , then the partial densities of states @xmath76 and @xmath77 will be different .
the total magnetic moment is then obtained by integrating these two densities of states up to @xmath1 .
if @xmath1 lies below the bottom edge of the lower band , there will be no net magnetic moment .
the moment becomes non - zero at the energy @xmath78 when @xmath1 moves above the bottom of the lower band .
it reaches its maximum when @xmath1 lies somewhere between the peaks in @xmath76 and @xmath77 ( assuming the two bands overlap ) , and thereafter decreases until it becomes zero at @xmath79 , the energy above which both sub - bands are filled .
thus , for any choice of the parameters @xmath26 , @xmath53 , and @xmath39 , there should be a finite range of @xmath1 , within which @xmath57 .
of course , this description is an oversimplification because the self - consistently determined @xmath76 and @xmath77 depend on the quantities @xmath38 ( @xmath17 ) , which themselves depend on @xmath0 . however , even with the oversimplification , the qualitative description remains correct .
the maximum value of @xmath0 depends , in part , on how much the two sub - bands overlap .
if the overlap is large , maximum of @xmath0 will be small , everything else being equal , while a small overlap will tend to produce a larger @xmath0 .
another reason why @xmath0 is reduced below @xmath49 is that there is generally a large electron transfer from the adatom onto the graphene sheet@xcite .
we also note that the energy range , @xmath80 , where @xmath57 is approximately equal to the width of the extra density of states due to the adatom .
the fermi energy @xmath1 can be controlled experimentally in several ways .
one is to apply a suitable gate voltage @xmath81 , which raises or lowers @xmath1 by an amount @xmath82 , where @xmath83 is the magnitude of the electronic charge .
another is by chemical doping : if one adds or subtracts charge carriers to the graphene - adatom system by doping with suitable molecules , this will also raise or lower @xmath1 as was done in ref . .
one could also add a small number of vacancies in the graphene , as also done in ref . .
this will reduce the number of charge carriers and hence lower @xmath1 . of course , vacancies would also change the graphene density of states ; so the present calculations would have to be modified to treat this situation .
the model used here treats only the effects of adatoms on graphene , but it does qualitatively reproduce the upper energy cutoff found in the experiments of ref . .
for example , nair _ et al . _
@xcite found an upper cutoff of around @xmath84 , which we can approximately obtain by assuming an on - site energy @xmath85 , @xmath86 and @xmath87 .
however , our model does not account for the onset energy found in ref .
@xcite of @xmath88 since in our model the bottom edge of the lower band occurs at @xmath89 . in summary , by
using a simple tight - binding model of adatoms on graphene we are able to calculate the total magnetic moment of graphene with a small concentration adatoms as a function of @xmath1 .
the model is expected to apply to the case of h adatoms , but could also be applicable to other adatom species , characterized by different model parameters . the fermi energy @xmath1 can be controlled experimentally by a suitable gate voltage .
our results show that , for realistic tight - binding parameters ( @xmath90 ) , the magnetic moment can be switched off at a relatively low voltage ( @xmath91 ) , in rough agreement with the experiments of ref . .
these results are potentially of much interest since they suggest that the magnetic moment of graphene with adatoms can be electrically controlled .
this work was supported by the center for emerging materials at the ohio state university , an nsf mrsec ( grant no .
dmr0820414 ) .
we thank r. k. kawakami for helpful discussions .
99 d. pesin and a. h. macdonald , nat . maters . * 11 * 409 - 416 ( 2012 ) .
n. tombros , c. jozsa , m. popinciuc , h. t. jonkman , and b. j. van wees , nature * 448 * , 06037 ( 2007 ) . k. m. mccreary , a. g. swartz , w. han , j. fabian , and r. k. kawakami , phys . rev
. lett . * 109 * , 186604 ( 2012 ) .
m. garnica , d. stradi , s. barja , f. calleja , c. diaz , m. alcami , n. martin , a. l. vazquez de parga , f. martin , r. miranda .
* 9 * 368 - 374 ( 2013 ) .
a. k. geim .
science * 324 * 1530 - 1534 ( 2009 ) .
f. m. hu , t. ma , h. lin , and j. e. gubernatis .
b. * 84 * , 075414 ( 2011 ) .
yun , m. lee , and y - c .
magn . mater . * 392 * 93 - 96 ( 2014 ) .
r. r nair , i - l tsai , m. sepioni , o. lehtinen , j. keinonen , a.v krasheninnikov , a. h castro neto , m. i katsnelson , a. k geim , i. v grigorieva , nat .
* 4 * 2010 ( 2013 ) .
n. a. pike and d. stroud phys .
b * 89 * 115428 ( 2014 ) .
o. v. yazyev and l. helm , phys . rev .
b * 75 * , 125408 ( 2007 ) .
a. l. rakhmanov , a. v. rozhkov , a. o. sboychakov , and f. nori .
b * * 85 * , 035408 ( 2012 ) .
k. nakada and a. ishii , solid state commun .
* 151 * , 13 ( 2011 ) .
k. t. chan , j. b. neaton , and m. l. cohen , phys .
b * 77 * , 235430 ( 2008 ) .
p. w. anderson , phys .
rev . * 124 * , 41 ( 1961 ) .
r. pariser and r. parr , j. chem .
* 21 * , 767 - 775 ( 1953 ) . k. r. lykke , k. k. murray , and w. c. lineberger , phys . rev .
* a * * 43 * , 6104 - 6107 ( 1991 ) .
t. o. wehling , m. i. katsnelson , and a. i. lichtenstein , chem .
lett . * 476 * 125 ( 2009 ) .
j. hobson and w. a. nierenberg , phys .
rev . * 86 * , 662 ( 1953 ) .
s. yuan , h. de raedt , and m. i. katsnelson , phys .
* b * * 82 * , 115448 ( 2010 ) . | we show that , in graphene with a small concentration of adatoms , the total magnetic moment @xmath0 can be switched on and off by varying the fermi energy @xmath1 , either by applying a gate voltage or by suitable chemical doping .
our calculation is carried out using a simple tight - binding model described previously , combined with a mean - field treatment of the electron - electron interaction on the adatom .
the values of @xmath1 at which the moment is turned on or off are controlled by the strength of the hopping between the graphene sheet and the adatom , the on - site energy of the adatom , and the strength of the electron - electron correlation energy u. our result is in qualitatively consistent with recent experiments by nair _
et al . _
[ nat .
commun . * 4 * , 2010 ( 2013 ) ] . the two - dimensional structure of graphene and the dirac - like dispersion relation of its electrons are the origin of many unusual properties , which may lead to novel electronic or spintronic applications @xcite .
for example , adatoms on graphene may develop magnetic moments which can be manipulated by an applied electric field in a manner similar to its other electric and optical properties @xcite .
recent experimental work by nair _
et al . _
shows that both @xmath2 defects and vacancies in graphene possess magnetic moments that can be switched on and off by chemical doping @xcite . in a recent paper
, we showed , using a tight - binding model , that certain non - magnetic adatoms , such as h , can create a non - zero magnetic moment on graphene @xcite . in this letter , we extend our calculation to show that the magnetic moment of graphene with adatoms can be switched on and off by varying the fermi energy .
this fermi energy can be controlled , in practice , by an applied voltage ; it can also be tuned by suitable chemical doping of the graphene , as in the experiments of ref . .
our calculations show that the onset and turn - off of the magnetic moment depend on the parameters characterizing the adatom , such as the hopping strength between the adatom and graphene , the on - site energy , and the electron - electron correlation energy . in the following section we briefly review our model , and its solution via mean field theory .
the model is appropriate when the adatom lies atop one of the @xmath3 atoms in graphene ( the so - called t site ) , as is the case for adsorbed h and several other adatom species@xcite . in section @xmath4
, we present numerical results showing the dependence of the magnetic moment on fermi energy for various model parameters . in section @xmath5
we give a concluding discussion . |
to meet the fast growing mobile data volume driven by applications such as smartphones and tablets , the traditional wireless network architecture based on a single layer of macro - cells has shifted to one composed of smaller cells such as pico / femto cells with more densely deployed access points ( aps ) . therefore , cloud radio access network ( c - ran ) @xcite has recently been proposed and drawn a great deal of attention . in a c - ran , the distributed aps , also termed remote radio heads ( rrhs ) , are connected to the baseband unit ( bbu ) pool through high bandwidth backhaul links , e.g. optical transport network @xcite , to enable centralized processing , collaborative transmission , and real - time cloud computing . as a result , significant rate improvement can be achieved due to reduced pathloss along with joint scheduling and signal processing .
however , with densely deployed aps , several new challenges arise in c - ran .
first , close proximity of many active aps results in increased interference , and hence the transmit power of aps and/or mobile users ( mus ) needs to be increased to meet any given quality of service ( qos ) .
second , the amount of energy consumed by a large number of active aps @xcite as well as by the transport network to support high - capacity connections with the bbu pool @xcite will also become considerable .
such facts motivate us to optimize the energy consumption in c - ran , which is the primary concern of this paper . in particular , both downlink ( dl ) and uplink ( ul ) transmissions are considered jointly .
the studied c - ran model consists of densely deployed aps jointly serving a set of distributed mus , where comp based joint transmit / receive processing ( beamforming ) over all active aps is employed for dl / ul transmissions . under this setup ,
we study a joint dl and ul mu - ap association and beamforming design problem to minimize the total energy consumption in the network subject to mus given dl and ul qos requirements .
the energy saving is achieved by optimally assigning mus to be served by the minimal subset of active aps , finding the power levels to transmit at all mus and aps , and finding the beamforming vectors to use at the multi - antenna aps .
this problem has not been investigated to date , and the closest prior studies are @xcite . however , the prior studies have all considered mu association and/or active ap selection problems for various objectives from the dl perspective .
in particular , the problems studied in @xcite can be treated as the dl - only version of our considered joint dl and ul problem , in which the transmit beamforming vectors and the active set of aps are jointly optimized to minimize the power consumption at all the aps .
note that the mu association and/or active ap selection based on dl only may result in inefficient transmit power of mus or even their infeasible transmit power in the ul considering various possible asymmetries between the dl and ul in terms of channel , traffic and hardware limitation .
furthermore , with users increasingly using applications with high - bandwidth ul requirements , ul transmission is becoming more important .
for example , the upload speed required for full high definition ( hd ) @xmath0p skype video calling is about @xmath1 mbps @xcite .
therefore , we need to account for both dl and ul transmissions while designing the mus association and active ap selection scheme . the ul - only mu association problem has also been considered extensively in the literature @xcite ; however , their solutions are not applicable in the context of this work due to their assumption of one - to - one mu - ap association .
it is worth noting that the joint mus association and active ap selection is mathematically analogous to the problem of antenna selection in large multiple - input multiple - output ( mimo ) systems @xcite , which aims to reduce the number of radio transmission chains and hence the energy consumption and signal processing overhead .
the connection between these two problems can be recognized by treating the c - ran as a distributed large mimo system . in terms of other related work , there have been many attempts to optimize the energy consumption in cellular networks , but only over a single dimension each time , e.g. power control @xcite , ap `` on / off '' control @xcite , and coordinated multi - point ( comp ) transmission @xcite . to avoid an infeasible power allocation ,
it was suggested in @xcite to gradually remove the mus that can not be supported due to their limited transmit power budgets .
in addition to achieving energy saving from mus perspective , @xcite proposed to switch off the aps that are under light load to save energy by exploiting the fact that the traffic load in cellular networks fluctuates substantially over both space and time due to user mobility and traffic burstiness .
cooperation among different cells or aps could be another possible way to mitigate the interference and achieve energy - efficient communication .
for example , if a certain cluster of aps can jointly support all the mus , the intercell interference can be further reduced especially for the cell - edge mus @xcite .
a judicious combination of these techniques should provide the best solution , and this is the direction of our work .
unfortunately , the considered joint dl and ul mu - ap association and beamforming design problem in this paper involves integer programming and is np hard as shown for a similar problem in ( * ? ? ?
* theorem 1 ) . to tackle this difficulty , two different approaches , i.e.
, group - sparse optimization ( gso ) and relaxed - integer programming ( rip ) , have been adopted in @xcite and @xcite , respectively , to solve a similar dl - only problem , where two polynomial - time algorithms were proposed and shown to achieve good performance through simulations . in particular , the gso approach is motivated by the fact that in the c - ran with densely deployed aps , only a small fraction of the total number of aps needs to be active for meeting all mus qos . however , due to the new consideration of ul transmission in this paper , we will show that the algorithms proposed in @xcite can not be applied directly to solve our problem , and therefore the methods derived in this paper are important advances in this field .
the contributions of this paper are summarized as follows : 1 . to optimize the energy consumption tradeoffs between the active aps and mus , we jointly study the dl and ul mu - ap association and beamforming design by solving a weighted sum - power minimization problem . to our best knowledge
, this paper is the first attempt to unify the dl and ul mu - ap association and beamforming design into one general framework .
2 . due to a critical scaling issue in the ul
receive beamforming design , the gso based algorithm @xcite and the rip based algorithm @xcite can not be applied to solve our joint dl and ul design directly . to address this issue
, we establish a virtual dl transmission for the original ul transmission in c - ran by first ignoring the individual ( per - ap and per - mu ) power constraints based on the celebrated ul - dl duality result @xcite .
consequently , the considered joint dl and ul problem without individual power constraints can be transformed into an equivalent dl problem with two inter - related subproblems corresponding to the original and virtual dl transmissions , respectively . with the equivalent dl - only formulation ,
we extend the gso based and rip based algorithms to solve the relaxed joint dl and ul optimization problem .
3 . considering the fact that the optimal solution to the ul sum - power minimization is component - wise minimum , there is no tradeoff among different mus in terms of power minimization in the ul .
consequently , we are not able to establish the duality result for the per - mu power constraints in the ul , which are thus difficult to incorporate into our developed algorithm . to resolve this issue , we propose a price based iterative method to further optimize the set of active aps while satisfying the per - mu power constraints .
finally , we verify the effectiveness of our proposed algorithms by extensive simulations from three perspectives : ensuring feasibility for both dl and ul transmissions ; achieving optimal network energy saving with mu - ap association ; and flexibly adjusting various power consumption tradeoffs between active aps and mus .
it is worth pointing out that as the baseband processing is migrated to a central unit , i.e. , bbu pool , the data exchanged between the aps and the bbu pool includes oversampled real - time digital signals with very high bit rates ( in the order of gbps ) . as a result ,
the capacity requirement for the backhaul / fronthaul links becomes far more stringent in the c - ran .
given finite backhaul capacity , the optimal strategy for backhaul compression and quantization has been studied recently in e.g. @xcite . in this paper , however , we focus on addressing the energy consumption ( including both transmission and non - transmission related portions ) issue in the c - ran , which is also one of the major concerns for future cellular networks , by assuming that the backhaul transport network is provisioned with sufficiently large capacity .
note that the optical network has been widely accepted as a good option to implement the high - bandwidth backhaul transport network @xcite .
the rest of this paper is organized as follows .
section [ sec : system model ] introduces the c - ran model , and the power consumption models for the aps and mus .
section [ sec : problem formulation ] presents our problem formulation , introduces the two existing approaches , namely gso and rip , and explain the new challenges in solving the joint dl and ul optimization .
section [ sec : joint ul dl ] presents our proposed algorithms based on the virtual dl representation of the ul transmission .
section [ sec : numerical ] shows numerical results .
finally , section [ sec : conclusion ] concludes the paper .
_ notations _ : boldface letters refer to vectors ( lower case ) or matrices ( upper case ) . for an arbitrary - size matrix @xmath2 ,
@xmath3 , @xmath4 , and @xmath5 denote the complex conjugate , conjugate transpose and transpose of @xmath2 , respectively .
the distribution of a circularly symmetric complex gaussian ( cscg ) random vector with mean vector @xmath6 and covariance matrix @xmath7 is denoted by @xmath8 ; and @xmath9 stands for `` distributed as '' .
@xmath10 denotes the space of @xmath11 complex matrices .
@xmath12 denotes the euclidean norm of a complex vector @xmath6 , and @xmath13 denotes the magnitude of a complex number @xmath14 .
we consider a densely deployed c - ran @xcite consisting of @xmath15 access points ( aps ) , denoted by the set @xmath16 .
the set of distributed aps jointly support @xmath17 randomly located mobile users ( mus ) , denoted by the set @xmath18 , for both downlink ( dl ) and uplink ( ul ) communications . in this paper , for the purpose of exposition , we consider linear precoding and decoding in the dl and ul , respectively , which is jointly designed at the bbu pool assuming the perfect channel knowledge for all mus .
the results in this paper can be readily extended to the case of more complex successive precoding / decoding , e.g. dirty - paper coding ( dpc ) @xcite and multiuser detection with successive interference cancelation ( sic ) @xcite , with fixed coding orders among the users .
we also assume that each ap @xmath19 , @xmath20 , is equipped with @xmath21 antennas , and all mus are each equipped with one antenna .
it is further assumed that there exist ideal low - latency backhaul transport links with sufficiently large capacity ( e.g. optical fiber ) connecting the set of aps to the bbu pool , which performs all the baseband signal processing and transmission scheduling for all aps .
the centralized architecture results in efficient coordination of the transmission / reception among all the aps , which can also be opportunistically utilized depending on the traffic demand .
we consider a quasi - static fading environment , and denote the channel vector in the dl from ap @xmath19 to mu @xmath22 and that in the ul from mu @xmath22 to ap @xmath19 as @xmath23 and @xmath24 , respectively .
let the vector consisting of the channels from all the aps to mu @xmath22 and that consisting of the channels from mu @xmath22 to all the aps be @xmath25 $ ] and @xmath26^{t}$ ] , respectively .
there are two main techniques for separating dl and ul transmissions on the same physical transmission medium , i.e. , time - division duplex ( tdd ) and frequency - division duplex ( fdd ) .
if tdd is assumed , channel reciprocity is generally assumed to hold between dl and ul transmissions , which means that the channel vector @xmath27 in the ul is merely the transpose of that @xmath28 in the dl , i.e. , @xmath29 . however , if fdd is assumed , @xmath30 s and @xmath27 s are different in general . in dl transmission ,
the transmitted signal from all aps can be generally expressed as @xmath31 where @xmath32 is the beamforming vector for all aps to cooperatively send one single stream of data signal @xmath33 to mu @xmath22 , which is assumed to be a complex random variable with zero mean and unit variance .
note that @xmath34
. then , the transmitted signal from ap @xmath19 can be expressed as @xmath35 where @xmath36 is the @xmath19th block component of @xmath37 , corresponding to the transmit beamforming vector at ap @xmath19 for mu @xmath22 .
note that @xmath38^{t}$ ] and @xmath39^{t}$ ] , @xmath40 .
from ( [ eq : dl transmit signal from ap n ] ) , the transmit power of ap @xmath19 in dl is obtained as @xmath41 we assume that there exists a maximum transmit power constraint for each ap @xmath19 , i.e. , @xmath42 the received signal at the @xmath22th mu is then expressed as @xmath43 where @xmath44 is the receiver noise at mu @xmath22 , which is assumed to be a circularly symmetric complex gaussian ( cscg ) random variable with zero mean and variance @xmath45 , denoted by @xmath46 . treating the interference as noise ,
the signal - to - interference - plus - noise ratio ( sinr ) in dl for mu @xmath22 is given by @xmath47 in ul transmission , the transmitted signal from mu @xmath22 is given by @xmath48 where @xmath49 denotes the transmit power of mu @xmath22 , and @xmath50 is the information bearing signal which is assumed to be a complex random variable with zero mean and unit variance . with @xmath51 denoting the transmit power limit for mt @xmath22 , it follows that @xmath52 the received signal at all aps is then expressed as @xmath53 where @xmath54 denotes the receiver noise vector at all aps consisting of independent cscg random variables each distributed as @xmath55 .
let @xmath56 denote the receiver beamforming vector used to decode @xmath57 from mu @xmath22 .
then the sinr in ul for mu @xmath22 after applying @xmath58 is given by @xmath59 let @xmath60 denote the @xmath19th block component in @xmath58 , corresponding to the receive beamforming vector at ap @xmath19 for mu @xmath22 .
we thus have @xmath61^{t}$ ] , @xmath40 . the total energy consumption in the c - ran comprises of the energy consumed by all aps and all mus .
from ( [ eq : dl transmit signal from all aps ] ) and ( [ eq : ul transmit signal from mu i ] ) , the total transmit power of all aps in dl and that of all mus in ul can be expressed as @xmath62 and @xmath63 , respectively . besides the static power consumption at each ap @xmath19 due to e.g. real - time a / d and d / a processing , denoted as @xmath64 , in c - ran with centralized processing
, the extensive use of high - capacity backhaul links to connect all aps with the bbu pool makes the power consumption of the transport network no more negligible @xcite .
for example , consider the passive optical network ( pon ) to implement the backhaul transport network @xcite .
the pon assigns an optical line terminal ( olt ) to connect to a set of associated optical network units ( onus ) , which coordinate the set of transport links connecting all the aps to the bbu pool , each through a single fiber . for simplicity
, the resulting power consumption in the pon can be modeled as @xcite @xmath65 where @xmath66 and @xmath67 are both constant and denote the power consumed by the olt and the transport link associated with ap @xmath19 , respectively .
moreover , we consider that for energy saving , some aps and their associated transport links can be switched into sleep mode @xcite ( compared with active mode ) with negligible power consumption ; thus , the total static power consumption of ap @xmath19 , denoted by @xmath68 , @xmath20 , can be saved if ap @xmath19 and its associated transport link are switched into sleep mode for both transmission in dl and ul . for convenience ,
we express the total static power consumption of all active aps as @xmath69 where @xmath70 , @xmath20 , is an indicator function for ap @xmath19 , which is defined as @xmath71 note that in practical pon systems , the olt in general can not be switched into sleep mode as it plays the role of distributor , arbitrator , and aggregator of the transport network , which has a typical fixed power consumption of @xmath72w @xcite .
we thus ignore @xmath66 since it is only a constant . from ( [ eq : on off condition ] ) , mu @xmath22 is associated with an active ap @xmath19 if its corresponding transmit and/or receive beamforming vector at ap @xmath19 is nonzero , i.e. , @xmath73 and/or @xmath74 . under this setup
, it is worth pointing out that each mu @xmath22 is allowed to connect with two different sets of aps for dl and ul transmissions , respectively , e.g. @xmath73 but @xmath75 for some @xmath20 , which is promising to be implemented in next generation cellular networks @xcite .
furthermore , from ( [ eq : on off condition ] ) , ap @xmath19 could be switched into sleep mode only if it does not serve any mu .
we aim to minimize the total energy consumption in the c - ran , including that due to transmit power of all mus ( but ignoring any static power consumption of mu terminals ) as well as that due to transmit power and static power of all active aps .
therefore , we consider the following weighted sum - power as our design metric : @xmath76 where @xmath77 is a weight to trade off between the total energy consumptions between all the active aps and all mus .
to minimize the weighted power consumption in ( [ eq : weighted sum power ] ) , we jointly optimize the dl and ul mu - ap association and transmit / receive beamforming by considering the following problem .
@xmath78 where @xmath79 and @xmath80 are the given sinr requirements of mu @xmath22 for the dl and ul transmissions , respectively . in the rest of this paper , the constraints in ( [ eq : uldl dl power constraint ] ) and ( [ eq : uldl ul power constraint ] )
are termed per - ap and per - mu power constraints , respectively .
problem ( p1 ) can be shown to be non - convex due to the implicit integer programming involved due to indicator function @xmath70 s in the objective . prior to solving problem ( p1 ) , we first need to check its feasibility .
since the dl and ul transmissions are coupled only by the objective function in ( [ eq : uldl objective function ] ) , the feasibility of problem ( p1 ) can be checked by considering two separate feasibility problems : one for the dl and the other for the ul , which have both been well studied in the literature @xcite and thus the details are omitted here for brevity . for the rest of this paper , we assume that problem ( p1 ) is always feasible if all aps are active . as mentioned in section [ sec : introduction ] , the problem of joint mu - ap association and transmit beamforming subject to mus qos and per - ap power constraints for power minimization in the dl - only transmission has been recently studied in @xcite using the approaches of gso and rip , respectively , where two different polynomial - time algorithms were proposed and shown to both achieve good performance by simulations .
in contrast , problem ( p1 ) in this paper considers both dl and ul transmissions to address possible asymmetries between the dl and ul in terms of channel realization , traffic load and hardware limitation .
furthermore , considering that mus are usually powered by finite - capacity batteries as compared to aps that are in general powered by the electricity grid , we study the power consumption tradeoffs between aps and mus by minimizing the weighted sum - power @xmath81 in ( p1 ) . therefore
, the problems considered in @xcite can be treated as special cases of ( p1 ) . in the following ,
we show that due to the new consideration of ul transmission , the algorithms proposed in @xcite based on gso and rip for solving the dl optimization can not be applied directly to solve ( p1 ) , which thus motivates us to find a new method to resolve this issue in section [ sec : joint ul dl ] .
given the fact that the static power , i.e. , @xmath82 , is in practice significantly larger than the transmit power at each ap @xmath19 , to minimize the total network energy consumption @xcite , it is conceivable that for the optimal solution of ( p1 ) only a subset of @xmath15 aps should be active . as a result , a `` group - sparse '' property can be inferred from the following concatenated beamforming vector : @xmath83,\cdots,[\hat{\mathbf{w}}^{\text{dl}}_n,\hat{\mathbf{v}}^{\text{ul}}_n]\right]\end{aligned}\ ] ] in which @xmath84 $ ] and @xmath85 $ ] , @xmath86 , i.e. , the beamforming vectors are grouped according to their associated aps . if ap @xmath19 is in the sleep mode , its corresponding block @xmath87 $ ] in ( [ eq : group sparsity ] ) needs to be zero . consequently , the fact that a small subset of deployed aps is selected to be active implies that the concatenated beamforming vector in ( [ eq : group sparsity ] ) should contain only a very few non - zero block components .
one well - known approach to enforce desired group sparsity in the obtained solutions for optimization problems is by adding to the objective function an appropriate penalty term .
the widely used group sparsity enforcing penalty function , which was first introduced in the context of the group least - absolute selection and shrinkage operator ( lasso ) problem @xcite , is the mixed @xmath88 norm . in our case ,
such a penalty is expressed as @xmath89\right\|.\end{aligned}\ ] ] the @xmath88 norm in ( [ eq : uldl penalty ] ) , similar to @xmath90 norm , offers the closest convex approximation to the @xmath91 norm over the vector consisting of @xmath92 norms @xmath93\right\|\right\}^{n}_{n=1}$ ] , implying that each @xmath94\right\|$ ] is desired to be set to zero to obtain group sparsity . more generally , the mixed @xmath95 norm has also been shown to be able to recover group sparsity with @xmath96 @xcite , among which the @xmath97 norm , defined as @xmath98 has been widely used @xcite . compared with @xmath88 norm ,
@xmath97 norm has the potential to obtain more sparse solution but may lead to undesired solution with components of equal magnitude . in this paper
, we focus on the @xmath88 norm in ( [ eq : uldl penalty ] ) for our study
. we will compare the performance of @xmath88 and @xmath97 norms by simulations in section [ sec : numerical ] . according to @xcite , at first glance
it seems that using the @xmath88 norm , problem ( p1 ) can be approximately solved by replacing the objective function with @xmath99 where @xmath100 can be treated as a convex relaxation of the indicator functions in ( [ eq : weighted sum power ] ) , and @xmath101 indicates the relative importance of the penalty term associated with ap @xmath19 .
however , problem ( p1 ) with ( [ eq : scaling problem objective ] ) as the objective function is still non - convex due to the constraints in ( [ eq : uldl dl sinr constraint ] ) and ( [ eq : uldl ul sinr constraint ] ) .
furthermore , since the ul receive beamforming vector @xmath58 s can be scaled down to be arbitrarily small without affecting the ul sinr defined in ( [ eq : ul sinr ] ) , minimizing ( [ eq : scaling problem objective ] ) directly will result in all @xmath58 s going to zero . to be more specific ,
let @xmath102 and @xmath103 denote the optimal solution of problem ( p1 ) with ( [ eq : scaling problem objective ] ) as the objective function .
then , it follows that @xmath104 and @xmath105 , preserves the `` group - sparse '' property where the non - zero block components correspond to the active aps .
two issues thus arise : first , the ul does not contribute to the selection of active aps ; second , the set of selected active aps based on the dl only can not guarantee the qos requirements for the ul . as a result ,
the @xmath88 norm penalty term in ( [ eq : scaling problem objective ] ) or more generally the @xmath95 norm penalty does not work for the joint dl and ul ap selection in our problem , and hence the algorithm proposed in @xcite , which involves only the dl transmit beamforming vector @xmath37 s , can not be modified in a straightforward way to solve our problem .
next , we reformulate problem ( p1 ) by introducing a set of binary variable @xmath106 s indicating the `` active / sleep '' state of each ap as follows .
@xmath107 where @xmath108 is a constant with arbitrary value .
note that the active - sleep constraints in ( [ eq : bigm onoff ] ) are inspired by the well - known big-@xmath109 method @xcite : if @xmath110 , the constraint ( [ eq : bigm onoff ] ) ensures that @xmath111 ; if @xmath112 , the constraint has no effect on @xmath113 and @xmath114 , as @xmath115 represents an upper bound on the term @xmath116 .
notice that @xmath117 can be chosen arbitrarily due to the scaling invariant property of ul receive beamforming vector @xmath58 s . with the active - sleep constraints in ( [ eq : bigm onoff ] )
, the equivalence between problems ( p1 ) and ( p2 ) can be easily verified . in @xcite , a similar problem to ( p2 )
was studied corresponding to the case with only dl transmission . for problem ( p2 ) without @xmath58 and @xmath118 and their corresponding constraints , the problem can be transformed to a convex second - order cone programming ( socp ) by relaxing the binary variable @xmath106 as @xmath119 , \forall n \in \mathcal{n}$ ] . under this convex relaxation ,
a bnc algorithm , which is a combination of the branch - and - bound ( bnb ) and the cutting plane ( cp ) methods @xcite , was proposed in @xcite to solve the dl problem optimally . however , the computational complexity of bnb is prohibitive for large networks in practice , which grows exponentially with the number of aps . to obtain polynomial - time algorithm with near - optimal performance , in @xcite
, the authors further proposed an incentive measure based heuristic algorithm to determine the set of active aps .
the incentive measure reflects the importance of each ap to the whole network and is defined as the ratio of the total power received at all mus to the total power expended for each ap .
however , with both dl and ul transmissions , it is observed that problem ( p2 ) can no longer be transformed to a convex form by relaxing @xmath106 s as continuous variables due to the constraints in ( [ eq : uldl ul sinr constraint ] ) .
furthermore , because of the scaling invariant property of ul receive beamforming vectors , solving the relaxed problem of ( p2 ) will result in all @xmath58 s going to zero , similar to the case of gso based solution .
particularly , the value of the relaxed indicator @xmath120 , will not be related to @xmath58 s , which in fact contributes to the penalty incurred in the objective due to the static power of ap @xmath19 , i.e. , @xmath121 .
finally , it is nontrivial to find an incentive measure that reflects the importance of each ap to both dl and ul transmissions .
in this section , we provide two efficient algorithms to approximately solve problem ( p1 ) based on the gso and rip approaches , respectively .
first , we consider the approach of gso and present a new method to address the joint dl and ul optimization . to obtain an efficient solution for problem ( p1 ) , we first assume that all the aps and mus have infinite power budget , i.e. , @xmath122 , @xmath123 and @xmath124 . the resulting problem is termed ( p1 - 1 ) .
an equivalent reformulation of problem ( p1 - 1 ) is then provided to overcome the receive beamforming scaling issue mentioned in section [ sec : problem formulation ] . then , we discuss the challenges of dealing with finite per - ap and per - mu power constraints and provide efficient methods to handle them .
first , we consider the following transmit sum - power minimization problem in the ul : @xmath125 from @xcite , it follows that problem ( [ eq : ul power minimization ] ) can be solved in a virtual dl channel as @xmath126 where @xmath127 is the virtual dl transmit beamforming vector over @xmath15 aps for mu @xmath22 .
denote @xmath128 , @xmath129 and @xmath130 , @xmath131 as the optimal solutions to problems ( [ eq : ul power minimization ] ) and ( [ eq : virtual dl power minimizatin ] ) , respectively .
then from @xcite it follows that @xmath128 and @xmath130 can be set to be identical , @xmath132 , and furthermore @xmath133 . by establishing a virtual dl transmission for the ul transmission based on the above ul - dl duality
, we have the following lemma . [ lemma:1 ] problem ( p1 - 1 ) is equivalent to the following problem .
@xmath134 for any given feasible solution to problem ( p3 ) , we can always find a corresponding feasible solution to problem ( p1 - 1 ) achieving the same objective value as that of problem ( p3 ) , and vice versa , similar as ( * ? ? ?
* proposition 1 ) ; thus , problems ( p1 - 1 ) and ( p3 ) achieve the same optimal value with the same set of optimal dl / ul beamforming vectors .
lemma [ lemma:1 ] is thus proved . since problem ( p3 )
is a dl - only problem that has the same `` group - sparse '' property as ( p1 - 1 ) , it can be approximately solved by replacing the objective function with @xmath135 comparing ( [ eq : reformulate problem objective ] ) and ( [ eq : scaling problem objective ] ) , we have successfully solved the scaling issue of ul receive beamforming vector , @xmath58 s , by replacing them with the equivalent dl transmit beamforming vector , @xmath136 s , since from ( [ eq : virtual dl power minimizatin ] ) it follows that the virtual dl sinr of each mu @xmath22 is no more scaling invariant to @xmath136 s . furthermore , since any arbitrary phase rotation of the beamforming vectors does not affect both ( [ eq : reformulate problem objective ] ) and the sinr constrains in ( [ eq : dl sinr r ] ) and ( [ eq : vdl sinr r ] ) , ( p3 ) with ( [ eq : reformulate problem objective ] ) as the objective function can be reformulated as a convex socp @xcite , which is given by @xmath137 where @xmath138 $ ] , @xmath139 $ ] , and @xmath140 s are auxiliary variables with @xmath141 and @xmath142 indicating that ap @xmath19 is in active and sleep mode , respectively
. notice that without @xmath88 norm penalty or @xmath143 , @xmath123 , problem ( p4 ) can be decomposed into two separate minimum - power beamforming design problems : one for the original dl transmission , and the other for the virtual dl transmission .
conventionally , the ul transmit sum - power minimization problem , as in ( [ eq : ul power minimization ] ) , has a convenient analytical structure and thus is computationally easier to handle , as compared to the dl minimum - power beamforming design problem , as in ( [ eq : virtual dl power minimizatin ] ) .
consequently , most existing studies in the literature have transformed the dl problem to its virtual ul formulation for convenience .
the motivation of exploiting the reverse direction in this work , however , is to overcome the scaling issue of ul receive beamforming in gso , so that we can solve the ap selection problem jointly for both dl and ul transmissions .
next , we present the complete algorithm for problem ( p1 - 1 ) based on gso , in which three steps need to be performed sequentially . 1 . _
identify the subset of active aps denoted as @xmath144_. this can be done by iteratively solving problem ( p4 ) with different @xmath145 s . notice that how to set the parameter @xmath145 s in ( p4 ) plays a key role in the resulting aps selection .
to optimally set the values of @xmath145 s , we adopt an iterative method similar as in @xcite , shown as follows . in the @xmath146th iteration , @xmath147 , @xmath148 s
are obtained by solving problem ( p4 ) with @xmath149 .
the @xmath150 s are derived from the solution @xmath151 s of the @xmath152th iteration as @xmath153 where @xmath154 is a small positive number to ensure stability .
notice that the initial values of @xmath155 s are chosen as @xmath156 where @xmath157 and @xmath158 are the beamforming vector solution of problem ( p4 ) with @xmath159 .
the above update is repeated until @xmath160 , @xmath123 , where @xmath117 is a small positive constant that controls the algorithm accuracy .
+ let @xmath161 $ ] denote the sparse solution after the convergence of the above iterative algorithm .
then the nonzero entries in @xmath162 correspond to the aps that need to be active , i.e. , @xmath163 .
_ obtain the optimal transmit / receive beamforming vectors @xmath164 and @xmath165 , @xmath131 , given the selected active aps_. this can be done by solving ( p4 ) with @xmath143 , @xmath166 and @xmath167 .
3 . _ obtain the optimal transmit power values of mus @xmath168 , @xmath131_. this can be done by solving problem ( [ eq : ul power minimization ] ) with @xmath169 , @xmath170 , which is a simple linear programming ( lp ) problem .
the iterative update given in ( [ eq : iterative update ] ) is designed to make small entries in @xmath171 converge to zero .
furthermore , as the updating evolves , the penalty associated with ap @xmath19 in the objective function , i.e. , @xmath172 , will converge to two possible values : @xmath173 in other words , only the active aps will incur penalties being the exact same values as their static power consumption , which has the same effect as the indicator function in problem ( p1 - 1 ) or ( p1 ) .
convergence of this algorithm can be shown by identifying the iterative update as a majorization - minimization ( mm ) algorithm @xcite for a concave minimization problem , i.e. , using @xmath174 function , which is concave , to approximate the indicator function given in ( [ eq : on off condition ] ) .
the details are thus omitted due to space limitations .
it is first observed that the per - ap power constraints in ( [ eq : uldl dl power constraint ] ) , i.e. , @xmath175 are convex .
therefore , adding per - ap power constraints to problem ( p1 - 1 ) does not need to alter the above algorithm .
thus , we focus on the per - mu power constraints in the ul transmission in this subsection .
again , we consider the following transmit sum - power minimization problem in the ul with per - mu power constraints : @xmath176 although it has been shown in @xcite that the sum - power minimization problem in the dl with per - ap power constraints can be transformed into an equivalent min - max optimization problem in the ul , we are not able to find an equivalent dl problem for problem ( [ eq : ul power minimization with per mu ] ) as in section [ sec : without ] which is able to handle the per - mu power constraints .
the fundamental reason is that the power allocation obtained by solving problem ( [ eq : ul power minimization ] ) is already component - wise minimum , which can be shown by the uniqueness of the fixed - point solution for a set of minimum sinr requirements in the ul given randomly generated channels @xcite .
the component - wise minimum power allocation indicates that it is not possible to further reduce one particular mu s power consumption by increasing others , i.e. , there is no tradeoff among different mus in terms of power minimization .
consequently , solving problem ( [ eq : ul power minimization with per mu ] ) requires only one additional step compared with solving problem ( [ eq : ul power minimization ] ) , i.e. , checking whether the optimal power solution to problem ( [ eq : ul power minimization ] ) satisfies the per - mu power constraints . if this is the case , the solution is also optimal for problem ( [ eq : ul power minimization with per mu ] ) ; otherwise , problem ( [ eq : ul power minimization with per mu ] ) is infeasible . next , we present our complete algorithm for problem ( p1 ) with the per - ap and per - mu power constraints .
compared to the algorithm proposed for problem ( p1 - 1 ) in section [ sec : without ] without the per - ap and per - mu power constraints , the new algorithm differs in the first step , i.e. , to identify the subset of active aps .
the main idea is that a set of candidate active aps is first obtained by ignoring the per - mu power constraints but with a new sum - power constraint in the ul ( or equivalently its virtual dl ) , i.e. , we iteratively solve the following problem similarly as in the first step of solving problem ( p1 - 1 ) in section [ sec : without ] .
@xmath177 the sum - power constraint in ( [ eq : sum user p4 ] ) is added to impose a mild control on the transmit powers of all mus in the ul . after obtaining the candidate set , the feasibility of the ul transmission is then verified .
if the candidate set can support the ul transmission with the given per - mu power constraints , then the optimal solution of ( p1 ) is obtained ; otherwise , one or more aps need to be active for the ul transmission . to be more specific ,
denote the set of candidate active aps obtained by iteratively solving problem ( p5 ) as @xmath178 .
problem ( [ eq : ul power minimization ] ) is then solved with @xmath178 , for which the feasibility is guaranteed due to the virtual dl sinr constraints in ( [ eq : vdl - socp sinr constraint ] ) .
we denote the obtained power allocation as @xmath179 , @xmath180 . *
if @xmath178 can support the ul transmission without violating any mu s power constraint , i.e. , @xmath181 the candidate set can be finalized as the set of active aps and the algorithm proceeds to find the optimal transmit / receive beamforming vectors similarly as that in section [ sec : without ] . * if @xmath178 can not support the ul transmission with the given mu s power constraints , we propose the following price based iterative method to determine the additional active aps .
specifically , in each iteration , for those aps that are not in the candidate set each will be assigned a price @xmath182 , which is defined as @xmath183 where @xmath184 .
the price @xmath185 is set to be the normalized ( by its corresponding static power consumption ) weighted - sum power gains of the channels from ap @xmath186 to all the mus that have their power constraints being violated .
the weights are chosen as the ratios of mus required additional powers to their individual power limits . according to the definition of @xmath185 in ( [ eq : price ] ) ,
the ap having smaller static power consumption and better channels to mus whose power constraints are more severely violated will be associated with a larger price .
the candidate set is then updated by including the ap that corresponds the largest @xmath185 as @xmath187 with updated @xmath178 , the feasibility of the ul transmission needs to be re - checked by obtaining a new set of power allocation , which will be used to compute the new @xmath185 s in next iteration if further updating is required .
the above process is repeated until all the mus power constraints are satisfied .
its convergence is guaranteed since problem ( p1 ) has been assumed to be feasible if all aps are active . combining with the algorithm in section [ sec : without ] , our complete algorithm for problem ( p1 ) based on gso is summarized in table [ table1 ] . for the algorithm given in table [ table1 ] , there are two problems that need to be iteratively solved , i.e. , problems ( [ eq : ul power minimization ] ) and ( p5 ) . since problem ( [ eq : ul power minimization ] )
can be efficiently solved by the fixed - point algorithm @xcite , the computation time is dominated by solving the socp problem ( p5 ) . if the primal - dual interior point algorithm @xcite is used by the numerical solver for solving ( p5 ) , the computational complexity is of order @xmath188 .
furthermore , since the convergence of the iterative update in steps 4)-5 ) , governed by the mm algorithm , is very fast ( approximately @xmath189-@xmath190 iterations ) as observed in the simulations , the overall complexity of the algorithm in table [ table1 ] is approximately @xmath191 . ' '' '' 1 .
set @xmath192 , initialize the set of candidate active aps as @xmath193 .
2 . obtain @xmath157 s and @xmath158 s by solving problem ( p5 ) with @xmath159 .
3 . set @xmath194 .
* repeat : * * @xmath195 . * set @xmath196 , @xmath123 . *
obtain @xmath197 $ ] by solving problem ( p5 ) with @xmath198 .
5 . * until * @xmath199 , @xmath123 or @xmath200 . 6 . set @xmath178 as @xmath201 .
* repeat : * * obtain @xmath179 , @xmath131 , by solving problem ( [ eq : ul power minimization ] ) with @xmath178 .
* set @xmath202 . *
set @xmath203 .
* set @xmath204
* until * @xmath205 .
9 . obtain @xmath164 and @xmath206 , @xmath131 , by solving ( p5 ) with @xmath143 , @xmath123 and @xmath207 .
10 . set @xmath208 , and compute the @xmath168 , @xmath131 , by solving problem ( [ eq : ul power minimization ] ) . ' '' '' [ table1 ] in this subsection , an alternative algorithm for problem ( p1 ) is developed based on rip by applying the same idea of establishing a virtual dl transmission for the original ul .
similar to the case with gso , the per - mu power constraints are first replaced with a sum - power constraint in the ul .
the resulting problem is further reformulated as a convex socp by relaxing the binary variables @xmath209 , which is given as follows .
@xmath210 note that instead of implementing the active - sleep constraints jointly for the actual dl and virtual dl as ( [ eq : bigm onoff ] ) in problem ( p2 ) , i.e. , @xmath211 , we divide them into two sets of coupled active - sleep constraints as in ( [ eq : bigm1 ] ) and ( [ eq : bigm2 ] ) via @xmath106 s .
for the non - relaxed problem of ( p6 ) with binary @xmath106 s , i.e. @xmath212 , it can be shown that these two formulations are equivalent . however , for the case of the relaxed problem ( p6 ) with continuous valued @xmath106 s , the separated active - sleep constraints are designed to avoid the situation that the difference between @xmath213 and @xmath214 is too large such that the optimal value of @xmath106 is dominated by either dl or ul transmission . to implement the big-@xmath109 method with active - sleep constraints @xcite , an appropriate upper bound for the term @xmath215 needs to be found . according to the ul - dl duality
, the minimum sum - power achieved is the same for the ul and its virtual dl transmissions .
therefore , @xmath214 can be chosen as the upper bound of @xmath216 .
finally , it is evident that the optimal value of problem ( p6 ) serves as a lower bound of its non - relaxed problem with binary @xmath106 s . in order to further tighten this lower bound , one way is to reduce the feasible set of design variables .
constraint in ( [ eq : red ] ) is introduced specifically to achieve this end , which can be shown to be redundant for the non - relaxed problem of ( p6 ) .
we adopt the same idea of incentive measure based ap selection as in @xcite to design a polynomial - time algorithm for problem ( p2 ) .
however , it remains to find an incentive measure that reflects the importance of each ap to both dl and ul transmissions based on problem ( p6 ) .
it is interesting to observe that after transforming the ul related terms to their virtual dl counterparts , the optimal relaxed binary variable solution of problem ( p6 ) becomes a good choice to serve this purpose .
let @xmath217 , @xmath218 and @xmath219 denote the optimal solution to problem ( p6 ) . intuitively , the ap that has larger static power consumption and worse channels to all mus is more desired to be switched into sleep mode from the perspective of energy saving . in ( p6 ) , for ap @xmath19 having larger @xmath82 , @xmath217 is desired to be smaller in order to achieve the minimum value of the objective function .
furthermore , it is practically valid that for dl power minimization problem , the optimal transmit power of aps that have worse channels to mus is in general smaller . as a result , solving problem ( p6 ) yields smaller @xmath220 and @xmath221 and thus smaller @xmath217 for ap @xmath19 that has worse channels to mus for both the dl and ul transmissions . to summarize , the ap that corresponds to smaller @xmath217 is more desired to be switched into sleep mode . an iterative process is then designed to determine the set of active aps based on @xmath217 s that are taken as incentive measures
. the process starts with assuming all aps are active . in each iteration , problem ( p6 ) is solved with a candidate set of active aps , and the ap corresponding to the smallest @xmath217 will be removed from the candidate set .
this process is repeated until one of the following conditions occurs : * the weighted sum - power can not be further reduced ; * problem ( p6 ) becomes infeasible ; * problem ( [ eq : ul power minimization with per mu ] ) becomes infeasible .
note that the feasibility checking for problem ( [ eq : ul power minimization with per mu ] ) is the same as that in algorithm i , which ensures the per - mu power constraints .
an overall algorithm for problem ( p1 ) based on rip is summarized in table [ table2 ] . for the algorithm given in table [ table2 ] , the computation time is dominated by solving the socp problem ( p6 ) . if the primal - dual interior point algorithm @xcite is used by the numerical solver for solving ( p6 ) , the computational complexity is of order @xmath188 .
furthermore , since the worst case complexity for the iteration in steps 2)-3 ) is @xmath222 , the overall complexity of the algorithm in table [ table2 ] is @xmath223 . ' '' '' 1 .
set @xmath192 , @xmath224 a sufficiently large value , and initialize the set of candidate active aps as @xmath193 .
* repeat : * * @xmath195 .
* solve problem ( p6 ) with @xmath110 , @xmath225 . *
set @xmath226 .
* set @xmath227 as the optimal value of problem ( p6 ) with @xmath228 and @xmath229 .
until * @xmath230 or problem ( p6 ) is infeasible or problem ( [ eq : ul power minimization with per mu ] ) is infeasible .
4 . obtain @xmath164 and @xmath206 , @xmath131 , by solving problem ( p6 ) with @xmath228 and @xmath229 .
5 . set @xmath208 , and compute the @xmath168 , @xmath131 , by solving problem ( [ eq : ul power minimization ] ) . ' '' '' [ table2 ]
in this section , we present numerical results to verify our proposed algorithms from three perspectives : ensuring feasibility for both dl and ul transmissions ; achieving network power saving with optimal mu - ap association ; and adjusting minimum power consumption tradeoffs between active aps and mus .
we consider two possible c - ran configurations : 1 .
homogeneous setup : all aps are assumed to have the same power consumption model with @xmath231w and @xmath232w , @xmath233 , if not specified otherwise .
heterogeneous setup : two types of aps are assumed , namely , high - power ap ( hap ) and low - power ap ( lap ) , where the static power consumption for hap and lap are set as @xmath234w and @xmath235w , respectively , and the transmit power budgets for hap and lap are set as @xmath1w and @xmath236w , respectively .
we assume that each ap @xmath19 , @xmath20 , is equipped with @xmath237 antennas . for the single - antenna mu , we set the transmit power limit as @xmath238w , @xmath170 . for simplicity , we assume that the sinr requirements of all mus are the same in the ul or dl . all the aps ( except haps under the heterogeneous setup ) and mus are assumed to be uniformly and independently distributed in a square area with the size of @xmath239km@xmath240@xmath239 km .
for all the simulations under heterogeneous setup , it is assumed that there are @xmath235 haps with fixed location at @xmath241 $ ] and @xmath242 $ ] , respectively .
we assume a simplified channel model consisting of the distance - dependent attenuation with pathloss exponent @xmath243 and a multiplicative random factor ( exponentially distributed with unit mean ) accounting for short - term rayleigh fading .
we also set @xmath244 if not specified otherwise , i.e. , we consider the sum - power consumption of all active aps and mus .
finally , we set the receiver noise power for all the aps and mus as @xmath245dbm .
first , we demonstrate the importance of active ap selection by jointly considering both dl and ul transmission in terms of the sinr feasibility in c - ran .
since feasibility is our focus here instead of power consumption , it is assumed that the selected active aps will support all mus for both the dl and ul transmissions .
the simulation results compare our proposed algorithms ( i.e. , algorithms i and ii ) with the following three ap selection schemes : * * ap initiated reference signal strength ( apirss ) based selection * : in this scheme , aps first broadcast orthogonal reference signals .
then , for each mu , the ap corresponding to the largest received reference signal strength will be included in the set of active aps .
note that this scheme has been implemented in practical cellular systems @xcite . *
* mu initiated reference signal strength ( muirss ) based selection * : in this scheme , mus first broadcast orthogonal reference signals .
then , for each mu , the ap corresponding to the largest received reference signal strength will be included in the set of active aps . note that since all mus are assumed to transmit reference signals with equal power and pathloss in general dominates short - term fading , the ap that is closest to each mu will receive strongest reference signal in general .
also note that in the previous apirss based scheme , if all aps are assumed to transmit with equal reference signal power ( e.g. , for the homogenous setup ) , the selected active aps will be very likely to be the same as those by the muirss based scheme . *
* proposed algorithm without considering ul ( paw / oul ) * : in this algorithm , the set of active aps are chosen from the conventional dl perspective by modifying our proposed algorithms . specifically , algorithm i is used here and similar results can be obtained with algorithm ii .
note that algorithm i without considering ul transmission is similar to that proposed in @xcite . with the obtained set of active aps , the feasibility check of problem ( p1 ) can be decoupled into two independent feasibility problems : one for the dl and the other for the ul , while the network feasibility is achieved only when both the ul and dl sinr feasibility of all mus are guaranteed . in fig .
[ fig : feasibilitydemo ] , we illustrate the set of active aps generated by different schemes under the heterogeneous setup , and also compare them with that by the optimal exhaustive search .
it is assumed that there are @xmath235 haps and @xmath246 laps jointly supporting @xmath246 mus .
the sinr targets for both dl and ul transmissions of all mus are set as @xmath246db .
first , it is observed that algorithm i and algorithm ii obtain the same set of active aps as shown in fig .
[ fig : feasibilitydemohetnetl12 ] , which is also identical to that found by exhaustive search .
second , it is observed that the @xmath235 haps are both chosen to be active in fig .
[ fig : feasibilitydemohetnethighsignal ] for the apirss based scheme .
this is due to the significant difference between hap and lap in terms of transmit power , which makes most mus receive the strongest dl reference signal from the hap .
the above phenomenon is commonly found in heterogenous network ( hetnet ) @xcite with different types of bss ( e.g. macro / micro / pico bss ) .
third , from fig .
[ fig : feasibilitydemohetnetclose ] , the active aps by the muirss based scheme are simply those closer to the mus , which is as expected . finally , in fig .
[ fig : feasibilitydemohetnetdl ] , only two laps are chosen to support all mus with the paw / oul algorithm .
this is because the algorithm does not consider ul transmission , and as a result fig .
[ fig : feasibilitydemohetnetdl ] only shows the most energy - efficient ap selection for dl transmission . to compare the feasibility performance , we run the above algorithms with different dl and ul sinr targets
it is assumed that @xmath247 and @xmath248 .
the results are summarized in table [ table4 ] and table [ table3 ] , where the number of infeasible cases for each scheme is shown over @xmath249 randomly generated network and channel realizations , for homogeneous setup and heterogeneous setup , respectively .
note that in these examples , algorithm i and algorithm ii have identical feasibility performance , since the system is infeasible only when the dl / ul sinr requirements can not be supported for given channels and power budgets even with all aps being active . from both table [ table4 ] and table [ table3 ]
, it is first observed that the three comparison schemes , i.e. , apirss based scheme , muirss based scheme and paw / oul , all incur much larger number of infeasible cases as compared to our proposed algorithms .
it is also observed that among the three comparison schemes , paw / oul has the best performance ( or the minimum number of infeasible cases ) when the dl transmission is dominant ( i.e. , @xmath250 ) ; however , it performs the worst in the opposite situation ( i.e. , @xmath251 ) .
this observation indicates that dl oriented scheme could result in infeasible transmit power of mus in the ul for the cases with stringent ul requirements . from the last two rows of table [ table3 ] , it is observed that the apirss based scheme performs worse than the muirss based scheme when the ul sinr target is high .
this is because that under heterogeneous setup , as shown in fig .
[ fig : feasibilitydemohetnethighsignal ] , mus are attached to the haps under apirss based scheme although the haps may be actually more distant away from mus compared with the distributed laps .
this imbalanced association causes much higher transmit powers of mus or even their infeasible transmit power in the ul .
there has been effort in the literature to address this traffic imbalance problem in hetnet .
for example in @xcite , the reference signal from picocell bs is multiplied by a factor with magnitude being larger than one , which makes it appear more appealing for mu association than the heavily - loaded macrocell bs .
.feasibility performance comparison under homogeneous setup [ cols="^,^,^,^,^,^,^ " , ] [ table3 ] next , we compare the performance of the proposed algorithms in terms of sum - power minimization in c - ran with the following benchmark schemes : * * exhaustive search ( es ) * : in this scheme , the optimal set of active aps are found by exhaustive search , which serves as the performance upper bound ( or lower bound on the sum - power consumption ) for other considered schemes . with any set of active aps , the minimum - power dl and ul beamforming problems can be separately solved .
since the complexity of es grows exponentially with @xmath15 , it can only be implemented for c - ran with small number of aps . *
* joint processing ( jp ) among all aps*@xcite : in this scheme , all the aps are assumed to be active and only the total transmit power consumption is minimized by solving two separate ( dl and ul ) minimum - power beamforming design problems .
* * algorithm i with @xmath97 norm penalty * : this algorithm is the same as that given in table [ table1 ] except that the sparsity enforcing penalty is replaced with @xmath97 norm as given in ( [ eq : infinity ] ) . in our simulations
, we consider the homogeneous c - ran setup with @xmath247 and plot the performance by averaging over @xmath252 randomly generated network and channel realizations .
the sinr requirements are set as @xmath253db for all mus @xmath254 .
[ fig : performancemu ] and fig .
[ fig : performancecomparepcdis ] show the sum - power consumption achieved by different algorithms versus the number of mus @xmath17 and ap static power consumption @xmath82 ( assumed to be identical for all aps ) , respectively . from both figures ,
it is observed that the proposed algorithms have similar performance as the optimal es and achieve significant power saving compared with jp .
it is also observed that the penalty term based on either @xmath88 or @xmath97 norm has small impact on the performance of algorithm i. finally , algorithm i always outperforms algorithm ii although the performance gap is not significant .
= 0.7w , @xmath123.,title="fig:",width=453 ] = 0.7 .,title="fig:",width=453 ] finally , we compare the sum - power consumption tradeoffs between active aps and all mus for the proposed algorithms as well as the optimal es , by varying the weight parameter @xmath255 in our formulated problems .
we consider a homogenous c - ran setup with @xmath247 and @xmath248 , where @xmath253db for all mus @xmath254 .
since it has been shown in the pervious subsection that algorithm i with @xmath88 norm and @xmath97 norm achieves similar performance , we choose @xmath88 norm in this simulation .
furthermore , since js assumes that all the aps are active , which decouples dl and ul transmissions and thus has no sum - power consumption tradeoffs between aps and mus , it is also not included . from fig . [
fig : tradeoff ] , it is first observed that for all considered algorithms , as @xmath255 increases , the sum - power consumption of active aps increases and that of all mus decreases , which is as expected .
it is also observed that algorithm i achieves trade - off performance closer to es and outperforms algorithm ii , which is in accordance with the results in figs .
[ fig : performancemu ] and [ fig : performancecomparepcdis ] .
in this paper , we consider c - ran with densely deployed aps cooperatively serving distributed mus for both the dl and ul transmissions .
we study the problem of joint dl and ul mu - ap association and beamforming design to optimize the energy consumption tradeoffs between the active aps and mus .
leveraging on the celebrated ul - dl duality result , we show that by establishing a virtual dl transmission for the original ul transmission , the joint dl and ul problem can be converted to an equivalent dl problem in c - ran with two inter - related subproblems for the original and virtual dl transmissions , respectively . based on this transformation , two efficient algorithms for joint dl and ul mu - ap association and beamforming design
are proposed based on gso and rip techniques , respectively . by extensive simulations
, it is shown that our proposed algorithms improve the network reliability / feasibility , energy efficiency , as well as power consumption tradeoffs between aps and mus , as compared to other existing methods in the literature .
s. tombaz , p. monti , k. wang , a. vastberg , m. forzati , and j. zander , `` impact of backhauling power consumption on the deployment of heterogeneous mobile networks , '' in _ proc .
ieee globecom _ ,
pp . 1 - 5 , dec . 2011 .
m. hong , r. sun , h. baligh , and z. luo , `` joint base station clustering and beamformer design for partial coordinated transmission in heterogeneous networks , '' _ ieee j. sel .
areas commun .
226 - 240 , feb . 2013 .
y. cheng , m. pesavento , and a. philipp , `` joint network optimization and downlink beamforming for comp transmissions using mixed integer conic programming , '' _ ieee trans . signal process .
16 , pp . 3972 - 3987 , aug
. 2013 .
m. rasti , a. r. sharafat , s. member , and j. zander , `` pareto and energy - efficient distributed power control with feasibility check in wireless networks , '' _ ieee trans .
inform . theory _
245 - 255 , jan . 2011 .
k. son , h. kim , y. yi , and b. krishnamachari , `` base station operation and user association mechanisms for energy - delay tradeoffs in green cellular networks , '' _ ieee j. sel .
areas commun .
29 , no . 8 , pp . 1525 - 1536 , sep .
d. gesbert , s. hanly , h. huang , s. s. shitz , o. simeone , and w. yu , `` multi - cell mimo cooperative networks : a new look at interference '' , _ ieee j. sel .
areas commun .
9 , pp . 1380 - 1408 , dec .
2010 .
s. h. park , o. simeone , o. sahin , and s. shamai , `` joint precoding and multivariate backhaul compression for the downlink of cloud radio access networks , '' _ ieee trans .
signal process .
22 , pp . 5646 - 5658 , nov . 2013 .
l. liu , r. zhang , and k. c. chua , `` achieving global optimality for weighted sum - rate maximization in the k - user gaussian interference channel with multiple antennas , '' _ ieee trans .
wireless commun .
5 , pp . 1933 - 1945 .
may 2012 .
g. auer , v. giannini , c. desset , i. godor , p. skillermark , m. olsson , m. imran , d. sabella , m. gonzalez , o. blume , and a. fehske , `` how much energy is needed to run a wireless network ? '' _ ieee trans .
wireless commun .
_ , vol . 18 , no .
40 - 49 , oct . 2011 .
m. yuan and y. lin , `` model selection and estimation in regression with grouped variables , '' _ journal of the royal statistical society : series b ( statistical methodology ) _ , vol .
49 - 67 , feb . 2006 .
q. ye , y. chen , m. al - shalash , c. caramanis , and j. g. andrews , `` user association for load balancing in heterogeneous cellular network , '' _ ieee trans .
wireless commun .
12 , no . 6 ,
2706 - 2716 , jun . 2013 .
j. sangiamwong , y. saito , n. miki , t. abe , s. nagata , and y. okumura , `` investigation on cell selection methods associated with inter - cell interference coordination in heterogeneous networks for lte - advanced downlink , '' in _
11th european wireless conference sustainable wireless technologies _ , pp . 1 - 6 | the cloud radio access network ( c - ran ) concept , in which densely deployed access points ( aps ) are empowered by cloud computing to cooperatively support mobile users ( mus ) , to improve mobile data rates , has been recently proposed .
however , the high density of active aps results in severe interference and also inefficient energy consumption .
moreover , the growing popularity of highly interactive applications with stringent uplink ( ul ) requirements , e.g. network gaming and real - time broadcasting by wireless users , means that the ul transmission is becoming more crucial and requires special attention . therefore in this paper ,
we propose a joint downlink ( dl ) and ul mu - ap association and beamforming design to coordinate interference in the c - ran for energy minimization , a problem which is shown to be np hard . due to the new consideration of ul transmission
, it is shown that the two state - of - the - art approaches for finding computationally efficient solutions of joint mu - ap association and beamforming considering only the dl , i.e. , group - sparse optimization and relaxed - integer programming , can not be modified in a straightforward way to solve our problem . leveraging on the celebrated ul - dl duality result
, we show that by establishing a virtual dl transmission for the original ul transmission , the joint dl and ul optimization problem can be converted to an equivalent dl problem in c - ran with two inter - related subproblems for the original and virtual dl transmissions , respectively . based on this transformation ,
two efficient algorithms for joint dl and ul mu - ap association and beamforming design are proposed , whose performances are evaluated and compared with other benchmarking schemes through extensive simulations . cloud radio access network , green communication , uplink - downlink duality , group - sparse optimization , relaxed - integer programming , beamforming . [ section ] [ section ] [ section ] [ section ] [ section ] [ section ] [ section ] |
the detector @xcite is located at the interaction point ( ip ) of the 2 @xcite , where 3.1 @xmath4 positrons from the low - energy ring ( ler ) collide with 9.0 @xmath4 electrons from the high - energy ring ( her ) .
the ler has two beam profile monitors : a visible synchrotron - light monitor ( @xmath5 ) @xcite in a high - coupling region , and an x - ray monitor ( @xmath6 ) @xcite at a separate location .
the her is equipped with one @xmath7 . in both rings , the vertical beam size at the slm
is measured using a companion interferometer @xcite .
the tracking system is used to measure the three - dimensional distribution of @xmath8 vertices @xcite and the transverse - boost distribution @xcite of the muon pairs .
this paper describes a first attempt at characterizing the phase space of the colliding beams by combining all the available information .
the strategy is outlined in fig .
[ fig : outline ] . in each ring ,
profile - monitor data are combined with measured lattice functions to extract the eigenmode emittances @xmath9 and predict the @xmath1 ip spot sizes .
the z - dependence of the luminosity @xmath10 , of the luminous size @xmath11 and of the boost angular divergence @xmath12 allow the determination , under high - luminosity conditions , of the overlap bunch length @xmath13 and of the vertical emittances @xmath14 and effective ip @xmath3-function @xmath15 .
together with the measured specific luminosity @xmath16 , they also provide constraints on the horizontal emittances @xmath17 and @xmath3 functions @xmath18 .
each set of observables ( profile monitors , luminous - region data ) offers a nearly complete description of the ip phase space . comparing results for overlapping parameters
is used to validate the techniques or identify inconsistencies , and combining all measurements should eventually yield a complete , and partially constrained , description .
[ cols="<,^,^,^,^",options="header " , ] [ tbl : capsigmaresults ] spot sizes were also recorded at somewhat higher currents . in the absence of collisions and with an @xmath19 current far below any potential @xmath20-cloud threshold
, the emittances should remain the same .
the observed values are comparable to the low - current results ( table [ tbl : characterisationdataprofileemittances ] ) . the increase in @xmath21 may be caused by thermal orbit distortions .
the already large horizontal her spot size appears to grow with current .
data recorded with high - current colliding beams exhibit a sizeable increase in the vertical @xmath6 and in both @xmath5 spot sizes , which is qualitatively consistent with the dynamic-@xmath3 effect and with beam - beam vertical blowup .
but the observed spot sizes result in inconsistent emittance estimates , presumably highlighting the fact that the additional focusing caused by the beam - beam interaction must imperatively be taken into account in the one - turn matrix .
the size of the luminous ellipsoid @xcite and the transverse - boost distribution @xcite of the colliding electron and positron are measured using @xmath8 events reconstructed in the detector .
the spatial variation of these observables is determined by the emittances , ip @xmath3-functions and waist locations of the colliding beams . in the horizontal plane ,
@xmath18 is much larger than the bunch lengths @xmath22 , so the horizontal beam parameters are only weakly apparent in the @xmath23-dependence of luminous - region variables .
but in the vertical plane , @xmath24 , resulting in an observable @xmath23-dependence of the luminosity , of the angular spread of the boost direction and of the vertical size of the luminous region . fitting an effective ip @xmath3-function to the first two observables and neglecting @xmath25-@xmath26 coupling yields similar results , in the range of 1216 mm @xcite ; fits to @xmath27 yield somewhat higher values , but with larger systematic uncertainties .
an effective vertical emittance can then be extracted from @xmath28 ; the more powerful boost technique allows to determine both the her and ler emittances , yielding @xmath292.59@xmath30-@xmath31 ( again under the no - coupling assumption ) @xcite . combining the emittance and @xmath32 results from the boost measurement yields estimates of @xmath33 in the range of 710@xmath34 m , that displays the expected anticorrelation with the specific luminosity ( fig .
[ fig : capsigyvlsp ] ) . because @xmath35 , the slope of this correlation provides a measurement of @xmath36 .
( boost method ) and @xmath16 measured by the bhabha luminosity monitor .
, width=220 ] further combining @xmath36 with the horizontal luminous size @xmath37 @xcite determines both @xmath19 and @xmath20 horizontal ip spot sizes ( fig .
[ fig : babarx ] ) .
these are then finally combined with the horizontal - boost angular spread @xmath38 and the horizontal angle - position correlation @xmath39 @xcite to extract @xmath40 .
the reduction in horizontal spot size caused by the dynamic-@xmath3 effect is strikingly apparent .
the value of @xmath41 extracted from data before the move to the half - integer , is consistent with low - current phase advance measurements recorded at that time . in the same period
however , the corresponding @xmath42 result is significantly smaller than the phase - advance measurements .
in addition , the horizontal ler emittance implied by fig .
[ fig : babarx ] is surprisingly large , possibly signaling an inconsistency in this preliminary analysis .
functions ( bottom ) in the ler ( open circles ) and the her ( black dots ) , extracted from luminous - region observables measured by .
the dotted line indicates the time when both @xmath25 tunes were moved close to the half - integer , resulting in a sizeable luminosity improvement.,width=307 ]
confronting ip beam sizes extrapolated from the profile monitors with a detailed phase - space characterization based on luminous - region observables would provide invaluable consistency checks , as well as restrict systematic uncertainties .
two main ingredients are required to this effect .
first , beam - beam focusing needs to be taken into account in the one - turn matrix of ref .
@xcite ; guidance can be provided here by beam - beam simulations .
second , @xmath25-@xmath26 coupling at the ip needs to be incorporated in the luminous - region analyses , as its impact has recently been shown @xcite to be significant . | we present a novel method to characterize the @xmath0 phase space at the ip of the slac b - factory , that combines single - beam measurements with a detailed mapping of luminous - region observables .
transverse spot sizes are determined in the two rings with synchrotron - light monitors and extrapolated to the ip using measured lattice functions .
the specific luminosity , which is proportional to the inverse product of the overlap ip beam sizes , is continuously monitored using radiative bhabha events .
the spatial variation of the luminosity and of the transverse - boost distribution of the colliding @xmath1 , are measured using @xmath2 events reconstructed in the detector .
the combination of these measurements provide constraints on the emittances , horizontal and vertical spot sizes , angular divergences and @xmath3 functions of both beams at the ip during physics data - taking .
preliminary results of this combined spot - size analysis are confronted with independent measurements of ip @xmath3-functions and overlap ip beam sizes at low beam current . |
in the present paper , we continue the study of topological properties of the group @xmath2 of all borel automorphisms of a standard borel space @xmath1 .
we consider two topologies , @xmath3 and @xmath4 , on @xmath2 which take their origins in ergodic theory .
they were defined and studied in the context of borel and cantor dynamics in @xcite .
recall that the topology @xmath3 is defined by the base of neighborhoods @xmath11 , where @xmath12 are borel probability measures on @xmath13 and @xmath14 .
obviously , @xmath3 is a direct analogue of the well known uniform topology on the group @xmath15 of all non - singular automorphisms of a measure space generated by the metric @xmath16 .
it is worthwhile to mention that , in fact , @xmath15 is formed by classes of automorphisms coinciding @xmath17-almost everywhere .
it allows one to neglect the behavior of automorphisms on sets of zero measure .
topological properties of @xmath18 were extensively studied in ergodic theory ( see , for example , @xcite ) .
the other topology , @xmath4 , on @xmath2 is defined by neighborhoods @xmath19 where @xmath20 are borel sets .
it was shown in @xcite that @xmath4 can be treated as an analogue of the weak topology @xmath21 which has been also widely used in ergodic theory .
note also that , in the context of cantor dynamics , @xmath4 is equivalent to the @xmath22-topology on the set of all homeomorphisms .
based on this observation , we call @xmath3 and @xmath4 the uniform and weak topologies on @xmath2 respectively .
our goal is to find out which of topological properties known in ergodic theory for @xmath15 hold for @xmath2 with respect to the topologies @xmath3 and @xmath4 .
for instance , it is important for many applications to know dense subsets in @xmath2 which consist of relatively simple " borel automorphisms .
it was shown in [ bdk1 ] that the set @xmath23 of periodic automorphisms is dense in @xmath24 but non - dense in @xmath5 .
therefore one needs to extend @xmath23 to produce a dense subset in @xmath2 with respect to @xmath4 .
the class @xmath25 of smooth borel automorphisms is a natural extension of periodic automorphisms . by definition , @xmath26 is smooth if there exists a borel subset in @xmath13 which intersects every @xmath26-orbit exactly once . in this paper
( see section 2 ) , we prove that the set of smooth borel automorphisms is dense in @xmath27
. this statement has been conjectured by a. kechris .
a number of papers in ergodic theory was devoted to the study of connectedness of @xmath15 in the weak and uniform topologies ( see , for example , @xcite ) .
in particular , it was proved that @xmath15 was path - connected and even contractible .
it is not hard to show that @xmath2 is not path - connected in @xmath3 because there is no continuous path connecting the identity with an automorphism with countable support . at first sight
, the situation seems to be different in borel dynamics .
but if we factorize @xmath2 ( like in ergodic theory ) by a closed normal subgroup consisting of automorphisms whose behavior can be neglected , we do produce a path - connected quotient group . from this point of view , it is natural to say that @xmath28 are equivalent if they are different on an at most countable set .
if @xmath6 denotes the group of all borel automorphisms with countable support , then @xmath29 is a hausdorff topological group with respect to the quotient topologies @xmath8 and @xmath10 . note that the group @xmath6 was also considered in @xcite .
it turns out that such a factorization improves topological properties @xmath2 in both quotient topologies @xmath8 and @xmath10 .
namely , the group @xmath30 becomes path - connected and @xmath31 possesses the so called _ rokhlin property _ , i.e. the action of @xmath9 on itself by conjugation is topologically transitive ( see @xcite for the rokhlin property in cantor dynamics ) .
in fact , we prove an even stronger result by showing that the conjugacy class of every aperiodic smooth automorphism is dense .
most definitions and notions used in this paper are mostly taken from the book @xcite .
we collected in section 1 definitions and facts which are used in the paper .
when we say that @xmath26 is an automorphism of @xmath32 , we always mean that @xmath26 is a borel automorphism
. we will also use the term automorphism " for elements of the quotient group @xmath9 .
let @xmath1 be a _ standard borel space _ with the @xmath33-algebra of borel sets @xmath34 .
this means , by definition , that @xmath32 is ( borel ) isomorphic to a polish space , i.e. a complete separable metric space .
recall several facts about standard borel spaces : ( i ) any two standard borel spaces are borel isomorphic ; ( ii ) if @xmath35 , then @xmath36 is either at most countable or has the cardinality continuum ; ( iii ) if borel sets @xmath37 have the same cardinality , then they are isomorphic .
denote by @xmath2 the group of all borel automorphisms of @xmath1 .
let @xmath38 and @xmath35 . the set @xmath39 is called the _ saturation _ of @xmath36 with respect to @xmath26 and denoted by @xmath40 ( or simply @xmath41 if @xmath26 is clear from the context ) .
a borel set @xmath42 is said to be _ wandering _ with respect to @xmath26 if @xmath43 . a borel set @xmath44
is called a _ complete section _ with respect to @xmath26 ( or simply a @xmath26-section ) if every @xmath26-orbit meets @xmath36 at least once , i.e @xmath45 .
a point @xmath46 from a borel set @xmath36 is called _ recurrent _ with respect to @xmath26 if there exists @xmath47 such that @xmath48 .
denote by @xmath49 and @xmath23 the sets of aperiodic and periodic automorphisms respectively .
we say that a transformation @xmath50 is _ smooth _ if there exists a complete borel section @xmath36 such that @xmath36 meets every @xmath26-orbit exactly once
. we will denote the class of smooth automorphisms by @xmath25 . obviously , any periodic borel automorphism is smooth . on the other hand ,
if @xmath13 is a compact metric space and @xmath26 is an aperiodic homeomorphism of @xmath13 , then @xmath26 can not be smooth .
[ technicals ] we will use the following basic statements taken from @xcite . 1 . _
( poincar recurrence lemma ) _ let @xmath50 and @xmath51 .
then there exists a wandering set @xmath52 such that for each @xmath53 the point @xmath46 returns to @xmath36 for infinitely many positive @xmath54 and also for infinitely many negative @xmath54 .
2 . let @xmath50 .
then @xmath13 can be partitioned into a disjoint union of borel sets @xmath55 where points from @xmath56 , have period @xmath57 and @xmath58 consists of aperiodic points .
let @xmath38 and let @xmath59 , be as in ( b ) .
then there exists a borel set @xmath60 such that @xmath61 and the union is disjoint .
* @xmath26-towers * [ towers_construction ] let @xmath62 .
assume that all points from @xmath36 are recurrent with respect to @xmath26 .
for @xmath63 , define @xmath64 as the smallest positive integer such that @xmath65 and @xmath66 .
let @xmath67 ( some of the @xmath68 s may be empty ) . notice that @xmath69 and @xmath70 consists of pairwise disjoint sets .
we call @xmath71 a @xmath26-_tower _ with base @xmath68 and top @xmath72 .
the set @xmath73 is called the @xmath74-th level of @xmath71 .
the height of @xmath71 is @xmath57 . since @xmath48 for infinitely many positive and negative @xmath54
, we have @xmath75 and @xmath76 the above relation shows that @xmath77 forms a partition of @xmath40 into @xmath26-towers @xmath78 .
notice that @xmath26 maps the union of tops of these towers onto the union of their bases .
given a partition @xmath79 on @xmath13 , a borel set @xmath80 is called a @xmath79-set if it is a union of atoms of @xmath79 .
the next lemma is one of the main tools in the study of borel automorphisms .
it is used in various problems related to finding a suitable approximation of an aperiodic automorphism , in particular , in the proof of the rokhlin lemma @xcite .
[ markers ] let @xmath38 be an aperiodic borel automorphism of a standard borel space @xmath32 , i.e. @xmath81
. then there exists a sequence @xmath82 of borel sets such that + ( i ) @xmath83 + ( ii ) @xmath84 + ( iii ) @xmath85 and @xmath86 are complete @xmath26-sections , @xmath87 , + ( iv ) for @xmath87 , every point in @xmath85 is recurrent , + ( v ) for @xmath87 , @xmath88 , + ( vi ) the base @xmath89 of every non - empty @xmath26-tower constructed over @xmath85 is an uncountable borel set , @xmath87 . for the proof , see ( * ? ? ?
* lemma 4.5.3 ) where ( i ) - ( iii ) have been proved in more general settings of countable borel equivalence relations .
it is shown in ( * ? ? ?
* chapter 7 ) that one can refine the choice of @xmath82 to get ( iv ) and ( v ) .
it is clear that one can remove an at most countable set of points from each @xmath85 to satisfy property ( vi ) .
a sequence of borel sets @xmath82 satisfying [ markers ] is called a _ vanishing sequence of markers_. recall now the definition of the uniform and weak topologies on @xmath2 following @xcite .
let @xmath90 denote the set of all borel probability measures on @xmath13 .
a measure @xmath91 is called continuous ( non - atomic ) if @xmath92 for all @xmath93 .
the dirac measure at @xmath93 is denoted by @xmath94 . for @xmath95 ,
define @xmath96 .
[ top ] the topologies @xmath3 and @xmath4 on @xmath0 are defined , respectively , by the bases of neighborhoods @xmath97 and @xmath98 where @xmath99 here @xmath100 , @xmath101 , @xmath102 , and @xmath103 .
it was shown in @xcite that @xmath2 is a hausdorff topological group with respect to these topologies .
more topological properties of @xmath2 and its subsets can be found in @xcite .
[ e0 ] if in the definition of @xmath3 one takes the set @xmath104 , then the obtained new topology is , in fact , equivalent to @xmath3 .
the proof of this fact is straightforward .
let @xmath6 be defined as a subset of @xmath2 consisting of all automorphisms with countable support , that is @xmath105 if @xmath106 where @xmath107 denotes the cardinality of @xmath36 .
note that @xmath6 is a normal subgroup closed with respect to the topologies @xmath3 and @xmath4 , see the proposition below .
therefore @xmath108 is a hausdorff topological group with respect to the quotient topologies @xmath8 and @xmath10 .
considering elements from @xmath9 , we will identify borel automorphisms which are different on a countable set , that is @xmath109 if @xmath110 .
in other words , @xmath109 if there exists @xmath111 such that @xmath112 .
this identification corresponds to the well known approach used in measurable dynamics when two automorphisms are also identified if they are different on a set of measure 0 .
@xmath6 is a normal closed subgroup in @xmath2 with respect to the topologies @xmath3 and @xmath4 . _
_ it is obvious that @xmath6 is a normal subgroup in @xmath2 , so it is enough to prove that it is closed in @xmath3 and @xmath4 . to do this
, suppose that there exists an automorphism @xmath113 .
then for any neighborhood @xmath114 there exists an automorphism @xmath115 , that is @xmath116 for all @xmath74 . since @xmath117 , we have that @xmath118 is uncountable .
let @xmath119 be a continuous measure on @xmath13 such that @xmath120 .
consider a neighborhood @xmath121 of @xmath122 .
then for @xmath123 we have that @xmath124 but @xmath125 , therefore @xmath126 and @xmath127 , a contradiction .
assume now that @xmath128 .
then @xmath129 is uncountable and @xmath122-invariant .
let @xmath130 be an uncountable borel subset of @xmath131 such that @xmath132 .
then for every @xmath133 we have that @xmath134 .
in particular , if @xmath135 , then we obtain that @xmath136 acts non - trivially on the uncountable set @xmath130 .
this contradicts the definition of @xmath6 .
@xmath137 + the bases of neighborhoods for @xmath8 and @xmath10 consists of the sets @xmath138 and @xmath139 , respectively .
the next proposition shows that , in fact , @xmath8 and @xmath10 are generated by neighborhoods @xmath140 and @xmath141 with continuous measures @xmath142 and uncountable sets @xmath143 . [ coun ] given a @xmath8-neighborhood @xmath144 and a @xmath10-neighborhood @xmath145 , there exist neighborhoods @xmath146 and @xmath147 of @xmath8 and @xmath10 , respectively , such that @xmath148 where measures @xmath149 are continuous and borel sets @xmath150 are uncountable . _
proof_. consider the countable set @xmath151 where @xmath152 .
let @xmath153 and assume that @xmath154 , @xmath155 .
define @xmath156 where @xmath157 . clearly , @xmath158 is a non - atomic borel probability measure on @xmath13 .
it remains to show that @xmath159 is a subset of @xmath140 . to do this
, it suffices to check that for every @xmath160 there exists @xmath161 such that @xmath162 .
let @xmath163 be the countable group of automorphisms of @xmath13 generated by @xmath26 and @xmath122 .
let @xmath164 be the @xmath163-orbit of @xmath36 .
define @xmath165 obviously , @xmath166 .
since @xmath167 and @xmath168 , we have that for @xmath155 , @xmath169 notice that if @xmath170 for some @xmath74 , then for any @xmath171 there exists @xmath166 such that @xmath172 .
the proof for the topology @xmath10 is similar.@xmath137
in this section , we prove that the @xmath4-closure of the set @xmath173 of smooth automorphisms is the entire group @xmath2 . moreover , it is shown that @xmath174 has the rokhlin property .
[ smooth ] @xmath175 .
moreover , each @xmath4-neighborhood of an aperiodic automorphism necessarily contains an aperiodic smooth automorphism . _ proof .
_ let @xmath50 . obviously , it suffices to consider the case when @xmath26 is aperiodic .
take a @xmath4-neighborhood @xmath176 of @xmath26 . without loss of generality
, we can assume that the sets @xmath177 form a partition of @xmath13 .
show that there exists a smooth aperiodic automorphism @xmath178 . for @xmath179 ,
consider the borel sets @xmath180 suppose that the collection @xmath181 contains exactly @xmath182 non - empty sets .
denote them by @xmath183 , @xmath184 .
then @xmath185 is a partition of @xmath13 which refines @xmath186 .
let @xmath187 here and below @xmath188 stands for @xmath189 .
clearly , @xmath190 is a partition of @xmath13 into borel sets . without loss of generality
, we may assume that all sets @xmath191 are non - empty .
for each @xmath192 , find a wandering subset @xmath193 ( see [ technicals ] ) such that all points from @xmath194 are recurrent .
therefore , by [ towers_construction ] , we can find a partition @xmath195 of @xmath196 into pairwise disjoint @xmath26-towers @xmath197 such that the union of bases of these towers is @xmath198 .
for short , we will write @xmath199 for a tower @xmath79 , denote by @xmath200 and @xmath201 its base and height , respectively . by construction , we have that @xmath202 and @xmath203 for some @xmath204 . since each @xmath205 , can be represented as a disjoint union of @xmath206 and @xmath196 , we obtain the partition @xmath207 the automorphism @xmath26 restricted to @xmath208 is smooth since @xmath209 is a wandering set for @xmath26 .
define @xmath210 on @xmath211 . to complete the proof , we need to define @xmath122 on @xmath212 such that @xmath213 for all @xmath214 .
fix the set @xmath196 and let @xmath215 .
then @xmath216 is countable and therefore @xmath26 , restricted to @xmath217 , is smooth .
set @xmath218 on @xmath219 .
for @xmath220 , denote by @xmath221 .
setting @xmath222 , we get a disjoint union @xmath223 note that the cardinality of each tower @xmath224 is continuum .
now , we define the automorphism @xmath122 on each tower @xmath225 such that @xmath226 and @xmath122 coincides with @xmath26 on each level of the tower @xmath227 except the top of @xmath227 . to do this
, we write down the base @xmath228 of @xmath227 as a disjoint union @xmath229 with uncountable borel sets @xmath230 .
let @xmath231 be an arbitrary borel isomorphism between @xmath232 and @xmath233 ( one can assume that @xmath234 by ( * ? ? ?
* theorem 7.25 ) ) . define @xmath235 for @xmath236 and @xmath237 for @xmath238 .
then @xmath122 is defined everywhere on @xmath13 . to prove that @xmath178 , we need to show that @xmath239 .
write down @xmath240 as a disjoint union of sets @xmath241 and @xmath130 where @xmath242 and @xmath243 .
it follows from the construction that if @xmath244 , then @xmath245 and @xmath246 .
therefore , by definition of @xmath122 , we have that @xmath247 .
it is clear that @xmath248 on @xmath241 .
the proof is complete .
@xmath137 [ den ]
\(1 ) we note that the set @xmath25 is dense in @xmath249 .
it follows from the fact that the set of periodic automorphisms is dense in @xmath249 ( * ? ? ?
* corollary 2.6 ) . on the other hand
, @xmath250 is nowhere dense in @xmath249 by ( * ? ? ?
* theorem 2.8 ) .
\(2 ) the set of aperiodic smooth automorphisms is not dense in @xmath27 because @xmath49 is a closed subset in @xmath5 ( * ? ? ?
* theorem 2.8 ) .
b. miller proved the following result @xcite which may be used to simplify the proof of theorem [ smooth ] : _ suppose that @xmath13 is a polish space , @xmath251 is a borel automorphism , and @xmath252 is a partition of @xmath13 into borel sets .
then there is a countable @xmath26-invariant set @xmath253 and an aperiodic smooth borel automorphism @xmath254 such that @xmath255 for every @xmath87 . _ [ nowd ]
@xmath23 is a closed nowhere dense subset of @xmath2 with respect to @xmath4 . _
proof_. we first show that the set @xmath23 is closed .
if we assume that there exists an automorphism @xmath256 , then @xmath26 must have an aperiodic point @xmath257 .
define @xmath258 and consider the @xmath4-neighborhood @xmath259 .
then @xmath42 necessarily contains a periodic automorphism @xmath260 such that @xmath261 . since @xmath262
, we obtain that @xmath263 for all @xmath264 .
therefore the point @xmath265 must be aperiodic for @xmath260 , a contradiction .
let @xmath266 be a @xmath4-neighborhood of a periodic automorphism @xmath260 .
we can assume that the sets @xmath267 form a partition of @xmath13 .
we will first show that @xmath268 contains a non - periodic automorphism . by [ technicals ] , @xmath13 is partitioned into @xmath260-towers @xmath269 such that @xmath260 has period @xmath57 on the set @xmath270 .
one can refine the partition @xmath271 to produce a new partition @xmath227 such that every @xmath272 is a @xmath227-set .
let @xmath273 be a @xmath260-tower from @xmath227 with uncountable base .
as in [ smooth ] , we can find an aperiodic automorphism @xmath26 defined on @xmath274 such that @xmath275 and @xmath276 .
define @xmath26 on @xmath277 by setting @xmath278 .
we see that @xmath279 for all @xmath74 , i.e. @xmath280 .
it is clear that there exists a borel set @xmath281 such that @xmath282 .
then @xmath283 contains no periodic automorphism .
thus , @xmath284 and we are done .
@xmath137 in contrast to [ den ]
, the situation for the quotient group @xmath9 is different .
it turns out that the set @xmath250 is a dense subset in @xmath174 .
[ dense ] the set of aperiodic smooth automorphisms is dense in @xmath174 . _
_ by [ smooth ] , we only need to show that each neighborhood @xmath285 of @xmath286 contains an aperiodic smooth automorphism @xmath122 .
it follows from [ technicals ] that @xmath13 is decomposed into an at most countable collection of @xmath260-invariant towers @xmath287 . by [ coun ]
, we can assume that all @xmath71 s are uncountable borel sets .
let @xmath288 be a refinement of @xmath289 , obtained by cutting the towers from @xmath289 , such that each @xmath290 is a @xmath288-set .
we partition each @xmath291 into a disjoint union @xmath292 of @xmath260-towers @xmath293 such that the base of @xmath293 is uncountable .
to define @xmath122 on @xmath79 , we apply the method used in the proof of [ smooth ] . for fixed @xmath294
, we set @xmath295 everywhere except the top of @xmath293 and set @xmath296 on the top where @xmath231 is a borel isomorphism mapping the top of @xmath293 onto the base of @xmath297 .
then @xmath122 is defined everywhere on @xmath13 and is aperiodic .
note that every @xmath291 is @xmath122-invariant . since all towers @xmath293 are of the same height , we have that @xmath122 maps @xmath79-atoms onto themselves .
it follows from this observation that @xmath298 , that is @xmath299 .
@xmath137 [ xx ] @xmath300 . _
proof_. this result is an easy consequence of [ dense ] . @xmath137
the next statement proves a borel version of the rokhlin property for @xmath174 .
note that this property was considered in the settings of measurable and cantor dynamics in @xcite .
( the rokhlin property ) the action of @xmath9 on itself by conjugation is transitive with respect to the topology @xmath10 .
moreover , @xmath301 holds for any @xmath302 . _
proof_. to prove this result it suffices to use [ xx ] together with the simple fact that any two aperiodic smooth automorphisms are conjugate .
the famous rokhlin lemma on approximation aperiodic automorphisms in the uniform topology was proved in the context of borel dynamics in @xcite .
we formulate here this statement in the following form .
[ rokhlin ] let @xmath26 be an aperiodic borel automorphism of a standard borel space @xmath1 and let @xmath303 @xmath102 , and @xmath304 .
then there exists a borel partition of @xmath13 into @xmath26-towers @xmath305 such that the following properties hold : \(i ) the height @xmath201 of each @xmath26-tower @xmath306 is greater than @xmath307 .
\(ii ) @xmath308 , @xmath309 , where @xmath200 is the base of @xmath306 .
_ the proof can be deduced from ( * ? ? ?
* theorem 2.5 ) .
@xmath137 let @xmath310 .
then for any @xmath311 and any @xmath3-neighborhood @xmath312 , there exists @xmath100 such that @xmath313 .
in other words , @xmath314 . _ proof
. _ apply [ rokhlin ] for @xmath311 , @xmath315 , @xmath316 , and @xmath317 .
we obtain a partition of @xmath13 into @xmath136-towers @xmath305 satisfying ( i ) , ( ii ) .
choose a sufficiently large @xmath318 such that @xmath319 , @xmath320 , where @xmath68 is the set supporting the tower @xmath71 .
therefore , we have that @xmath321 for all @xmath320 , where @xmath322 is the base of @xmath71 and @xmath323 is its height .
since @xmath122 is aperiodic , we can find @xmath324 disjoint @xmath122-towers @xmath325 such that the height of @xmath326 is @xmath323 and @xmath326 has the same cardinality as @xmath71 for all @xmath327 .
denote by @xmath328 the base of @xmath326 and let @xmath329 be the support of @xmath326 .
let @xmath330 be a borel isomorphism which maps @xmath322 onto @xmath328 , @xmath327 , and let @xmath331 be a borel isomorphism which sends @xmath332 onto @xmath333 .
define the automorphism @xmath26 as follows : @xmath334 then @xmath26 is defined everywhere on the set @xmath13 .
it is not hard to see that @xmath335 hence @xmath336 , @xmath320 , and therefore @xmath313 .
in this section , we prove that @xmath9 is path - connected in the topology @xmath8 .
we first show that the group @xmath2 does not possess this property .
_ let @xmath260 be an arbitrary involution in @xmath2 , that is @xmath337 and @xmath338 for all @xmath93 .
we will show that @xmath260 can not be connected with the identity @xmath339 by a continuous path , i.e. there exists no continuous map @xmath340\to { aut(x,\mathcal{b})}$ ] such that @xmath341 .
assume that the converse is true and let @xmath342 be such a path .
choose @xmath343 in @xmath13 such that @xmath344 .
consider the @xmath3-neighborhood @xmath345 of @xmath260 .
notice that @xmath346 contains only those automorphisms from @xmath2 which map @xmath347 to @xmath348 and @xmath348 to @xmath347 . since , by assumption , @xmath342 is continuous , there exists @xmath349 such that @xmath350)\subset u$ ] .
set @xmath351\,:\ , f(s)\in u,\;t\leq s\leq 1\}.\ ] ] clearly , @xmath352 .
consider now the neighborhood @xmath353 of @xmath354 .
if @xmath355 , then there exist @xmath356 and @xmath357 such that @xmath358 and @xmath359)\subset u(f(t_0^*))$ ] .
we obtain that @xmath360 and therefore , @xmath361 and @xmath362 .
a similar relation holds for @xmath363 .
thus , @xmath364 and therefore @xmath365 can not be the infimum . hence @xmath366 and @xmath367 , which is a contradiction .
@xmath137 [ sect ] let @xmath38 and let @xmath35 be a complete @xmath26-section such that every point from @xmath36 is recurrent .
if a borel set @xmath368 contains @xmath36 , then @xmath368 is also a complete @xmath26-section which consists of recurrent points . _
proof_. we first prove separately that every periodic automorphism @xmath260 and every aperiodic automorphism @xmath26 can be connected with the identity by a continuous path ( see [ per ] and [ aper ] respectively ) . by [ technicals ] , these two results will give the proof for any automorphism .
recall that , by [ coun ] , it is sufficient to deal with continuous measures only .
_ by [ technicals ] , we have the decomposition of @xmath374 where @xmath375 is a @xmath260-tower . without loss of generality , we can assume that the @xmath322 s are uncountable borel sets and therefore they all are isomorphic to the unit interval @xmath376 .
let @xmath377 be a borel isomorphism .
for each @xmath322 , define the map @xmath378\rightarrow \mathcal{b}\upharpoonright b_k$ ] by @xmath379 define now the path @xmath371\rightarrow { aut_0(x,\mathcal{b})}$ ] as follows : @xmath383 it is clear that @xmath372 and @xmath373 and we need to show only that @xmath384 is continuous . to do this ,
fix @xmath385 $ ] and consider the map @xmath386 on [ 0,1 ] where @xmath381 is non - atomic and @xmath241 is defined in [ e0 ] .
[ aper ] let @xmath394 be an arbitrary aperiodic automorphism .
then there exists a continuous map @xmath395 \rightarrow ( { aut_0(x,\mathcal{b})},\tau_0)$ ] such that @xmath396 and @xmath397 .
moreover , for all @xmath398 , the automorphism @xmath399 is periodic .
_ by [ markers ] , choose a vanishing sequence of markers @xmath400 with @xmath401 . without loss of generality , we can assume that the set @xmath402 is uncountable for all @xmath54 .
take a sequence of real numbers @xmath403 such that @xmath404 and @xmath405 .
let @xmath406 be a borel isomorphism .
define the function @xmath407 \rightarrow \mathcal{b}$ ] as follows : @xmath408 observe that for each @xmath409 there exists @xmath410 such that @xmath411 . by [ sect ]
, we get that @xmath412 is a @xmath26-section which consists of recurrent points .
we also notice that @xmath413 for any non - atomic @xmath381 and @xmath414 $ ] .
fix @xmath416 $ ] . by [ towers_construction ] , define the set @xmath417 , where @xmath418 and @xmath419 .
clearly , @xmath420 may be empty for some @xmath54 . by construction
, the entire space @xmath13 is partitioned into @xmath26-towers with bases @xmath420 .
it remains to prove that the map @xmath422 sending @xmath423 $ ] to @xmath9 is continuous .
note that for every @xmath264 the family @xmath424 is continuous , i.e. for a non - atomic measure @xmath425 , @xmath426 indeed , this fact follows from continuity of @xmath412 and from the relation @xmath427 next , show that for @xmath428 $ ] we have @xmath429 let @xmath430 for definiteness .
suppose that @xmath431 .
then @xmath46 belongs to a @xmath26-tower constructed over @xmath432 , that is @xmath433 for some @xmath264 and @xmath434 . if @xmath435 , then by construction of @xmath436 and @xmath437 , we have that @xmath438 .
therefore , @xmath439 .
fix @xmath414 $ ] .
consider a neighborhood @xmath440 of @xmath441 where measures @xmath442 are continuous . by definition of @xmath441
, we have that @xmath443 and these unions are disjoint .
take @xmath444 such that @xmath445 for @xmath446 . by continuity of @xmath424
, we can find a neighborhood @xmath447 $ ] such that for any @xmath448 one has @xmath449 for @xmath446 , @xmath450 , and @xmath434 .
therefore , we have @xmath454 for all @xmath448 .
thus , for all @xmath448 , we obtain that @xmath455 this means that @xmath456 and the proof is completed .
@xmath137 + let @xmath38 and let @xmath457 denote the @xmath26-orbit of @xmath93 .
recall the definition of the _ full group _
@xmath458 $ ] generated by @xmath459 : @xmath460 = \ { \gamma \in { aut(x,\mathcal{b})}\ \vert\
\gamma x \in orb_t(x),\ \forall x\in x\}.\ ] ] then every @xmath461 $ ] defines a borel function @xmath462 such that @xmath463 .
it follows easily from [ top ] that @xmath464 $ ] is @xmath3-closed in @xmath2 .
_ proof_. the proofs of [ per ] and [ aper ] show that the constructed paths @xmath384 and @xmath399 connecting the identity with @xmath26 belong to the full group @xmath464 $ ] .
@xmath137 + _ acknowledgement_. we are thankful to a.h .
dooley , a. kechris , j. kwiatkowski , and b. miller for numerous discussions of these results .
the first - named author thanks the university of new south wales for the warm hospitality and the australian research council for its support . | let @xmath0 be the group of all borel automorphisms of a standard borel space @xmath1 .
we study topological properties of @xmath2 with respect to the uniform and weak topologies , @xmath3 and @xmath4 , defined in @xcite .
it is proved that the class of smooth automorphisms is dense in @xmath5 .
let @xmath6 denote the group of borel automorphisms with countable support .
it is shown that the topological group @xmath7 is path - connected with respect to the quotient topology @xmath8 .
it is also proved that @xmath9 has the rokhlin property in the quotient topology @xmath10 , i.e. the action of @xmath9 on itself by conjugation is topologically transitive . |
accretion disks are believed to lose mass via powerful outflows in many astrophysical environments such as active galactic nuclei ( agn ) ; many types of interacting binary stars , non - magnetic cataclysmic variables ( nmcvs ) , for instance ; and young stellar objects ( ysos ) .
magnetically driven winds from disks are the favored explanation for the outflows in many of these environments .
blandford & payne 1982 ( see also pelletier & pudritz 1992 ) showed that the centrifugal force can drive a wind from the disk if the poloidal component of the magnetic field , @xmath1 makes an angle of @xmath2 with respect to the normal to the disk surface .
generally , centrifugally - driven mhd disk winds ( magnetocentrifugal winds for short ) require the presence of a sufficiently strong , large - scale , ordered magnetic field threading the disk and the poloidal magnetic field to be comparable to the toroidal magnetic field , @xmath3 ( e.g. , cannizzo & pudritz 1988 , pelletier & pudritz 1992 ) . additionally , magnetocentrifugal winds require some thermal assistance to flow freely and steadily from the surface of the disk , to pass through a slow magnetosonic surface ( e.g. , blandford & payne 1982 )
. many authors have studied magnetocentrifugal winds from a keplerian disk ( e.g. , ouyed & pudritz 1997 , ustyugova et al .
1999 , krasnopolsky , li & blandford 1999 , and references therein ) .
these studies are either analytic , looking for stationary , often self - similar solutions or numerical , looking for both stationary and time - dependent solutions .
however in numerical simulations of magnetocentrifugal winds , the lower boundary is between the slow magnetosonic surface and the alfv@xmath4n surface .
this specification requires setting the mass loss rate in advance ( e.g. , bogovalov 1997 ) .
simulations of magnetocentrifugal winds considering the whole disk ( even regions below the slow magnetosonic surface ) and not requiring an _ ad hoc _ mass loss rate , are only now becoming feasible .
this requires , however , an accurate treatment of the radiation and gas pressure effects , among other physical processes . thermal expansion and the radiation force have been suggested as other mechanisms that can drive disk winds .
these mechanisms can produce powerful winds without the presence of a magnetic field .
winds are likely thermally driven in x - ray - irradiated accretion disks in systems such as x - ray binaries and agns ( e.g. , begelman , mckee & shields 1983 , woods et al .
these winds require the gas temperature to be so high that the gas is not gravitationally bound . in such cases ,
the radiation driving is probably not important because the gas is fully ionized , at least in the hottest regions , and the radiation force is only due to electron scattering .
however in the regions where gas is cooler and not fully ionized the radiation force will be enhanced by spectral lines and play an important role in controlling the dynamics of the flow .
in fact murray et al . (
1995 ) designed a line - driven disk wind model specifically for agns .
radiation - driven disk winds have been extensively modeled ( e.g. , pereyra , kallman & blondin 1997 ; proga , stone & drew 1998 , hereafter psd i ; proga 1999 ; proga , stone & drew 1999 , hereafter psd ii ) .
these recent studies showed that radiation pressure due to spectral lines can drive winds from luminous disks .
this result has been expected ( e.g. , vitello & shlosman 1988 ) .
these studies , in particular those by psd i , also showed some unexpected results .
for example , the flow is unsteady in cases where the disk luminosity dominates the driving radiation field . despite the complex structure of the unsteady disk wind , the time - averaged mass loss rate and terminal velocity scale with luminosity , as do steady flows obtained where the radiation is dominated by the central object . to calculate the line force for disk winds , psd i adopted the method introduced by castor , abbott & klein ( 1975 , hereafter cak ) and further developed by friend & abbott ( 1986 ) , pauldrach , puls & kudritzki ( 1986 ) for one - dimensional radial flows within the context of a wind from hot , luminous ob stars . to extend the cak method to model multi - dimensional disk winds
, it is necessary to accommodate the effects of the three - dimensional velocity field and the direction - dependent intensity .
owocki , cranmer & gayley ( 1996 ) showed that these effects can lead to qualitatively different results compared to those obtained from a one - dimensional treatment in the case of a rapidly rotating star .
the most difficult aspect of calculating the line force due to a disk is in the evaluation of the integral involving the velocity gradient tensor over the entire solid angle occupied by the radiating surface .
psd i used an angle - adaptive quadrature to ensure an accurate result .
however , computational limitations required that they simplified the integrand , retaining only the dominant terms in the velocity gradient tensor .
psd ii generalized the psd i method and introduced a new quadrature that avoids any simplification of the integrand .
this allowed them to evaluate the radiation force for completely arbitrary velocity fields within the context of the cak formalism .
they applied this method to recalculate several disk wind models first discussed in psd i. these more physically accurate models show that psd i s more approximate method was very robust .
the psd ii calculations predict total mass loss rates and velocities marginally different from those published in psd i. additionally , psd ii find that models which display unsteady behavior in psd i are also unsteady with the new method .
the largest change caused by the new method is in the disk - wind opening angle : winds driven only by the disk radiation are more polar with the new method while winds driven by the disk and central object radiation are typically more equatorial . in this paper
, we will numerically check how strong , ordered magnetic fields can change disk winds driven by the line force for a given disk luminosity .
we assume that the transport of angular momentum in the disk is dominated by local disk viscosity , for instance due to the local shear instability in weakly magnetized disk ( balbus & hawley 1997 ) . instead of adding the magnetic fields to the psd ii model and solving consistently the equations of mhd ,
we simply start by adding some of the effects due to the magnetic fields , namely ( 1 ) the azimuthal force so the wind conserves the specific angular velocity along the streamlines and ( 2 ) the force perpendicular to the field lines so the geometry of the disk wind is controlled by the magnetic field .
we thus adopt the popular concept that the magnetic field dominates outside the disk and at least near the disk surface , one can think of the magnetic field lines as rigid wires that control any flow ( e.g. , blandford & payne 1982 ; pelletier & pudritz 1992 ) .
such an approach is clearly simplistic and can be applied only to sub - alfv@xmath4nic flows .
nevertheless , the results presented here provide a useful exploratory study of line - driven disk winds in the presence of a strong , large - scale field magnetic threading the disk .
the organization of the paper is as follows . in section 2
we describe our numerical calculations ; in section 3 we present our results ; in section 4 we conclude with a brief discussion .
to calculate the structure and evolution of a wind from a disk , we solve the equations of hydrodynamics @xmath5 @xmath6 where @xmath7 is the mass density , @xmath8 the velocity , @xmath9 the gravitational acceleration of the central object , @xmath10 the total radiation force per unit mass , and @xmath11 the total magnetocentrifugal force .
the term with @xmath11 ensures that gas flows along an assumed direction and conserves its specific angular velocity .
the gas in the wind is isothermal with a sound speed @xmath12 .
we solve these equations in spherical polar coordinates @xmath13 . the geometry and assumptions needed to compute the radiation field from the disk and central object are as in psd ii ( see also psd i ) .
the disk is flat , keplerian , geometrically - thin and optically - thick .
we specify the radiation field of the disk by assuming that the temperature follows the radial profile of the so - called @xmath14-disk ( shakura & sunyaev 1973 ) , and therefore depends only on the mass accretion rate in the disk , @xmath15 , and the mass and radius of the central object , @xmath16 and @xmath17 . in particular , the disk luminosity , @xmath18 . in models where the central object radiates ,
we take into account the stellar irradiation of the disk , assuming that the disk re - emits all absorbed energy locally and isotropically .
we express the central object luminosity @xmath19 in units of the disk luminosity @xmath20 .
see psd i and psd ii for further details .
our numerical algorithm for evaluating the line force is described in psd ii . for a rotating flow
, there may be an azimuthal component to the line force even in axisymmetry .
however we set this component of the line force to zero because it is rather weak as compared to other components and is not of great importance ( e.g. , psd ii ) .
then we assume that the rotational velocity @xmath21 is determined by the azimuthal component of the magnetocentrifugal force . the description of our calculation of the magnetocentrifugal force follows .
we choose the simplest geometry of the magnetic field and the flow : the poloidal component of the magnetic field , @xmath22 and of the velocity , @xmath23 are parallel to one another , and the inclination angle @xmath24 between @xmath23 and the disk midplane is fixed for all locations and times . in other words , we constrain the geometry of the flow to straight cones of pre - specified , radius- and time - independent opening angle . in practice , we impose this geometry in the following way : ( i ) we evaluate the physical accelerations and the curvature terms in eq .
2 , hereafter collectively referred to as @xmath25 where @xmath26 represents the curvature terms appearing on the left hand side of eq . 2 when @xmath27 is expressed in the spherical polar coordinate system ( e.g. , shu 1992 ) : @xmath28 ( ii ) we calculate the component of @xmath29 perpendicular to a streamline : @xmath30 where @xmath31 ( iii ) and finally we subtract @xmath32 from @xmath29 so the total remaining force is tangent to the streamline . the disk outflow that conserves its specific angular velocity ,
@xmath33 along a streamline is simply the flow that corotates with the disk , i.e. , @xmath34 is constant and equals to the disk rotational velocity , @xmath35 at the footpoint of the streamline on the disk at the radius , @xmath36 . to ensure corotation of the wind
, we introduce the azimuthal force : @xmath37 summarizing , our magnetocentrifugal force is @xmath38 as in psd ii , we use the zeus-2d code ( stone & norman 1992 ) to numerically integrate the hydrodynamical equations ( 1 ) and ( 2 ) . we did not change the transport step in the code and so the changes we made are just to the source step as outlined above .
we explore disk wind models for three cases : ( i ) a wind corotates with the disk ( @xmath39 and @xmath40 ) ; ( ii ) the wind geometry is fixed by adopting a fixed inclination angle between the poloidal component of the velocity and the disk midplane ( @xmath41 and @xmath42 ) ; and ( iii ) a combination of ( i ) and ( ii ) ( @xmath41 and @xmath43 ) . in case i with @xmath44 , we decided to put an additional constraint on the radial velocity to produce outward flows . near the disk midplane , fluctuations of the density and velocity can occur ( e.g. , psd i ) .
additionally , the radial component of the radiation force is negative near the disk midplane and the central object ( e.g. , icke 1980 , psd i ) .
thus @xmath45 can be less than zero ; gravity can dominate and pull the flow toward the center . to avoid this
we apply the condition : @xmath46 . this condition does not change the mass loss rate , but does reduce the wind opening angle . to produce transonic flows in cases ii and iii
, we need to increase the density along the disk midplane , @xmath47 , from @xmath48 ( as used by psd i ) up to @xmath49 . for low @xmath47 , the wind velocity does not depend on @xmath7 ,
the mass loss rate is proportional to @xmath47 and may be higher than the mass accretion rate . for @xmath50
the mass loss rate is dramatically lower , does not change with @xmath47 , and the flow is subsonic near the disk midplane .
these are desired properties of the outflow ; we expect gas near the midplane to be nearly in hydrostatic equilibrium .
equally important , we assume that the disk is in a steady state so the mass loss rate should be lower than the mass accretion rate
. in reality , gas near the disk midplane ( inside the disk ) is in hydrostatic equilibrium for the densities lower than those we assumed because the radiation force is lower as the radiation becomes isotropic .
we would like to stress that we treat the region very close to the midplane as a boundary condition and do not attempt to model the disk interior .
as in psd i and psd ii , we calculate disk winds with model parameters suitable for a typical nmcv ( see psd i s table 1 and our table 1 ) .
we vary the disk and central object luminosity and the inclination angle .
we hold all other parameters fixed , in particular the parameters of the cak force multiplier : @xmath51 , @xmath52 and @xmath53 ( see psd i ) .
nevertheless we can use our results to predict the wind properties for other parameters and systems such as agn by applying the dimensionless form of the hydrodynamic equations and the scaling formulae as discussed in psd i and proga ( 1999 ) .
psd i and psd ii showed that radiation - driven winds from a disk fall into two categories : 1 ) intrinsically unsteady with large fluctuations in density and velocity , and 2 ) steady with smooth density and velocity .
the type of the flow is set by the geometry of the radiation field , parametrized by @xmath54 : if the radiation field is dominated by the disk ( @xmath55 ) then the flow is unsteady , and if the radiation is dominated by the central object ( @xmath56 ) then the flow is steady .
the geometry of the radiation field also determines the geometry of the flow ; the wind becomes more polar as @xmath54 decreases .
however the mass - loss rate and terminal velocity are insensitive to geometry and depend more on the total system luminosity , @xmath57 .
we recalculated some of the psd ii models to check how inclusion of the magnetocentrifugal force will change line - driven disk winds .
figure 1 compares the density in the wind in two models from psd ii where they used a generalized cak method ( top panels ) , models using psd ii s method but conserving specific angular velocity our case i ( middle panels ) , and models using psd ii s method , conserving specific angular velocity and having fixed inclination angle of the stream lines our case iii ( bottom panels ) .
models in case iii ( bottom panels ) are with the inclination angles approximately equal to those in the psd ii case ( the top panels ) .
the left column shows the results for a model with @xmath58 and @xmath44 , while the right column shows the results with @xmath59 and @xmath60 .
the former corresponds to the fiducial @xmath44 model while the latter corresponds to the fiducial @xmath60 model discussed in detail in psd i and psd ii .
we start with describing the @xmath44 wind model ( left column panels ) .
the most striking yet expected difference between psd ii s case ( top ) and our case i ( middle ) is a decrease of the wind opening angle , @xmath61 from @xmath62 to @xmath63 when the specific angular momentum conservation is replaced by the specific angular velocity conservation .
this result is caused by a much stronger centrifugal force in our case i than in the psd ii case which is also reflected in an increase of the radial velocity by more than one order of magnitude ( table 1 ) . another big difference between the two cases is in the mass loss rate , @xmath64 which increases from @xmath65 to @xmath66 .
an unchanged property in these two cases is that the wind is unsteady .
the quantitative changes between psd i s case ( top ) and case iii ( bottom ) are similar or smaller than the changes described above ( see table 1 ) .
the most pronounced change caused by holding the inclination angle fixed is that the wind becomes steady .
however this is not surprising because models with the fixed , rigid geometry are pseudo one - dimensional and the flow has one degree of freedom in space . in psd
i and psd ii models where the wind geometry is calculated self - consistently , a strong radial radiation force is required to produce a steady outflow . the x=1 wind model ( top right hand panel of figure 1 ) is an example of a steady outflow from psd ii .
this model remains steady in the two other cases shown in the figure and in case i listed in table 1 ( run ic ) .
the model changes from case to case mainly in the radial and rotational velocities which increase when the azimuthal force is added ( middle and bottom panel ) .
for example , the radial velocity increased from 3500 @xmath67 in psd ii s run c ( top panel ) to 28000 @xmath67 in our run iiic ( bottom panel ) . as in the x=0 wind model , adding the azimuthal force also reduces the wind opening angle ( middle panel ) .
however an increase in the mass loss rate from psd ii s case ( top panel ) to our case ii ( bottom panel ) is only a factor of @xmath68 .
the mass loss rate is practically the same in case ii and case iii ( run iic and iiic ) .
models presented in figure 1 both for @xmath44 and @xmath60 suggest that an increase in the mass loss rate due to conservation of the specific angular velocity decreases with the disk luminosity . our other results , shown in table 1 , confirm this .
for example , the x=0 model ib with @xmath69 has @xmath64 higher than the corresponding psd ii s model b by a factor of @xmath70 while an increase of @xmath71 for models with @xmath72 is by a factor @xmath73 ( i.e. , runs a and ia ) .
the mass loss rate of the line - driven disk wind also increases by imposing the wind geometry such that @xmath74 as comparison of our results in case ii and from psd ii reveals .
once the geometry is fixed the mass loss rate does not change with adding the azimuthal force the corresponding models in our case ii and iii have the same @xmath71 .
purely line - driven x=0 winds have streamlines perpendicular to the disk midplane near the midplane because the total horizontal force is negligible in comparison with the vertical force due to lines .
this property of disk winds has been assumed in analytic studies of line - driven disk winds ( e.g. , vitello & shlosman 1988 ) .
we calculated a model assuming @xmath75 ( run iia ) and find that the mass loss rate is very similar to the corresponding psd ii s run a. this confirms then that the mass loss rate in line - driven disk winds for @xmath44 is determined in the regions where the matter flows still perpendicularly to the disk .
we have studied winds from accretion disks with very strong organized magnetic fields and the radiation force due to lines .
we use numerical methods to solve the two - dimensional , time - dependent equations of hydrodynamics .
we have accounted for the radiation force using a generalized multidimensional formulation of the sobolev approximation . to include the effects of the strong magnetic field we have added to the hydrodynamic equation an approximate magnetocentrifugal force .
our approach of treating the latter allows us to study cases where the wind geometry is controlled entirely by the magnetic field with the lines of the poloidal component of the field , @xmath76 being straight and inclined at a fixed angle to the disk midplane where the specific angular velocity is conserved along the streamlines due to the magnetic azimuthal force , or both .
our simulations of line - driven disk winds show that inclusion of the azimuthal force , @xmath77 which makes the wind corotate with the disk , increases the wind mass loss rate significantly only for low disk luminosities .
as expected such winds have much higher velocities than winds with zero azimuthal force . fixing the wind geometry
effectively reduces the flow from two - dimensional to one - dimensional .
this in turn stabilizes winds which are unsteady when the geometry is allowed to be determined self - consistently .
the inclination angle between the poloidal velocity and the normal to the disk midplane is important .
if it is higher than zero it can significantly increase the mass loss rate for low luminosities , and increase the wind velocity for all luminosities .
the presence of the azimuthal force does not change the mass loss rate if the geometry is fixed .
it is very intriguing that the mass loss rate can be enhanced by the magnetocentrifugal force the most for low luminosities where line force is weak and at the same time the mass loss rate remains a strong function of the disk luminosity for all luminosities .
in particular , there is no wind when @xmath78 as in purely line - driven case ( e.g. , psd i ; proga 1999 ) .
model iia with @xmath79 is steady and has the same @xmath80 as the corresponding psd ii model .
the mass loss rate is then the same regardless of the time behavior .
this is consistent with psd i conclusion that the mass loss rate depends predominately on the total system luminosity
. drew & proga ( 1999 ) applied results from psd i , psd ii and proga 1999 to nmcv .
in particular , they compared mass loss rates predicted by the models with observational constraints .
they concluded that either mass accretion rates in high - state nmcv are higher than presently thought by a factor of 2 - 3 or that radiation pressure alone is not quite sufficient to drive the observed hypersonic flows .
if the latter were true then an obvious candidate to assist radiation pressure in these cases is mhd ( e.g. , drew & proga 1999 ) .
an increases of the mass loss rate due to inclusion of a magnetocentrifugal force ( our case i ) brings our predictions close to the observational estimates for nmcvs .
however we should bear in mind that at the same time , the radial velocity of the wind increases above the observed velocities in nmcvs that are of the order a few thousand @xmath67 .
the increase of the wind velocity may not be large if the magnetic field is moderately strong and spins up the disk wind only close to its base where the mass loss rate is determined . then farther away from the disk
the wind would gain less or no angular momentum and velocities would be similar to those in the case without the magnetic field at all .
outflows which conserve specific angular velocity have a higher rotational velocity than those which conserve specific angular momentum .
the former also have higher terminal velocities due to stronger centrifugal force .
additionally in the wind where the specific angular velocity is conserved the rotational velocity is comparable to the terminal velocity while in the wind where the specific angular momentum is conserved the rotational velocity decreases asymptotically to zero with increasing radius and is therefore much lower than the terminal velocity .
thus we should be able to distinguish these two kinds of winds based on their line profiles .
for example , highly rotating winds should produce emission lines much broader than slowly rotating , expanding winds , if we see the disk edge - on .
so far we have discussed changes of the line - driven wind caused by inclusion of the magnetocentrifugal force .
however we can also anticipate some changes in the magnetocentrifugal wind if we add the line force .
for example , there may be a difference in the wind collimation caused by the line force .
the centrifugal force decollimates the magnetocentrifugal outflow near the disk as @xmath24 must be greater than @xmath81 .
in such a case the magnetic field collimates the outlow only beyond the alfv@xmath4n surface where the pinch force exerted by the toroidal component of the field operates . in the case with a significant line driving
, the magnetic field does not have to be inclined at the angle @xmath2 . then in such case , outflows are more collimated than pure magnetocentrifugal outflows from the start and may end up more collimated far away from the disk .
let us now discuss some results of other related works .
recently ogilvie & livio ( 1998 ) studied a thin accretion disk threaded by a poloidal magnetic field .
their purpose was to determine qualitatively how much thermal assistance is required for the flow to pass through the slow magnetosonic surface .
they found that a certain potential difference must be overcome even when @xmath82 and that thermal energy is not sufficient to launch an outlow from a magnetized disk .
ogilvie & livio suggested that an additional source of energy , such as coronal heating may be required .
psd i showed the radiative line force can drive disk winds .
thus the missing energy may be in the radiation field .
our calculations here show that if the disk luminosity is sufficiently high the line and magnetocentrifugal forces produce strong transonic winds regardless of the wind geometry .
then the magnetocentrifugal force can assist the line force in producing disk winds our approach to the problem or the line force or / and the thermal force can well assist mhd ogilvie & livio s approach ( see also wardle & k@xmath83ngil 1993 ; cao & spruit 1994 , for instance ) . in seeking a steady state solution of a line - driven disk wind , vitello & shlosman ( 1988 ) also found that the flow must overcome a potential difference
an increase of the vertical gravity component with height .
they suggested that for the radiation force to increase with height above the disk midplane and overcome this potential difference a very particular variation in the ionization state of the gas is required .
there are a number of limitations of our treatment of magnetic fields which are worthy of mention .
our models do not include the whole richness of mhd because we approximate the lorentz force by the magnetocentrifugal force ( eqs 3 - 8 ) instead of solving self - consistently the equation of motion and the induction equation .
we have included effects of a magnetic field on the disk winds in a simplistic manner that mimics some effects of a very strong , organized magnetic field where @xmath84 near the disk surface .
the magnetic field is treated as a rigid wire that controls the flow geometry outside the disk .
more quantitatively , this corresponds to the situation where outside the disk , at least near the disk photosphere , the magnetic field dominates i.e. , the magnetic pressure exceeds the disk gas pressure @xmath85 . inside the disk
however the situation may be different .
we assume that the disk is in a steady state that is also stable .
in such a case , the regions near the disk midplane are likely supported by the gas pressure and the magnetic pressure should be smaller than the gas pressure , i.e. , @xmath86 , otherwise the disk may be unstable ( stella & rosner 1984 ) . using the system parameters adopted here : @xmath87 and @xmath88 , we find that the @xmath89 condition yields a maximum value for the disk magnetic fields strentgh of the order of @xmath90 g. our assumption that poloidal magnetic field lines are straight excludes any magnetic tension in the @xmath91 plane . in our cases
i and iii we assume that the specific angular velocity , @xmath34 is conserved along a streamline from the footpoint of the line to infinity or rather to the outer boundary of the computational domain . however in a steady state axisymmetric mhd flow , the quantity which is conserved is the total angular momentum per unit mass which can be written as @xmath92 where @xmath93 is the position of the the alfv@xmath4n point of the flow on the streamline , and @xmath94 is the mass load another constant along the streamline ( e.g. , pelletier & pudritz 1992 ) .
the total angular momentum has contributions from both the flowing rotating gas and the twisted magnetic field .
this means that as the material angular momentum , @xmath95 increases due to corotation with the disk , the toroidal component of the magnetic field must increase so the total angular momentum is conserved . then our approximation in cases i and iii corresponds to the situation where near the disk
the total angular momentum is dominated by the contribution from the twisted magnetic field and the alfv@xmath4nic surface is beyond our computational domain . in other words
our approximation is valid in the region where the outflow is sub - alfv@xmath4nic .
consequently , our approach does not allow us to study collimation of the wind by the magnetic field because this happens beyond the fast magnetosonic surface where the flow is super - alfv@xmath4nic .
psd ii s models and ours in case ii correspond to the other extreme case where the total angular momentum has no contribution from the twisted magnetic field and the material angular momentum is conserved .
our treatment of magnetic fields does not include effects of the magnetic pressure .
the magnetocentrifugal driving presupposes the existence of a strong poloidal magnetic field comparable to the toroidal magnetic field near the disk surface , @xmath3 .
however when the poloidal magnetic field is weaker than the toroidal magnetic field , @xmath96 the magnetic pressure may be dynamically important in driving disk winds ( e.g. , uchida & shibata 1985 ; pudritz & norman 1986 , shibata & uchida 1986 , contopoulos 1995 ; kudoh & shibata 1997 ; ouyed & pudritz 1997 and references therein ) .
for @xmath97 , there is initially the buildup of the toroidal magnetic field by the differential rotation of the disk that in turn generates the magnetic pressure of the toroidal field .
the magnetic pressure then gives rise to a self - starting wind . to produce a steady outflow driven by the magnetic pressure a steady supply of the advected toroidal magnetic flux at the wind base
is needed , otherwise the outflow is likely a transient ( e.g. , k@xmath83nigl 1993 , contopoulos 1995 , ouyed & pudritz 1997 ) .
however it is not clear whether the differential rotation of the disk can produce such a supply of the toroidal magnetic flux to match the escape of magnetic flux in the wind and even if it does whether such a system will be stable ( e.g. , contopoulos 1995 , ouyed & pudritz 1997 and references therein ) .
concluding , we would like to stress that a further development of models of radiation - driven winds from disks should include taking into account magnetic field but it is equally important to consider adding the radiation force to models of mhd winds from luminous disks . both kinds of models are quite well understood now and if merged they could allow us to study better disk winds in systems such as cataclysmic variables and agns .
as we mentioned above thermal assistance may not be sufficient to launch wind from a magnetized disk and we should consider not only coronal heating but also line driving .
acknowledgements : we thank john cannizzo , janet drew , achim feldmeier , tim kallman , scott kenyon , mario livio , and james stone for comments on earlier drafts of this paper .
we also thank steven shore and an anonymous referee for comments that helped us clarify our presentation .
this work was performed while the author held a national research council research associateship at nasa / gsfc .
computations were supported by nasa grant nra-97 - 12 - 055 - 154 .
= = 1 = .5 in @xmath98 for all models from psd ii and our models in case i , the last column lists the wind opening angle , @xmath61 . while for models in cases ii and iii , the last column lists @xmath99 ( marked with @xmath100 ) , where @xmath24 is the assumed inclination angle between the poloidal component of the velocity and the normal to the disk midplane ; note that then @xmath101 . | we study the two - dimensional , time - dependent hydrodynamics of radiation - driven winds from luminous accretion disks threaded by a strong , large - scale , ordered magnetic field .
the radiation force is mediated primarily by spectral lines and is calculated using a generalized multidimensional formulation of the sobolev approximation .
the effects of the magnetic field are approximated by adding into the equation of motion a force that emulates a magnetocentrifugal force .
our approach allows us to calculate disk winds when the magnetic field controls the geometry of the flow , forces the flow to corotate with the disk , or both . in particular , we calculate models where the lines of the poloidal component of the field are straight and inclined to the disk at a fixed angle .
our numerical calculations show that flows which conserve specific angular velocity have a larger mass loss rate than their counterparts which conserve specific angular momentum .
the difference in the mass loss rate between these two types of winds can be several orders of magnitude for low disk luminosities but vanishes for high disk luminosities .
winds which conserve angular velocity have much higher velocities than angular momentum conserving winds . fixing the wind geometry stabilizes winds which are unsteady when the geometry is derived self - consistently . the inclination angle between the poloidal velocity and
the normal to the disk midplane is important .
non - zero inclination angles allow the magnetocentrifugal force to increase the mass loss rate for low luminosities , and increase the wind velocity for all luminosities .
the presence of the azimuthal force does not change the mass loss rate when the geometry of the flow is fixed .
our calculations also show that the radiation force can launch winds from magnetized disks .
the line force can be essential in producing magnetohydrodynamical ( mhd ) winds from disks where the thermal energy is too low to launch winds or where the field lines make an angle of @xmath0 with respect to the normal to the disk midplane . in the latter case
the wind will be less decollimated near its base by the centrifugal force and its collimation far away from the disk can be larger than the collimation of its centrifugally - driven mhd counterpart .
= = = 1=1=0pt = 2=2=0pt |
below we give a brief sketch of the field theoretical approach to the conductivity . here
we are led by the representation given in refs . [ ] .
the main quantity to be computed is the disorder averaged two particle green s function .
the disorder potential @xmath2 of the strength @xmath3 is supposed to have zero mean @xmath4 and gaussian correlator @xmath5 . for random gap disorder ,
the disorder averaged two particle green s function reads @xmath6\rangle^{}_v & \\ & \displaystyle = \sum_{m , m^\prime , n , n^\prime } [ \sigma^{}_1]^{}_{mn}[\sigma^{}_1]^{}_{n^\prime m^\prime}\langle\phi^1_{r^\prime m^\prime}\bar\phi^1_{rm}\phi^2_{rn } \bar\phi^2_{r^\prime n^\prime}\rangle^{}_{\phi } , & \end{aligned}\ ] ] where tr@xmath7 is taken on the extended dirac space , @xmath8 is a four component superfield @xmath9 , consisting of a complex @xmath10 and a grassmann @xmath11 field . in our notation @xmath12
are usual pauli matrices and @xmath13 the @xmath14 unity matrix .
the field averaging is defined as @xmath15\cdots e^{-{\cal s}},\ ] ] with the action @xmath16 where @xmath17 , @xmath18 , @xmath19 the @xmath20 unity matrix , and @xmath21 with the chemical potential @xmath22 , the non - random hamiltonian @xmath23 and the random hamiltonian @xmath24 then @xmath25 is invariant under the global symmetry transformation @xmath26 where @xmath27 is given by the following matrix @xmath28 with two scalar fields @xmath29 and @xmath30 , which obey grassmann statistics , i.e. @xmath31 and @xmath32 .
we decouple the interaction term in eq .
( [ eq : avact ] ) by a hubbard stratonovich transformation .
integrating out superfields @xmath8 yields an action in terms of composite supersymmetric hubbard
@xmath33 @xmath34 a non trivial vacuum of this theory @xmath35 is found from the saddle point condition and turns out to be degenerated with respect to the transformation @xmath27 defined in eq .
( [ eq : simitr ] ) : @xmath36 where @xmath37 ( @xmath38 ) commutes ( anticommutes ) with @xmath27 , and vanishes under the graded trace @xmath39 . on the saddle point manifold , the action represents a fermionic
linear @xmath0model [ ] @xmath40 the field @xmath38 represents the order parameter for the spontaneous breaking of the symmetry generated by @xmath27 .
expanding eq .
( [ eq : nlsm1 ] ) up to second order in @xmath38 as @xmath41 and using the exact relation for grassmann fields @xmath42 yields @xmath43 , & \end{aligned}\ ] ] with matrix green s function @xmath44 and @xmath45^{-1}_{rr^\prime},\ ] ] where @xmath46 $ ] is the scattering rate @xcite
. eventually , we rewrite eq .
( [ eq : nlsm ] ) in terms of scalar grassmann fields @xmath47 and obtain a nonlocal fermionic theory @xmath48 = \frac{g\eta}{2}\sum_{j=1,2}\sum_{rr^\prime}\left[\varphi^{}_{jr^\prime } \delta^{}_{r^\prime r } ( i\epsilon - d\nabla^2)\varphi^\prime_{jr } \right . &
\\ \label{eq : initfullact } & \displaystyle \left.- 2\eta^2(-1)^j \sum_{s=\pm}{\rm tr}^{}_2\left\{sg^{}_{s , rr^\prime } g^{}_{s , r^\prime r}\right\ } \varphi^{}_{jr^\prime}\varphi^\prime_{jr^\prime}\varphi^{}_{jr}\varphi^\prime_{jr}\right ] , & \hspace{3mm}\end{aligned}\ ] ] which describes the diffusion of dirac electrons . for @xmath49 ,
the diffusion coefficient reads @xmath50
below we investigate the scaling properties of action eq .
( [ eq : initfullact ] ) at large distances . for this purpose
we expand the fourier transform of the vertex function @xmath51 to the leading order in momenta of the fields .
this changes action eq .
( [ eq : initfullact ] ) to @xmath52 = \sum_{j , q}~\varphi^{}_{jq}(d^\prime q^2+i\epsilon^\prime)\varphi^\prime_{jq } & \\ & \displaystyle -i\lambda \sum_{j , kqpt } ( -1)^{j}\delta^{}_{k - q , p - t } ( p - t)^2 \varphi^{}_{jk}\varphi^\prime_{jq}\varphi^{}_{jp}\varphi^\prime_{jt } , & \;\;\;\end{aligned}\ ] ] with the shorthands @xmath53 , @xmath54 , @xmath55 , and @xmath56 . the interaction strength is defined as @xmath57 the zeroth order term in field momenta of the interaction part is zero due to its locality , while the first order term vanishes by the symmetry .
the frequency @xmath58 is supposed to be small and sent to zero in the dc limit .
for this reason we do not distinguish between @xmath59 and @xmath58 . in order to find the infrared behavior of action eq .
( [ eq : initact ] ) we follow the usual prescription of the wilson rg transformation @xcite : we decompose grassmann fields into fast @xmath60 and slow @xmath61 modes .
the idea is to integrate out fast modes and to obtain an action which mimics action eq .
( [ eq : initact ] ) but contains solely slow fields . to the second order in dc perturbation theory this action reads @xmath62 \approx { \cal s}^{}_0[\varphi^{}_s ] + { \cal s}^{}_{\rm int}[\varphi^{}_s ] + \langle { \cal s}^{}_{\rm int}[\varphi^{}_s,\varphi^{}_f]\rangle^{dc}_{f } & \\ \label{eq : newaction } & \displaystyle - \frac{1}{2 } \langle { \cal s}^{}_{\rm int}[\varphi^{}_s,\varphi^{}_f]{\cal s}^{}_{\rm int}[\varphi^{}_s,\varphi^{}_f]\rangle^{dc}_{f } , & \end{aligned}\ ] ] with @xmath63 & = & \displaystyle -i\lambda \sum_{j , kqpt } ( -1)^{j } \delta^{}_{k - q , p - t } ( p - t)^2 \\ & & \displaystyle \varphi^{}_{sjk}\varphi^\prime_{sjq}\varphi^{}_{sjp}\varphi^\prime_{sjt } , \label{eq : intact}\end{aligned}\ ] ] and @xmath64 $ ] represents terms which contain both slow and fast fields .
the averaging operator reads @xmath65~\cdots~e^{-{\cal s}^{dc}_0[\varphi^{}_f]},\ ] ] with the free dc action @xmath66 = d^\prime \sum_{j , r}\nabla\varphi^{}_{jrf } \nabla\varphi^\prime_{jrf}.\ ] ] obviously , this construction guarantees @xmath67 and defines the dc propagator ( in fourier representation ) @xmath68 the derivation of the renormalization group equations for the frequency @xmath58 , diffusion coefficient @xmath69 and interaction strength @xmath70 of the action eq .
( [ eq : initact ] ) is a challenging task .
diagrams which have to be evaluated arise by merging vertices shown in figs .
[ fig : eps ] , [ fig : diff ] and [ fig : lambda ] .
the renormalization of the energy @xmath58 comes from one loop diagrams which emerge by averaging vertices depicted in fig .
[ fig : eps ] over fast fields .
as shown in appendix [ app : eps ] , the evaluation of the diagrams yields a result that does not develop any divergences in the infrared , since the vertex is proportional to the squared loop momentum , and the propagator to inverse squared loop momentum : @xmath71 where @xmath72 denotes an upper cutoff .
this expression is not a renormalization group equation in the strict sense , since it does not contain the running cutoff parameter @xmath73 .
it represents a kind of a finite size effect which disappears in the continuous limit . .
, height=56 ] .
, height=56 ] .
, height=56 ] the renormalization of the diffusion coefficient can be obtained by integrating out fast fields in vertices depicted in fig .
[ fig : diff ] .
the evaluation of the functional integral is presented in appendix [ app : diffcoeff ] and yields the following renormalization of the free action : @xmath74,\ ] ] where @xmath75 and @xmath76 are expansion coefficients of zeroth and second order in momentum @xmath77 of the vertex function @xmath78}{k^2 q^2 ( k+q+p)^2}.\ ] ] there is no linear term in this expansion , since @xmath79 is symmetric with respect to the sign mirroring of @xmath77 , i.e. @xmath80 . in the appendices [ app : mass ] and
[ app : diff ] is shown that both coefficients @xmath75 and @xmath76 vanish .
this is a very important result , since it guarantees the preservation of the gapless diffusive mode and the realty of the diffusion coefficient even for the complex interaction strength .
therefore , the only running parameter is the interaction strength @xmath70 .
its renormalization is due to the one loop diagrams which emerge after integrating out fast fields in fig .
[ fig : lambda ] .
lengthy and elaborate calculations presented in appendix [ app : lambda ] lead to the remarkably simple renormalization group equation @xmath81 for each fermionic channel .
this equation is easily solved with the same starting value @xmath82 in both channels , @xmath83 given in eq .
( [ eq : intstrength ] ) @xmath84 both real and imaginary parts of the interaction scale down to zero but not equally fast . at large scales ,
the imaginary part of @xmath70 becomes dominant and therefore generates a genuine real interaction .
the rg flow for the grassmann field @xmath30 is depicted in fig .
[ fig : rgflow ] . fig .
[ fig : rgfparam ] shows the rg landscape in the parametric space spanned by the real and imaginary part of the interaction @xmath85 .
the rg trajectories represent a set of excentric circles with diameter @xmath83 , each attracted to the gaussian fixed point at @xmath86 and @xmath87 .
this can be seen best if we substitute @xmath88 in eq .
( [ eq : rgint ] ) .
then we get @xmath89 the right hand side of this system of differential equations represents indeed a parametrized circle . independently from the choice of the initial value both real and imaginary parts of @xmath90 ( analogously for @xmath91 ) become equal at the length obtained from the condition : @xmath92 with @xmath93 denoting the mean free path , which gives @xmath94 \approx l\exp\left[\frac{3g^2}{32}\frac{\eta}{\mu}\right]\;\;{\rm for}\;\;\mu<\eta.\ ] ] here we used definitions of the bare diffusion coefficient eq .
( [ eq : diffcoeff ] ) and interaction strength eq .
( [ eq : intstrength ] ) . at half filling , i.e. for @xmath95 , this scale is infinite . .
solid ( red ) line shows the real part and dashed ( blue ) line the imaginary part of @xmath90 .
the cross over scale is shown by the vertical ( black ) dotted line at @xmath96.,width=264 ] in the parametric space spanned by @xmath97 and @xmath98 .
the flow in the upper halfplane corresponds to the grassmann field @xmath30 , that in the lower plane to @xmath29.,width=302 ]
our ultimate task is to determine the scaling behavior of the dc conductivity . for this
we need to compute the corrections to the conductivity which arise due to the doping .
the dc conductivity is either determined from the einstein relation @xmath99 ( @xmath100 is the density of states at the fermi level ) or calculated from the kubo formula , @xmath101 where the two - particles green s function takes contributions from both channels @xmath102 into account : @xmath103 here , the functional integral should be performed over the full action eq .
( [ eq : initact ] ) : @xmath104 \cdots e^{-{\cal s}[\varphi]},\ ] ] with @xmath105 . to the leading order in @xmath85 , the two particles green s function
is approximated as @xmath106 where @xmath107 is given in eq .
( [ eq : initact ] ) .
the functional integration is to be performed over the free action only .
as shown in appendix [ app : cond ] , we eventually arrive at the following one loop rg equation for the conductivity @xmath108.\ ] ] further progress can be made if we exploit eq .
( [ eq : sollambda ] ) : @xmath109 unpleasant constants can be eliminated by rescaling @xmath110 and using definition of the diffusion coefficient @xmath111 .
this finally yields @xmath112 the integration of this equation is simple and we obtain the following asymptotic expression for the conductivity @xmath113.\ ] ] at half filling , i.e. @xmath114 , the conductivity does not flow , i.e. it is scale invariant . at large scales ,
i.e. for @xmath115 , @xmath116 given in ( [ eq : dellenmain ] ) , the conductivity grows bi - logarithmically as function of the sample size @xmath117 .
for this reason @xmath116 can be associated with an intermediate localization scale . due to the infrared asymptotic freedom of the underlying model
, this result should be asymptotically correct in all loops .
function corresponding to the conductivity in eq .
( [ eq : runcond]).,width=264 ]
the scaling properties of the conductivity are usually given by the @xmath1 function @xmath118 with rescaled logarithmic length @xmath119 .
the conductivity from eq .
( [ eq : runcond ] ) generates the @xmath1 function as depicted in fig .
[ fig : betafunc ] .
its shape reveals a striking resemblance of the numerically determined @xmath1 function of graphene at the dirac point with random scalar potential disorder @xcite : it starts at the value of the universal minimal conductivity ; it is strictly positive ; it reveals a distinct maximum related to the length @xmath116 ; it does not have any fixed points besides @xmath120 and @xmath121 .
finally , it does not depend on any quantities apart from the conductivity @xmath0 itself , in line with the one
parameter scaling hypothesis .
however , one might wonder whether the above mentioned bi
logarithmic asymptotics of the conductivity compares well with the predicted logarithmic growth at the dirac point @xcite .
it is indeed not difficult to reproduce the @xmath1 function of ref .
[ ] by applying the wilson rg transformation directly to the two particle green s function and exploiting the scaling properties of the disorder strength @xmath3
. instead , in our approach we keep @xmath3 scale invariant .
this assumption suits well for weak disorder , provided the sample size is much smaller than disorder generated intrinsic length @xmath122 $ ] @xcite , which corresponds to the common experimental situation . under these circumstances the scale invariance of the conductivity was demonstrated both numerically and analytically @xcite . on the other hand , in finite samples
the conductivity grows logarithmically as function of the chemical potential : @xmath123 in accordance with ref .
in conclusion , we have presented a scaling analysis of the diffusion coefficient and the dc conductivity of doped graphene with a random gap .
for this purpose we have used an alternative field theory and investigated its scaling properties . on this basis
we derived an invariant diffusion coefficient and an astonishingly simple expressions for the scaling of the conductivity that reproduces the distinct shape of the @xmath1 function of disordered graphene , found previously in numerical calculations @xcite .
we acknowledge financial support by the dfg grant zi 305/51 .
for sake of simplicity we use below the following notation for the interaction part of the action eq .
( [ eq : initact ] ) : @xmath124 the renormalization of @xmath58 is due to diagrams which are obtained by contracting fast fields in fig .
[ fig : eps ] .
contracting is possible in the only way , i.e. the diagrams have one fold degeneracy . to take grassmann statistics correctly into account it is necessary to permute fast fields through , such that they form normal ordered pairs @xmath125 : @xmath126&\\ { \nonumber}&\displaystyle = 2i\lambda \sum_{j}(-1)^j\int d1d2d3d4 ~(3 - 4)^2 \delta(1 + 3 - 2 - 4)&\\ { \nonumber}&\displaystyle \left[\varphi^{}_{j1s } \varphi^\prime_{j2s } \langle\dot\varphi^{}_{j3f } \dot\varphi^\prime_{j4f}\rangle - \varphi^{}_{j1s}\varphi^\prime_{j4s}\langle\dot\varphi^{}_{j3f } \dot\varphi^\prime_{j2f}\rangle \right].&\end{aligned}\ ] ] the fields to be contracted are marked with black dotes .
then , the contractions can be performed and we obtain : @xmath127 , & \end{aligned}\ ] ] where @xmath128 .
first contribution is zero because @xmath129 second contribution is finite and cutoff dependent : @xmath130 the renormalization factor then reads : @xmath131 which reduces for @xmath132 to @xmath133 .
lifting it into the exponent and absorbing into the action gives eq .
( [ eq : freqren ] ) .
the renormalization of the diffusion coefficient comes from the diagrams constructed from vertices depicted in fig [ fig : diff ] .
every diagram is twice degenerated , i.e. the analytical expression reads @xmath134.&\end{aligned}\ ] ] we permute fields through and perform functional integrations in order to get @xmath135 & \\ { \nonumber}&\displaystyle = 4(i\lambda)^2 \sum_{ij}(-1)^{i+j } \delta^{}_{ij}\delta^{}_{ij}\delta^{}_{ij } \int d1d2d3d4 \int d5d6d7d8 & \\ { \nonumber}&\displaystyle \times(3 - 4)^2(7 - 8)^2 \delta(1 + 3 - 2 - 4)\delta(5 + 7 - 6 - 8 ) & \\ { \nonumber}&\displaystyle \varphi^{}_{1si}\varphi^\prime_{6sj } ~ \pi(2)\pi(4)\pi(8 ) [ \delta(5 - 2)\delta(3 - 8)\delta(7 - 4 ) & \\ { \nonumber } & - \delta(7 - 2)\delta(3 - 8)\delta(5 - 4 ) ] & \\ { \nonumber}&\displaystyle = 4(i\lambda)^2 \sum_i \int d1~\varphi^{}_{1i } \varphi^\prime_{1i}\int d2d4~(2 - 1)^2 & \\ { \nonumber}&\displaystyle \times [ ( 2 - 1)^2-(4 - 1)^2]\pi(2)\pi(4)\pi(2 + 4 - 1 ) & \\ { \nonumber}&\displaystyle = \sum_{i , p } \gamma(p)\varphi^{}_{ip}\varphi^\prime_{ip}.&\end{aligned}\ ] ] the function @xmath79 is defined in eqs .
( [ eq : gam1 ] ) and ( [ eq : gam2 ] ) .
the way , how momenta in arguments of @xmath136functions are integrated out , is not unique .
therefore we can shift integration variables when necessary .
first we evaluate the expression for the mass : @xmath137}{k^2 q^2 ( k+q)^2 } & \\ { \nonumber}&\displaystyle = \int\frac{d^2k}{(2\pi)^2}\int\frac{d^2q}{(2\pi)^2 } \left [ \frac{k^2}{q^2(k+q)^2 } - \frac{1}{(k+q)^2 } \right].&\end{aligned}\ ] ] shifting in the second term @xmath138 we have @xmath139 where @xmath140 .
first term is conveniently evaluated using feynman parametrization : @xmath141 ^ 2 } = ( \ast).&\end{aligned}\ ] ] performing shift @xmath142 we get @xmath143 ^ 2 } & \\ { \nonumber}&\displaystyle = \frac{1}{4\pi } \int\frac{d^2k}{(2\pi)^2 } k^2 \int_0 ^ 1 dx ~ \frac{1}{k^2x(1-x ) } & \\ { \nonumber}&\displaystyle = \frac{1}{4\pi}\int_0 ^ 1 dx \left[\frac{1}{x}+\frac{1}{1-x}\right ] \int\frac{d^2k}{(2\pi)^2 } & \\ { \nonumber}&\displaystyle = \frac{1}{2\pi } \int^1_{e^{-\ell } } \frac{dx}{x } \int\frac{d^2k}{(2\pi)^2 } = \frac{\ell}{2\pi}\int\frac{d^2k}{(2\pi)^2 } , & \end{aligned}\ ] ] i.e. the very same result .
therefore , the mass is zero , the diffusive goldstone mode is preserved .
next we evaluate the renormalization of the diffusion coefficient : @xmath144 \right|_{p=0}. & \;\;\;\end{aligned}\ ] ] we start with first term . using feynman parametrization
we have @xmath145 ^ 2 } \right|_{p=0}. & \end{aligned}\ ] ] next shift @xmath146 : @xmath147 ^ 2}\right|_{p=0 } & \\ { \nonumber}&\displaystyle = \frac{1}{4\pi}\left.\frac{\partial^2}{\partial p^2}\int\frac{d^2q}{(2\pi)^2 } \frac{(q+p)^4}{q^2 } \int_0 ^ 1 dx~ \frac{1}{(q+p)^2x(1-x)}\right|_{p=0 } & \\ { \nonumber}&\displaystyle = \frac{1}{4\pi}\left.\frac{\partial^2}{\partial p^2}\int\frac{d^2q}{(2\pi)^2 } \frac{(q+p)^2}{q^2}\right|_{p=0 } \int_0 ^ 1 dx~ \frac{1}{x(1-x ) } & \\ & \displaystyle { \nonumber}= \frac{1}{4\pi}\int\frac{d^2q}{(2\pi)^2 } \frac{2}{q^2}\int_0 ^ 1 dx~ \frac{1}{x(1-x ) } = 2\left(\frac{\ell}{2\pi}\right)^2 . &
\end{aligned}\ ] ] it is important to recognize that shifting of the integration variables with respect to the external momentum @xmath77 does not affect the final result .
this is because the integration over the momentum conserving @xmath136function is not unique . indeed ,
if we shift @xmath148 in the above integral we obtain the same result : @xmath149 where we skipped integration over the momentum @xmath150 . after mirroring sign of @xmath151 we have @xmath152 = ( \ast ) . &
\end{aligned}\ ] ] here we have to shed some light on the structure of the first term : @xmath153 thus we continue : @xmath154 & \\ { \nonumber}&\displaystyle = 2\frac{\ell}{2\pi}\int\frac{d^2q}{(2\pi)^2}\frac{1}{q^2}=2\left(\frac{\ell}{2\pi}\right)^2 , & \end{aligned}\ ] ] i.e. the very same result as before .
this knowledge can now be used for evaluating second term : @xmath155 here we are allowed to shift @xmath156 , which yields @xmath157 & \\ { \nonumber}&\displaystyle = \int\frac{d^2q}{(2\pi)^2}\int\frac{d^2k}{(2\pi)^2}~\frac{2}{q^2(k+q)^2 } & \\ { \nonumber}&\displaystyle + \int\frac{d^2q}{(2\pi)^2}\int\frac{d^2k}{(2\pi)^2}~\frac{1}{(k+q)^2}\left[8\frac{(\hat e_p\cdot k)^2}{k^4 } - \frac{2}{k^2}\right ] & \\ { \nonumber}&\displaystyle + 8 \int\frac{d^2q}{(2\pi)^2}\int\frac{d^2k}{(2\pi)^2 } \frac{(\hat e_p\cdot q)(\hat e_p\cdot k)}{q^2k^2(k+q)^2 } . & \end{aligned}\ ] ] first and second terms can be easily evaluated after shifting @xmath158 and @xmath159 , respectively .
both give the same contribution @xmath160 calculated above .
the last term is more cumbersome : @xmath161 ^ 2 } .
& \end{aligned}\ ] ] again , we shift @xmath162 , which gives @xmath163 ^ 2 } = & \\ { \nonumber}&\displaystyle -\int\frac{d^2q}{(2\pi)^2}\frac{(\hat e_p\cdot q)^2}{q^2}\int_0 ^ 1 dx \int\frac{d^2k}{(2\pi)^2 } \frac{8x}{[k^2+x(1-x)q^2]^2 } & \\ { \nonumber}&\displaystyle = -8 \int\frac{d^2q}{(2\pi)^2}\frac{(\hat e_p\cdot q)^2}{q^4 } \frac{1}{4\pi}\int_0 ^ 1 \frac{dx}{1-x } & \\ { \nonumber}&\displaystyle = \frac{1}{\pi}\int\frac{d^2q}{(2\pi)^2}\frac{1}{q^2}\int^{e^{-\ell}}_1\frac{dy}{y } = -2\left(\frac{\ell}{2\pi}\right)^2 = -{\rm i } , & \end{aligned}\ ] ] where at some point we have substituted @xmath164 . summing over all contributions to @xmath165 we
obtain @xmath166 the renormalization of the diffusion coefficient to order @xmath167 is therefore zero : @xmath168
the renormalization of the interaction is due to one loop diagrams which arise after merging vertices depicted in fig .
[ fig : lambda ] . we consider both contributions separately .
it has a two fold degeneration , i.e. there are two possibilities for contracting fast fields : @xmath169.&\end{aligned}\ ] ] next , we permute and contract grassmann fields : @xmath170&\\ { \nonumber}&\displaystyle = ( i\lambda)^2 \sum_{ij}(-1)^{i+j } \delta^{}_{ij}\delta^{}_{ji}\int d1d2d3d4\int d5d6d7d8 & \\ { \nonumber}&\displaystyle ( 3 - 4)^2(7 - 8)^2\delta(1 + 3 - 2 - 4)\delta(5 + 7 - 6 - 8 ) & \\ { \nonumber}&\displaystyle \times\left[\delta(7 - 2)\delta(5 - 4 ) - \delta(5 - 2)\delta(7 - 4)\right]&\\ { \nonumber}&\displaystyle \varphi^{}_{s1i}\varphi^{}_{s3i}\varphi^\prime_{s6j}\varphi^\prime_{s8j}\pi(2)\pi(4 ) . &
\end{aligned}\ ] ] performing summations and integrations we arrive at @xmath171\pi(4)\pi(4 - 3 - 1).&\end{aligned}\ ] ] after reordering fields in first term and renaming variables we finally obtain @xmath172 now we expand vertex function up to the second order in fields momenta : @xmath173 \left.\frac{1}{{d^\prime}^2}\int\frac{d^2q}{(2\pi)^2}~\frac{(q - k)^2(q - p)^2}{q^2(q - k - t)^2}\right|_{k , p , t=0}. & \;\;\;\end{aligned}\ ] ] * order @xmath174 : the factor is zero , since @xmath175 after differentiation with respect to @xmath150 .
* order @xmath176 : the expansion factor is also zero , because : @xmath177\;\;(\ast ) & \end{aligned}\ ] ] rewrite first term : @xmath178 plugging this back into @xmath179 we see that @xmath180 & \\ { \nonumber}&\displaystyle = -\int\frac{d^2q}{(2\pi)^2}~\left[2\frac{\hat e^{}_k\cdot \hat e^{}_t}{q^2 } - 2\frac{\hat e_t\cdot \hat e^{}_k}{q^2 } \right ] = 0 . &
\end{aligned}\ ] ] * order @xmath181 : the expansion factor reads : @xmath182 thus , the order @xmath181 in expansion is @xmath183 * order @xmath184 : the corresponding factor is zero : @xmath185 * order @xmath186 : the factor in expansion is @xmath187 here , it is possible to shift @xmath188 and we obtain @xmath189 * order @xmath190 : the corresponding expansion coefficient reads @xmath191 concluding , the leading order momentum expansion of the vertex function reads @xmath192 and the contribution to the renormalization of the interaction strength @xmath193
each diagram has the degeneracy one , and we may write corresponding expressions as @xmath194 . &
\end{aligned}\ ] ] after permuting fields , performing summations and integrations , and renaming variables we obtain following expressions : @xmath195 while evaluating contributions iia and iic , it suffices to take only the most divergent part from the integral over loop momentum 5 into account .
this yields @xmath196 for contribution ib , it is necessary to perform a full integration over the loop momentum with help of the feynman parametrization : @xmath197 where we replace momentum @xmath198 .
hence , contribution iib reads @xmath199 the evaluation of the contribution iid goes analogously to the evaluation of the contribution i with the result @xmath200 an apparent problem with the momentum dependence of the vertex function can be cured if we remember that in dc limit @xmath201 and @xmath202 are not independent but @xmath203 . using this property
we can mirror momenta @xmath204 and @xmath205 and obtain @xmath206 the sing change is due to permuting grassmann variables .
after renaming variables we get the topological structure of initial interaction term . summing up all contributions
gives @xmath207 therefore , the rg equation for the interaction strength acquires the form @xmath208 it is therefore convenient to distinguish between interactions in each channel : @xmath209 in continuous limit this gives eq .
( [ eq : rgint ] ) .
the dc conductivity is calculated from kubo formula eq .
( [ eq : kubo ] ) with the two - particles green s function defined in eqs .
( [ eq:2pgf ] ) and ( [ eq : avop ] ) .
the action is slightly changed using the acquired knowledge , as @xmath210 = { \cal s}^{}_0[\varphi ] + { \cal s}^{}_{\rm int}[\varphi],\end{aligned}\ ] ] with @xmath211 = \sum_{ij=1,2}\int d1d2 ~\delta^{}_{i , j}\delta(1 - 2)~\varphi^{}_{i1}(i\epsilon + d^\prime \nabla^2)\varphi^\prime_{j2},\ ] ] and @xmath212 & = & \sum^{}_{j=1,2 } ( -1)^j\lambda^{}_{j}\int d1d2d3d4~\delta(1 + 3 - 2 - 4 ) \\ { \nonumber}&\times & ( 3 - 4)^2 \varphi^{}_{j1}\varphi^\prime_{j2}\varphi^{}_{j3}\varphi^\prime_{j4}.\end{aligned}\ ] ] we evaluate eq .
( [ eq : pt1ord ] ) denoting @xmath213 where @xmath214 \cdots e^{-{\cal s}^{}_0[\varphi]}.\ ] ] here , the proportionality factor is @xmath215 .
the zeroth order contribution to the conductivity is calculated as usual : @xmath216 and further with @xmath217 : @xmath218 i.e. the usual universal dc conductivity of dirac electron gas .
the evaluation of the second term is more cumbersome .
respecting all possible ( four in total ) contraction combinations yields : @xmath219\rangle^0 & \\ { \nonumber}&\displaystyle = \sum_{ij\alpha}(-1)^{\alpha } i\lambda^{}_\alpha\int\frac{d^2p}{(2\pi)^2 } & \\ { \nonumber}&\displaystyle \int d1d2d3d4~(3 - 4)^2~\delta(1 + 3 - 2 - 4)&\\ { \nonumber}&\displaystyle \left [ \dot\varphi^{}_{iq } \ddot\varphi^\prime_{jp } \ddot\varphi^{}_{\alpha 1 } \dot\varphi^\prime_{\alpha2 } \dddot\varphi^{}_{\alpha3 } \dddot\varphi^\prime_{\alpha4 } + \dot\varphi^{}_{iq } \ddot\varphi^\prime_{jp } \ddot\varphi^{}_{\alpha 1 } \dddot\varphi^\prime_{\alpha2 } \dddot\varphi^{}_{\alpha3 } \dot\varphi^\prime_{\alpha4 } \right . & \\ { \nonumber}&\displaystyle \left .
+ \dot\varphi^{}_{iq } \ddot\varphi^\prime_{jp } \dddot\varphi^{}_{\alpha 1 } \dddot\varphi^\prime_{\alpha2 } \ddot\varphi^{}_{\alpha3 } \dot\varphi^\prime_{\alpha4 } + \dot\varphi^{}_{iq } \ddot\varphi^\prime_{jp } \dddot\varphi^{}_{\alpha 1 } \dot\varphi^\prime_{\alpha2 } \ddot\varphi^{}_{\alpha3 } \dddot\varphi^\prime_{\alpha4 } \right ] & \\ { \nonumber}&\displaystyle = i\sum_{ij\alpha}(-1)^{\alpha } \lambda^{}_\alpha \delta^{}_{i\alpha}\delta^{}_{j\alpha}\delta^{}_{\alpha\alpha } & \\ { \nonumber}&\displaystyle \int\frac{d^2p}{(2\pi)^2 } \int d1d2d3d4~(3 - 4)^2~\delta(1 + 3 - 2 - 4 ) & \\ { \nonumber}&\displaystyle \left [ -\delta(2-q)\delta(1-p)\delta(3 - 4)k^{}_0(q)k^{}_0(p)k^{}_0(3 ) \right.&\\ { \nonumber } & + \delta(4-q)\delta(1-p)\delta(3 - 2)k^{}_0(q)k^{}_0(p)k^{}_0(3 ) & \\ { \nonumber}&\displaystyle -\delta(4-q)\delta(3-p)\delta(1 - 2)k^{}_0(q)k^{}_0(p)k^{}_0(2 ) & \\ { \nonumber}&\displaystyle \left.+ \delta(q-2 ) \delta(3-p ) \delta(1 - 4 ) k^{}_0(q)k^{}_0(p)k^{}_0(4 ) \right ] . &
\end{aligned}\ ] ] while contributions from the first and third terms vanish , both other give equal finite contributions : @xmath220}{\displaystyle(d^\prime q^2+i\epsilon)^2 } \int \frac{d^2k}{(2\pi)^2}~\frac{k^2+q^2}{d^\prime k^2 + i\epsilon}.\ ] ] the integral @xmath221 behaves well in dc limit ( @xmath222 ) as it does not develop any ir divergences .
the corresponding contribution disappears in continuous limit and does not affect the conductivity at large scales .
therefore , the main contribution arises from the following expression : @xmath223}{d^\prime k^2+i\epsilon}.\ ] ] in dc limit , the integral diverges logarithmically .
the corresponding conductivity correction reads @xmath224\ell,\ ] ] and the full expression for the renormalized conductivity : @xmath225\ell.\ ] ] continuous limit of this expression yields eq .
( [ eq : condrgeq ] ) .
novoselov , a.k .
geim , s.v .
morozov , d. jiang , m.i .
katsnelson , i.v .
grigorieva , s.v .
dubonos , a.a .
firsov , nature * 438 * , 197 ( 2005 ) .
y. zhang , y .- w .
tan , h.l .
stormer , p. kim , nature * 438 * , 201 ( 2005 ) .
k. ziegler , j. phys .
* 45 * , 335001 ( 2012 ) .
k. ziegler , phys .
b * 55 * , 10661 ( 1997 ) ; k. ziegler and g. jug , z. phys .
b * 104 * , 5 ( 1997 ) ; k. ziegler , phys . rev . lett .
* 80 * , 3113 ( 1998 ) .
k. ziegler , phys .
102 * , 126802 ( 2009 ) ; phys .
b * 79 * , 195424 ( 2009 ) .
a. sinner and k. ziegler , phys .
b * 84 * , 233401 ( 2011 ) .
k. ziegler , preprint arxiv:1207.3701 m. bocquet , d. serban , and m. r. zirnbauer , nucl .
b * 578 * , 628 ( 2000 ) .
r. shankar , rev .
* 66 * , 129 ( 1994 ) .
d. s. l. abergel , v. apalkov , j. berashevich , k. ziegler , and t. chakraborty , adv . phys . * 59 * , 261 ( 2010 )
. j. h. bardarson , j. tworzydo , p. w. brouwer , c. w. j. beenakker phys .
lett . * 99 * , 106801 ( 2007 ) .
a. schuessler , p. m. ostrovsky , i. v. gornyi , and a. d. mirlin , phys .
b * 79 * , 075405 ( 2009 ) .
j. h. bardarson , m. v. medvedyeva , j. tworzydo , a. r. akhmerov , and c. w. j. beenakker , phys .
b * 81 * , 121414(r ) ( 2010 ) .
i. l. aleiner and k. b. efetov , phys .
lett . * 97 * , 236801 ( 2006 ) . | we investigate the scaling properties of the recently acquired fermionic non
linear @xmath0model which controls gapless diffusive modes in a two dimensional disordered system of dirac electrons beyond charge neutrality .
the transport on large scales is governed by a renormalizable nonlocal field theory .
for zero mean random gap , it is characterized by the absence of a dynamic gap generation and a scale invariant diffusion coefficient .
the @xmath1 function of the dc conductivity , computed for this model , is in perfect agreement with numerical results obtained previously .
= 10000 transport in systems whose band structure has a node structure ( e.g. dirac points ) , as it appears in graphene and on the surface of 3d topological insulators , has been the subject of intense research recently .
the experimental observation of transport in graphene is characterized by a minimal conductivity at the charge neutrality ( or dirac ) point and by a linearly increasing conductivity with increasing ( electron or hole ) charge density @xcite .
thus , experimentally it is easy to distinguish the very robust minimal conductivity which is contrasted by the disorder - dependent conductivity away from the dirac point . the behavior at the dirac point has also been predicted by field theory , showing that the minimal conductivity is quite independent ( or very weakly dependent ) on disorder .
for instance , a nonlinear sigma model approximation @xcite as well as perturbation theory in terms of disorder strength clearly shows a very weak disorder dependence @xcite .
this has led to the claim that ballistic transport can not be distinguished from diffusive transport at the dirac point
. however , transport properties away from the dirac point are theoretically not easily accessible .
there are several attempts , based on a classical boltzmann approach , which predict the experimentally observed linearly increasing conductivity as we go away from the dirac point . however , it was only recently that a more general field - theoretical approach , based on the kubo formalism , was suggested to describe the transport of 2d dirac fermions within a unified theory @xcite , using a four - body hamiltonian .
in this paper we focus on disorder due to a random gap .
this case is particularly interesting because it can lead to a metal - insulator transition when the average gap is equal to a critical value @xcite .
starting from a nonlinear sigma model that controls the diffusive modes , we study the renormalization of the interaction of these modes as well as the renormalization of the diffusion coefficient and the conductivity at the dirac point and away from it . |
star formation is one of the most fundamental processes of the evolution of galaxies .
stars are formed by the contraction of molecular interstellar medium ( ism ) , and then a new ism containing heavy element is dispersed into interstellar space by stellar wind and supernova explosions when stars end their lives . by repeating these processes , ism , galaxies , and the universe evolve .
therefore , the physics of star formation is very important in the understanding of the evolution of galaxies . in the central regions of disk galaxies
, we often find intense star - formation activities ( e.g. , @xcite ) where not only the star formation rate ( sfr ) but also the star formation efficiency ( sfe ; @xcite ) , which is an intensive parameter defined as the fraction of the sfr surface density in the surface mass density of molecular gas , are also enhanced .
these star formation activities in central regions of galaxies are different from a simply scaled - up version of that in the disk region , and it is are referred to as starburst @xcite . the nuclear starburst is prominent not only because they show high sfr but also because they show elevated sfe .
therefore , for understanding the nuclear starburst we should reveal what causes the high sfe star formation .
it is essential to study _ dense _ molecular gas , because stars are formed from the dense cores of molecular clouds .
in fact , recent studies of the dense molecular medium in galaxies based on the observations in the hcn(@xmath1 ) emission , a dense molecular gas tracer ( @xmath3 @xmath4 a few @xmath5 @xmath6 ) due to its large dipole moment ( @xmath11 debye , whereas @xmath12 debye ) , demonstrate the intimate connection between dense molecular gas and massive star formation in galaxies . after the pioneer work by @xcite ,
a tight correlation between the hcn(@xmath1 ) intensity and fir continuum luminosities has been shown among many galaxies including nearby ones ( @xcite , @xcite ) , and high redshift quasar host galaxies ( @xcite , @xcite ) , although some deviations from the correlation are reported toward extreme environments ( @xcite ; @xcite , @xcite ) . a spatial coincidence between dense molecular gas traced by the hcn(@xmath1 ) emission and massive star - forming regions
was also reported ( @xcite , @xcite ) .
in addition , @xcite reported that hcn(@xmath1)/co(@xmath1 ) integrated intensity ratio in the brightness temperature scale , hereafter @xmath8 , correlates well with sfe .
this suggests that sfe is controlled by the fraction of dense molecular gas to the total amount of molecular contents traced by @xmath8 ( e.g. , @xcite , @xcite , @xcite , @xcite ) . however , these previous reports on the relationship between dense gas fraction and sfe are derived from coarse spatial resolution ( @xmath13 ) data of moderately distant galaxies ( @xmath14 mpc ) .
it means the results show a correlation on global scales of galaxies . in our galaxy , a massive star - forming region
is associated with the giant molecular clouds ( gmcs ) or their associations ( giant molecular associations ; gmas ) .
a typical scale of a gmc is a few 10 pc ( @xcite ) and that of a gma is a few 100 pc .
therefore , we should investigate a correlation between dense gas fraction and sfe on this scale . to address this issue ,
observations with high - angular resolution of the central region of nearby galaxies are required . in this paper
, we present co(@xmath1 ) , hcn(@xmath1 ) , and 95 ghz continuum observations toward the central regions of m 83 ( ngc 5236 ) using the nobeyama millimeter array ( nma ) .
m 83 is a nearby , face - on , barred , grand - design spiral galaxy hosting an intense starburst at its center .
the distance to m 83 is estimated to be 4.5 mpc @xcite ; therefore , 1@xmath15 corresponds to 22 pc .
we can then discuss the relation between @xmath8 and the sfe at the center of m 83 on a gma scale with a few arcsec resolution observations , which can be accomplished with the nma .
the high - resolution ( @xmath16 23@xmath15 @xmath2 500 pc ) co line observations of the central region of m 83 have been performed repeatedly with various telescopes and in various transitions .
we summarized them in table 1 .
for example , @xcite mapped the distribution of co(@xmath1 ) emission for the center and bar at a resolution of @xmath17 using the nro 45-m telescope .
they observed a large concentration of molecular gas in the nucleus and an extended ridge along the major axis of the bar .
@xcite reported the co(@xmath1 ) observation toward the center with the nma .
the distribution of co(@xmath1 ) emission shows twin peak structure at the nucleus . @xcite
performed co(@xmath18 ) and co(@xmath19 ) emission observations toward the center and the bar using the submillimeter array ( sma ) .
their co maps revealed two gas ridges on the leading side of the stellar bar and a nuclear gas ring of @xmath2 300 pc in diameter .
the distribution of co(@xmath18 ) emission is not coincident with the nuclear starburst , which is depicted by a three - color composite image made from @xmath20/wfpc2 data , f300w , f547 m , and f814w .
the authors mentioned that the nuclear starburst in m 83 is lopsided , mostly on the receding side ( south ) of the dynamical center .
table 1.previous high - resolution co observations of m 83 . + crlll + transition & authors & telescope & map size & resolution + + j=10 & @xcite & nro 45-m & [email protected] @xmath22 1@xmath21 & 16@xmath15 + & @xcite & nma & [email protected] @xmath22 [email protected] & 12@xmath15 @xmath22 6@xmath15 + & @xcite & nma & [email protected] @xmath22 [email protected] & 12@xmath15 @xmath22 6@xmath15 + & @xcite & nro 45-m & 6@xmath21 @xmath22 6@xmath21 & 16@xmath15 + & this work & nma & 2@xmath21 @xmath22 1@xmath21 & 7@xmath15 @xmath22 3@xmath15 + + j=21 & @xcite & jcmt 15 m & & 22@xmath15 + & @xcite & jcmt 15 m & [email protected] @xmath22 2@xmath21 & 21@xmath15 + & @xcite & sest 15 m & 10@xmath21 @xmath22 10@xmath21 & 23@xmath15 + & @xcite & sma & [email protected] @xmath22 1@xmath21 & [email protected] @xmath22 [email protected] + + j=32 & @xcite & cso 10 m & & 22@xmath15 + & @xcite & jcmt 15 m & [email protected] @xmath22 [email protected] & 14@xmath15 + & @xcite & jcmt 15 m & [email protected] @xmath22 [email protected] & 14@xmath15 + & @xcite & smto 10 m & 2@xmath21 @xmath22 2@xmath21 & 22@xmath15 + & @xcite & sma & [email protected] @xmath22 0@xmath216 & [email protected] @xmath22 [email protected] + & @xcite & cso 10 m & [email protected] @xmath22 2@xmath21 & 22@xmath15 + & @xcite & aste 10 m & 5@xmath21 @xmath22 5@xmath21 & 22@xmath15 + + j=43 & @xcite & jcmt 15 m & [email protected] @xmath22 [email protected] & 11@xmath15 + & @xcite & jcmt 15 m & [email protected] @xmath22 [email protected] & 11@xmath15
+ + + it is difficult to estimate the star formation activity accurately in the central starburst region because of strong extinction .
@xcite presented a photometric catalog of 45 massive star clusters in the nuclear starburst region of m 83 , observed with @xmath20/wfpc2 .
they revealed that the ages of star clusters in the central 300 pc are younger than 10 myr old , and the cluster age distribution has an extremely sharp peak at 5 7 myr .
however , the cluster ages in the north side of the nucleus are unknown since no clusters were found due to strong extinction by dust
. they also mentioned the original starburst had started at the southern end of the nuclear starburst region , and propagated northward .
the propagation of star formation is also reported by @xcite and by @xcite , who quantify the age gradient better using near - ir data . to the north side of the nucleus
, the h@xmath23/h@xmath24 ratio is higher than that to the south side ( see figure 4 in @xcite ) .
this means that dust extinction to the north side is stronger than that to the south side .
the strong dust extinction indicates the existence of huge amounts of dust and molecular gas .
in addition , the extinction on the north side has not yet been accurately estimated since it is too strong , and several authors suggests the existence of a highly - obscured nuclear starburst ( @xcite , @xcite , @xcite ) .
therefore , it is necessary to evaluate the strength of extinction to investigate the relationship between molecular gas and true star formation in the central region of m 83 .
the goals of this paper are : ( 1 ) to obtain distributions of total molecular gas and dense molecular gas through co(@xmath1 ) line and hcn(@xmath1 ) line emission in the central 1.5 kpc region of m 83 with a spatial resolution of @xmath2 100 pc , ( 2 ) to evaluate the extinction and reveal its distribution , and to obtain the extinction - corrected sfr and sfe .
( 3 ) to examine the correlation between dense gas fraction traced by @xmath8 and the sfe on a gma scale .
aperture synthesis observations in the co(@xmath1 ) line , the hcn(@xmath1 ) line , and the 95 ghz continuum emission towards the central region ( @xmath2 1.5 kpc ) of m 83 were carried out with the nma during the periods from december 2003 to april 2004 , and from december 2005 to april 2006 .
co(@xmath1 ) line observations were made toward 3 adjacent field - of - views ( fovs ) to make a mosaic of the center of the galaxy .
hcn(@xmath1 ) line and 95 ghz continuum observations were made toward only a central fov .
figure 1 shows the each fov superposed on the v - band image obtained with the vlt @xcite .
the nma consists of six 10-m antennas equipped with cryogenically cooled receivers employing superconductor - insulator - superconductor mixers in double side band operation .
three antenna configurations ( ab , c , and d ) were used during the observations . the backend used was the ultra wide - band correlator @xcite . a radio source , j1337 - 129 , was observed every @xmath2 20 minutes for amplitude and phase calibrations , and the passband shape of the system was determined from observations of a strong continuum source 3c 273 .
the flux density of j1337 - 129 was measured at several times during an observing run based on that of a quasar 3c 345 .
its flux is determined by flux measurements of saturn and uranus .
the overall error of the absolute flux scale was estimated to about @xmath2520% .
the raw visibilities were edited and calibrated using the nro uvproc - ii package @xcite , and then final images were created using the imagr task in the nrao aips package . for some spectral line analysis , data also have the continuum emission . in this case , we subtracted the continuum emission from the visibilities using the uvproc - ii task lcont at first .
we use the subtracted continuum emission in hcn(@xmath1 ) data as the 95 ghz continuum emission .
therefore , this flux is actually an average flux in two separated bands at 88.741 @xmath25 0.5 ghz and 100.741 @xmath25 0.5 ghz .
mosaic images in the co(@xmath1 ) line were made using miriad @xcite .
achieved sensitivities and spatial resolutions are summarized in table 2 and 3 .
( 80mm,75mm)figure-1.eps table 2.observation parameters and results of m 83 with the nma .
+ llll + observed line & & co(@xmath1 ) & hcn(@xmath1 ) + + rest frequency & [ ghz ] & 115.271204 & 88.631604 + band width & [ mhz ] & 512 & 1024 + band resolution & [ mhz ] & 2 & 8 + velocity resolution & [ km s@xmath26 & 5.2 & 27.1 + sideband & & upper & lower + fov & [ @xmath15 ] & 59 & 77 + number of fovs & & 3 ( mosaic ) & 1 + observed date & & 2003/12 - 2004/4 & 2005/12 - 2006/4 + synthesized beam size & [ @xmath15 ] & 7.2 @xmath22 3.4 & 7.1 @xmath22 3.1 + equivalent @xmath28 & [ k ( jybeam@xmath29)@xmath26 & 3.76 & 7.15 + rms noise ( channel map ) & [ mjybeam@xmath26 & 85 & 7.2 + & [ mk ] & 320 & 51.5 + rms noise ( intensity map ) & [ jybeam@xmath29kms@xmath26 & 2.7 & 0.52 + & [ k km s@xmath26 & 10.2 & 3.72 + total flux within fov & [ jykms@xmath26 & [email protected] @xmath22 10@xmath30 & 60@xmath2510 + + & & + r.a .
( j2000 ) & & + decl .
( j2000 ) & & + + + @xmath31 reference of the central position of the fov @xcite table 3.summary of the 95 ghz continuum in m 83 .
+ llc + synthesized beam & [ @xmath15 ] & 8.0 @xmath22 3.4 + equivalent @xmath28 & [ k ( jybeam@xmath29)@xmath26 & 5.00 + rms noise & [ mjybeam@xmath26 & 0.85 + & [ mk ] & 4.25 + total flux within fov & [ mjy ] & 30 + + + here , 95 ghz means a combination of [email protected] ghz and [email protected] ghz .
figure 2 shows an integrated intensity map and a velocity field in the co(@xmath1 ) line , an integrated intensity map in the hcn(@xmath1 ) line , and a map of the 95 ghz continuum emission .
our high resolution co(@xmath1 ) mosaic map depicts the presence of molecular ridges along the leading sides of the stellar bar and nuclear twin - peaks structure .
this structure is seen in the previous studies in the co(@xmath1 ) ( @xcite ) , and in the co(@xmath19 ) and the co(@xmath18 ) @xcite .
strong non - circular motion along the molecular ridges was also detected in the velocity field .
this strong non - circular motion is consistent well to that detected in the co(@xmath19 ) velocity field ( @xcite ) .
the motion might play an important role in feeding large amount of molecular gas into the starburst nucleus .
we determined the dynamical center from the co velocity field using the aips task gal ( see table 4 ) .
we find the distributions of the hcn(@xmath1 ) line intensity , and the 95 ghz continuum flux are confined to a single peak , corresponding to the northern peak seen in the co map .
this means that dense molecular gas traced by the hcn(@xmath1 ) line and current star formation traced by the 95 ghz continuum emission are distributed only at north side of the center , although less dense molecular gas seen in the co(@xmath1 ) line emission are widely distributed in the central region .
the peak position of hcn(@xmath1 ) line intensity is consistent spatially with the youngest star clusters found by @xcite and @xcite .
( 170mm,181mm)figure-2.eps here , we summarized the positions of several peaks and nuclei determined from near - ir and our mm - wave data in figure 3 and table 4 .
the dymanical center determined from our co velocity field is near the symmetry center of the outer k - band isophotes , whereas the visible nucleus ( that is also the location of the k - band photometric peak in table 4 ) is offset from the dynamical center and the peaks of hcn(@xmath1 ) line intensity , 95 ghz continuum emission , and the extinction - corrected h@xmath23 luminosity ( see section 4.1 ) .
the position of the hidden mass concentration discussed in @xcite and @xcite seems to be coincident with the peaks of hcn(@xmath1 ) line intensity and the 95 ghz continuum emission , although the spatial resolutions of our mm - wave data ( @xmath32 ) are insufficient to make a detailed comparison .
we compare our hcn(@xmath1 ) map with that obtained by @xcite .
the peak positions of the hcn(@xmath1 ) line emission are almost coincident between these two maps , whereas the flux values of the peak position are different .
when our map was convolved to @xmath33 , which is the beam size of their map , the peak flux was measured as 12.5 jykms@xmath29 .
this value is about 30% lower than that of their data .
it is unclear what causes such a significant discrepancy .
note that our hcn(@xmath1 ) line flux measured with the nma is consistent well with that measured with the nro 45-m telescope for central 22@xmath15 region ( see section 3.2 ) .
( 80mm,81mm)figure-3.eps table 4.peaks/nuclei in the central region of m 83 + llll + r.a . & decl . & description & reference + + 13 37 00.90 & -29 51 56.7 & 2mass , ( the central position of the fov ) & @xcite + 13 37 00.73 & -29 51 57.9 & dynamical center determined from co velocity field & this work + 13 37 00.38 & -29 51 53.5 & 95 ghz continuum peak & this work + 13 37 00.52 & -29 51 53.5 & peak of extinction - corrected h@xmath23 luminosity & this work + + 13 37 00.95 & -29 51 55.5 & k - band photometric peak , visible nucleus & @xcite + 13 37 00.57 & -29 51 56.9 & the symmetry center of the outer k - band isophotes & @xcite + 13 37 00.46 & -29 51 53.6 & hidden mass concentration & @xcite + + + we found that the co(@xmath1 ) spectrum at the center of m 83 obtained with the nma is different from that obtained with the nro 45-m telescope @xcite at the same resolution of 16@xmath15 as shown in figure 4 .
this corresponds to missing flux of the interferometry , and we evaluated that the missing flux was @xmath2 30 % . in order to correct the missing flux and obtain the true flux value in the co(@xmath1 ) emission with the nma , we combined the nma co(@xmath1 ) data with the nro 45-m co(@xmath1 ) data .
( 80mm,87mm)figure-4.eps we employed the latest co(@xmath1 ) data ( fukuhara et al . in prep . )
obtained with the nro 45-m using the on - the - fly ( otf ) method .
the combination of two co(@xmath1 ) data sets was performed using miriad .
figure 5 shows a combined co(@xmath1 ) image and an co(@xmath1 ) image only with the nma data .
the results and specification of the combining are summarized in table 5 .
compared to the nma - only image , the flux was properly recovered at the center , and the co - emitting area is enlarged in the combined image .
that is , emission from diffuse components of molecular gas could be reproduced appropriately .
the 45-m data made the resultant noise level lower and the synthesized beam size slightly wider .
however , the overall structure at the center , such as the twin - peaks ( or ring - like ) is little affected by this combining process . for the hcn ( @xmath1 ) line , we did not combine the nma data with single dish data .
it is because we could not find any missing flux in the hcn(@xmath1 ) flux for the central 22@xmath15 region .
the flux obtained with the nma , 38 @xmath25 3 jykms@xmath29 , coincides well with that obtained with the nro 45-m , 39 @xmath25 2 jykms@xmath29 ( hirota et al . in prep ) .
this means that there is no significant diffuse component of hcn(@xmath1 ) emission .
the difference in missing flux between co(@xmath1 ) flux and hcn(@xmath1 ) flux suggests hcn emitting region is more restricted than co emitting region .
this is because hcn(@xmath1 ) emission is originated from dense molecular gas ( @xmath3 @xmath4 a few @xmath5 @xmath6 ) associating with star - forming region , whereas co(@xmath1 ) emission is originated from low - dense ( @xmath3 @xmath2 @xmath34 @xmath6 ) component of molecular gas which is distributed ubiquitously .
( 170mm,93mm)figure-5.eps table 5.parameters of co(@xmath1 ) images before / after the 45-m data combine + llcc + & & nma only & nma + 45-m + + rms noise ( channel map ) & [ mk ] & 320 & 273 + & [ mjybeam@xmath26 & 85 & 80 + rms noise ( intensity map ) & [ k km s@xmath26 & 10.2 & 8.53 + & [ jybeam@xmath29 km s@xmath26 & 2.7 & 2.5 + synthesized beam size & [ @xmath15 ] & 7.2 @xmath22 3.4 & 7.5 @xmath22 3.6 + peak flux &
[ k km s@xmath26 & 350 & 442 + & [ jybeam@xmath29 km s@xmath26 & 93 & 120 + equivalent @xmath28 & [ k ( jybeam@xmath29)@xmath26 & 3.76 & 3.41 + + + we compare the combined co(@xmath1 ) image with the nma hcn(@xmath1 ) image in figure 6 .
this corresponds to the comparison between the total amount of the molecular gas and the dense molecular gas which will host star formation activity soon .
the peak of the hcn(@xmath1 ) intensity is almost coincident with the northern peak of co(@xmath1 ) intensity , whereas little hcn(@xmath1 ) emission was seen at the southern peak of co(@xmath1 ) .
smaller fraction of the dense gas at the southern peak suggests that this region may be in a post star formation phase .
( 80mm,81mm)figure-6.eps we made an intensity ratio map from the hcn(@xmath1 ) data and the combined co(@xmath1 ) data . to improve the data quality and to adjust the beamsize we convolved the images to have the same @xmath35 ( 160 pc @xmath22 160 pc ) resolution before calculating the ratio .
the resultant @xmath8 map is shown in figure 7 .
the peak of @xmath8 value is 0.11 @xmath25 0.01 .
the region with high @xmath8 is concentrated to the north side of the dynamical center of the galaxy .
the high ratio region has an elongation southward toward the dynamical center .
( 80mm,76mm)figure-7.eps to compare our millimeter - wave images with an optical image obtained with the @xmath20/wfpc2 , we made the maps of the combined co(@xmath1 ) intensity , the hcn(@xmath1 ) intensity , the 95 ghz continuum , and @xmath8 , respectively , superposed on the hst v - band image ( f547 m ) obtained from @xmath20/wfpc2 archival data .
these maps are shown in figure 8 .
the v - band image shows the distribution of optically luminous young star clusters , which is referred to as `` optical starburst . ''
the optical starburst region is distributed on the south side of the center , and is confined between the twin peaks in the co(@xmath1 ) emission line .
the distribution of the hcn(@xmath1 ) integrated intensity is shifted northward with respect to that of the optical starburst . moreover ,
even the 95 ghz continuum , which is believed to trace current star formation as well as h@xmath23 , is also clearly shifted northward from the optical starburst region and the peak is close to that in the hcn(@xmath1 ) line .
this means the `` optical starburst '' seen in the v - band does not trace current star - forming region .
the fact is also reported on the basis of near - ir imaging ( @xcite , @xcite ) .
as described in the previous section , we found that the optical starburst region of m 83 is not spatially coincident with current star - forming region traced by the 95 ghz continuum .
this spatial inconsistency is probably due to strong extinction . to confirm this possibility
, we estimate the magnitude and distribution of extinction .
some previous researches conclude that extinction at the center of m 83 is not small .
@xcite reported that v - band extinction @xmath36 spreads from 0.5 to 9.2 mag in the central 12@xmath15 ( @xmath2 250 pc ) region .
@xcite reported that extinction to the north of the center ( @xmath37 or 330 pc ) is stronger than that to the south .
this suggests that @xmath36 is different from place to place and it is very large .
therefore , we should estimate the @xmath36 with enough accuracy . to estimate the h@xmath23 extinction , @xmath38
, we use the pa@xmath23 emission .
we employed an h@xmath23 image obtained with the ctio 1.5-m telescope @xcite and a pa@xmath23 image obtained with archival data of the _ hst_/nicmos camera ( p.i
. m. rieke , proposal i.d .
the pa@xmath23 image is convolved to be the same resolution as that of the h@xmath23 , @xmath39 , although the resolution of the pa@xmath23 image is @xmath40 .
note that the fov of the pa@xmath23 image is narrower than that of nma observations .
we adopted a metallicity - dependent intrinsic ratio h@xmath23/pa@xmath23 = 8.45 in the assumption of electron temperature @xmath41 = 10000 k for @xmath42 @xmath6 @xcite .
this means that without any extinction @xmath43 where @xmath44 means extinction - corrected h@xmath23 luminosity , and @xmath45 means that of pa@xmath23 . from an extinction curve we used , @xmath46 from combination of these two relations , we got formula to calculate @xmath38 .
@xmath47 since @xmath48 and @xmath49 can be calculated from each observed set of data , @xmath38 could be obtained .
figure 9 shows the obtained @xmath38 map and an extinction - corrected h@xmath23 luminosity map for the central 20@xmath15 region of m 83 .
the maximum @xmath38 is about 4 mag , and its location is almost coincident with the nuclear peak of the hcn(@xmath1 ) emission .
@xmath38 @xmath2 4 mag corresponds to @xmath36 @xmath2 5 mag .
thus , the optical light is extinct by a factor of 100 around the hcn(@xmath1 ) peak .
in addition , the peak of the extinction - corrected h@xmath23 luminosity coincides well with those of the @xmath38 and the co(@xmath1 ) line , the hcn(@xmath1 ) line , and the 95 ghz continuum emission . on the other hand ,
extinction is almost negligible to the south of the center .
this inhomogeneous extinction shows why the optical starburst is clearly visible to the south of the center but almost invisible in the north side .
this suggests the existence of deeply buried ongoing starburst with strong extinction , which is already reported by several authors ( e.g. , @xcite , @xcite ) , near the peaks in the hcn(@xmath1 ) line and the 95 ghz continuum emission .
( 160mm,171mm)figure-9.eps in order to discuss the relationship between dense molecular gas and star formation , we estimate the sfr in the central @xmath50 ( @xmath2 500 pc ) region of m 83 .
we used various sfr indicators , i.e. our 95 ghz ( 3 mm ) continuum emission , 5 ghz ( 6 cm ) continuum emission , infrared ( ir ) luminosity , and extinction - corrected h@xmath23 luminosity .
all of them should be consistent when all corrections are applied properly .
continuum emission at 95 ghz is expected to be an extinction - free tracer of sfr .
this is because continuum emission in this wavelength is dominated by thermal free - free emission , and is therefore directly converted to the lyman photon rate @xcite . by assuming that the observed 95 ghz continuum is dominated by the thermal free - free emission
, a lyman continuum rate or h@xmath23 luminosity can be evaluated by the following formula ( @xcite , @xcite ) , @xmath51 where @xmath52 is the distance to the galaxy , @xmath41 is the electron temperature , @xmath53 is the frequency , and @xmath54 is the flux density of the thermal free - free continuum emission . we assume that the observed 95 ghz continuum flux within the central @xmath50 region of m 83 , @xmath54 = 30 mjy , is dominated by the thermal free - free continuum emission .
then , we derived @xmath55(95 ghz ) = 9.2 @xmath22 @xmath56 erg s@xmath29 . using this @xmath55(95 ghz ) , we can derive an sfr by adopting the relation between @xmath57 luminosity and sfr ( @xcite , @xcite ) , @xmath58 the resultant sfr from our 95 ghz continuum flux is 0.73 @xmath25 0.21 @xmath59 .
the error is estimated from the signal - to - noise(s / n ) ratio of the 95 ghz continuum map . in order to evaluate the validity of the derived sfr from 95 ghz continuum flux , we computed the sfr from ir luminosity and 5 ghz radio continuum flux . according to @xcite , the sfr derived from ir luminosity within the central 22@xmath15 region is 0.37 @xmath59 , which is almost half the value of the sfr based on the 95 ghz continuum .
we then derived the sfr from 5 ghz radio continuum flux .
non - thermal radio luminosity is related to the observed radio continuum flux density as @xmath60^{-1 } \right ] \ , { \rm w } \ , { \rm hz}^{-1}\end{aligned}\ ] ] ( @xcite ) , where @xmath52 is the distance , @xmath61 the observed flux density at the frequency of @xmath53 , and @xmath23 the non - thermal continuum spectral index ( @xmath2 0.8 ) .
this non - thermal radio luminosity is related to an sfr as @xmath62 ( @xcite ) . from these equations and the 5 ghz radio continuum map produced by @xcite
, we obtained the sfr of 0.73 @xmath25 0.15 @xmath59 within the central 22@xmath15 region , which is the same value as the sfr based on the 95 ghz continuum .
sfr can be calculated from h@xmath23 luminosity as shown in equation ( 5 ) .
however , the h@xmath23 emission often suffers from extinction by interstellar dust .
in fact , there is up to 4 mag of h@xmath23 extinction in the central region of m 83 as described in the previous subsection .
therefore , there is no doubt that appropriate correction of extinction is indispensable .
here we must verify what data should be used to correct extinction .
pa@xmath23 seems to be very useful , but can not cover the entire fov of co(@xmath1 ) and hcn(@xmath1 ) image .
recently , the spitzer / mips 24 @xmath63 m image has begun to be employed for calibration of sfr ( e.g. m 51 ; @xcite ) .
an archival mips 24 @xmath63 m image ( p00059 , george , rieke , starburst activity in nearby galaxies ) covers the entire disk of m 83 , and the spatial resolution of the image is about [email protected] .
the formula to calibrate h@xmath23 luminosity using 24 @xmath63 m image is as follows @xcite .
@xmath64 where @xmath48 means observed h@xmath23 luminosity , and @xmath65 means that of 24 @xmath63 m .
at the center of m 83 , h@xmath23 extinction derived from mips 24 @xmath63 m image is about 3 mag at a resolution of [email protected] .
considering the difference in spatial resolution , the derived h@xmath23 extinction from mips 24 @xmath63 m image seems to be consistent with that from the pa@xmath23 image .
using equation ( 5 ) , the resultant sfr from extinction - corrected h@xmath23 is 0.24 @xmath25 0.05 @xmath59 within the central 22@xmath15 region .
this value is close to that from ir luminosity , but 3 times smaller than that from 95 ghz and 5 ghz continuum flux .
the sfrs derived from various indicators are summarized in table 6 .
it is unclear what causes such a significant discrepancy among these sfrs .
table 6.star formation rates from various indicators within the central 22@xmath15 region .
+ ccc + indicator & sfr & reference of emission data + + 95 ghz ( 3 mm ) continuum & 0.73 @xmath25 0.21 & this work + 5 ghz ( 6 cm ) continuum & 0.73 @xmath25 0.15 & ( 1 ) + ir luminosity & 0.37 & ( 2 ) + extinction - corrected h@xmath23 & 0.24 @xmath25 0.05 & ( 3 ) + + + reference. ( 1 ) @xcite , ( 2 ) @xcite , ( 3 ) @xcite in order to compare the sfe in the central region directly with our @xmath8 data , we need reliable sfr data with an adequate sensitivity and an adequate spatial resolution higher than @xmath9 . for the ir luminosity data and the 5 ghz continuum data , their spatial resolutions ( @xmath66 10@xmath15 ) are inadequate .
in addition , our 95 ghz continuum data is unfavorable for comparing @xmath8 since the area where 95 ghz continuum emission is detected in adequate s / n ratio ( more than 5 @xmath67 ) is narrower than that of @xmath8 .
then , we use the sfr data based on the extinction - corrected h@xmath23 luminosity using mips 24 @xmath63 m data in order to calculate the sfe in the central region .
the sfe is calculated as follows , @xmath68 @xmath69 is the surface mass density of sfr , and @xmath70 is that of molecular gas . @xmath70 is calculated as follows , @xmath71 here , the @xmath72/@xmath73 conversion factor ( @xmath74 ) is assumed to be @xmath75 @xmath76@xmath77 @xcite .
@xmath78 is derived from the combined co(@xmath1 ) intensity map .
figure 10 shows the calibrated @xmath69 map and the sfe map .
both maps are convolved to the resolution of @xmath35 ( 160 pc @xmath22 160 pc ) to match @xmath8 data .
the peak value of the sfr is @xmath79 . the peak value of sfe ( i.e. the site where most active star formation is supposed to occur ) is @xmath80 and it is located north of the center .
the highest sfe region is spatially coincident well with the highest sfr region .
( 170mm,89mm)figure-10.eps here , we compare @xmath8 with the sfe within the central 1 kpc region of m 83 .
figure 11 shows the map of @xmath8 superposed on that of sfe .
the spatial correlation between these two maps seems to be roughly good , but @xmath8 map seems to trail the skirt toward the northeast .
this skirt - like structure of @xmath8 is not seen in the sfe map , and is coincident with the dust lane seen in the vlt v - band image @xcite .
( 80mm,81mm)figure-11.eps in addition , we examine the correlation between @xmath8 and the sfe . for the region where @xmath8 exceeds 0.02
, the values of @xmath8 and the sfe are obtained for each separation of @xmath81 , which corresponds to half of the spatial resolution of each map .
figure 12 shows a plot of @xmath8 vs. the sfe in each region . the correlation between @xmath8 and the sfe
is clearly seen .
these @xmath8 correspond to the dense gas fraction in the central region of m 83 on a 160 pc ( corresponding to gma ) scale .
the dense gas fraction would be translated as the number of star - forming dense cores per unit gas mass .
then , the correlation derived between @xmath8 means that an outbreak of extensive star formation ( high - sfe star formation ) such as the nuclear starburst requires the generation of a large number of star - forming dense cores within gmas ( e.g. , @xcite , @xcite , @xcite , @xcite ) .
we compare our results with that shown by @xcite .
we converted the sfr in m 83 , which is estimated from extinction - corrected h@xmath23 luminosity , to the total ir ( 8 to 1000 @xmath63 m ) luminosity using the following formula @xcite , @xmath82 then , we adapted the vertical axis of our @xmath8 vs. sfe plot to that of figure 5a in @xcite .
figure 13 shows the composite of the @xmath8 vs. sfe plot for the center of m 83 and for ulirgs , lirgs , and normal spirals @xcite .
the correlation between @xmath8 and the sfe in the central region of m 83 almost seems to coincide with that of @xcite sample .
this suggests that the correlation between @xmath8 and sfe on a gma ( @xmath2 160 pc ) scale found in the nuclear starburst region of m 83 is the origin of the global correlation on a galactic ( a few kpc ) scale shown by @xcite .
in other words , @xmath8 ( dense gas fraction ) and sfe on a galactic scale are averages of those parameters on a gma scale .
low-@xmath8 ( less dense ) gmas would be dominant in a low - sfe galaxy , whereas high-@xmath8 ( dense ) gmas are possibly dominant in a high - sfe galaxy .
this is consistent with the prediction of a three - dimensional , high - resolution hydrodynamic simulation @xcite .
they showed the sfr and sfe are sensitive to increasing average density of molecular gas .
the average gas density just corresponds to @xmath8 .
( 80mm,70mm)figure-12.eps ( 80mm,80mm)figure-13.eps
we have performed aperture synthesis high - resolution ( @xmath0 ) observations in the co(@xmath1 ) line , the hcn(@xmath1 ) line , and the 95 ghz continuum emission toward the central region ( @xmath2 1.5 kpc ) of the nearby barred spiral galaxy m 83 with the nma .
a summary of this work is as follows . 1 .
the size of the co(@xmath1 ) map is 2@xmath21 @xmath22 1@xmath21 ( 3 pointings mosaic observations ) .
the synthesized beam size and the resultant rms noise level of the intensity map are 7@xmath15 @xmath22 3@xmath15 and 2.7 jybeam@xmath29kms@xmath29 , respectively .
our high resolution co(@xmath1 ) mosaic map with the highest sensitivity and the highest spatial resolution to date depicts the presence of molecular ridges along the leading sides of the stellar bar and nuclear twin peak structure .
in addition , we combined the nma co(@xmath1 ) data with the nro 45-m co(@xmath1 ) data .
the combined co(@xmath1 ) map first reveals the high - resolution distribution of molecular gas containing diffuse components in the central region of m 83 .
the size of the hcn(@xmath1 ) map and the 95 ghz continuum map is 77 @xmath15 .
the synthesized beam size is 7@xmath15 @xmath22 3@xmath15 for the hcn(@xmath1 ) map and 8@xmath15 @xmath22 3@xmath15 for the 95 ghz continuum map , respectively .
the resultant noise level is 0.52 jybeam@xmath29kms@xmath29 for the hcn(@xmath1 ) intensity map and 0.85 mjybeam@xmath29 for 95 ghz continuum map , respectively .
we found the distribution of the hcn(@xmath1 ) line emission which traces dense molecular gas shows nuclear single peak structure , and coincides well with that of the 95 ghz continuum emission which traces massive starburst .
however , the peaks of the hcn(@xmath1 ) line and the 95 ghz continuum emission are not associated with the optical starburst traced by the hst v - band image .
3 . using the h@xmath23/pa@xmath23 ratio , an extinction map of the center of m 83 is obtained .
the highest extinction is @xmath83 mag ( @xmath7 mag ) , and which is spatially coincides with the peak of extinction - corrected h@xmath23 luminosity and that of the hcn(@xmath1 ) line emission .
this suggests the existence of deeply buried ongoing starburst due to strong extinction near the peaks of the hcn(@xmath1 ) line and the 95 ghz continuum emission .
4 . we found that @xmath8 correlates well with the extinction - corrected sfe using the mips 24 @xmath63 m data in the central region of m 83 at a resolution of @xmath9 ( @xmath10 pc ) .
that is , the sfe is controlled by dense gas fraction traced by @xmath8 on a gma scale .
in addition , the correlation between @xmath8 and the sfe in the central region of m 83 seems to be almost coincident with that of @xcite sample .
this suggests that the correlation between @xmath8 and the sfe on a gma ( @xmath2 160 pc ) scale found in the nuclear starburst region of m 83 is the origin of the global correlation on a few kpc scale shown by @xcite .
we would like to acknowledge the referee for his invaluable comments .
we are deeply indebted to the nro staff for the operation of the telescopes and their continuous efforts to improve the performance of the instruments .
we are grateful to f. comern for sending us the v - band image of m 83 obtained with vlt , m. fukuhara for providing us his co(@xmath1 ) image obtained with the nro 45-m telescope , and a. hirota for his hcn(@xmath1 ) data obtained with the nro 45-m .
k. m. was financially supported by a grant - in - aid for jsps fellows .
this study was partly supported by the mext grant - in - aid for scientific research on priority areas no .
this research has made use of 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 this work is based on observations made with the nasa / esa hubble space telescope , obtained from the data archive at the space telescope science institute .
stsci is operated by the association of universities for research in astronomy , inc . under nasa contract
nas 5 - 26555 .
this work is based on observations made with the spitzer space telescope , which is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa . | we present aperture synthesis high - resolution ( @xmath0 ) observations in co(@xmath1 ) line , hcn(@xmath1 ) line , and 95 ghz continuum emission toward the central ( @xmath2 1.5 kpc ) region of the nearby barred spiral galaxy m 83 with the nobeyama millimeter array .
our high - resolution co(@xmath1 ) mosaic map depicts the presence of molecular ridges along the leading sides of the stellar bar and nuclear twin peak structure . on the other hand
, we found the distribution of the hcn(@xmath1 ) line emission which traces dense molecular gas ( @xmath3 @xmath4 a few @xmath5 @xmath6 ) shows nuclear single peak structure and coincides well with that of the 95 ghz continuum emission which traces massive starburst .
the peaks of the hcn(@xmath1 ) line and the 95 ghz continuum emission are not spatially coincident with the optical starburst regions traced by the hst v - band image .
this suggests the existence of deeply buried ongoing starburst due to strong extinction ( @xmath7 mag ) near the peaks of the hcn(@xmath1 ) line and the 95 ghz continuum emission .
we found that the hcn(@xmath1)/co(@xmath1 ) intensity ratio @xmath8 correlates well with extinction - corrected sfe in the central region of m 83 at a resolution of @xmath9 ( @xmath10 pc ) .
this suggests that sfe is controlled by dense gas fraction traced by @xmath8 even on a giant molecular cloud association ( gma ) scale .
moreover , the correlation between @xmath8 and the sfe in the central region of m 83 seems to be almost coincident with that of the @xcite sample .
this suggests that the correlation between @xmath8 and the sfe on a gma ( @xmath2 160 pc ) scale found in m 83 is the origin of the global correlation on a few kpc scale shown by @xcite . |
accurate description of the dynamics of relativistic gravitating systems remains one of the challenging problems in general relativity .
there is great interest in particular on the description of systems involving very compact objects , as for example black holes .
it is expected that these kind of systems will be describable in terms of an appropriate isolated model ; which in a relativistic theory of gravity is expressed by an asymptotically flat spacetime . in this description
the gravitational radiated energy will be encoded in the behavior of the physical field at future null infinity . in numerical calculations
these ideas have been used in attacking the problem of the time symmetric asymptotically flat initial data corresponding to a two equal mass black hole system@xcite@xcite .
it would be nice to be able to compare these calculations with some analytic solution of the field equation ; but in trying to do this one is faced with the fact that there is up to now only one known vacuum asymptotically flat radiating family of solutions , the so called robinson - trautman metrics@xcite .
these spacetimes represent the dynamics of a single black hole which settles down , in the asymptotic future@xcite , to the schwarzschild geometry . in the process of two black holes falling into each other in a head - on collision
, one can distinguish two natural eras : one in which the two black holes are separated and falling , and the other in which they already form a single black hole which is decaying to a stationary state .
since the robinson - trautman metrics represent a black hole , one could estimate the total gravitational radiation emitted after the formation of the black hole using these geometries . in doing
this one needs to associate the data of the two black hole collision system to the robinson - trautman data .
we do this by matching , at an arbitrary retarded time , the radiation content of these metrics with the radiation calculated from the quadruple formula with newtonian dynamics at the moment of collapse .
let us consider the newtonian problem of two particles of mass @xmath0 , which start at rest at an initial distance @xmath1 .
one can estimate the power of the energy carried away by gravitational waves when they are at a distance d assuming newtonian dynamics and the quadrupole approximation . more specifically ; the quadrupole radiation formula relates the power of the energy radiated with the time derivatives of the quadrupole : @xmath2 where @xmath3 @xmath4 and we are using geometric units in which the newton s gravitational constant and the speed of light have the unit value .
then , the total power radiated is @xmath5 where @xmath6 is the total mechanical mass and d the distance at observation .
in a previous paper@xcite it was found the general asymptotic behavior of the robinson - trautman metrics in the asymptotic future .
the leading behavior of these metrics is governed by the quadrupole structure of the spacetime .
more explicitly , the whole geometry of these metrics is determined by a scalar @xmath7 , where u , can be considered a retarded time and @xmath8 the coordinates of a 2-sphere . according to ref .
@xcite , for an axisymmetric quadrupole structure , the asymptotic behavior of this scalar , for @xmath9 is @xmath10 where a is a constant @xmath11 is the asymptotic mass of the spacetime in the regime @xmath9 , and @xmath12 is a spherical harmonic function .
this geometry undoubtedly represents the dynamics of a single black hole ; which becomes evident when in the limit for @xmath9 , i.e. , v=1 , one recognizes the schwarzschild metric . as we have described above , there are two natural eras in the process of the head - on collision : before and after the formation of the remaining black hole .
since the robinson - trautman geometries describe the dynamics of a single black hole , it is natural to try to study , with these geometries , the final stage of the process .
the time derivative of the bondi shear of the robinson - trautman sections of future null infinity has the following asymptotic behavior : @xmath13 where now @xmath14 is a spin weight 2 spherical harmonic function and @xmath15 means higher order terms of the time exponential . with this information
one can calculate the flux of gravitational bondi radiation at future null infinity in first order , obtaining : @xmath16 to relate the robinson - trautman geometry with the problem of the head - on collision of two black holes , we identify the bondi power at some retarded time with the power calculated from the quadrupole formula at the moment when the two black holes touch ; that is when @xmath17 . in this way
we determine a appearing in the above formulae . with the bondi flux
one can calculate in first order the total energy radiated , namely : @xmath18 let @xmath19 denote de mass of the spacetime at the _ initial _ retarded time ; then one has the relation @xmath20 therefore , the ratio radiation over initial mass , can be expressed by @xmath21 identifying the bondi with the quadrupole flux one would obtain @xmath22 with @xmath23 and @xmath24 . in this expression
we identify m with @xmath19 ; that is we map the newtonian total mass to the total mass of the spacetime at the initial retarded time .
in reference @xcite it was presented the initial data corresponding to a time symmetric 2-black hole system . there
the adm mass @xmath25 was calculated , and a magnitude @xmath26 ( there denoted with capital l ) was defined , having the interpretation of some measure of initial distance between the black holes .
more precisely @xmath26 was defined as the length of the minimal line from the `` ring '' at minimum radius from one throat to the respective `` ring '' in the other throat .
if one wants to compare calculations coming from different realizations of the same physical situation , namely the head - on collision of two equal mass black holes , one must provide a relation between the physical quantities involved .
the two fundamental physical quantities involved in this situation are the total mass of the system and the separation of the black holes . as we have done in the previous section ,
the identification of the masses is straight forward , since the natural thing to do is to identify the invariantly defined total masses in each case .
there remains however the relation among the notion of distance involved in each case .
the question is : is there a natural way to identify the misner magnitude @xmath26 with the newtonian notion of distance appearing in the estimates of the total gravitational energy radiated using the quadrupole formula approach ?
a simple identification of @xmath26 with d would be too naive in the regime of small distances .
we seek then for some relation between @xmath26 and d such that it captures the most important physical aspects of both systems . in order to do this
let us recall that in reference @xcite the time symmetric initial data for a two equal mass black hole system was represented by the 3-geometry : @xmath27 where @xmath28 is the line element of a flat 3-geometry , @xmath29 and @xmath30 are the euclidean distances from the field point to the deleted points in the flat 3-geometry which represent the location of the first and second black hole respectively , and @xmath31 and @xmath32 are parameters which characterize the spacetime . if in the conformal , scalar appearing in the above line element , one eliminates the third term , the constant time schwarzschild line element is obtained , with mass @xmath31 expressed in terms of isotropic coordinates .
according to the arguments of reference @xcite the mass of , let us say , the first black hole is @xmath33 where now @xmath34 is the euclidean distance between the black holes ; while the total mass of the spacetime is @xmath35 this expression agrees with the first relativistic correction to the newtonian total mass ; since in first order one has @xmath36 therefore it seems that the newtonian distance , for large @xmath37 , should be identified with the coordinate distance of the flat 3-geometry of the time symmetric problem .
so using the line element @xmath38 shown above for the region between the black holes , and identifying the radius of each black hole with double their respective masses , we define the distance between the black holes by @xmath39 which in terms of the total mass m for a two equal mass system can be expressed by @xmath40 then , defining @xmath41 and @xmath42 one has the relation @xmath43 it is observed that with respect of this variable , the black holes do not touch exactly at @xmath44 but when @xmath45 ; which corresponds to @xmath46 .
this picture is in agreement with the numerical calculations since it is found in reference @xcite that for @xmath47 a common apparent horizon surrounds both holes . which corresponds to our @xmath47 .
] we relate d to the misner magnitude @xmath26 by setting @xmath48 .
one can estimate an upper limit for the total energy radiated away in the form of gravitational waves when two very separated black holes collide , from the area theorem ; which using the relation between total mass and single black hole mass shown above and in terms of the quantities already defined , establishes that @xmath49
in ref . @xcite the numerical calculation of the total energy radiated for the time symmetric head - on collision of two equal mass black holes was reported .
they improved previous calculations @xcite with the help of latest generation of supercomputers , and further developed analytic and numerical techniques . in the next graph
we plot their recent@xcite data , our robinson - trautman ( rt ) estimate for the total energy radiated and the upper limit from the area theorem . and @xmath50 respectively .
the vertical bar at approximately @xmath51 corresponds to the region where a single black hole is formed .
the rt estimate is plotted in terms of the relation @xmath52 explained in the main text.,width=377,height=226 ] it can be observed that the rt estimate keeps bellow the supercomputer data ; but this is no surprise since one had expected to only estimate , by this mean , the total energy radiated after the formation of the resulting black hole .
the energy radiated in the first stage , when the black holes fall from @xmath1 to @xmath37 , before the formation of the resulting black hole , can be estimated from the quadrupole formula and using the newtonian dynamics ; that is @xmath53 from which it is deduced that @xmath54 with @xmath55 , @xmath56 and m is the total mass of the system . then adding to the rt estimate the quadrupole newtonian estimate one obtains the total energy radiated during the entire process in the quadrupole approximation , which is shown in the next graph .
this remarkable agreement between our estimates and the supercomputer calculations allows one to think that the global quantities , as the total energy radiated , for a gravitating system like the one under consideration , is not very dependent on the model realization of the system ; since , in particular , the robinson - trautman approach is not time symmetric .
therefore the supercomputer results obtained from the time symmetric two black holes head - on collision system might have a wider range of validity than the one assumed up to now .
it is important to note that our estimates can be calculated for values of l corresponding to distance over mass relation above the threshold of the formation of the remaining black hole ; which from reference @xcite it is deduced to be around l=2.49 .
so our estimates are complementary of those calculated in reference @xcite in which , using a completely different approach , they have estimated with high precision the total energy radiated for small l ; while their method overestimates the total energy in two orders of magnitude in the vicinity of @xmath57 ( which correspond to their @xmath58 ) .
the total gravitational energy radiated in the large separation regime was represented through semianalytic method in reference @xcite by their equation ( 6 ) .
the approach followed in that reference requires of a series of extrapolation in the applicability of known equations .
first of all they use an equation for the total energy radiated calculated in the test particle limit@xcite ( eq .
( 1 ) ) ; that is , one mass much smaller that the other mass ( @xmath59 ) falling from infinity , and then they extrapolate the equation to the equal mass case ( @xmath60 ) .
this approach needs from several correcting factors . using an altered quadrupole radiation formula they suggest a correcting factor , appearing in their eq .
( 4 ) , that takes into account the effects associated to the fact that the infall is not from infinity but from a finite distance .
the alteration in the quadrupole formula comes from a substitution of the radial velocity by a `` nonunique '' of reference @xcite . ]
it is observed that the integral in their eq .
( 4 ) is calculated up to the upper limit @xmath61 , which is consistent with the test particle limit approach , but it overestimates the total gravitational energy radiation in the equal mass case .
therefore they have to introduce another correcting factor smaller than one , there called @xmath62 , which is intended to take into account the black hole structure of both objects . these points were also studied in references @xcite and @xcite .
it is important to emphasize that we use , without alterations , the quadrupole radiation formula for the first stage before the collision , up to the point in which the two horizons touch each other ( @xmath63 ) ; this in contrast to the case of a test particle crossing one horizon ( @xmath64 ) .
then , in order to take into account the black hole structure of the remaining object we use the robinson - trautman geometry , which radiation is to be added to the quadrupole estimate ; this in contrast to the reducing factor @xmath62 . the present estimate of the total gravitational energy radiated is based on a couple of simple ideas with direct physical interpretation ; which is a contribution to the understanding of the global aspects of the head - on black hole collision .
it would be interesting to reproduce these calculations taken into account post - newtonian techniques , as those used in reference @xcite , to see whether these effects will make any substantial difference to our calculations .
a question that motivated us for this work was how a global quantity , as the total energy radiated , could be dependent on the details of the modeling of a physical situation .
on purpose we use a completely different approach , and the result is that this global quantity does not appear to be very dependent on the details of the modeling . as a final comment
let us point out that since the time symmetric initial data obviously incorporates incoming radiation in the system , it is important to be able to know whether this incoming radiation will have an important contribution to the outgoing radiation . since the robinson - trautman geometry can be considered to represent pure outgoing radiating spacetimes , and also the quadrupole estimate only takes into account retarded fields , the agreement of our estimates with the numerical calculations , of the time symmetric problem , suggests that the incoming radiation contribution to the outgoing one is negligible .
it is a pleasure to thank richard price for a series of illuminating discussions , for pointing out to us reference @xcite and for a detailed exposition of their work .
we are also very grateful to peter anninos and collaborators for sending us their recent calculations . | we report on calculations of the total gravitational energy radiated in the head - on black hole collision , where we use the geometry of the robinson - trautman metrics .
pacs numbers : 04.30.db , 04.25.-g , 04.25.dm , 04.70.bw |
magnetic cataclysmic variables ( mcvs ) are close binary stars , consisting of the donor star ( red dwarf ) and magnetized accreting star ( white dwarf ) @xcite .
generally mcvs can be divided into polars and intermediate polars according to the magnitude of their magnetic fields .
some mcvs have the extremely fast - rotating white dwarf .
they have the special place in this division .
it is called superpropellers @xcite .
we can find the intensive outflow of material from these systems because of the additional angular momentum from the rotating magnetosphere to the material . in this case
the accretion disk is not formed .
the possible representative object of superpropellers is ae aqr .
the analysis @xcite has shown that the most part of the strange behavior of this system can be connected with the superpropeller regime of the white dwarf .
on the other hand , the observational data show that , despite the very fast rotation of the magnetosphere in this system , some part of the flow from the inner lagrange point @xmath0 retains in the roche lobe of the accretor . in this paper
we analyze possible physical mechanisms that can reduce the transfer efficiency of the additional angular momentum from the rotating magnetosphere to the material .
the parameters of the system under consideration were adopted from @xcite and correspond to ae aqr .
the donor star has mass @xmath1 and effective temperature @xmath2 k. the accretor star has mass @xmath3 and effective temperature @xmath4 k. the orbital rotation period is @xmath5 h and the orbital separation is @xmath6 .
the estimated magnetic field on the white dwarf surface is @xmath7 mg . one more important characteristic is the proper rotation period @xmath8 s of the white dwarf .
we use the cartesian coordinate system ( @xmath9 , @xmath10 , @xmath11 ) rotating with the orbital angular velocity @xmath12 around the center of mass of the binary system .
we assume that the accreting star has a proper dipolar magnetic field . in the case of asynchronous rotation of the accretor
its magnetic field @xmath13 depends on time , but it is potential : @xmath14 .
we can split the total magnetic field @xmath15 into the background field @xmath13 of the accretor and the field @xmath16 induced by currents in the gas dynamic flows , @xmath17 .
we describe the flow structure using the following equations @xcite : @xmath18 @xmath19 @xmath20,\ ] ] @xmath21 = n^2 \left ( \gamma - \lambda \right ) + \frac{\eta}{4 \pi } \left ( \nabla \times { \bf b } \right)^2,\ ] ] where @xmath22 is the density , @xmath23 the velocity , @xmath24 the pressure , @xmath25 the roche potential , @xmath26 the velocity of the magnetic field lines , @xmath27 the entropy per unit mass , @xmath28 the number density , @xmath29 the proton mass , @xmath30 the magnetic diffusivity .
the dependencies of the radiative heating @xmath31 and cooling @xmath32 functions on the temperature @xmath33 are complicated ( see @xcite for details ) .
the density , entropy , and pressure are related by the equation of state of an ideal gas : @xmath34 , where @xmath35 is the specific heat capacity of the gas at constant volume and @xmath36 is the adiabatic index .
note that our model was developed for the simulation of plasma flows in strong background magnetic fields .
it is based on the averaged characteristics of plasma flow in the frame of the wave mhd turbulence @xcite . in our approximation
the influence of the magnetic field of the accretor @xmath37 is described by the force @xmath38 ( see eq . ): @xmath39 where @xmath40 corresponds to the specific time of the penetration of the magnetic field into plasma and @xmath41 denotes components of velocity perpendicular to @xmath42 .
we use 3d parallel mhd code for the numerical simulations @xcite . in the fig .
[ fg1 ] we present the results of 3d simulations in the frame of the basic model for two values of magnetic field induction of the accretor . in this approximation
the rotation of the white dwarf is so fast that there is no accretion in the system .
the flowing material is just captured by the rotating magnetosphere of the accretor , acquires additional angular momentum and flows away from the roche lobe .
the results show that in the frame of the basic model the material does nt remain in the roche lobe of the accretor even in the case of a relatively weak magnetic field .
the similar results for the ae aqr were also received by other authors in the frame of the quasi - particle method @xcite .
let us investigate the behavior of the stream in the vicinity of the inner lagrange point @xmath0 in the ballistic approximation .
the similar analysis was done by lubow and shu @xcite for the non - magnetic case .
therefore we use their approach but taking into account the magnetic field and rotation of the accretor . the formal basis for this assumption is a supersonic flow pattern , that allows us to neglect the effects of pressure . for this purpose we can use only the equation of motion . in the vicinity of the inner lagrange point @xmath0
this equation can be rewritten in the following form : @xmath43 the left panel of fig .
[ fg2 ] shows the trajectories of test particles obtained in the ballistic approximation in the vicinity of the inner lagrange point @xmath0 for different magnetic fields .
the test particles trajectory in pure hydrodynamic case is in agreement with the corresponding results of lubow and shu @xcite .
the other trajectories for the case of the non - zero magnetic field confirm the results of our 3d mhd simulations .
for the weak magnetic field ( @xmath44 mg ) the trajectories are not strongly different from the non - magnetic case .
but in the case of strong magnetic fields ( @xmath45 mg ) the trajectories significantly deviate from the pure hydrodynamic trajectory and for @xmath46 mg even do not come to the roche lobe of the white dwarf .
our calculations show that for the widely adopted parameters of the magnetic field in the ae aqr system the material almost immediately flows away from the roche lobe of the white dwarf .
so we need to consider some physical mechanisms keeping material in the roche lobe .
one of the mechanisms can be connected with the partial ionization of plasma .
in fact , in the case of very week degree of ionization we should obtain the hydrodynamic solution when an accretion disk forms in the system . in the case of partial ionization of plasma the effect of ambipolar diffusion can be very effective
. the correspondent solution should be intermediate between pure hydrodynamic and ideal mhd ones . after some simple manipulations we can obtain the following expression for averaged electromagnetic force acting on plasma from the external magnetic field : @xmath47 where @xmath48 is the unit vector along the external magnetic field @xmath42
, @xmath49 @xmath50 describes larmor frequency for protons , @xmath9 is the degree of ionization , @xmath51 characteristic time between collisions of ions with neutrals .
analysis of the corresponding solutions shows that the effect of ambipolar diffusion can play an important role in the system for very low degrees of ionization in a range from @xmath52 to @xmath53 .
but the estimation of the equilibrium degree of ionization from the saha equation gives values in a range from @xmath54 to @xmath55 in the inner lagrange point @xmath0 of the system ae aqr .
so this mechanism does nt work in the considered system .
the other mechanism can be connected with relativistic effects in the system due to the very fast rotation of the accretor . to estimate these possible effects we should consider the electromagnetic field of the rotating magnetic dipole in the vacuum .
the corresponding solution of maxwell equations in the laboratory rest frame can be written in the form : @xmath56 @xmath57 + \frac{1}{r^2 c } \left [ 3 { \bf n } \left ( { \bf n } \cdot \dot{{\bf m } } \right ) - \dot{{\bf m } } \right ] + \frac{1}{r c^2 } \left [ { \bf n } \times \left ( { \bf n } \times \ddot{{\bf m } } \right ) \right],\ ] ] where @xmath58 and @xmath59 is the vector of magnetic moment . the characteristic spatial scale for the electromagnetic field , is the light radius @xmath60 . in the ae aqr system the boundary of the light cylinder is located farther than the inner lagrange point but closer than the center of the donor .
so the relativistic effects actually can play an important role in this system .
we can consider the pressure of the magneto - dipole radiation as one of the mechanisms that can keep the material in the roche lobe of the accretor . to include this effect we should add the corresponding force determined by the last terms in eqs . and into our numerical model .
but simple estimations show that in our case in the lagrange point @xmath0 this force is about @xmath61@xmath62 times smaller than the gravitational force of the accretor .
therefore we can neglect this effect , too .
another relativistic effect is the retard of rotating magnetic field lines .
in fact , in relativistic case the electromagnetic field of the rotating magnetic dipole , is determined by the magnetic moment @xmath63 at a retarded time @xmath64 .
in particular , this means that the magnetic field lines rotate not like a solid body .
the retard effect increases with the distance from the acrretor . at the boundary of the light cylinder the velocity of the field lines rotation equals the speed of light @xmath65 . therefore outside the light cylinder the magnetic field in exact solution
can not transfer effectively the angular momentum to the material . in our model
we can describe roughly this effect using an empirical dependence for angular velocity of magnetic lines on radius . the right panel of the fig .
[ fg2 ] shows a result of calculations in the ballistic approximation in the case of exponential dependence for angular velocity of magnetic lines rotation on radius .
it is seen that in the frame of this approach we can easy find the model parameters when the sufficient part of material stays in the roche lobe of the white dwarf .
the results of 3d simulations are shown in fig .
the figure demonstrates the distribution of the density and the velocity field in the equatorial plane of the binary system .
we can see that at initial stage ( left panel of fig .
[ fg3 ] ) the inflowing material forms a ring around the accretor . the gradual accumulation of material in the ring leads to increase of the pressure gradient .
therefore the subsequent evolution leads to the transformation of the original ring into a disk . at a particular time moment ( right panel of fig .
[ fg3 ] ) the material overfills the roche lobe and quasi - periodical overflows begin .
so , the relativistic retard of rotating magnetic field lines can be considered as an effective mechanism for retention of material in the roche lobe of the accretor in mcvs with fast rotating magnetized white dwarfs .
in this work we investigate the flow structure in mcvs taking into account the effects of strong magnetic field and fast rotation of the white dwarf .
the ae aqr system is considered as an example .
we analyzed possible physical mechanisms that can prevent the material to outflow from the binary system . among these mechanisms
are the ambipolar diffusion , the pressure of magneto - dipole radiation and the relativistic retard of rotating magnetic field lines .
our analysis has shown that the last mechanism can be the most effective .
both ballistic calculations and mhd 3d simulations show the formation of fairly stable ring - like structures around the accretor in the ae aqr system caused by presence of the relativistic retard of rotating magnetic field lines .
the gradual accumulation of material in the ring leads to increase of the pressure gradient , and , as a result , to the transforming of the original ring into the disk .
+ the work was supported by the basic research program of the presidium of the russian academy of sciences , russian foundation for basic research ( projects 11 - 02 - 00076 , 12 - 02 - 00047 , 13 - 02 - 00077 , 13 - 02 - 00939 ) , federal targeted program science and science education for innovation in russia 2009 - 2013 . | the flow structure in magnetic cataclysmic variables was investigated taking into account the effects of strong magnetic field and fast rotation of the white dwarf .
we modeled the ae aqr system as a unique object that has the rotation period of the white dwarf is about 1000 times shorter than the orbital period of the binary system .
observations show that in spite of fast rotation of the white dwarf some part of the stream from the inner lagrange point comes into the roche lobe region .
we analyzed possible mechanisms preventing material to outflow from the system . |
background from muon - induced neutrons is one of the most important limitations to detector sensitivity for rare event searches and is largely responsible for the existence of the field of physics called ` underground physics ' and of deep underground laboratories across the world . in wimp
dark matter detectors nuclear recoils of kev energies originating in neutron elastic scattering mimic wimp - nucleus interactions . for double - beta decay experiments high - energy neutrons ( about a few mev and above )
produce background gamma - rays via inelastic scattering while thermal neutrons contribute to the gamma - ray background via neutron capture accompanied by gamma - ray emission .
neutrons at mev energies and above also mimic neutrino detection in scintillators via inverse beta decay posing a severe threat to low - energy neutrino experiments ( reactor and geo neutrinos ) .
neutrons at sub - gev and gev energies , although rare , constitute the background for proton decay and atmospheric neutrino experiments .
neutrons from radioactivity originate in spontaneous fission ( of @xmath13u mainly ) and ( @xmath14,n ) reactions on low and intermediate z isotopes ( z@xmath1530 ) .
their energies are limited to about 10 mev .
neutrons from cosmic - ray muons have spectra extending to gev energies .
although the flux of muon - induced neutrons deep underground is far below that from radioactivity , they can be responsible for a significant background component and limit sensitivity of detectors to rare events due to the following reasons .
firstly , neutrons from radioactivity in rock can be well shielded by hydrogen - rich material , while using ultra - pure materials in detector construction reduces the background from all other components . for muons , however , any material in shielding or detector ( except hydrogen ) is a target for neutron production .
secondly , as muon - induced neutrons have higher energies , it is much more difficult to moderate and/or absorb them .
they can travel far from the muon track or their point of origin reaching detectors from large distances and reducing the efficiency of an anticoincidence system .
any high - a material is also a good target for secondary neutron production from primary neutrons .
all this makes the measurements and calculation of muon - induced neutron rate an important task for designing and constructing sensitive detectors for rare event searches .
neutrons are produced by muons via 4 main processes : i ) negative muon capture ( relevant only to low - energy , stopping muons , or for shallow depths less than 100 m w. e. ) ; ii ) direct muon - induced spallation of a nucleus ; iii ) photoproduction of neutrons or photon - induced spallation ( mainly in electromagnetic cascades initiated by muons ) ; iv ) hadroproduction of neutrons ( mainly in hadronic or nuclear cascades originated by muons ) .
the relative contribution of different processes in different models have been investigated in refs .
@xcite .
there were many attempts to measure the neutron fluxes produced by high - energy muons in the laboratories at surface ( using accelerators ) and underground ( see , for example , @xcite ) and more experiments are planned @xcite .
reliable simulations of the expected effects have become possible only recently with the appearance of powerful monte carlo codes based on advanced theoretical models , such as fluka @xcite and geant4 @xcite .
several measurements of neutron yield in liquid scintillator at different depth @xcite are in agreement with fluka @xcite and geant4 @xcite within a factor of two or better as discussed in refs .
three of them carried out at depths more than 50 m w. e. @xcite show higher neutron yield than fluka and geant4 predictions .
the lvd experiment @xcite reported smaller neutron yield in scintillator than predicted by either fluka or geant4 .
neutron yield in lead has been reported in ref . @xcite as 0.016 n/@xmath1/(g/@xmath3 ) for a mean muon energy of about 310 gev .
this value is two ( four ) times higher than simulations carried out with the fluka ( geant4 ) code @xcite .
significant excess of neutrons in lead over model predictions was observed also in ref .
@xcite .
the evaluation of neutron yields from the experimental data , however , is not straightforward and requires , in its turn , detailed monte carlo simulation of the setup and all physical processes involved .
this is difficult to do now for early experiments since not all details of the setups are known . at the time when these experiments were carried out , the theoretical models and computer codes were not developed to the extent that would allow accurate simulations of the expected effects .
so any comparison between old data and simulations should be taken with caution .
hence there is an urgent need for new experimental data on neutron yields in different materials , as well as neutron energy spectra , lateral distributions etc . supported by accurate monte carlo simulations with widespread multi - purpose codes or toolkits . in this paper
we describe the measurement of the total neutron yield from cosmic - ray muons carried out at the boulby underground laboratory ( boulby mine , north yorkshire , uk ) .
this is the first experiment with a large mass of lead as a target for which direct measurements and detailed monte carlo are compared directly enabling the test of the models .
our results are relevant to many sensitive underground experiments for rare event searches that use or will use lead as a shielding against gamma - rays from rock and hence expect a large background from neutrons produced by muons and their secondaries in lead . due to the relatively simple setup , the systematic uncertainties from geometry , trigger effects etc .
are reduced to the minimum allowing , for the first time , accurate calculations of the expected neutron rate using geant4 version 8.2 . in section [ detector ]
we describe the detector and data acquisition used in the measurements of the muon - induced neutron flux .
section [ results ] shows our results .
monte carlo simulations of the detector setup and physical processes are presented in section [ sim ] .
we discuss and compare our results to simulations in section [ discussion ] .
the conclusions are given in section [ conclusions ] .
the measurements were carried out at the boulby underground laboratory at a depth of about 1070 m or 2850 m w. e. the liquid scintillator of an active veto system working also in anticoincidence with the wimp dark matter detector zeplin - ii @xcite was used for muon and neutron detection .
the veto detector ( figure [ fig - veto ] ) is a hollow structure surrounding the zeplin - ii experiment on five sides .
the lower part of the detector is hemispherical in shape with an inner radius of 0.35 m and an outer radius of 0.65 m giving a distance between walls of 0.3 m. the upper part is cylindrical in shape with the same inner and outer radii and a height of 0.36 m. the vessel is filled with liquid scintillator based on mineral oil containing approximately 25% of phenyl - o - xylylethane ( produced by elgin ) .
this was chosen primarily for its high flash point due to safety constraints on materials used at boulby .
the scintillator has a density of 0.89 g/@xmath16 , average atomic number of 4.75 and average atomic weight of 8.33 .
the wavelength of maximum emission is 425 nm and the attenuation length exceeds 2 m. the light output , as specified by the manufacturer , is 57% of that for anthracene .
our calibration carried out using a @xmath17co gamma - ray source and single photoelectron pulses , revealed a light yield of about 30 photoelectrons per mev .
the volume of the detector is 0.82 m@xmath18 giving a total mass of liquid scintillator of 0.73 tonnes .
the scintillator is viewed from above by ten 20 cm etl hemispherical photomultiplier tubes ( pmts ) .
the inner surfaces of the detector are covered in aluminium with a coefficient of reflection greater than 0.9 to maximise light collection .
more details about the veto detector and its performance relevant to dark matter searches with zeplin - ii can be found in @xcite .
the veto detector ( together with the zeplin - ii liquid xenon detector ) is surrounded by a ` castle ' made of lead and designed to shield the zeplin - ii dark matter detector from gamma - rays from rock .
the thickness of the lead shielding ranges from 15 cm on top to 22.5 cm below the veto and on four sides ( figure [ fig - veto ] ) .
the total weight of lead is about 50 metric tonnes making it an excellent target for neutron production .
another important feature of the setup is the presence of pure and gd - impregnated wax and polypropylene ( about 0.2% of gd by weight on average ) on top of the veto detector under the castle roof .
the inner surface of the veto vessel ( close to the zeplin - ii detector ) was also covered with a paint mixed with gd salt .
the purpose of this was to shield the zeplin - ii target from neutrons from rock .
as we see later , gd did not help in detecting neutron capture events in the time window used in our measurements .
the conclusion may be different for the zeplin - ii dark matter detector where the main neutron background is assumed to be originated just around the detector ( pmts , stainless steel vessel etc . ) .
part of the data run was carried out without the top ( roof ) lead and wax sections , reducing the total lead target mass by about 12% .
this did not have a large impact on neutron production : the neutron event rate was slightly reduced for the time periods without the roof section .
the relatively small reduction can be explained by the fact that wax on top of the detector , that could efficiently absorb neutrons , was also removed in those runs together with lead .
all materials of and around the veto detector and their exact locations within the underground laboratory were put into the simulation code based on geant4 toolkit @xcite as will be described later .
the principle of neutron detection was based on the delayed coincidences between the first , high - energy , pulse from a muon or muon - induced cascade and the delayed second , low - energy , pulse from neutron capture gamma(s ) . the electronics and data acquisition system were designed for this purpose . signals from the ten pmts were passed through discriminators and a coincident unit .
the unit generated a logic pulse if the amplitudes of the analogue pulses from three or more pmts exceeded an average amplitude of one photoelectron within 150 ns .
the logic pulses were used in the off - line analysis to identify ` true ' pulses and reject noise and other background pulses caused by radioactivity close to a single pmt ( for instance on the pmt window ) without coincidences with any other pmt .
analogue signals from all ten pmts were summed together by an adder and the sum signal was fed into a waveform digitiser sampling at a rate of 500 mhz .
the sum pulse was digitised at two different amplitude ranges : 0.5 v and 5 v. the first range was used primarily for low - energy pulses from neutron capture gamma - rays , while the large range was important for identifying muon events .
the adder had a range of about 2 v. all pulses with amplitude exceeding this limit were truncated .
the gamma - ray spectrum from natural radioactivity extends to a few mev corresponding to less than 1 v ( the energy calibration procedure is described in section [ results ] ) .
taking into account the non - uniform light collection , the gamma - ray spectrum extends to almost 2 v in amplitude .
the minimum thickness of the scintillator in the vertical direction is 30 cm or 26.7 g/@xmath3 .
a muon crossing this thickness of scintillator deposits on average about 50 mev which is far above the saturation level of the adder .
so practically all muon pulses were saturated ( truncated ) with the exception of some fraction of events with muons cutting edges of the detector or events caused by low - energy secondaries .
this feature was used in the selection procedure for muon events . in
the ` data ' run the trigger to the daq was provided by large analogue pulses exceeding about 1.2 v. this corresponded approximately to 10 mev energy deposition in the detector .
the rate of recorded events in the data run was a few events / hour consisting of muons and a tail of gamma events seen at higher energies due to their location close to pmts .
the 5 @xmath1s and 195 @xmath1s time periods before and after the trigger pulse , respectively , were digitised allowing for the delayed pulse from neutron capture to be recorded .
the logic pulse , generated by the coincidence unit if at least three pmt hits were above the threshold equivalent to an average amplitude of a single photoelectron , was also recorded on the same waveform 100 ns after the start of the analogue pulse .
only analogue signals accompanied by logic pulses were considered in the off - line data analysis .
event waveforms were parameterised using the data reduction code similar to those written for the neutron measurements from radioactivity in rock described in ref .
@xcite and for the zeplin - ii experiment @xcite . for each pulse on the waveform , the area ( proportional to the charge and , hence , deposited energy ) , amplitude , width , mean time ,
full width at half maximum ( fwhm ) and arrival time were recorded . in addition , the charge in the first 20 ns from the beginning of the pulse and the time during which the amplitude of the pulse exceeded a pre - defined threshold of 1.7 v were calculated .
the last parameter was particularly important for identifying muon events since an accurate measurement of the muon energy deposition was impossible due to the adder saturation .
( even without this effect it would be impossible for pmts and daq to cover with equal precision the mev range relevant for gamma - rays from neutron capture and tens of mev or higher energy depositions from muons and cascades . )
only pulses that exceeded 0.02 v in amplitude ( equivalent to approximately 200 kev energy deposition ) and were accompanied by a logic pulse , were selected for further analysis .
figure [ fig - pulse ] shows a typical waveform for a muon - induced neutron event .
figure [ fig - muon ] displays the time period of about 8 @xmath1s around the trigger .
the muon pulse occurs at a trigger time @xmath19 ns .
it is characterised by a nearly flat top part ( effect of adder saturation resulting in pulse truncation ) , large width ( both mean time and fwhm ) and the presence of many afterpulses during the first 15 - 20 @xmath1s after it .
logic pulses are those with a width of about 0.4 @xmath1s and an amplitude of 0.6 v. as the dynamic range of the pmts was chosen for the best sensitivity to mev energy depositions , higher energy muon pulses saturated pmts resulting in a large number of afterpulses .
some afterpulses from different pmts coincided in time producing logic pulses .
the presence of this ` forest ' of afterpulses made impossible any analysis of neutron captures that could occur within the first 15 @xmath1s after a muon .
only delayed pulses in the time period of 15 - 195 @xmath1s after the trigger were analysed by the data reduction code .
this selection suppresses the detection of gammas from neutron capture on gd .
the two isotopes of gd have very high neutron capture cross - section and , hence , the small mean time delay for neutron capture , about 15 @xmath1s for 0.2% of gd , while the analysis of secondary pulses from neutron capture started only 15 @xmath1s after the muon signal .
a typical pulse from neutron capture gamma - ray(s ) is shown in figure [ fig - pulse ] at a time of about 90 @xmath1s . due to the large time window and small amplitude of the analogue pulse , only the logic pulse can be seen . from all pulses found on the waveform in the time window 15 - 195 @xmath1s after the muon trigger ,
only 20 with highest charge were recorded .
this had some effect on the number of detected neutrons since simulations showed that multiplicities larger than 20 could also be observed .
hence the maximum detected multiplicity is 20 for 15 - 195 @xmath1s time window .
this effect will be studied later .
two types of detector calibration were carried out : energy calibration with a gamma - ray source and calibration with a neutron source .
energy calibrations with @xmath17co gamma - ray source were performed at the beginning of the long data run and at the end of the experiment .
they allowed us to determine the energy scale for events , namely , the conversion from the area of a pulse to the energy deposited in the detector . in the energy calibration run only 1 @xmath1s of the waveform was recorded since only one pulse without any delayed coincidences was expected from gamma - rays .
the amplitude threshold for the daq was reduced to about 0.4 v allowing triggering on the logic pulse and detecting analogue pulses with smaller amplitudes .
figure [ fig - encal ] shows the energy spectra of events from @xmath17co gamma - ray source collected at the beginning and at the end of the data run .
no gamma - ray line is seen in the spectra due to poor energy resolution ( large non - uniformity of the light collection ) of the detector but the shoulders correspond to the combination of the two compton edges and two full absorption peaks of the @xmath17co gamma - ray lines at 1.173 mev and 1.333 mev .
the dashed histograms show the normalised simulated spectra .
the simulations were carried out using the geant4 toolkit @xcite taking into account the geometry of the veto and surroundings and the position of the @xmath17co source .
the energy depositions of photons in scintillator ( through secondary electrons ) were recorded .
the light collection was not simulated but was taken into account by applying a gaussian smearing to the energy deposition in each event .
the standard deviation ( or @xmath20 ) for the smearing was described by the equation : @xmath21 suggested by birks @xcite and used also in the description of the energy calibration results from a small scintillator cell for neutron background measurements at boulby @xcite .
the parameters @xmath14 and @xmath22 , as well as the conversion factor ( for the data ) from the measured pulse area to the energy scale , were determined from a comparison between the measured spectra and simulations .
the two dashed histograms of simulated events differ only by the normalisation factor that determines the vertical scale .
they use the same values for the parameters @xmath14 and @xmath22 . the two data sets ( solid histograms )
are plotted using slightly different pulse area to energy conversion factors , namely : 0.36 mev/(v@xmath23ns ) for the first run and 0.34 mev/(v@xmath23ns ) for the second run showing good stability of the detector over several months of running .
the agreement between simulation and both data sets in the energy range of interest ( 0.5 - 1.5 mev ) demonstrates the reliability of the energy calibration and of the simulation model .
the difference of 6% in the pulse area to energy conversion factors between the two data runs determines the uncertainty in the energy scale .
the difference between data and simulations at low energies is due to the energy threshold ( not simulated ) .
the difference at high energies is probably due to the strong enhancement of the signal if the energy deposition occurs close to the pmts and/or to a possibility of detecting energy depositions of two photons from the same decay ( not taken into account in simulations ) . as will be shown later , more reliable energy calibrations , though with larger uncertainty in the energy scale can be obtained directly from the data using the 2.22 mev gamma - rays from neutron capture on hydrogen .
neutron calibration with an am - be source of 0.1 gbq @xmath14-activity has been carried out before the beginning of the data run .
ideally , the calibration of the experiment with a neutron source would allow the evaluation of detector efficiency to neutrons .
the outcomes of calibration carried out in this experiment , however , were limited due to the reasons outlined below .
firstly , neutrons from the am - be source have on average lower energies than those produced by cosmic - ray muons .
secondly , neutrons are produced by muons everywhere , while the source position was fixed ( on top of the veto detector , just below the castle roof ) .
the neutron capture may thus occur on different materials during the neutron calibration and data runs .
thirdly , the trigger in the neutron calibration run was different from that in the data run .
in the data run the high - energy muon energy deposition was the natural trigger , whereas in the neutron calibration run , in the absence of muons , the low - energy trigger from neutron - induced proton recoils was used . finally , the neutron event rate during the calibration run was a few tens of events per second while the maximum rate handled by the daq ( with almost 100% dead time ) was about 20 hz ( in any run that recorded 200 @xmath1s waveform for each event ) .
hence , the main aims of the neutron calibration were : i ) to demonstrate that the experiment is sensitive to neutrons ; ii ) to show that the monte carlo models are accurate enough in describing neutron transport and detection by comparing measured and simulated time delay distributions of neutron - capture signals .
we point out that similar restrictions to calibration accuracy apply to other experiments that measured muon - induced neutron flux and calibrated their detectors using neutron sources .
figure [ fig - ncal ] shows measured and simulated time delay distributions between the pulses in the events relative to the trigger pulse in the neutron calibration run .
simulation of this run was carried out using geant4 taking into account the geometry of the setup , position of the source , neutron interactions and capture .
the initial energy spectrum of neutrons from am - be source was calculated using the sources4 @xcite code .
two free parameters were used to tune the simulated distribution to match the data : ( i ) the normalisation constant or the total number of neutrons ; ( ii ) the flat ` background ' component that was seen in the data but was not simulated .
the natural background during the calibration run was small and was not responsible for the flat component of the time delay distribution .
instead , the flat component was due to random coincidences between either proton recoil pulses from two neutrons or gamma - ray pulses from two neutron captures occurring because of the high event rate .
the good agreement between measured and simulated time delay distributions gives evidence for the reliability of the geant4 toolkit to model neutron interactions and capture at mev and sub - mev energies .
the expected time delay distribution is not purely exponential since neutron capture occurred on several spatially separated targets ( hydrogen , gd , copper , steel , lead etc . ) .
for this reason , only the comparison between data and monte carlo is shown without any fit .
the data on muon - induced neutrons were collected from august 2006 until april 2007 .
the veto detector was running in parallel with the zeplin - ii experiment .
time periods when the zeplin - ii detector ( together with the veto system ) was exposed to the calibration sources were excluded from the data analysis .
the total live time of the experiment was 204.8 days .
muons were selected as follows : ( i ) trigger pulse area higher than 50 v@xmath23ns corresponding to an energy threshold of about 14 mev ( assuming proper reconstruction of the muon energy deposition , i.e. no saturation of pmts or daq ) ; ( ii ) fwhm greater than 40 ns ; ( iii ) time during which the pulse amplitude exceeded 1.7 v , greater than 10 ns .
only events with a trigger pulse area exceeding 70 v@xmath23ns satisfy all selection criteria rising the energy threshold to about 20 mev ( assuming no saturation ) .
( in the pulse area to energy conversion above we used the more reliable energy calibration using the 2.22 mev gamma - rays from neutron capture on hydrogen in the data run that will be described below . )
this allows rejection of all background gamma - rays keeping more than 90% of muons @xcite .
it also excludes a significant number of events when low - energy secondaries associated with muons are detected .
more details on muon flux simulations and muon detection efficiency can be found in refs .
@xcite . during the experiment 10832
muons were detected translating to a rate of @xmath24 per day in agreement with previous measurements @xcite .
comparing the measured rate with the monte carlo predictions gives the total muon flux at boulby as ( @xmath25 ( stat ) @xmath26 ( syst))@xmath27 @xmath28 s@xmath29 . the systematic error of the muon flux is due to the uncertainty in the energy scale .
the above value is slightly smaller than reported in ref .
@xcite mainly due to the more accurate three - dimensional monte carlo of muon transport in the vicinity of and in the detector ( with the account of all interactions ) . in ref .
@xcite a one - dimensional simplified model was used .
the small reduction in the muon flux compared to previous measurements @xcite is equivalent to a small increase in the estimated depth ( column density ) of the boulby laboratory where the experiment was carried out . the vertical depth reconstructed from the present observations ( assuming flat surface relief above the lab ) is @xmath30 m w. e. ( see ref .
@xcite for details on the depth reconstruction procedure ) . note that the present measurements of the muon flux have been carried out in a new laboratory area at boulby located about 200 m from the old lab where previous experiment @xcite was performed .
the muon flux reported here and in ref .
@xcite for the boulby underground laboratory is defined as the flux through a sphere with unit cross - sectional area .
this definition may be different from those used for some other underground laboratories .
secondary , delayed , pulses were selected using the following criteria : ( i ) energy higher than 2 v@xmath23ns or 550 kev ; ( ii ) presence of logic pulse ; ( iii ) time delay relative to the muon pulse from 20 ns to 190 ns .
the measured and simulated energy spectra of delayed pulses are shown in figure [ fig - spn ] .
the energy scale for the measured spectrum was chosen to achieve the best visual agreement with the position of the simulated peak that corresponds to a combination of full absorption and compton edge of 2.22 mev gamma - rays from neutron capture on hydrogen the main target for neutron capture in this experiment .
when the energy calibration from the @xmath17co run is used , the peak appears at the energy of about @xmath31 mev . for neutron data
, this peak serves as an independent and more accurate energy calibration .
this is due to different positions of gamma - induced events within the detector in the calibration and data runs . in the gamma calibration run
, the source was located between the veto detector and the main zeplin - ii target , close to the bottom of the veto where the light collection was unfavourable . in the data run ,
neutrons from muons were captured evenly in the detector volume .
hence the peak in the spectrum at around 2 mev provides an alternative and , to a certain extent , more reliable energy calibration . superimposing this spectrum onto the simulated one , taking into account the smearing of energy deposition due to finite energy resolution , provides the pulse area to energy conversion factor .
this conversion is shifted by about 20% relative to the @xmath17co calibration data due to the different location of energy depositions as explained above .
using the peak from neutron capture , the energy thresholds for neutron capture gammas was determined as @xmath32 mev , and that for muons as @xmath33 mev .
the uncertainty in the energy resolution and the statistical uncertainty in the peak position are responsible for the error in the conversion factor .
this will be converted later on to the systematic uncertainty in the neutron yield .
this error also gives the systematic uncertainty in the muon flux .
the time delay distribution of pulses in the events relative to trigger ( muon ) pulses is presented in figure [ fig - tdn ] together with the simulations .
both experimental data sets ( with roof section on and off , hereafter called roof - on and roof - off ) were combined together .
details of the simulations will be described below .
only the time window 40 - 190 @xmath1s after the muon trigger is shown for experimental data .
we neglected all events in 0 - 20 @xmath1s time window because of the large number of afterpulses .
in addition we did not consider events that occur at 20 - 40 @xmath1s after the trigger .
this is because the neutron capture at that time may happen with a non - negligible probability on gd ( about 18% according to our simulations as described below ) which is difficult to simulate accurately due to two reasons .
firstly , gd loading and its distribution within wax is not known precisely .
secondly , gammas from neutron capture on gd are not described accurately enough in the geant4 model framework . to avoid ambiguities in the data interpretation
, we restricted the time window to 40 - 190 @xmath1s after the trigger ( 11% probability of capture on gd according to our simulations ) .
neutron capture on gd was included in our simulations but the reduced time window allowed us to minimise possible errors associated with aforementioned uncertainties .
we do not expect a significant change in the neutron rate even if captures on gd were not accurately simulated .
the total number of secondary ( delayed ) pulses with energy deposition greater than 0.55 mev in the time window of 40 - 190 @xmath1s after the muon trigger was measured as 1037 whereas the number of muon triggers was 10832 giving the rate of @xmath34 pulses / muon .
a similar graph for gamma events ( small energy deposition , small fwhm and no truncated amplitude ) is shown in figure [ fig - tdb ] .
the distribution is flat proving that the delayed pulses in gamma events are due to rare random coincidences whereas the quasi - exponential shape of delayed pulses in muon events is due to neutron captures .
random coincidences between gamma background pulses should also be present in muon events with the same rate as in gamma - induced events .
the distribution shown in figure [ fig - tdn ] is in fact the sum of an exponential ( or several exponentials ) due to neutron capture and a flat background component due to random background coincidences .
gamma - induced events allow us to determine the flat background component or the rate of background pulses in muon events .
the number of background events in 40 - 190 @xmath1s time window after the gamma trigger ( figure [ fig - tdb ] ) is 351 for 21461 gamma triggers giving a rate of background pulses of @xmath35 per event .
so the true rate of neutron pulses in muon events is the difference between the total rate of secondary pulses in these events and the rate of pulses due to random background coincidences ( flat component ) .
hence the neutron rate , defined as the average number of detected neutrons per detected muon , is obtained as @xmath0 neutrons / muon in 40 - 190 @xmath1s time window .
the threshold for energy deposition from neutron capture gammas was 0.55 mev .
the error is purely statistical .
systematic uncertainty is mainly due to the uncertainty in the pulse area to energy conversion and will be added to the predicted neutron yield .
figure [ fig - mult ] shows multiplicity distributions for both types of events ( muons solid histogram , and gammas dashed histogram ) , namely the number of events as a function of the number of delayed pulses in an event .
both distributions are normalised to the total number of events of a particular type .
there is a bigger fraction of muon events with non - zero secondary ( neutron ) multiplicity , than the fraction of gamma events ( non - zero multiplicity is due to random coincidences ) .
large neutron multiplicity is expected in some muon events due to enhanced neutron production in lead , proving the neutron origin of delayed pulses .
measured distribution has a maximum multiplicity of 16 .
this number is smaller than the maximum multiplicity cut described in section 2 , since in the final analysis the reduced time window of 40 - 190 @xmath1s was used compared to data reduction procedure ( 15 - 190 @xmath1s ) . for about 40% of the exposure the detector was running without the roof of the castle reducing the lead target mass for neutron production and the mass of gd - loaded wax for neutron moderation and capture .
the two sets of runs ( roof - on and roof - off ) were analysed separately to estimate the effect of the roof .
the random background event rates were determined separately from corresponding time distributions of gamma events . because of the absence of the roof section of the castle ( shielding ) the mean number of secondaries per gamma event with roof off was almost twice that number with roof on .
after subtraction of random background component the neutron rates for two sets of runs were obtained as : @xmath36 neutrons / muon ( roof - on ) and @xmath37 neutrons / muon ( roof - off ) .
measurement of the neutron rate using liquid scintillator at boulby is hard to interpret without the full monte carlo that had been developed using geant4 toolkit .
the presence of several potential targets for neutron production and neutron capture requires a detailed model of the detector , its surroundings and physics processes .
similar considerations are true for any other experiment with similar goals .
the main advantage of our measurements is in the use of a single detector with well controlled energy threshold and systematics that makes its simulation a relatively easy task compared to larger modular detectors .
the modeling of the muon - induced neutrons has been carried out in two stages . in the first stage , muons were sampled using the muon generator musun @xcite according to their energy spectrum and angular distribution at the boulby underground laboratory .
these distributions were obtained by propagating muons from the surface through the boulby rock using the code music @xcite .
the calculated absolute muon flux was normalised to the present muon flux measurements which , as mentioned above , agree well with earlier work @xcite .
the mean muon energy at boulby was calculated as @xmath38 gev .
muons were sampled on the surfaces of a rectangular parallelepiped that surrounded the main experimental hall of the underground laboratory where the detector was placed .
the distance from the surface of parallelepiped to the cavern walls was 7 m on four sides , 10 m above and 5 m below the detector .
this ensured that high - energy neutrons produced in rock far away from the laboratory hall and the detector could still reach it and be detected ( for more discussion about neutron production by muons see ref .
two million muons were generated in this way and their parameters ( energy , position , direction cosines and sign ) were recorded and passed to geant4 on the second stage .
neutron production , transport and detection were simulated using geant4 @xcite version 8.2 .
the laboratory hall , lead and wax shielding , gd , veto and zeplin - ii detector were included in the simulations .
detailed description of the simulations , various validation tests and comparison with other simulation work will be presented in a separate paper @xcite . here
we present only the results important for the interpretation of the experimental data .
the physics models and particle production thresholds ( cuts ) were essentially those described in ref .
@xcite , although other configurations were studied as described in ref .
all hadronic and electromagnetic processes were taken into account in the simulations .
the hadronic interactions were simulated using the quark - gluon string model above 6 gev , an intra - nuclear binary cascade model at lower energies and a pre - equilibrium de - excitation stage below 70 mev .
neutron interactions below 20 mev were treated using high - precision data - driven model
. the production thresholds ( ` cuts ' ) considered in these simulations were a few tens of kev for gammas and @xmath391 mev for electrons and positrons in all materials .
no thresholds were applied to neutron tracking .
muon parameters were read from the musun output file and passed to geant4 for further muon transport .
stochastic muon interactions and continuous energy loss due to ionisation were simulated .
all muons and secondary particles produced by the muons and in muon - induced cascades ( showers ) were transported and their energy depositions in the veto detector and interaction times were stored in memory .
finally , for each muon event we recorded the energy depositions for different time bins covering 01 @xmath1s ( 10 bins ) , 1200 @xmath1s ( 199 bins ) and 200500 @xmath1s ( 30 bins ) .
this allowed us to select events imposing the same cuts as for real data sets and plot distributions in a similar way as for the data run for direct comparison . in total ,
about 120 million muons were sampled , each of the pre - recorded muons was sampled about 60 times , but transported differently using different random numbers .
this corresponded to a run time of about 960 days .
the statistics are about 4.7 times that for real data .
comparison between data and simulations is shown in figures [ fig - spn ] , [ fig - tdn ] , [ fig - mult ] . reasonable agreement between measured and simulated spectra of delayed gammas , shown in figure [ fig - spn ] , allowed us to establish the energy scale and evaluate the threshold for muons and delayed gammas .
we estimated the uncertainty in the energy scale as about 20% that leads to an uncertainty in the energy threshold for delayed gammas of about 0.1 mev .
changing the pulse area to energy conversion or energy resolution by this factor destroys the agreement in the peak shape and position .
this does not exceed the shift in the energy scale if the @xmath17co spectrum is used for energy calibration , confirming the scale of the systematic uncertainty .
the uncertainty in the energy threshold for muons may be a little higher since a saturation effect can suppress slightly the measured energy deposition from muons .
we estimated the total systematic uncertainty in the energy scale for muons as about 25% .
this is consistent with agreement between muon rates reported here and in ref .
@xcite where the suppression of energy deposition for muons occured at much higher energies . large suppression of muon energy deposition at threshold , leading to higher effective energy threshold , would manifest itself in a smaller muon rate measured in the present experiment .
systematic uncertainties in energy scales for muons and delayed gammas were converted into the uncertainty in the simulated neutron rates and will be reported below .
figure [ fig - tdn ] shows the simulated time delay distributions compared to the experimental data described above .
two sets of simulations ( roof - on and roof - off ) are plotted separately to show good agreement in shape at 40 - 190 @xmath1s used for data analysis . at small time
delays a visible difference is due to the presence of gd in the roof shielding in the run with the roof section resulting in the enhanced capture of neutrons on gd at small times .
the same energy thresholds for muon and delayed gamma energy depositions were applied to the simulated events as for data .
the energy resolution was chosen to match the energy spectrum of delayed gammas ( see figure [ fig - spn ] ) .
as for the neutron calibration run , tuning of simulated spectra was done using two free parameters : i ) the total number of neutrons ; ii ) the flat component due to random background coincidences ( not simulated ) .
the second parameter is bound by @xmath40 standard deviations from the measured background rate based on gamma - induced events ( see section [ results ] ) .
figure [ fig - tdn ] shows good agreement between measured and simulated shapes of the time delay distribution .
note that the absolute normalisation was chosen to reach good visual agreement and hence no conclusion about absolute neutron rate can be drawn based on this figure .
the rate of simulated pulses from neutron capture gammas in 40 - 190 @xmath1s can be directly compared to the measurements .
this does not require any assumption about the shape of the time delay distribution other than a general agreement between measured and simulated distributions ( shown in figure [ fig - tdn ] ) .
measured and simulated neutron rates are compared in table [ tab1 ] for the two runs separately and combined together . only statistical errors are shown in the table .
systematic uncertainty is mainly due to the 20% uncertainty in the energy scale as discussed above .
this was converted into the uncertainty of 0.009 neutrons / muon in the simulated neutron rate in the 40 - 190 @xmath1s time window with a threshold of 0.55 mev for delayed pulses .
table [ tab1 ] shows a factor of 1.8 difference between measured and simulated neutron rates for both runs , geant4 predicting higher neutron rate than measured in the present experiment . figure [ fig - mult ] shows the simulated neutron multiplicity distribution ( runs with roof - on ) compared to the data ( all runs ) .
a smaller rate of events is observed for almost all neutron multiplicities .
no more than 20 pulses on each waveform were recorded in the data run . since a restricted time window of 40 - 190 @xmath1s
was used in the analysis , the effective multiplicity cut was found to be 16 delayed pulses ( see figure [ fig - mult ] ) . taking the fraction of simulated events with multiplicity larger than 16 we estimated the fraction of missed neutrons if the maximum multiplicity cut of 16 was applied , as less than 2% .
this is smaller than the statistical error of the measurements and does not affect the results . a smaller observed neutron rate compared to simulations can be due to two main factors : i ) smaller neutron yield than predicted in geant4 models ; ii ) enhanced absorption of neutrons or gammas before neutrons or neutron capture gammas can reach the detector .
the second argument is highly unlikely because of the following reasons .
firstly , the measured time delay distribution matches well the simulated one for the time window chosen .
the shape of the time delay distribution depends on the neutron capture cross - section and , hence , is affected by the geometry and neutron capture models .
good agreement between data and simulations proves that the neutron capture and secondary gammas are described reasonably well by the geant4 models .
secondly , a similar agreement in shape was seen in the neutron calibration data proving again a good model for neutron capture and gamma transport and detection .
finally , neutron transport at mev and sub - mev energies was simulated with geant4 and mcnpx @xcite and they were found to be in good agreement .
hence , the most probable explanation of the observed deficit of muon - induced neutrons lies in the geant4 model(s ) involved in neutron production .
the ratio of roof - off / roof - on neutron rates is consistent with unity for both data ( within two standard deviations ) and simulations .
lead in the roof section contributes about 12% to the total mass of lead .
lead is responsible for about 91% of neutrons in the run with roof - on and about 90% of neutrons in the run with roof - off ( results from our simulations ) .
the fact that neutron rates are similar in both sets of data despite a 12% difference in lead mass can be explained by the presence of gd - loaded wax in the roof section for runs with roof - on .
this enhances neutron moderation and capture above the detector preventing neutrons ( gammas ) being detected in scintillator in runs with roof - on .
thus , enhanced neutron production in lead in the roof section was compensated by the efficient absorption of these neutrons before they can reach the scintillator . since lead contributes about 90% to the detected neutron rate
, we can conclude that geant4 probably overestimates the neutron yield in lead .
this result is somewhat surprising since geant4 was found to underproduce neutrons relative to fluka monte carlo @xcite .
alternative geant4 hadronic models result in similar neutron yields . for the most widespread geant4 models
the difference is within 20% @xcite .
recently developed chips models @xcite predict even higher flux of neutrons and , hence , are probably inconsistent with the present measurements .
more detailed studies will follow @xcite .
the time window of 40 - 190 @xmath1s used in the data analysis contains about 42% of the total number of neutrons according to the simulations ( 41% with roof section and 44% without roof section ) , assuming a well known geometry .
hence , the total neutron rate for an infinite time window can be evaluated as @xmath41 neutrons / muon for the simulated geometry .
this figure is still geometry dependent . for practical use
this figure has to be converted to the neutron yield in a specific material , defined as the average number of neutrons produced by a muon along 1 g/@xmath3 of its path .
the most reliable and quasi model independent way of doing this is based on the following assumptions : 1 ) the geometry , neutron moderation and capture are described correctly in geant4 ; this is proven by the time delay distributions of data and calibration neutrons and by the agreement reached in ref .
@xcite for mev and sub - mev neutron transport ; 2 ) most neutrons ( about 90% ) detected in our scintillator are produced in lead - this follows from simulations ; 3 ) the ratio of measured - to - simulated neutron rates per detected muon is thus the same as for raw neutron yields in lead .
using the fact that the measured neutron rate is 1.8 times smaller than geant4 predictions , and the calculated neutron yield in the same model in lead is @xmath42 neutrons / muon/(g/@xmath3 ) for a muon energy of 260 gev , we found the yield in lead , reconstructed from our measurements , to be @xmath2 neutrons / muon/(g/@xmath3 ) , for a mean muon energy of about 260 gev .
this value is 3.6 times smaller than the expected neutron yield from fluka simulations @xcite .
although our lead target and detector are of relatively small size , the fractional contribution of physical processes to the measured neutron rate does not differ much from contributions of these processes to the total neutron yield ( see , for instance , @xcite for discussion ) , allowing us to convert our measured rate to the total neutron yield as described above .
our simulations show that 2.2% of detected neutrons are coming from the muon spallation process , 63% were initiated by photons or electrons and 34.8% were produced by hadrons either via capture or hadron inelastic scattering .
although most neutrons detected in our scintillator were produced in lead , cascades could be originated also in the rock around the setup , thus increasing the fraction of detected neutrons from photons and hadrons .
the hypothetical contribution of a muon - induced neutron background for dark matter searches was checked using data from the zeplin - ii science run @xcite with 225 kg@xmath23days exposure .
zeplin - ii was operated in coincidence with the active veto system described here .
the data did not contain any neutron - induced nuclear recoil event in the target ( zeplin - ii ) in coincidence with a large muon signal in the active veto showing that for detectors with a few kilograms of target mass this background is not a severe threat at a depth of about 3 km w. e.
the first measurements of the muon - induced neutron flux at the boulby underground laboratory were presented .
the experiment was carried out with an 0.73 tonne liquid scintillator .
the delayed coincidence method was used to detect a muon ( the first pulse in an event ) and gamma - rays resulted from neutron capture on hydrogen or other elements ( secondary , delayed , pulses ) .
the muon - induced neutron rate was measured as @xmath0(stat . )
neutrons / muon using neutron - capture signals above 0.55 mev in a time window of 40 - 190 @xmath1s after the muon trigger .
accurate monte carlo simulations of the neutron production , transport and detection in a precisely modeled laboratory and experimental setup were carried out using the geant4 toolkit .
the simulations gave a rate of @xmath43(stat.)@xmath44(syst . )
neutrons / muon for the same selection criteria .
the simulated result is 1.8 times higher than the measured value and the difference largely exceeds all statistical and systematic uncertainties .
as the vast majority of neutrons detected in the current setup were produced in lead we evaluated from our measurements the total neutron yield in lead as @xmath2 neutrons / muon/(g/@xmath3 ) for mean muon energy of about 260 gev .
this work has been supported by the ilias integrating activity ( contract no . rii3-ct-2004 - 506222 ) as part of the eu fp6 programme in astroparticle physics .
we acknowledge the financial support from the particle physics and astronomy research council ( pparc , currently science and technology facility council
stfc ) , the us department of energy ( grant numbers de - fg03 - 91er40662 and de - fg03 - 95er40917 ) , the us national science foundation ( grants phy-0139065 and phy-0653459 ) and from fundao para a cincia e tecnologia ( project poci / fp/81928/2007 ) .
we are grateful to the nuffield foundation for the support of j. blockley through undergraduate research bursary ( grant urb/34132/2007 ) .
we also wish to thank cleveland potash ltd for assistance .
a. fass , a. ferrari and p. r. sala .
_ proceedings of the montecarlo 2000 conference _
( lisbon , october 2326 , 2000 ) , ed . a.kling , f.barao , m.nakagawa , l.tvora , p.vaz ( springer - verlag , berlin , 2001 ) , p. 159 ; a. fass , a. ferrari , j. ranft and p. r. sala , _ ibid . _ p. 995 .
w. b. wilson et al . , _ sources-4c : a code for calculating ( alpa , n ) spontaneous fission and delayed neutron sources and spectra _ , american nuclei society / radiation protection and shielding division ( 2002 ) .
p. antonioli , c. ghetti , e. v. korolkova , v. a. kudryavtsev and g. sartorelli .
_ astroparticle phys .
_ , * 7 * ( 1997 ) 357 ; v. a. kudryavtsev , e. v. korolkova and n. j. c. spooner , _ phys .
b _ , * 471 * ( 1999 ) 251 .
m. bauer , j. jochum and s. scholl .
5th intern .
workshop on the identification of dark matter _
( edinburgh , uk , 6 - 10 september 2004 ) , ed . by n. j. c. spooner and vitaly kudryavtsev ( world scientific , singapore , 2005 ) , 494 .
.measured and simulated rates of delayed pulses due to neutron capture given in neutrons / muon : 2nd column ( roof - on ) for runs with roof section ; 3rd column ( roof - off ) for runs without roof section ; 4th column ( ratio ) ratio of rate without roof section to that with section ; 5th column ( combined ) the data from all runs are combined together . delayed pulses with energy deposition greater than 0.55 mev were counted in the 40 - 190 @xmath1s time window after the muon trigger .
only statistical error is shown for simulation results .
systematic uncertainty of simulations is largely dominated by the energy scale uncertainty ( pulse area to energy conversion factor ) and is equal to 0.009 for the two sets of runs and for the combined data . for the ratio of rates in the two runs , the systematic uncertainty cancels out . [ cols="^,^,^,^,^",options="header " , ] .
pulses with about 0.6 v amplitude are logic pulses .
muon events are characterised by a large number of afterpulses accompanied by logic pulses preventing reliable neutron detection 20 @xmath1s after the trigger . ]
co energy calibration runs : upper solid histogram first calibration in august 2006 before the beginning of the data run ; lower solid histogram
second calibration in march 2007 , close to the end of the data run .
dashed histograms show normalised simulated spectra . ] | we present the first measurements of the muon - induced neutron flux at the boulby underground laboratory .
the experiment was carried out with an 0.73 tonne liquid scintillator that also served as an anticoincidence system for the zeplin - ii direct dark matter search .
the experimental method exploited the delayed coincidences between high - energy muon signals and gamma - rays from radiative neutron capture on hydrogen or other elements .
the muon - induced neutron rate , defined as the average number of detected neutrons per detected muon , was measured as @xmath0 ( stat . )
neutrons / muon using neutron - capture signals above 0.55 mev in a time window of 40 - 190 @xmath1s after the muon trigger .
accurate monte carlo simulations of the neutron production , transport and detection in a precisely modeled laboratory and experimental setup using the geant4 toolkit gave a result 1.8 times higher than the measured value .
the difference greatly exceeds all statistical and systematic uncertainties . as the vast majority of neutrons detected in the current setup
were produced in lead we evaluated from our measurements the neutron yield in lead as @xmath2 neutrons / muon/(g/@xmath3 ) for a mean muon energy of about 260 gev . * measurements of neutrons produced by high - energy muons at the boulby underground laboratory * h. m. arajo@xmath4 , j. blockley@xmath5 , c. bungau@xmath6 , m. j. carson@xmath5 , h. chagani@xmath5 , e. daw@xmath5 , b. edwards@xmath4 , c. ghag@xmath7 , e. v. korolkova@xmath5 , v. a. kudryavtsev@xmath5 , p. k. lightfoot@xmath5 , a. lindote@xmath8 , i. liubarsky@xmath6 , r. lscher@xmath9 , p. majewski@xmath5 , k. mavrokoridis@xmath5 , j. e. mcmillan@xmath5 , a. st . j. murphy@xmath7 , s. m. paling@xmath5 , j. pinto da cunha@xmath8 , r. m. preece@xmath9 , m. robinson@xmath5 , n. j. t. smith@xmath9 , p. f. smith@xmath9 , n. j. c. spooner@xmath5 , t. j. sumner@xmath10 , r. j. walker@xmath10 , h. wang@xmath11 and j. white@xmath12 @xmath10 _ blackett laboratory , imperial college london , uk _ + @xmath9 _ particle physics department , stfc rutherford appleton laboratory , uk
_ + @xmath5 _ department of physics & astronomy , university of sheffield , uk _ + @xmath7 _ school of physics , university of edinburgh , uk _
+ @xmath8 _ lip coimbra & department of physics of the university of coimbra , portugal
_ + @xmath11 _ department of physics & astronomy , university of california , los angeles , usa _ + @xmath12 _ department of physics , texas a&m university , usa _ + |
today , material research is focused toward a technology base whereby miniature devices operating at high speed , over a wide range of frequencies and exhibiting multifunctional properties can be developed routinely and efficiently@xcite .
titanium dioxide tio@xmath3 is a well known wide band gap oxide semiconductor belonging to the group iv - vi semiconductors .
it is described in terms of a model structure of ti@xmath4 and o@xmath5 ions .
it has large dielectric constant and dielectric anisotropy with various crystal structures .
tio@xmath6 is known to be an n - type semiconductor with a large energy gap in the range @xmath7 depending on sample preparation@xcite .
previously@xcite , thin tio@xmath0 films of thickness in the range @xmath8 were deposited on substrates of ( 100 ) of lanthanum aluminate laalo@xmath1 .
bulk tio@xmath6 does not order magnetically at room temperature . however , spontaneous magnetization in tio@xmath0 films@xcite occurred in the temperature range @xmath9 .
the spontaneous magnetization was attributed to the oxygen deficiency induced during the growth of the films .
it was argued that oxygen deficiency caused an imbalance in the ionic charge neutrality of the film thereby creating electron donor impurity ions of ti@xmath10 and ti@xmath11 in addition to the usual ti@xmath4 ions .
ferromagnetism , i.e. net magnetic moments in the ti@xmath10 and ti@xmath11 ions , was a result of superexchange interactions between these ions via the oxygen ions .
calculations indicate that of all possible superexchange permutations , double exchange may be the most predominant interaction in view of the filled electron bands of the oxygen ions@xcite .
it is well known@xcite that such double exchange favors ferromagnetism .
our purpose is to argue that the _ same _ magnetic ions and oxygen deficiency that gave rise to magnetism in tio@xmath0 films also play a vital role in the electronic _ transport properties_. specifically , the existence of donor ti@xmath10 and ti@xmath11 donor ions in our films allow for electronic excitations into conduction bands .
the number of carriers in the conduction band varies only mildly with magnetic field and more strongly with temperature .
indeed , the measurements reported here of the hall resistivity is approximately linear in the magnetic field and the the magneto - resistivity is only very mildly field dependent .
the magneto- and hall- resistivity have very similar temperature dependence reflecting the variation of the number of mobile electron transport carriers with temperature .
measurements of transport properties of tio@xmath6 films as well as the theoretical analysis are discussed in the next section .
the films were deposited employing a pulsed laser ablation deposition ( pld ) technique at an oxygen pressure @xmath2 mtorr .
final arguments are presented in the concluding section .
pulsed laser ablation deposition ( pld ) techniques were employed with a weakly paramagnetic magnetic tio@xmath6 target .
thin films of oxygen deficient tio@xmath0 were produced on ( 100 ) lanthanum aluminate laalo@xmath12 substrates . during the pld process
, the substrate temperature was fixed at @xmath13 while varying the oxygen pressure in the range @xmath14 .
films deposited at @xmath15 exhibited spontaneous magnetization and were therefore selected for measurements of magneto - transport film properties . the thickness of the films were in the range @xmath8 .
the details of the magnetic and crystallographic characterization as well as other impurities and contaminants in the films were previously reported@xcite . for steady currents and with the magnetic intensity @xmath16 directed normal to the film , the resistance matrix @xmath17 in the plane of the film
may be written as@xcite @xmath18 wherein @xmath19 and @xmath20 represent , respectively , the magneto - resistivity and the hall resistivity .
the resistance matrix @xmath17 of the tio@xmath0 films was measured by a conventional four - terminal technique using a gold - plated resistance sample puck from the quantum design physical property measurement system ( ppms ) .
in that the ferromagnetic moment per unit volume is small on the scale of applied magnetic intensities , @xmath21 , the magnetic field intensity may be identified with the magnetic induction @xmath22 , i.e. @xmath23 . as a consequence of such a weak magnetization @xmath24 , the anomalous hall resistivity is negligible .
the original hall expression for @xmath20 is sufficiently accurate .
the magneto - resistivity is here described by the drude model .
altogether , we may analyze the data in terms of the simple conventional expressions @xmath25 wherein @xmath26 is the density per unit volume of carriers in the conduction band . in fig .
[ fig1 ] is shown measurements of @xmath19 at zero magnetic intensity and @xmath20 at @xmath27 tesla is plotted as a function temperature @xmath28 .
at zero magnetic field intensity and the hall resistivity @xmath20 at @xmath29 tesla versus inverse temperature .
the temperature variations are quite similar , i.e. the hall angle @xmath30 is fairly uniform in temperature.,scaledwidth=45.0% ]
one notes that the temperature variations of @xmath19 and @xmath20 are quite similar as they are in other semiconductors@xcite the hall angle @xmath31 , as defined by @xmath32 is experimentally fairly uniform in temperature . for two temperatures , @xmath33 and @xmath34 . in the high temperature regime ,
the linear behavior in magnetic field @xmath35 is accurately obeyed . in the low temperature regime
there are experimental deviations from linear behavior.,scaledwidth=45.0% ] the hall resistivity @xmath20 as a function of magnetic field were measured at fixed temperature with applied field sweeps in the range @xmath36 .
the data for a high and low temperature are plotted in fig .
[ fig2 ] . for high temperatures ,
the experimental data is in excellent agreement with the linear magnetic field behavior for @xmath20 in eq.([transport2 ] ) . for low temperatures ,
the agreement with the linear behavior in magnetic field is only fair .
nevertheless , it is experimentally clear that the carriers are n - type and the carrier densities may be determined by @xmath37 quite accurately for high temperatures and somewhat less accurately for low temperatures .
the resulting variation of the density of carriers @xmath26 with temperature @xmath28 is shown in fig .
[ fig3 ] . as a function of temperature taken from hall resistivity data in a fixed magnetic field of @xmath29 tesla . also shown ( solid curve ) is the fit to eqs.([transport5 ] ) and ( [ transport6 ] ) with @xmath38 and @xmath39.,scaledwidth=45.0% ] the two striking features of the resulting inferred density of carriers are as follows : ( i ) the density of carriers does not appear to vanish in the low temperature limit @xmath40 .
in fact , there appears to be a low density of carriers in the ground state tio@xmath41 system of approximately @xmath42 at liquid helium temperatures .
this suggests that the electron chemical potential is placed very slightly into the conduction band by the ti@xmath10 donors .
the double exchange mechanism lowers that part of the conduction band containing electrons with spins parallel to those localized spins producing the magnetization@xcite .
these effects allow for the low density of ground state carriers .
( ii ) the density of carriers as a function of temperature exhibits a clear maximum .
we attribute this maximum in the number of carriers to the decrease of the magnetization @xmath43 to values below the saturation value @xmath44 as the temperature is raised .
theoretically , one finds from spin wave theory that @xmath45/t^{3/2}=a_0 $ ] .
for the temperature range here of experimental interest , we find that @xmath46 $ ] .
if the rise in the carrier band lowest energy state is proportional to the the deviation of the magnetization from saturation@xcite , i.e. if @xmath47 , then a reasonable phenomenological expression for the density of carriers , both in the ground state and in the thermally excited states , is @xmath48\ , \label{transport6}\ ] ] wherein the electron thermal wavelebgth @xmath49 @xmath50 is a dimensionless constant and @xmath51 is derived from the boltzman factor @xmath52 on the right hand side of eq.([transport6 ] ) .
that the density of carriers is the sum of ground state and thermally activated terms is at the heart of our phenomenology . as can be seen from the data in fig .
[ fig3 ] , the model gives rise to a good fit to the experimental data . of the magneto - resistance in eq.([transport7 ] ) is plotted as the function of magnetic field .
the blue squares and red circles represent , respectively , the measurements taken at temperatures @xmath53 and @xmath54.,scaledwidth=45.0% ] finally in fig . [ fig4 ] , the magneto - resistivity is plotted as a function of magnetic field in both the high and low temperature regimes .
one may view the fractional change in magneto - resistivity , @xmath55 as a measure of how strongly the carrier density @xmath26 and the carrier lifetime @xmath56 vary with magnetic field . in the high temperature regime ,
the fractional change in the magneto - resistivity is quite small @xmath57 . in the low temperature regime
the fractional change in magneto - resistivity is again fairly small @xmath58 .
altogether , the magnetic field variations of the magneto - resistivity obeys @xmath59 in all regimes .
that the deviations are small provides experimental support for the hall - drude model of electronic transport in eq.([transport2 ] ) for the oxygen deficient tio@xmath41 ferromagnetic semiconductor .
measurements of the magneto- and hall- resistivity have been measured in oxygen deficient tio@xmath0 magnetic semiconductor films .
the data can be understood on the basis of the hall - drude model of resistivity from mobile carriers .
the anomalous magnetic moment contribution to the hall resistivity is small in most regimes since the ferromagnetic moment @xmath60 resides mainly on the dilutely distributed ions ti@xmath10 and ti@xmath11 and the magnitude is small @xmath21 .
the drude relation time @xmath56 is only weakly dependent upon the magnetic field @xmath61 and the temperature @xmath28 so that both the hall resistivity @xmath20 and the normal resistivity @xmath19 vary similarly with temperature .
the experimental data can thereby be described in terms of a temperature dependent mobile carrier density @xmath62 .
apart from an expected zeeman internal magnetic field splitting of the conduction band wherein the carrier spins align themselves parallel to the spins of the local ionic ferromagnetic moments , there also appears to be a small but finite mobile carrier density that persists even in the @xmath40 limit .
the theory of such quantum ground state carriers in double ferromagnetic exchange semiconductors is worthy of further investigation . | tio@xmath0 films were deposited on ( 100 ) lanthanum aluminates laalo@xmath1 substrates at a very low oxygen chamber pressure @xmath2 mtorr employing a pulsed laser ablation deposition technique . in previous work , it was established that the oxygen deficiency in these films induced ferromagnetism . in this work
it is demonstrated that this same oxygen deficiency also gives rise to semiconductor titanium ion impurity donor energy levels .
transport resistivity measurements in thin films of tio@xmath0 are presented as a function of temperature and magnetic field .
magneto- and hall- resistivity is explained in terms of electronic excitations from the titanium ion donor levels into the conduction band . |
charge symmetry violation ( csv ) , in itself , is an interesting physical phenomenon . while the charge symmetry ( cs ) implies that the interaction between two neutrons or two protons are equal , but , in nature , this is found to be only approximately true .
the violation of cs automatically violates charge independence ( ci ) , however , the converse might not be always true @xcite .
it is possible to have cs even if the ci is violated which actually is a higher symmetry .
the csv , at the qcd level , is caused by the splitting of @xmath5-@xmath6 quark masses .
experimentally csv can be observed at various levels .
for instance , in @xmath2 interaction , the effect of csv is traditionally inferred from the difference of the @xmath7 and @xmath8 scattering lengths in the @xmath9 state .
the most recent scattering data @xcite observes that the amount of csv in the @xmath9 state is @xmath10 , where the superscript @xmath11 indicates the ` nuclear ' effect obtained after the electromagnetic ( em ) corrections
. other convincing evidence of csv @xmath2 interaction comes from the binding energy difference of mirror nuclei which is known as okamoto - nolen -schifer ( ons ) anomaly @xcite .
the modern manifestation of csv includes difference of neutron - proton form factors , hadronic correction to @xmath12 @xcite and the observation of the decay of @xmath13 etc @xcite . in the present work we focus on the hadronic sector , and , in particular
, we attempt to construct csv potential for the @xmath2 interaction in one boson exchange ( obe ) model by invoking momentum dependent @xmath0-@xmath4 mixing .
the fact that the neutron and proton masses are not degenerate , the various isospin pure resonant states like @xmath0-@xmath4 or @xmath14-@xmath15 can , in reality , mix without violating any conservation principles dictated by other symmetries . in particular , the @xmath0-@xmath4 mixing seems to be a viable mechanism for the generation of significant amount of csv @xcite .
the earlier construction of csv potential involved on - shell mixing of the @xmath0 and @xmath4 meson states @xcite . a whole class of phenomenon including the difference of @xmath8-@xmath7 scattering length , binding energy difference of @xmath16-@xmath17 , or ons anomoly in general
could be successfully explained via @xmath0-@xmath4 mixing @xcite .
however , in ref.@xcite such a success was severely criticised on ground that the on - shell mixing amplitude differs quite significantly as one extrapolates the results from the @xmath3 ( or @xmath4 ) pole to the space - like region which is relevant for the construction of the csv potential .
goldman , henderson and thomas @xcite showed the strong @xmath18 dependence of @xmath0-@xmath4 mixing to csv potential using a simple quark model .
similar results were reported in ref.@xcite . in ref .
@xcite it was shown that such momentum dependencies suppress the contribution of @xmath0-@xmath4 mixing and also changes the sign of the mixing amplitude as one moves away from the @xmath3 and @xmath4 poles .
on the other hand miller @xcite and coon _ et al _ @xcite have advanced counter arguments that would restore the traditional role of @xmath0-@xmath4 mixing .
the issue is still unresolved .
informative summaries of the controversial point of views can be found in refs .
subsequently , several other calculations were also performed including qcd sum - rule @xcite with varied conclusions .
recently machleidt and mther @xcite have discussed various csv mechanism to estimate the @xmath9 scattering length .
therefore , the issues , including @xmath0-@xmath4 mixing as the origin of the csv force , seem to be quite open which provide part of the motivation of the present work .
here we revisit the problem of @xmath0-@xmath4 mixing and invoke the mechanism adopted in @xcite _
i. e. _ the mixing is driven by the neutron - proton mass difference .
although the driving mechanism is same , the main difference of our work with that presented in ref.@xcite resides in the treatment of the external legs . this is another source of csv due to the non - degenerate nucleon mass .
this , as we shall see , has serious consequence which even modify the central part of the csv potential .
we highlight the importance of @xmath19 tensor interaction and show how this can counter balance the weakening of the strength of csv interaction even when one extrapolates @xcite the results from on - shell to off - shell .
the paper is organised as follows .
in section ii we present the formalism where three momentum dependent @xmath0-@xmath4 mixing amplitude is used for the construction of csv potential .
the numerical results including the contributions of external legs and the @xmath19 tensor coupling to csv potential are discussed in section iii .
finally , we summarize in section iv .
to calculate the @xmath0-@xmath4 mixing amplitude we use the following vector meson - nucleon interactions : _ nn & = & g_|_^_ [ olag ] + _ nn & = & g_| ^ _ [ rlag ] where @xmath20 is the ratio of vector to tensor couplings . @xmath21 and @xmath22 denote nucleon and meson fields respectively .
the tensor coupling of @xmath1 is not included in the present calculation because it is negligible in comparison to the vector coupling .
all the parameters used in the present calculation are taken from those given by the bonn group @xcite . )
-isoscalar ( @xmath4 ) mesons that contributes to the csv @xmath2 interaction .
[ fig1 ] ] we now proceed to calculate the csv potential using the lagrangian described above .
the corresponding feynman diagrams are shown in fig .
[ fig1 ] . here
the csv is represented by the crossed blobs @xmath23 .
this apart , the external legs , depending upon whether we have proton or neutron , serve as another source of csv as mentioned in the introduction due to their non - degenerate mass .
following fig.[fig1 ] we write the matrix element as follows : ^nn_(q ) & = & [ |u_n(p_3)_^u_n(p_1 ) ] _ ^(q ) + & & ^_(q^2 ) _ ^(q ) [ |u_n(p_4)_^ u_n(p_2 ) ] .
[ ma0 ] the momentum space @xmath2 potential @xmath24 can be obtained by taking the limit @xmath25 of @xmath26 . here
@xmath27 , @xmath28 $ ] denote the vertex factors and @xmath29 is the mixing amplitude ( _ i.e. _ self - energy ) driven by the difference between proton and neutron loops ( _ see _ fig.[fig2 ] ) : ^_(q^2 ) & = & ^(p)_(q^2 ) -^(n)_(q^2 ) .
[ se ] the origin of the relative sign in the above equation is due to the different signs involved in the coupling of @xmath1 and @xmath0 with @xmath30 and @xmath31 ( _ see _ eqs .
( [ olag])-([rlag ] ) ) .
it is to be noted that the @xmath19 vertex factor will have a relative sign depending upon whether it couples to @xmath30 or @xmath31.this sign flip has been included .
-@xmath30 mass difference .
[ fig2 ] ] the polarization tensor of @xmath0-@xmath1 mixing due to @xmath32 excitations is calculated using standard feynman rules : i^(n)_(q^2 ) & = & tr . [ pol0 ] here @xmath33 is the usual feynman propagator given by g_n(k ) & = & .
[ np ] after performing the trace of eq.([pol0 ] ) , one may write the polarization tensor as ^(n)_(q^2 ) & = & q^ [ pol1 ] where @xmath34 . the current conservation yields @xmath35 as @xmath36 . from dimensional
counting , it is clear that the integral in eq.([pol0 ] ) is ultraviolet divergent .
we use dimensional regularization @xcite to isolate the divergent parts of the integral in eq.([pol0 ] ) obtaining , ^vv(n)_(q^2 ) & = & -q^2 + [ vv0 ] + ^tv(n)_(q^2 ) & = & - q^2 [ tv0 ] , where @xmath37 is an arbitrary renormalization constant ; @xmath38 is the euler - mascheroni constant .
@xmath39 contains the singularity ; @xmath40 as @xmath41
. since the mixing amplitude is the difference between proton and neutron loops contribution the divergent parts of the above expressions cancel out yielding , ^vv_(q^2 ) & = & ^vv(p)_(q^2 ) -
^vv(n)_(q^2 ) = ^1_0 dx ( 1-x)x ( ) q^2 , + [ vv1 ] + ^tv_(q^2 ) & = & ^tv(p)_(q^2)- ^tv(n)_(q^2 ) = ^1_0 dx ( ) q^2 , [ tv1 ] the full mixing amplitude thus becomes , ( q^2 ) & = & ^vv_(*q * ) + ^tv_(*q * ) = q^2 ^ 1_0 ( ( 1-x)x + ) ( ) dx .
[ ma ] eq.([ma ] ) displays the four - momentum dependence of the @xmath0-@xmath1 mixing amplitude in terms of three parameters @xmath42 , @xmath43 and @xmath44 .
we obtain @xmath45 mev@xmath46 and @xmath47 mev@xmath46 .
these are within the limit of experimentally extracted values ( @xmath48 mev@xmath46 ) @xcite .
upto now our results are same as that of ref.@xcite .
note that most of the earlier efforts to understand the role of @xmath0-@xmath1 mixing in csv potential was based on the assumption of constant on - shell value for the mixing amplitude @xcite . to calculate the csv potential we have to use the mixing amplitude in spacelike region @xmath49 . as a result
the mixing amplitude becomes @xmath50 dependent _ i.e. _ @xmath51 , where we find _ (*q * ) - ( 2 + 3c_)(m_p / m_n)*q*^2 -^2 .
[ mix ] to calculate the csv potential we take the non - relativistic ( nr ) limits of eq.([ma0 ] ) .
the relativistic energy @xmath52 is expanded in powers of @xmath18 and @xmath53 keeping the lowest order in @xmath54 _ i.e. _ @xmath55 . here
, @xmath56 is the average three momentum of the nucleon .
the three momentum transfer is denoted by @xmath57 ( _ see _ fig.[fig1 ] )
. also taking the nr limit of dirac spinor and keeping terms @xmath58 and @xmath59 we obtain u_n(_1 ) ( 1 - - ) ( c 1 + + ) , [ sp ] where @xmath60 is the nucleon spin . the relevant expressions which will be needed to construct the momentum space potential are the following : & & [ ex1 ] + i ( ) , [ ex3 ] + where @xmath61 . a straight forward calculation using eqs .
( [ ex1])-([ex4 ] ) and eq.([ma0 ] ) together with the nr meson propagators leads to the momentum space csv potential due to @xmath0-@xmath4 mixing : v^nn_(*q * ) & = & - + & & .
[ mpt ] here @xmath62 and @xmath63 is the total spin of the interacting nucleon pair .
we define @xmath64 , @xmath65 and @xmath66 .
the spin dependent parts of the momentum space potential as found in eq.([mpt ] ) appear because of the contribution of the external legs shown in fig.[fig1 ] . on the other hand ,
@xmath67 and @xmath68 arise due to expansion of the relativistic energy @xmath52 .
note that the potential derived in eq.([mpt ] ) contains both class @xmath69 and class @xmath70 potentials , and both of these potentials break the charge symmetry of @xmath2 interactions .
the first part of this potential represents class @xmath69 @xmath2 interaction which differentiates between @xmath8 and @xmath7 systems but vanishes for @xmath71 system . on the other hand , last part of eq.([mpt ] ) is class @xmath70 @xmath2 interaction which exists for @xmath71 system only . in the present paper ,
we focus on the class @xmath69 @xmath2 potential . from eq.([mpt ] ) we extract a piece which , in coordinate space , gives rise to @xmath72-function potential .
in momentum space it is given by v^nn_ & = & g_g_t^+_3 .
+ [ dfp ] the problem of contact term can be avoided by using form factors , for which @xmath73 is repleced with @xmath74 .
g_i g_i(*q*^2)=g_i ( ) [ ff ] the cut - off parameters @xmath75 governs the range of the suppression , which can be directly related to the hadron size .
the values of @xmath75s are determined from the fit of the two - nucleon empirical data @xcite .
the spin independent central part neglecting the contributions due to external legs and the @xmath19 tensor coupling , reduces to v^0_(*q * ) = - , [ vc ] which is same as obtained in ref.@xcite . in coordinate space , treating the on - shell mixing amplitude to be constant one obtains v^0_(*r * ) = - , + [ vcr ] where @xmath76 and @xmath77 . with form factors eq.([vcr ] ) reduces to v^0_(*r * ) & = & - , [ vcrf ] where @xmath78 .
eq.([vcrf ] ) represents the csv potential with constant mixing amplitude neglecting the contribution of external legs .
it is to be noted that in the linit @xmath79 , eq.([vcrf ] ) reduces to eq.([vcr ] ) .
if we include the contribution of the external legs and @xmath19 tensor coupling , simplifies to the central part , v^0nn_(*r * ) & = & - t^+_3 . [ vcnn ] in the above equation the first term in the bracket is same as one would have obtained from eq.([vc ] ) by taking the momentum dependent mixing amplitude as in eq.([mix ] ) , while the second term contains the contribution coming from the dirac spinors of the external lines .
the latter , clearly involves @xmath19 vector and tensor interactions , and , as we shall see , the term containing the tensor coupling ( @xmath80 ) is significantly larger compared to the vector interaction at distances below @xmath81 or so .
we leave out the coordinate space contact terms from eq.([vcnn ] ) and eq.([vrw ] ) .
we also drop the term @xmath67 from eq.([mpt ] ) while deriving the total coordinate space potential as it is not important in the present context .
however , it should be noted that to fit the @xmath9 and @xmath82 phase shifts simultaneously this term is necessary as @xmath53 gives the operator @xmath83 in coordinate space .
moreover , we use the @xmath18 dependent mixing amplitude instead of constant on - shell value and for this we consider terms upto @xmath59 .
taking all this into consideration we obtain , after some algebraic manipulations , the coordinate space csv potential as v^nn_(*r * ) & = & - t^+_3 .
[ vrw ] the spin - spin , tensor and spin - orbit interaction terms are explicitly containted in @xmath84 and @xmath85 which are as follows : v_vv(x ) & = & y_0(x ) + y_0(x)(**_1_2 ) + & - & y_1(x)s_12 ( ) - y_2(x)*l * [ vvv ] + v_tv(x ) & = & y_0(x ) + y_0(x)(**_1_2 ) + & - & y_1(x)s_12 ( ) - 2y_2(x)*l * [ vtv ] .
where , y_1(x ) & = & ( 1 + + ) y_0(x ) + y_2(x ) & = & ( + ) y_0(x ) + s_12 ( ) & = & 3(_1)(_2 ) - ( _ 1_2 ) the first part of eq.([vrw ] ) represents the central part without contributions from external legs .
in addition , the last two terms of eq.([vrw ] ) are the contributions coming from the external nucleon legs as discussed earlier .
it is also to be noted that the central part also receives contributions due to the presence of the first terms in eq.([vvv ] ) and eq.([vtv ] ) .
the tensor contribution @xmath86 of @xmath3-meson is contained in the third term of eq.([vrw ] ) .
eq.([vrw ] ) does not include form factors .
it diverges near the core .
this divergence can be removed by incorporating form factors as in eq.([ff ] ) .
thus the complete csv potential with form factors reduces to v^nn_(*r * ) & = & - .
[ vrwf ] note that the above equation contains the contribution of eq.([dfp ] ) .
the csv @xmath2 potential given in eq.([vrwf ] ) can be used to calculate the difference between @xmath8 and @xmath7 scattering lengths at @xmath9 state .
the difference between scattering lengths , @xmath87 , and the difference between csv @xmath8 and @xmath7 potential , @xmath88 , are related by a = -a^2m^_0 v_
u^2_0(r ) dr where @xmath89 and @xmath90 is the zero energy wave function , normalized to approach @xmath91 as @xmath92 and @xmath93 . to calculate @xmath94 we use the following zero energy wave function @xcite : u_0(r)&=&- , + & & [ zewf ] where @xmath95 , @xmath96 and @xmath97 is the effective range . in the present calculation
we take @xmath98 fm .
in this section we present our results .
first we show the momentum space central potential ( _ see _ eq.[vc ] ) in fig.[fig3 ] considering the three momentum dependent mixing amplitude . in this figure dotted curve represents the central potential with the form factor in the space - like region . in contrast , the solid curve represents the same without the form factor @xcite . in fig.[fig4 ] the importance of the relativistic correction to the central part potential in momentum space is displayed .
this correction , as expected , is marginal at low momentum ( below @xmath99 mev ) transfer . in the short distance regime _
i.e. _ near the core region , the relativistic correction becomes significant which is clearly seen in fig.[fig4 ] . in fig.[fig5
] we show the central part of the potential due to both on - shell and off - shell mixing amplitudes .
it is seen that the contribution of the off -shell @xmath0-@xmath1 mixing amplitude to the @xmath2 potential is opposite in sign relative to the contribution obtained from using the on - shell value .
this , again , is consistent with the observation made in ref.@xcite .
the individual contribution of different parts of the central potential given in eq.([vcnn ] ) is presented in fig.[fig : vcnn ] .
clearly the contribution of @xmath19 tensor coupling to the csv potential is found to be much larger than the contribution of the first part ( _ i.e. _ the central part without external legs and @xmath19 tensor contribution ) .
it is to be noted that @xmath19 tensor contribution is present only when the external legs are taken into account . the @xmath9 state csv potential for @xmath7 system due to @xmath0-@xmath1 mixing is shown in fig.[fig6 ] .
the importance of the central part with relativistic correction ( dashed curve ) and tensor contribution ( dashed- dotted curve ) are clearly revealed .
the magnitude of the contribution of tensor coupling is comparable with that of the central part with relativistic correction in the core region . on the other hand , magnitude of the contribution of tensor coupling
is found to be much larger than the contribution of the central part ( dotted curve ) without relativistic correction in the core region . in the dynamical region , it is seen that all the contributions are comparable . the solid curve in this figure represents the total contribution together with the relativistic correction . in fig.[fig7 ]
we present the csv potential at @xmath9 state both for the @xmath7 and @xmath8 system .
the solid and dashed curves , respectively , show the csv potential for @xmath7 and @xmath8 system taking the contribution of tensor coupling of @xmath3-meson .
the same are presented by dotted and dot - dashed curves without considering the tensor coupling .
the @xmath9 csv potential with form factors is displayed in fig.[fig8 ] .
it is seen that the inclution of @xmath100 modifies the csv potential dramatically .
it is to be noted that , with its inclusion , the csv potential changes its sign .
the difference in scattering length @xmath94 when calculated with the potential of eq.([vrwf ] ) is also markedly different from that calculated ignoring the term @xmath100 .
the results of @xmath94 with and without @xmath100 are shown in table [ dsc ] .
.the difference between @xmath7 and @xmath8 scattering lengths at @xmath9 . [ cols="<,^,^",options="header " , ]
in the present work we have constructed the csv potential within the framework of obe model and studied the role of three momentum dependence @xmath0-@xmath1 mixing amplitude in csv .
we find that the inclusion of the contributions coming from the external legs are important because of the strength of the @xmath19 tensor interactions .
it is seen that unlike the previous finding @xcite where the charge symmetry violation at the external legs were ignored , the strength of the csv interaction could be significantly larger even when the off - shell amplitude for the @xmath0 -@xmath4 mixing is considered .
it is important to note that contribution from the spinors also modifies the central part of the two body potential as shown in eqs.([vcnn ] ) and ( [ vrw ] ) .
furthermore , we present results both for the central and non - central part of the csv potential .
we also have calculated difference of @xmath9 scattering lengths between @xmath7 and @xmath8 systems and explicitly show the contribution of @xmath100 .
it is to be noted that @xmath94 changes sign with the inclusion of the fourier transform of @xmath100 .
it would be interesting to apply the potential presented here to calculate various other csv observables to delineate the role of tensor interaction further . | we construct the charge symmetry violating ( csv ) nucleon - nucleon potential induced by the @xmath0-@xmath1 mixing due to the neutron - proton mass difference driven by the @xmath2 loop . analytical expression for for the two - body csv potential
is presented containing both the central and non- central @xmath2 interaction .
we show that the @xmath3@xmath2 tensor interaction can significantly enhance the charge symmetry violating @xmath2 interaction even if momentum dependent off - shell @xmath0-@xmath4 mixing amplitude is considered .
it is also shown that the inclusion of form factors removes the divergence arising out of the contact interaction .
consequently , we see that the precise size of the computed scattering length difference depends on how the short range aspects of the csv potential are treated . |
we thank m. a. idiart for fruitful discussions and suggestions . this work has been partially supported by brazilian agencies cnpq , finep and fapergs . | the thermal growth of silicon oxide films on si in dry o@xmath0 is modelled as a dynamical system , assuming that it is basically a diffusion - reaction phenomenon .
relevant findings of the last decade are incorporated , as structure and composition of the oxide / si interface and o@xmath0 transport and reaction at initial stages of growth .
the present model departs from the well established deal and grove framework ( deal , b.e . and grove , a. s. general relationship for the thermal oxidation of silicon _ j. appl
. phys . _ * 36 * , 3770 - 3778 ( 1965 ) ) indicating that its basic assumptions , steady - state regime and reaction between o@xmath0 and si at a sharp oxide / si interface are only attained asymptotically .
experimental growth kinetics by various authors , obtained for a wide range of growth parameters are shown to collapse into one single curve when the scaling properties of this model equations are explored .
2 silicon oxide films thermally grown on single - crystalline si ( c - si ) wafers in dry o@xmath0 are still the most common materials used as gate dielectrics in si - based metal - oxide - semiconductor ( mos ) structures . as device dimensions shrink below 0.25 @xmath1 m , the oxide films are forced to scale down to thicknesses of 5 nm and less .
reliability of highly integrated si devices is critically dependent on the characteristics of the vitreous oxide film , like thickness uniformity , defect density , dielectric strength , and others , as well as on those of the oxide / si interface , like roughness , electronic states density and others @xcite .
the dielectric performance of ultrathin films of silicon oxide thermally grown on si and the structural and electronic properties of the oxide / si interface have been intensively studied experimentally , theoretically , and computationally @xcite in synergism with the development of the semiconductor industry .
the understanding of film growth kinetics , however , did not make any significant progress since the early model of deal and grove @xcite , which subsisted as the only fundamental approach .
the deal and grove model , and more specifically the resulting linear - parabolic law , agrees with the observed growth kinetics in dry o@xmath0 only above a certain thickness @xcite .
this is an expected fact from the model itself , since its basic assumption is that growth is promoted by interstitial diffusion of the oxidant species , the o@xmath0 molecule , through a previously grown oxide layer , thick enough to guarantee steady state regime .
besides , reaction between o@xmath0 and si is assumed to take place solely at an abrupt oxide / si interface , producing stoichiometric sio@xmath0 .
this initial oxide thickness was estimated @xcite from experimental data as being between 20 and 30 nm , well above the thickness range of interest for present and future use in microelectronics si - devices .
description of the growth kinetics in the lower thickness range was addressed by many authors , within the deal and grove model framework , adding new terms to the linear - parabolic expression in order to fit experimental data @xcite .
although fitting the experimental curves , and so providing useful analytical expressions capable of reproducing the whole thickness interval , the extra terms added to the deal and grove expression did not have a well - defined physical meaning , even though their dependence on some processing parameters , like temperature for instance , could be explored @xcite .
none of these models has been shown to be clearly correct and none of them has gained widespread acceptance .
one is left then with several empirical expressions which can be used to model the growth kinetics in dry o@xmath0 in the thin and ultrathin film regimes below 20 nm , as well as in further stages of growth , but with no satisfactory physical explanation @xcite . according to b. deal @xcite `` such an expression may well fit the data , but provides little guidance or insight into mechanisms involved or effects of process variations '' .
recent investigations showed that as film thickness decreases below 20 nm , the contribution to film growth due to reaction away from the interface region becomes increasingly significant @xcite .
furthermore , there are strong theoretical and experimental evidences of the existence of a reactive layer formed by sub - oxides ( also called si - excess ) near the oxide / si interface @xcite . in the present work the thermal growth of silicon oxide films on si
is modelled as a dynamical system , assuming that it is basically a diffusion - reaction phenomenon .
since steady - state regime is not imposed , an initial oxide thickness is not required .
therefore , the model is expected to describe the whole oxide thickness interval .
diffusion - reaction equations have been used to describe different systems @xcite but , to our knowledge , this is the first time they are applied to silicon oxide growth kinetics . the diffusing species is taken to be o@xmath0 , to model what has been largely demonstrated by isotopic substitution experiments @xcite .
growth is promoted by reaction of o@xmath0 with si , not necessarily at the oxide / si interface , but wherever the two species meet .
the growth kinetics can then be obtained at any temperature by specifying the diffusivity of the o@xmath0 molecule in the silica network , @xmath2 , the reaction rate between o@xmath0 and si , @xmath3 , and the o@xmath0 pressure in the gas phase .
growth in one dimension is considered , which is the most common experimental situation , meaning that as a face of c - si is exposed to o@xmath0 , the silica film grows in the direction perpendicular to this face .
the proposed description for the growth kinetics is contained in the following coupled partial differential equations : @xmath4 where @xmath5 is the concentration of the i - species ( @xmath6 , @xmath7 ) in the solid phase , in units of number of atoms per unit volume , @xmath8 , normalised by the concentration of si in c - si , @xmath9 .
thus , the state of any species at any time is characterised by a density function @xmath10 , where @xmath11 is the coordinate in the direction of growth , @xmath12 indicating the surface of the sample , and @xmath13 is the elapsed growth time .
the first equation describes the o@xmath0 concentration rate and it is essentially a balance equation . at a given position the concentration of oxygen may vary due to the net flux of o@xmath0 through fully ( sio@xmath0 in oxide bulk ) or partially ( sio@xmath14 , @xmath15 , near the interface @xcite ) oxidised si , modelled by the diffusion term with a diffusion constant @xmath2 , or due to reaction to produce silicon oxide , characterised by a reaction rate constant @xmath3 .
the second equation describes the si concentration rate , which is only due to reaction with o@xmath0 , implying that si is an immobile species @xcite .
the silicon oxide concentration , @xmath16 , does not explicitly appear in these equations , but it is in direct connection with @xmath17 through the expression @xmath18 reflecting conservation of si species in the si + o@xmath19 sio@xmath0 reaction .
it is assumed that the o@xmath0 diffusivity at a certain temperature does not change during oxidation , and therefore @xmath2 is constant throughout the whole growth process .
furthermore , based on the fact that oxygen diffuses mainly through oxide , one neglects volume changes due to the above chemical reaction . in order to complete the description in terms of equations ( [ edif ] ) , it is necessary to provide the initial and boundary conditions .
they are @xmath20 where @xmath21 is the o@xmath0 concentration in the gas phase and @xmath22 is the ratio between the accessible free volume for o@xmath0 in the silica network and the unit volume of the solid @xcite .
initially there is a pure , c - si substrate from @xmath12 to @xmath23 , whose surface is exposed to a gaseous medium , o@xmath0 , with a relative concentration at this surface represented in the second condition by @xmath24 .
reaction of these two species produces silicon oxide , not necessarily stoichiometric , as the solid - phase reaction si + o@xmath19 sio@xmath0 may occur in several steps .
as time proceeds , o@xmath0 diffuses through the oxide network , reaches si , reacts producing more oxide , and consequently pushes the oxide / si interface further in the positive @xmath11 region .
different values of @xmath2 , @xmath3 and @xmath25 imply different kinetics .
however , the values of @xmath2 and @xmath3 may be used to define `` natural '' units for the system , resulting in a set of adimensional equations with only one parameter , namely @xmath25 @xcite . under the following transformations : @xmath26 the resulting adimensional set of differential equations and initial and boundary conditions are @xmath27 leaving @xmath25 as the only parameter .
this is a powerful statement .
it predicts that in all experiments performed at the same gas pressure , oxide growth kinetics and o@xmath0 and si concentrations versus depth curves ( profiles ) at different temperatures collapse to the same curves , provided that appropriate units are used : @xmath28 for time and @xmath29 for thickness . in order to obtain profiles and growth kinetics , eqs.([eadi ] ) must be solved .
those are partial , nonlinear , coupled differential equations and can not be analytically solved at all times . on the other hand , they can be easily solved by finite differences methods , iterating for as long as it takes to reproduce the growth time of interest .
the calculation gives then the profiles @xmath30 of all species and the kinetics .
typical profiles at different growth times for a given value of @xmath25 are shown in figure 1 .
the o@xmath0 profiles are not straight lines as they are assumed to be in the steady - state regime @xcite . strictly speaking ,
steady - state is never attained as long as the oxide / si interface moves deeper into the c - si substrate .
furthermore , this interface is not sharp , in accordance with a wealth of experimental evidences @xcite .
still in strict accordance with theoretical predictions @xcite and experimental observations @xcite , a finite width region formed by si in various lower oxidation states ( + 1 , + 2 , + 3 , sub - oxides or si - excess ) early develops between the stoichiometric sio@xmath0 film and the c - si substrate . in this way
, o@xmath0-si reaction can take place in the whole region where non - fully oxidised si is available .
profiles in the solid phase at various temperatures and one pressure , for different oxidation times .
the solid , dashed , and dot - dashed lines represent , respectively , the profiles at increasing oxidation times . ]
fig .
1 shows that the two hypotheses of the deal and grove model are not valid : i ) steady - state regime is never attained ;
ii ) reaction between o@xmath0 and si leading to growth does not take place at a sharp oxide / si interface . for high enough temperatures and pressures , and long enough times
, the o@xmath0 profile approaches a straight line , and the thickness of the near - interface , sub - oxide region is much smaller than the oxide thickness .
this is the observed situation at oxide thicknesses well above 20 nm . in this limit , steady state and abrupt interface
are plausible approximations . in other words ,
asymptotically the deal and grove assumptions are valid , whereas at initial stages they are not .
the kinetics in the initial growth regime , which was described as `` anomalous '' in the deal and grove framework , naturally emerges from the present model as follows .
let us consider the quantity that is mainly focused in practical situations : the oxide thickness @xmath31 .
we calculate it through the relation : @xmath32 the oxide growth kinetics @xmath33 is completely determined by @xmath34 , whose evolution is ruled by eqs.([eadi ] ) and therefore is determined by a unique parameter @xmath25 , the concentration of o@xmath0 in the gas phase .
however , due to the particular form of the differential equations , together with boundary and initial conditions , the system presents a further scaling property .
suppose two different growth kinetics , with o@xmath0 concentrations in the gas phase given by @xmath35 and @xmath36 , then the following relations apply @xmath37 therefore , besides collapsing for different temperatures ( i.e. different @xmath2 and @xmath3 ) when natural units of the system are used , kinetic curves for different @xmath25 collapse to one single curve due to the symmetry present in the form of the diffusion - reaction model equations together with initial and boundary conditions . in summary ,
the theoretical prediction is that all kinetic curves reduce to a single one .
these predictions were tested with kinetic curves taken from references @xcite , corresponding to oxidations in joule - effect heated and rapid thermal processing furnaces in a variety of growth parameters ( temperature , o@xmath0 pressure , and time ) leading to oxide thicknesses ranging from 1 to 100 nm . in fig .
2 we show the collapse of the experimental kinetics ( symbols , colours ) into the theoretical curve ( line , black ) . for the kinetics curves from reference @xcite , with five points in each curve
, we first used the scaling predicted by eqs.([scal ] ) , obtaining four different curves , one for each temperature .
we then scaled these curves using natural units for @xmath11 and @xmath13 according to eqs.([natu ] ) until they overlapped the theoretical curve .
the curves from reference @xcite contained much more experimental points , extending from 0.1 to 1000 min , and from 1 to 80 nm . for these curves
the pressure scaling predicted by eqs.([scal ] ) is not perfect at long growth times , probably due to an increase in diffusivity .
they were then collapsed one by one into the theoretical curve using eqs.([natu ] ) first , and then eqs.([scal ] ) .
the same procedure was used to collapse the kinetics from references @xcite into the theoretical curve .
the asymptotic depart from the theoretical curve is an expected result , since the present model assumes a constant diffusivity throughout the whole growth process , despite many evidences that this is not the case : i ) transport of the diffusing species through the o - excess region near the surface @xcite , through the si - excess region near the oxide / si interface @xcite , and through the stoichiometric sio@xmath0 in the bulk of the growing oxide film have different diffusivities ; ii ) the growth of the defective , near - surface and near - interface regions saturate within a few nm @xcite , while the bulk , stoichiometric oxide grows continuously .
so , interstitial diffusion of o@xmath0 through the bulk of the growing oxide gives a relative contribution that increases as the width of this region becomes dominant ; iii ) oxidations performed in dry o@xmath0 flow during long time intervals ( hours ) may suffer from water vapor contamination , which accelerates the oxide growth @xcite .
taking the dimensional oxide thickness @xmath38 , and defining @xmath39 , the asymptotic solution to eqs.([eadi ] ) predicts the inverse growth rate @xmath40 , with @xmath41 constant , which is very similar to the asymptotic approximation of the deal and grove model @xcite @xmath42 . in fig .
3 the inverse growth rate is plotted as a function of oxide film thickness for different experimental kinetics @xcite . and theoretical inverse growth rate ( inset ) . ]
a deviation from a straight line in the thick film range is apparent , corresponding to a variable , increasing diffusivity , contrary to the theoretical prediction shown in the inset .
thus , the thermal growth of silicon oxide films on c - si in dry o@xmath0 is promoted by diffusion of o@xmath0 through the growing oxide with variable diffusivity .
similar physical arguments could support a variable diffusion rate .
nevertheless , even using constant , effective values for the diffusivity and reaction rate , modelling the growth as a diffusion - reaction process where conditions such as initial thickness , steady - state regime , and reaction solely at an abrupt interface are abandoned allows to reproduce the observed kinetics , especially in the lower thickness ( ultrathin film ) regime which is the one of present and future practical interest .
realistic model evaluations of variable diffusivity and reaction rate stay beyond the scope of the present work , although this seems to be the only route to a quantitative description of the complete growth kinetics .
the present approach opens the possibility of incorporating new and determinant facts on o@xmath0 transport at initial stages of thermal oxide growth and on structure of the oxide / si interface , that were discovered after the proposal of deal and grove @xcite and not yet integrated into growth kinetics models .
some were discussed here , like the graded , si - excess nature of the oxide / si interface and the incorporation of freshly arriving o@xmath0 into different regions of the growing oxide , while several others may also be included in a diffusion - reaction approach , such as the role played by oxygen excess centers @xcite in the near - surface region on oxygen transport which has been recently investigated @xcite . 99 buchanan , d. a. and lo , s. h. growth , in _ the physics and chemistry of sio@xmath0 and the si - sio@xmath0 interface-3 _ vol 96 - 1 ( eds .
massoud , h. z. , poindexter , e.h . , and helms , c.r . ) 3 - 14 ( the electrochemical society , pennington , 1996 ) .
feldman , l.c . ,
gusev , e.p . , and garfunkel , e. in _ fundamental aspects of ultrathin dielectrics on si - based devices _
( eds .
garfunkel , e. , gusev , e.p , and vul , a. ) 1 - 24 ( kluwer academic publishers , dordrecht , 1998 ) .
schuegraf , k. f. and hu , c. reliability of thin sio@xmath0 _ semicond .
sci .
technol . _ * 9 * , 989 - 1006 ( 1994 ) .
pasquarello , a. , hybertsen , m.s . , and car , r. interface structure between silicon and its oxide by first principles molecular dynamics _ nature _ * 396 * , 58 - 60 ( 1998 ) deal , b.e . and grove , a. s. general relationship for the thermal oxidation of silicon _ j. appl
. phys . _ * 36 * , 3770 - 3778 ( 1965 ) .
massoud , h. z. , plummer , j. d. , and irene , e. a. thermal oxidation in dry oxygen growth - rate enhancement in the thin regime _ j. electrochem
. soc . _ * 132 * , 2685 - 2693 ( 1985 ) .
han , c. -j . and helms , c.r .
parallel oxidation mechanism for si oxidation in dry o@xmath0 _ j. electrochem .
soc . _ * 134 * , 1297 - 1302 ( 1987 ) .
rigo , s. in _ instabilities in silicon devices _ vol .
1 ( eds . g. barbottin and a. vapaille ) 5 - 100 ( elsevier publishers , amsterdam , 1986 ) ( 1986 ) .
plummer , j. d. in _ the physics and chemistry of sio@xmath0 and the si - sio@xmath0 interface-3 _ vol 96 - 1 ( eds .
massoud , h. z. , poindexter , e.h . , and helms , c.r . )
129 - 142 ( the electrochemical society , pennington , 1996 ) .
deal , b.e . in _ the physics and chemistry of sio@xmath0 and the si - sio@xmath0 interface
_ , ( eds . helms , c. r. and deal , b.e . ) 5 - 16 ( plenum press , new york , 1988 ) .
han , c .- j . and helms , c.r .
@xmath43o tracer study of si oxidation in dry o@xmath0 using sims _ j. electrochem .
soc . _ * 35 * , 1824 - 1832 ( 1988 ) .
trimaille , i. , stedile , f.c . ,
ganem , j .-
j . , baumvol , i.j.r . , and rigo , s. in _ the physics and chemistry of sio@xmath0 and the si - sio@xmath0 interface-3 _ vol 96 - 1 ( eds .
massoud , h. z. , poindexter , e.h . , and helms , c.r . )
59 - 71 ( the electrochemical society , pennington , 1996 ) .
pennington ( 1996 ) .
rochet , f. , rigo , s. , froment , m. , danterroches , c. , maillot , c. , roulet , h. , and dufour , g. the thermal oxidation of silicon .
the special case of the growth of very thin films _ adv .
phys . _ * 35 * , 237 - 274 ( 1986 ) .
trimaille i. and rigo , s. use of @xmath43o isotopic labeling to study the thermal dry oxidation of silicon as a function of temperature and pressure _ appl .
surf .
sci . _
* 39 * , 65 - 80 ( 1989 ) .
lu , h.c . ,
t. gustafsson , t. , gusev , e.p .
, and garfunkel , e. an isotopic labeling study of the growth of thin oxide films on si(100 ) _ appl .
phys .
lett . _ * 67 * , 1742 - 1744 ( 1995 ) .
hamann , d.r .
diffusion of atomic oxygen in sio@xmath0 _ phys .
rev .
lett . _ * 81 * , 3447 - 3450 ( 1998 ) .
mott , n.f . ,
rigo , s. , rochet , f. , and stoneham , a.m. oxidation of silicon _ phil . mag .
_ * b60 * , 189 - 212 ( 1988 ) .
stoneham , a.m. , grovenor , c.r.m . , and cerezo , a. oxidation and the structure of the silicon oxide interface _ phil . mag . _ * b55 * , 201 - 210 ( 1986 ) .
himpsel , f.j .
, mcfeely , f.r .
, taleb - ibrahimi , a. , yarmoff , j.a . and hollinger , g. microscopic structure of the sio@xmath0/si interface _ phys . rev .
_ * b38 * , 6084 - 6096 ( 1988 ) .
jackman , t.e . , macdonald , j.r . ,
feldman , l.c . ,
silverman , p.j . and
stensgaard , i. , ( 100 ) and ( 110 ) si - sio@xmath0 interface studies by mev ion backscattering _ surf .
sci . _ * 100 * , 35 - 42 ( 1980 ) .
mattis , d.c . and
glasser , m.l .
the uses of quantum field theory in diffusion - limited reactions _ rev .
mod . phys._*70 * , 979 - 1001 ( 1998 ) .
larralde , h. , araujo , m. , havlin , s. , and stanley , h.e .
diffusion reaction kinetics for a + b(static ) @xmath44 c(inert ) for one - dimensional systems with initially separated reactants _ phys .
rev .
_ * a46 * , r6121-r6123 ( 1992 ) barkema , g.t . ,
howard , m.j . , and cardy , j.l .
reaction - diffusion front for a + b @xmath44 @xmath45 in one dimension _ phys .
rev . _
* e53 * , r2017-r2020 ( 1996 ) .
pretorius , r. , strydom , w. , mayer , j.w . , and
comrie c. @xmath46si tracer studies of the oxidation of si , cosi and ptsi _ phys .
rev . _ * b22 * , 1885 - 1891 ( 1980 ) .
see also baumvol , i.j.r . , krug , c. , stedile , f.c .
, gorris , f. , and schulte , w.h . on the immobility of si during thermal growth of ultrathin silicon oxide films on si(111 ) in o@xmath0 , submitted for publication in phys .
rev .
lett .
, january , 1999 .
gusev , e.p . ,
lu , h.c . ,
gustafsson , t. , and garfunkel , e. growth mechanisms of thin silicon oxide films on si(100 ) studied by medium - energy ion scattering _ phys .
rev . _ * b52 * , 1759 -1775 ( 1995 ) .
gusev , e.p . ,
lu , h.c . , gustafsson , t. , and garfunkel , e. the initial oxidation of silicon : new ion scattering results in the ultra - thin regime _ appl.surf .
sci . _ * 104/105 * , 329 - 334 ( 1996 ) .
stedile , f.c . ,
baumvol , i.j.r . ,
oppenheim , i.f . ,
trimaille , i. , ganem , j .-
j . , and rigo , s. thickness of the sio@xmath0/si interface and composition of silicon oxide thin films : effect of wafer cleaning procedures _ nucl .
instrum .
meth . _
* b118 * , 493 - 498 ( 1996 ) .
kamigaki , y. and itoh , y. thermal oxidation of silicon in various oxygen partial pressures _ j. appl
. phys . _ * 48 * , 2891 - 2896 ( 1977 ) .
ganem , j .- j . ,
rigo , s. , and trimaille , i. modellization of the silicon rapid thermal oxidation in the initial stages according to the silicon fragments model _ microelec .
eng . _
* 22 * 35 - 38 ( 1993 ) .
stesman , a. and scheerlink , f. generation aspects of the delocalized intrinsic ex defect in thermal sio@xmath0 _ j. appl
. phys . _ * 75 * , 1047 - 1058 ( 1994 ) . |
quasar absorption lines have developed into a key tool for probing the evolving population of galaxies from high redshift to the present day . since selecting galaxies based on their absorption , rather than emission properties
circumvents selection effects associated with malmquist bias , this is a potentially powerful technique for tracing the evolution of gas repositories over the bulk of the hubble time .
one of the most challenging aspects of quasar absorption line research is the association between different types of absorbers and their galactic counterparts ; without this crucial information we are lacking an infra - structure within which to interpret the results .
this crux is perhaps best highlighted by the work which has focussed on damped lyman alpha systems ( dlas ) .
dlas have been variously associated with galactic disks ( prochaska & wolfe 1997 , 1999 ) , low surface brightness galaxies ( jimenez et al .
1999 ) , dwarfs ( nulsen et al .
1998 ) , merging protogalactic clumps ( haehnelt et al . 1998 ) and galactic winds ( schaye 2001 ) .
a key diagnostic to the nature of high redshift dlas is the combined study of low redshift absorbers whose counterpart galaxies can be investigated in tandem . establishing an understanding of low redshift galaxy absorption properties through
the profiling of abundances , luminosities , morphologies and star formation histories is essential for extrapolation to high redshifts where only absorption line abundances are available .
unfortunately , the direct identification of dla galaxies has seen relatively slow progress .
some 20 dlas now have tentative counterparts ( e.g. , le brun et al .
1997 ; turnshek et al .
2001 ; nestor et al . 2001
; bowen , tripp & jenkins 2001 ) , although only approximately 25% of these have actually been confirmed spectroscopically .
clearly , increasing the number of spectroscopic confirmations is a priority , one which has been highlighted by the results of rosenberg & schneider ( 2002 ) who warn against simply assigning the identification to the nearest bright galaxy .
the characterisation of strong mg ii absorbers at @xmath9 has been significantly more successful than for classical dlas , with the identification of a galaxy candidate within 40 kpc in almost every case ( steidel , dickinson & persson 1994 ; steidel et al .
1997 ) , many of which have been spectrscopically confirmed .
modelling of the absorption line profiles of mg ii systems has been undertaken in a similar vein to dlas ( e.g. , lanzetta & bowen 1992 ; charlton & churchill 1998 ; bond et al .
2001b ) , including models of radial infall to explain the almost unit covering factor ( but see also charlton & churchill 1996 ) .
most recently , steidel et al . (
2002 ) have directly investigated the connection between absorption and emission kinematics in 5 mg ii galaxies and found that the absorption profiles seem to be consistent with rotation , although the kinematics can not be explained by a simple disk model . to our knowledge
, the kinematics of absorber - galaxy systems have been investigated in only 2 other studies ( barcons , lanzetta , & webb 1995 ; lanzetta et al .
although the former study also finds kinematics consistent with disk rotation , lanzetta et al ( 1997 ) intriguingly find the absorption gas moving counter to the rotation direction of the @xmath10 galaxy disk . here
we present the results of a multi - faceted investigation of the kinematics of a mg ii absorber at @xmath11 towards q1331 + 17 which combines high resolution qso absorption line data with slit spectroscopy and hst imaging of the galaxy counterpart .
this absorber not only represents an eighth system for which the connection between absorption and emission kinematics can be explored at intermediate redshift , but additionally allows us to investigate the presence of superbubbles recently suggested by bond et al .
( 2001a ) .
bond et al .
( 2001b ) have proposed that a subset of strong mg ii absorbers may be associated with superbubbles or winds ; shells of expanding gas scooped up in the ism of absorbing galaxies ( e.g. bond et al .
2001a ; rauch et al .
the @xmath12 mg ii system towards q1331 + 17 has been previously investigated by bond et al .
( 2001a ) as a particular case study for bubbles due to the striking pair - wise structure of absorption lines seen in keck hires spectra , first highlighted by churchill , vogt & steidel ( 1995 ) .
large - scale outflows associated with such winds have been observed in a variety of environments over a range of redshifts , such as lyman break galaxies ( pettini et al .
2001 , 2002 ) , high redshift lensed galaxies ( frye , broadhurst and benitez 2002 ) and local starburst galaxies ( heckman et al .
the existence of such superbubbles that burst out of the galaxy may be a significant source of metal enrichment in the intergalactic medium ( giroux & shull 1997 ; ellison et al .
2000 ; ferrara et al.2001 ) . the uves data presented here , although of only slightly superior resolution than the hires data used by bond et al .
( 2001a ) , reveal even more strikingly the pairwise symmetries of the absorption features , which we discuss in the context of superbubbles .
the paper is organized as follows . in
2 we describe multi - object spectroscopy of candidate galaxy counterparts , echelle spectroscopy of the qso itself and the re - analysis of archival wfpc2 data .
determination of the redshift of galaxy g5 and a compilation of absorption line profiles are presented in 3 . in 4 , we discuss the connection between absorption and emission kinematics and the implication for the superbubble scenario .
we adopt @xmath13 km / s / mpc , unless otherwise stated .
q1331 + 17 is a quasar well known to purveyors of absorption line systems , since it is a bright ( @xmath14 ) background target with a @xmath15 dla which has been widely studied in the literature ( e.g. , prochaska & wolfe 1999 ; pettini et al .
1997 ; lu et al .
in addition , there is a lower redshift ( @xmath16 ) mg ii system which has been provisionally identified with an edge - on @xmath17l@xmath18 disk at an impact parameter of 3.86 arcsec from the qso ( le brun et al .
1997 ; bond et al .
2001a ) based on hst imaging .
there is also a mg ii system at @xmath19 ( steidel & sargent 1992 ) and a c iv system at @xmath20 ( sargent , steidel & boksenberg 1988 ) .
finally , we have identified a new c iv system at @xmath21 in the uves data presented in 2.2 .
we have obtained ground - based spectroscopic observations of galaxies in the qso field in order to confirm the absorber identification for the @xmath16 mg ii system and study its internal kinematics .
in addition to the galaxy spectra , we have obtained high resolution spectra of the qso to study the absorber s metal lines and compare their kinematics with those of the associated galaxy .
a total of 292 minutes of fors2 service - mode spectroscopic observations were carried out on 18 may 2001 , 14 june 2001 , 16 june 2001 , and 14 july 2001 .
specifically , five 51-minute exposures and one 37-minute exposure were taken for the two slits adjacent to the qso .
grism 600ri+19 and 1@xmath22 slits were used , resulting in a dispersion of 1.30 / pix and average resolution of r@xmath171200 . for slits near the qso ,
the wavelength range covered was 5500 - 8150 .
figure [ fors_mask]a shows an outline of the two central slits on a rotated version of the hst wfpc2 archival image of the field . for consistency with earlier work , we have adopted the galaxy names used by le brun et al.(1997 ) and bond et al .
( 2001a ) .
the end of the top slit covered g4 , and the bottom slit covered g3 and g5 , the latter of which has been previously identified by le brun et al .
( 1997 ) as the preferred absorbing galaxy candidate .
g5 is located 3.86 arcsecs from the qso which translates to 28.3 @xmath23 kpc in our adopted cosmology .
note that the fine alignment of these two slits near the qso was done with the aid of the wfpc image , which was re - sampled and superimposed on the fors2 image used for mask design .
the data were reduced using iraf .
a bias frame was first subtracted from each spectroscopic exposure .
there were 12 pixel offsets in the object positions along the slit in different frames , as well as in the position of the slit on the ccd . to account for these shifts ,
the multi - slit spectra were registered so that the objects were positioned in the same pixel along the spatial direction for every frame .
subimages corresponding to individual slits were extracted and then reduced individually .
two - dimensional wavelength - calibrated spectra from each slit and exposure were created using iraf tasks @xmath24 , and @xmath25 . the wavelength calibration was first determined based on a henear comparison spectrum , and a second - order correction around the region of interest was found by comparing the wavelength of sky lines in the spectrum with theoretical values listed in osterbrock et al.(2000 ) available from the vizier database ( ochsenbein , bauer , & marcout 2000 ) . once the wavelength solution was determined , the sky lines were subtracted interactively from the raw spectra using the iraf task @xmath26 .
cosmic rays were tagged on the sky - subtracted spectra using a sigma - clipping algorithm , and replaced via linear interpolation of neighbouring pixels ( a necessary step before wavelength re - sampling ) .
the sky - subtracted and cosmic - ray free spectra were re - sampled into two - dimensional wavelength - calibrated spectra , using the previously - determined wavelength solution .
the individual wavelength - calibrated and sky - subtracted spectra were co - added using iraf task @xmath27 with cosmic - ray pixels masked to produce the final two - dimensional spectra .
no emission lines or continuum were detected in g3 or g4 , which is not surprising due to their proximity to the qso and faint magnitudes ( @xmath28 and 24.47 from our re - analysis of the hst data , see section 2.3 ) .
however , g5 shows emission lines at 6505 and 5674
( see figure [ em_mos ] ) .
a faint trace of continuum from g5 is detected redwards of 6600 ( not seen in figure [ em_mos ] due to the limited wavelength range ) .
the identification of these lines is discussed in [ gal_conf ] and summarized in table [ line_ids ] .
two hours of uves observations of q1331 + 17 were obtained on february 17 and 23 , 2002 in service mode .
the 390 + 564 dichroic setting was employed in order to give almost complete wavelength coverage from 3300 to 6650 with small gaps between the 3 ccds .
the spectra were extracted using a customised version of the uves pipeline ( ballester et al .
2000 ) following the steps outlined in ellison , ryan and prochaska ( 2001 ) . by using a 1 arcsecond slit ( aligned at the parallactic angle )
combined with 2@xmath292 binning , an average resolution of @xmath30 was achieved .
once converted to a linear vacuum heliocentric scale , the two individual exposures were combined with a weight proportional to the s / n ( the two spectra were taken under slightly different seeing and moon conditions ) and normalised by dividing through by a smoothed continuum .
the final s / n in the combined spectrum varies between approximately 40 ( per pixel ) in the blue and 70 in the red .
we have re - reduced the archival hst wfpc2 data of q1331 + 17 to re - assess the absolute magnitude of g5 and its error estimate , since there are various ( discrepant ) values available in the literature .
the archival wfpc2 data were obtained through the cadc and consisted of six 600s f814w images , as well as two 900s f702w images and two 600s f702w images .
individual exposures were cleaned from cosmic rays by comparing different exposures of the same mask using ppp ( yee , ellingson & carlberg 1996 ) .
additional cosmic ray residuals were tagged and masked by hand .
single images in f814w and f702w were produced by co - adding the individual exposures in each filter .
all images in the f814w filter were used . for the f702w filter
only the last three images were co - added , since the first f702 exposure had a large number of cosmic rays covering galaxy g5 .
we measured the counts in a 0.8 @xmath29 3.2 arcsecond rectangular aperture which was centred on g5 and whose long dimension was aligned with the major axis of the object .
the sky level was estimated in similar rectangular - shaped apertures placed to the east and west of g5 at the same distance from the qso as the g5 aperture , but adjusted in width to avoid contamination by light from the qso diffraction spikes .
we also experimented with an alternative method of sky subtraction which was to rotate the image 180@xmath31 around the qso and then subtract it from the original image .
both methods of sky subtraction produced consistent results .
in addition to g5 , aperture counts were also measured for a number of galaxies in the field and converted to apparent magnitudes using the published stsci instrumental zero points . for g5 ,
we determine ( on the ab magnitude scale ) @xmath32 and @xmath33 ( c.f . @xmath34 and @xmath35 from le brun et al .
1997 ) , where the uncertainties include both the poissonian count statistics and an estimate of the systematics in aperture choice and sky subtraction .
this is in good agreement with the magnitude of g5 derived from ground based observations ( @xmath36 = 23.25 steidel , private communication ) , but differs by almost 2 magnitudes from the value of le brun et al .
the magnitudes of other measured galaxies in the field are consistent with those of le brun et al .
we applied a k - correction appropriate for an scd galaxy ( coleman , wu & weedman 1980 ) , but this has a negligible effect since the @xmath37 band at the redshift of g5 corresponds nearly identically to the rest - frame b - band . therefore , converting to our adopted cosmology yields @xmath38 .
adopting @xmath39 ( efstathiou , ellis & peterson 1998 , adjusted to our cosmology and magnitude system ) this equates to @xmath40 compared with the value of @xmath41 determined ( for our cosmology ) from the le brun et al .
( 1997 ) magnitude .
two emission lines and a trace of faint continuum were found in the spectrum of g5 ( figures [ em_mos ] and [ 1d ] ) . there is a prominent line with a noticeable velocity gradient across the galaxy at @xmath42 ( figure [ em_mos]b , d ) , as would be expected from a large rotating disk galaxy . a second , broader and fainter line
is found at @xmath43 ( figure [ em_mos]a , c ) .
this second line is also slightly spatially displaced from the line at 6505 ( a _ conservative _ estimate of the offset between the line centers is 2 pixels ) , suggestive that the two lines may not be coming from the same physical location , although the spatial offset is a marginal one .
more convincingly , there is a complete lack of kinematic structure in the 5674 line when cross - correlated with a line template ( see 3.2 ) , whereas the 6505 line shows a clear rotation curve .
there are 4 previously known absorbers in the line of sight to q1331 + 17 ( see le brun et al .
1997 for a summary ) , therefore there are several possible galaxies which may be causing the emission lines .
furthermore , we have searched our uves spectrum ( which has extended blue coverage ) for additional absorbers and identify one new c iv system at @xmath44 .
table [ line_ids ] lists the rest - frame wavelengths that each of the two emission lines would have at the redshift of the four absorbers known in the spectrum of q1331 + 170 .
these rest - frame wavelengths were then compared with the lists of main emission lines from nebular objects in kinney et al .
( 1993 ) and osterbrock ( 1989 ) .
we now discuss each emission line in turn and present arguments for the line identification in each case . _
( a ) the emission line at 6505 .
_ there is a very good wavelength match for the 6505 line with [ oii ] @xmath453727 at @xmath46 , in agreement with the galaxy identification for this system given by le brun et al .
( 1997 ) and bond et al .
( 2001a ) .
at the moderate resolution of our spectra ( r @xmath47 ) we do not expect the [ oii ] @xmath453727 doublet ( @xmath48 3726.032 , 3728.815 , @xmath49 ) to be resolved .
however , a comparison with sky lines in the spectrum shows that this emission line has a broader profile than would be expected from a single unresolved line .
a simple simulation shows that the width of the emission line at 6505 is in fact consistent with an unresolved [ oii ] @xmath453727 doublet . coincidentally , there is a mg ii absorption system at @xmath19 which puts mg ii @xmath452796 at precisely the same observed wavelength as [ oii ] @xmath453727 at @xmath50 .
the absorption doublet can in fact be seen in the qso continuum in figure [ em_mos ] , with the bluer component lined up with the emission line . however , there is no sign of emission at the wavelength that would correspond to the 2803 component of the mg ii doublet .
although the @xmath51-value ratio of mgii 2796:2803 is 2:1 , our spectrum is sufficiently sensitive that we would have resolved and detected the redder doublet feature were it present .
furthermore , galaxies generally show mg ii in absorption rather than emission ( e.g. , kinney et al . 1993 ) .
thus , we rule out mg ii at @xmath52 as the cause for the emission line ( see also the discussion on the emission line at 5674 ) .
finally , there is a near match between the he ii @xmath45 2733 line at the redshift of the z=1.374 c iv system .
interestingly , there is a second he ii line , at @xmath53=2390 that would also be a near match to the observed emission line at 5674 .
however , there are several pieces of evidence against this identification .
first , these he ii lines are generally very weak , and rarely seen in local galaxy spectra ( kinney et al . 1993 ) .
most importantly , however , the relative wavelength ratios of the helium lines do not match the observed ratios ( even accounting for possible sky line contamination ) ; since these emission lines originate from the same species , this would be a fundamental requirement for this match .
_ ( b ) the emission line at 5674 .
_ in contrast to the above , there is no bright emission line expected at this wavelength for _ any _ of the redshifts of the absorbers towards this qso .
the line has a fwhm of 15 ; which corresponds to an effective @xmath54 fwhm once the 5.2 instrumental profile is subtracted in quadrature .
this is equivalent to @xmath55 km / s , and could be caused by either an unresolved multiplet , or from physical broadening of the line ( e.g. , from agn activity ) .
although there are no obvious identifications for this emission line associated with the known intervening absorbers , there are some known fainter lines which may match .
the first feasible possibility is [ neiv ] 2423 , 2425 at @xmath56 , which would be expected @xmath1730 ( @xmath57 km / s ) blueward of the observed line .
although usually faint in galaxies , this transition can be relatively strong in some narrow line radio galaxies ( mccarthy 1993 ) .
however , agn activity is usually accompanied by other strong emission lines , such as mg ii which we had previously discounted as the identification of the 6505 line to the lack of a doublet feature .
the second close match is the cii ] multiplet 23242328 at @xmath58 which would be expected @xmath1717 ( 900 km / s ) to the red of the observed line .
the ground state doublet cii ] 2324 , 2325 has an @xmath51-value ratio that is @xmath17 18:1 , therefore the doublet blend is unlikely to account for the broadened line , since the emission is so faint .
the excited state transition cii ] @xmath452326 is seen in some agn ( e.g. , mccarthy 1993 ) , but again we see no associated mg ii emission .
finally , as discussed above , there is a reasonable match with he ii @xmath452390 , but this is also excluded ( see previous paragraph ) .
therefore , none of these possibilities presents a very convincing case , suggesting that this emission line is associated with an object not known in absorption .
finally , it is possible that this feature might be due to an internal reflection in the spectrograph however , the fact that it is coincident with g5 , and that it is noticeably more extended than a psf makes this an unlikely possibility .
it is , of course , plausible that the detected emission lines do not correspond to a known absorber . given that we have argued ( based on spatial and kinematic differences ) that the two detected emission lines arise in different objects , an unequivocal identification of the 6505 line as [ oii ] @xmath453727 is not possible .
lilly et al .
( 1995 ) explored the problems associated with redshift confirmation for single emission line objects in the cfrs and defined a break index to test for [ oii ] @xmath453727 , effectively measuring the presence of a balmer break .
although our spectrum is not of sufficient quality to reliably perform this test , there is a small rise in flux to the red of the emission line , suggestive of a possible break that is qualitatively similar to figure 2 of lilly et al .
finally we note that the other main emission lines regularly observed in moderate redshift galaxies can be confidently ruled out - h@xmath59 @xmath456563 would have to be blueshifted and the [ oiii]@xmath484959 , 5007 doublet would be resolved in a clear 1:3 ratio . if the observed emission line at 6505 were h@xmath60 @xmath454861 , we would also have observed [ oiii]@xmath48 4959 , 5007 . to summarise this subsection , after an exhaustive analysis of possible emission line matches ,
we have identified the line at 6505 as [ oii ] @xmath453727 emission from g5 at the redshift of the @xmath61 absorber .
this confirms the galaxy identification of le brun et al .
( 1997 ) and bond et al .
( 2001a ) .
the exact systemic redshift depends on which wavelength is taken as the central point of the rotation curve . since our s / n is low , this is difficult to determine accurately , so we adopt @xmath62 for the rest of the paper .
we find that the additional emission line seen at 5674 is not at the same redshift , but are unable to find a convincing match with any known absorption galaxy .
regardless of the identification , if we accept this line as real ( rather than an artifact ) it implies that the image of g5 is a superposition of a galaxy associated with the @xmath63 mg ii absorber , and a galaxy or agn at a different redshift .
each line of pixels in the g5 spectrum was cross - correlated with a template of the [ oii ] @xmath453727 doublet , which was constructed by co - adding two arc lines with the separation corresponding to 3726.032 , 3728.815 in a 1 to 1.3 ratio at a redshift of @xmath63 .
error bars were estimated from a simulation in which random gaussian noise was added to each line of the spectrum 100 times , and the cross - correlation with the template was performed on each noisy spectrum .
the resulting velocity points and errors are shown in figure [ em_abs ] and exhibit a linear velocity gradient ( a similar test on the unidentified emission line at 5674 exhibits no velocity structure ) .
while the velocities in the upper region of the emission line ( above pixel 25 ) are reliable , the lower region lies in a part of the spectrum which originally had strong sky emission , and the observed [ oii ] line and its velocity are therefore suspect .
in particular , the line profiles for the lower portion of the emission line are skewed towards shorter wavelengths , therefore giving a cross - correlation velocity which is likely blue - shifted from the true velocity ( compare the lower portion of figure 2d to the lower panel of figure [ em_abs ] ) .
we perform a least squares fit to the velocity points , shown as a diagonal line in figure [ em_abs ] . the errors on the least squares fit yield a formal error of @xmath6432 km / s on the rotation velocity .
however , in practice we are restricted to quoting a lower limit to @xmath65 by the low s / n at the lower end of the curve , possible contamination by sky line residuals and by the fact that we may well not be seeing out to the flat part of the rotation curve .
we also note that convolution of the emission line with the seeing profile will also serve to under - estimate @xmath65 by blending signal from stronger emission towards the inner part of the disk at lower velocities .
the straightforward interpretation of these data is that g5 is a large disk galaxy with @xmath4 km / s ( determined from the upper portion of the emission line ) and possibly as large as 350 km / s if the lower extension of the [ oii ] line is real .
as a second estimate of the rotation velocity , we calculate @xmath65 directly from the tully - fisher relation ( e.g. , tully & fisher 1977 ) , based on the absolute magnitude of the galaxy , as derived from our re - appraisal of the hst data . in order to compare
with published tully - fisher relations locally and at intermediate redshift , we convert the @xmath66 calculated for our preferred cosmology to those adopted by vogt et al .
( 1997 ) and kannappan , fabricant & franx ( 2002 ) , i.e. , @xmath67 = 0.05 and @xmath68=75 km / s / mpc .
we further correct this value of @xmath69 by + 0.14 mags to convert to the vega scale and by @xmath70 mags following the prescription of tully & fouqu ( 1985 ) to correct for internal extinction for an edge - on galaxy .
the resulting magnitude after these corrections is @xmath71 .
we derive the expected velocity for a galaxy of this magnitude using the tully - fisher relation from kannappan , fabricant & franx ( 2002 ) ( kff ) .
we calculate the maximum rotation velocity by taking the unweighted inverse fit from table 1 of kff @xmath72 , where @xmath73 is the 50% hi 21-cm linewidth .
the linewidth is related to the rotation velocity @xmath74 by @xmath75 with scatter 25 km / s ( kff appendix b ) . using our @xmath71 value for g5
, we obtain @xmath76 km / s , and @xmath77 km / s . in order to reproduce the observed ( minimum ) rotation velocity of @xmath78 km / s , the same relationship from kff yields an absolute @xmath79-band magnitude of @xmath80 , i.e. , 1.26 magnitudes brighter than the value that we measure ( once the edge - on dust correction of tully & fouqu 1985 is made ) .
if the lower extension to the rotation curve is real , then the rotation velocity could be as large as 350 km / s , which would imply @xmath81 , i.e. , require an additional 3.4 magnitudes of extinction after the edge - on correction in order to match the tully - fisher relation . therefore , there is a significant discrepancy between our measured luminosity and rotation curve with the local tully - fisher relation , which would be further exacerbated by including the lower extension of the emission line .
there is still considerable debate over the possible evolution of the tully - fisher relation at @xmath82 .
vogt et al .
( 1996 , 1997 ) found that out to @xmath83 there is relatively little evolution ( @xmath84 mag brightening ) for large disk galaxies to z@xmath171 , consistent with the hst wfpc2 morphological study of lilly et al . , ( 1998 ) .
other groups , however , have found significant luminosity evolution in small , blue galaxies ( e.g. , phillips et al .
1997 , rix et al .
1997 , simard & pritchet 1998 , malln - ornelas et al .
1999 ) , and even in large spirals ( e.g. , rigopoulou et al .
ziegler et al . (
2002 ) have suggested that the extent of luminosity evolution depends on rotation velocity with significant corrections required for galaxies with @xmath85 km / s . despite this still murky picture
, we can nonetheless make a clear statement about the case of g5 .
all suggestions of evolution in the tully - fisher relation advocate brightening ( for a given @xmath65 ) at higher redshifts ; if present this would further increase the discrepancy between our measured luminosity , and that predicted by the tully - fisher relation for our measured @xmath65 . in order to determine column densities for the metal species ,
we have used the voigt profile fitting program vpfit to determine column densities for the individual absorption components .
we have adopted the common strategy of tying doppler widths ( @xmath86-values ) and redshifts of components for different species and obtain good simultaneous fits for mg i , mg ii and fe ii indicating that these species occupy the same velocity space .
fit parameters are given in table [ vp_fits ] and we note that since several of the mg ii components are mildly saturated , we quote the total n(mg ii ) as a lower limit . in principle , hst spectra are required in order to determine the n(h i ) of this system .
however , in practice , the lyman limit of the higher redshift dla prevents us from measuring the ly@xmath59 absorption in the lower redshift system .
however , as discussed by bond et al.(2001a ) , although the mg ii @xmath452796 and fe ii @xmath452600 equivalent widths ( ew ) in this absorber give a 50% probability that this is a dla ( rao & turnshek 2000 ) , the absence of extended mg ii saturation would be unusual for a _ bona fide _ damped system . due to the lack of n(h
i ) constraint , we are unable to determine abundances for this absorber
. however , we return to the issue of relative column densities below .
the absorption profiles of this system are extended over approximately 300 km / s .
we adopt the systemic absorption redshift to be the same as that determined for the [ oii ] @xmath453727 emission line , i.e. , @xmath87 .
figure [ uves_data ] shows the mg ii and fe ii lines covered by the uves data .
the slightly higher resolution of the uves data compared with the hires spectra of bond et al .
( 2001a ) reveals even more striking symmetries in the absorption features .
however , the clearest examples are now not the pairs that are separated by @xmath88 km / s , but those with much larger velocity differences e.g. , the strongest components at @xmath89 km / s and @xmath90 km / s . in figure [ symm ]
we illustrate qualitatively some of these symmetries with the mg ii @xmath452803 transition .
the top panel shows the spectrum on both a velocity scale relative to @xmath91 and also shifted by 150 km / s to the blue demonstrating the keen similarities in terms of velocities between components 6 10 and 13 17 , as marked by the horizontal bar .
the middle panel shows the spectrum both relative to @xmath91 and also flipped in velocity space to illustrate the ` mirror ' symmetries .
notice how the pairs of lines at @xmath92 km / s ( components 6 , 7 & 8 compared with 13 , 14 & 15 ) overlap , including the broadening towards zero velocity .
the bottom panel illustrates another example of such a mirror symmetry , this time between components 9 and 10 compared with 16 and 17 . with a single line of sight , we are not able to further constrain the possible geometry of this system ( c.f .
rauch et al . 2002 ) , but since patterns of this sort may be expected if the line of sight intersects an expanding shell , we speculate that the components 6 17 may arise as a consequence of bubbles ( see also bond et al.2001a ) .
we can , however , assess whether the observed symmetries could plausibly arise by chance due to a random placement of lines by using the two point correlation function ( tpcf ) and column density distribution function ( cddf ) for mg ii systems from churchill , vogt & charlton ( 2002 ) .
the tpcf quantifies the probability of finding 2 absorption components with a separation @xmath93v and is parameterised by a double gaussian of the form @xmath94 from their large sample of mg ii systems , churchill , vogt & charlton ( 2002 ) have determined @xmath95 = 54 km / s and @xmath96 = 166 km / s .
the cddf describes the number of components with a given column density and is a power law of the form @xmath97 with @xmath98 for mg ii components . by combining the tpcf and the cddf we can assess the probability that two pairs of components in a given system will have the same @xmath93v and matching column densities , within some tolerance ( we make no distinction between ` shifted ' symmetries and ` mirror ' symmetries in these calculations ) .
we produce 10000 realizations of 2 pairs of mg ii lines and apply a relatively relaxed tolerance that @xmath93v @xmath99 10 km / s and @xmath93logn @xmath99 0.3 dex .
the results are shown in figure [ test_symmetry ] . in the top panel ,
we plot the fraction of pairs that have the same @xmath93v , as a function of @xmath93v . since small values of @xmath93v are more probable , this fraction is larger for smaller velocity separations . in the middle panel , we plot the fraction of pairs which have matching @xmath93v _ and _ logn , again as a function of @xmath93v .
the typical @xmath93v between components of the symmetric pairs observed in q1331 + 17 is @xmath99 15 km / s , so the results of these simulations ( figure [ test_symmetry ] , middle panel ) would imply that in a few percent of cases we may expect to find _ one set _ of matching pairs .
however , this is a strong function of column density ; since the cddf is a power law with negative slope , high column density components are intrinsically rarer and therefore matches that include a high column density component will also be rare . indeed , several of the symmetric components in q1331 + 17 have relatively large column densities , log n(mgii)@xmath10012.7 . in the bottom panel of figure [ test_symmetry ]
we have plotted a histogram showing the absolute number of matches ( for 10000 realizations ) as a function of column density .
clearly , matched components with moderately large n(mg ii ) should be extremely rare .
moreover , these simulations only consider matches between two pairs of lines whereas in the spectrum of q1331 + 17 , we see multiple pairings , sometimes with more than two components .
while the probability of chance pairing of two lines is , as shown by our simulations , already small , the pairing of whole line complexes is even more unlikely ; its probability can be approximated by the probability of the pairing of n individual line pairs , and therefore decreases quadratically compared to the already small probability of chance pairings of individual lines .
therefore , the likelihood of observing symmetries such as those seen in this mg ii system are vanishingly small ( @xmath990.02% ) .
steidel et al . ( 2002 ) have recently studied the absorption and emission kinematics of 5 mg ii absorbers associated with inclined disks .
they find that 4/5 of these systems are consistent with absorbing gas that is co - rotating with the main galactic disk ( although the profiles are not consistent with pure disk rotation ) . in the mg ii absorber towards q1331 + 17 , we find that the absorption components fall approximately central to the systemic redshift of the emission line , but with a large spread ( @xmath17 100 km / s ) to either side ( figure [ em_abs ] ) .
although steidel et al . ( 2002 ) find one example of absorption gas located at the centre of the rotation curve , the velocity spread of the absorption components is very small . whilst we do not detect the turnover of the rotation curve ( and even with the conservative assumption that the curve flattens beyond the limit of the detected rotation ) it is clear from figure [ em_abs ] that there is _ no absorption at the extrapolated disk rotation velocity determined for g5_. in fact , figure [ em_abs ] implies that if all of the absorbing gas were associated with a disk , then a significant amount would have to be at counter - rotating velocities .
interestingly , lanzetta et al .
( 1997 ) have reported a similar incidence of counter - rotating gas in a @xmath101 lyman limit system associated with a 0.4@xmath5 spiral inclined at @xmath102 .
although several examples of counter - rotation between stars and gas have been studied in the local universe ( e.g. , galletta 1987 ; bertola , buson & zeilinger 1992 ; ciri , bettoni & galletta 1995 ; jore et al .
1996 ; rubin , graham & kenney 1992 ) , kinematic de - coupling remains a relatively rare phenomenon among spirals ( kannappan & fabricant 2002 ) . given the even higher scarcity of outer disk vs. inner disk gas decoupling ( e.g. , braun , walterbos , kennicutt 1992 )
, it would be extremely surprising that two ( q1331 + 17 and q0850 + 44 , lanzetta et al .
1997 ) out of 8 ( i.e. , including the 5 additional systems from steidel et al.2002 and the system of barcons et al . 1995 ) galaxy absorbers need to appeal to this explanation .
a more likely explanation in the case of the @xmath87 mg ii absorber towards q1331 + 17 is that internal non - disk kinematics play a dominant role , consistent with the scenario of bubbles or winds .
alternatively , the absorption could be associated with a companion to g5 ( i.e. , at the same redshift ) which would make its relative velocity entirely independent of g5 s rotation curve .
zaritsky et al .
( 1997 ) have shown that the velocity distribution between local spirals and their satellites extends to several hundred km / s , although most are within 200 km / s ( but mostly along the disk s minor axis ) .
lanzetta et al .
( 1997 ) also raise the possibility that a faint nearby galaxy could be causing the absorption in the case of q0850 + 44 .
their argument is further supported by the absorber s very low metallicity ( @xmath103 ) , which is much lower than would normally be expected for a 0.4@xmath5 based on the metallicity - luminosity relation ( kobulnicky & zaritsky 1999 ) .
unfortunately , due to the impossibility of an n(h i ) measurement for the absorber in q1331 + 17 , we can not make a similar comparison for the mg ii system studied here .
although g5 is fainter than the typical mg ii galaxy ( steidel , persson & dickinson 1994 ) , its luminosity and impact parameter are consistent with the gas cross section relation derived by steidel ( 1995 ) , @xmath104 . whilst the correlation between mg ii equivalent width and
luminosity is weak ( churchill et al .
2000b ) , it is nonetheless striking that the absorber towards q1331 + 17 is one of the strongest compared with the sample of churchill et al .
( 2000a ) , yet it is at the faint end of the luminosity distribution .
this could be a further symptom of superbubbles which would increase the density of local gas and consequently increase the n(mg ii ) for a given metallicity .
this possibility could be tested by identifying the absorbing galaxies associated with other potential superbubble mg ii candidates presented in bond et al .
( 2001b ) .
a re - assessment of the archival wfpc2 imaging has led to a significant downward revision of g5 s luminosity to @xmath105 .
compared with the rotation velocity of @xmath4 km / s , this is inconsistent with the tully - fisher relation , both locally and at higher redshifts .
we can speculate upon various possibilities that would reconcile this inconsistency .
the identification of a second emission line , indicates that a second galaxy is present coincident ( at least to some degree ) with g5 .
although the s / n of g5 in the wfpc2 image is low , there is evidence for a faint ` kink ' , see figure [ fors_mask]b .
this will undoubtedly affect the luminosity of g5 , although it would lead to an even fainter absolute magnitude .
therefore , appealing to luminosity evolution ( 3.2 ) or blending with another coincident object only exacerbate the tully - fisher discrepancy .
it is possible that our internal extinction correction may be inaccurate and large amounts of dust are known to exist both in and above the plane of disk galaxies ( e.g. , howk & savage 1999 ) .
however , the total extinction correction required to reconcile the luminosity with the tully - fisher relation is approximately 3 times the maximum given by tully & fouqu ( 1985 ) , more if the true rotation velocity is significantly larger than our current lower limit of 210 km / s .
we also note that none of the 5 galaxies studied by steidel et al .
( 2002 ) exhibit such a discrepancy with the tully - fisher relation .
it is also possible that g5 truly lies a long way from the classical tully - fisher relation .
for example , there is some evidence that low surface brightness ( lsb ) galaxies deviate from the classical tully - fisher relation in the sense that some are apparently under - luminous for the rotation velocities , e.g. , oneil et al .
however , more recent work suggests that the majority of discrepant points may suffer from h i contamination from nearby galaxies ( chung et al .
our interpretation of superbubbles ( see discussion in the next section ) signals active physical processes such as intense star formation or even merging which may impact significantly on the relation between luminosity and rotation velocity .
intense bursts of star formation will make a galaxy even brighter for a given @xmath65 , again exacerbating the tully - fisher discrepancy we observe , although the global effects on the tully - fisher relation of merging or starburst galaxies are unknown .
unfortunately , these data are not of sufficient quality to determine a star formation rate with a meaningful accuracy that would allow us to explore these possibilities in any quantative detail . finally , it is important to stress again that g5 is a faint galaxy with weak emission lines , so measurement errors are relatively large .
similarly , the tully - fisher exhibits a significant scatter .
as calculated above , the absolute @xmath79-band magnitude error would have to be under - estimated by a factor of @xmath17 6 , but the large error on the rotation velocity places it within a 1 @xmath106 scatter from the tully - fisher relation of kannappan , fabricant & franx ( 2002 ) .
the pair - wise absorption pattern noted by bond et al . ( 2001a ) is even more striking in the uves data presented here ( see figures [ uves_data ] and [ symm ] ) and is indeed reminiscent of the expected signature of expanding gas shells . whereas bond et al .
focussed on line pair separations of @xmath17 30 km / s , the uves data show even more striking symmetries between components with @xmath7 km / s , such as between components 6 , 7 & 8 with components 13 , 14 & 15 .
it is interesting to note that several local cases of absorption in qsos behind readily identifiable galaxies may be associated with disturbed gas , including bubbles , shells and mergers ( e.g. , norman et al . 1996 ; womble , junkkarinen & burbidge 1992 and references therein ) .
bond et al .
( 2001a ) used cloudy models to infer the energetics of the starburst that may be driving the bubble and the implied metallicity of the gas .
these models place limits on the energy injected into the bubble and its age , assuming a metallicity and an empirically motivated range of column densities and shell thicknesses . with a @xmath107 larger by a factor of 5 than those studied by bond et al .
( 2001a ) , these models would be shifted to energies some 300 times larger for a given age ( see bond et al .
2001a for the formalism of these calculations ) . the combined effect of metallicity and starburst luminosity governs the shell thickness and observed mg ii column density .
extrapolation of the models in bond et al .
( 2001a ) implies total luminosities of between 1 10 @xmath108 ergs @xmath109 corresponding to an ob association of some 500 5000 stars , and a metallicity of 1/10 solar .
given that this upper limit is similar to the largest ob associations seen at @xmath8 ( e.g. mckee & williams 1997 ) , this implies that lower metallicities are unlikely .
higher metallicities could be reproduced by proportionately lower energies .
the ratios of mg i / fe ii and mg i / mg ii contain information pertaining to the ionization state ( and hence , density ) of the gas and may provide clues to the nature of the absorption . in figure [ density ]
we plot the component by component mg i / fe ii and mg i / mg ii ratios ( along with the velocity - scaled absorption profiles to guide the eye ) . ideally , we would like to compare n(mg i)/n(mg ii ) and indeed there is an interesting trend that the components that exhibit the symmetries described above ( i.e. , those components we have tentatively identified as arising in bubble shells ) generally have higher n(mg i)/n(mg ii ) .
although churchill , vogt & charlton ( 2002 ) have recently shown that n(mg
i)/n(mg ii ) correlates with n(mg ii ) , the distribution of n(mg i)/n(mg ii ) in figure [ density ] is independent of this relation .
however , since many of the mg ii components are partially saturated , these ratios are often upper limits .
the ratio of n(mg i)/n(fe ii ) is unaffected ( in this case ) by saturation , although we have no _ a priori _ knowledge that the mg / fe ratio is identical in every component ( e.g. , ledoux , bergeron & petitjean 2002 ; ellison 2000 ) there is recent evidence that the abundance ratios , at least in dlas , show remarkable uniformity ( prochaska 2003 ; churchill et al .
nonetheless , the n(mg i)/n(fe ii ) is almost universally higher in the components proposed to be bubbles , indicative of large densities .
column densities of highly ionized species such as c iv would help to constrain the ionization parameter , however the only existing hst spectra of q1331 + 17 are of very poor quality ( bechtold , private communication ) .
rauch et al .
( 2002 ) find similarly high ratios of n(mg i)/n(mg ii ) and n(mg i)/n(fe ii ) in an absorber at @xmath110 which they have interpreted as due to superbubbles .
in addition , similar column density ratios of these species have been found in low redshift mg ii absorbers known to be undergoing strong galaxy - galaxy interactions ( norman et al .
1996 ) . in figure [ cloudy ] we present representative cloudy94 ( ferland et al .
1998 ) models of n(mg i)/n(mg ii ) and n(mg i)/n(fe ii ) to illustrate that large values of this ratio are indicative of high densities . however , modelling the precise density of absorbers depends on the n(h i ) which is unknown for this system .
moreover , given that the bubbles may consist of shocked gas that is not in thermal equilibrium and that the effective ionizing radiation spectrum is unknown , precise modelling of relative abundances is outside the scope of this paper .
we conclude that , although the nature of this absorber can never be unequivocally determined solely from its absorption properties , the double - peaked profile of the redward components , their striking symmetries and high n(mg i)/n(mg ii ) and n(mg i)/n(fe ii ) ratios and large mg ii equivalent width support the interpretation of superbubbles put forward by bond et al . ( 2001a ) , but the larger velocity separations investigated here imply larger energies and enriched gas .
we have presented complementary quasar absorption line spectra and spatially resolved emission line data for a candidate mg ii absorbing galaxy at @xmath11 .
our main conclusions are : 1 . we have obtained a spectroscopic redshift for galaxy g5 based on the detection of the [ oii ] @xmath453727 line .
the galaxy , located at 3.86 arcsec ( 28.3 @xmath23 kpc ) from qso q1331 + 17 , is at a redshift @xmath3 .
there is a second unidentified emission line which does not match any expected line at @xmath3 or for any other of the known qso absorption systems . in combination with its distinct spatial profile , this indicates a second object , at a different redshift is ( at least partially ) coincident with g5 .
2 . a re - analysis of archival hst data leads us to determine an absolute luminosity @xmath105 for this galaxy .
if g5 is an edge - on disk , our detection of [ oii ] @xmath453727 indicates a rotation velocity of @xmath111210 km / s .
this is inconsistent with the established tully - fisher relation both locally and at intermediate redshifts .
the qso absorption is aligned with the center of the emission line , i.e. , at @xmath112 .
there is no absorbing gas at the extension of the rotation curve , but there is a significant amount at negative velocities .
we suggest that this is most likely explained due to internal motions , such as winds or bubbles which may be associated either with non - disk component of g5 or an as yet unidentified companion .
4 . the double - peaked absorption profiles , striking symmetries and large n(mgi)/n(mgii ) and n(mgi)/n(feii ) ratios are indicative of expanding superbubbles .
this is consistent with observations at low redshift which find that extended absorption halos are often associated with disrupted gas , for example in shells , mergers or interactions .
superbubbles may also explain why the absorber towards q1331 + 17 has a relatively large mg ii equivalent width relative to the luminosity of its absorbing galaxy .
the ob association driving the bubble may have a total energy of 1 10 @xmath108 ergs cm @xmath113 and contain as many stars as the largest seen at the present day .
this indicates that the gas has been enriched to at least 1/10 solar metallicity .
sle is grateful to the ias and princeton university for hosting a visit , during which part of this work was carried out .
mjs acknowledges support from proyecto fondecyt de incentivo a la cooperacin internacional no .
7000529 for visits to chile .
gmo is partially supported by fundacin andes and by proyecto fondecyt regular 2000 no .
the authors thank nick bond , joop schaye , chris howk , howard yee and sebastin lpez for valuable input and chris churchill , pat hall , david bowen and todd tripp for stimulating discussions . | we present a comparative analysis of the galaxy emission and qso absorption kinematics of a @xmath0 mg ii system and its candidate absorbing galaxy ( g5 ) located 3.86 arcsecs ( 28.3 @xmath1 kpc ) from the qso .
we have obtained a spectrum of the galaxy candidate , previously identified as a luminous edge - on disk , and detect the [ oii ] @xmath2 doublet at a systemic redshift of @xmath3 . from slit spectroscopy of this galaxy ,
we find @xmath4 km / s , and possibly as large as 350 km / s .
plotted on the same velocity scale , the systemic redshift of the galaxy coincides with the center of the absorption system , although the absorption components span more than 100 km / s in either direction .
however , once the sense of the rotation is taken into account , there is no absorbing gas at the projected velocity of the disk rotation curve .
this implies incompatibility with a simple disk scenario .
moreover , a re - analysis of archival hst data reveals that the galaxy is only 0.3 @xmath5 , considerably less luminous than previously reported in the literature .
this is incompatible with the established tully - fisher relation at this redshift , unless approximately 2 magnitudes of total extinction are invoked .
careful inspection of the archival hst data reveal that g5 may well be composed of two galaxies , although the quality of the data do not permit a detailed investigation of this .
this possibility is further supported by the identification of a second faint emission line at @xmath6 , whose distinct spatial and velocity profile indicates that it arises in a different galaxy at a different redshift .
analysis of the absorption lines shows evidence for superbubbles in the interstellar medium ( ism ) of the absorbing galaxy , based on the striking symmetry between components and large mg i / mg ii and mg i / fe ii ratios , indicative of large densities . the large velocity separations between line pairings , @xmath7 km / s indicates that these bubbles may be powered by ob associations comparable to the largest observed at @xmath8 and that the gas is probably enriched to at least 1/10 solar metallicity .
this is consistent with observations at low redshift that extended mg ii halos are often seen in galaxies that contain disturbed gas .
superbubbles may also explain why the absorber has a relatively large mg ii equivalent width relative to the luminosity of the associated galaxy ( or galaxies ) . |
constraint logic programming ( clp ) @xcite is a natural and well understood extension of logic programming ( lp ) in which term unification is replaced by constraint solving over a specific domain .
this brings a number of theoretical and practical advantages which include increased expressive power and declarativeness , as well as higher performance for certain application domains .
the resulting clp languages allow applying efficient , incremental constraint solving techniques to a variety of problems in a very natural way : constraint solving blends in elegantly with the search facilities and the ability to represent partially determined data that are inherent to logic programming . as a result ,
many modern prolog systems offer different constraint solving capabilities .
one of the most successful instances of clp is the class of constraint logic languages using _ finite domains _
( @xmath0 ) .
finite domains refer to those constraint systems in which constraint variables can take values out of a finite set , typically of integers ( i.e. , a _ range _ ) .
they are very useful in a wide variety of problems , and thus many prolog systems offering constraint solving capabilities include a finite domain solver . in such systems ,
domain ( range ) definition constraints as well as integer arithmetic and comparison constraints are provided in order to specify problems . since the seminal paper of van hentenryck et al .
@xcite , many fd solvers adopt the so - called `` glass - box '' approach .
our fd kernel also follows this approach , based on a unique primitive called an _
indexical_. high - level constraints are then built / defined in terms of primitive constraints .
an indexical has the form ` x in r ` , where @xmath1 is a range expression ( defined in [ fig : range : syntax ] ) . intuitively , `
x in r ` constrains the @xmath0term ( @xmath0variable or integer ) ` x ` to belong to the range denoted by the term ` r ` . in the definition of the range special expressions
are allowed .
in particular , the expressions ` max(y ) ` and ` max(y ) ` evaluate to the minimum and the maximum of the range of the @xmath0variable ` y ` , and the expression ` dom(y ) ` evaluates to the current domain of ` y ` .
constrains are solved partially in an incremental using consistency techniques @xcite which maintain the constraint network in some coherent state ( depending on the arc - consistency algorithm used ) .
this is done by monotone domain shrinking and propagation . when all constraints are placed and all values have been propagated a call is typically made to a _ labeling _ predicate which performs an enumeration - based search for sets of compatible instantiations for each of the variables that remain not bound to a single value .
we refer to @xcite for more details regarding indexicals and finite domain constraint solving . in this paper , we present a new free library for constraint logic programming over finite domains , included with the ciao prolog system @xcite . the library is entirely written in prolog , leveraging on ciao s module system and code transformation capabilities in order to achieve a highly modular design without compromising performance .
we describe the interface , implementation , and design rationale of each modular component .
the library meets several design goals : a high level of modularity , allowing the individual components to be replaced by different versions ; high - efficiency , being competitive with other @xmath0 implementations ; a glass - box approach , so the user can specify new constraints at different levels ; and a prolog implementation , in order to ease the integration with ciao s code analysis components .
the core is built upon two small libraries which implement integer ranges and closures .
on top of that , a _ finite domain variable _ datatype is defined , taking care of constraint reexecution depending on range changes .
these three libraries form what we call the _ @xmath0kernel _ of the library .
this @xmath0kernel is used in turn to implement several higher - level finite domain constraints , specified using indexicals .
together with a labeling module this layer forms what we name the _
@xmath0 solver_. a final level integrates the clp(@xmath0)paradigm with our @xmath0solver . this
is achieved using attributed variables and a compiler from the clp(@xmath0)language to the set of constraints provided by the solver .
it should be noted that the user of the library is encouraged to work in any of those levels as seen convenient : from writing a new range module to enriching the set of @xmath0constraints by writing new indexicals .
one of the first clp(@xmath0)implementations is the chip system @xcite .
this commercial system follows a typical black - box approach : it consists of a complete solver written in c and interfaces in an opaque manner to a prolog engine .
this makes it difficult for the programmer to understand what is happening in the core of the system .
also , no facilities are provided for tweaking the solver algorithms for a specific application . more recent clp(@xmath0)systems such as those in sicstus @xcite , gnu prolog @xcite , and b - prolog @xcite are built instead following more the glass - box approach .
the basic constraints are decomposed into smaller but highly optimized primitives ( typically indexicals ) . consequently
, the programmer has more latitude to extend the constraints as needed .
however , even if such systems can be easily modified / extended at the interface level ( e.g. , both sicstus and b - prolog provide way to define new global constraints ) they are much harder to modify at the implementation level ( e.g. , it is not possible to replace the implementation of range ) .
the ciao clp(@xmath0)library that we present has more similarities with the one recently developed for swi prolog @xcite .
both are fully written in prolog and support unbound ranges .
the swi library is clearly more complete than ciao s ( e.g. , it provides some global constraints and always terminating propagation ) , but it is designed in a monolithic way : it is implemented in a single file , mixing different language extensions ( using classical prolog ` term_expansion ` mechanisms ) while the ciao library is split in more around 20 modules with a clear separation of the different language extensions @xcite .
summarizing , our library differs in a number of ways from other existing approaches : * first , along with more recent libraries it differs from early systems in that it is written entirely in prolog . this dispenses with the need for a foreign interface and opens up more opportunities for automatic program transformation and analysis .
the use of the meta - predicates ` setarg/3 ` and ` call/1 ` means that the use of prolog has a minimal impact on performance . *
second , the library is designed as a set of separate modules .
this allows replacing a performance - critical part like the range code with a new implementation better suited for it . *
third , the library supports the `` glass - box '' approach fully , encouraging the user to access directly the low - level layers for performance - critical code without losing the convenience of the high - level clp paradigm .
again , the fact that the implementation is fully in prolog is the main enabler of this feature . *
lastly , we have prioritized extensibility , ease of modification , and flexibility , rather than micro - optimizations and pure raw speed . however , we argue that our design will accommodate several key optimizations like the ones of @xcite without needing to extend the underlying wam . the rest of the paper proceeds as follows . in sec . [
sec : structure ] we present the architecture of the library and the interface of the modules . in sec .
[ sec : glass - box - effect ] we discuss with an example how to use the glass box approach at different levels for better efficiency in a particular problem , with preliminary benchmarks illustrating the gains . finally , in sec .
[ sec : conclusions ] we conclude and discuss related and future work .
the ciao clp(@xmath0)library consists of seven modules grouped into three logical layers plus two specialized prolog to prolog translators . in the definition of these modules and interfaces we profit from ciao s module system @xcite and ciao s support for assertions @xcite , so that every predicate is correctly annotated with its types and other relevant interface - related characteristics , as well as documentation .
the translators are built using the ciao _ packages _ mechanism @xcite , which provides integrated and modular support for syntax modification and code transformations .
a description of the user interface for the library along with up - to - date documentation may be found in the relevant part of the ciao manual . the global architecture
is illustrated in fig .
[ fig : global - architecture ] .
the kernel layer provides facilities for range handling and propagation chains , which are used for defining finite domain variables which , as mentioned before , are different from the standard logical variables .
the @xmath0layer defines a finite set of constraints such as ` a+b = c/3 ` , using indexicals .
these constraints are translated form their indexical form to a set of instructions of the kernel layer .
labeling and branch - and - bound optimization search modules complete the finite domain solver .
the clp(@xmath0)constraints are translated to @xmath0constraints by a clp(@xmath0)compiler .
we use attributed variables to attach a finite domain variable to every logical variable involved in clp(@xmath0)constraints .
thus , the clp(@xmath0)layer is thin and of very low overhead .
( fd_var ) at ( 0.5 , 1 ) @xmath0term ; ( fd_propags ) at ( -1.5 , -0.8 ) propagators ; ( fd_range ) at ( 1.5 , -0.8 ) range ; ( fd_var ) ( fd_range ) ; ( fd_var )
( fd_propags ) ; ( kernel ) at ( -0.25,0.3 ) ; ( n ) at ( -1.75,1.5 ) * @xmath0kernel * ; ( fd_const ) at ( 4 , -1.25 ) [ right ] @xmath0constraints ; ( labeling ) at ( 4,0.25 ) [ right ] labeling ; ( optim ) at ( 4,1.75 ) [ right ] b&b optimization ; ( optim ) to [ out=180,in=25 ] ( kernel ) ; ( optim ) to [ out=-45,in=45 ] ( fd_const ) ; ( labeling ) to [ out=180,in=0 ] ( kernel ) ; ( labeling ) to [ out=-90,in=120 ] ( fd_const ) ; ( fd_const ) to [ out=180,in=-20 ] node[below=13,pos=0.6 ] idx compiler ( kernel ) ; ( solver ) at ( 2.25,0 ) ; ( n ) at ( 6.5,-2.25 ) * @xmath0solver * ; ( clp_rt ) at ( 2.25,3.65 ) clp(@xmath0)run time ; ( clp_rt ) ( solver ) node[midway , auto = left ] clp(@xmath0)compiler ; ( solver ) at ( 2.25,0.5 ) ; ( n ) at ( -2.5,3.5 ) * clp(@xmath0 ) * ; the finite domain kernel is the most important part of the library .
its implementation freely follows the design of the gnu prolog @xmath0solver ( @xcite provides a general overview of this solver ) .
a finite domain variable is composed of a range and several propagation chains . when the submission of a constraint modifies the range of a finite domain variable , other finite domain variables depending on that range are updated by firing up constraints stored in propagation chains .
the propagation events are executed in a synchronous way , meaning that a range change will fail if any of its dependent constraints can not be satisfied
. the kernel implements arithmetic over ranges ( pointwise operations , union , intersection complementation , ... ) and management of propagation chains , amounting to the delay of prolog goals on arbitrary events .
these two elements are used to implement the two basic operations of a finite domain variable : ` tell ` and ` prune ` .
the first one attempts to constrain a variable into a particular range , while the second one ( prune/2 ) removes a value form the range of a variable .
the variable code inspects the new and old ranges and wakes up the suspended goals on a given variable .
all the data structures are coded in an object - oriented style .
efficient access and in - place update are implemented by using the ` setarg/3 ` primitive .
we took special care to use ` setarg/3 ` in a _
safe _ way to avoid undesired side effects , such as those described by tarau @xcite .
range handling is one of the most important parts of the library , given the high frequency of range operations .
indeed , the library supports three implementations for ranges : the standard one using lists of closed integer intervals ; an implementation using lists of open ( i.e. , unbounded ) intervals ; and a bit - based implementation which despite allowing unbound ranges is more suitable for problems dealing with small ranges .
indeed , the user is encouraged to implement new range modules which are better suited to some particular problems .
the interface that a range module must implement is split into two parts .
the first one , shown in fig .
[ fig : range : syntax ] , deals with range creation and manipulation .
each of the operations defined in the figure has a corresponding predicate .
for instance , bounds addition ` t+t ` is implemented by the predicate ` bound_add/3 ` , and similarly for the rest of the predicates .
note that it is a convention of the interface that any operation that tries to create an empty range will fail .
this is better for efficiency and we found no practical example yet where this would be inconvenient .
[ fig : range : preds ] lists the rest of the predicates that a range implementation must provide .
they are mainly used for obtaining information about a range and are instrumental for the labeling algorithms .
lllllll ` r ` & & @xmath2 & @xmath3 & ` t .. t ` & @xmath4 & ( interval ) + & & & @xmath5 & ` { t } ` & & ( singleton ) + & & & @xmath5 & ` r \/
r ` & & ( union ) + & & & @xmath5 & ` r /\ r ` & & ( intersection ) + & & & @xmath5 & ` - r ` & & ( complementation ) + & & & @xmath5 & ` r + n ` & & ( pointwise addition ) + & & & @xmath5 & ` r - n ` & & ( pointwise subtraction ) + & & & @xmath5 & ` r * n ` & & ( pointwise multiplication ) + + ` t ` & & : : = & & ` min(y ) ` & & ( minimum ) + & & & @xmath5 & ` max(y ) ` & & ( maximum ) + & & & @xmath5 & ` dom(y ) ` & & ( domain ) + & & & @xmath5 & ` val(y ) ` & & ( value ) + & & & @xmath5 & @xmath6 & & ( arithmetic expression ) + & & & @xmath5 & ` n ` & & ( bound ) + [ cols= " < , < " , ] the differences go from negligible to more than 50% . in a different benchmark ( bridge ) , the closed interval version was 25% faster than the open one .
we now focus on the different possibilities that the library allows for @xmath0constraint programming . in the queens program ,
the main constraint of the problem is expressed by the ` diff/3 ` constraint : .... diff(x , y , i ) : - x # \= y , x # \= y+i , x+i # \= y. .... where ` i ` will be always an integer .
however , the compiler can not ( yet ) detect that ` i ` is an integer , and may perform some unnecessary linearization .
we may skip the compiler and define ` diff ` using directly the @xmath0constraints : .... diff(x , y , i):-
fd_constraints:'a<>b'(~w(x),~w(y ) ) , fd_constraints:'a<>b+t'(x , y , i ) , fd_constraints:'a<>b+t'(y , x , i ) . ....
the speedup is considerable , getting close to 50% speedup in some cases .
indeed , the compiler should be improved to produce this kind of code by default .
the user may notice that the above three constraints may be encoded by using just two indexicals .
for instance one can use the following definition for ` diff/3 ` : .... diff(x , y , i):-
idx_diff(~w(x ) , ~w(y ) , i ) .
idx_diff(x , y , i ) + : x in -{val(y ) , val(y)+c(i ) , val(y)-c(i ) } , y in -{val(x ) , val(x)+c(i ) , val(x)-c(i)}. .... again , the improvement is up to 40% from the previous version .
however , the constraint ` diff ` can be improved significantly by using directly the kernel delay mechanism ( val chain ) and @xmath0variable operations . in particular , we use the optimized kernel ` prune/2 ` operation that removes a single element form the range of a variable : .... diff(x , y , i):- wrapper(x , x0 ) , wrapper(y , y0 ) fd_term : add_propag(y , val , ' queens : cstr'(x0 , y0 , i ) ) , fd_term : add_propag(x , val , ' queens : cstr'(y0 , x0 , i ) ) .
% y is always a singleton .
cstr(x , y , i):- fd_term : integerize(y , y0 ) , fd_term : prune(x , y0 ) , y1 is y0 + i , fd_term : prune(x , y1 ) , y2 is y0 - i , fd_term : prune(x , y2 ) .
.... we reach around 80% speedup from the first version , and this result is optimal regarding what the user can do .
additional speedups can be achieved , but not without going beyond our glass - box approach .
indeed , our clp(@xmath0)compiler is simpler given that we are working on a new translator that directly generates custom kernel constraints from clp(@xmath0)constraints .
the ciao clp(@xmath0)library described is distributed with the latest ciao version , available at http://ciaohome.org . although included in the main distribution , it lives in the ` contrib ` directory , as it should be considered at a beta stage of development . even if we did not include yet important optimizations that should improve significantly the performance of the library
, the current results are encouraging .
the library has been used successfully internally within the ciao development team in a number of projects .
the modular design and low coupling of components allow their easy replacement and improvement .
indeed , every individual piece may be used in a glass - box fashion .
we expect that the use of prolog will allow the integration with ciao s powerful static analyzers . at the same time
, the clear separation of run - time and compile - time phases allows the modification and the improvement of the translation schemes in an independent manner .
indeed , the advantages of this design have already been showcased in @xcite , where a prolog to javascript cross - compiler was used to provide a js version of the library and which only required replacing a few lines of code . using this
cross - compiler clp(@xmath0)programs can be run on the server side or on the browser side unchanged . regarding future work , we distinguish two main lines : the kernel and the clp(@xmath0)compiler . for the kernel ,
the first priority is to finish settling down its interface . while we consider it mature , some optimizations
like avoiding reexecution may require that we include more information in our @xmath0variable structure , range modification times , etc .
indeed , we would like to support more strategies for propagators than the current linear one .
support for some global constraints is on the roadmap , and will likely mean the addition of more propagation chains .
the library features primitive but very useful statistics .
however we think it is not enough and we are working on an @xmath0instrumentation package that will provide detailed statistics and profiling .
this is key in order to extract the maximum performance from the library .
once we get detailed profiling information from a wide variety of benchmarks , a better range implementation will be due . regarding the clp(@xmath0)compiler , the current version should be considered a proof of concept .
indeed , we are studying alternative strategies including the generation of custom kernels or specialized @xmath0constraints for each particular program in contrast to the current approach of mapping a clp(@xmath0)program to a fixed set of primitive constraints .
ciaopp ciao s powerful abstract interpretation engine could be used in the translation , providing information about the clp(@xmath0)program to the clp(@xmath0)compiler so it can generate an optimal kernel of @xmath0code for that program . in this sense
, we think that we will follow the ciaopp approach of combining inference with user - provided annotations in the new clp(@xmath0)compiler .
the authors would like to thank the anonymous reviewers for their insightful comments .
the research leading to these results has received funding from the madrid regional government under cm project p2009/tic/1465 ( prometidos ) , and the spanish ministry of economy and competitiveness under project tin-2008 - 05624 _ doves_. the research by rmy haemmerl has also been supported by picd , the programme for attracting talent / young phds of the montegancedo campus of international excellence . 10 jaffar , j. , maher , m. : onstraint lp : a survey .
jlp * 19/20 * ( 1994 ) 503581 , saraswat , v. , deville , y. : design , implementation and evaluation of the constraint language cc(fd ) .
journal of logic programming * 37*(13 ) ( 1998 ) 139164 dib , m. , abdallah , r. , caminada , a. : arc - consistency in constraint satisfaction problems : a survey . in : second international conference on computational intelligence ,
modelling and simulation .
( 2010 ) 291296 hermenegildo , m.v . ,
bueno , f. , carro , m. , lpez , p. , mera , e. , morales , j. , puebla , g. : n overview of ciao and its design philosophy .
theory and practice of logic programming * 12*(12 ) ( january 2012 ) 219252 http://arxiv.org/abs/1102.5497 .
dincbas , m. , hentenryck , p.v . , simonis , h. , aggoun , a. : he constraint logic programming language chip . in : proceedings of the 2nd international conference on fifth generation computer systems .
( 1988 ) 249264 carlsson , m. , ottosson , g. , carlson , b. : an open - ended finite domain constraint solver . in : proceedings of the9th international symposium on programming languages : implementations , logics , and programs : including a special trach on declarative programming languages in education .
plilp 97 , london , uk , uk , springer - verlag ( 1997 ) 191206 d. diaz , s.a . ,
codognet , p. : on the implementation of gnu prolog . theory and practice of logic programming * 12*(12 ) ( january 2012 ) 253282 codognet , p. , diaz ,
d. : compiling constraints in clp(fd ) . j. log . program .
* 27*(3 ) ( 1996 ) 185226 zhou , n.f . :
programming finite - domain constraint propagators in action rules .
. log . program .
* 6*(5 ) ( september 2006 ) 483507 triska , m. : the finite domain constraint solver of swi - prolog . in schrijvers ,
t. , thiemann , p. , eds . : functional and logic programming .
volume 7294 of lecture notes in computer science .
springer berlin / heidelberg ( 2012 ) 307316 morales , j.f . , hermenegildo , m.v . ,
haemmerl , r. : odular extensions for modular ( logic ) languages . in : 21th international symposium on logic - based program synthesis and transformation ( lopstr11 ) , odense , denmark ( july 2011 ) to appear .
daz , d. , codognet , p. : inimal extension of the wam for clp(fd ) . in : proceedings of the tenth international conference on logic programming , budapest , mit press ( june 1993 ) 774790 cabeza , d. , hermenegildo , m. : ew module system for prolog . in : international conference on computational logic ,
number 1861 in lnai , springer - verlag ( july 2000 ) 131148 hermenegildo , m. , puebla , g. , bueno , f. , lpez - garca , p. : ntegrated program debugging , verification , and optimization using abstract interpretation ( and the ciao system preprocessor ) .
science of computer programming * 58*(12 ) ( 2005 ) 115140 tarau , p. : inprolog 2006 version 11.x professional edition user guide .
binnet corporation .
( 2006 ) available from ` http://www.binnetcorp.com/ ` .
schrijvers , t. , triska , m. , demoen , b. : or : extensible search with hookable disjunction . draft .
available from http://users.ugent.be/~tschrijv/tor/ ( 2012 ) morales , j.f .
, haemmerl , r. , carro , m. , hermenegildo , m.v .
: ightweight compilation of ( c)lp to javascript . theory and practice of logic programming , 28th intl .
conference on logic programming ( iclp12 ) special issue ( 2012 ) to appear .
.... queens(n , l , lab , const ) : - length(l , n ) , domain(l , 1 , n ) , safe(l , const ) , labeling(lab , l ) . | we present a new free library for constraint logic programming over finite domains , included with the ciao prolog system . the library is entirely written in prolog , leveraging on ciao s module system and code transformation capabilities in order to achieve a highly modular design without compromising performance .
we describe the interface , implementation , and design rationale of each modular component .
the library meets several design goals : a high level of modularity , allowing the individual components to be replaced by different versions ; high - efficiency , being competitive with other @xmath0 implementations ; a glass - box approach , so the user can specify new constraints at different levels ; and a prolog implementation , in order to ease the integration with ciao s code analysis components .
the core is built upon two small libraries which implement integer ranges and closures .
on top of that , a _ finite domain variable _ datatype is defined , taking care of constraint reexecution depending on range changes .
these three libraries form what we call the _
@xmath0kernel _ of the library .
this @xmath0kernel is used in turn to implement several higher - level finite domain constraints , specified using indexicals .
together with a labeling module this layer forms what we name _ the @xmath0 solver_. a final level integrates the clp(@xmath0)paradigm with our @xmath0solver .
this is achieved using attributed variables and a compiler from the clp(@xmath0)language to the set of constraints provided by the solver .
it should be noted that the user of the library is encouraged to work in any of those levels as seen convenient : from writing a new range module to enriching the set of @xmath0constraints by writing new indexicals . |
one of the ultimate goals in investigating central heavy - ion collisions at very high energies is to analyze the properties of highly excited and deconfined matter .
direct probes , like electromagnetic signals @xcite , are considered as useful carriers of nearly undisturbed information from the transient hot reaction stages .
pairs of electrons and muons are experimentally accessible , and will be measured also in forthcoming heavy - ion experiments @xcite . these types of probes have proven useful at super proton synchrotron ( sps ) energies at cern in the resonance region below the @xmath1 , @xmath2 @xcite , and in the @xmath3 region @xcite ; also in the so - called continuum region in between a yet unexplained excess seems to be found @xcite .
however , recent estimates @xcite of dileptons stemming from semileptonic decays of open charm and bottom mesons produced in the same central collisions of heavy ions show that these represent the dominating dilepton source at energies envisaged at the relativistic heavy ion collider ( rhic ) in brookhaven and at the large hadron collider ( lhc ) in cern . in particular , the dilepton signal from thermalized , strongly interacting matter in the intermediate reaction stages is estimated to be up to two orders of magnitude below the background of correlated decays of open charm or bottom meson pairs in a wide invariant mass region .
even strong energy loss effects of the heavy quarks in the hot and dense medium @xcite do not basically change this situation : the decay dileptons might be suppressed down to the drell - yan background , but it is questionable whether such strong energy loss effects can happen in reality @xcite . in this situation at least two issues can be seen .
first , it is necessary to extend the traditional strategy for searching a thermal dilepton signal from deconfined matter .
the usual approach relies on the analysis of the invariant mass ( @xmath4 ) spectra .
one can expect , however , that the double differential dilepton spectrum as a function of the transverse pair momentum @xmath5 and transverse pair mass @xmath6 contains much more information which can be used to impose special kinematical gates for finding a window to observe the thermal dileptons in the continuum region in spite of the large decay background .
second , one can try to employ the initial hard production of open charm and bottom followed by semileptonic decays to probe deconfined matter by energy loss effects .
a similar idea for hard jet production is widely under discussion now @xcite . in the present paper we focus on both of the above issues .
we compare the so - called @xmath0 scaling property of the dielectrons stemming from deconfined matter with pairs resulting from correlated open charm and bottom decays . as well - known , under certain conditions the dilepton yield from a thermalized quark - gluon plasma depends only on the transverse mass , i.e. , there is no dependence on the variable @xmath5 at fixed @xmath0 in a given rapidity interval @xmath7 . at the same time the corresponding @xmath5 dependence of dileptons from heavy quark decays has qualitatively a different behavior : it is peaked at small values of @xmath5 and drops rapidly at increasing values of the transverse momentum . to get realistic estimates of the @xmath0 scaling we perform monte carlo simulations for both the correlated heavy quark decays and the thermal production including the acceptance of the detectors phenix at rhic and alice at lhc . to strengthen the @xmath0 scaling violation for dileptons from decays we select only such leptons with individual transverse momentum @xmath8 with appropriate value of @xmath9 .
we find that , while the thermal dilepton yield from the plasma is less affected by the kinematical restriction , the contributions from decays at large transverse pair momentum is reduced drastically . actually this leads to a threshold like behavior of the transverse momentum spectrum at fixed value of @xmath0 for dileptons from heavy quark decays in contrast to the approximate plateau of the thermal yield which extends up to the kinematical limit at @xmath10 . as a result ,
an opportunity is found which allows to separate the thermal dilepton signal from the background of heavy quark decays on the basis of the @xmath0 scaling property .
we also study to what extent the single electron and dilepton spectra are modified by the energy loss effects of heavy quarks when traversing deconfined matter . in doing
so we concentrate on calculable processes in the first few fm / c of the collision course .
there are , of course , other processes in the later reaction stages that might considerably affect the momenta of the charm and bottom mesons @xcite , and also thermal radiation of semihard probes can stem from later stages .
for instance , in ref .
@xcite the hypothesis has been formulated , resting on an analysis of present pb + pb data at sps energy , that all hadron spectra can be parametrized by a common flow velocity and freeze - out temperature .
this , if correct for higher energies too , would imply that charm and bottom spectra are much stronger modified than assumed in the following , and as a consequence the chances to identify a thermal dilepton source would be better . our paper is organized as follows . in section 2 we present our calculation procedure of charm and bottom spectra and the resulting single electron and dielectron spectra .
section 3 is devoted to the thermal dileptons at midrapidity and the @xmath0 scaling properties .
our conclusions can be found in section 4 .
in our approach we utilize the leading order qcd processes @xmath11 and @xmath12 for heavy quark - antiquark ( @xmath13 ) production and simulate higher order corrections by an appropriate @xmath14 factor .
we have checked that such a procedure reproduces within the needed accuracy the more involved next - to - leading order calculations @xcite .
the number of @xmath13 pairs , produced initially with transverse momenta @xmath15 at rapidities @xmath16 in central @xmath17 collisions can be calculated by @xmath18 \frac{d \hat \sigma_q^q}{d \hat t } \right\},\end{aligned}\ ] ] where @xmath19 are elementary cross sections ( see for details @xcite ) , @xmath20 with @xmath21 denote the parton structure functions , @xmath22 and @xmath23 .
as heavy quark masses we take @xmath24 1.5 gev and @xmath25 4.5 gev .
we employ the hera supported structure function set mrs d- @xcite from the pdflib at cern .
nuclear shadowing effects are not included since they will be considered separately .
in addition , for our present goals we expect no significant modification by shadowing in the large transverse momentum region according to results of ref .
the overlap function for central collisions is @xmath26 with @xmath27 fm and @xmath28 . from a comparison with results of ref .
@xcite we find the scale @xmath29 and @xmath30 2 as most appropriate .
energy loss effects ( cf . @xcite and further references therein ) have been found important @xcite since the momenta of back - to - back moving heavy quarks are degraded and the resulting dileptons in correlated semileptonic decays get less invariant mass .
this process causes a considerable reduction of the number of high invariant mass dileptons from correlated @xmath31 and @xmath32 decays . to model the energy loss effects of heavy quarks in expanding matter we assume that a heavy quark has the same rapidity as the longitudinal flow at the space - time point of the initial quark production .
therefore , with respect to the fluid s local rest frame the heavy quark has essentially only a transverse momentum @xmath33 which may depend on the proper local time @xmath34 in accordance with the energy loss @xmath35 in transverse direction .
this picture implies a very fast thermalization of gluons and light quarks .
unfortunately a sufficiently general treatment of the energy loss rate @xmath35 for heavy quarks passing through expanding qcd matter has not been developed so far .
recently the total energy loss of a high - energy parton propagating transversely through expanding deconfined matter has been derived as @xmath36 @xcite , supposed the medium cools according to a power law and the initial time is short .
the numerical factor @xmath37 is @xmath38 for a parton created inside the medium .
within such a scenario the stopping power depends only on the final temperature at which the deconfined medium is left , as indicated by the subscript @xmath39 . we denote this stopping scenario as model i and take @xmath40 from @xcite .
this model is appropriate for lhc conditions , where the initial temperature is very high and where the life time of deconfined matter is large enough that most energetic heavy quarks can leave this stage before @xmath41 is reached . for rhic conditions we adopt the results of ref .
@xcite for @xmath35 in a static medium but apply them in the simplest generalization to a medium with time dependent density .
this gives evolution equations for the transverse momentum @xmath42 of quarks propagating the distance @xmath43 in the transverse direction .
the scenario is denoted as model ii , which is described in more detail in refs .
@xcite . strictly speaking , due to the finite size of the medium , the large mass of the heavy quarks and other approximations inherent in such approaches , the models i and ii provide at best semi - quantitative estimates .
nevertheless one can hope to cover various limiting cases to get some insight in the importance of the energy loss effects . to get the spectra of charm and bottom quarks after energy loss we use a monte carlo simulation with a uniform distribution of the random initial position and random orientation of @xmath13 pairs in the transverse plane . in model
i the total energy loss is determined by the transverse distance @xmath44 of a created heavy quark to the boundary of the system , the quark initial energy and the temperature when leaving the system .
most quarks experience an energy loss @xmath45 and only a few ones @xmath46 .
( the transverse expansion at early times can be neglected , and we take as averaged transverse radius of parton matter @xmath47 fm . ) in model ii we integrate the evolution equations for the heavy quark momentum @xmath48 and position @xmath43 together with the time evolution of the temperature @xmath49 and fugacities @xmath50 ( see below ) as described in refs .
the quantity @xmath43 is used to check whether the considered heavy quark propagates still within deconfined matter ; if it leaves the deconfined medium it does not longer experience an energy loss .
( the energy loss in a mixed and hadron phase might be incorporated in line with the approach of @xcite . )
we consider a heavy quark to be thermalized if its transverse mass @xmath51 during energy loss becomes less than the averaged transverse mass of thermalized light partons at given temperature .
these heavy quarks are redistributed according to the transverse mass distribution @xmath52 with an effective temperature @xmath53 mev @xcite .
we employ a delta function like fragmentation scheme for the heavy quark conversion into @xmath54 and @xmath55 mesons .
it results in the same transverse momentum of the heavy meson as the heavy parent quark has had previously .
inclusive single transverse momentum spectra of electrons / positrons from semileptonic open charm and bottom decays , i.e. , @xmath56 , and dilepton spectra from correlated decays as well , i.e. , @xmath57 , are obtained from a monte carlo code which utilizes the inclusive primary electron energy distribution as delivered by jetset 7.4 .
the heavy mesons are randomly decayed in their rest system and the resulting electrons then boosted appropriately .
the average branching ratio of @xmath56 is taken as 12 ( 10)% @xcite .
the neglect of secondary electrons is justified since we consider here single electron spectra with @xmath58 1 gev and dilepton spectra with @xmath59 2 gev .
the average energies of the secondary electrons , in particular from @xmath55 decays , are too small to affect the spectra in these kinematical regions noticeable .
the following acceptance cuts are utilized in our monte carlo calculations .
phenix @xcite can register electrons in the pseudorapidity window @xmath60 with azimuthal coverage 0 - 90 and 135 - 225 degrees , while alice @xcite is azimuthally symmetric and covers @xmath61 . as minimum transverse electron momentum we use @xmath62 1 gev .
a maximum transverse momentum cut is not imposed , even at alice there is probably one .
we consider here only the electron and @xmath63 channels and the midrapidity region since the energy loss effects are strongest there @xcite .
the results of our calculations of the transverse momentum spectra of inclusive single electrons at rhic conditions are depicted in figs .
1a and b. as already pointed out in ref .
@xcite , a measurement of single electrons with @xmath58 1.5 gev at rhic can give a clean and background - free charm signal in heavy - ion collisions .
energy loss effects of the heavy quarks modify this signal . as seen in figs .
1a and b such modifications , according to the energy loss described above , are measurable .
for both the charm and the bottom decay the transverse single electron spectra become softer .
this leads to a noticeable change of the slope of the spectra .
one can fit the distributions by @xmath64 in the interval 2.5 gev @xmath65 4.5 gev and finds @xmath66 646 ( 512 ) mev in case of no stopping ( stopping according to model ii ) for charm , and @xmath66 851 ( 640 ) mev for bottom .
the superposition of charm and bottom decay spectra can be parametrized by @xmath66 711 ( 574 ) mev . therefore , in comparing the @xmath33 spectra of single electrons in pp and aa collisions at the same beam energies one can reveal the stopping effect .
( a direct measurement of the @xmath67 @xmath33 distributions would be more useful , but it is unclear whether the @xmath67 mesons can be identified via their hadronic decay modes in such high - multiplicity environments . ) in extreme case , when the stopping is so large that thermalization with the surrounding matter happens , then a pronounced thermal bump at @xmath68 2.5
gev appears .
all of these considerable modifications of the transverse electron spectra in aa collisions in comparison with pp collisions can be helpful for identifying the creation of a hot and dense parton system .
it should be noted that the energy loss effects affect the single - electron @xmath33 spectra much stronger than another fragmentation scheme ( e.g. the below described peterson scheme ) would do .
the invariant mass dilepton spectra resulting from both correlated charm and bottom
decays are displayed in figs .
2a and b for rhic and lhc energies . also shown is the distribution of thermal dileptons according to eq . ( 9 ) below .
as pointed out earlier @xcite the @xmath4 spectra are by far dominated by the dileptons from correlated open charm and bottom decays .
a strong energy loss according to the above models reduces the decay contributions , so that ( i ) at rhic charm and bottom decay dileptons are in the same order of magnitude as the drell - yan background @xcite , and ( ii ) at lhc still bottom will dominate the drell - yan background , and charm is slightly below bottom @xcite . while without any detector acceptances usually the predictions of the thermal yield ( dealt with in the next section ) is in the same order of magnitude as the drell - yan background @xcite , it turns out that both the acceptances of phenix and alice suppress the thermal signal stronger than the drell - yan background .
at the first sight such a situation looks quite unfavorable for a measurement of the thermal signal even at such high energies as achieved at rhic and lhc .
nevertheless we show below that one can select special kinematical conditions superimposed to the detector cuts to analyze the more informative double differential dilepton spectrum with the aim to find a window for the thermal signal .
in order to estimate the thermal dilepton yield from deconfined matter we restrict ourselves to the lowest order processes @xmath69 , i.e. , the electromagnetic annihilation of quarks and antiquarks .
contributions from qcd processes like @xmath70 and @xmath71 have been considered @xcite and appear to be not very significant ( typically they increase the yield by a factor less than 2.5 ) .
the rate of dilepton production per space - time volume can be obtained within a kinetic theory approach as @xmath72 & & \times ( 2\pi)^4 \ , \delta^{(4 ) } ( p_1 + p_2 - p_+ - p_-),\end{aligned}\ ] ] where @xmath73 are the four momenta of incoming quarks ( @xmath74 1 , 2 ) and outgoing leptons ( @xmath75 ) .
the quantity @xmath76 denotes the square of the amplitude of the process , summed over initial and final state particles .
the quantities @xmath77 stand for the distribution functions of quarks and antiquarks . integrating eq .
( 3 ) over the momenta of the leptons one gets a widely used expression @xcite for the dilepton rate per four - momentum of the pair , @xmath78 , @xmath79 the pair s four momentum can be expressed as @xmath80 with @xmath6 as transverse mass , @xmath4 as invariant mass , and @xmath7 as rapidity of the pair .
the relative quark velocity reads @xmath81 , while the total cross section is @xmath82 with quark charge factor @xmath83 for u , d quarks , and @xmath84 is the fine structure constant . for the calculation of dilepton spectra in the region of large invariant mass one can utilize the boltzmann approximation for the incoming partons @xmath85 where @xmath86 is the four - velocity of the medium .
the above parton distributions are assumed to be in thermal equilibrium in momentum space , but not necessarily in chemical equilibrium . only for chemical equilibrium
the fugacities @xmath87 become unity .
the space - time evolution of the produced deconfined matter is governed in our approach by the longitudinal scaling - invariant expansion accompanied by quark and gluon chemical equilibration processes @xcite .
within such a model the space - time volume reads @xmath88 with @xmath34 as proper time and flow rapidity @xmath89 . integrating the rate eq .
( 4 ) over the parton momenta and space - time evolution results in the dilepton spectrum @xmath90 the physical information encoded in the spectrum eq .
( [ m_perp ] ) can be inferred from the following approximation . for large values of @xmath0
the main contribution to the time integral stems from early , hot stages at @xmath91 . under the assumption that the equation of state behaves like @xmath92 ( here @xmath93 and @xmath94 stand for the pressure and energy density , respectively )
one can derive from energy - momentum conservation the relation @xmath95 , @xmath96^{1/4}$ ] with @xmath97 as degeneracy factors of quarks and gluons . since near @xmath98
one can approximate @xmath99 and @xmath100 , one arrives at @xmath101 with @xmath102 and @xmath103 as modified bessel function of @xmath104th order .
for the possible initial conditions at lhc ( see below ) eq .
( [ m_perp_a ] ) represents a fairly good approximation , also for small values of @xmath105 .
( [ m_perp_a ] ) has the structure @xmath106 , therefore one can infer from it the value of @xmath107 by measuring the transverse rate at two distinct values of @xmath0 .
afterwards , the combination @xmath108 can be extracted . if one could constrain by other means the initial time of the thermalized era , @xmath98 , then even eq .
( [ m_perp_a ] ) allows to estimate the initial fugacity @xmath109 . due to
particle production processes the comoving entropy density does not longer serve as a link between initial and final states in the evolution dynamics .
our choice of initial conditions for produced deconfined matter is based on the estimates of refs .
@xcite for the minijet plasma with a suitable parametrization of the soft component @xcite which are similar to the self - screened parton cascade model @xcite .
we take as main set of parameters for the initial temperature @xmath110 550 ( 1000 ) mev , for gluon fugacity @xmath111 0.5 , and for light quark fugacity @xmath112 of the parton plasma formed at rhic ( lhc ) at initial time @xmath113 0.2 fm / c . to achieve an upper optimistic limit of the thermal dilepton yield we consider below also higher values of the initial quark and gluon fugacities ( @xmath114 0.5 , 1 , and @xmath111 1 ) which lead to different final states . for the sake of definiteness
we assume full saturation at confinement temperature @xmath115 170 mev and a quadratic time dependence of @xmath116 @xcite .
to get realistic dilepton spectra which can be measured in relativistic heavy - ion collisions one should take into account the experimental conditions related to the acceptance of detectors . to do
so we employ monte carlo simulations with proper distributions of created leptons with respect to individual rapidities @xmath117 and transverse momenta @xmath118 . in case of the thermal emission of leptons from deconfined matter the needed distribution can be obtained directly from eq .
( 3 ) . due to the energy - momentum conservation
the exponential function , originating from the boltzmann distribution , can be rewritten as @xmath119 . since also @xmath120 the integration in eq .
( 3 ) over the parton momenta yields @xmath121 where @xmath122 and the flux factor becomes @xmath123 .
then the space - time integrated distribution of thermal leptons reads @xmath124 with dilepton transverse mass @xmath125 . to get the invariant mass spectra we also use the kinematical relationship
@xmath126 $ ] , where @xmath127 denote the azimuthal angles of the leptons in the transverse plane .
as pointed out above , due to the dominant contribution of correlated open charm and bottom decays via @xmath128 into the high invariant mass dilepton spectra it is very unlikely that the thermal dilepton production from quark fusion can be observed by using the distribution @xmath129 alone even when imposing tricky cuts .
in this situation the more detailed information of the full spectrum @xmath130 is expected to allow for better chances to extract the wanted thermal signal . as known since some time @xcite and as directly seen in eq .
( 6 ) the dilepton yield from an equilibrium quark - gluon plasma depends only on the transverse mass , i.e. , @xmath131 ; that is , it scales with @xmath0 .
the main conditions for this famous @xmath0 scaling are @xcite ( i ) local thermalization of the source of the dileptons , ( ii ) no scale other then the temperature , and ( iii ) predominant boost - invariant flow of matter . the most convincing way to see the validity of the @xmath0 scaling is to consider the @xmath5 dependence of the dilepton spectrum at fixed value of @xmath0 . in case of the boost - invariantly expanding quark - gluon plasma one gets for @xmath132 indeed a straight line between @xmath133 0 and @xmath0
. a detailed analysis of the transverse momentum dependence of the dilepton spectra is performed in ref .
@xcite , where the possible origins of a @xmath0 scaling violation related to transverse flow and contributions from the hadron gas are studied .
the transverse expansion of deconfined matter affects hardly the @xmath0 scaling , while for hadron matter the life time is strongly reduced , so that its contribution to the total spectrum is moderate . in fact , our previous studies @xcite show that the conditions for the @xmath0 scaling are satisfied for radiation from thermalized matter expected to be created in central collisions at rhic and lhc . at the same time
the dileptons from correlated open charm and bottom decay should strongly violate the @xmath0 scaling if the @xmath134 and @xmath135 pairs are dominantly produced by hard initial parton collisions .
the qualitative argument is as follows : similar to the drell - yan process the hard charm and bottom pairs are created mainly back - to - back ( unless hard gluon radiation spoils this correlation ) , i.e. , the pair momentum is preferential @xmath136 , and consequently also the @xmath137 spectrum at fixed @xmath138 is strongly peaked at 0 .
the subsequent semileptonic decays cause a partial randomization of the lepton directions but , due to the lorentz boost effect , both the leptons still are concentrated at small values of @xmath5 . as a result one can expect that in the large @xmath5 region the thermal signal sticks out , while the semileptonic decay products are sharply dropping . to quantify this effect we perform monte carlo calculations generating thermal dielectrons in accordance with the distribution eq .
( 9 ) . as above
we take into account the acceptances of phenix at rhic and alice at lhc .
first we consider the simplified case without any momentum degradation of heavy quarks , i.e. , energy loss effects are not included and the @xmath139 function fragmentation scheme is employed .
this is obviously the most favorable case for the relative contribution from decay dileptons to the full dilepton yield . in order to get some threshold effect for the decay dileptons in dependence of @xmath5 at
fixed @xmath0 one needs to select leptons with individual transverse momenta greater than the maximum electron energy in the rest system of the decaying open charm or bottom meson .
typically such a threshold value is given by 1 ( 2 ) gev for charm ( bottom ) .
the results of our calculations of the @xmath5 spectra for rhic conditions are displayed in fig .
we select here dileptons with transverse mass within the interval 2 gev @xmath140 3 gev .
the value of the minimum transverse momentum of leptons is chosen as @xmath141 1 gev .
the detector acceptance destroys the ideal @xmath0 scaling , i.e. , the thermal yield gets some structure reflecting the @xmath33 cut and the geometry .
one can see the threshold like behavior of the dilepton spectrum from charm decays : there is a rapid dropping of the yield at @xmath142 1.3 gev .
on the contrary the thermal dilepton spectrum sticks out up to the kinematical boundary @xmath10 .
unfortunately such a value of @xmath141 1 gev is not large enough to cause a similar suppression effect for leptons from bottom decays .
nevertheless as pointed out in ref .
@xcite and as seen in fig .
3a the contribution from bottom decays is suppressed ( this is due to the narrow rapidity gap covered by phenix ) but it is clearly above the thermal yield for the initial quark fugacity @xmath143 0.1 . to estimate the chances to see a thermal dilepton signal over the background from bottom
decays we perform variations of the initial quark fugacity . due to the factor @xmath144 in the thermal rate eq .
( 9 ) , even a comparatively small increase of the initial quark phase space saturation can result in a noticeable enhancement of thermal dileptons compared to the decay background .
this can be seen in fig .
3a for the initial quark fugacities @xmath143 0.5 and 1 . for the latter value
the thermal signal is above the bottom decay contribution . for lhc conditions one can obtain the threshold like effect in the @xmath5 spectrum even for the dielectrons from the massive bottom decays . in fig .
3b we show our results for dileptons within the transverse mass interval 5.25 @xmath140 5.75 gev .
the electron transverse momentum cut @xmath141 2 gev causes a sufficiently strong threshold effect both for charm and bottom , while the dielectrons from thermalized matter exhibit the approximate @xmath0 scaling even for such a value of @xmath9 .
it gives therefore the opportunity to measure the thermal dilepton signal on the basis of the @xmath0 scaling at lhc conditions .
actually such measurements could be performed for lepton pairs with high enough transverse momentum but smaller invariant mass .
in particular , as follows from fig .
3b , for a fixed value of @xmath145 5.5 gev the proper region for transverse pair momenta is about 4 gev @xmath146 5.5 gev
. this window increases considerably with increasing initial quark fugacity @xmath147 of thermalized parton matter .
we perform also calculations of the @xmath5 spectra at fixed @xmath0 for decay dileptons including the peterson fragmentation function @xcite , where the function @xmath148^{-1}\ ] ] ( @xmath149 0.06 ( 0.02 ) for charm ( bottom ) @xcite ) prescribes the probability to change the spatial momentum of a heavy quark from @xmath150 to @xmath151 during hadronization .
independently we also study the influence of energy loss of heavy quarks in the deconfined medium .
both these effects cause a suppression of the dilepton yield from open charm and bottom , in particular in the small-@xmath5 ( large-@xmath4 ) region . since the suppression effect due to peterson s fragmentation is quite small ( typically within factor 2 )
we do not display it separately . on the contrary ,
the degradation of the heavy quark momenta due to energy loss in the deconfined medium causes a considerably suppression of the contribution of decay leptons in the @xmath5 spectrum at fixed value of @xmath0 . in fig .
4 we show the results of our calculations for lhc conditions with the same kinematical cuts as in fig . 3b . as seen in this plot ,
both charm and bottom are considerably suppressed due to energy loss effects , in particular in the region of smaller values of @xmath5 .
these calculations exploit the model i. in case of model ii the suppression effect is much stronger . with respect to
the wanted thermal dilepton signal the suppression of the @xmath5 spectra from bottom decays appears most important .
the proper @xmath5 window to extract thermal dileptons becomes greater , e.g. , 3 gev @xmath146 5.5 gev for initial quark fugacity @xmath143 0.5 .
therefore , the energy loss effect obviously helps to measure thermal dielectrons with alice at lhc on the basis of the @xmath0 scaling behavior . to clarify whether the proposed signal is experimentally feasible we estimate the background from @xmath152 decays within the same kinematical gates . for this background study
we also perform monte carlo simulations of @xmath153 decays utilizing the distribution @xmath154 with @xmath155 200 mev in accordance with present sps results for central pb + pb collisions at 158 agev beam energies @xcite .
the slope parameter @xmath156 is a convolution of local freeze - out temperature and transverse flow .
the latter one is expected to increase from sps to rhic to lhc due to higher rapidity densities , and it might cause a slightly higher value of @xmath156 . we choose here
the fictitious values @xmath157 2500 ( 500 ) for lhc ( rhic ) .
each @xmath152 is three - body decayed in its rest frame with the branching ratio of 1.2% @xcite and then the resulting @xmath63 pair is boosted and filtered .
we find the individual @xmath152 decay dileptons with 5.25 gev @xmath140 5.75 gev three orders of magnitude below the thermal dilepton signal and only at @xmath142 5.2 gev .
the flattening of the pion spectrum at large values of @xmath33 , due the cronin effect and hard scatterings , is not covered by our parametrization of the @xmath158 spectrum ; it will enhance the pion decay contribution .
otherwise some quenching @xcite at large values of @xmath158 can counteract and balance such effects .
the finite experimental momentum resolution will cause a decay contribution also at somewhat smaller values of @xmath5 .
however , these effects will hardly close the window for the thermal signal .
the next - to - leading order drell - yan contribution is also estimated to be strongly suppressed at larger values of @xmath5 ( say , for @xmath142 1 gev ) because of the chosen large @xmath33 cut .
the most prominent danger however comes from the combinatorical background from uncorrelated hadron decays : only after experimentally removing this background together with the background from the conversion dileptons , emerging from energetic photons in the detector material , one has a chance to see the interesting signal .
fully analog considerations also apply at rhic .
however , selecting electron pairs with such large values of @xmath159 5.5 gev and choosing @xmath141 2 gev results in prohibitory small counting rates which seem to make such a strategy not feasible . for
the above choice of @xmath159 2.5 gev and choosing @xmath141 1 gev the dileptons from @xmath152 decays produce here a contribution which is three orders of magnitude above the thermal signal , however only in the interval @xmath142 1.95 gev .
therefore , one remains in a situation as depicted in fig .
3a and discussed above .
in summary we analyze the chances of observing a thermal dilepton signal in high - energy heavy - ion collisions . exploiting currently estimated energy loss effects of heavy quarks passing through deconfined matter we find that the invariant mass spectra of dileptons , as measurable in the detector facilities phenix and alice , offer hardly a chance to find thermal dileptons with @xmath59 2 gev .
even if in optimistic energy loss scenarios the dileptons from correlated charm and bottom decays compete with the drell - yan background , the thermal signal is significantly smaller . since the energy loss is a poorly quantified effect we stress the need to make attempts to verify it , for instance via the single electron @xmath33 spectra which should reflect the transverse momentum degradation of parent charm and bottom mesons .
the double differential rate @xmath160 with @xmath0 in a narrow interval and with a suitable @xmath9 cut on the individual leptons , however , seems to allow for a window at large values of the pair transverse momentum where the thermal yield shines out . at rhic
this is probably not feasible , while at lhc the estimated high initial temperature of the parton matter causes a stronger thermal signal .
the basic issue is here whether the combinatorical background of many other strong sources can be subtracted with sufficient accuracy . stimulating discussions with r. baier , e.v .
shuryak , m. tannenbaum , r. vogt , x.n .
wang , and g.m .
zinovjev are gratefully acknowledged .
o.p.p . thanks for the warm hospitality of the nuclear theory group in the research center rossendorf .
this work is supported in parts by bmbf grant 06dr829/1 and usffr grant 2.5.1/41 .
shuryak , phys .
b 78 ( 1978 ) 150 phenix conceptual design report , 1993 , unpublished alice technical proposal , cern / lhcc/ 95 - 71 , 1995 g. agakishev et al .
( ceres collaboration ) , phys .
lett . 75 ( 1995 ) 1272 m. gonin et al .
( na50 collaboration ) , nucl .
phys . a 610 ( 1996 ) 404c
a. drees , nucl .
phys . a 610 ( 1996 ) 536c s. gavin ,
mcgaughey , p.v .
ruuskanen , and r. vogt , phys .
c 54 ( 1996 ) 2606 e.v .
shuryak , phys .
c 55 ( 1997 ) 961 b. kmpfer , o.p .
pavlenko , and k. gallmeister , fzr-188 ( 1997 ) , phys .
lett . in print
z. lin , r. vogt , and x.n .
wang , nucl - th/9705006 , phys .
c in print r. baier , private communication , + r. baier , yu.l .
dokshitzer , a.h .
mueller , and d. schiff , to be published x.n .
wang , and z. huang , phys .
c 55 ( 1997 ) 3047 b. svetitsky , and a. uziel , phys . rev .
d 55 ( 1997 ) 2616 b. kmpfer , o.p .
pavlenko , a. peshier , m. hentschel , and g. soff , j. phys .
g 23 ( 1997 ) 2001 b. kmpfer , and o.p .
b 391 ( 1997 ) 185 a.d .
martins , w.j .
stirling , and r.g .
roberts , phys .
b 306 ( 1993 ) 145 v. emelyanov , a. khodinov , s.r . klein , r. vogt , nucl - th/9706085 m.g .
mustafa , d. pal , and d.k .
srivastava , nucl - th/9706001 , phys .
c in print , + m.g .
mustafa , d. pal , d.k .
srivastava , and m. thoma , nucl - th/9711059 r. baier , yu.l .
dokshitzer , s. peigne , and d. schiff , phys .
b 345 ( 1995 ) 277 r. baier , yu.l .
dokshitzer , a.h .
mueller , s. peigne , and d. schiff , nucl .
b 483 ( 1997 ) 291 , b 484 ( 1997 ) 265 b.g . zakharov , hep - th/9704255 r.m .
barnett et al .
( particle data group ) , phys .
d 54 ( 1996 ) 1 m. tannenbaum , heavy ion physics 4 ( 1996 ) 139 b. kmpfer , o.p .
pavlenko , a. peshier , and g. soff , phys .
c 52 ( 1995 ) 2704 p.v .
ruuskanen , in _ quark gluon plasma _ , world scientific , singapore ( 1990 ) , ( ed . ) r. hwa , p. 519
biro , e. van doorn , b. mller , m.h .
thoma , and x .-
wang , phys .
c 48 ( 1993 ) 1275 , + p. levai , b. mller and x .-
wang , phys .
c 51 ( 1995 ) 3326 k.j .
eskola , k. kajantie , and p.v .
ruuskanen , phys .
b 332 ( 1994 ) 191 , + k.j .
eskola , nucl .
a 590 ( 1995 ) 383c k.j .
eskola , b. mller , and x .-
wang , phys .
b 374 ( 1996 ) 20 l. mclerran , and t. toimela , phys .
d 31 ( 1985 ) 545 m. asakawa , c.m .
ko , and p. levai , phys .
70 ( 1993 ) 398 c. peterson , d. schlatter , i. schmitt , and p.m. zerwas , phys .
d 27 ( 1983 ) 105 m. cacciari , hep - ph/9708282 p.g .
jones et al .
( na49 collaboration ) , nucl .
phys . a 610 ( 1996 ) 188c | we analyze the opportunity to observe thermal dileptons emitted off deconfined matter resulting in heavy - ion collisions at rhic and lhc energies .
special kinematical conditions provided by the detector systems phenix and alice , and the so - called @xmath0 scaling behavior of thermal dilepton spectra are taken into account .
our considerations include energy loss effects of the fast heavy quarks in deconfined matter , which for themselves can help to identify the creation of a hot and dense parton medium .
due to a threshold like effect for decay leptons we find a window at large transverse pair momentum and fixed transverse mass where the thermal signal can exceed the background of dileptons from correlated semileptonic decays of charm and bottom mesons . -3truemm
-3truemm -13 mm 250truemm 155truemm |
systematic investigations of young stellar objects ( ysos ) in a star - forming cloud including comparative studies of theoretical predictions and observations are important for understanding the stellar / substellar formation processes . examples of well - studied star - forming clouds are the orion trapezium cluster , ngc 2024 , and the @xmath7 ophiuchus and taurus - auriga clouds
. recent observations of young clusters have been used to determine circumstellar disc frequencies , disc mass distributions , initial mass functions ( imfs ) and evolutional stages of the objects in the clusters .
it has been shown that _
jhkl _ and _ hkl _ colour - colour diagrams are particularly effective in identifying the presence of circumstellar discs ( e.g. @xcite ; @xcite ; @xcite ; @xcite ) .
the spectral indices @xcite or the slopes of observed spectral energy distributions ( seds ) of ysos from near- to mid - infrared wavelengths are used to classify seds , and the classification scheme is often related to the evolutionary stages of ysos ( e.g. @xcite ; @xcite ) .
the distribution of circumstellar disk masses in young clusters can be measured from millimetre continuum emission ( e.g. @xcite for ngc 2024 ) .
determining imfs of young clusters requires evolutionary models ( e.g. @xcite ; @xcite ) and infrared spectroscopy for spectral type identifications ( e.g. @xcite for the @xmath7 ophiuchus cloud ) .
@xcite presented results from a very large three - dimensional ( 3-d ) smoothed particle hydrodynamics ( sph ) simulation of the collapse and fragmentation of a 50@xmath8 turbulent molecular cloud to form a stellar cluster .
the calculation resolved circumstellar discs down to @xmath9 au in radius and binary stars as close as 1 au . although some observational predictions , such as the imf and binary fraction , may be gleaned directly from a hydrodynamical simulation of stellar cluster formation , the principal observable characteristics ( optical , near - ir , ir , and sub - millimetre images and spectra ) require further detailed radiative transfer modelling .
the density distribution of such hydrodynamical calculations is very complicated , and the corresponding radiative transfer must be also performed in full 3-d .
there are two basic approaches to 3-d radiative transfer problems : grid based methods ( e.g. finite differencing , short- and long- characteristic methods ) , and particle ( photon ) based methods , i.e. monte carlo .
examples of the first kind are @xcite , @xcite and @xcite .
those of the second kind include @xcite , @xcite , @xcite , @xcite and @xcite .
the advantages of the second approach are , for example , the flexibility to treat a complex density distribution and a complex scattering function .
readers are referred to @xcite and @xcite for more extensive discussion on the advantages and disadvantages of these two different methods .
the temperature of the interstellar and circumstellar dust in the cluster must be calculated in order to determine the source function of the dust emission .
the radiative equilibrium temperature of the dust particles can be found using the monte carlo method ( e.g. @xcite ; @xcite ; @xcite ; and @xcite ) .
the technique used by @xcite takes into account the fractional photon absorptions between two events of a ` photon packet ' ; hence , it works well even in the limit of low opacity .
@xcite used the immediate re - emission technique , in which radiative equilibrium is forced at each interaction with the dust .
a photon is re - emitted immediately after an absorption event using a product of the dust opacity and the difference between the planck function with a current temperature and that with a new temperature corresponding to the radiative equilibrium of the dust that absorbed the photon . unlike the method of @xcite ,
this does not require a temperature iteration if the opacity is independent of temperature .
alternatively , @xcite used the ray - splitting method showing its effectiveness for both low and high optical depth media . here
we aim to simultaneously resolve the dust on parsec and sub - stellar - radius spatial scales , whilst including multiple radiation sources . to overcome the resolution problem ,
we have implemented an adaptive mesh refinement ( amr ) scheme in the grid production process of the torus radiative transfer @xcite .
see also @xcite , @xcite and @xcite for a similar gridding scheme in a radiative transfer problem .
the method described by @xcite is used to compute dust temperatures in our models .
the objectives of this paper are : 1.to compute observable quantities by solving the radiative transfer problem using a complex density distribution from the sph calculation by @xcite ; 2.to analyse the predicted observational properties of the cluster generated in the simulation of @xcite at a distance of 140 pc which corresponds to the distance of nearby star - forming regions such as taurus - auriga and the @xmath10ophiuchus cloud ( e.g. @xcite ) . in section 2 , we describe the details of our models .
the results of the model calculations are given in section 3 .
the conclusions are summarised in section 4 .
we calculate the main observable quantities of the low mass ( 50 @xmath8 ) cluster formation simulation by @xcite .
there are four basic steps in this process : 1 .
create the amr grid from the sph particle positions and the density at each particle .
the density is assigned to the grid cells during the grid construction .
2 . calculate the luminosity and the effective temperature of each light source ( in the sph calculation ) based on its mass using an evolutional model of young stars ( @xcite ; @xcite ) .
3 . compute the temperature of the dust in the cluster using the monte carlo radiative equilibrium model of @xcite .
4 . compute the seds and the images using the monte carlo radiative transfer code , torus .
bate et al .
( @xcite , @xcite , @xcite ) presented a numerical sph simulation of star cluster formation resolving the fragmentation process down to the opacity limit ( @xmath110.005m@xmath12 ) .
the initial conditions consist of a large - scale , turbulent molecular cloud with a mass of 50 @xmath8 and a diameter of 0.375 pc ( 77400 au ) .
the initial temperature of the cloud was 10 k , and its corresponding mean thermal jeans mass was 1 @xmath8 ( i.e. the cloud contained 50 thermal jeans masses ) .
the free - fall time of the cloud was @xmath13 .
similar to the method used by @xcite , they imposed an initial supersonic turbulent velocity field on the cloud by generating a divergence - free random gaussian velocity field with a power spectrum @xmath14 where @xmath15 is the wave number .
this was chosen to reproduce the observed larson scaling relations @xcite for molecular clouds .
the total number of particles used in the simulations was @xmath16 , making it one of the largest sph calculations ever performed .
approximately 95000 cpu hours on the sgi origin 3800 of the united kingdom astrophysical fluids facility ( ukaff ) were spent on the calculation .
we use the output from the final time step of the sph calculation as the input for the radiative transfer code .
the time of the data dump is 0.27 myr , at which the cluster contains 50 point masses ( stars and brown dwarfs ) .
each sph data point contains the position ( @xmath17 , @xmath18 , @xmath19 ) , the velocity components ( @xmath20 , @xmath21 , @xmath22 ) and the density ( @xmath7 ) .
the velocity information is not used in our calculation since we are only interested in continuum emission , but these data may be used in future calculations of molecular line emission .
the total masses contained in the stars and the molecular gas are @xmath23 and @xmath24 respectively .
the sph data provides the masses of the stellar objects ( 23 stars and 27 brown dwarfs ) , ranging from 0.005 to 0.731 @xmath8 ( see table [ tab : catalog ] ) . since the monte carlo radiative transfer code requires the luminosity ( @xmath25 ) , the effective temperature ( @xmath26 ) and the radius ( @xmath27 ) of each star , we computed them indirectly from evolutionary models . for a given age and mass of each star and brown dwarf
, the luminosity and the temperature are interpolated from the 1998 updated version of data by @xcite and @xcite available on their website ( http://www.mporzio.astro.it/~dantona/prems.html ) .
although the actual age of the stars and brown dwarfs in the sph data of @xcite ranges from @xmath112000 to @xmath1170000 years old , we make the pragmatic assumption that all the objects are 0.25 myr old because we believe that at young ages the stellar radii are overestimated ( c.f . , @xcite ) , leading to unrealistically high luminosities .
adopting models with ages of 0.25myr provides more plausible radii ( and therefore luminosities ) , while the change in temperature over this timespan is modest ( @xcite ) .
we take the spectral energy distribution of the individual stars to be blackbodies .
although we recognise that the spectra of the objects ( particularly at the lowest masses ) may be significantly structured due to molecular opacity , the vast majority of the stellar flux is reprocessed by the circumstellar dust and the precise form of the input spectrum is unimportant . | we present three - dimensional monte carlo radiative transfer models of a very young ( @xmath0 years old ) low mass ( @xmath1 ) stellar cluster containing 23 stars and 27 brown dwarfs .
the models use the density and the stellar mass distributions from the large - scale smoothed particle hydrodynamics ( sph ) simulation of the formation of a low - mass stellar cluster by bate , bonnell and bromm . using adaptive mesh refinement ,
the sph density is mapped to the radiative transfer grid without loss of resolution .
the temperature of the ism and the circumstellar dust is computed using lucy s monte carlo radiative equilibrium algorithm . based on this temperature ,
we compute the spectral energy distributions of the whole cluster and the individual objects .
we also compute simulated far - infrared spitzer space telescope ( _ sst _ ) images ( 24 , 70 , and 160 @xmath2 bands ) and construct colour - colour diagrams ( near - infrared _ hkl _ and _ sst _ mid - infrared bands ) .
the presence of accretion discs around the light sources influences the morphology of the dust temperature structure on a large scale ( up to a several @xmath3 au ) .
a considerable fraction of the interstellar dust is underheated compared to a model without the accretion discs because the radiation from the light sources is blocked / shadowed by the discs .
the spectral energy distribution ( sed ) of the model cluster with accretion discs shows excess emission at @xmath4 330 @xmath5 and @xmath6 , compared to that without accretion discs . while the former is caused by the warm dust present in the discs , the latter is cause by the presence of the underheated ( shadowed ) dust .
our model with accretion discs around each object shows a similar distribution of spectral index ( 2.220 @xmath2 ) values ( i.e. class 0iii sources ) as seen in the @xmath7 ophiuchus cloud .
we confirm that the best diagnostics for identifying objects with accretion discs are mid - infrared ( @xmath4 310 @xmath5 ) colours ( e.g. _ sst _ irac bands ) rather than _ hkl _ colours . [ firstpage ] radiative transfer stars : formation circumstellar matter infrared : stars brown dwarfs accretion , accretion discs . |
18ifxundefined [ 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 @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevb.81.224418 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.067206 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.104.097203 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop `` '' , ( , , ) @noop * * , ( ) @noop _ _ ( , , ) @noop * * , ( )
this work was partially supported by anr-07-nano-034 _ dynawall _ and anr-08-blan-0199 _ micromanip_. vu was supported by the research programmes of gaav ( project no .
kan400100701 ) and european regional development fund ( ceitec - cz.1.05/1.1.00/02.0068 ) . | we investigate numerically the transverse versus vortex phase diagram of head - to - head domain walls in co / cu / py spin valve nano - stripes ( py : permalloy ) , in which the co layer is mostly single domain while the py layer hosts the domain wall .
the range of stability of the transverse wall is shifted towards larger thickness compared to single py layers , due to a magnetostatic screening effect between the two layers .
an approached analytical scaling law is derived , which reproduces faithfully the phase diagram .
0.5 in 0.5 in the fundamental study of magnetic domain walls ( dws ) has found a playground in lithographically - defined nano - stripes . in the range of widths from a few tens to a few hundreds of nanometers their complexity
is dramatically reduced with respect to extended thin films , while still retaining a few internal degrees of freedom leaving a rich physics . for stripes with in - plane
magnetization , head - to - head dws are of transverse ( tw ) or vortex ( vw ) typeb@xcite , and may be characterized by a chirality , asymmetry and/or polarity .
their propagation under the stimulus of a magnetic field or a current of spin - polarized electrons is intrinsically precessional and reveals effects such as dw inertia@xcite and a so - called walker limit beyond which periodic dw transformations occur@xcite . due to their small size and fast dynamics , magnetic domain walls ( dws )
are promising candidates in the area of information processing and storage@xcite .
most reports so far have considered dws in a single layer of magnetically - soft materials , such as permalloy ( py ) or cofeb alloys .
magnetic trilayers f1/nm / f2 ( with @xmath0 a ferromagnet , and nm a non - magnetic material ) also deserve attention as they play a key role in spintronics , especially in devices using giant magneto - resistance . these stacking are called spin valve for a metal spacer layer , and pseudo spin valve or magnetic tunnel junction for an insulator spacer layer .
f1 is a soft layer in which the dw moves , while f2 is a reference layer that is supposed to remain uniformly magnetized .
unusually - large current - driven dw mobilities with speed exceeding have been reported in such stacks@xcite .
tentative explanations point at either out - of - plane spin accumulation and torque@xcite , oersted field@xcite or magnetostatic coupling between the layers that may modify the statics or dynamics of the dw@xcite .
while the role of magnetostatic coupling is well established in thin films with the formation of quasi - walls in the supposedly uniformly - magnetized layer@xcite , it has been scarcely addressed@xcite in stripes . in this letter
we determine the phase diagram of head - to - head dws in spin valves with in - plane magnetization ( tw versus vw ) , combining numerical simulation and analytical modeling . for this purpose
, we consider co / cu / py nano - stripes with various widths @xmath1 and thickness of the individual layers @xmath2 , @xmath3 and @xmath4 . the dw is located in the py layer , while the co layer is initialized with uniform magnetization .
we predict that the stability of tws is enhanced towards larger py thickness compared to single layers@xcite , due to the screening in the co layer of the stray field arising from the dw in py .
we used finite - differences micromagnetic codes , both oommf@xcite and our home - made gl - fft code@xcite .
the cell size in all simulations presented here is @xmath5{nm^3}$ ] ; fo significant difference is found for a cell size @xmath6{nm^3}$ ] .
effects of the finite length of the stripe in the simulations were reduced by either fixing at both ends the magnetic moments parallel to the stripe direction , or by compensating the surface magnetic charges at both ends , however with no constraint on magnetization .
the length @xmath7 of the stripes was set equal or larger than about @xmath8 , for which finite - size effects are identically small under either of the above procedures .
material parameters are @xmath9{t}$ ] , @xmath10{pj / m}$ ] , @xmath11{t}$ ] , @xmath12{pj / m}$ ] , and zero magnetocrystalline anisotropy in both layers .
no exchange coupling between py and co is considered either .
different initial conditions were chosen in py to set a tw or vw in the py layer , while magnetization in co is set uniform along the stripe direction ; both layers are then set free to evolve during the simulation .
we considered head - to - head dws ( two domains facing each - other , see a ) with no loss of generality , as tail - to - tail dws are equivalent under time - reversal symmetry ( @xmath13 ) . like for single layers , tws and vws coexist as ( meta)stable states in a large range of geometrical parameters .
the iso - energy curve separating the areas where either a tw or a vw is lowest in energy , was determined for each fixed @xmath1 through a parabolic interpolation of the energy difference for three different values of @xmath4 .
we checked that this procedure yields a phase diagram of the single py layer in very close agreement with existing reports@xcite .
( a ) cross - sectional sketch of a head - to - head domain wall and the associated charge in a spin valve .
plane views of a vortex wall and transverse wall in a @xmath14-wide ( b ) single layer of py(@xmath15 ) and ( c ) a co(5 nm)/cu(5 nm)/py(5 nm ) spin valve .
( d ) thickness - integrated maps of magnetization from ( c ) , see .
the red ( blue ) color indicates the positive ( negative ) transverse component of magnetization , normalized to the magnetization of the layer under focus in ( b ) and ( c ) , and to half the magnetization of co in ( d).,width=328 ] we first discuss qualitatively the difference between a dw in a single layer and in a spin valve .
the head - to - head dw in py is associated with a total magnetic charge @xmath16a .
while vws are similar in both systems , clear differences are found for tws . in a single layer , tws
are symmetric for small thickness and width .
going towards a larger thickness or width , the tw becomes asymmetric , reducing the magnetostatic energy by spreading the magnetic charges@xciteb ; this corresponds to the onset of zig - zag walls in continuous films .
the transition from symmetric to asymmetric tw is of second order ( continuous ) , so that for a given geometry only one type of tw exists , either symmetric or asymmetric .
asymmetric tws are found for a large range of values of width and thickness , although they are the state of lowest energy ( over the vw ) only in a narrow range of values . in a spin valve
the co layer deviates from uniform magnetization due to the stray field arising from the dw in py , which in turn creates a stray field acting on the py layer .
this magnetic screening effect lowers the energy compared to a single layer , as already pointed out@xcite .
the stray field arising from the dw in py is parallel or antiparallel to the initial direction of magnetization of co , depending whether the left or right side of the dwa is considered .
thus , the magnetic configuration of co is expected to be asymmetric , and so will be the stray field arising from co and acting on the py layer .
a tw in a spin valve is therefore expected to be always asymmetric by nature , due to the unidirectional magnetization in the co layer .
c shows a vw and a tw in a spin valve , with the same stripe width and py thickness as in b. the magnetic screening in co is clear from the non - uniformity of magnetization .
as expected from the symmetry arguments given above the tw is asymmetric , to a larger extent than in the single layer case@xcite .
both the vw and tw have a larger width , associated with the emergence of tails .
indeed , as a dw profile results from the balance of magnetostatic with exchange energy , the decrease of magnetostatic energy allows to reduce exchange energy via the increase of the dw width .
this is consistent with simple analytical models@xcite derived for charged ( head - to - head ) domain walls in extended spin - valve thin films .
the flux closure taking place between py and co can also be illustrated by the map of magnetization @xmath17 integrated over the two layers along their normal d : @xmath18 this map shows similarities with a vw , which highlights the principle of reduction of the total energy : the flux is better closed than in a single layer tw , while avoiding the cost associated with the vortex core in a single layer vw .
in this view the dw tails of the individual layers are also absent .
phase diagram of head - to - head domain walls in stripes , with the boundary between symmetric and asymmetric tws ( @xmath19 ) , and tw and vw ( @xmath20 ) in a single layer , and asymmetric tw and vw in a spin valve ( @xmath21 ) . from bottom to top , the three lines are : a @xmath22 fit to the tw - vw data for a single layer , a thickness translation of the former to fit the tw - vw data for a spin valve , and another thickness translation with @xmath23 , see and text for details.,width=321 ] as the share of magnetostatic energy is larger in a tw than in a vw@xcite , the inter - layer closure of flux is lowers the energy of the tw more than that of the vw .
this suggests that the tw should be the ground state in a larger range of geometrical parameters in a spin valve than in a single layer .
this is confirmed by a set of simulations with varying @xmath4 and @xmath1 , illustrated in for @xmath24 .
notice that in a single layer tws with opposite asymmetries ( left or right ) are degenerate in energy , while this degeneracy is lifted in a spin valve , due to the unidirectional magnetization in the underlying co. in the phase diagram we considered only the tw with the lowest energy .
we now derive a simple analytical model to grasp the main features of this diagram .
mcmichael and donahue already noticed that the iso - energy line in a single layer follows the power law @xmath25 , with @xmath26 the dipolar exchange length , @xmath27 the dipolar constant , and @xmath28 a constant .
this law can be derived qualitatively by balancing the energies at play in each domain wall@xcite .
these authors argued that the energy of the tw can be estimated from the lateral demagnetizing coefficient scaling like @xmath29 , and the volume of the tw of the order of @xmath30 ( thus @xmath31 as a whole ) , while that of the vw is related to the energy of the vortex core , scaling with @xmath32 .
the exchange energy in the @xmath33 nel sub - walls can be ignored as the total length of these walls is identical in a tw and a vw .
the numerical value @xmath34 must be provided by simulations@xcite .
this scaling model can be adapted to a spin valve , however the above arguments must first be discussed in more detail .
it shall first be noted that in any head - to - head dw , either tw or vw , most of the energy is of magnetostatic origin and related to the head - to - head charges , so that the above argument must be refined .
simulation and magnetic force microscopy of a head - to - head dw in a single layer show that the total charge @xmath35 is nearly uniformly spread over the area of the dw@xcite , which is @xmath36 for a tw , and @xmath37 for a vw .
the associated densities of surface charges are then @xmath38 and @xmath39 .
the resulting volume density of stray field energy scales with @xmath40 and extends over a typical distance @xmath1 above and below the stripe .
this results in an excess of magnetostatic energy in the tw over the vw of @xmath41 .
we now consider a spin valve . in the limit @xmath42 , which is suitable for @xmath43 and the boundaries of the phase diagram of practical interest@xcite
, the deformation in co reaches its maximum while not being enough for a full screening of the charges arising from the dw in py .
the above scaling laws can be rewritten accordingly and the following law is derived again with some approximations ( see below ) , making also use of the partial flux closure in spin valves to refine the lateral demagnetizing coefficient@xcite : @xmath44 this suggests that the phase diagram for a spin valve is based on that of a single layer , shifted towards higher thicknesses by a value @xmath45 .
this is in very good agreement with the results of numerical simulations .
quantitatively , the analytical formula for @xmath45 is not simple , implying both power and logarithmic functions .
however , thanks to the slow variation of the latter , @xmath45 can be approximated faithfully with @xmath46 .
for @xmath24 the analytical model thus yields @xmath47 , while @xmath45 determined from simulations is @xmath48 .
this discrepancy is linked with the magnetic screening in the co layer being partial .
this can be understood as in the scaling model we neglect the magnetostatic energy stored in the volume of the cu spacer , locus of a magnetic field arising between the charges of opposite sign in py and in co. shows @xmath45 resulting from micromagnetic simulations as a function of @xmath3 for @xmath49 and @xmath43 . for large @xmath3
the problem tends towards the case of a single layer because the screening is no more effective .
the decay is close to exponential , with a fit providing the extrapolation @xmath50 .
this figure is now in excellent agreement with the analytical model , which suggests the following empirical law : @xmath51 where @xmath52 is derived from .
additional micromagnetic simulations show that @xmath53 increases with @xmath1 , @xmath54 and @xmath55 for @xmath56 and @xmath57 , respectively .
this trend can be understood as the initial stray field ( to be reduced through screening ) arising from the dw extends outside the stripe over a volume scaling with @xmath58 , while the magnetic field arising between the charges of opposite sign in py and in co ( a side cost of the screening effect ) applies in a volume scaling with @xmath59 .
thus , the relative gain of energy through screening decreases with decreasing the stripe width . then , can be applied for other co thicknesses , or more generally to other couples of magnetic materials if substituting the proper spontaneous magnetization values in , and fitting the decay of @xmath45 like in .
one must finally keep in mind that for @xmath60 a very effective screening of the head - to - head dw charge is possible , so that deviations from are expected . in that case , the energy of the tw is drastically reduced .
the iso - energy line becomes very flat beyond this limit , an effect that starts to be visible for the largest widths in .
thickness shift of the iso - energy line in a spin valve compared to a single layer , for @xmath49 and @xmath43.,width=276 ] to conclude , we have investigated with numerical simulation and analytical scaling laws the phase diagram of head - to - head domain walls in f1/nm / f2 spin valves , with a domain wall in one layer and no domain wall in the other layer .
we showed that the range of stability of transverse versus vortex walls is enhanced due to a magnetic screening effect .
the iso - energy line in spin valves is translated towards a larger thickness with respect to single layers , by a value decreasing approximately exponentially with the spacer thickness .
this enhanced stability and larger width provide a magnetostatic contribution to the high mobility of dws in trilayers , provided that the damping in co is not too large@xcite .
effects on the domain wall inertia and automotion@xcite are also expected . |
one of the most intriguing questions in the study of the formation of planets , and of terrestrial planets in particular , is how water is transported to their surfaces , and whether or not water is a common ingredient during their formation and early evolution .
spatially and spectrally resolved observations of water in extrasolar planetary systems and protoplanetary disks would be instrumental in resolving these questions . unfortunately
, detecting water in extrasolar planetary systems remains difficult @xcite and water in protoplanetary disks has been elusive , with only a few definitive ice @xcite and vapor @xcite detections .
recently , however , the spitzer infrared spectrograph ( irs ) has begun to reveal warm water vapor emission lines from disks around classical t tauri stars ( ctts , carr & najita 2008 ) . in this letter , we present high - resolution spectroscopy of numerous water vapor emission lines from protoplanetary disks around two ctts : dr tau and as 205a .
spectra were obtained using both the spitzer - irs and nirspec @xcite at the keck ii telescope .
accretion rates onto these stars are high , which causes the inner disks to be significantly hotter than in more quiescent systems .
dr tau is known to have variable mass accretion , with measured rates between @xmath7 and @xmath8 @xcite . as 205a has an accretion rate of @xmath9 and is the primary component of a 1.3 triple system , with the secondary being a spectroscopic binary @xcite .
however , it is likely that the physical separation of as 205a and b is significantly larger than the projected separation and that , therefore , as 205b does not affect the ( inner ) disk of as 205a @xcite .
the spitzer irs short - high ( sh ) module spectrum of as 205a was taken as part of the `` cores to disks '' ( c2d ) legacy program @xcite , while the spectrum of dr tau , observed as part of the guaranteed time observations @xcite , was obtained from the spitzer archive .
the 2d basic calibrated data images were reduced using the c2d pipeline as described in @xcite . to confirm the detections as well as measure the gas emission line profiles
, we obtained high resolution ( @xmath10=@xmath11 ) l - band spectra using nirspec in a single grating setting , of which two orders , centered at 2.92@xmath0 m and 3.16@xmath0 m , respectively , have identifiable h@xmath1o and oh emission features .
we also collected co @xmath4=1@xmath50 m - band spectra ( @xmath65@xmath0 m ) as part of an on - going survey ( c.f .
@xcite , @xcite ) .
the data were reduced in a standard way .
division by the spectra of standard stars ( hr 1620 ( a7v ) for dr tau ; hr 5993 ( b1v ) for as 205a ) was used to correct for telluric absorption , which can be poor when the absorption is high .
thus , regions of the spectrum with @xmath1270% atmospheric transmission were removed .
fluxes were calibrated using the photometry and spectral types of the standards , and are generally accurate to 20% .
the nirspec lines are narrow ( fwhm@xmath13 35 km s@xmath14 ) , spatially unresolved , and centered at the stellar @xmath15 @xcite to within our uncertainty ( a few km s@xmath14 ) all consistent with disk emission . because the disks are optically thick and heated from above , the line emission likely arises from the upper disk atmosphere @xcite .
the irs spectra of as 205a and dr tau show a large number of water emission lines ( see figure [ irs_h2o ] ) , the vast majority pure rotational transitions with quantum numbers in the range 2050 .
the @xmath16=600 resolution of the sh module results in unresolved , and significantly blended , lines .
consequently , the identification with water is achieved via a simple , isothermal lte ring model computed using the hitran 2004 database @xcite .
the ring model consists of a single temperature ( @xmath17 ) and column density ( @xmath18 ) gas in keplerian rotation .
model parameters include @xmath17 , @xmath18 , solid angle ( @xmath19 ) , stellar mass ( @xmath20 ) , inclination ( @xmath21 ) and characteristic radius ( @xmath22 ) .
in addition , the local line broadening is assumed to be gaussian , with specified width ( @xmath23 ) . @xmath20 and
@xmath21 were fixed to values taken from the literature @xcite .
the spectrally unresolved irs data can not completely constrain the remaining model parameters , but do require high @xmath17 ( @xmath24 k ) and @xmath18 ( @xmath25 @xmath26 for @xmath23@xmath62 km s@xmath14 ) .
figure [ irs_h2o ] compares the irs spectra with a model in which @xmath17 and @xmath18 are constrained using the nirspec data described below , but where @xmath19 has been increased by a factor of 2 .
the addition of co@xmath1 ( and , for as 205a , oh ) improves our fits , as shown by the as 205a model in fig .
[ irs_h2o ] . with
@xmath17 and @xmath19 fixed , we find h@xmath1o : co@xmath1=40 for dr tau and 20 for as 205 .
detections of these and other molecules will be discussed in detail elsewhere . while tropospheric water vapor blocks significant portions of incoming radiation near 3@xmath0 m , there are a handful of rovibrational lines that are optically thin in the atmosphere but readily excited at temperatures typical of the inner @xmath61 au of circumstellar disks .
this is illustrated in figure [ nirspec ] , in which spectral regions with @xmath1270% atmospheric transmission have been removed , yet many prominent ( line - to - continuum@xmath6510% ) h@xmath1o emission features remain .
these are primarily @xmath27 lines , with upper level energies of @xmath65,00010,000 k. a few weaker ( line - to - continuum@xmath61% ) features are nominally detected from 3.133.18 @xmath0 m and are used as model constraints .
we also detect the p4.5 ( 2.9344@xmath0 m ; @xmath28@xmath6@xmath29 k ) and p9.5 ( 3.1788@xmath0 m ; @xmath28@xmath6@xmath30 k ) oh doublets .
the m - band spectra include portions of the @xmath2co/@xmath3co @xmath4=1@xmath50 and @xmath2co v=2@xmath51 p- and r - branches ( figure [ rotfit ] ) .
we fit all spectra with the ring models described in [ sec : spitzer ] .
because the co lines span a large range of excitation energies , across which they transition from optically thick to optically thin , @xmath18 , @xmath17 , and @xmath19 can be determined uniquely .
also , a constraint on @xmath31 is provided by the @xmath3co lines ( assuming a @xmath2co:@xmath3co ratio , here fixed at 77 , c.f .
thus , we began by fitting the co emission by comparison with an excitation diagram of the @xmath4=1@xmath50 transitions ( fig .
[ rotfit ] ) , as described in @xcite . for our nominal h@xmath1o and
oh fits , we then fixed @xmath19 and @xmath17 to the co - derived values to determine @xmath18 a reasonable , though not ideal , assumption , given the similarity in line shapes ( see fig .
[ rotfit ] ) . our nominal ring model is shown in figure [ nirspec ] , with best - fit parameters in table [ fit_parameters ] , and uncertainties of order @xmath32@xmath6@xmath33 , @xmath34@xmath6@xmath35 and @xmath36@xmath6@xmath37 .
we have fixed @xmath38 au to roughly match the emission profiles since the lineshape does not affect the integrated flux ; we discuss lineshapes below .
@xmath19 corresponds to areas of 0.2 and 0.4 au@xmath39 for dr tau and as 205 , respectively , at distances of 140 and 120 pc .
uncertainties in the nominal h@xmath1o and oh fits are linked to co uncertainties , but if @xmath17 is instead allowed to vary , the h@xmath1o spectra are consistent with @xmath17@xmath69001200/9001100 k for dr tau / as 205a and ranges in @xmath18 of about 1 order of magnitude . for oh
, we can only constrain @xmath17 900 k. the model results depend crucially on the assumed local line width ( @xmath23 ) , which is unknown . for our nominal model ,
we have adopted @xmath40 km s@xmath14 the sound speed for h@xmath1 at 1000 k. however , we have also tested values between @xmath60.3 and 10 km s@xmath14 .
we find that the opacity needs to remain similar from fit to fit , so that a change in @xmath41 requires a proportional change in @xmath18 . to maintain the flux level ,
the total number of molecules must remain similar , so an increase in @xmath23 implies a decrease in @xmath19 .
accordingly , the co : h@xmath1o ratio changes by no more than a factor of @xmath6 2 between models .
( if @xmath42 is not fixed , however , this ratio may change ) .
in addition , @xmath17 is only weakly affected by @xmath23 , since it is set by the overall shape of the spectrum , rather than by the absolute flux levels .
nirspec spectrally resolves the molecular emission , allowing an investigation of the disk gas kinematics . in fig .
[ rotfit ] we compare the unblended h@xmath1o transition at 2.931 @xmath0 m with a co line composite constructed from an average of all @xmath4=1@xmath50 lines . in general , keplerian disks should produce double - peaked emission profiles , with peak separation set by the inner and outer emission radii ( @xmath43 , @xmath44 ) , as well as the temperature gradient @xcite .
the co and h@xmath1o line profiles , however , are more gaussian , or even lorentzian in shape , with relatively wide wings and a narrow peak .
the narrow core could be produced by a ring with a very large @xmath43 ( @xmath4510 au ) if the peak - to - peak separation is equal to the nirspec resolution ( given the @xmath20 and @xmath21 in table [ fit_parameters ] ) .
however , this is difficult to reconcile with the high @xmath17 derived from the model fits and the observed line wings .
another way to produce a single peak would be to have the line emission extend to significant radii ( @xmath44 of order 7 au , for a temperature profile that is constant with radius ) , such that the contribution from low - velocity portions of the disk infills the center of the line profile .
however , this model produces too much flux to match observations .
additionally , passive disk models predict a steep temperature decline with radius .
a more promising solution may be that some flux is produced via fluorescence , in which molecules in the flared outer disk ( typically out to @xmath65 au ) intercept and re - radiate inner disk ir continuum , thus filling in the line profiles at low velocity @xcite . with canonical disk flaring and temperature structure , we find that a resonance fluorescence model with 100% scattering efficiency can indeed produce fluxes comparable to those of thermal emission ; a more thorough investigation is left as future work
. in any case , the amount of flux in the line profile wings is consistent with emission arising in a keplerian disk with a small inner radius . with this assumption
, the line wing velocity can be used to estimate @xmath43 . because turbulent motion and disk structure can both have non - negligible effects on the line shapes , we adopt here the velocity , @xmath46 , at 2.2 times the half width at half maximum a value found to be appropriate for the moderate - inclination disk around gm aur @xcite . to compute @xmath46 we use gaussian fits to the co and h@xmath1o lines discussed above , as well as the average of two - gaussian fits to the oh emission doublets , deconvolved with the nirspec instrument response function .
if turbulent velocities are low , radii associated with @xmath46 will be overestimates , and so we also compute @xmath4 at 3@xmath23 , which should represent a lower limit to @xmath43 ( see table [ fit_parameters ] ) .
we have convincingly detected h@xmath1o emission in the 1020 @xmath0 m region with spitzer - irs as well as h@xmath1o and oh emission near 3 @xmath0 m with nirspec , arising from the disk atmospheres surrounding two ctts . by combining these data with constraints from spectrally resolved co @xmath4=1@xmath50 emission , we find that the excitation temperatures are typical of terrestrial planet forming regions ( @xmath61000 k ) . additionally , line wing velocities imply inner emission radii no larger than 1 au for all molecules . therefore , further observations and analyses of these species may provide a new window into terrestrial planet forming regions .
observations of h@xmath1o in disks may also provide constraints on disk evolution and water transport . in a disk in which water vapor transport
is controlled by diffusion , the presence of a condensation front ( ` snow line ' ) can potentially dry out an inner disk in as little as @xmath47 yr , for typical disk viscosities @xcite .
the snow line for the early solar system is estimated to be near 3 au , but would be further out for these high accretion rate stars @xcite .
therefore , the presence of significant water vapor inside 1 au may be evidence for inward radial migration , or upward mixing , of icy solids , followed by evaporation . with both solid migration and diffusion controlling water transport ,
water vapor concentrations become sensitive to an array of parameters , including disk viscosity and planetesimal growth rates , and water concentrations may trace disk evolution @xcite .
the detection of strong oh emission is also interesting , for oh is known to be an important ingredient to the chemistry in the inner regions of protoplanetary disks .
it also acts as a coolant near the disk surface ( e.g. * ? ? ?
* ) , and photolysis of oh to o(@xmath48d ) has been invoked to explain the observed strengths of the 6300 o i line @xcite .
detections of oh in these and other disks will be discussed in greater detail in mandell et al .
, in prep .
an understanding of the local line broadening in disks will be crucial for obtaining accurate absolute column densities of molecular species , as smaller local line widths require lower column densities and vice versa .
nevertheless , we find that molecular ratios are more robust , and remain similar when utilizing different disk models with the nominal fits having co : h@xmath1o@xmath61015 and h@xmath1o : oh@xmath634 , provided @xmath49 .
stronger constraints on both @xmath17 and @xmath18 will require high line - to - continuum observations across a wider wavelength range , but the available data reveal that h@xmath1o : co is similar to , or slightly higher than that in the dense clouds out of which disks form @xcite .
vertically integrated h@xmath1o : oh ratios have been predicted to be much higher ( @xmath50 ) near 1 au @xcite , but significant oh abundances are expected near the c@xmath51/c / co transition zone at large scale heights @xcite . also , it is important to keep in mind that disk transport is not included in current chemical models .
our analysis of the l- and m - band spectra thus far has assumed thermal excitation of the vibrationally excited states , but this need not be the case
. @xmath52=1 emission can also be driven by the stellar uv output or by the absorption of ir photons from the star or disk , followed by fluorescence @xcite . for oh
, vibrationally excited emission can also result from the ( photo)dissociation of h@xmath1o followed by prompt emission @xcite .
in fact , fluorescence may be necessary to explain observed line profile shapes ( see [ sec : nirspec ] ) , and could explain the surprisingly strong oh emission .
a better understanding of these processes may therefore prove essential for relating observed line fluxes to disk column densities .
if emission in the line wings truly represents emission from the disk inner radius , relatively large inner emission radii are implied for co , oh , and h@xmath1o . both dr tau and as 205a
have had their dust inner rim sizes measured with the keck interferometer . at @xmath53 and @xmath54 au , respectively , they are consistent with the location of dust sublimation @xcite , while the inner emission radii we derive are at least 3 times larger .
perhaps these high accretion - rate stars can stir up the inner disk enough to erase the vertical temperature gradients necessary for formation of emission lines .
another possibility may be that replenishment rates can not keep up with photo - dissociation at the disk surface .
the ease with which h@xmath1o and oh are now detected in the terrestrial planet - forming zones of certain disks holds the promise of more extensive and detailed studies .
although the ground based echelle observations presented here probe only the inner disk atmospheres , future high spectral resolution follow - up studies of spitzer - detected water emission from protoplanetary disks will provide constraints at a variety of disk radii and vertical depths . additionally , spectro - astrometric observations are being used to directly constrain emitting location and structure of the molecular gas @xcite .
eventually , a suite of constraints combined with non - lte radiative transfer models will allow us to rigorously address such far - reaching questions as : how does the water vapor abundance vary within and between disks ?
what constraints can be put on models for the chemistry and transport of volatiles in disks , and what implications will these have for the early evolution of the solar system ?
acke , b. , van den ancker , m. e. , & dullemond , c. p. 2005 , , 436 , 209 akeson , r. l. , et al .
2005 , , 622 , 440 andrews , s. m. , & williams , j. p. 2007
, , 659 , 705 blake , g. a. , & boogert , a. c. a. 2004 , , 606 , l73 bonev , b. p. , mumma , m. j. , disanti , m. a. , dello russo , n. , magee - sauer , k. , ellis , r. s. , & stark , d. p. 2006
, , 653 , 774 boogert , a.c.a .
2004 , , 154 , 359 carr , j. & najita , j. 2008 , science , in press carr , j. s. , tokunaga , a. t. , & najita , j. 2004 , , 603 , 213 ciesla , f. j. , & cuzzi , j. n. 2006 , icarus , 181 , 178 clarke , c. j. & bouvier , j. 2000 , mnras , 319 , 457 dello russo , n. , disanti , m. , magee - sauer , k. , gibb , e. , mumma , m. j. , barber , r. , & tennyson , j. 2004 , icarus , 168 , 186 dullemond , c. p. , hollenbach , d. , kamp , i. , & dalessio , p. 2007 , protostars and planets v , 555 ehrenreich , d. , hbrard , g. , lecavelier des etangs , a. , sing , d. k. , dsert , j
.- m . , bouchy , f. , ferlet , r. , & vidal - madjar , a. 2007 , , 668 , l179 eisner , j. a. , hillenbrand , l. a. , white , r. j. , akeson , r. l. , & sargent , a. i. 2005 , , 623 , 952 evans , n. j. , ii , et al.2003 , , 115 , 965 gullbring , e. , calvet , n. , muzerolle , j. , & hartmann , l. 2000 , , 544 , 927 herbig , g.h . & bell , k.r .
1988 , lick obs . bull .
1111 horne , k. , & marsh , t. r. 1986 , , 218 , 761 houck , j. r. , et al.2004 , , 5487 , 62 johns - krull , c.m .
& gafford , a.d .
2002 , , 573 , 685 kamp , i. & dullemond , c.p .
2004 , , 615 , 991 kessler - silacci , j.e .
2006 , , 639 , 275 kitamura , y. , momose , m. , yokogawa , s. , kawabe , r. , tamura , m. & ida , s. 2002 , , 581 , 357 lahuis , f. et al .
2006 , `` c2d spectroscopy explanatory supplement , '' ( pasadena : spitzer science center ) lecar , m. , podolak , m. , sasselov , d. , & chiang , e. 2006 , , 640 , 1115 malfait , k. , waelkens , c. , waters , l. b. f. m. , vandenbussche , b. , huygen , e. , & de graauw , m. s. 1998 , , 332 , l25 markwick , a. j. , ilgner , m. , millar , t. j. , & henning , t. 2002 , , 385 , 632 mclean , i. s. , et al .
1998 , , 3354 , 566 najita , j. , carr , j. s. , & mathieu , r. d. 2003 , , 589 , 931 pontoppidan , k. m. , blake , g. a. , van dishoeck , e. f. , smette , a. , ireland , m. i. & brown , j. 2008 , , submitted rothman , l. s. , et al .
2005 , journal of quantitative spectroscopy and radiative transfer , 96 , 139 salyk , c. , blake , g. a. , boogert , a. c. a. , & brown , j. m. 2007 , , 655 , l105 stevenson , d.j . & lunine , j.i .
1988 , icarus , 75 , 146 terada , h. , tokunaga , a. t. , kobayashi , n. , takato , n. , hayano , y. , & takami , h. 2007 , , 667 , 303 thi , w. f. , et al .
2001 , , 561 , 1074 lll & dr tau & as 205a + co @xmath46 , @xmath56 [ kms@xmath14 ] & 27 , 32 & 35 , 40 + co @xmath43 , @xmath58 [ au ] & 0.8 , 0.6&0.5 , 0.4 + h@xmath1o @xmath46 , @xmath56 [ kms@xmath14 ] & 24 , 28 & 36 , 42 + h@xmath1o @xmath43 , @xmath58 [ au ] & 1.0 , 0.7 & 0.4 , 0.3 + oh @xmath46 , @xmath56 [ kms@xmath14 ] & 28 , 32&34 , 39 + oh @xmath43 , @xmath58 [ au ] & 0.7 , 0.5&0.5 , 0.4 + @xmath22[au ] & 3 & 3 + @xmath23 [ km s@xmath14 ] & 2 & 2 + @xmath20 [ @xmath59 & 0.76 & 1.2 + @xmath60 $ ] & 67&47 + @xmath61 [ sr ] & @xmath62 & @xmath63 + @xmath64 [ k ] & 1000 & 1000 + @xmath31 [ @xmath26 ] & @xmath66 & @xmath67 + @xmath68 [ @xmath26 ] & @xmath69 & @xmath70 + @xmath71 [ @xmath26 ] & @xmath72 & @xmath72 + | we present detections of numerous 1020@xmath0 m h@xmath1o emission lines from two protoplanetary disks around the t tauri stars as 205a and dr tau , obtained using the infrared spectrograph on the spitzer space telescope .
follow - up 35@xmath0 m keck - nirspec data confirm the presence of abundant water and spectrally resolve the lines .
we also detect the p4.5 ( 2.934@xmath0 m ) and p9.5 ( 3.179@xmath0 m ) doublets of oh and @xmath2co/@xmath3co @xmath4=1@xmath50 emission in both sources .
line shapes and lte models suggest that the emission from all three molecules originates between @xmath60.5 and 5 au , and so will provide a new window for understanding the chemical environment during terrestrial planet formation .
lte models also imply significant columns of h@xmath1o and oh in the inner disk atmospheres , suggesting physical transport of volatile ices either vertically or radially ; while the significant radial extent of the emission stresses the importance of a more complete understanding of non - thermal excitation processes . |
glassy te tl alloys ( te and tl being the early and late transition metal , respectively ) have been extensively studied in recent decades @xcite and the interest in these alloys further increased after the discovery of te tl base bulk metallic glasses @xcite .
these studies revealed several unusual phenomena @xcite , which has lead to the development of novel concepts for the calculation of their properties @xcite . in te
tl alloys the composition range for the formation of the amorphus state by rapid quenching from the melt is quite wide and in favourable cases it spans from @xmath14 at .
% of tl component .
such a broad composition range enables a detailed study of the changes in the electronic band structure and properties on alloying , through comparison between the model and experiment @xcite . in nonmagnetic amorphous te tl alloys ,
several properties which are related to the electronic density of states ( dos ) show simple , sometimes linear variation with tl content @xcite .
these simple variations of the properties correlate with ultraviolet photoemission spectroscopy ( ups ) results for the same alloy systems @xcite , which showed that dos at the fermi level ( @xmath15 ) , @xmath16 , is dominated by te d states .
accordingly , in amorphous te tl alloys , the effect of alloying with tl can be approximated with the dilution of amorphous te @xcite .
so far , a majority of results on te tl alloys has been obtained for zr tl metallic glasses , rendering a comparison between alloy systems based on different te ( eg .
ti , zr , hf ) rarely possible @xcite .
this is particularly true for superconductivity , with only a few results for superconducting transition temperatures @xmath17 of ti @xcite and hf
base @xcite metallic glasses .
here we report , to our knowledge first , systematic study of superconductivity in hf fe and hf cu metallic glasses .
our results show that the variation of @xmath17 with @xmath13 in hf@xmath0fe@xmath1 and hf@xmath0cu@xmath1 glassy alloys is quite similar to that observed in corresponding zr@xmath0fe@xmath1 and zr@xmath0cu@xmath1 metallic glasses @xcite .
in particular @xmath17 decreases with @xmath13 and the rate of decrease is much faster for @xmath18 fe than for cu .
the magnitudes of @xmath17 in hf
base alloys are about two times lower than those in corresponding zr base alloys @xcite
. the dressed density of states at the fermi level , @xmath19 , also decreases with increasing @xmath13 .
thus , superconductivity in hf base alloys is consistent with `` split band '' electronic structure of glassy te tl alloys @xcite .
hf@xmath0fe@xmath1 @xmath20 , hf@xmath0cu@xmath1 @xmath3 , and ti@xmath4cu@xmath5 amorphous ribbons were prepared by melt spinning of master alloys with the predetermined concentration in either pure ar @xcite or he @xcite atmosphere .
the ribbons were typically @xmath21 m ( hf fe ) and @xmath22 m ( all other alloys ) thick and their amorphousness was verified by x ray diffraction @xcite . about @xmath23 mm
long samples for resistivity measurements were glued by ge
varnish on the sample holder of a @xmath24he cryostat inserted into @xmath25 t superconducting magnet .
the current and voltage wires were glued with silver paste onto the samples .
the resistivity measurements were performed by low frequency ( @xmath26hz ) ac method with rms current @xmath27ma in the temperature range @xmath28k in magnetic field @xmath7 t , perpendicular to the broad surface of the ribbon and to the current direction .
the temperature was measured with calibrated cernox thermometer situated close to the samples .
the resistivity was determined from the measurements of resistance , length , mass and density of samples @xcite . due to finite width of the silver paste contacts the uncertainty in the absolute resistivity values was about 5% .
this uncertainty propagated into the values of the density of states @xmath19 .
some data relevant to our samples are given in table [ tbl : tableone ] .
lccccc alloy & @xmath29 ( @xmath30 cm ) & @xmath31 ( @xmath32k@xmath33 ) & @xmath17 ( k ) & @xmath34 & @xmath19 + & & & & ( koe / k ) & ( states / ev atom ) + hf@xmath35fe@xmath36 & 206 & @xmath37 & 1.86 & 38 & 2.27 + hf@xmath38fe@xmath39 & 200 & @xmath40 & 1.12 & 35 & 2.14 + hf@xmath41cu@xmath42 & 206 & @xmath43 & 1.36 & 28.5 & 1.64 + hf@xmath44cu@xmath45 & 212 & @xmath37 & 0.82 & 26.1 & 1.40 + hf@xmath46cu@xmath46 & 210 & @xmath47 & 0.44 & 23.7 & 1.16 +
figure [ fig : rdt ] shows the variation of resistance with temperature for ti@xmath4cu@xmath5 and selected hf
cu , fe glassy alloys . as is usual for glassy te@xmath0tl@xmath1 alloys with high resistivity ( @xmath48 cm ) @xcite all our samples had negative temperature coefficients of resistivity ( tcr ) .
the literature values for resistivities and tcrs of corresponding hf cu and ti@xmath4cu@xmath5 @xcite agree quite well with our results ( table [ tbl : tableone ] ) . in particular , our @xmath49k resitivities , @xmath50 , are a few percent higher than the room temperature resistivites , @xmath51 , of other authors @xcite .
figure [ fig : tc ] shows the variations of normalized resistance @xmath52 with temperature below @xmath53k for all studied alloys .
all hf cu , fe samples become superconducting within the explored temperature range ( @xmath54k ) . except for hf@xmath38fe@xmath39 alloy ,
all other samples show very narrow superconducting transitions with typical widths ( from 0.1 to @xmath55 ) @xmath56k , which can be regarded as an indication of good quality ( homogeneity ) of studied samples @xcite .
the transition width for hf@xmath38fe@xmath39 alloy , @xmath57k , is somewhat larger but not unusual for amorphous alloys .
as illustrated in figure [ fig : tc ] the resistive transition of this alloy became narrower in applied field , which allowed reliable determination of the variation of the upper critical field with temperature , @xmath58 , also for this alloy .
the values of superconducting transition temperatures ( defined as midpoints of resistive transitions ) are given in table [ tbl : tableone ] .
sample ti@xmath4cu@xmath5 showed no sign of superconductivity down to @xmath59k , which is consistent with the reported @xmath60k for this alloy @xcite . in figure
[ fig : tcx ] we compare the variations of zero - field @xmath17s with concentration @xmath13 for our hf@xmath0fe@xmath1 and hf@xmath0cu@xmath1 alloys , with the literature results for zr fe and zr cu alloys @xcite .
also shown are the maximum @xmath17s ( i.e. the highest attainable @xmath17 in given te@xmath0cu@xmath1 series ) of presumably amorphous ti cu , hf cu and zr cu thin films , obtained by low temperature ion implantation @xcite , which seem to extrapolate quite well the results for metallic glasses to lower cu contents .
since there are no previous results for @xmath17 of hf fe , cu glassy alloys , we can compare our results only with that for splat cooled hf@xmath41ni@xmath42 foil @xcite with @xmath61k .
judging by the relation between @xmath17s of similar zr ni and zr cu alloys @xcite , @xmath17 of hf@xmath41ni@xmath42 alloy @xcite is consistent with @xmath62k for our hf@xmath41cu@xmath42 alloy .
the transition temperatures of pure crystalline ( hcp ) zr and hf ( figure [ fig : tcx ] ) are about an order of magnitude lower than maximum @xmath17s of zr cu and hf cu amorphous alloys .
this is qualitatively consistent with the observed @xcite and calculated @xcite higher @xmath16 in dilute amorphous te tl alloys than those of pure crystalline ( hcp ) te metals .
as seen from figure [ fig : tcx ] the variations of @xmath17 with @xmath13 in zr fe , cu and hf fe , cu amorphous alloys are qualitatively very similar , the main difference is that @xmath17s of hf fe , cu alloys are about two times lower than those of corresponding zr fe , cu alloys . like in zr
base alloys the rate of decrease of @xmath17 with @xmath13 in hf
base alloys is much faster for fe than for cu alloy .
this is due to onset of magnetic correlations such as the spin fluctuations and/or formation of magnetic moments / clusters which cause strong pair - breaking @xcite .
lower @xmath17s of hf fe , cu alloys , compared to those of zr fe , cu , are consistent with a decrease of @xmath16 on going from zr to hf ( due to the increase of the bandwidth ) , but may also be affected @xcite by the different debye temperatures of zr and hf base alloys .
unfortunately , there are no measurements of the low temperature specific heat ( ltsh ) of hf base metallic glasses @xcite which are necessary in order to explain the difference between @xmath17s of zr base and hf base alloy systems . in the absence of ltsh ,
useful information about nature of superconductivity in metallic glasses can be obtained from the measurements of upper critical field @xmath58 @xcite .
the variation of @xmath63 with temperature in te
tl metallic glasses is usually well described by the werthamer - helfand - hohenberg theory @xcite and a fit of experimental results to the model enables one to determine the spin - orbit interaction parameter , @xmath64 , and the maki paramagnetic limitation parameter @xmath31 @xcite
. however , such fits yield reliable results for the above parameters ( especially @xmath64 ) only if the measurements extend to sufficiently low temperature , @xmath65 @xcite .
the @xmath58 variations for our hf
fe , cu alloys are shown in figure [ fig : hc2 ] .
@xmath63 was defined with @xmath55 , but as illustrated for hf@xmath35fe@xmath36 alloy the variation of @xmath63with @xmath66 for @xmath67 criterion was virtually the same . due to low @xmath17s of alloys we have studied , our measurements are limited to @xmath68 which is not sufficient for the accurate estimate of both @xmath64 and @xmath31 .
instead , we can use rather well defined initial slopes of @xmath63 , @xmath34 , for our alloys ( table [ tbl : tableone ] ) in order to estimate their dressed densities of states , @xmath19 @xcite .
@xmath19 can be calculated from @xcite : @xmath69 where @xmath70 is boltzmann constant , @xmath71 the avogadro number , @xmath72 the molecular weight , @xmath73 the mass density , and @xmath74 .
the product @xmath75 can be expressed via resistance @xmath76 , length @xmath77 , and mass @xmath78 of the sample , @xmath79 @xcite .
the values of @xmath19 for hf
cu , fe alloys calculated by using ( [ eq : ngamma ] ) decrease with increasing fe , cu content ( table [ tbl : tableone ] ) in the same fashion as @xmath19 in corresponding zr
fe , cu alloys @xcite . however , the magnitudes of @xmath19 in hf
base alloys are some 1012% lower than those in corresponding zr base alloys @xcite . as in zr
base alloys @xcite a clear correlation exists between the values of @xmath19 and @xmath17 . in metallic glasses
the values of @xmath19 calculated from ( [ eq : ngamma ] ) usually agree well with those obtained more directly from the coefficient of a linear term in ltsh , @xmath80 @xcite : @xmath81 the dressed density of states is enhanced by many body interactions in respect to a band ( bare ) density of states , @xmath16 .
in particular , @xmath82 , where @xmath83 and @xmath84 are the electron - phonon and electron - spin fluctuations interaction parameters , respectively .
since our hf
base alloys were paramagnetic @xcite with magnetic susceptibilities well below of those for corresponding zr base alloys @xcite we expect @xmath84 to be small for hf
fe , and negligible for hf cu alloys . for a reliable estimate of @xmath83
the ltsh measurements are required @xcite . since at present no ltsh results for hf
base glassy alloys exist @xcite we can not make accurate estimates of @xmath16 for studied alloys . in amorphous te tl alloys the electron - phonon enhancement factor
can also be estimated from the temperature variation of the thermopower , @xmath85 @xcite .
such an estimate for amorphous hf@xmath46cu@xmath46 alloy yields @xmath86 , nearly the same as that obtained for zr@xmath46cu@xmath46 alloy @xcite .
another estimate of @xmath83 in hf
cu glassy alloys can be obtained by dividing our results for @xmath19 with the calculated values of @xmath16 for amorphous hf@xmath0cu@xmath1 alloys @xcite .
the ratio between our @xmath19 and ( interpolated ) values of @xmath16 decreased from about 1.4 ( @xmath87 ) to @xmath88 ( @xmath89 ) .
this calculation gave @xmath16 values for amorphous hf
cu alloys about 10% lower than the values of @xmath16 in corresponding zr cu alloys @xcite . in principle
, @xmath83 can also be obtained from the approximate proportionality between @xmath83 and @xmath16 derived for disordered transition metal alloys of a given series @xcite which was found applicable to several 4d ( zr , mo)-base metallic glasses @xcite .
there , the coefficient of @xmath83 vs. @xmath16 variation for 4d and 5d series was found to be quite similar @xcite , whereas that for 3d series was sizeably smaller .
this result is in qualitative agreement with the estimates of @xmath83 from @xmath85 @xcite for equiatomic ti
cu , zr cu and hf cu amorphous alloys .
thus , the electron - phonon enhancement in hf
cu glassy alloys is probably quite similar to that in zr
cu alloys and the main reason for lower @xmath17s in the former system may be higher ionic mass of hf ( lower debye temperature , @xmath90 @xcite ) and lower @xmath16 @xcite .
near absence of superconductivity in ti
base metallic glasses @xcite , also confirmed by us ( figures [ fig : tc ] and [ fig : tcx ] ) , is puzzling . since in these systems
both @xmath19 and @xmath90 are higher than those in corresponding zr base and hf base metallic glasses @xcite , an inefficient electron - phonon coupling is required to explain their low @xmath17s @xcite .
the first systematic study of superconductivity in hf based metallic glasses has been reported . a clear correlation between the values of @xmath17 and the dressed density of states @xmath19 has been established . with exception of hf@xmath38fe@xmath39 alloy ,
higher @xmath19 corresponds to higher @xmath17 . more rapid suppression of @xmath17 with @xmath13 in hf@xmath0fe@xmath1 alloys than in hf@xmath0cu@xmath1 is probably caused by magnetic effects . in general , the variations of @xmath17 and @xmath19 in hf
fe , cu metallic glasses with fe , cu content are qualitatively the same as those in corresponding zr fe , cu glassy alloys which is consistent with very similar electronic structures of these alloys . considerably lower values of @xmath17 in hf based metallic glasses than those in corresponding zr based alloys are probably due to the lower debye temperatures @xmath90 , and electronic densities of states @xmath16 in former system . for a more detailed insight into the superconductivity of hf based metallic glasses the additional measurements of the low temperature specific heat ( yielding @xmath90 ) and perhaps tunnelling experiments ( giving more directly electron - phonon coupling ) are required .
the samples hf@xmath41cu@xmath42 and hf@xmath44cu@xmath45 have been prepared by dr .
l. varga and dr .
i. bakonyi from research institute for solid state physics and optics , hungarian academy of sciences .
this work was supported by the croatian ministry of science , education and sports projects nos .
119 - 1191458 - 1023 and 119 - 1191458 - 1019 . | systematic study of electrical resistivity of hf@xmath0fe@xmath1 @xmath2 , hf@xmath0cu@xmath1 @xmath3 , and ti@xmath4cu@xmath5 metallic glasses has been done in the temperature range @xmath6k , and in magnetic fields @xmath7 t .
all hf base alloys are superconducting with @xmath8k , which is well above @xmath9 of pure crystalline hf ( @xmath10k ) . from the initial slopes of the upper critical fields , @xmath11 , and resistivities we determined the dressed electronic densities of states , @xmath12 , for all alloys . both @xmath9 and @xmath12 decrease with increasing @xmath13 ( fe , cu content ) .
the results are compared with those for corresponding zr base metallic glasses and ion - implanted hf films . |
@xmath0 carinae is the most luminous , evolved stellar object that can be closely studied @xcite .
its immense luminosity ( @xmath10 , * ? ? ? * ) and relative proximity ( @xmath11 , * ? ? ? * ) make it possible to test and constrain specific theoretical models of extremely massive stars ( @xmath12 ) using high - quality data @xcite .
is thus one of the most intensely observed stellar systems in the galaxy , having been the focus of numerous ground- and space - based observing campaigns at multiple wavelengths from radio @xcite to gamma - rays @xcite .
unfortunately , the dusty homunculus nebula that formed during s `` great eruption '' in the 1840s enshrouds the system , complicating direct observations of the central stellar source @xcite .
nevertheless , ground- and space - based , multi - wavelength observations obtained over the past two decades strongly indicate that is a highly eccentric ( @xmath13 ) colliding wind binary ( cwb ) with a 5.54-year orbital period ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* hereafter c11 ) .
the consensus view is that the primary star , , is a luminous blue variable ( lbv ) @xcite .
radiative transfer modeling of _ hubble space telescope_/space telescope imaging spectrograph ( /stis ) spatially - resolved spectroscopic observations suggests that has a current mass @xmath14 , and a current - day stellar wind with a mass - loss rate of @xmath15 and terminal speed of @xmath16 ( * ? ? ?
* ; * ? ? ?
* hereafter h01 , h06 ) .
the companion star , , continues to evade direct detection since dwarfs its emission at most wavelengths ( damineli et al .
2000 ; h01 ; h06 ; smith et al . 2004 ; c11 ) . as a result , s stellar parameters , evolutionary state , orbit , and influence on the evolution of
are poorly known .
currently , the best constraints on the stellar properties of come from photoionization modeling of the `` weigelt blobs '' , dense , slow - moving ejecta in the vicinity of the binary system @xcite .
@xcite initially characterized as a mid - o to wn supergiant .
more recent work by @xcite tightly constrains the effective temperature of ( @xmath17 ) , but not its luminosity ( @xmath18 ) , resulting in a larger range of allowed stellar parameters .
constraints on the wind parameters of come from extended x - ray monitoring by the _ rossi x - ray timing explorer _ ( _ rxte _ ) , _ chandra _ , _ xmm _ , and _ suzaku _ satellites @xcite . the periodic nature , minimum around periastron , and hardness ( up to 10 kev ) of s _ rxte _ light curve are all characteristics of a highly eccentric cwb , the variable x - ray emission arising in a wind - wind collision ( wwc ) zone formed between the two stars ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* hereafter pc02 , o08 , p09 , p11 , and c10 , respectively ) .
the hardness of the x - rays requires that have a high wind terminal speed of @xmath19 , and detailed modeling of the momentum balance between the two shock fronts suggests a mass - loss rate of @xmath20 ( pc02 ; o08 ; p09 ; p11 ) .
proper numerical modeling of s wwc remains a challenge , mainly because it requires a full three - dimensional ( 3-d ) treatment since orbital motion , especially during periastron , can be important or even dominant , affecting the shape and dynamics of the wwc region ( o08 ; p09 ; p11 ) .
most hydrodynamical simulations of have been two - dimensional ( 2-d ) , neglecting the effects of orbital motion for simplicity ( * ? ? ?
* ; * ? ? ?
* pc02 ; henley 2005 ) . until very recently ( o08 ; p11 ) , fully 3-d simulations were computationally impractical .
when it comes to reproducing observational diagnostics of from hydrodynamical simulations of its colliding winds , the focus has been almost exclusively on x - rays .
one drawback of this is that models for at other wavelengths have been mostly phenomenological , offering only qualitative explanations for the vast array of complicated observations .
this makes it difficult to test and refine models , or quantitatively constrain the physical parameters of the system .
an excellent example is the lack of consensus regarding s orbital orientation .
the majority favor an orbit in which is behind during periastron ( damineli 1996 ; pc02 ; corcoran 2005 ; hamaguchi et al .
2007 ; nielsen et al . 2007a
; damineli et al . 2008b ; henley et al . 2008 ; o08 ; p09 ; p11 ; groh et al . 2010b ; richardson et al . 2010 ) . but others place on the near side of at periastron @xcite . some claim that the orientation is not well established at all @xcite .
a more fundamental drawback of focusing purely on x - rays or other spatially - unresolved data is that any derived orbit is ambiguous with respect to its _ absolute _ orientation on the sky ; _ any orbital orientation derived solely from fitting the x - ray data can be rotated on the sky about the observer s line - of - sight and still match the observations_. a degeneracy also exists in the orbital inclination , with models that assume @xmath21 and @xmath22 both capable of fitting the observed _ rxte _ light curve .
however , high - resolution spatial information is not enough ; moderate - resolution spectral data is needed to determine the velocity structure of the emitting gas , and thus fully constrain the orbital orientation . as a result ,
using x - ray data alone , it is impossible to determine the 3-d alignment or misalignment of the orbital axis with the homunculus polar axis , or the direction of the orbit .
it is commonly _ assumed _ that the orbital axis is aligned with the homunculus polar axis , but to date , neither modeling nor observations have unambiguously demonstrated that this is the case . knowing the orbit s orientation is key to constraining theories for the cause of the great eruption and formation of the homunculus ( i.e. single star versus binary interaction / merger scenarios , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
furthermore , a precise set of orbital parameters would lay the groundwork for orbital modeling to determine the stellar masses . in a recent attempt to characterize s interacting winds , ( * ? ? ?
* hereafter g09 ) presented an analysis of _
hst_/stis observations taken between 1998 and 2004 , identifying spatially - extended ( up to @xmath23 ) , velocity - resolved forbidden emission lines from low- and high - ionization species .
_ hst_/stis imaging spectroscopy is ideal for moving beyond x - ray signatures as it uniquely provides the spatial ( @xmath24 ) _ and _ spectral ( @xmath25 ) resolution necessary to separate the spectra of s central source and nearby circumstellar ejecta from those of the homunculus and other surrounding material .
moreover , this high - ionization forbidden emission is phase - locked , which strongly suggests it is regulated by the orbital motion of .
observed spectro - images are also highly dependent on the pa of the _ hst_/stis slit , implying that they contain valuable geometrical information .
this opens up the possibility of obtaining s orbital , stellar , and wind parameters through proper modeling of the extended forbidden line emission .
this paper presents a detailed 3-d dynamical model for the broad components of the high - ionization forbidden line emission observed in using the _ hst_/stis .
our model is based on the results of full 3-d smoothed particle hydrodynamics ( sph ) simulations of s colliding winds ( @xmath26 and @xmath27 ) .
radiative transfer calculations performed with a modified version of the splash code @xcite are used with idl routines to generate synthetic spectro - images of [ iii ] emission line structures at various orbital phases and stis slit pas ( @xmath28 ) . through a parameter study that varies the orbital inclination , the angle that the line - of - sight makes with the apastron side of the semi - major axis , and the pa on the sky of the projected orbital axis
, we are able to tightly constrain , for the first time , the _ absolute _ ( 3-d ) orientation and direction of s orbit ( @xmath29 ) , showing in particular that the orbital axis is closely aligned in 3-d with the inferred polar axis of the homunculus nebula .
a discussion of the results and this derived orientation are in @xmath30 .
@xmath31 summarizes our conclusions and outlines the direction of future work .
we begin ( @xmath32 ) with a brief summary of the observations used in this paper .
this paper uses the same /stis data described in g09 ( see their table 1 ) . extracted portions of observations recorded from march 1998 to march 2004 with the stis ccd ( @xmath33 scale ) with medium dispersion gratings ( @xmath25 ) , in combination with the @xmath34 slit from 1,640 to 10,100 , were used to sample the spectrum of the central core of .
phases of the observations , @xmath35 , are relative to the _ rxte _ x - ray minimum at 1997.9604 @xcite : @xmath36 .
solar panel orientation requirements constrained the stis slit pas .
the reduced stis ccd spectra available through the stsci archives ( http://archive . stsci.edu/prepds/etacar/[http://archive .
stsci.edu/prepds/etacar/ ] ) were used .
compass directions ( nw = north - west , etc . ) describe the spatial extent of the emission .
all wavelengths are in vacuum and velocities are heliocentric . for details , see g09 and references therein .
figure [ fig1 ] shows examples of resolved , broad emission from [ iii ] @xmath37 . in the top center panel , faint spatially- and velocity - resolved emission
can be seen against bright nebular emission and dust - scattered stellar radiation .
strong , narrow line emission from the sw of is also seen .
the bright continuum close to the wavelength of the wind line of interest is subtracted on a spatial row - by - row basis in order to isolate the fainter broad forbidden line emission ( top right panel of figure [ fig1 ] ) .
all spectro - images in this paper have been processed identically using a portion of the spectrum with no bright narrow- or broad - line contamination .
row - by - row subtraction across the stellar position is less successful due to insufficient correction in the data reduction for small tilts of the spectrum on the ccd ( g09 ) .
the bottom right panel of figure [ fig1 ] contains the same spectro - image of [ iii ] @xmath37 as the top right panel , but with masks that remove narrow line contamination from the weigelt blobs , and contamination to the red due to the nearby [ ii ] @xmath384666 wind line .
the masks are intended to remove `` distracting '' emission features not relevant to the work in this paper , displaying only the observed broad structures that are the focus of the modeling ( outlined in white ) . to determine the spatial and spectral ranges where [ ii ] @xmath384666 contaminates the [ iii ] @xmath37 image , we use as a template the uncontaminated spectro - image of the bright [ ii ] @xmath384815 line ( bottom center panel of figure [ fig1 ] ) .
we also compare the [ iii ] @xmath37 images to those of the [ iii ] @xmath39 and [ ii ] @xmath40 lines , which form in nearly identical conditions as they have very similar critical densities and ips ( table 3 of g09 ) .
a detailed discussion of the observations and masking procedure for all spectro - images modeled in this paper is in appendix a ( available in the online version of this paper - see supporting information ) .
three key features observed in spectro - images of the high - ionization forbidden line emission can be used to constrain the 3-d orientation of s orbit .
each provides important clues about the nature and orbital variation of s interacting winds .
we briefly summarize these below .
figure [ fig2 ] displays observed spectro - images of [ iii ] @xmath384659 , recorded at slit pa = + 38@xmath43 on 2003 may 5 ( @xmath42 ) , showing spatially - extended ( up to @xmath44 ) emission in the form of very distinct , nearly complete arcs that are entirely blue - shifted , up to @xmath45 .
several weaker [ iii ] lines nearby in the spectrum , while blended , show no evidence of a spatially - extended ( @xmath46 ) red component , and no other high - ionization forbidden lines show an extended red component at this position ( g09 ) .
to illustrate this , the bottom right panel of figure [ fig2 ] contains a continuum - subtracted spectro - image of [ ii ] @xmath385756 , taken at the same orbital phase and slit pa .
the ip ( @xmath47 ) and critical density ( @xmath48 ) of [ ii ] @xmath385756 are very similar to that of [ iii ] @xmath384659 ( @xmath49 and @xmath50 ) , meaning that their spatially - extended emission forms in nearly identical regions and via the same physical mechanism .
the [ ii ] @xmath385756 line is strong and located in a spectral region with very few other lines . unlike [ iii ] @xmath384659 , there is no contamination to the red of [ ii ] @xmath385756 , making it ideal for demonstrating that there is no spatially - extended , red - shifted , broad high - ionization forbidden line emission . of particular importance are the asymmetric shape and intensity of these emission arcs .
both arcs extend to nearly the same spatial distance from the central core , but the upper arc is noticeably dimmer and stretches farther to the blue , to @xmath45 .
the lower arc is brighter , but only extends to @xmath51 .
every high - ionization forbidden line observed at this orbital phase and slit pa exhibits these asymmetries in intensity and velocity ( figures @xmath52 of g09 ) , indicating that they are independent of the ip and critical density of the line and intrinsic to the shape and distribution of the photoionized extended wind material in which the lines form . as such
, the shape and asymmetry of the blue - shifted emission arcs may be used to help constrain the orbital orientation . ) .
top left panel : _
hst_/acs hrc image of the central @xmath53 of with the @xmath53 long portion of the stis slit at pa = + 38@xmath43 overlaid .
the locations of weigelt blobs b , c , and d are indicated , as is the direction of north .
top right panel : continuum - subtracted spectro - image of [ iii ] @xmath384659 .
emission appears as a pair of completely blue - shifted arcs .
color is proportional to the square root of the intensity and the velocity scale is @xmath54 .
weak emission to the red is entirely due to contamination by [ ii ] @xmath384666 , while weak , narrow emission is centered at @xmath55 .
bottom left panel : same spectro - image , but with masks that remove the line contamination .
bottom right panel : continuum - subtracted spectro - image of [ ii ] @xmath385756 included to demonstrate that at this phase and stis pa , the high - ionization forbidden lines show no evidence of broad , spatially - extended , red - shifted emission.,width=321 ] let us focus next on shifts of the high - ionization forbidden emission with orbital phase .
figure 11 of g09 shows sets of six spectra of [ iii ] @xmath384659 and @xmath384702 recorded at select phases at slit @xmath57 . during the two minima at @xmath58 and 1.040 ( rows 1 and 5 of figure 11 of g09 ) ,
the broad [ iii ] emission is absent .
by @xmath59 ( row 2 ) and @xmath60 ( row 6 , @xmath61 months after the x - ray minimum ) , it strongly reappears .
images at @xmath62 are similar to those at @xmath59 and 1.122 . by @xmath63 ,
emission above @xmath64 disappears .
this phase dependence in the [ iii ] emission , especially the disappearance during periastron , strongly indicates that it is tied to the orbital motion of .
another important observational feature is the presence of spatially - extended _ red - shifted _ emission to the nw , in addition to extended blue - shifted emission to the se , at phases far from periastron .
this is in contrast to the entirely blue - shifted arcs seen for slit pa = @xmath41 , indicating that the stis slit at pa = @xmath56 is sampling different emitting portions of s extended wind structures , even at similar orbital phases .
the final set of observations focuses on changes with slit pa leading up to periastron .
figures 12 and 13 of g09 illustrate the behavior of the [ iii ] emission for various slit pas at select phases .
the key point to take away from this particular data set is that away from periastron , the spatially - extended [ iii ] emission is almost entirely blue - shifted for positive slit pas = @xmath65 to @xmath66 , but is partially red - shifted for negative pas , most notably pa = @xmath67 .
table 3 of g09 lists the various forbidden emission lines observed in .
however , there are several important reasons to focus particular attention on [ iii ] @xmath384659 emission .
first , the physical mechanism for the formation of forbidden lines is well understood ( appendix b , online version ) .
moreover , emission from forbidden lines is optically thin .
thus , forbidden line radiation can escape from a nebula much more easily than radiation from an optically thick resonance line , which is emitted and reabsorbed many times @xcite .
this is of great value when studying an object so enshrouded by circumstellar material since one does not have to model complicated radiative transfer effects .
[ iii ] is considered rather than a lower ionization line of [ ii ] because _ the bulk of the _ [ iii ] _ emission arises in regions that are directly photoionized by _
( verner et al . 2005 ; g09 ; mehner et al . 2010 ) .
no intrinsic [ iii ] emission is expected from .
this is based on detailed theoretical models by h01 and h06 , which show that in s envelope , fe@xmath68 only exists in regions where the electron density is two - to - four orders of magnitude higher than the critical density of the [ iii ] @xmath384659 line .
since fe@xmath69 needs @xmath49 radiation ( or collisions ) to reach the fe@xmath68 state , most of the [ iii ] emission arises in areas directly photoionized by . in comparison
, fe@xmath69 only needs 7.9 ev radiation or collisions to ionize to fe@xmath70 , which can form in the wind of ( without s influence ) and/or near the wind - wind interaction regions in areas excited by mid - uv radiation filtered by s dense wind .
collisions and photoexcitation to upper fe@xmath70 energy levels can also populate many metastable levels . as a result , [ ii ]
emission is far more complicated and originates from lower excitation , lower density regions on much larger spatial scales ( h01 ; h06 ; g09 ; mehner et al .
because of this complexity , the modeling of [ ii ] emission is deferred to future work .
the ip of 16.2 ev required for fe@xmath68 also means that the [ iii ] emission forms in regions where hydrogen is ionized , but helium is still neutral .
in contrast , both ar@xmath68 and ne@xmath68 have ips greater than @xmath71 ev .
therefore , [ iii ] and [ iii ] emission arise in areas where helium is singly ionized . modeling [ iii ]
is thus more straightforward as one does not have to worry about the two different types of ionization structure possible ( one due to h and one due to he ) , which depend on the spectrum of ionizing radiation and the abundance of helium @xcite .
[ iii ] @xmath384659 is chosen over the similar emission line of [ ii ] @xmath385756 in order to avoid potential complications due to intrinsic [ ii ] emission in the extended wind of .
[ ii ] emission can form in a broad zone between 100 and 1000@xmath72 ( @xmath73 au ) of and is sensitive to its wind temperature and mass - loss rate ( h01 ) . by focusing on [ iii ] , one avoids having to include any intrinsic forbidden line emission from .
the modeling in this paper focuses solely on the _ broad _ [ iii ] emission features that are thought to arise in the dense , moderate velocity ( @xmath74 ) extended primary wind and wwc regions .
the much narrower ( @xmath75 ) emission features that form in the weigelt blobs and other dense , slow - moving equatorial ejecta from s smaller eruption in the 1890s @xcite are not modeled .
any possible effects of local dust formation are neglected for simplicity as not enough information is available to realistically include them at this time @xcite . as all observed
high - ionization forbidden lines show the same basic spatial and temporal features in their broad emission , with any differences in size or location attributable to differences in the ip and critical density of the specific line of interest , focusing solely on [ iii ] @xmath384659 should not significantly bias the overall conclusions , which should extend to the other high - ionization forbidden lines as well .
values of the transition probability ( @xmath76 ) , statistical weights ( @xmath77 ) , and line transition frequency ( @xmath78 ) for [ iii ] @xmath384659 all come from @xcite .
values for the collision strengths ( @xmath79 ) come from @xcite .
a solar abundance of iron is also used , consistent with the works of h01 ; @xcite ; and h06 .
the numerical simulations in this paper were performed with the same 3-d sph code used in o08 .
the stellar winds are modeled by an ensemble of gas particles that are continuously ejected with a given outward velocity at a radius just outside each star . for simplicity ,
both winds are taken to be adiabatic , with the same initial temperature ( @xmath80 k ) at the stellar surfaces and coasting without any net external forces , assuming that gravitational forces are effectively canceled by radiative driving terms ( o08 ) .
the gas has negligible self - gravity and adiabatic cooling is included . in a standard @xmath81 cartesian coordinate system
, the binary orbit is set in the @xmath82 plane , with the origin at the system center - of - mass and the orbital major axis along the @xmath83-axis .
the two stars orbit counter - clockwise when looking down on the orbital plane along the @xmath84 axis .
simulations are started with the stars at apastron and run for multiple consecutive orbits . by convention , @xmath85 ( @xmath86 )
is defined to be at periastron passage . to match the @xmath87 scale of the stis observations at the adopted distance of 2.3 kpc to the system @xcite , the outer simulation boundary is set at @xmath88 from the system center - of - mass , where @xmath89 is the length of the orbital semi - major axis ( @xmath90 au ) .
particles crossing this boundary are removed from the simulation .
we note that the sph formalism is ideally suited for such large - scale 3-d simulations , as compared to grid - based hydrodynamics codes @xcite .
table [ tab4.1 ] summarizes the stellar , wind , and orbital parameters used in the modeling , which are consistent with those derived from the observations ( pc02 ; h01 ; h06 ; c11 ) , with the exception of the wind temperature of .
the effect of the wind temperature on the dynamics of the high - velocity wind collision is negligible ( o08 ) .
note that the simulation adopts the higher mass - loss rate for derived by @xcite , @xcite , h01 , and h06 , rather than the factor of four lower mass - loss rate assumed when analyzing x - ray signatures ( pc02 ; o08 ; p09 ; p11 ; c11 ) .
effects of the adopted value of the primary mass - loss rate on the forbidden line emission will be the subject of a future paper .
the publicly available software splash @xcite is used to visualize the 3-d sph code output .
splash differs from other tools because it is designed to visualize sph data using sph algorithms .
there are a number of benefits to using splash , and the reader is referred to @xcite and the splash userguide for these and discussions on the interpolation algorithms .
.stellar , wind , and orbital parameters of the 3-d sph simulation [ cols="<,^,^",options="header " , ] the 3-d sph simulation provides the time - dependent , 3-d density and temperature structure of s interacting winds on spatial scales comparable to those of the _ hst_/stis observations .
this forms the basis of our 3-d dynamical model .
unfortunately , it is currently not possible to perform full 3-d simulations of in which the radiative transfer is properly coupled to the hydrodynamics @xcite . instead
, the radiative transfer calculations here are performed as post - processing on the 3-d sph simulation output .
this should not strongly affect the results so long as the material that is photoionized by responds nearly instantaneously to its uv flux .
in other words , as long as the timescale for the recombination of @xmath91 is very small relative to the orbital timescale , the calculations should be valid .
the recombination timescale is @xmath92 seconds , where @xmath93 is the recombination rate coefficient ( in units of @xmath94 ) at temperature @xmath95 and @xmath96 is the electron number density . for @xmath97
k , @xmath98 @xcite .
since densities near the critical density of the line are the primary focus , @xmath99 ; this gives @xmath100 , which is much smaller than the orbital timescale , even during periastron passage , which takes approximately one month .
any possible time - delay effects should not strongly affect the results as the light - travel - time to material at distances of @xmath101 in is only about one week .
synthetic slit spectro - images are generated using a combination of interactive data language ( idl ) routines and radiative transfer calculations performed with a modified version of the splash @xcite code .
here , we outline the basic procedure , with the specifics discussed in the following subsections . nearly all of the required calculations are performed within splash , which reads in the 3-d sph code output for a specific orbital phase , rotates the data to a desired orbital orientation on the sky , and computes the line - of - sight velocity for all of the material .
next is the computation of the ionization volume created by where fe@xmath70 is photoionized to fe@xmath68 .
it is assumed that any material within this volume is in collisional ionization equilibrium , and that no material with @xmath102 k emits .
all other material in the photoionization volume has its forbidden line emissivity calculated using equation ( [ 5.23 ] ) , derived below .
material located outside the photoionization volume does not emit . the intensity @xmath103 is computed by performing a line - of - sight integration of the emissivity at each pixel , resulting in an image of the spatial distribution of the [ iii ] intensity projected on the sky .
this is done for each velocity @xmath104 used along the dispersion axis of the synthetic slit in the velocity range of interest , resulting in a set of intensity images .
the individual [ iii ] @xmath384659 intensity images are then combined and convolved with the _
hst_/stis response using idl routines in order to create a synthetic position versus velocity spectro - image for comparison to the observations . with the two stars orbiting in the @xmath82 plane ,
the 3-d orientation of the binary relative to the observer is defined by the orbital inclination @xmath1 , the prograde direction angle @xmath2 that the orbital plane projection of the observer s line - of - sight makes with the apastron side of the semi - major axis is the same as the angle @xmath35 defined in figure 3 of o08 . ] , and the position angle _ on the sky _ of the @xmath84 orbital axis , pa@xmath105 ( figure [ orientation ] ) . that the observer s line - of - sight makes with the @xmath84 orbital axis , and the equatorial projection angle @xmath2 of the line - of - sight relative to the apastron side of the semi - major axis @xmath83 .
the background orthogonal planes show slices of density from the 3-d sph simulation at apastron in the @xmath82 orbital plane and the @xmath106 plane perpendicular to the orbital plane and major axis . bottom :
diagram defining the position angle on the sky , @xmath107 , of the @xmath84 orbital axis ( blue ) , measured in degrees counter - clockwise of north ( n ) . the binary orbit projected on the sky
is shown in yellow , as are the semi - major ( red ) and semi - minor ( green ) axes .
the small inset to the left is a wfpc2 image of the homunculus ( credit : nasa , esa , and the hubble sm4 ero team ) included for reference.,width=291 ] following the standard definition for binary systems , @xmath1 is the angle that the line - of - sight makes with the @xmath84 axis .
an @xmath108 is a face - on orbit with the observer s line - of - sight along the @xmath84 axis and the two stars orbiting counter - clockwise on the sky .
an @xmath109 places the line - of - sight in the orbital @xmath82 plane , while @xmath110 is a face - on orbit with the two stars orbiting clockwise on the sky . for @xmath109 ,
a value of @xmath111 has the observer looking along the @xmath112 axis ( along the semi - major axis on the apastron side of the system ) , while a @xmath113 has the observer looking along the @xmath114 axis . in the conventional notation of binary orbits , the ` argument of periapsis ' @xmath115 .
pa@xmath105 defines the position angle on the sky of the @xmath84 orbital axis and is measured in degrees counter - clockwise of n. a @xmath116 aligns the projected orbital axis with the homunculus polar axis @xcite and has @xmath84 pointing nw on the sky .
a @xmath117 places the orbital axis perpendicular to the homunculus polar axis with @xmath84 pointing ne .
a @xmath118 @xmath119 ) is also aligned with ( perpendicular to ) the homunculus polar axis and has @xmath84 pointing se ( sw ) on the sky .
based on the models of s envelope by h01 and h06 , it is assumed that fe@xmath70 is initially the dominant ionization state of iron in primary wind material located @xmath120 from , and that hydrogen and helium are initially neutral in the primary wind for distances of @xmath121 and @xmath122 , respectively .
the calculation of the volume of material photoionized from fe@xmath70 to fe@xmath68 by follows that presented in @xcite for symbiotic star systems . for cases where hydrogen or helium alone determine the ionization structure , the boundary between neutral hydrogen ( h@xmath69 ) and ionized hydrogen ( h@xmath70 ) can be expressed analytically . as s extended wind consists mostly of hydrogen and helium ( figure 9 of h01 ) , and
because the [ iii ] emission arises in regions where hydrogen is ionized but helium is not , such a calculation should provide a reasonable approximation for the size and shape of the photoionization boundary between fe@xmath70 and fe@xmath68 .
this , however , is an upper limit for the photoionization volume since the ip of 16.2 ev required for fe@xmath68 is slightly larger than the 13.6 ev needed to ionize h@xmath69 . in our calculation , the primary star is separated by a distance @xmath123 from a hot companion star that emits spherically symmetrically @xmath124 photons s@xmath125 that are capable of ionizing h@xmath69 .
the primary has a standard spherically symmetric mass - loss rate of @xmath126 , where @xmath127 is the hydrogen number density , @xmath128 is the mean molecular weight , @xmath129 is the mass of a hydrogen atom , and @xmath130 is the terminal speed of the primary wind . in a small angle @xmath131 around the direction @xmath2 ,
the equilibrium condition for recombination - ionization balance is @xmath132 + where @xmath133 measures the distance from the ionizing secondary star in the direction @xmath2 , @xmath134 is the boundary between h@xmath69 and h@xmath70 , @xmath135 is the total hydrogenic recombination coefficient in case b at electron temperature @xmath136 , and @xmath96 is the electron number density @xcite . the electron density can be written as @xmath137 where @xmath138 + and @xmath139 is the abundance by number of he relative to h. expressing both @xmath133 and @xmath140 in units of @xmath123 , defining @xmath141 , and making the further approximation that @xmath135 is constant @xcite , equation ( [ 5.24 ] ) can be written as @xmath142 where @xmath143 + and @xmath144 in the modified splash routine , equation ( [ 5.27 ] ) is solved for a specified @xmath124 . however , since equation ( [ 5.27 ] ) does not account for the effects of the wwc , we assume that mass is conserved and that all primary wind material normally within the cavity created by is compressed into the walls of the wwc region , with the ionization state of the wwc zone walls before being photoionized the same as that of the primary wind ( i.e. iron is in the fe@xmath70 state ) . the stellar separation is computed directly from the 3-d sph simulation .
we use the same value for the abundance by number of he to h as h01 and h06 , @xmath145 . a constant value of @xmath146 at @xmath147 k is also used @xcite .
based on @xcite and @xcite , the value of @xmath124 is chosen to be that of an o5 giant with @xmath148 , which according to @xcite has a hydrogen ionizing flux of @xmath149 photons s@xmath125 . any shielding effects due to the presence of the weigelt blobs or other dense , slow - moving equatorial ejecta are not presently included .
the third column of figure [ fig5 ] illustrates the time variability of the photoionization region created by for slices in the @xmath82 orbital plane .
the region is largest around apastron , when is farthest from s dense wind .
as moves closer to , the photoionization zone becomes smaller and more wedge - shaped due to the gradual embedding of in s wind . during periastron ,
becomes completely enshrouded in s thick wind , preventing material at large distances ( @xmath150 ) from being photoionized .
as emerges after periastron , the photoionization region is restored and grows as moves back toward apastron .
this plunging into , and withdrawal from , s wind by leads to the illumination of distant material in very specific directions as a function of phase . for a simple two - level atom ,
the volume emissivity @xmath151 ( erg @xmath152 s@xmath125 sr@xmath125 ) of a forbidden line is @xmath153 + where @xmath154 is the number density of atoms in the excited upper level and @xmath155 is the frequency of the line transition @xcite . using the total number density of element of interest @xmath156 in ionization state @xmath1 , @xmath157 @xcite , together with equation ( b5 )
, one finds @xmath158 + solving for @xmath154 and using equation ( b4 ) for @xmath159 gives @xmath160 \right\}^{-1 } \ , \label{5.17}\ ] ] + where @xmath161 , defined in equation ( b6 ) , is the critical density of the line . by defining @xmath162 as the fraction of element @xmath156 in ionization state @xmath1 , @xmath163 as the abundance of element @xmath156 relative to all nucleons @xmath164 , and @xmath165 as the ratio of nucleons to electrons @xcite ,
@xmath166 becomes @xmath167 therefore , using equations ( [ 5.17 ] ) and ( [ 5.18 ] ) , @xmath168 \right\}^{-1 } \ .
\label{5.19}\end{aligned}\ ] ] based on equation ( [ 5.19 ] ) , emission from a specific forbidden line is concentrated only in regions that ( 1 ) contain the right ionization state of the element of interest ( in our case , fe@xmath68 ) , ( 2 ) are near the critical density of the line ( for [ iii ] , @xmath50 ) , and ( 3 ) are near the right temperature ( for [ iii ] , @xmath169 ) .
this important point is crucial to understanding the forbidden line emission observed in . to compute the emissivity using splash
, the line profiles are assumed to have a gaussian thermal broadening , so that equation ( [ 5.19 ] ) is weighted by an exponential with the form @xmath170 \nonumber\\ \nonumber\\ & & \times \left\{1 + \exp\left(\frac{h \nu_{21}}{kt}\right ) \left[\frac{g_{1}}{g_{2 } } + \frac{a_{21}g_{1}t^{1/2}}{\beta \omega_{12 } n_{e } } \right ] \right\}^{-1}\ , \label{5.22 } \ ] ] + where equation ( b8 ) has been used for @xmath161 , @xmath171 is the line - of - sight velocity of material in the 3-d sph simulation for a specified orientation of the binary orbit relative to the observer , @xmath172 is a thermal velocity dispersion ( taken to be 25 km s@xmath125 ) , and @xmath104 is the bin size used along the dispersion axis of the synthetic slit in the wavelength range of interest , expressed in thermal velocity units ( 25 km s@xmath125 ) .
to relate the physical density in the sph simulation to the electron number density , we use @xmath173 , with @xmath174 the mean molecular weight per free electron .
as the focus is on forbidden lines of trace metals , the value of @xmath96 should not be impacted by the ionization balance of the metals @xcite . moreover , since far - uv radiation ( or high - energy collisions ) is necessary for the formation of the high - ionization forbidden lines , wherever [ iii ] emission occurs , hydrogen should also be ionized . for such regions dominated by h@xmath70 , we thus take @xmath175 .
the volume emissivity then takes the final general form , @xmath176 \nonumber\\ \nonumber\\ & & \times \left\{1 + \exp\left(\frac{h\nu_{21}}{kt}\right ) \left[\frac{g_{1}}{g_{2 } } + \frac{a_{21}g_{1}t^{1/2}m_{\mathrm{h}}}{\beta \omega_{12 } \rho_{sph } } \right ] \right\}^{-1 } .
\label{5.23 } \ ] ] + the appropriate value of @xmath177 is found assuming collisional ionization equilibrium , with the fraction of fe@xmath68 as a function of @xmath95 based on the ion fraction data from @xcite .
these data show that at @xmath178 k , the fractional abundance of fe@xmath68 is @xmath179 , with the remaining fe being collisionally ionized to higher states .
the amount of fe@xmath68 decreases even more at higher temperatures .
the model therefore assumes zero [ iii ] emission from material with @xmath102 k. since the [ iii ] emission is optically thin , the intensity _
i _ is simply the integral of the volume emissivity @xmath151 along the line - of - sight , @xmath180 + @xcite .
the intensity is computed in splash by performing a line - of - sight integration of @xmath151 through the entire 3-d simulation at each pixel , resulting in an image of the intensity in the [ iii ] @xmath384659 line centered at a particular velocity ( wavelength ) , as an observer would see it projected on the sky .
this is done for multiple velocities ( @xmath104 ) in @xmath181 intervals for line - of - sight velocities from @xmath182 to @xmath183 .
the resulting series of [ iii ] intensity images are combined using idl routines to create a synthetic position versus velocity spectro - image .
the idl code reads in the splash images and rotates them to a specified pa on the sky corresponding to a desired pa of the _ hst_/stis slit , assuming that the orbital axis is either aligned with the homunculus polar axis at a @xmath116 , or rotated relative to the homunculus axis by some specified angle .
each image is cropped to match the @xmath184 width of the stis slit , with the slit assumed centered exactly on the central source .
since position information is only available along one direction , each image has its intensity values integrated along each row of pixels within the slit width .
this produces a single ` slice ' of the [ iii ] @xmath384659 intensity along the slit , centered at a specific velocity .
these slices are combined to create the synthetic spectro - image .
the resulting model spectro - images have spatial ( @xmath185 ) and spectral ( 25 km s@xmath125 ) resolutions that are better than the _ hst _ observations .
the synthetic spectro - images are therefore convolved with the response of _
hst_/stis in order to match its spatial ( @xmath184 ) and spectral ( @xmath186 ) resolutions .
point - spread functions ( psfs ) for stis generated using the tiny tim program @xcite are used for the convolution in the spatial direction , while a gaussian is used for the convolution in the spectral direction .
color in the model spectro - images scales as the square root of the intensity , with the colorbar ranging from 0 to @xmath187 of the maximum intensity , the same as used for displaying the observations .
figure [ unconvolved ] shows that there is a noticeable difference between the unconvolved ( left ) and convolved ( right ) model spectro - images , especially in the central @xmath188 .
details in the unconvolved spectro - images that arise near the inner wwc zone are completely unresolved in the convolved images , resulting in a bright , central streak of emission extending from negative to positive velocities .
this is consistent with the observations and indicates that _
hst_/stis lacks the spatial resolution needed to resolve the details of s inner wwc zone .
@xmath384659 both unconvolved ( left ) and convolved with the response of _ hst_/stis ( right ) , assuming @xmath189 , @xmath111 , and @xmath116 .
roman numerals i and ii indicate the top and bottom of the slit , respectively .
color is proportional to the square root of the intensity , and the velocity scale is @xmath54.,width=321 ]
in the discussions below , it is assumed for simplicity that phase zero of the spectroscopic cycle ( from the observations ) coincides with phase zero of the orbital cycle ( periastron passage ) . in a highly - eccentric binary system like , the two values are not expected to be shifted by more than a few weeks @xcite .
such a time shift would only cause a small change of @xmath190 in the derived best value of @xmath2 , which will not affect the overall conclusions .
the first two columns of figures [ fig4 ] and [ fig5 ] show , respectively , the density and temperature in the @xmath82 orbital plane from the 3-d sph simulation . the binary system has undergone multiple orbits , as indicated by the dense arcs of wind material in the outer ( @xmath191 ) regions . on the @xmath192 ( periastron ) side of the system
are narrow cavities carved by in s dense wind during each periastron passage .
the cavities are most easily seen in the temperature plots as they contain warm ( @xmath193 ) , low - density wind material from .
bordering these cavities are compressed , high - density shells of primary wind that form as a result of the wwc .
the insets of figure [ fig5 ] illustrate the creation of one of these narrow wind cavities and its bordering dense shell of primary wind .
their spiral shape is due to the increased orbital speeds of the stars during periastron .
this is in contrast to phases around apastron ( rows b , c and d ) when orbital speeds are much lower and the current wwc zone maintains a simple , axisymmetric conical shape ( o08 ; p09 ; p11 ) .
the increasing orbital speeds when approaching periastron causes the postshock gas in the leading arm of the wwc zone to be heated to higher temperatures than the gas in the trailing arm ( insets of rows d and e ) , an affect also found by p11 .
following periastron , moves back to the @xmath112 ( apastron ) side of the system , its wind colliding with and heating the dense primary wind that flows unimpeded in the @xmath194 direction .
as pointed out by p11 , the arms of the wwc region shortly after this ( @xmath195 ) are so distorted by orbital motion that the leading arm collides with the trailing arm from before periastron , leading to additional heating of the postshock gas in the trailing arm ( compare @xmath196 , figure [ fig5 ] , row f ) .
the leading arm of the wwc zone ( including the portion that collides with the trailing arm ) helps to form a dense , compressed shell of primary wind that flows in the @xmath112 and @xmath197 directions after periastron passage . because the wind of collides with a dense wall of postshock primary wind with high inertia , the primary wind controls the overall rate of expansion of the resulting spiral ( p11 ) .
the overall stability of the expanding shell of primary wind on the apastron side of the system depends on the shock thickness ( vishniac 1983 ; wnsch et al . 2010 ; p11 ) .
portions of the shell moving in the @xmath197 direction appear to be the most stable due to the increased amount of primary wind that borders it in this direction .
however , our simulations show that at @xmath198 , the upper portion of the shell expanding in the @xmath199 direction starts to fragment .
eventually , the wind of is able to plough through the shell , causing it to separate from the leading arm of the wwc zone ( @xmath200 , row c of figure [ fig5 ] ) .
this produces a pair of dense ` arcs ' of primary wind on the apastron side of the system .
the arc expanding in the @xmath194 direction is the remnant of the leading arm of the wwc region from the previous periastron passage .
the arc in the @xmath201 direction is the remnant of the trailing arm of the wwc region from just _ prior _ to the previous periastron passage .
multiple pairs of these arcs are visible in figures [ fig4 ] and [ fig5 ]
. they are also quite spatially extended ; by the time the system is back at periastron , the arcs from the previous periastron are up to @xmath202 from the central stars ( in the @xmath82 plane ) . as the arcs expand further , they gradually mix with the surrounding low - density wind material from .
the fragmentation of the shell and formation of the arcs is mainly due to the above - mentioned collision of the leading arm of the wwc zone with the trailing arm just after periastron .
this collision produces instabilities at the interface between the two arms ( p11 ) , which , together with the shell s expansion , the pressure from s high - velocity wind , and the lack of bordering primary wind material for support , causes the shell to break apart in the @xmath199 direction @xmath203 years after periastron .
while the term ` shell ' is used to describe the layer of compressed primary wind that flows in the direction of apastron following periastron passage , this is not related to the ` shell ejection ' event discussed in the past literature @xcite .
the scenario proposed by @xcite and advocated by @xcite to explain s spectroscopic events is a qualitative single - star model wherein some sort of thermal or surface instability is presumed to induce significant mass loss from approximately every 5.54 years .
a shell ejection as the result of a latitude - dependent disturbance in the wind of a single star has also been proposed @xcite .
@xcite suggested that the shell event may be triggered by the close approach of a secondary star , but this is still mainly a single - star scenario requiring some kind of surface instability in the primary .
in contrast to these ideas , the formation of the outflowing shell of primary wind seen in the 3-d sph simulations is a natural , unavoidable consequence in any high - eccentricity , massive cwb in which the primary star has a significant mass - loss rate ( significant both in terms of overall mass - loss rate and mass - loss rate relative to the companion star ) .
it is not so much an ` ejection ' event as it is a chance for the normal primary wind to flow in the direction of apastron for a brief time ( @xmath204 months ) while performs periastron passage .
eventually , returns to the apastron side and its high - velocity wind collides with and compresses this primary wind material , forming a dense shell that continues to propagate in the direction of apastron until it breaks apart at its weakest point , resulting in a pair of dense ` arcs ' , as described above .
figure [ fig4 ] presents slices in the orbital plane from the 3-d sph simulation at @xmath42 , used to model the blue - shifted emission arcs seen at stis slit @xmath3 .
the colors show log density , log temperature , photoionization zone created by , and square root of the modeled emissivity of the [ iii ] @xmath384659 line .
the photoionization region is confined to the same side of the system as , the dense wind of preventing the far ( @xmath192 ) side of the system from being photoionized .
all of the [ iii ] emission is thus concentrated in two spatially - distinct regions on the apastron side of the system : * the strongest [ iii ] emission originates near the walls of the current wwc zone in the inner @xmath205 of . *
faint , spatially - extended ( out to @xmath206 ) [ iii ] emission arises in the arcs of dense primary wind formed during the previous periastron passage .
the [ iii ] emission occurs in these two areas because they are the regions within the photoionization zone that are near the critical density of the [ iii ] line and at the appropriate temperature .
figure [ fig5 ] illustrates how the density , temperature , photoionization region , and [ iii ] @xmath384659 emissivity change with orbital phase .
the six phases shown correspond to those observed in figure 11 of g09 . during periastron passage ( @xmath58 and 1.040 ,
rows a and e ) , becomes deeply embedded in the dense wind of and the photoionization region shrinks considerably , down to only a few semi - major axes across .
this leads to an effective ` shutting - off ' of the forbidden line emission at large distances ( @xmath150 ) .
by @xmath59 and 1.122 ( rows b and f ) , has moved far enough from to re - establish the large photoionization zone and spatially - extended forbidden line emission .
the photoionization region is largest , and the [ iii ] most spatially extended , around apastron ( row c , @xmath62 ) . by @xmath63 ( row d ) , starts to become embedded in the dense wind of , the photoionization region in the @xmath207 quadrant decreasing slightly in size .
far - uv radiation from leads to highly ionized regions that extend outward from its low density wind cavity in the direction of system apastron . during periastron passage
, the disappearance of the extended high - ionization forbidden emission can be attributed to the wrapping of the dense primary wind around , which traps its far - uv radiation and prevents it from photoionizing the outer wind structures responsible for the observed emission .
while is embedded in s wind , the latter flows unimpeded in the direction of apastron , eventually forming dense arcs of material that expand outward as completes periastron passage and the inner regions of the wwc zone regain their near conical shape .
these cooler , dense arcs of expanding primary wind also produce high - ionization forbidden line emission when photoionized by . observed temporal variations of the high - ionization forbidden lines can thus be linked to the orbital motion of in its highly eccentric orbit , which causes different portions of the wwc regions and extended arcs of primary wind from earlier cycles to be photoionized .
as the two stars move closer to or farther from each other , orbital motion also leads to changes in the density and temperature of the inner wwc zone , which in turn modifies the size and shape of s photoionization volume , and thus the overall shape , location , and intensity of the high - ionization forbidden line emission , even at phases away from periastron ( rows b , c , d , and f of figure [ fig5 ] ) .
forbidden line emission from the compressed , inner wwc region increases in intensity until the critical density is reached or the temperature exceeds that at which the appropriate ions can exist .
forbidden emission from the extended arcs of primary wind decreases in intensity as the arcs gradually expand and mix with the surrounding low - density wind material from .
the blue - shifted emission arcs in figure [ fig2 ] represent distinct , well - defined structures observed at a specific orbital phase and slit pa .
as such , they provide a natural basis for modeling and can be used to constrain the orbital orientation .
we have performed a parameter study in @xmath1 , @xmath2 , and @xmath107 with the goal of determining which set(s ) , if any , of orientation parameters result in synthetic spectro - images that closely match the observations in figure [ fig2 ] .
the value of @xmath2 was varied in @xmath208 increments for values of @xmath209 , with @xmath1 varied in @xmath210 increments over the range @xmath211 for each @xmath2 value , and @xmath107 varied in @xmath210 increments over the range @xmath212 for each pair of @xmath213 values .
we find that _ only _ synthetic spectro - images generated for @xmath214 are able to reasonably match the observations taken at @xmath42 , @xmath215 .
however , an ambiguity exists in @xmath1 and @xmath107 , with values of @xmath216 , @xmath217 to @xmath218 ( top row of figure [ fig9 ] ) and @xmath219 , @xmath220 to @xmath221 ( bottom row of figure [ fig9 ] ) both capable of producing entirely blue - shifted arcs that resemble those observed .
our best morphological fits are obtained using values of @xmath222 , @xmath223 , @xmath224 or @xmath189 , @xmath223 , @xmath225 , which lie near the centers of these two derived ranges of best - fit orientation parameters .
the observed morphology and asymmetries in brightness and @xmath171 of the spatially - extended emission arcs determine the allowed ranges of the orbital orientation parameters .
the parameters @xmath1 and @xmath226 primarily control the brightness asymmetry and spatial orientation of the blue- and red - shifted emission components on the sky .
figure [ fig9 ] shows that the synthetic spectro - images at slit @xmath215 for @xmath222 and @xmath227 ( at nearly identical @xmath2 and @xmath226 ) are remarkably similar .
yet , the two orbital orientations are drastically different , resulting in unique distributions on the sky of the blue- and red - shifted emission .
the red component is the most notable , which extends to the se for @xmath222 , but to the nw for @xmath189 .
the asymmetry in brightness between the two blue - shifted arcs is also controlled by @xmath1
. only values of @xmath216 or @xmath219 produce a brightness asymmetry , with other @xmath1 resulting in arcs that are approximately equal in brightness .
the value of @xmath226 determines the final orientation of the projected @xmath228 axis , and thus of the individual emission components .
because the lower blue - shifted arc in the observations is brighter than the upper arc , the extended blue emission projected on the sky must be brighter in directions to the sw ( where the lower half of the slit is located , position ii in figure [ fig9 ] ) , and dimmer in directions to the ne ( position i ) .
this limits @xmath107 to values between @xmath229 and @xmath221 .
more importantly , @xmath1 and @xmath226 determine the 3-d orientation of the orbital axis . if @xmath222 and @xmath217 to @xmath221 , the orbital axis is inclined toward the observer and _ is not aligned _ with the homunculus polar axis .
the two stars would also orbit counter - clockwise on the sky .
however , if @xmath189 , then , because @xmath230 , the orbital axis _ is _ closely aligned _ in 3-d _ with the homunculus polar axis , and the stars orbit clockwise on the sky .
figure [ fig6 ] shows that the value of @xmath2 ( for fixed @xmath189 and @xmath116 ) determines the @xmath171 of the material exhibiting [ iii ] emission .
according to our 3-d model , the spatially - extended [ iii ] emission originates in the expanding arcs of dense primary wind formed during the previous periastron passage .
when @xmath231 ( top row of figure [ fig6 ] ) this emission is mostly blue - shifted because the emitting material within s photoionization zone is moving mostly toward the observer .
the model spectro - image consists of spatially - extended , blue - shifted arcs because the slit at @xmath215 primarily samples the blue component , which stretches from ne to sw on the sky .
in contrast , the extended red component is _ not _ sampled ; only a small amount of red - shifted emission in the very central @xmath232 core falls within the slit .
moreover , only values of @xmath214 produce significant amounts of spatially - extended ( @xmath233 in total length ) blue - shifted emission with the correct observed asymmetry in @xmath171 , wherein the dimmer upper arc extends @xmath234 more to the blue than the brighter lower arc .
figure [ fig9 ] demonstrates that our 3-d dynamical model and derived best - fit orbital orientation(s ) reproduce all of the key features in the observations taken at @xmath42 , @xmath215 .
both synthetic and observed spectro - images contain concentrated , velocity - extended emission in the central @xmath232 core .
the completely blue - shifted emission arcs extend to roughly the same spatial distances ( @xmath235 ) .
the asymmetry in brightness between the upper and lower arcs is matched as well .
the asymmetry in @xmath171 is also reproduced and of the same magnitude . however , the arcs in the model images stretch a bit farther to the blue , to @xmath236 ( upper arc ) and @xmath237 ( lower arc ) , versus @xmath45 and @xmath238 in the observations .
the synthetic images thus show emission at velocities slightly above the value used in the sph simulation for s wind terminal speed .
this is a minor discrepancy though , likely due to the extended emitting material receiving an extra ` push ' from the fast wind of ( see subsection 4.3.4 ) .
[ [ results - for - orbital - orientations - with - a - mathbfbtheta - approx-180circ ] ] results for orbital orientations with a @xmath239 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + returning back to figure [ fig6 ] , one sees that the spectro - image for @xmath240 fails to match the observations .
there is a distinct lack of any spatially - extended , blue - shifted emission . instead
, the extended emission arcs are entirely _ red - shifted _ due to the quite different @xmath171 of the emitting material . when @xmath240 , is between the observer and during most of the orbit .
therefore , all of the emitting material within s photoionization zone is on the far side of the system and moving _ away from _ the observer .
the extended forbidden line emission is thus mostly red - shifted .
the slit at @xmath215 now primarily samples this red component , producing a spectro - image consisting of entirely red - shifted , spatially - extended arcs , the direct _ opposite _ of the observations . changing the value of @xmath1 and/or @xmath107 when @xmath240 does not result in a better match to the observations since most of the material photoionized by and exhibiting forbidden line emission is still moving away from the observer .
this is true for all values of @xmath2 near 180@xmath43 ; none of the model images generated for @xmath241 , regardless of the assumed @xmath1 and @xmath107 , matches the @xmath42 , @xmath215 observations .
example model spectro - images for these orientations can be found in @xcite .
we therefore find that an orbital orientation that places on the near side of at periastron , such as that favored by @xcite and others , is explicitly ruled out .
[ [ results - for - orientations - with - mathbfbtheta - approx-90circ - and - mathbf270circ ] ] results for orientations with @xmath242 and @xmath243 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + model spectro - images for @xmath113 and @xmath244 also fail to match the observations .
the 2-d images of the modeled [ iii ] intensity on the sky in figure [ fig6 ] show that in both cases , one half of the stis slit is empty ( from center to position ii for @xmath113 and from i to the center for @xmath245 )
. therefore , emission is absent to one side spatially in the spectro - images .
changing the value of @xmath1 has no effect on this since a change in @xmath1 at these @xmath2 values is equivalent to a rotation about the @xmath83 axis in the 2-d intensity images . instead , a rotation of nearly @xmath246 in @xmath107 is needed in order to place emitting material in both halves of the slit . however , such a rotation does not result in a match to the observations .
synthetic spectro - images for @xmath247 and @xmath248 all suffer from the same problems , namely , either no blue - shifted ring - like emission feature ( if @xmath249 or @xmath250 ) , or arc - like emission to only one side spatially ( if @xmath251 or @xmath252 ) @xcite .
this is true regardless of the @xmath1 and @xmath107 assumed . using the 3-d dynamical model we generated synthetic spectro - images for a slit @xmath57 at each of the phases in figure 11 of g09 , @xmath58 , 0.213 , 0.407 , 0.952 , 1.040 , and 1.122 .
the main goals were to determine if the binary scenario and 3-d model could explain the phase dependence observed in the high - ionization forbidden lines , and whether the set of @xmath57 observations could resolve the ambiguity between the sets of @xmath253 and @xmath254 best - fit orientations found in the above subsection .
figure [ fig9_2 ] illustrates the results of investigating the two best - fit @xmath1 regimes .
it is clear that synthetic spectro - images for @xmath255 to @xmath256 _ do not _ match the observations taken at slit @xmath57 ( top row of figure [ fig9_2 ] ) .
there are a number of discrepancies , but the most important and obvious is that the blue- and red - shifted emission components in the synthetic spectro - image extend in the wrong spatial directions , with the blue emission stretching to the nw and the red to the se , the _ opposite _ of the observations .
in contrast , synthetic images generated for @xmath4 to @xmath257 ( bottom row of figure [ fig9_2 ] ) _ are _ capable of matching the observations .
the morphology of the @xmath57 observations breaks the degeneracy in @xmath1 and fully constrains , in 3-d , s orbit .
because red - shifted emission is observed to extend spatially in directions to the nw , the red emission component of [ iii ] on the sky must be oriented such that it too stretches nw and falls within the top half of the @xmath57 slit .
the synthetic images of the [ iii ] intensity on the sky ( first two columns of figures [ fig9_2 ] and [ fig12 ] ) illustrate how the slit at @xmath57 samples the emitting regions of s extended interacting winds in very different directions compared to slit @xmath215 .
this is why the observed spectro - images are so different between the two slit pas , even at similar orbital phases .
we find that _ only _ orbital orientations with @xmath4 to @xmath5 , @xmath231 to @xmath208 , @xmath258 to @xmath9 are able to simultaneously match the observations taken at both slit @xmath215 _ and _ @xmath56 .
figure [ fig12 ] shows that the 3-d dynamical model and derived orbital orientation reproduce the overall observed shape and velocity structure of the [ iii ] emission at each phase for slit @xmath57 .
figure [ fig12 ] further illustrates how s extended interacting winds change with time .
the orientation of the blue component does not appear to change , always stretching from ne to sw on the sky .
however , the spatial extent does change , growing larger going from periastron to apastron as the wind structures flow outward . moving from apastron back to periastron , the intensity and spatial extent of the blue component
eventually start to decrease as the expanding arcs of primary wind drop in density and mix with the surrounding wind from .
the red - shifted emission changes in a way similar to that of the blue component , but points mainly nw .
ccl @xmath259 & pa & figure + + 0.601 & @xmath65 & [ fig13]a + 0.738 & @xmath67 & [ fig13]b + 0.820 & @xmath260 & [ fig13]c + 0.930 & @xmath261 & [ fig13]d + 0.970 & @xmath262 & [ fig13]e + 0.984 & @xmath263 & [ fig13]f + 0.995 & @xmath66 & [ fig13]g + 1.001 & @xmath260 & @xmath264 + 1.013 & @xmath265 & @xmath264 + 1.068 & @xmath266 & [ fig13]h + + synthetic spectro - images were generated at the ten combinations of phase and slit pa in table [ tab5.3 ] , assuming the best - fit orientation @xmath189 , @xmath223 , and @xmath225 .
the goal was to determine how well the 3-d model and derived binary orientation could reproduce observations at a variety of other phases and slit pas .
model images for @xmath267 and @xmath268 are nearly identical to those at @xmath58 and @xmath269 in figure [ fig12 ] and add no new information .
we therefore focus this discussion on the phases before and after periastron .
figure [ fig13 ] presents the results .
the overall match between synthetic and observed spectro - images is quite good .
the model reproduces all of the key spatial features , namely , extended ( @xmath270 ) [ iii ] emission that is almost entirely blue - shifted for positive slit pas , but that is partially red - shifted for negative slit pas . observed variations in emission with phase are also reproduced .
model images at @xmath271 ( [ fig13]a ) and @xmath272 ( [ fig13]b ) demonstrate how a large change in slit pa results in very different spectro - images .
the spatial distributions of the emission on the sky are very similar at these two phases , as expected since the system is near apastron when orbital velocities are their lowest . however , the slit at @xmath273 samples the emission in very different directions compared to @xmath274 , producing a spectro - image that resembles those taken at @xmath57 in figure [ fig12 ] .
the partial ring of emission observed at @xmath275 , @xmath276 ( [ fig13]f ) , is also noteworthy .
the match between model and observations is very good , the emission arc to the sw present in both , as well as the much shorter arc to the ne between @xmath277 and @xmath278 .
two effects appear to be causing the ne arc to be shorter .
first , due to their outward expansion , the extended arcs of primary wind have dropped in density and started to mix with the surrounding low - density wind from .
second , the photoionization of material in directions to the ne by has diminished due to its clockwise orbital motion and gradual embedding in s wind .
these lead to a decrease in the amount of extended [ iii ] emission to the ne .
this , combined with the slit pa , produces a partial arc of blue - shifted emission to the ne in the spectro - image . by @xmath279 ( [ fig13]g ) ,
the extended emission has vanished , implying that the ionizing flux of photons from shuts off sometime between @xmath275 and 0.995 . at @xmath280 ( [ fig13]h )
, there is still no observed spatially - extended emission .
yet , the 2-d model images of the intensity on the sky show that the extended [ iii ] emission has started to return in directions to the ne .
this is visible in the model spectro - image as a bright , blue - shifted bulge that points ne .
unfortunately , the observational data in the central @xmath188 is of insufficient quality to tell if such an emission bulge was detected .
nevertheless , the 3-d model predicts that should start to emerge from s dense wind and begin to restore the spatially - extended , high - ionization forbidden line emission at @xmath281 . during most of the orbit , the wind of collides with the dense arcs of primary wind on the apastron side of the system where the spatially - extended forbidden line emission forms ( see rows b through d of figure [ fig5 ] ) . because the arcs are bordered by a large , low - density wind cavity created earlier by , they have almost no support against s high - velocity wind , which is able to drive portions of the arcs into the cavity at velocities just above their outflow speed , the terminal speed of the primary wind .
an orientation that places the observer on the apastron side of the system thus results in model spectro - images containing blue - shifted emission at velocities slightly higher than the wind terminal velocity of the primary . because the observations do not contain blue - shifted emission at speeds more than @xmath282
, the wind speed(s ) of and/or used in the sph simulation may be a bit too large .
the launching of the two winds at their terminal velocity in the simulation may also be having a small effect on the results .
better constraints on s wind terminal speed and improved 3-d simulations with proper driving of the stellar winds should help resolve these issues .
improved modeling coupled with future observations of the extended forbidden emission in s central arcsecond also has the potential to further constrain the wind terminal velocity of . the main discrepancies between synthetic spectro - images assuming @xmath189 , @xmath223 , @xmath225 and the observations for slit @xmath57 are : ( 1 ) the extended blue - shifted component is located @xmath283 farther to the blue in the model images than in the observations , and ( 2 ) the red - shifted component in the model images is @xmath234 broader than observed .
both issues are likely due to the reasons discussed above concerning the wind terminal velocities of and .
there has been much speculation about the binary orientation , and many papers published offering suggestions ( see section @xmath285 ) .
the detailed modeling of section @xmath286 tightly constrains the observer s line - of - sight to angles of @xmath287 prograde of the semi - major axis on the apastron side of the system , placing behind during periastron .
given the uncertainties in the stellar and wind parameters of used in the 3-d simulations , and the assumption that phase zero of the orbital and spectroscopic periods coincide , values of @xmath288 may also be possible .
a study in the stellar / wind parameters of , or improved observational constraints , is needed to further refine the value of @xmath2 .
more importantly , the results bound , for the first time , the orientation of s orbital plane , with the orbital axis closely aligned in 3-d with the inferred polar axis of the homunculus at an @xmath289 and @xmath290 . figure [ fig14 ] illustrates the orientation of s binary orbit on the sky relative to the homunculus . with @xmath291 , @xmath292 , and @xmath293 , the resulting projected orbit on the sky
has moving _ clockwise _ relative to , with approaching from the sw prior to periastron , and receding to the ne afterward . , @xmath223 , and @xmath225 , which lies near the center of our best - fit range of orbital parameters .
the @xmath84 orbital axis ( blue ) is closely aligned with the homunculus polar axis in 3-d .
orbits clockwise on the sky relative to ( black arrows in inset ) , and apastron is on the observer s side of the system . the semi - major axis ( @xmath112-axis , red ) runs from nw to se on the sky , while the semi - minor axis ( @xmath114-axis , green ) runs from sw to ne .
north is up.,width=309 ] in their recent work modeling s _ rxte _ light curve , o08 and p09 investigated values of @xmath1 in the range @xmath294 , and each obtained a best - fit value of @xmath295 , which they state is consistent with the orbital axis of the binary being aligned with the homunculus polar axis .
however , the inclination angle of the homunculus , as defined in @xcite and @xcite , corresponds to the tilt of the polar axis out of the line of sight ( figure 5 of davidson et al .
since this is the same quantity as the inclination defined for binary orbits , the polar axis of the homunculus is tilted from the plane of the sky by @xmath296 .
therefore , an @xmath291 is required in order for the orbital axis to be aligned with the homunculus polar axis , _ not _ @xmath297 .
the spatially - unresolved nature of the x - ray observations and overall top - bottom symmetry of the problem are what allowed o08 and p09 to reasonably reproduce the _ rxte _ light curve using an @xmath295 .
however , an @xmath291 works as well .
moreover , with either @xmath1 , _ any _ pa on the sky of the orbital axis produces a good fit to the _ rxte _ light curve .
unfortunately , this ambiguity in @xmath1 and pa of the orbital axis has not been sufficiently discussed in the literature , and most have simply assumed that the orbital and homunculus polar axes are aligned . yet , knowing the value of @xmath1 is crucial for proper interpretation of spatially - resolved observational diagnostics and , most importantly , determining the stellar masses .
taking into account this ambiguity in @xmath1 inherent to 3-d models of x - ray data , our derived ranges of @xmath1 and @xmath2 agree very well with those of o08 , p09 , and p11 .
they are also consistent with the @xmath1 and @xmath298 values derived by @xcite in their attempts to understand the origin of the high - velocity ( up to @xmath299 ) absorption wing detected in i @xmath300 during s 2009 periastron passage .
our binary orientation is in complete agreement with simple models proposed to explain the available radio @xcite and ii @xmath301 @xcite observations as well .
further support for our results and derived orbital orientation comes from recent work by @xcite , who mapped the flux of the broad , blue - shifted component of [ iii ] @xmath384659 and found that most of the emission originates in the inner @xmath302 region of ( see their figure 8) .
the synthetic spectro - images and 2-d spatial maps of [ iii ] @xmath384659 in section @xmath286 show that most of the emission should indeed originate in the inner @xmath302 .
according to our 3-d model , this emission forms in / near the current wwc zone in regions photoionized by and manifests itself in spectro - images as a bright , central streak in the inner @xmath188 that spans a wide range of velocities from blue to red .
this emission appears as a bright central streak because _ hst_/stis lacks the resolution necessary to resolve the inner wwc region s complex features .
@xcite additionally found that the spatial distribution of the blue - shifted component of [ iii ] @xmath384659 in the central @xmath184 of is not as sharp as a stellar point source and is detectably elongated to the ne and sw ( see their figures 8 and 9 ) .
model images displaying the 2-d spatial distribution on the sky of the blue - shifted component of [ iii ] @xmath384659 ( first column of figures [ fig9 ] , [ fig12 ] , and [ fig13 ] ) show that the emission _ is _ elongated to the ne and sw for our suggested orbital orientation , even in the central @xmath184 . while detailed mapping with _
hst_/stis is needed to verify the results of @xcite , they do strongly suggest that our 3-d model and binary orientation are correct .
all other orbital orientations , such as that with @xmath303 favored by @xcite ; and @xcite , produce spectro - images that are in strong disagreement with the @xmath42 , @xmath215 observations , lacking any spatially - extended , blue - shifted emission and containing significant amounts of extended , red - shifted emission . with the main source of photons capable of producing high - ionization forbidden line emission
, there is no realistic situation with @xmath303 , for any @xmath1 and @xmath107 , that can produce spectro - images like those observed since there will always be an extended red - shifted component .
hypothesizing some method to absorb or diminish the red - shifted emission does not help since there would then be almost _ no _ spatially - extended emission detected at slit @xmath215 .
there is no known source of ionizing photons in the system that permits a @xmath303 orientation .
photons generated in the wwc region will not work simply because the wwc zone is on the opposite side of the system during most of the orbit and s dense wind would easily absorb any such photons before they reach the periastron side .
even if some contrived process could produce significant amounts of spatially - extended , blue - shifted emission on the periastron side of the system , this would not result in a match to the observations as the spectro - images would then contain both red- _ and _ blue - shifted extended emission .
there would have to be some way to absorb or diminish the extended , optically - thin red - shifted emission while allowing the blue - shifted emission component to reach the observer .
the problem is that extended , red - shifted emission is observed at other slit pas , even for similar orbital phases . therefore , the absorption / diminution would have to be restricted to very specific , extended regions on the sky , which seems unrealistic . given all of these issues , as well as those concerning the observed phase dependence of the emission , an orientation that places apastron on our side of the system seems unavoidable .
while a detailed review of s spectral variability is beyond the scope of this paper , it now seems well established that observed , phase - dependent , narrow ( @xmath75 ) high - ionization forbidden emission lines form in weigelt blobs b , c , and d @xcite .
illumination of the weigelt blobs by is required for the formation of these narrow lines , which are present during most of s 5.5-year orbit , and fade during periastron passage , gradually returning to their ` normal ' strength a few months afterward @xcite .
our proposed orientation and direction for s binary orbit are consistent with , and provide an explanation for , the time variability of the narrow - line emission features seen in the weigelt blobs .
@xcite found that weigelt blobs b , c , and d are located within @xmath302 to @xmath304 nw of the central stellar source near the homunculus equatorial plane ( see figure [ fig2 ] ) .
based on proper motions and observed blue - shifted emission , the weigelt blobs are located on the same side of as the observer and are thought to have been ejected sometime around the smaller eruption of 1890 that also formed the little homunculus @xcite . since the weigelt blobs are on the same side of the system as the observer , and because is required for the formation of the narrow , high - ionization forbidden lines that form in the blobs during most of the orbit , one easily concludes that must be on the same side of the system as the observer during most of the orbit .
moreover , the weigelt blobs are located in the nw quadrant of the system . for our proposed orbital orientation , during most of the orbit ,
the low - density wind cavity created by is open toward and pointing nw , in the direction of the blobs .
the low - density cavity thus provides a path for the ionizing photons from to reach the weigelt blobs and produce the high - ionization emission . over the course of the 5.5-year orbit ,
gradually moves from se to nw on the sky when going from periastron to apastron , and then from the nw back to the se when moving from apastron to periastron .
the orbital motion of and gradual embedding in the wind of close to periastron leads to changes in the spatial extent and direction of the photoionization region , which causes observed variations in the narrow , high - excitation emission in the blobs . during periastron passage , the ionizing flux from shuts off , and
the high - excitation emission from the blobs fades .
it takes several months after periastron for to restore the large photoionization volume that faces nw , and thus the high - excitation emission in the blobs .
while omitting some of the details , this qualitative picture is consistent with the known behavior of the weigelt blobs .
the proposed clockwise direction of s orbit is further supported by the images of @xcite , which show excess uv emission to the sw of just before periastron , and to the ne just after .
a complete review of all of the available observations of and how they support the derived orientation and direction of the binary orbit is obviously not possible here .
however , we note that the orientation shown in figure [ fig14 ] and clockwise motion on the sky of are consistent with all known observations of to date . a variety of models , ranging from the interaction of stellar winds with differing speeds and densities , to binary interactions , and even a binary merger , have been proposed for explaining both the great eruption and the bipolar shape of the resulting nebula @xcite .
observational studies of the homunculus have firmly established that its polar axis is orientated on the sky at a @xmath306 , and that its inclination is @xmath307 @xcite . as shown in figure [ fig14 ] , the orbital axis of the binary for our best - fit range of orientation parameters is closely aligned in 3-d with the homunculus polar axis .
such an alignment has important implications for theories for the formation of the homunculus and/or the present - day shape of s wind .
_ hst_/stis long - slit spectral observations of the homunculus by smith et al .
( 2003a ) indicate that the reflected stellar spectrum over the poles has stronger and broader absorption in h@xmath308 , implying a denser , polar outflow .
very large telescope ( vlt)/very large telescope interferometer ( vlti ) observations of the optically thick stellar wind of by @xcite and @xcite also indicate that s wind could be prolate in shape and aligned at the same pa on the sky as the homunculus .
many have interpreted these observational studies as evidence for the current - day wind of being prolate , with a bipolar form and orientation similar to the homunculus .
thus , one explanation offered for both the homunculus formation and the nature of the present - day wind is that is a rapid rotator ( owocki & gayley 1997 ; maeder & meynet 2000 ; dwarkadas & owocki 2002 ; smith et al .
2003a ; owocki , gayley , & shaviv 2004 ) ; the effects of gravity darkening on radiation - driven wind outflows from a rapidly rotating star suggested as a way of explaining both the high polar densities and velocities inferred in s extended wind .
however , using 2-d radiative transfer models , @xcite and @xcite find that the density structure of s wind can be sufficiently disturbed by , thus strongly affecting the observed uv spectrum , optical hydrogen lines , and mimicking the effects of fast rotation in the interferometric observables .
@xcite further show that even if is a fast rotator , models of the interferometric data are not unique , and both prolate- and oblate - wind models can reproduce the interferometric observations .
these prolate- and oblate - wind models additionally suggest that the rotation axis of would not be aligned with the homunculus polar axis . further complicating the situation are the recent results of @xcite , who find no evidence for higher wind velocities at high stellar latitudes in h@xmath309 p cygni profiles obtained during s normal state ( see their section 5 ) .
together , these new results challenge the idea that the present - day wind of is latitudinally - dependent and prolate in shape . in this context , it is noteworthy that the results in section 4 match the observations so well , given that the winds from both stars are assumed to be spherical in the 3-d simulations .
new 3-d simulations with a latitudinal - dependent wind for , and comparison of the resulting spectro - images to the observations , may provide additional clues as to the nature of s present - day wind .
it is important to keep in mind though that these new results do not necessarily mean that s wind was not latitudinally - dependent in the past .
latitudinal - dependent mass - loss from a rapidly - rotating could have played a role in the formation of the homunculus .
however , the argument for this possibility should not be made based on the above - mentioned observations of s current - day wind .
if s rotation axis is aligned with the orbital axis , and the star is a rapid rotator , then the extended wind should be prolate and aligned at nearly the same pa as the homunculus .
however , the orbital axis and rotation axis of do not have to be aligned .
this is another common assumption , with important implications for theories proposed to explain s spectroscopic events @xcite .
future observations using diagnostics that are not significantly affected by are needed to determine whether or not is a rapid rotator , and whether its rotation axis is aligned with the orbital axis .
alignment of the orbital and homunculus polar axes strongly suggests that binarity played a role in the great eruption , and possibly the smaller eruption that later formed the little homunculus .
a binary merger seems unlikely given the multiple known large eruptions of the system @xcite .
the most likely situation involves some sort of interaction between and at periastron .
@xcite suggested a scenario in which a rotating loses mass due to increased angular momentum caused by an interaction with .
models for the periastron - passage triggering of s massive eruptions , and the requirement of a binary to explain the bipolar shape of the homunculus , have also been proposed by soker and colleagues @xcite .
more recently , @xcite presented a revised historical light curve for , showing that two 100-day peaks observed in 1838 and 1843 just before the great eruption coincide to within weeks of periastron , provided the orbital period then was shorter than the current period by @xmath310 .
the beginning of s lesser outburst in 1890 also seems to have occurred around periastron @xcite .
based on these findings and other considerations , @xcite proposes that a stellar collision occurred at periastron before and during s great eruption , with plunging deeply into the bloated photosphere of .
is close alignment of the orbital axis and polar axis of the homunculus evidence for such a scenario ?
only detailed theoretical modeling can answer this question , but given the now apparent alignment of the two axes , binary interaction scenarios should be seriously considered as possible explanations for s multiple , massive eruptions .
the system not only has a centuries - long history of variation , including two major eruptions , it exhibits shorter - term variability with a 5.5-year period @xcite .
this implies that s numerous spectroscopic events are also related to periastron passage @xcite .
contrary to many expectations , s 2009.0 spectroscopic event was considerably different than that of earlier cycles ( c11 ; * ? ? ?
the observed x - ray minimum in 2009 was substantially shorter than the minima in 1998 and 2003 , by approximately one month ( c10 ; c11 ) .
the emission strength of h@xmath308 and various x - ray lines also appears to have decreased by factors of order two ( c11 ) .
recent _ hst _
wfpc2 observations by @xcite show that the minimum in the uv was much deeper ( again by about a factor of two ) for the 2009 event than for the 2003 event . while still the subject of debate , the behavior of [ ii ] @xmath384686 might have been different in 2009 as well , with a second ` outburst ' occurring @xmath311 after the first @xcite .
the cause of this sudden change in the behavior of s spectroscopic event is still poorly understood .
one proposed explanation is that the mass - loss rate of has suddenly decreased by at least a factor of two ( c10 ; c11 ; * ? ? ?
* ) . however , this is far from confirmed and the exact reasons for such a drop are unknown .
the results in this paper suggest a possible way to observationally test the idea that s mass - loss rate has recently changed .
all of the _
hst_/stis observations modeled were taken in or before 2004 .
the model in this paper assumes a mass - loss rate of @xmath312 for , which is based on earlier observational and modeling studies ( davidson & humphreys 1997 ; h01 ; h06 ) . the good match between the synthetic spectro - images and the observations suggests that the mass - loss rate of was @xmath313 at the time the observations were taken .
the results in section 4 though are strongly dependent on the mass - loss rate assumed for and the ionizing flux of photons assumed for .
if the ionizing flux of photons from remains constant , but the mass - loss rate of drops by a factor of two or more , the size of the photoionization region should increase considerably .
therefore , the spatial extent and flux of the observed high - ionization forbidden lines should also drastically change .
the phase dependence of the forbidden emission would likely differ , with the high - ionization emission vanishing at later phases ( compared to earlier orbital cycles ) when going into periastron , and reappearing at earlier phases afterward . observed variations with slit pa should be different too , possibly showing longer spatially - extended structures , and even new components at different @xmath171 .
changes in the wind terminal velocity of could be similarly investigated using the maximum observed @xmath171 of the spatially - extended emission .
multi - epoch observations coupled with improved 3-d radiative transfer modeling of the high - ionization forbidden line emission would also help in determining if there is a significant change in s ionizing flux of photons , mass - loss rate , or wind terminal speed .
constraints on the ionizing flux of photons from could be compared to stellar models for a range of o @xcite and wr @xcite stars , allowing one to obtain a luminosity and temperature for .
a major goal of this paper has been to use the 3-d dynamical model and available _
hst_/stis observations of high - ionization forbidden lines to constrain the absolute 3-d orientation and direction of s binary orbit .
the spatially - resolved spectroscopic observations obtained with the /stis provide the crucial spatial _ and _ velocity information needed to help accomplish this . since the forbidden lines do not have an absorption component
, they provide a clear view throughout s extended interacting winds .
each high - ionization forbidden line arises _ only _ in regions where ( 1 ) photoionization by can produce the required ion , ( 2 ) the density is near the line s critical density , and ( 3 ) the material is at the appropriate temperature .
our results are therefore not as dependent on the many complicated details and effects encountered when modeling other forms of emission and/or absorption , such as x - rays or spectral features of i , i , or other species ( h01 ; h06 ; nielsen et al . 2007b ; p11 ) . synthetic spectro - images of [ iii ] emission line structures generated using 3-d sph simulations of s binary wwc and radiative transfer codes were compared to the available /stis
observations for a variety of orbital phases and stis slit pas .
the model spectro - images provide important details about the physical mechanisms responsible for the observed high - ionization forbidden line emission , as well as the location and orientation of the observed emitting structures .
below , we summarize our key conclusions and outline the direction of future work .
1 . large - scale ( @xmath314 ) 3-d sph simulations of s binary colliding winds show that during periastron passage , the dense wind of flows unimpeded in the direction of system apastron ( figure [ fig5 ] ) .
following periastron , s high - velocity wind collides with this primary wind material , creating dense shells that expand outward during s 5.5-year orbital cycle .
these shells eventually fragment , forming a pair of dense , spatially - extended ` arcs ' of primary wind on the apastron side of the system .
after several orbital cycles , the arcs drop in density and mix with the surrounding low - density wind material from . 2 . during
most of s orbit , far - uv radiation from photoionizes a significant , spatially - extended region on the apastron side of the system , including portions of the dense arcs of primary wind formed following periastron ( figure [ fig5 ] ) .
portions of the current wwc region and primary wind just beyond it are also photoionized by .
the density and temperature of the material in these regions are ideal for producing [ iii ] emission ( figure [ fig5 ] ) .
3 . during periastron passage ,
becomes enveloped in the dense wind of .
this significantly reduces the size of the photoionization region , to approximately a few semi - major axes in diameter . since the large photoionization region has ` collapsed ' , the spatially - extended ( @xmath46 ) high - ionization forbidden line emission vanishes during periastron passage ( @xmath315 ) . it does not reappear until completes periastron passage and restores the extended photoionization region , which can take several months
synthetic spectro - images of [ iii ] @xmath384659 show that most of the [ iii ] emission originates in the central @xmath302 , in agreement with figure 8 of @xcite .
our 3-d model shows that this emission forms in photoionized material near the current wwc zone , which is not spatially - resolved by /stis .
spatially - extended ( @xmath316 ) [ iii ] emission arises in photoionized portions of the expanding arcs of primary wind formed just after periastron passage .
only spectro - images generated for lines - of - sight angled @xmath317 to @xmath7 prograde of the semi - major axis on the apastron side of the system are able to match the observations taken at @xmath42 , slit @xmath215 .
however , there is an ambiguity in the orbital inclination @xmath1 , with models for @xmath318 to @xmath256 , @xmath217 to @xmath218 and @xmath319 to @xmath5 , @xmath220 to @xmath221 both able to produce entirely blue - shifted arcs ( figure [ fig9 ] ) .
model spectro - images of multi - phase observations obtained at slit @xmath57 break the above degeneracy in @xmath1 , fully constraining s orbital orientation parameters . given the uncertainties in some of the stellar / wind parameters used in the 3-d modeling ,
_ our suggested best - fit range of orientation parameters for the binary system are _
@xmath4 to @xmath5 , @xmath6 to @xmath7 , @xmath258 to @xmath9 .
therefore , _ the orbital axis of the binary system is closely aligned in 3-d with the homunculus polar axis , with apastron on the observer s side of the system and orbiting clockwise on the sky relative to _ ( figure [ fig14 ] ) .
all other orbital orientations , including that in which is in front of at periastron ( @xmath240 ) , are explicitly ruled out as they produce model spectro - images that lack entirely blue - shifted , spatially - extended arcs , containing instead significant amounts of unobserved , spatially - extended red - shifted emission at @xmath42 , slit @xmath215 ( figure [ fig6 ] ) .
the 3-d model predicts that the orientation on the sky of the blue - shifted component of [ iii ] @xmath384659 should not vary much during s orbit , stretching from ne to sw .
however , the spatial extent of the emission should grow larger as the system moves from periastron to apastron , and decrease in size close to periastron
. the orientation and spatial extent of the red - shifted component should behave in a way similar to that of the blue - shifted component , but pointing nw on the sky .
the apparent alignment of the orbital axis and homunculus polar axis implies a link between binarity and s numerous massive eruptions .
binary interaction scenarios should be seriously considered as possible explanations for the great eruption and formation of the bipolar homunculus , and possibly the smaller eruption in 1890 that formed the little homunculus . 10 .
future detailed 3-d simulations and radiative transfer calculations , together with high - resolution spatial mapping of the high - ionization forbidden line emission with _
hst_/stis , have the potential to further constrain the temperature and luminosity of .
the strong dependence of the forbidden line emission on the mass - loss rate of and ionizing flux of photons from may provide clues as to the nature of s long - term variability , specifically , the cycle - to - cycle variations of the spectroscopic events . the 3-d dynamical modeling in this paper and
the available _
spectral observations have increased our understanding of the system .
much work remains , however .
efforts are underway to improve both the 3-d hydrodynamical modeling and 3-d radiative transfer approach .
new 3-d simulations that include radiative cooling , radiation - driven stellar winds , gravity , and effects like radiative inhibition and braking are on the horizon .
preliminary results of detailed 3-d radiative transfer calculations using ` simplex ` @xcite are promising as well . with these new tools at our disposal , it should be possible to further constrain the stellar , wind , and orbital parameters of , setting the stage for orbital modeling to determine the stellar masses .
t. i. m. and c. m. p. r. were supported through the nasa graduate student researchers program .
t. i. m. and j. h. g. also thank the max planck society for financial support for this work .
all sph simulations were performed on nasa advanced supercomputing ( nas ) facilities using nasa high end computing ( hec ) time .
o. acknowledges partial support from nasa atp grant nnk11ac40 g .
t. r. g. acknowledges the hospitality of the mpifr during the completion of this paper .
99 abdo , a. a. , et al .
2010 , apj , 723 , 649 abraham , z. , falceta - gonalves , d. , dominici , t. p. , nyman , l .- . , durouchoux , p. , mcauliffe , f. , caproni , a. , and jatenco - pereira , v. 2005a , a&a , 437 , 977 abraham z. , falceta - gonalves d. , dominici t. , caproni a. , and jatenco - pereira v. 2005b , mnras , 364 , 922 abraham z. and falceta - gonalves d. 2007 , mnras , 378 , 309 abraham , z. and damineli , a. 1999 , eta carinae at the millennium , 179 , 263 bate , m. r. and bonnell , i. a. 1997 , mnras , 285 , 33 bate , m. r. , bonnell , i. a. , and price , n. m. 1995 , mnras , 277 , 362 benz , w. , cameron , a. g. w. , press , w. h. , and bowers , r. l. 1990 , apj , 348 , 647 bryans , p. , badnell , n. r. , gorczyca , t. w. , laming , j. m. , mitthumsiri , w. , and savin , d. w. 2006 , apjs , 167 , 343 corcoran , m. f. 2011 , bulletin de la societe royale des sciences de liege , 80 , 578 ( c11 ) corcoran , m. f. , hamaguchi , k. , pittard , j. m. , russell , c. m. p. , owocki , s. p. , parkin , e .
r. , and okazaki , a. 2010 , apj , 725 , 1528 ( c10 ) corcoran , m.f . 2005 ,
apj , 129 , 2018 corcoran , m. f. , et al .
2004 , apj , 613 , 381 corcoran , m. f. , fredericks , a. c. , petre , r. , swank , j. h. , and drake , s. a. 2000 , apj , 545 , 420 cox , p. , mezger , p. g. , sievers , a. , najarro , f. , bronfman , l. , kreysa , e. , and haslam , g. 1995 , a&a , 297 , 168 crowther , p. a. 2007 , ara&a , 45 , 177 damineli , a. et al . 2008a , mnras , 384 , 1649 damineli , a. et al .
2008b , mnras , 386 , 2330 damineli , a. , kaufer , a. , wolf , b. , stahl , o. , lopes , d. f. , and de arajo , f. x. 2000 , apjl , 528 , l101 damineli , a. 1996 , apjl , 460 , l49 davidson , k. 2005 , the fate of the most massive stars , 332 , 101 davidson , k. , smith , n. , gull , t. r. , ishibashi , k. , and hillier , d. j. 2001 , aj , 121 , 1569 davidson , k. 1999 , eta carinae at the millennium , 179 , 304 davidson , k. , ishibashi , k. , gull , t. r. , and humphreys , r. m. 1999 , eta carinae at the millennium , 179 , 227 davidson , k. 1997 , newa , 2 , 387 davidson , k. et al .
1997 , aj , 113 , 335 davidson , k. and humphreys , r.m .
1997 , ara&a , 35 , 1 davidson k. , ebbets d. , weigelt g. , humphreys r. m. , hajian a. r. , walborn n. r. , and rosa m. 1995 , apj , 109 , 1784 dopita , m. a. and sutherland , r. s. 2003 , diffuse matter in the universe , springer : heidelberg duncan , r. a. and white , s. m. 2003 , mnras , 338 , 425 dwarkadas , v. and owocki , s.p .
2002 , apj , 581 , 1337 falceta - gonalves d. and abraham , z. 2009 , mnras , 399 , 1441 falceta - gonalves d. , jatenco - pereira v. , and abraham z. 2005 , mnras , 357 , 895 feast m. , whitelock p. , and marang f. 2001 , mnras , 322 , 741 fernndez - lajs , e. et al . 2010 , new astronomy , 15 , 108 groh , j. 2011 , bulletin de la societe royale des sciences de liege , 80 , 590 groh , j. h. , et al .
2010b , a&a , 517 , a9 groh , j. h. , madura , t. i. , owocki , s. p. , hillier , d. j. , & weigelt , g. 2010a , apjl , 716 , l223 gull , t.r .
, et al . 2009 ,
mnras , 396 , 1308 ( g09 ) hamaguchi , k. et al .
2007 , apj , 663 , 522 hartman , h. 2003 , ph.d .
dissertation , lund observatory , lund university , sweden henley , d. b. , corcoran , m. f. , pittard , j. m. , stevens , i. r. , hamaguchi , k. , and gull , t. r. 2008 , apj , 680 , 705 henley , d.b .
2005 , ph.d .
dissertation , university of birmingham hillier , d. j. et al .
2006 , apj , 642 , 1098 ( h06 ) hillier , d. j. , davidson , k. , ishibashi , k. , and gull , t. 2001 , apj , 553 , 837 ( h01 ) hillier , d. j. and miller , d. l. 1998 , apj , 496 , 407 hillier , d.j .
and allen , d.a .
1992 , a&a , 262 , 153 hillier d. j. 1988 , apj , 327 , 822 humphreys , r. m. , davidson , k. , and koppelman , m. 2008 , aj , 135 , 1249 iben , i. , jr . 1999 ,
eta carinae at the millennium , 179 , 367 ignace , r. , bessey , r. , and price , c. s. 2009 , mnras , 395 , 962 ignace , r. and brimeyer , a. 2006 , mnras , 371 , 343 ishibashi , k. , et al .
2003 , aj , 125 , 3222 ishibashi , k. , corcoran , m. f. , davidson , k. , swank , j. h. , petre , r. , drake , s. a. , damineli , a. , and white , s. 1999 , apj , 524 , 983 kashi , a. and soker , n. 2010 , apj , 723 , 602 kashi , a. and soker , n. 2009 , mnras , 397 , 1426 kashi , a. and soker , n. 2008 , mnras , 390 , 1751 kashi , a. and soker , n. 2007 , newa , 12 , 590 krist , j. and hook , r. 1999 , the tiny tim user s guide ( baltimore : stsci ) madura , t. i. 2010 , ph.d .
dissertation , university of delaware maeder , a. and meynet , g. 2000 , ara&a , 38 , 143 martins , f. , schaerer , d. , and hillier , d. j. 2005 , a&a , 436 , 1049 mehner , a. , davidson , k. , martin , j. c. , humphreys , r. m. , ishibashi , k. , and ferland , g. j. 2011 , arxiv:1106.5869 mehner , a. , davidson , k. , ferland , g. j. , and humphreys , r. m. 2010 , apj , 710 , 729 mihalas , d. 1978 , stellar atmospheres ( new york : w. h. freeman ) monaghan , j. j. 2005 , reports on progress in physics , 68 , 1703 nahar , s. n. 1997 , phrva , 55 , 1980 nahar , s. n. and pradhan , a. k. 1996 , a&as , 119 , 509 nielsen , k. e. , ivarsson , s. , and gull , t. r. 2007a , apjs , 168 , 289 nielsen , k. e. , corcoran , m. f. , gull , t. r. , hillier , d. j. , hamaguchi , k. , ivarsson , s. , and lindler , d. j. 2007b , apj , 660 , 669 nussbaumer , h. and vogel , m. 1987 , a&a , 182 , 51 okazaki , a. t. , owocki , s. p. , russell , c. m. p. , and corcoran , m. f. 2008 , mnras , 388 , l39 ( o08 ) osterbrock , d. e. 1989 , astrophysics of gaseous nebulae and active galactic nuclei , university science books , california owocki , s. p. , gayley , k. g. , and shaviv , n. j. 2004 , apj , 616 , 525 owocki , s. p. and gayley , k. g. 1995 , apj , 454 , l145 paardekooper , j - p .
2010 , ph.d .
dissertation , leiden observatory , leiden university , netherlands parkin , e. r. , pittard , j. m. , corcoran , m. f. , and hamaguchi , k. 2011 , apj , 726 , 105 ( p11 ) parkin , e. r. , pittard , j. m. , corcoran , m. f. , hamaguchi , k. , and stevens , i. r. 2009 , mnras , 394 , 1758 ( p09 ) pittard , j.m . and corcoran , m.f . 2003 , rmxac , 15 , 81 pittard , j.m . and corcoran , m.f . 2002 , a&a , 383 , 636 ( pc02 ) pittard , j.m .
2000 , ph.d .
dissertation , university of birmingham pittard , j. m. 1998 , mnras , 300 , 479 price , d. j. 2007 , publications of the astronomical society of australia , 24 , 159 price , d. j. 2004 , ph.d . dissertation ,
institute of astronomy & churchill college , university of cambridge richardson , n. d. , gies , d. r. , henry , t. j. , fernndez - lajs , e. , and okazaki , a. t. 2010 , aj , 139 , 1534 smith , n. and frew , d. j. 2011 , mnras , 415 , 2009 smith , n. 2011 , mnras , 415 , 2020 smith , n. 2009 , arxiv:0906.2204 smith , n. 2006b , apj , 644 , 1151 smith , n. and owocki , s. 2006 , apj , 645 , l45 smith , n. et al .
2004 , apj , 605 , 405 smith , n. , davidson , k. , gull , t. r. , ishibashi , k. , and hillier , d. j. 2003a , apj , 586 , 432 smith , n. , gehrz , r. d. , hinz , p. m. , hoffmann , w. f. , hora , j. l. , mamajek , e. e. , and meyer , m. r. 2003b , aj , 125 , 1458 soker , n. 2007 , apj , 661 , 490 soker , n. 2004 , apj , 612 , 1060 tavani , m. , et al .
2009 , apjl , 698 , l142 teodoro , m. , et al .
2011 , arxiv:1104.2276 van boekel , r. , et al . 2003 , a&a , 410 , l37 van genderen , a. m. , sterken , c. , allen , w. h. , and walker , w. s. g. 2006 , journal of astronomical data , 12 , 3 verner , e. , bruhweiler , f. , and gull , t. 2005 , apj , 624 , 973 vishniac , e.t . 1983 , apj , 274 , 152 weigelt , g. , et al .
2007 , a&a , 464 , 87 weigelt , g. and ebersberger , j. 1986 , a&a , 163 , l5 white , s. m. , duncan , r. a. , chapman , j. m. , and koribalski , b. 2005 , the fate of the most massive stars , 332 , 126 whitelock , p. a. , feast , m. w. , marang , f. , and breedt , e. 2004 , mnras , 352 , 447 williams , p. m. 2008 , revista mexicana de astronomia y astrofisica conference series , 33 , 71 wnsch , r. , dale , j.e . ,
palous , j. , and whitworth , a.p .
2010 , mnras , 407 , 1963 zanella , r. , wolf , b. , and stahl , o. 1984 , a&a , 137 , 79 zethson , t. , johansson , s. , hartman , h. , and gull , t.r .
2011 , a&a , submitted zethson , t. 2001 , ph.d .
dissertation , lunds universitet , sweden zhang , h. 1996 , a&as , 119 , 523 | we present a three - dimensional ( 3-d ) dynamical model for the broad [ iii ] emission observed in @xmath0 carinae using the _ hubble space telescope_/space telescope imaging spectrograph ( /stis ) .
this model is based on full 3-d smoothed particle hydrodynamics ( sph ) simulations of s binary colliding winds .
radiative transfer codes are used to generate synthetic spectro - images of [ iii ] emission line structures at various observed orbital phases and stis slit position angles ( pas ) . through a parameter study that varies the orbital inclination @xmath1 , the pa @xmath2 that the orbital plane projection of the line - of - sight makes with the apastron side of the semi - major axis , and the pa on the sky of the orbital axis , we are able , for the first time , to tightly constrain the absolute 3-d orientation of the binary orbit . to simultaneously reproduce the blue - shifted emission arcs observed at orbital phase 0.976 , stis slit @xmath3 , and the temporal variations in emission seen at negative slit pas , the binary needs to have an @xmath4 to @xmath5 , @xmath6 to @xmath7 , and an orbital axis projected on the sky at a @xmath8 to @xmath9 east of north .
this represents a system with an orbital axis that is closely aligned with the inferred polar axis of the homunculus nebula , in 3-d .
the companion star , , thus orbits clockwise on the sky and is on the observer s side of the system at apastron .
this orientation has important implications for theories for the formation of the homunculus and helps lay the groundwork for orbital modeling to determine the stellar masses .
[ firstpage ] hydrodynamics binaries : close stars : individual : eta carinae stars : mass - loss stars : winds , outflows line : formation |
@xmath7 is one of the most interesting hadrons and attracting much interest from several view points .
@xmath7 is the lightest negative - parity baryon , even though it has one valence strange quark , which is heavier than the up and down quarks . among the @xmath8 baryons ,
@xmath7 is isolated from the others , much lighter than the non - strange counterpart @xmath9 .
it has no spin - orbit partner in the vicinity , as the lowest spin @xmath10 state is @xmath11 .
furthermore , the structure of @xmath7 remains mysterious .
on one hand , @xmath7 is interpreted as a flavor - su(3)-singlet three - quark state in conventional quark models . on the other hand ,
@xmath7 could be interpreted as an antikaon - nucleon @xmath12 molecular bound state ( b.e .
@xmath13 30 mev ) .
the binding energy of @xmath12 implies a strong attractive force between @xmath14 and @xmath15 @xcite , which may cause a new type of dense hadronic matter , kaonic nuclei or kaonic nuclear matter @xcite .
we also expect that such @xmath12 bound states with large binding energies can be regarded as compact 5-quark states .
the 5-quark picture of @xmath16 has advantages that all five quarks can be placed in the lowest - lying @xmath17 state to form a negative - parity baryon , and also that it requires no spin - orbit partner of @xmath7 .
the property of @xmath7 can therefore be an important clue to new paradigm in hadron physics .
we here study properties of @xmath16 using the lattice qcd formulation .
lattice qcd is nowadays employed as a very powerful tool for nonperturbative analysis directly based on qcd , and expected to cast light on the nature of @xmath7 in a model - independent way .
though several lattice qcd studies on @xmath7 have been performed so far @xcite , most of them are based on quenched qcd and few lattice qcd studies succeeded in reproducing the mass of @xmath7 .
moreover , little has been discussed on the lattice about the possible mixing of flavor - su(3)-octet and -singlet components induced by the symmetry breaking .
several possible reasons for the failure of reproducing @xmath7 in lattice qcd were suggested through these studies , such as missing meson - baryon components due to quenching , exotic ( non-3-quark type ) structure of @xmath7 , or insufficiency of the lattice volume in the simulations . resolving such difficulties requires unquenched lattice qcd calculation on a larger lattice volume with varieties of interpolating operators . in this paper
, we aim at clarifying the properties of @xmath7 with two - flavor full lattice qcd , adopting the `` octet '' and `` singlet '' baryon operators to construct correlation matrices , which enables us to extract the low - lying spectrum as well as the mixing between octet and singles components in @xmath7 .
we adopt the renormalization - group improved gauge action and the @xmath2-improved quark action .
simulation parameters are listed in table [ simparam ] .
we adopt three different @xmath3 s , @xmath4 , 1.95 and 2.10 , and corresponding lattice spacings are @xmath5 , 0.1555 and 0.1076 fm @xcite , which are determined so that the empirical @xmath18-meson mass is reproduced . the hopping parameters for strange quark @xmath19
are set to be 0.1431 , 0.1393 , and 0.1373 at @xmath3=1.80 , 1.95 , and 2.10 so that the kaon mass is reproduced .
we note here that , if we adopt @xmath20-meson mass as an input , @xmath19 would be slightly different from those determined by the kaon mass .
the dynamics of @xmath0 resonances would be closely connected to kaon dynamics , and hence we adopt kaon mass as an input in this study .
we employ four different hopping parameters ( @xmath6 ) for each cutoff .
corresponding pion masses range approximately from 500mev to 1.1 gev at each @xmath3
. .[simparam ] list of the present simulation parameters . @xmath3 and @xmath21 are the parameters for the renormalization - group improved gauge action and the @xmath2-improved quark action .
@xmath22 and @xmath23 denote the spatial and temporal lattice extents , and the lattice spacings @xmath24 and the strange - quark hopping parameters @xmath19 are determined so that the empirical @xmath18-meson mass and kaon mass are reproduced @xcite . [
cols="<,<,<,<,<,^",options="header " , ] masses of the ground and the 1st - excited states in the chiral limit for the positive ( upper ) and negative ( lower ) parity @xmath0 states plotted as functions of the lattice spacing @xmath24 .
the horizontal lines indicate the masses of the experimentally observed @xmath0 states .
, title="fig : " ] masses of the ground and the 1st - excited states in the chiral limit for the positive ( upper ) and negative ( lower ) parity @xmath0 states plotted as functions of the lattice spacing @xmath24 .
the horizontal lines indicate the masses of the experimentally observed @xmath0 states .
, title="fig : " ] fig .
[ massvsa ] shows the chirally extrapolated masses of the lowest two states in each parity v.s .
the lattice spacing @xmath24 .
except for the 1st - excited positive - parity state at the largest @xmath24 , all the data show little lattice - cutoff dependence .
the horizontal lines indicate the masses of the experimentally observed @xmath0 states .
the positive - parity ground state agrees well with the mass of @xmath25 , while the lowest two negative - parity states lie around @xmath26 .
since our calculations contain dynamical u and d quarks , scattering states could come into spectra .
@xmath27 and @xmath28 thresholds are drawn in fig .
[ posmass ] and fig.[negmass ] as dashed lines .
the energy of the 1st - excited state at the lightest u- and d - quark masses could be contaminated by such scattering states .
several strategies have been adopted to discriminate resonances from scattering states : study of volume dependence of the masses , and the spectral weights or boundary condition dependence of the eigen - energies .
they were tested and applied to the pentaquark baryon @xcite .
there are two meson - baryon channels relevant in the present calculation , @xmath29 and @xmath30 . in terms of the valence
quark contents , we have 5 different thresholds ( 2 for @xmath30 , 3 for @xmath29 ) . in this paper , in order to distinguish resonance states from all these scattering states , we impose the following boundary condition on the quark fields , @xmath31 under such boundary condition for quark fields , a hadronic state @xmath32 which contains @xmath33 valence quarks obeys @xmath34 and can have spatial momenta , @xmath35 as a result , only states which consist of 3k valence quarks can be zero - momentum states .
since other quark combinations inevitably have non - vanishing spatial momenta , their energies are raised up .
as long as we employ three - quark operators for baryon creation / annihilation , hadronic states appearing in scattering states should contain one or two valence quark(s ) , since sea - quark pairs themselves can not carry flavors .
lowest two eigen - energies in the @xmath36 channel at @xmath37 and @xmath38 under the normal ( periodic ) and the twisted boundary conditions .
the circles ( squares ) are for the ground ( 1st - excited ) states .
the lower and upper solid lines respectively represent the threshold energies of @xmath28 and @xmath27 states evaluated with normal / twisted boundary condition . ]
now we turn back to our full - qcd calculations .
we impose the following boundary condition to all the valence quark fields ( @xmath39 ) ; @xmath40 we plot in fig .
[ compare ] the eigen - energies under the periodic and the twisted boundary conditions , which are obtained at @xmath37 and @xmath38 .
the open and filled circles ( squares ) are for the ground ( 1st - excited ) states .
the lower and upper solid lines respectively represent the threshold energies of @xmath28 and @xmath27 states with normal / twisted boundary condition .
the threshold energies are raised up in the case of twisted boundary condition , whereas the lowest two eigen - energies remain unchanged .
thus we conclude that the observed states are insensitive to boundary conditions , and contaminations from meson - baryon scattering states are small .
namely the results indicate that these states are regarded as compact resonance states .
similar results are obtained also for the positive - parity states .
their energies are insensitive to the boundary conditions .
( the energies of the positive - parity 1st - excited states at the smallest @xmath41 are higher than p - wave thresholds ( with relative momenta @xmath42 ) of @xmath28 or @xmath27 states . )
the chiral unitary approach @xcite has suggested that @xmath7 is not a single pole but a superposition of two independent resonance poles .
the structure of @xmath7 is now attracting much interest , and desired to be clarified in a model independent manner .
we investigate the flavor structure of the ground and the 1st - excited states obtained from the cross correlators of two operators . in order to clarify the flavor structure of the low - lying states ,
we evaluate the overlaps of the obtained states with the source and sink operators .
we define the magnitudes of the singlet ( @xmath43 ) and the octet ( @xmath44 ) components in the @xmath45-th state ( @xmath46 for the ground state and @xmath47 for the 1st excited state ) by @xmath48 actually , we can only obtain the ratio between @xmath49 and @xmath50 since the overall factors in @xmath51 can not be determined in our setups .
we evaluate @xmath52 and @xmath53 defined as @xmath54 both @xmath55 and @xmath56 vanish when the su(3)@xmath57 symmetry is exact , showing that the ground ( 1st excited ) state is purely flavor singlet ( octet ) in the limit . for the positive - parity states ,
we similarly define @xmath58 in this case , we exchange the denominator and the numerator since the ground state is flavor octet and the 1st excited state is flavor singlet in the su(3)@xmath57 limit .
@xmath59 are plotted as a function of @xmath60 .
solid lines denote quadratic functions used in chiral extrapolation .
, title="fig : " ] @xmath59 are plotted as a function of @xmath60 .
solid lines denote quadratic functions used in chiral extrapolation .
, title="fig : " ] @xmath59 are plotted as a function of @xmath60 .
solid lines denote quadratic functions used in chiral extrapolation .
, title="fig : " ] @xmath61 are plotted as a function of @xmath60 .
solid lines denote quadratic functions used in chiral extrapolation .
, title="fig : " ] @xmath61 are plotted as a function of @xmath60 .
solid lines denote quadratic functions used in chiral extrapolation .
, title="fig : " ] @xmath61 are plotted as a function of @xmath60 .
solid lines denote quadratic functions used in chiral extrapolation .
, title="fig : " ] we show @xmath62 and @xmath63 as a function of pion - mass squared for each lattice cutoff in fig .
[ gpos ] and fig .
as is expected from the symmetry , such mixing coefficients cross zero at the flavor symmetric limit ( @xmath64 ) .
the data show smooth quark - mass dependence toward the chiral limit in all the lattice - cut - off cases and the dependences are almost lattice - cut - off independent .
we find that the magnitude of operator mixing gets larger for larger flavor - symmetry breaking .
it is interesting that the 1st - excited state in positive - parity channel is dominated by the flavor - singlet component showing almost no contaminations of octet components .
we perform quadratic fits of these mixing coefficients as @xmath65 and list the extrapolated values in table [ hadronicmass2 ] . even in the chiral limit , @xmath66 in negative - parity channel
remain less than 0.6 , which indicates that the ground ( 1st - excited ) state is dominated by the singlet ( octet ) component in the present quark - mass range . on the other hand ,
the mixings of the singlet and octet components are generally weaker in the positive parity @xmath0 s .
it is natural because the splitting of the two states are large in the su(3)@xmath57 .
nevertheless , it is interesting to see that the ground state , corresponding to @xmath25 , contains significant ( about 10% ) flavor singlet component in the chiral limit , which was not expected from the simple quark model picture . in practice
, we need to interpret such overlap coefficients into physical amplitudes . in the euclidean - time range we consider @xmath67 are satisfied .
we here assume that each state can be expanded in terms of the singlet and octet basis states as @xmath68 we note that @xmath69 s in this expression are no longer energy eigenstates in qcd , and they are restricted to elements that interpolating operators have overlaps with ; @xmath70 .
( actually we can introduce other singlet and octet bases , for example , whose spatial distributions are different .
if we adopt more interpolating fields such as those with spatial derivatives or with exotic spinor structures , we can introduce much more basis states in this argument . )
we here extract @xmath71 s , using the conditions , @xmath72 and @xmath73 , giving @xmath74 with @xmath75 .
@xmath76 can be estimated from the 2-point correlators , since optimized operators satisfy @xmath77 using the normalization relations , @xmath78 we determine @xmath79 , the amplitude of @xmath80-plet component in @xmath45-th state as follows : @xmath81 the amplitudes @xmath82 obtained on the finest lattice are plotted in fig .
[ fcompon ] . in the present quark - mass range ,
ground ( 1st - excited ) state is dominated by flavor - singlet ( octet ) components .
@xmath79 obtained at @xmath83 are plotted as a function of squared pion mass @xmath60 .
@xmath71 are normalized so that @xmath84 . , title="fig : " ] @xmath79 obtained at @xmath83 are plotted as a function of squared pion mass @xmath60 .
@xmath71 are normalized so that @xmath84 . , title="fig : " ] so far , we have argued the internal structures from the view point of flavor contents .
it would be interesting and useful to interpret these flavor - mixing amplitudes in terms of a hadronic basis , since hadronic - basis expansion of unstable particles could be directly compared with experiments and sometimes beneficial for model calculations .
each flavor element can be decomposed into hadronic bases as @xmath85 in this paper we restrict ourselves only to the @xmath27 and @xmath28 bases .
the su(3 ) clebsh - gordan coefficients tell us @xcite @xmath86 and the following expressions follows ; @xmath87 here , @xmath88 and @xmath89 are unknown coupling constants corresponding to two possible su(3 ) constructions of the meson - baryon octet states , @xmath90 and @xmath91 m\right)$ ] , respectively .
we finally reach the probabilities of @xmath27 and @xmath28 states in @xmath45-th state , @xmath92 and @xmath93 , @xmath94 ^ 2 \\ p_{\sigma\pi}^i & = & \left[\sum_{r=1,8}m_{ir}w_{\sigma\pi , r}\right]^2.\end{aligned}\ ] ] defining @xmath95 by @xmath96 we obtain @xmath97 and @xmath98 as functions of the angle @xmath95 , where we substitute @xmath52 and @xmath53 extrapolated to the chiral limit at @xmath37 ( the finest lattice ) .
we simply adopt @xmath99 , since @xmath76 does not deviate from unity , and is almost quark - mass independent ; @xmath1001.002(29 ) , 1.015(21 ) , 1.030(29 ) , and 1.024(29 ) at @xmath1010.1357 , 0.1367 , 0.1374 , and 0.1382 at this @xmath3 .
the relative signs in @xmath71 are determined so that @xmath102 .
we focus on the meson - baryon components of the low - lying states in negative - parity channel , since they are of interest here .
we show in fig .
[ mbcompon ] , @xmath97 and @xmath98 normalized as @xmath103 as a function of @xmath95 . though @xmath104 s are determined from those obtained at @xmath105 , the behaviors do not change significantly even if we adopt extrapolated values on coarser lattices .
now that the only unknown parameter is @xmath95 representing @xmath88 and @xmath89 in the @xmath106 channel .
unfortunately , the couplings , @xmath88 and @xmath89 , should be determined up to including their signs in order to estimate the meson - baryon components in each state in this strategy .
they can in principle be determined from lattice qcd computations and their determination is left for forthcoming studies .
@xmath97 and @xmath98 of the ground state ( upper panel ) and of the 1st excited state ( lower panel ) as functions of @xmath95 .
they are determined from the @xmath61 s obtained at @xmath105 ( the finest lattice ) .
, title="fig : " ] @xmath97 and @xmath98 of the ground state ( upper panel ) and of the 1st excited state ( lower panel ) as functions of @xmath95 .
they are determined from the @xmath61 s obtained at @xmath105 ( the finest lattice ) .
, title="fig : " ]
an unquenched lattice qcd study for the low - lying @xmath0 baryon states in the @xmath107 , @xmath108 and @xmath109 channel has been performed , focusing especially on the masses and structures of the two lowest negative - parity @xmath0 s .
we have constructed @xmath1 cross correlators of the `` octet '' and `` singlet '' baryonic operators , and diagonalized them so that we can properly extract the information of the low - lying two states . in our measurements , we have adopted the 2-flavor gauge configurations generated by cp - pacs collaboration with the renormalization - group improved gauge action and the @xmath2-improved quark action .
we have performed simulations at three different @xmath3 s , @xmath4 , 1.95 and 2.10 , whose corresponding lattice spacings are @xmath5 , 0.1555 and 0.1076 fm , employing four different hopping parameters ( @xmath6 ) for each cutoff , corresponding to the pion masses ranging from 500 mev to 1.1 gev . it is shown that the mass of the ground - state @xmath25 is reproduced well at all three @xmath3 s , while the 1st - excited positive - parity @xmath0 state lies much higher than the experimentally observed @xmath110 .
the same tendency was reported in ref .
@xcite , and also in the case of the roper resonance , @xmath111 , which is the non - strange su(3 ) partner of @xmath112 @xcite .
it should also be noted that no @xmath29 or @xmath30 scattering states are seen in the calculation .
our results for negative - parity @xmath0 states indicate that there are two @xmath113 states nearly degenerate at around @xmath114gev , while no state as low as @xmath7 is observed . we have revealed the flavor structures of these states from the lattice data for the first time .
it is found that the lowest negative - parity @xmath0 state is dominated by flavor - singlet component .
the second state , which is less than 100 mev above the ground state , is predominantly flavor octet .
thus we find that the first two negative - parity @xmath0 s have different flavor structures .
they , however , come significantly heavier than the experimentally observed lowest - mass state , @xmath16 .
one naive possibility is that the obtained states do not correspond to the physical @xmath16 , but describe excited @xmath0 states .
in fact , the 2nd and 3rd negative - parity states lie at 1670 mev and 1800 mev , both of which are three - star states in the particle data group classification .
the present lattice data are consistent with these excited states . in the non - relativistic quark model approach ,
each of these states is classified as a flavor octet p - wave baryon .
we , however , have shown that the lower state is dominated by a flavor - singlet component .
so , the results here predict one flavor - singlet state and one flavor - octet state in the vicinity .
the reason for missing @xmath16 state will be its poor overlap with three - quark operators . in fact , the inclusion of dynamical quarks does not strongly enhance the signals of possible meson - baryon scattering states .
( see the later discussions . )
if @xmath7 is predominantly a meson - baryon molecular state , such overlaps would be naturally small .
the other possibility is , of course , that the lowest and the 2nd - lowest states describe physical @xmath16 and @xmath26 but the masses have been overestimated . considering that our simulation contains two - flavor dynamical quarks ,
the failure of obtaining a light @xmath0 state could be attributed either to ( 1 ) strange - quark quenching , ( 2 ) insufficient lattice volume or ( 3 ) lack of chiral symmetry .
in fact , ( 1 ) seems to make the masses of octet baryons in positive parity channel slightly ( @xmath115 ) overestimated in the present setups @xcite . on the other hand , the deficiencies,(2 ) and ( 3 ) ,
may cause the lowest state not properly reproduced , supposing that the main component of @xmath7 is a meson - baryon molecular state . in order to check
whether this conjecture is correct , simulations with light dynamical quarks ( @xmath116 mev ) and larger volume ( @xmath117 fm ) will be required . upon the above conjecture
, we can further consider one interesting scenario that these states both correspond to the physical @xmath16 .
then , the results may support its double pole structure proposed by the chiral unitary approach @xcite . in our results ,
the lowest two states are almost degenerate at all the @xmath3 s ( lattice spacing ) and @xmath41 s ( quark masses ) .
namely , the obtained two states are the signature of the double - pole resonance , but the mass has not yet been reproduced because of the deficiency stated above .
it is also important to note the missing meson - baryon scattering states . despite that dynamical up and down quarks
are included and the meson - baryon thresholds appear around / below the obtained eigen - energies , we have found no clear signal of the meson - baryon scattering states . this fact is supported by the observation that the effective mass plots under the normal and twisted boundary conditions show no prominent differences .
we show in fig.[bccompare2 ] the effective mass plots of the ground state negative - parity @xmath0 under the normal and twisted b.c.s .
they are obtained on the @xmath118 lattice with the largest hopping parameter .
thus we conclude that no scattering states appear in the present spectrum . also in ref .
@xcite , in which excited - state nucleon spectrum was systematically and extensively investigated with two - flavors of dynamical quarks , no clear signal of scattering states was found and the importance of multi - quark operators was raised .
a similar situation can be found in the computation of the wilson loops , whose expectation values give us the potential between a ( heavy fundamental ) quark and an antiquark . in the presence of dynamical quarks ,
such an interquark potential should saturate and flatten at some interquark distance , where the confining string is broken and a quark - antiquark pair is created .
however , no one has ever observed successfully such a `` string breaking '' effect in the wilson loop computations @xcite .
one possible reason for this phenomenon is that the wilson loop itself has poor overlaps with such broken strings . in the @xmath0 spectrum
, we may similarly conjecture that the scattering states or broad resonances , which are sensitive to boundary conditions , have little overlap with the 3-quark interpolating field operators .
more detailed investigation would be needed for clarification of scattering states purely induced by dynamical quarks .
the flavor structures of the @xmath0 states have been very well clarified using the variational method .
the octet and the singlet components are mixed when the flavor - su(3 ) symmetry is broken . actually , the ground ( 1st - excited ) state is dominated by singlet ( octet ) component , and the contamination by another representation is at most 20% ( 5% when squared ) in our present analysis . from these findings
, we expect that the flavor - su(3 ) symmetry is not largely broken .
a similar conclusion was also derived for the study of the meson - baryon coupling constants in lattice qcd @xcite . because the su(3 ) breaking effect seems small , the analyses without su(3 ) mixings adopted so far @xcite make sense to some extent .
the mixings , however , get larger towards the chiral limit , variational analyses could be essentially needed when we adopt much lighter quarks .
we also find the meson - baryon contents in each state strongly depend on the meson - baryon couplings and their signs .
precise determination of the meson - baryon contents will require reliable determination of the couplings up to signs , for which further lattice qcd calculations may be helpful .
all the numerical calculations were performed on nec sx-8r at cmc , osaka university and bluegene / l at kek .
the unquenched gauge configurations employed in our analysis were all generated by cp - pacs collaboration @xcite .
this work was supported in part by the yukawa international program for quark - hadron sciences ( yipqs ) , by the japanese society for the promotion of science under contract number p-06327 and by kakenhi ( 17070002 , 19540275 , 20028006 and 21740181 ) .
j. j. sakurai , annals phys .
* 11 * , 1 ( 1960 ) .
r. h. dalitz , t. c. wong and g. rajasekaran , phys .
153 * , 1617 ( 1967 ) .
y. akaishi and t. yamazaki , phys .
c * 65 * , 044005 ( 2002 ) .
t. yamazaki and y. akaishi , phys .
b * 535 * ( 2002 ) 70 .
y. akaishi , a. dote and t. yamazaki , phys .
b * 613 * , 140 ( 2005 ) [ arxiv : nucl - th/0501040 ] .
w. melnitchouk _ et al .
_ , phys .
d * 67 * , 114506 ( 2003 ) [ arxiv : hep - lat/0202022 ] .
y. nemoto , n. nakajima , h. matsufuru and h. suganuma , phys .
d * 68 * , 094505 ( 2003 ) [ arxiv : hep - lat/0302013 ] .
t. burch , c. gattringer , l. y. glozman , c. hagen , d. hierl , c. b. lang and a. schafer , phys . rev .
d * 74 * , 014504 ( 2006 ) [ arxiv : hep - lat/0604019 ] .
n. ishii , t. doi , m. oka and h. suganuma , prog .
suppl . * 168 * , 598 ( 2007 ) [ arxiv:0707.0079 [ hep - lat ] ] .
a. ali khan _ et al . _
[ cp - pacs collaboration ] , phys .
d * 65 * , 054505 ( 2002 ) [ erratum - ibid .
d * 67 * , 059901 ( 2003 ) ] m. luscher and u. wolff , nucl .
b * 339 * , 222 ( 1990 ) .
s. perantonis and c. michael , nucl .
b * 347 * , 854 ( 1990 ) .
b. blossier , m. della morte , g. von hippel , t. mendes and r. sommer , jhep * 0904 * , 094 ( 2009 ) [ arxiv:0902.1265 [ hep - lat ] ] . s. sasaki , t. blum and s. ohta , phys . rev .
d * 65 * , 074503 ( 2002 ) [ arxiv : hep - lat/0102010 ] .
s. sasaki and t. yamazaki , phys .
d * 74 * , 114507 ( 2006 ) [ arxiv : hep - lat/0610081 ] .
n. mathur _ et al .
_ , phys .
b * 605 * , 137 ( 2005 ) [ arxiv : hep - ph/0306199 ] .
f. csikor , z. fodor , s. d. katz and t. g. kovacs , jhep * 0311 * , 070 ( 2003 ) [ arxiv : hep - lat/0309090 ] .
n. mathur _ et al .
_ , phys .
d * 70 * , 074508 ( 2004 ) [ arxiv : hep - ph/0406196 ] .
n. ishii , t. doi , h. iida , m. oka , f. okiharu and h. suganuma , phys .
d * 71 * , 034001 ( 2005 ) [ arxiv : hep - lat/0408030 ] .
t. t. takahashi , t. umeda , t. onogi and t. kunihiro , phys .
d * 71 * , 114509 ( 2005 ) [ arxiv : hep - lat/0503019 ] .
d. jido , j. a. oller , e. oset , a. ramos and u. g. meissner , nucl .
a * 725 * ( 2003 ) 181 [ arxiv : nucl - th/0303062 ] . c. amsler
_ et al . _
[ particle data group ] , phys .
b * 667 * , 1 ( 2008 ) .
j. m. bulava _ et al .
_ , phys .
d * 79 * , 034505 ( 2009 ) [ arxiv:0901.0027 [ hep - lat ] ] .
u. m. heller , k. m. bitar , r. g. edwards and a. d. kennedy , phys .
b * 335 * , 71 ( 1994 ) [ arxiv : hep - lat/9401025 ] .
b. bolder _ et al . _ , phys .
d * 63 * ( 2001 ) 074504 [ arxiv : hep - lat/0005018 ] .
g. erkol , m. oka and t. t. takahashi , phys .
d * 79 * , 074509 ( 2009 ) [ arxiv:0805.3068 [ hep - lat ] ] . | low - lying @xmath0 baryons with spin 1/2 are analyzed in full ( unquenched ) lattice qcd .
we construct @xmath1 cross correlators from flavor su(3 ) `` octet '' and `` singlet '' baryon operators , and diagonalize them so as to extract information of two low - lying states for each parity .
the two - flavor cp - pacs gauge configurations are used , which are generated in the renormalization - group improved gauge action and the @xmath2-improved quark action .
three different @xmath3 s , @xmath4 , 1.95 and 2.10 , are employed , whose corresponding lattice spacings are @xmath5 , 0.1555 and 0.1076 fm . for each cutoff , we use four hopping parameters , ( @xmath6 ) , which correspond to the pion masses ranging about from 500 mev to 1.1 gev .
results indicate that there are two negative - parity @xmath0 states nearly degenerate at around 1.6 gev , while no state as low as @xmath7 is observed . by decomposing the flavor components of each state , we find that the lowest ( 1st - excited ) negative - parity state is dominated by flavor - singlet ( flavor - octet ) component .
we also discuss meson - baryon components of each state , which has drawn considerable attention in the context of multi - quark pictures of @xmath7 . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.